Citation
The role of subcellular compartments and structures in eukaryotic gene expression

Material Information

Title:
The role of subcellular compartments and structures in eukaryotic gene expression
Creator:
Zambetti, Gerard Paul, 1958-
Publication Date:
Language:
English
Physical Description:
xii, 219 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
Cells ( jstor )
Cytochalasins ( jstor )
Cytoskeleton ( jstor )
DNA ( jstor )
Hela cells ( jstor )
Histones ( jstor )
Messenger RNA ( jstor )
Polyribosomes ( jstor )
RNA ( jstor )
Signals ( jstor )
Gene Expression Regulation ( mesh )
RNA, Messenger ( mesh )
Transcription, Genetic ( mesh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1989.
Bibliography:
Includes bibliographical references (leaves 206-218).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Gerard Paul Zambetti.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
22253045 ( OCLC )
ocm22253045
0025463579 ( ALEPH )

Downloads

This item has the following downloads:

E7JEYIHGD_AM9AO5.xml

AA00012891_00001.pdf

roleofsubcellula00zamb_0040.txt

AA00012891_00001_0184.txt

AA00012891_00001_0007.txt

AA00012891_00001_0223.txt

roleofsubcellula00zamb_0141.txt

roleofsubcellula00zamb_0123.txt

roleofsubcellula00zamb_0190.txt

AA00012891_00001_0087.txt

roleofsubcellula00zamb_0136.txt

roleofsubcellula00zamb_0183.txt

AA00012891_00001_0003.txt

AA00012891_00001_0144.txt

roleofsubcellula00zamb_0088.txt

AA00012891_00001_0231.txt

AA00012891_00001_0039.txt

roleofsubcellula00zamb_0225.txt

roleofsubcellula00zamb_0053.txt

AA00012891_00001_0048.txt

roleofsubcellula00zamb_0193.txt

AA00012891_00001_0057.txt

AA00012891_00001_pdf.txt

roleofsubcellula00zamb_0034.txt

roleofsubcellula00zamb_0210.txt

AA00012891_00001_0151.txt

AA00012891_00001_0177.txt

AA00012891_00001_0049.txt

AA00012891_00001_0086.txt

roleofsubcellula00zamb_0026.txt

roleofsubcellula00zamb_0188.txt

roleofsubcellula00zamb_0137.txt

AA00012891_00001_0053.txt

AA00012891_00001_0189.txt

roleofsubcellula00zamb_0027.txt

roleofsubcellula00zamb_0184.txt

AA00012891_00001_0169.txt

roleofsubcellula00zamb_0021.txt

roleofsubcellula00zamb_0214.txt

AA00012891_00001_0158.txt

AA00012891_00001_0035.txt

roleofsubcellula00zamb_0010.txt

roleofsubcellula00zamb_0134.txt

roleofsubcellula00zamb_0020.txt

AA00012891_00001_0016.txt

roleofsubcellula00zamb_0109.txt

roleofsubcellula00zamb_0211.txt

AA00012891_00001_0043.txt

AA00012891_00001_0011.txt

roleofsubcellula00zamb_0191.txt

AA00012891_00001_0228.txt

AA00012891_00001_0164.txt

roleofsubcellula00zamb_0003.txt

roleofsubcellula00zamb_0139.txt

roleofsubcellula00zamb_0181.txt

AA00012891_00001_0127.txt

AA00012891_00001_0160.txt

AA00012891_00001_0066.txt

roleofsubcellula00zamb_0009.txt

roleofsubcellula00zamb_0231.txt

roleofsubcellula00zamb_0033.txt

roleofsubcellula00zamb_0087.txt

AA00012891_00001_0197.txt

roleofsubcellula00zamb_0164.txt

roleofsubcellula00zamb_0189.txt

AA00012891_00001_0211.txt

roleofsubcellula00zamb_0201.txt

AA00012891_00001_0115.txt

AA00012891_00001_0179.txt

roleofsubcellula00zamb_0064.txt

roleofsubcellula00zamb_0217.txt

roleofsubcellula00zamb_0204.txt

roleofsubcellula00zamb_0074.txt

roleofsubcellula00zamb_0132.txt

roleofsubcellula00zamb_0058.txt

roleofsubcellula00zamb_0199.txt

AA00012891_00001_0056.txt

roleofsubcellula00zamb_0180.txt

AA00012891_00001_0214.txt

AA00012891_00001_0182.txt

AA00012891_00001_0096.txt

AA00012891_00001_0166.txt

roleofsubcellula00zamb_0212.txt

roleofsubcellula00zamb_0075.txt

roleofsubcellula00zamb_0155.txt

roleofsubcellula00zamb_0219.txt

AA00012891_00001_0021.txt

roleofsubcellula00zamb_0146.txt

AA00012891_00001_0192.txt

AA00012891_00001_0105.txt

AA00012891_00001_0149.txt

AA00012891_00001_0220.txt

AA00012891_00001_0216.txt

roleofsubcellula00zamb_0102.txt

AA00012891_00001_0025.txt

roleofsubcellula00zamb_0200.txt

roleofsubcellula00zamb_0233.txt

roleofsubcellula00zamb_0113.txt

AA00012891_00001_0020.txt

roleofsubcellula00zamb_0078.txt

roleofsubcellula00zamb_0129.txt

roleofsubcellula00zamb_0106.txt

AA00012891_00001_0080.txt

AA00012891_00001_0065.txt

roleofsubcellula00zamb_0207.txt

AA00012891_00001_0042.txt

AA00012891_00001_0186.txt

AA00012891_00001_0119.txt

roleofsubcellula00zamb_0178.txt

roleofsubcellula00zamb_0119.txt

AA00012891_00001_0045.txt

AA00012891_00001_0183.txt

AA00012891_00001_0234.txt

AA00012891_00001_0092.txt

roleofsubcellula00zamb_0206.txt

AA00012891_00001_0136.txt

AA00012891_00001_0103.txt

roleofsubcellula00zamb_0173.txt

roleofsubcellula00zamb_0121.txt

AA00012891_00001_0100.txt

roleofsubcellula00zamb_0005.txt

roleofsubcellula00zamb_0213.txt

E7JEYIHGD_AM9AO5_xml.txt

roleofsubcellula00zamb_0224.txt

AA00012891_00001_0204.txt

AA00012891_00001_0168.txt

roleofsubcellula00zamb_0176.txt

roleofsubcellula00zamb_0046.txt

roleofsubcellula00zamb_0172.txt

AA00012891_00001_0135.txt

roleofsubcellula00zamb_0209.txt

roleofsubcellula00zamb_0073.txt

AA00012891_00001_0040.txt

AA00012891_00001_0044.txt

roleofsubcellula00zamb_0112.txt

AA00012891_00001_0022.txt

AA00012891_00001_0072.txt

AA00012891_00001_0012.txt

roleofsubcellula00zamb_0122.txt

AA00012891_00001_0108.txt

AA00012891_00001_0205.txt

roleofsubcellula00zamb_0223.txt

roleofsubcellula00zamb_0047.txt

AA00012891_00001_0051.txt

AA00012891_00001_0002.txt

roleofsubcellula00zamb_0126.txt

AA00012891_00001_0200.txt

AA00012891_00001_0031.txt

AA00012891_00001_0014.txt

roleofsubcellula00zamb_0174.txt

AA00012891_00001_0075.txt

AA00012891_00001_0147.txt

AA00012891_00001_0079.txt

roleofsubcellula00zamb_0060.txt

roleofsubcellula00zamb_0157.txt

roleofsubcellula00zamb_0130.txt

AA00012891_00001_0201.txt

AA00012891_00001_0219.txt

AA00012891_00001_0088.txt

AA00012891_00001_0102.txt

roleofsubcellula00zamb_0052.txt

AA00012891_00001_0078.txt

roleofsubcellula00zamb_0035.txt

AA00012891_00001_0187.txt

roleofsubcellula00zamb_0012.txt

AA00012891_00001_0074.txt

roleofsubcellula00zamb_0154.txt

roleofsubcellula00zamb_0192.txt

roleofsubcellula00zamb_0081.txt

AA00012891_00001_0010.txt

roleofsubcellula00zamb_0187.txt

AA00012891_00001_0178.txt

roleofsubcellula00zamb_0050.txt

roleofsubcellula00zamb_0031.txt

roleofsubcellula00zamb_0149.txt

roleofsubcellula00zamb_0156.txt

AA00012891_00001_0029.txt

AA00012891_00001_0150.txt

AA00012891_00001_0170.txt

AA00012891_00001_0052.txt

AA00012891_00001_0004.txt

AA00012891_00001_0226.txt

roleofsubcellula00zamb_0159.txt

roleofsubcellula00zamb_0068.txt

AA00012891_00001_0054.txt

AA00012891_00001_0131.txt

AA00012891_00001_0062.txt

roleofsubcellula00zamb_0160.txt

AA00012891_00001_0026.txt

roleofsubcellula00zamb_0166.txt

AA00012891_00001_0176.txt

roleofsubcellula00zamb_0202.txt

AA00012891_00001_0133.txt

roleofsubcellula00zamb_0095.txt

AA00012891_00001_0161.txt

AA00012891_00001_0076.txt

roleofsubcellula00zamb_0150.txt

roleofsubcellula00zamb_0045.txt

AA00012891_00001_0196.txt

roleofsubcellula00zamb_0085.txt

roleofsubcellula00zamb_0015.txt

AA00012891_00001_0121.txt

AA00012891_00001_0036.txt

roleofsubcellula00zamb_0179.txt

roleofsubcellula00zamb_0014.txt

AA00012891_00001_0155.txt

AA00012891_00001_0124.txt

roleofsubcellula00zamb_0144.txt

AA00012891_00001_0038.txt

AA00012891_00001_0071.txt

roleofsubcellula00zamb_0082.txt

roleofsubcellula00zamb_0177.txt

AA00012891_00001_0199.txt

AA00012891_00001_0030.txt

AA00012891_00001_0180.txt

AA00012891_00001_0129.txt

roleofsubcellula00zamb_0049.txt

AA00012891_00001_0191.txt

roleofsubcellula00zamb_0002.txt

AA00012891_00001_0028.txt

roleofsubcellula00zamb_0093.txt

AA00012891_00001_0024.txt

roleofsubcellula00zamb_0011.txt

AA00012891_00001_0118.txt

roleofsubcellula00zamb_0083.txt

roleofsubcellula00zamb_0097.txt

AA00012891_00001_0225.txt

roleofsubcellula00zamb_0084.txt

roleofsubcellula00zamb_0220.txt

roleofsubcellula00zamb_0128.txt

AA00012891_00001_0083.txt

AA00012891_00001_0015.txt

roleofsubcellula00zamb_0028.txt

AA00012891_00001_0091.txt

roleofsubcellula00zamb_0036.txt

AA00012891_00001_0101.txt

AA00012891_00001_0027.txt

AA00012891_00001_0132.txt

roleofsubcellula00zamb_0203.txt

AA00012891_00001_0055.txt

AA00012891_00001_0117.txt

roleofsubcellula00zamb_0153.txt

AA00012891_00001_0032.txt

AA00012891_00001_0019.txt

AA00012891_00001_0209.txt

roleofsubcellula00zamb_0094.txt

roleofsubcellula00zamb_0195.txt

AA00012891_00001_0107.txt

AA00012891_00001_0167.txt

AA00012891_00001_0139.txt

roleofsubcellula00zamb_0125.txt

AA00012891_00001_0046.txt

AA00012891_00001_0146.txt

AA00012891_00001_0037.txt

roleofsubcellula00zamb_0030.txt

AA00012891_00001_0229.txt

AA00012891_00001_0018.txt

AA00012891_00001_0142.txt

AA00012891_00001_0212.txt

AA00012891_00001_0175.txt

roleofsubcellula00zamb_0198.txt

roleofsubcellula00zamb_0007.txt

roleofsubcellula00zamb_0163.txt

AA00012891_00001_0008.txt

AA00012891_00001_0068.txt

AA00012891_00001_0181.txt

AA00012891_00001_0084.txt

roleofsubcellula00zamb_0072.txt

AA00012891_00001_0156.txt

roleofsubcellula00zamb_0107.txt

roleofsubcellula00zamb_0108.txt

roleofsubcellula00zamb_0110.txt

roleofsubcellula00zamb_0117.txt

AA00012891_00001_0106.txt

roleofsubcellula00zamb_0013.txt

AA00012891_00001_0063.txt

roleofsubcellula00zamb_0067.txt

AA00012891_00001_0194.txt

roleofsubcellula00zamb_0197.txt

AA00012891_00001_0165.txt

roleofsubcellula00zamb_0151.txt

AA00012891_00001_0141.txt

roleofsubcellula00zamb_0023.txt

AA00012891_00001_0138.txt

AA00012891_00001_0069.txt

AA00012891_00001_0185.txt

AA00012891_00001_0059.txt

roleofsubcellula00zamb_0227.txt

roleofsubcellula00zamb_0131.txt

roleofsubcellula00zamb_0138.txt

roleofsubcellula00zamb_0140.txt

roleofsubcellula00zamb_0048.txt

roleofsubcellula00zamb_0054.txt

roleofsubcellula00zamb_0105.txt

AA00012891_00001_0094.txt

AA00012891_00001_0005.txt

roleofsubcellula00zamb_0145.txt

roleofsubcellula00zamb_0092.txt

roleofsubcellula00zamb_0161.txt

roleofsubcellula00zamb_0127.txt

AA00012891_00001_0122.txt

AA00012891_00001_0162.txt

AA00012891_00001_0233.txt

roleofsubcellula00zamb_0167.txt

roleofsubcellula00zamb_0038.txt

roleofsubcellula00zamb_0100.txt

AA00012891_00001_0090.txt

roleofsubcellula00zamb_0170.txt

roleofsubcellula00zamb_0228.txt

AA00012891_00001_0195.txt

roleofsubcellula00zamb_0120.txt

AA00012891_00001_0060.txt

AA00012891_00001_0085.txt

AA00012891_00001_0111.txt

AA00012891_00001_0130.txt

AA00012891_00001_0058.txt

AA00012891_00001_0198.txt

roleofsubcellula00zamb_0221.txt

roleofsubcellula00zamb_0070.txt

roleofsubcellula00zamb_0099.txt

roleofsubcellula00zamb_0041.txt

AA00012891_00001_0073.txt

AA00012891_00001_0070.txt

AA00012891_00001_0210.txt

AA00012891_00001_0193.txt

roleofsubcellula00zamb_0162.txt

roleofsubcellula00zamb_0079.txt

roleofsubcellula00zamb_0215.txt

roleofsubcellula00zamb_0059.txt

AA00012891_00001_0190.txt

roleofsubcellula00zamb_0186.txt

roleofsubcellula00zamb_0185.txt

AA00012891_00001_0140.txt

AA00012891_00001_0145.txt

AA00012891_00001_0125.txt

AA00012891_00001_0047.txt

AA00012891_00001_0095.txt

roleofsubcellula00zamb_0208.txt

roleofsubcellula00zamb_0062.txt

AA00012891_00001_0174.txt

AA00012891_00001_0034.txt

roleofsubcellula00zamb_0103.txt

roleofsubcellula00zamb_0222.txt

roleofsubcellula00zamb_0142.txt

roleofsubcellula00zamb_0032.txt

AA00012891_00001_0089.txt

AA00012891_00001_0093.txt

AA00012891_00001_0050.txt

AA00012891_00001_0230.txt

roleofsubcellula00zamb_0061.txt

AA00012891_00001_0023.txt

AA00012891_00001_0203.txt

roleofsubcellula00zamb_0165.txt

AA00012891_00001_0033.txt

roleofsubcellula00zamb_0158.txt

roleofsubcellula00zamb_0218.txt

AA00012891_00001_0082.txt

AA00012891_00001_0123.txt

roleofsubcellula00zamb_0025.txt

AA00012891_00001_0218.txt

roleofsubcellula00zamb_0022.txt

roleofsubcellula00zamb_0051.txt

roleofsubcellula00zamb_0229.txt

roleofsubcellula00zamb_0194.txt

roleofsubcellula00zamb_0039.txt

roleofsubcellula00zamb_0111.txt

AA00012891_00001_0157.txt

AA00012891_00001_0013.txt

roleofsubcellula00zamb_0004.txt

roleofsubcellula00zamb_0006.txt

AA00012891_00001_0137.txt

roleofsubcellula00zamb_0135.txt

roleofsubcellula00zamb_0037.txt

roleofsubcellula00zamb_0230.txt

roleofsubcellula00zamb_0024.txt

AA00012891_00001_0104.txt

AA00012891_00001_0232.txt

roleofsubcellula00zamb_0124.txt

roleofsubcellula00zamb_0196.txt

AA00012891_00001_0207.txt

AA00012891_00001_0215.txt

roleofsubcellula00zamb_0017.txt

roleofsubcellula00zamb_0096.txt

roleofsubcellula00zamb_0069.txt

roleofsubcellula00zamb_0114.txt

roleofsubcellula00zamb_0143.txt

AA00012891_00001_0153.txt

AA00012891_00001_0077.txt

AA00012891_00001_0148.txt

AA00012891_00001_0173.txt

AA00012891_00001_0110.txt

roleofsubcellula00zamb_0044.txt

roleofsubcellula00zamb_0057.txt

AA00012891_00001_0163.txt

AA00012891_00001_0159.txt

roleofsubcellula00zamb_0063.txt

roleofsubcellula00zamb_0098.txt

roleofsubcellula00zamb_0056.txt

roleofsubcellula00zamb_0086.txt

AA00012891_00001_0099.txt

roleofsubcellula00zamb_0089.txt

AA00012891_00001_0041.txt

AA00012891_00001_0143.txt

roleofsubcellula00zamb_0216.txt

roleofsubcellula00zamb_0104.txt

roleofsubcellula00zamb_0226.txt

AA00012891_00001_0227.txt

AA00012891_00001_0134.txt

AA00012891_00001_0064.txt

AA00012891_00001_0188.txt

roleofsubcellula00zamb_0076.txt

roleofsubcellula00zamb_0066.txt

roleofsubcellula00zamb_0019.txt

AA00012891_00001_0217.txt

AA00012891_00001_0001.txt

roleofsubcellula00zamb_0077.txt

AA00012891_00001_0171.txt

roleofsubcellula00zamb_0042.txt

roleofsubcellula00zamb_0018.txt

roleofsubcellula00zamb_0175.txt

AA00012891_00001_0009.txt

roleofsubcellula00zamb_0080.txt

AA00012891_00001_0221.txt

AA00012891_00001_0081.txt

AA00012891_00001_0213.txt

AA00012891_00001_0222.txt

AA00012891_00001_0006.txt

AA00012891_00001_0152.txt

roleofsubcellula00zamb_0133.txt

AA00012891_00001_0067.txt

AA00012891_00001_0172.txt

AA00012891_00001_0114.txt

roleofsubcellula00zamb_0055.txt

roleofsubcellula00zamb_0118.txt

AA00012891_00001_0097.txt

roleofsubcellula00zamb_0000.txt

roleofsubcellula00zamb_0182.txt

AA00012891_00001_0206.txt

AA00012891_00001_0202.txt

AA00012891_00001_0061.txt

roleofsubcellula00zamb_0008.txt

roleofsubcellula00zamb_0091.txt

AA00012891_00001_0116.txt

roleofsubcellula00zamb_0101.txt

roleofsubcellula00zamb_0071.txt

roleofsubcellula00zamb_0168.txt

roleofsubcellula00zamb_0148.txt

AA00012891_00001_0154.txt

roleofsubcellula00zamb_0029.txt

AA00012891_00001_0128.txt

AA00012891_00001_0109.txt

AA00012891_00001_0224.txt

roleofsubcellula00zamb_0016.txt

roleofsubcellula00zamb_0232.txt

roleofsubcellula00zamb_0171.txt

roleofsubcellula00zamb_0090.txt

roleofsubcellula00zamb_0205.txt

roleofsubcellula00zamb_0001.txt

AA00012891_00001_0017.txt

roleofsubcellula00zamb_0116.txt

roleofsubcellula00zamb_0147.txt

roleofsubcellula00zamb_0065.txt

AA00012891_00001_0126.txt

roleofsubcellula00zamb_0169.txt

AA00012891_00001_0120.txt

roleofsubcellula00zamb_0115.txt

roleofsubcellula00zamb_0043.txt

roleofsubcellula00zamb_0152.txt

AA00012891_00001_0098.txt

AA00012891_00001_0208.txt

AA00012891_00001_0112.txt

AA00012891_00001_0113.txt


Full Text














THE ROLE OF SUBCELLULAR COMPARTMENTS AND STRUCTURES
IN EUKARYOTIC GENE EXPRESSION












By

GERARD PAUL ZAMBETTI


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA
1989















ACKNOWLEDGEMENTS

I would like to acknowledge the dedication of Drs.

Janet and Gary Stein to their work. Their commitment to

science has provided us with a well-equipped lab and the

funds to pursue our research. I would also like to extend

special thanks to the Stein's for their constant support and

guidance during my graduate studies. In addition, I would

like to thank Drs. Bert Flanegan, Tom Rowe and Dick Moyer

for their helpful advice as members of my graduate

committee.

My experiences with my fellow coworkers in the Stein's

lab have been very interesting, rewarding and fulfilling. I

would like to acknowledge and thank Marie Giorgio, Linda

Green, Farhad Marashi, Brian O'Donnell and Mary Beth Kennedy

for their technical guidance and assistance, and Ken Wright

and Joe Teixeira for their help dealing with the computers.

I thank Pat Naples for her appreciation and pledge of care

for "the coffee mug." I would like to thank Dave Collart,

Tim Morris and Charles Stewart for their friendship and for

their advice on raising bees, fishing, and golfing,

respectively. I would like to thank Andre van Wijnen for his

friendship, scientific expertise on and off the "Foos" ball

field, and for keeping an eye on Pat Naples and my coffee









mug. I would particularly like to thank the following

individuals for their tremendous inspiration: April, Emilie,

Ashley, Danny, Shannon, John, Aubrey, Stephan, Alan, Brooke,

Kate, Sarah and Nikki.

I would like to thank the members of the Department of

Immunology and Medical Microbiology. I will always remember

the football games in which Dr. Ed Siden led our team to an

undefeated record against the Pathology team starring Dr.

Ward Wakeland, and the late night "Nerf-ball" baseball games

in the hallway with Mark Labow, Paul Hermonat, Brian Masters

and Dick Wright.

Lastly, I would like to thank my family. The Lord has

truly blessed me with a loving and supportive family and I

fully appreciate all that they have done and sacrificed for

me during my studies.


iii















TABLE OF CONTENTS


PAGE

ACKNOWLEDGEMENTS........................................ ii

LIST Of FIGURES... ...................................... vi

LIST OF TABLES................................. .......... ix

KEY TO ABBREVIATIONS .....................................

ABSTRACT........................ .......................... xi

CHAPTERS

1) INTRODUCTION
General Background.. .............................. 1
Histone Genes and Proteins.......................... 2
Histone Gene Regulation: Transcriptional and Post-
Transcriptional Control........................ 3
Subcellular Compartments........................... 10
The Cytoskeleton: Microtubules, Intermediate
Filaments and Microfilaments........................ 12
Cytoskeletal-mRNA Interactions..................... 16
Nonmembrane-Bound and Membrane-Bound Polyribosomes:
The Signal Hypothesis........................... 19
Overview of Project............................... 22

2) GENERAL METHODS

Conditions for Enzyme Reactions.................... 28
Mammalian Cell Culture............................ 28
Plasmid DNA....................................... 30
DNA Isolation and Purification.................... 35
Isolation of Mammalian RNA........................ 40
Two-Dimensional Gel Electrophoresis and Immuno-
Blotting Analysis................................ 48
Northern Blot Analysis............................. 49
Sl Nuclease Protection Analysis.................... 52
In Vitro Nuclear Run-on Transcription Analysis..... 55
Radiolabeling DNA for Northern Blot Analysis....... 58
Site-Directed Mutagenesis.......................... 61
Transfection of DNA into Cells..................... 71
Establishment of Polyclonal and Monoclonal Stably
Transformed Cell Lines........................... 73









3) SUBCELLULAR LOCATION OF HISTONE mRNAs ON CYTOSKELETON
ASSOCIATED NONMEMBRANE-BOUND POLYSOMES IN HELA S3
CELLS
Introduction..... .................................... 75
Results.............................. .. ....... .... 77
Discussion............ .. ............ ............. 90

4) SUBCELLULAR LOCATION OF HISTONE mRNA PLAYS A ROLE IN
THE POSTTRANSCRIPTIONAL REGULATION OF HISTONE
GENE EXPRESSION
Introduction....................................... 96
Results ........ ................................ 97
Discussion.... ........ ......... .... .............. 126

5) DIFFERENTIAL ASSOCIATION OF MEMBRANE-BOUND AND
NONMEMBRANE-BOUND POLYSOMES WITH CELL CYTOSTRUCTURE
Introduction..................................... 130
Results............ ............. .................. 131
Discussion......................................... 146

6) HETEROGENEOUS PATTERN FOR CYTOSKELETAL-mRNA
INTERACTIONS
Introduction. ..................................... 153
Results........................................... 154
Discussion ........................................ 166

7) THE INFLUENCE OF THE CYTOSKELETON ON THE REGULATION
OF c-Fos GENE EXPRESSION
Introduction. .................. ................... 168
Results................. .......................... 171
Discussion ......................... ................ 190

8) SUMMARY AND FUTURE CONSIDERATIONS..................... 196

REFERENCES.............................................. 206

BIOGRAPHICAL SKETCH..................................... 219















LIST OF FIGURES


Figure Page

2-1 Outline of the cloning scheme for the construction of
the signal peptide-histone fusion gene............... 34

2-2 Outline of procedure for the isolation of subcellular
polysomal fractions................................. 45

3-1 Northern blot analysis of H3 histone, H4 histone and
HLA-B7 mRNA in subcellular fractions................. 81

3-2 Two-dimensional gel electrophoresis of cytoskeleton
and soluble associated proteins...................... 85

3-3 Northern blot analysis of H3 histone, H4 histone and
HLA-B7 mRNAs isolated from subcellular fractions of
cells treated with metabolic inhibitors.............. 89

3-4 Northern blot analysis of H3 histone, H4 histone and
HLA-B7 mRNAs associated with the cytoskeleton and
soluble fractions from HeLa cells treated with
metabolic inhibitors................................ 92

4-1 Schematic diagram of endogenous H3 histone mRNA ST519
and the beta-lactamase signal peptide-H3 histone
fusion mRNA ........................................ 100

4-2 Expression of the signal peptide-histone fusion gene
in HeLa cells....................................... 103

4-3 The subcellular localization of the signal peptide-
histone fusion mRNA................................. 106

4-4 The stability of the signal peptide-histone fusion
mRNA following inhibition of DNA synthesis.......... 109

4-5 Schematic diagram of wild type and mutated signal
peptide-histone fusion mRNAs....................... 112

4-6 Distribution of SPH3E1, SPH3Elalpha, and SPH3E1ATG
mRNAs in nonmembrane-bound and membrane-bound
polysomal fractions................................. 115

4-7 Effects of hydroxyurea treatment on SPH3E1 mRNA...... 119

vi















4-8 Effects of hydroxyurea treatment on SPH3E1ATG
mRNA ................................................. 121

4-9 Effects of hydroxyurea treatment on SPH3Elalpha
mRNA ................................................. 123

4-10 Quantitation of SPH3E1, SPH3Elalpha, SPH3E1ATG
mRNA during hydroxyurea treatment.................. 125

5-1 Northern blot analysis of the cytoskeleton and
soluble phase distribution of H3 histone, H4 histone
and HLA-B7 mRNAs in cytochalasin D treated cells.... 135

5-2 Northern blot analysis of the cytoskeleton and
soluble phase distribution of c-fos mRNA in
cytochalasin D, puromycin, CD/puro, and
CD/cycloheximide treated cells ..................... 140

5-3 Northern blot analysis of the cytoskeleton and
soluble phase distribution of nonmembrane-bound
and membrane bound polysomal mRNAs in cytochalasin D,
puromycin, CD/puro and CD/cycloheximide treated
cells................................................ 145

6-1 Cytochalasin D and puromycin cotreatment does not
release SPH3E1 or SPH3E1ATG- mRNA from the
cytoskeleton........................................ 157

6-2 The wild type signal peptide-histone fusion mRNA is
not released from the cytoskeleton in CD and
puromycin cotreated polyclonal HeLa cell cultures... 161

6-3 Endogenous membrane-bound polysomal mRNAs are
released from the cytoskeleton during cotreatment
with CD and puromycin in SPH3E1 and SPH3E1ATG-
expressing cell cultures........................... 163

7-1 Effects of cytochalasin D and puromycin treatment on
steady state levels of c-fos and chorionic
gonadotropin mRNA................................... 175

7-2 Densitometric analysis of steady state levels of
c-fos mRNA during cytochalasin D and puromycin
treatments......................................... 177


vii















7-3 Time course of c-fos mRNA accumulation during
cytochalasin D and puromycin treatments in
exponentially growing HeLa cells................... 181

7-4 Effect of cytochalasin D on transcription of c-fos
in HeLa cells...................................... 184

7-5 Effect of protein synthesis inhibition on
cytochalasin D induction of c-fos transcription in
HeLa cells........................... .... .............. 188


viii















LIST OF TABLES
Table Page

3-1 Quantity of histone and HLA-B7 mRNAs in the
subcellular fractions.................................. 82

4-1 QuAntitation of SPH3E1, SPH3Elalpha and SPH3E1ATG
mRNA in nonmembrane-bound and membrane-bound
polysome fractions ........ ....... ...... ............ 116

5-1 The percent of cytoskeleton and soluble phase
associated mRNAs in cells treated with
cytochalasin D..................................... 136

5-2 Percent distribution of nonmembrane-bound and
membrane bound polysomal RNAs in the Csk and Sol
fractions from CD, Puro, CD/puromycin and
CD/cycloheximide treated cells...................... 141

6-1 The percent of cytoskeleton associated mRNAs
isolated from cytochalasin D, puromycin and
CD/puromycin treated cells.......................... 158















KEY TO ABBREVIATIONS

BSA Bovine Serum Albumin

cpm Counts per minute

dCTP Deoxycytidine-5'-triphosphate

ddCTP Dideoxycytidine-5'-triphosphate

DNA Deoxyribonucleic acid

DNase Deoxyribonuclease

DTT Dithiothreitol

EDTA Ethylenediaminetetraacetic acid

Hepes N-2 hydroxyethyl piperazine-N'-2
ethanesulfonic acid

HU Hydroxyurea

MCi Microcurie

Mg Microgram

Al Microliter

Am Micrometer

AM Micromolar

OD Optical density

PVS Polyvinyl sulfate

RNA Ribonucleic acid

rpm Revolutions per minute

SDS Sodium dodecylsulfate

SSC Standard saline citrate buffer

Tris (hydroxymethyl)aminomethane

x















Abstract of Dissertation Presented to the Graduate School of the
University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy


THE ROLE OF SUBCELLULAR COMPARTMENTS AND STRUCTURES
IN EUKARYOTIC GENE EXPRESSION

By

Gerard Paul Zambetti

December 1989


Chairperson: Janet Stein
Major Department: Immunology and Medical Microbiology


Histone mRNAs naturally reside on nonmembrane-bound

polysomes that are associated with the cytoskeleton. To

study whether the subcellular location plays a role in the

coupling of histone mRNA stability to DNA synthesis, a

signal peptide-histone fusion gene was constructed that

encodes a histone mRNA that is directed to membrane-bound

polysomes. The signal peptide-histone fusion mRNA when

localized in the foreign subcellular compartment is no

longer destabilized when DNA synthesis is inhibited. A

single nucleotide substitution in the signal peptide-histone

fusion gene by site-directed mutagenesis resulted in the

localization of the histone fusion mRNA on nonmembrane-bound

polysomes and restored the coupling of mRNA stability with









DNA synthesis. These results indicate that the subcellular

location of histone mRNA plays an important role in the

posttranscriptional regulation of histone gene expression.

We have determined that the association of histone mRNA

with the cytoskeleton is dependent on the integrity of actin

filaments. Disruption of the microfilaments with

cytochalasin D efficiently released histone mRNA from the

cytoskeleton into the soluble phase. In contrast, membrane-

bound polysomal mRNAs remained attached to the cytoskeleton

during cytochalasin D treatment. Subsequent studies have

demonstrated that membrane-bound polysomal mRNAs are

attached to the cytoskeleton at two different sites: 1) a

cytochalasin D sensitive site and 2) a puromycin sensitive

site.

Disruption of the cytoskeleton with cytochalasin D

resulted in a rapid and marked stimulation in transcription

of the c-fos proto-oncogene and an increase in c-fos mRNA

cellular levels. Transcription of several other genes was

unaffected by the cytochalasin D treatment. These results

suggest that the nucleus can respond to the structural

organization of the cytoskeleton and selectively adjust the

regulation of gene expression accordingly.


xii















CHAPTER 1

INTRODUCTION



General Background

Histones are small basic proteins that are involved in

the packaging of eukaryotic DNA into chromatin. It is

estimated that more than 2 meters of DNA are contained

within the limited area of the nucleus. The condensation of

the DNA to such a great extent is due primarily to its

interactions with the histone proteins. In addition to their

structural properties histones play a functional role in

controlling gene transcription. The way in which the histone

proteins are arranged along a region of DNA appears to

affect the transcriptional properties of that DNA domain.

The complex nature of the regulation of histone protein

synthesis at the transcriptional, posttranscriptional, and

posttranslational levels makes these genes an important

model for studying eukaryotic gene expression. However, one

of the most important reasons for studying the regulation of

histone gene expression is that the synthesis of the

majority of histone proteins (cell cycle dependent histones)

occurs during DNA replication. Elucidation of the mechanisms

coupling histone protein synthesis with DNA synthesis may











yield a better understanding of the processes controlling

proliferation, differentiation, cell aging and neoplasia.



Histone Genes and Proteins

There are five major classes of histone proteins (H4,

H2B, HIA, H3, and H1) ranging in molecular weight from

11,000 to 21,000 Daltons. They are basic proteins that have

been highly conserved throughout evolution. For example,

there are only two amino acids out of 102 which differ

between histone H4 from pea seedlings and calf thymus, which

suggests a critical role for these proteins in cell survival

(DeLange et al., 1969). The histone proteins are encoded by

a family of moderately repeated genes (Wilson and Melli,

1977; Kedes, 1979). In humans, the histone genes are

arranged in clusters that are not organized in simple tandem

repeats as found in the sea urchin and other lower

eukaryotic genomes (Heintz et al., 1981; Sierra et al.,

1982; Carozzi et al., 1984). The clusters are localized on

chromosomes 1, 6, and 12 and are interspersed with

repetitive elements (Green et al., 1984b; Collart et al.,

1985). It has been speculated that these repetitive elements

may play a role in recombination and gene duplication during

the evolution of this multi-gene family. The general

structure of the human histone genes is less complex than

many other eukaryotic genes since the histone protein coding

region is not interrupted by introns and the encoded mRNAs











are not polyadenylated (Adensik and Darnell, 1972). However,

the genes coding for histone proteins that are synthesized

in the absence of DNA synthesis (cell cycle independent

histones) generally contain introns and code for mRNAs that

are poly-adenylated (Engel et al., 1982; Harvey et al.,

1983; rush et al., 1985; Wells and Kedes, 1985). Subsequent

discussions will generalize what is true only for the cell

cycle dependent histones unless otherwise indicated.



Histone Gene Requlation: Transcriptional and Post-
Transcriptional Control


The stringent requirement to package newly replicated

DNA into chromatin, the invariable 1:1 molar ratio of

histone protein to DNA and the inability of most eukaryotic

cells to store unbound histone proteins implies a temporal

and functional coupling of histone protein synthesis with

DNA replication (Prescott, 1966; Allfrey et al., 1963).

Support for this hypothesis first came from studies on

synchronized cell cultures of Euplotes eurystomus in which

the synthesis of histone protein was found to occur only

during the S phase of the cell cycle, the time when DNA

replication takes place (Prescott, 1966). Furthermore, the

quantity of histone protein within the cell doubled during S

phase. The temporal coupling of histone protein synthesis

with DNA synthesis has since been confirmed and well

documented by a variety of direct experimental approaches











(Robbins and Borun, 1967; Spaulding et al., 1966; Stein and

Borun, 1972). In vitro translation studies using polysomal

RNA isolated from cells at various stages of the cell cycle

indicated that the histone mRNA was present exclusively

during S phase (Spalding et al., 1966; Gallwitz and Mueller,

1969; Jacobs-Lorena et al., 1972; Gallwitz and Briendl,

1972; Perry and Kelley, 1973; Borun et al., 1975).

Quantitation of histone mRNA content during the cell cycle

by Northern blot analysis, using radiolabeled histone cDNA

and genomic clones, demonstrated that histone mRNA steady

levels closely parallel the rates of histone protein

synthesis as well as the rates of DNA replication (Heintz et

al., 1983; Plumb et al., 1983a; Baumbach et al., 1984). The

timing of histone protein synthesis during DNA synthesis is

not coincidental and represents a functional relationship

between these two processes. This is supported by the

following observations: 1) there is a coordinate induction

of histone protein synthesis and DNA synthesis during

stimulation of quiescent cells to proliferate (DeLisle et

al., 1983; Green et al., 1984a); 2) inhibition of DNA

synthesis with compounds such as hydroxyurea or cytosine

arabinoside results in a loss of histone mRNA with a

concomitant decrease in histone protein synthesis (Breindl

and Gallwitz, 1973; Butler and Mueller, 1973; Borun et al.,

1975; Baumbach et al., 1984) and 3) terminal differentiation

is marked by a down regulation in histone protein synthesis











in conjunction with the decrease in proliferative activity

(Bird et al., 1985; Challoner et al., 1989; Stein et al.,

1989).

The expression of histone genes during the cell cycle

is regulated at both the transcriptional and

posttranscriptional levels. Analysis by hybrid selection of

RNA radiolabeled in synchronized HeLa cells demonstrates a

burst of 3- to 5- fold in transcription of the histone genes

within the first two hours of S phase, followed by a

reduction to a basal level that remains constant until the

following S phase (Plumb et al., 1983a). Similar results

were obtained when histone gene transcription during the

cell cycle was measured by nuclear run-on analysis (Sive et

al., 1984; Baumbach et al., 1987). Steady state levels of

histone mRNA, however, were elevated 15-20 fold during S

phase and closely paralleled the rates of DNA synthesis

(Plumb et al., 1983a; Heintz et al., 1983; Baumbach et al.,

1984). Maximal accumulation of histone mRNA occurred at 5

hours into S phase and returned to a basal level as DNA

synthesis ended. The timing and extent of the increase in

histone transcription can not account for the marked

increase in histone mRNA steady state levels and must be a

consequence of a posttranscriptional control mechanism.

Posttranscriptional regulation of histone gene expression is

further supported by DNA synthesis inhibitor studies. The

estimated half life of histone mRNA in exponentially growing











HeLa cells is 45-60 minutes (Sittman et al., 1983; Plumb et

al., 1983). Hydroxyurea, or any other compound or treatment

that inhibits DNA synthesis studied to date, results in a

rapid and selective destabilization of histone mRNA (half

life of 10-15 minutes) with little or no effect on histone

gene transcription (Sittman et al., 1983; Plumb et al.,

1983; Baumbach et al., 1987).

A long-standing observation concerning histone mRNA

turnover during inhibition of DNA synthesis is that the

mechanism is completely dependent on protein synthesis

(Butler and Mueller, 1973; Gallwitz, 1975; Stahl and

Gallwitz, 1977; Stimac et al., 1983; Baumbach et al., 1984;

Helms et al., 1984). Pretreatment of cells with inhibitors

of protein synthesis such as high salt, cycloheximide or

puromycin followed by the addition of a DNA synthesis

inhibitor prevents the destabilization of histone mRNAs and

actually results in an increase in their steady state levels

due to the continued transcription of the histone genes.

This observation led Butler and Mueller and others to

hypothesize that the histone proteins can autoregulate their

own synthesis through a feedback loop operative at the

posttranscriptional level (Butler and Mueller, 1973; Stein

and Stein, 1984; Wu and Bonner, 1985). The model predicts

that at the natural end of S phase or when DNA synthesis is

interrupted with inhibitors a transient accumulation of

unbound histone protein may arise due to a loss in high-











affinity binding sites for the newly synthesized histone

proteins. Unbound histone protein may then interact with

histone mRNA-containing polysomes resulting in a

conformational change in the mRNA that is more accessible to

ribonuclease. Growing support for this model comes from a

cell-free mRNA decay system in which the destabilization of

histone mRNA is accelerated when incubated in the presence

of histone proteins (Peltz and Ross, 1987). This result

appears to be specific for histone mRNA since the rates of

decay for gamma globin mRNA, c-myc mRNA, total poly(A)" RNA

or total poly(A)+ RNA are unaffected by the presence of

histone proteins.

Gene transfer studies demonstrated that the minimal

sequence necessary for coupling histone mRNA stability with

DNA replication lies within the 3' terminus of the mRNA

(Stauber et al., 1986; Levine et al., 1987; Pandey and

Marzluff, 1987). Removal of only 60 nucleotides from the 3'

end of a murine H3 histone mRNA releases the mutant mRNA

from destabilization during inhibition of DNA synthesis

(Levine et al., 1987). Conversely, the last 30 nucleotides

of histone H3 mRNA when fused to the 3' end of human gamma

globin mRNA are sufficient to confer instability to the

chimeric mRNA when DNA synthesis is inhibited, whereas wild

type gamma globin mRNA remains stable (Pandey and Marzluff,

1987). Contained within the minimal target sequence for

destabilization is a stem-loop structure which is common to











all known cell cycle dependent histone mRNAs (Hentschel and

Birnstiel, 1981). The presence of the stem-loop structure is

not sufficient by itself to couple mRNA stability with DNA

replication. The target sequences must be present in the

proper position at the 3' terminus of the mRNA to be

operative. Aberrant histone mRNA transcripts containing the

target sequences at an internal position are fully stable

during the inhibition of DNA synthesis (Alterman et al.

1985; Luscher et al., 1985; Pandey and Marzluff, 1987;

Levine et al., 1987). Furthermore, moving the stem-loop

structure 380 nucleotides downstream from the translation

stop codon abolishes the coupling of histone mRNA stability

with DNA replication despite the retention of the stem loop

structure at the 3'terminus (Graves et al., 1987).

An additional constraint on the histone

posttranscriptional regulatory mechanism is that translation

must proceed to within close proximity of the natural

termination codon for destabilization to occur during

inhibition of DNA synthesis (Graves et al., 1987). Frame-

shift mutations in histone mRNA resulting in the premature

termination of translation more than 300 nucleotides from

the 3' terminus stabilize the messenger RNA when DNA

synthesis is interrupted (Graves et al., 1987). Other frame-

shift mutants that allow translation to continue past the

natural UGA stop codon into the 3' untranslated region are

also stable when DNA synthesis is inhibited (Capasso et al.,











1987). Correcting for translation read through in these

mutants by inserting stop codons in all three reading frames

at the natural translation termination site restores the

coupling of mRNA stability to DNA replication.

Detailed studies on deletion mutants suggest that the

histone nascent polypeptide is not involved in the

posttranscriptional regulatory process (Capasso et al.,

1987). Systematic removal of approximately 100 nucleotide

blocks spanning nearly the entire protein coding region had

no effect on the destabilization of these mutant histone

mRNAs when DNA synthesis was inhibited. Consistent with this

result, the deletion of the ATG translation start codon

which prevents the synthesis of histone protein did not

prevent the destabilization of the mRNA during inhibition of

DNA replication. Furthermore, as discussed above, the

minimal target sequences containing the stem-loop structure

confer instability to gamma globin mRNA when DNA synthesis

is interrupted (Pandey and Marzluff, 1987). The globin-

histone chimeric mRNA, although regulated as wild type

histone mRNA, does not encode any histone protein sequences.

Destabilization of histone mRNA appears to occur in a

stepwise manner initiating at the 3' terminus (Ross et al.

1986). The first detectable histone mRNA degradation product

lacks 5 nucleotides from the 3' terminus (Ross and Kobs,

1986; Ross et al., 1986). This degradation product is

detectable in the in vitro mRNA decay system as well as in









10

whole cells (Ross et al., 1986). It is unclear whether this

product arises from a single endonucleolytic scission or

from multiple cuts removing one or more of the nucleotides

at a time. Degradation continues rapidly in a 3' to 5'

direction carried out by an exonuclease which is part of the

ribosome complex (Ross et al., 1987). These results are

consistent with the requirement for translation to proceed

near the 3' terminus for the destabilization of the histone

mRNA during inhibition of DNA synthesis.



Subcellular Compartments

The appearance of histone mRNA on polysomes within

minutes after transcription and the equally rapid transfer

of the newly synthesized histone proteins into the nucleus

where they complex with DNA imply that histone protein

synthesis may be compartmentalized in a region located just

outside the nucleus (Spaulding et al., 1966; Borun et al.,

1967; Robbins and Borun, 1967). The asymmetric distribution

of eukaryotic mRNAs in the cytoplasm supports this

possibility. Morphological studies initially demonstrated

that polyribosomes are concentrated in areas surrounding the

nucleus in mouse 3T3 cells (Fulton et al., 1980). Later,

localization of specific mRNAs by in situ hybridization

techniques provided direct evidence that mRNAs are

differentially distributed throughout the cell. Tubulin mRNA

is preferentially located in the periphery of the cytoplasm









11

whereas vimentin mRNA is concentrated in perinuclear regions

in chick myoblast and fibroblast cells (Lawrence and Singer,

1986). Furthermore, actin mRNA is concentrated in the

lamellipodia, structures involved in cell locomotion

(Lawrence and Singer, 1986). The requirement for actin

protein in the lamellipodia during cell movement (Wang,

1984) suggests a functional relationship between where the

mRNA is localized and where the encoded protein is needed

(Lawrence and Singer, 1986).

Localization of mRNAs in distinct areas of the

cytoplasm may occur due to an association of the mRNA with

an underlying structure. This structure could serve as an

anchor to concentrate mRNAs in particular regions of the

cytoplasm. Consistent with this reasoning is the existence

of a complex, proteinaceous structure in the cytoplasm of

most eukaryotic cells that is referred to as the

cytoskeleton. The cytoskeleton plays a role in cell shape,

cell motility and the intracellular transport of

macromolecules and organelles (Clarke and Spudich, 1977;

Goldman et al., 1978; and reviewed in Nielsen et al., 1983).

In addition, there is growing support for the hypothesis

that the cytoskeleton plays a key role in regulating the

translation process (Lenk et al., 1977; Lenk and Penman,

1979; Fulton et al., 1980; Cervera et al., 1981; van Verooij

et al., 1981; Nielsen et al., 1983; Howe and Hershey, 1984).











The Cytoskeleton: Microtubules. Intermediate Filaments
and Microfilaments


The cytoskeleton structure is an interconnected system

of three major classes of protein fibers, namely the

microtubules, intermediate filaments and microfilaments.

Each system represents a heterogenous population of

proteins. The microtubules, for example, are composed of 13

protofilaments of the well characterized alpha and beta

tubulin proteins (Amos and Klug, 1974). These filaments are

25 nm in diameter and appear to contain a hollow core.

Several microtubule associated proteins (MAPs) have also

been identified which appear to facilitate the assembly of

the microtubules and promote their interactions with other

cellular organelles (Sloboda et al., 1976, Bulinski and

Borisy, 1980; Bloom et al., 1984; Binder et al., 1985). It

has long been recognized that in mitotic cells the

microtubules reorganize into the mitotic spindle including

the astral, kinetochore and continuous fibers which mediate

the segregation of the chromosomes (Weber et al., 1975).

During interphase the role of the microtubules is less clear

and somewhat controversial. Studies on agents that cause the

depolymerization of the microtubules, such as colchicine and

colcemid, suggest that the microtubules play a role in cell

shape and the intracellular transport of macromolecules and

organelles (Hayden and Allen, 1984; Vale et al., 1985).











There are at least five major classes of protein that

constitute the intermediate filaments: 1) desmin, 2)

vimentin, 3) keratin, 4) neurofilaments and 5) glial

filaments. Biochemical and immunological studies

demonstrated that these proteins are generally expressed in

a celltype specific manner (for review see Lazarides,

1980). Desmin is an acidic protein with a molecular weight

of 53,000 Daltons and is predominantly found in intermediate

filaments from cardiac, smooth and skeletal muscle cells

(Lazarides and Hubbard, 1976; Hubbard and Lazarides, 1979).

The localization of desmin at the Z line in muscle suggests

a role for this protein in the coordination of the

contractile actions of the muscle fiber (Lazarides and

Hubbard, 1976; Izant and Lazarides, 1977).

Vimentin is found principally in intermediate filaments

from cells of mesenchymal origin. Immunological and

biochemical characterization of vimentin indicated that the

protein is 52,000 Daltons and widely conserved throughout

evolution (Brown et al., 1976; Bennett et al., 1978; Franke

et al., 1979). A characteristic feature of the vimentin

intermediate filaments is their sensitivity to colchicine

which erroneously led researchers to believe initially that

they were disaggregation products of the microtubules

(Goldman and Knipe, 1973; Holltrop et al., 1974; Blose and

Chacko, 1976). The collapse of vimentin into filamentous

bundles around the nucleus during colchicine treatment









14

suggests that these filaments are in close association with

the microtubules.

The keratin filaments, also known as tonofilaments, are

predominantly found in epithelial cells. The keratin

proteins are a family of approximately 30 distinct

polypeptides ranging in molecular weight from 40,000-65,000

Daltons (Steinert, 1975; Steinert and Gullino, 1976; Sun and

Green, 1978; Moll et al., 1982). The keratins can be clearly

divided into acidic and basic groups based on two-

dimensional electrophoretic analysis. Although only a few

keratin subtypes are expressed in any one epithelial cell,

there is always a coordinated synthesis of at least one

class each of the acidic and basic keratin subunits (Moll et

al., 1982; Woodcock-Mitchell et al., 1982; Fuchs and

Marchuk, 1983). The combination of keratin subtypes

expressed by an epithelial cell is characteristic of the

tissue origin as well as the differentiation state of the

cell (Roop et al., 1983; Sun et al., 1983). In addition,

some epithelially derived tumors express different keratin

subtypes than those found in the normal counterpart and

these differences may be useful in diagnostic tests

(Summerhayes et al., 1981; Osborn, 1983).

Little is known about the function and organization of

the intermediate filaments from neuronal and glial cells.

The major cytoskeletal element of neuronal axons and

dendrites are the neuronal filaments which are generally









15

composed of three large polypeptides with molecular weights

of 200,000, 150,000, and 68,000 Daltons (Hoffman and Lasek,

1975; Delacourte et al., 1980). Glial filaments are

specifically found in astrocytes and other types of glial

cells and consist of a single protein, the glial fibrillary

acidic protein (GFA), with a molecular weight of 51,000

Daltons (Schachner et al., 1977; Goldman et al., 1978).

Although the intermediate filament proteins from the

different cell types are heterogeneous in their chemical

composition, they assemble into similar morphological

structures consisting of filaments that are 10 nm in

diameter that are highly resistant to extraction with

detergents. It has been proposed that the intermediate

filaments contain a constant and variable region, analogous

to the structure of antibodies. The constant domain would

conserve the morphology whereas the variable domain would

confer functional specificity (Lazarides, 1980).

The microfilaments are the thinnest of the cytoskeletal

filament systems, with an average diameter of 6 nm, and are

composed primarily of actin protein. It has been estimated

that approximately 5-10% of total cellular protein is actin

which exists in an equilibrium between monomeric and

filamentous forms (Hartwig and Stossel, 1975).

Microfilaments from non-muscle cells contain two predominant

species of actin protein, namely beta- and gamma-actin. The

non-muscle actin proteins are quite similar to muscle











specific alpha-actin with respect to molecular weight

(42,000 Daltons), amino acid sequence and polymeric

structure (Tilney and Detmers, 1975; Hartwig and Stossel,

1975; Kane, 1975; Garrels and Gibson, 1976; Whalen et al.,

1976; Rubenstein and Spudich, 1977). These results suggest

that beta- and gamma-actin protein in non-muscle cells may

play a role in contraction processes such as cell motility ,

chromosome movement and exocytosis (Berl et al., 1973;

Huxley, 1973; Sanger, 1975).



Cytoskeletal-mRNA Interactions

Early studies using stereo-electron microscopic

analysis of whole cells indicated that polysomes are

associated with the cytoskeleton structure (Wolosewick and

Porter, 1977). Closer examination of the cell interior, by

removing membranes and soluble cytoplasmic proteins with

detergent, reveals that many of the ultrastructural features

and nearly all the polysomes are retained on the

cytoskeletal framework (Lenk et al., 1977). Furthermore, the

non-random topological distribution of the polysomes is

preserved in the cytoskeleton preparation (Fulton et al.,

1980). The association of the polysomes with the

cytoskeleton does not appear to be due to trapping since

monosomes and independent ribosomal subunits are efficiently

removed by the detergent extraction (Lenk et al., 1977;

Cervera et al., 1981; Ornelles et al., 1986). These results









17

indicate that actively translated mRNAs are associated with

the cytoskeleton and this association may be requisite for

protein synthesis. Consistent with this possibility, the

inhibition of host protein synthesis coincides with the

release of host mRNA from the cytoskeletal structure and the

attachment of viral mRNA to the cytoskeleton in poliovirus

infected and adenovirus infected cells (Cervera et al.,

1981; Lemieux and Beaud, 1982). During development,

maternally inherited mRNAs participate in protein synthesis

only at the time they become associated with the

cytoskeleton in sea urchin oocytes (Moon et al., 1983).

The region of the polysome involved in its attachment

to the cytoskeleton is not known. Previous results indicate

that the association of polysomal mRNA with the cytoskeleton

is through the mRNA or mRNP particle and not through the

ribosome (Lenk et al., 1977; Lenk and Penman, 1979; Cervera

et al., 1981; Lemieux and Beaud, 1982). Disruption of

polysomes by heat shock, high salt, sodium fluoride or

pactamycin treatment releases the ribosomal subunits but not

the mRNA from the cytoskeleton into the soluble phase (Lenk

et al., 1977; Cervera et al., 1981; van Venrooij et al.,

1981; Howe and Hershey, 1984). Immunofluorescence studies

indicate that the cap-binding protein is associated with the

cytoskeleton suggesting that the attachment of eukaryotic

mRNAs to the cytoskeleton may occur via the 5' mRNA cap

structure (Zumbe et al., 1982). Earlier studies suggest that











the association of eukaryotic mRNAs with the cytoskeleton

may occur through the poly-A tract (Milcarek and Penman,

1974). However, the association of uncapped poliovirus mRNAs

(Bonneau et al., 1985) and poly-A- histone mRNAs (Jeffery,

1984; Bonneau et al., 1985) indicates that the attachment

site might also exist in an internal region contained within

the 5' and 3' ends of the mRNA. Isaacs and Fulton (1987)

have presented evidence suggesting that myosin heavy chain

in embryonic chicken muscle cells is co-translationally

assembled into the cytoskeleton. The myosin nascent

polypeptide would therefore serve as an additional

cytoskeleton attachment site for the polysomal myosin mRNA.

Based on these results it appears that several mechanisms

may operate in the attachment of a specific polysomal mRNA

or mRNP to the cytoskeleton and these attachment sites may

vary depending on the species of mRNA.

The component of the cytoskeleton to which the mRNA is

attached is not known although the data accumulated to date

suggest that the microfilaments may be involved in the mRNA-

cytoskeleton association. Removal of tubulin from the

cytoskeleton by using cold calcium-containing buffers during

the isolation procedure has no effect on the attachment of

mRNA with the cytoskeleton; this result suggests that the

microtubules do not play a role in the association of mRNA

with the cytoskeleton (Lenk et al., 1977; Howe and Hershey,

1984; Ornelles et al., 1986). Disruption of vimentin









19

filaments with heat shock or indirectly with colcemid has no

effect on protein synthesis as determined by two-dimensional

electrophoresis indicating that translation is not dependent

on the integrity of the intermediate filaments (Welch and

Feramisco, 1985). Other reports suggest that the attachment

of mRNA to the cytoskeleton is mediated by the

microfilaments (Lenk et al., 1977; Howe and Hershey, 1984;

Ornelles et al., 1986). Treatment with cytochalasins, either

cytochalasin B or D, to disrupt the actin-containing

microfilaments results in the release of poly-A+ mRNA from

the cytoskeleton into the soluble phase. At least in the

case of cytochalasin D treatment, the release of poly-A+

mRNA from the cytoskeleton occurs in a dose dependent manner

and stoichiometrically reflects the degree of inhibition of

protein synthesis (Ornelles et al., 1986). These results

indicate that intact microfilaments are required for the

association of mRNA to the cytoskeleton; however, it can not

be concluded that the actin fibers directly bind the mRNA.



Nonmembrane-Bound and Membrane-Bound Polyribosomes:
The Signal Hypothesis


As described above, nearly 100 percent of the actively

translated polysomes are associated with the cytoskeleton.

Studies on vesicular stomatitis virus infected HeLa cells

demonstrated that membrane-bound polysomal mRNA coding for

the viral G glycoprotein and "free" polysomal mRNA coding









20
for the polymerase, nucleocapsid and matrix proteins are all

attached to the cytoskeleton during translation (Cervera et

al., 1981). The association of free polysomes with the

cytoskeleton has also been demonstrated by electron

microscopy (Wolosewick and Porter, 1979). Free polysomes are

generally the class of polysomes involved in the synthesis

of intracellular proteins whereas membrane-bound polysomes

are generally involved in the synthesis of exported proteins

that are destined for the cell surface or secretion (for

review see Emr et al., 1980). The association of free

polysomal mRNAs with the cytoskeleton suggests that the

polysome is anchored to a scaffold, albeit a dynamic

structure, and not freely diffusible within the cytoplasm.

The term "free polysome" is operationally defined and is a

misnomer when viewed in light of the biochemical and

microscopic studies on the cytoskeleton. The proper

reference to free polysomes is nonmembrane-bound polysomes.

The mechanism by which mRNAs assemble into membrane-

bound polysomes has been described in the signal hypothesis

(Blobel and Dobberstein, 1975). Many types of signal

peptides exist that target proteins to various subcellular

membrane compartments such as the endoplasmic reticulum,

mitochondria and chloroplasts in eukaryotes as well as the

cell membrane in prokaryotes (Date et al., 1980; Schatz and

Butow, 1983; Muller and Blobel, 1984; Cline, 1986). The work

presented here is focused on events that occur at the









21

endoplasmic reticulum and on what is generally known for the

synthesis of proteins that are localized on the cell surface

or secreted from the cell. Most cell surface or secreted

proteins are initially synthesized with a segment of amino

acid residues at the N-terminus that is not usually found in

the mature protein. This segment is commonly referred to as

a signal peptide. Although signal peptides are not highly

conserved in sequence, they are fairly conserved in

structure and based on the compilation of 277 signal

peptides ranging from bacteria to higher eukaryotes, a

consensus structure has been deduced (Watson, 1984). In

general, signal peptides are 25-30 amino acids in length.

They contain a charged amino acid within the first five

residues, which is followed by a stretch of at least nine

hydrophobic amino acids, and terminate at a small uncharged

amino acid such as alanine, serine or glycine. It is

believed that the hydrophobic domain of the signal peptide

facilitates the transfer of the protein across the lipid

environment of specific intracellular membranes whereas the

small, uncharged amino acid is thought to direct the

protease to cleave the peptide from the mature protein

(Hortin and Boime, 1981).

Presumably, translation of mRNA encoding exported

proteins is initiated on nonmembrane-bound polysomes.

Following the synthesis of approximately 70 amino acids, the

point at which the entire signal peptide has emerged from









22
the ribosome, the signal recognition complex (SRP) binds to

the signal peptide and arrests translation (Walter and

Blobel, 1981). The SRP is an 11S ribonucleic acid-protein

complex composed of six non-identical polypeptides and one

small cytoplasmic RNA molecule (Walter and Blobel, 1980;

Walter'and Blobel, 1982; Walter and Blobel, 1983). The RNA

molecule of SRP is 300 ribonucleotides in length and shares

approximately 80% homology with a middle repetitive human

DNA family (Ullu et al., 1982; Li et al., 1982). With the

SRP bound to the signal peptide, the complex can then

associate with the membrane of the endoplasmic reticulum

through the interaction with the SRP receptor which is also

known as the docking protein (Gilmore et al., 1982a; Gilmore

et al., 1982b; Meyer et al., 1982a; Meyer et al., 1982b).

The SRP then dissociates from the polysome with the

resultant release of the translational block. As translation

resumes the protein is co-translocated from the cytoplasm

into the endoplasmic reticulum. During translocation, a

signal peptidase which is located in the lumen of the

endoplasmic reticulum removes the signal peptide from the

protein (Maurer and Mckean, 1978; Walter et al., 1984).



Overview of Project

The overall aim of this work is to better define the

mechanisms) involved in coupling histone protein synthesis

with DNA replication. Specifically, we have focused our











attention on the posttranscriptional regulation of histone

gene expression. The evidence cited above suggests that the

cytoplasm of eukaryotic cells is a highly organized region

which contains subcellular compartments that concentrate

translation factors as well as specific mRNAs (Fulton et

al., 1980; Howe and Hershey, 1984; Lawrence and Singer,

1986; Ornelles et al., 1986). Traditionally, the idea of

subcellular compartments refers to organelles such as the

nucleus, endoplasmic reticulum, and mitochondria. For

example, a popular biochemical definition of

compartmentation is "a simple regulatory mechanism in which

the enzymes for catabolism and biosynthesis are physically

separated by membrane-bound cell organelles" (Rawn, 1983). A

broader definition of subcellular compartmentation includes

any specialized region of the cell regardless of the

presence of a membrane and may be extended to such

components and regions as 1) nonmembrane-bound and membrane-

bound polysomes, 2) cytoskeleton and soluble phase, and 3)

spatial location (i.e., near the nucleus versus near the

plasma membrane).

Histone mRNAs and the factors that are involved in

their selective degradation during the inhibition of DNA

synthesis may be localized in a particular subcellular

compartment. Consistent with this reasoning, the subcellular

compartment may influence the posttranscriptional control of

histone gene expression. Based on this possibility, we have











specifically asked the question: Does the subcellular

location of histone mRNA play a role in the destabilization

of histone mRNA during inhibition of DNA synthesis? To

address this question, we have applied a combination of

molecular and cellular approaches first, to determine the

natural location of histone mRNA and second, to alter the

subcellular compartment to study whether the foreign

location affects histone mRNA stability. The rationale for

this approach is that directing histone mRNA to an altered

subcellular location may physically remove the message from

the factors that are involved in the selective degradation

of histone mRNA during inhibition of DNA synthesis.

Our initial studies demonstrated that histone mRNAs are

translated predominantly on nonmembrane-bound polysomes that

are associated with the cytoskeleton. The subcellular

location of the mRNA was altered biologically by

constructing a histone fusion gene that contains sequences

coding for the signal peptide of the beta-lactamase gene of

E. coli pBR322 plasmid DNA. The encoded fusion histone mRNA

was directed to membrane-bound polysomes and was not

destabilized during inhibition of DNA synthesis.

Inactivation of signal peptide function by a single

nucleotide substitution in the ATG translation start codon

of the signal peptide resulted in the association of the

mutated fusion mRNA with nonmembrane-bound polysomes and

restored the coupling of the mRNA stability with DNA











replication. These results indicated that the subcellular

location plays a role in the posttranscriptional control of

histone gene regulation.

During the course of studying histone mRNA subcellular

localization it was necessary to examine histone mRNA-

cytoskeletal interactions in detail. Previously, Ornelles et

al. (1986) reported that total poly-A+ mRNA, as a

heterogenous population, is released from the cytoskeleton

in HeLa cells by cytochalasin D treatment. We demonstrated

by Northern blot analysis that cytochalasin D treatment of

HeLa cells readily releases histone mRNA, which is not poly-

adenylated, from the cytoskeleton into the soluble phase.

Surprisingly, mRNA coding for the cell surface HLA-B7 class

I antigen remained associated with the cytoskeleton during

cytochalasin D treatment. Subsequent studies demonstrated

that HLA-B7 mRNA expresses at least two distinct

cytoskeletal attachment sites: a cytochalasin D sensitive

site and a puromycin sensitive site. The puromycin sensitive

site appears to be a feature of membrane-bound polysomal

mRNAs and is most likely the nascent polypeptide and/or

ribosome associated with the protein substructure of the

endoplasmic reticulum.

Cytochalasin D and puromycin cotreatment, conditions

that efficiently release HLA-B7 membrane-bound polysomal

mRNA and histone nonmembrane-bound polysomal mRNA from the

cytoskeleton into the soluble phase, did not dissociate the











signal peptide-histone fusion mRNA from the cytoskeleton.

The nucleotide sequences of the signal peptide-histone

fusion mRNA encoding the signal peptide contain a

cytochalasin D and puromycin insensitive cytoskeletal

attachment site which is not expressed by histone mRNA or

HLA-B7 mRNA. These results provide evidence for the

heterogeneic pattern for cytoskeleton-mRNA interactions.

Lastly, we investigated the role of the cytoskeleton in

the regulation of gene expression by studying the effects of

disrupting the microfilaments with cytochalasin D on the

transcription and steady state mRNA levels of several genes.

We determined that the transcriptional and

posttranscriptional regulation of the c-fos proto-oncogene

is quite sensitive to the perturbation of the cytoskeleton

structure. Transcription of the c-fos gene is selectively

stimulated with a concomitant increase in c-fos mRNA steady

state levels within minutes of the addition of cytochalasin

D. Consistent with previous results, the transcription of

beta-actin was also elevated in cytochalasin D treated

cells. These findings suggests that the cell monitors the

organization of the cytoskeleton and selectively regulates

gene expression accordingly.















CHAPTER 2

GENERAL METHODS



Materials

[a-3P]dCTP (-3,000 Ci/mmol), [a- P]dATP (-3,000

Ci/mmol), [gamma-32P]ATP (-3,000 Ci/mmol) and [a- S]dATP

(>1,000 Ci/mmol) radionucleotides were purchased from

Amersham, Arlington Heights, IL; polyvinylsulfonic acid

(PVS), polyethylene glycol 8,000 (PEG), XAR-5 x-ray film

were obtained from Eastman Kodak Co., Rochester, NY;

ultrapure electrophoresis grade agarose and Zeta-probe nylon

membranes were from Bio Rad, Richmond, CA; formaldehyde

solution (37% w/w) was from Fisher Scientific, Fair Lawn,

NJ; nitrocellulose was from Schleicher and Schuell, Keene,

NH; RPI scintillator ppo-popop concentrated liquid

scintillator (Liquifluor) was from Research Products

International, Elk Grove Village, IL; formamide was from

Bethesda Reasearch laboratories (BRL), Gaithersburg, MD;

acrylamide, calf intestine alkaline phosphatase, nuclease

Sl, E. coli DNA polymerase I large fragment (Klenow enzyme)

and most restriction endonucleases were obtained from

Boehringer Mannheim Biochemicals (BMB), Indianapolis, IN; T4

DNA ligase was from New England Bio Labs (NEB), Beverly, MA;











T4 polynucleotide kinase (cloned) was from United States

Biochemical Corp. (USB), Cleveland, OH; cytochalasin D,

polyoxyethylene sorbitan monopalmitate (Tween 40), sodium

deoxycholate (NaDOC), sodium dodecyl sulfate (SDS), Triton

X-100, and DNase I (DN-EP) were purchased from Sigma, St.

Louis,'MO; lysozyme was purchased from Worthington

Biochemicals, Freehold, N.J.; fetal bovine, calf bovine and

horse sera were obtained from Gibco Laboratories, Grand

Island, NY or Flow Laboratories, McLean, VA; cycloheximide

was kindly provided by Upjohn, Kalamazoo, Michigan.



Conditions for Enzyme Reactions

The buffers and incubation conditions for each of the

restriction endonucleases and DNA modifying enzymes used in

this study were as described by the specific recommendations

of the manufacturers unless otherwise indicated.



Mammalian Cell Culture

HeLa Monolayer Cell Cultures

HeLa S3 suspension cells (human cervical carcinoma cell

line; ATCC ccl 2.2) were seeded into EMEM (Eagles-modified

minimum essential medium) containing 5% fetal calf serum, 5%

horse serum, 100 U/ml penicillin, 100 Ag/ml streptomycin and

1 mM glutamine at 3x106 cells per 10 cm tissue culture dish

and incubated at 370C under 5% CO2. Cells were maintained at

sub-confluent densities by splitting the cultures 1:10 into

fresh medium every 3 to 5 days.











HeLa Suspension Cell Cultures

HeLa S3 cells were grown and maintained in suspension

at 3-6x105 cells/ml in SMEM (Joklik-modified minimum

essential medium) supplemented with 7% calf serum, 100 U/ml

penicillin, 100 Cg/ml streptomycin and 1 mM glutamine in a

warm room at 370C. Monoclonal and polyclonal HeLa S3 cell

lines, which were isolated during G418 selection of

transfected monolayer cultures, were maintained in SMEM

containing 5% fetal calf serum, 5% horse serum, 100 U/ml

penicillin, 100 ig/ml streptomycin and 1 mM glutamine at 3-

6x105 cells/ml.



WI-38 Monolayer Cell Cultures

WI-38 cells (normal, human diploid fibroblast-like

cells; ATCC ccl 75) at passage 28, were grown as monolayers

in EMEM containing 10% fetal calf serum, 100 U/ml

penicillin, 100 ig/ml streptomycin and 1 mM glutamine at

37C under 5% CO2. Cell cultures were made quiescent by

maintaining cultures in EMEM containing 0.5% fetal calf

serum, 100 U/ml penicillin, 100 gg/ml streptomycin and 1 mM

glutamine at 370C under 5% CO2 for 48 hours.



Metabolic Inhibitor Treatments

Inhibition of DNA synthesis was usually performed with

1-5 mM hydroxyurea for 30 to 60 minutes. The drug was always

prepared fresh in IX spinner salts (Gibco) as a 100 mM stock

solution.











Protein synthesis was inhibited with 10 gg/ml

cycloheximide for varying lengths of time. Cycloheximide

stock solutions [500 gg/ml] were prepared fresh in IX

spinner salts. Protein synthesis was also inhibited with 0.4

mM puromycin for varying lengths of time. Puromycin stock

solutions [54.4 mg/ml] were prepared in dimethyl sulfoxide

(DMSO) and stored at -200C.

To study the dependence of protein synthesis on the

destabilization of histone mRNA during the inhibition of DNA

synthesis, we pre-treated cells with 10 gg/ml cycloheximide

for 5 minutes and then added 1 mM hydroxyurea for an

additional 30 minutes.

Disruption of the cytoskeleton was generally achieved

by treating the cells with 5 to 60 gg/ml cytochalasin D for

15-20 minutes at 370C. The cytochalasin D stock solutions

[5-10 mg/ml] were prepared in DMSO and stored at -200C.



Plasmid DNA

The human histone genes used in this study were

previously isolated by Sierra et al. from a lambda Ch4A

bacteriophage library (1982). The histone genes were

subsequently subcloned into Escherichia coli pBR322 plasmid

DNA and propagated in E. coli strain HB101 bacterium (Sierra

et al., 1982; Plumb et al., 1983; Stein et al., 1984).

Recombinant plasmids ST519 and FF435C each contain a

distinct genomic DNA fragment encoding a human H3 histone











gene. The recombinant plasmid FO108A contains an entire

human H4 histone gene which includes the promoter, coding

and 3' non-transcribed regions.

An HLA-B7 cDNA clone, pDP001, was generously supplied

by Dr. Sherman Weissman, Yale University, New Haven,

Connecticut (Sood et al., 1981). The pDP001 plasmid DNA was

constructed from the human cell line RPMI 4265 and contains

a 1400 bp insert coding for HLA-B7 in the PstI site of

pBR322 plasmid DNA. Therefore, the plasmid confers

tetracycline resistance and not ampicillin resistance to E.

coli HB101 bacteria.

The phCGa plasmid DNA was generously provided by Dr.

John Fiddes, California Biotechnology Inc., Mountain View,

California and Dr. John Nilson, Case Western Reserve

University, Cleveland, Ohio. The phCGa plasmid is a cDNA

clone containing a 621 bp fragment coding for the alpha

subunit of human chorionic gonadotropin inserted into the

HindIII site of pBR322 and confers ampicillin resistance

(Fiddes and Goodman, 1979).

The beta-lactamase signal peptide-globin fusion gene,

pSP125e, was kindly provided by Dr. Vishwanath Lingappa,

University of California at San Francisco. The pSP125e clone

contains the signal peptide from the beta-lactamase gene of

E. coli plasmid pBR322 fused in frame to the coding sequence

for chimpanzee alpha globin. This fusion gene was subcloned

into the pSP64 vector (Promega Biotech) as a HindIII-PstI









32
fragment and is under control of the SP6 promoter (Simon et

al., 1987).

To study the regulation of histone mRNA localized on

membrane-bound polysomes, we constructed a beta-lactamase

signal peptide-histone fusion gene as outlined in Figure 2-

1. The first intermediate, pSPpSTdeltaHH/E, involved the

placement of the E. coli pBR322 beta-lactamase signal

peptide-globin fusion gene from pSP125e under

transcriptional control of the cell cycle-dependent H3

histone regulatory region of pST519deltaH. The 840 bp

EcoRI/HindIII fragment from pST519deltaH (Marashi et al.,

1986) was isolated electrophoretically and blunt-ended using

the Klenow fragment of DNA polymerase I. The signal peptide-

globin fusion construct was digested with BglII restriction

endonuclease, which cuts once in the untranslated leader

region of the fusion gene, blunt-ended with Klenow fragment

and dephosphorylated using calf intestinal alkaline

phosphatase. The 840 bp blunt-ended fragment from pST519 was

ligated into the BglII blunt-ended site of pSP125e using T4

DNA ligase. The construction of the second intermediate,

pSPpSTdeltaHH/EpST, involved the substitution of the

chimpanzee globin coding sequences in pSPpSTdeltaHH/E with

the H3 histone coding sequences from pST519. The

pSPpSTdeltaHH/E DNA was digested with NcoI, which cuts at

the signal peptide-globin coding region junction, and

treated with calf intestinal phosphatase. The pST519 DNA was









Figure 2-1. Outline of the cloning scheme for the
construction of the signal peptide-histone fusion gene.

Restriction endonuclease sites are as follows: B,
BglII; E, EcoRI; H, HindIII; N, NcoI; P, PstI; S, SmaI.
Abbreviations: S.P., E. coli pBR322 beta-lactamase signal
peptide coding sequences; H3REG, transcriptional H3 histone
regulatory region including CAAT and TATAA consensus
sequences; CAP, histone H3 mRNA transcription start site; H3
struct, sequences coding for H3 histone protein; ATG,
translation start codon; TAA, translation stop codon; SP6,
bacteriophage SP6 promoter.













ATG
SI--S.P-- N-Globin--- nSP125e


,1


/2 Leader'

CAP
STATAA I H
T 795bp 125bp |20bp I


EcolHind
Isolate E/H fragment
Klenow


pST519AH
lr\


//2 Lead


SP6
CAP
~ TATAA j,.1 j
H3 1 1-- ,I S.P4GLOBIN I pSPpSTAHH/E


(ATG)
E CAP N H
kH3 Reg- H3 Struct.-- pST519


Nco I ,
Phosphatase

Ncol


Ligate


.SP



pS STAHH/EpST




3
HpaI
Isolate 1.1 Kb fragment
pUC8 Hinc Lgate


6
6 H3Cap H
HISeqa ATG T fp H pp, T/
43 Seq. ,I PP-" 4 -4H3 Seq. I I .IpSPpSTAHH/EpST


,SPH3


i-


40pEnhaAcer-

H3 SQ. 214H3 SEQ-?*p -
N*210bp "


EcoR I
EcoRI Isolate Enhancer cassette
Phosphatase

Ligate


H3 Struct.


P SPE
>F pSPH3E


SV0ES


t -









35

digested with NcoI, which cuts at the ATG translation start

codon, and ligated to NcoI/dephosphorylated pSPpSTdeltaHH/E

DNA using T4 DNA ligase. The signal peptide-histone fusion

gene was subcloned into pUC8 plasmid DNA by lighting the 1.1

kbp HpaI fragment from pSPpSTdeltaHH/EpST into the HincII

site of pUC8 to form pSPH3. The signal peptide-histone

fusion gene in pSPH3 was transcriptionally enhanced by

placing the SV40 viral enhancer element from pSVE108A

(Shiels et al., 1985) into the EcoRI site in the polylinker

of pSPH3 to form pSPH3El (containing one enhancer element)

and pSPH3E2 (containing 2-3 enhancer elements).



DNA Isolation and Purification

Plasmid DNA Preparations

Plasmid DNAs used in the experiments described here

were harbored in Escherichia coli strain HB101 bacteria (F-,

recA 13) unless otherwise indicated. Bacteria containing the

appropriate plasmid were grown in medium containing

antibiotics (usually 25-50 Ag/ml ampicillin or tetracycline)

as specified by the genotype of the plasmid or bacterium.

The following procedure is the routine method for

isolating pBR322 derived plasmid DNAs and includes an

amplification step with chloramphenicol for higher yields of

DNA (Maniatis et al., 1982a). Plasmids that are derived from

pUC8 DNA are also prepared as follows; however, due to their

high copy number the amplification step is unnecessary. The









36

standard protocol for the isolation of pUC8 related plasmid

DNAs is the same as described below except that the 500 ml

culture is inoculated directly and allowed to grow overnight

into the late log stage.

Starter cultures were prepared by inoculating 5 ml LB

mediumn(Luria-Bertani medium: 1% Bacto tryptone, 0.5% Bacto

yeast extract, 1% NaCl) supplemented with 0.1% glucose with

25-50 ul of bacterial glycerol stock and incubated overnight

at 370C. The 5 ml starter culture was transferred to 500 ml

pre-warmed LB medium supplemented with glucose and incubated

at 370C until the optical density at 590 nm (0.D.590) of the

culture reached 0.55 units. Plasmid DNA was amplified by

treating the culture with 200 gg/ml chloramphenicol

(prepared as a 20 mg/ml stock solution in 95% ethanol and

filter sterilized) at 370C for 12-16 hours. The bacteria

were centrifuged at 5,000 rpm, 40C for 10 minutes in a

Beckman JA-10 rotor and J2-21 centrifuge. The cell pellet

was resuspended in 10 ml solution I (50 mM glucose, 25 mM

Tris pH 8.0, 10 mM EDTA, 5 mg/ml lysozyme) and incubated at

room temperature for 5 minutes. The cells were lysed at 0C

with 20 ml solution II (1% SDS, 0.5 M NaOH) for 5 minutes.

The cell lysate was treated with 15 ml 5 M potassium

acetate, pH 4.8 and kept on ice for 5 minutes. The lysate

was transferred to a 50 ml polycarbonate tube and

centrifuged at 16,000 rpm for 20 minutes at 40C in Beckman

JA-20 rotor to pellet the bacterial DNA and cell debris. The









37

supernatant was transferred to a 150 ml Corex tube (Corning)

and nucleic acids were precipitated from solution with 30 ml

isopropanol at room temperature for 15 minutes. The DNA was

pelleted by centrifugation at 5,000 rpm in Beckman JS-7.5

rotor for 30 minutes at 200C. The pellet was briefly dried

under vacuum and resuspended in 8 ml 10:1 TE buffer (10 mM

Tris pH 8.0, 1 mM EDTA). The DNA solution was adjusted to a

final concentration of 1 mg/ml of CsC1 (refractive index of

1.3860) and 600 ig/ml ethidium bromide. The DNA solution was

then transferred to a 16 x 76 mm Quick-Seal tube (Beckman)

and centrifuged to equilibrium at 45,000 rpm in a Beckman

Type-50 Ti rotor and Beckman L5-65 ultracentrifuge for 36-48

hours at 200C. The supercoiled plasmid DNA was removed from

the cesium chloride gradient by puncturing the side of the

tube just below the DNA band with a syringe containing a 22

gauge needle. The DNA was extracted repeatedly with 1-

butanol to remove the ethidium bromide and to reduce the

volume of the solution. Cesium chloride was removed by

dialysis against 10:1 TE buffer at 40C. The DNA was

quantitated by absorbance at 260 nm and stored at -200C.

Routinely, the yield for pBR322 derived plasmids was 500-800

Ag DNA per 500 ml amplified culture.



M13 Bacteriophage DNA Preparations

M13 is a filamentous DNA bacteriophage that is specific

for male E. coli bacteria. M13 phage DNA is useful for site











directed mutagenesis and dideoxy DNA sequencing reactions

since it can be readily isolated as a double-stranded

covalently closed circular DNA replicative form (RF) from

the infected bacteria and as a circular, single-stranded DNA

genome (SS) from the mature phage particle. In addition, M13

phage DNA has been altered genetically to facilitate cloning

and subsequent manipulations. One such derivative, M13mpl8,

is small in size, maintained in high copy number

(approximately 200 RF molecules per cell), contains a

convenient polylinker for the insertion of foreign DNA (up

to 25 kb), and expresses beta-galactosidase which is

inactivated by the insertion of DNA into the polylinker

site.



Single-stranded templates. Single-stranded M13 DNA

templates were prepared from individual "plaques" which

formed on a lawn of E. coli strain JM103 bacteria

(delta(lac-proAB), supE / F'[traD36, proA+, proB+, lacIq

lacZdeltaM15]) that were infected with M13 bacteriophage or

transfected with M13 double stranded, covalently closed

circular DNA. A single plaque was removed from the agar

plate with a sterile pasteur pipette and placed in 2 ml YT

medium (0.8% Bacto-tryptone, 0.5% Bacto-yeast extract, 0.5%

NaC1). Two drops of an overnight JM103 bacterial culture

were added to the phage and incubated at 370C with agitation

for 6 hours. Longer incubation times result in deletions











within the phage DNA. The cell culture (1.5 ml) was

transferred to a microcentrifuge tube and spun in an

Eppendorf centrifuge at 12,000 x g for 10 minutes at 40C.

The supernatant (1.2 ml) was transferred to a sterile

microcentrifuge tube, 300 ul of 20% PEG, 2.5 M NaC1 were

added and the sample was incubated at room temperature for

15 minutes. The remaining 200-300 ul of supernatant were

removed and stored at -200C as a phage stock. The phage

suspension was centrifuged at 12,000 x g for 5 minutes at

room temperature and the supernatant was decanted. The

pellet was resuspended in 150 ul TES buffer (20 mM Tris-HCl

pH 7.5, 10 mM NaC1, 0.1 mM Na2EDTA). The sample was

extracted with phenol as follows: 1) added 50 ul phenol and

vortexed for 2 seconds; 2) incubated at room temperature for

5 minutes; 3) vortexed for 2 seconds; and centrifuged at

12,000 x g for 4 minutes at room temperature. DNA was

precipitated from 130 ul of aqueous phase with 5 ul 3 M

sodium acetate and 340 ul 95% ethanol on crushed dry ice for

15 minutes and collected by centrifugation at 12,000 x g,

for 15 minutes at 40C. The supernatant was discarded and the

pellet was washed in 70% ethanol at room temperature. The

sample was centrifuged at 12,000 x g, room temperature for 5

minutes. The DNA pellet was dried under vacuum, resuspended

in 80 ul TES buffer and stored at -20C.











Double-stranded templates. E. coli strain JM101

(delta(lac-proAB), supE / F' [proA+, DroB', lacIq,

lacZdeltaM15]) bacteria were grown overnight in 5 ml YT

medium at 370C with agitation. The overnight culture (1 ml)

was transferred to 100 ml prewarmed YT medium and incubated

at 37C until the O.D.590 reached 0.4 units. To assure

exponential growth the culture was again diluted 1:100 into

500 ml prewarmed YT medium and grown to an O.D.590 of 0.4

units. The culture was infected with 500 ul of M13

bacteriophage supernatant which had been prepared from an

isolated plaque as described above for the preparation of

M13 single-stranded DNA. The infected cells were incubated

at 370C with agitation for no more than 8 hours. The cells

were harvested and RF M13 DNA was isolated as described

above for the preparation of plasmid DNA. Routinely, the

yield was 300 Ag of supercoiled M13 DNA per 500 ml culture.



Isolation of Mammalian RNA

Isolation of Total Cellular RNA

The following procedure for the isolation of total

cellular RNA was developed by Dr Mark Plumb in the Stein's

laboratory primarily for work with HeLa cells. However, this

procedure has been successfully used without modification

for the isolation of total cellular RNA from human HL-60

cells, mouse 3T3-L1 cells, rat osteoblast primary cell

cultures, and Trypanosome cruzii cells.










Exponentially growing HeLa cells (5 x 107 cells) were

harvested by centrifugation at 1,500 rpm for 5 minutes at

370C in an IEC rotor. The cells were washed in phosphate-

buffered saline (PBS; 150 mM NaCl, 10 mM sodium phosphate pH

6.8) and resuspended in 3 ml lysis buffer (2 mM Tris-HCl pH

7.4, 1 mM EDTA) containing 5 gg/ml PVS. Cells were lysed in

the presence of 2.4% SDS and 88 pg/ml of proteinase K at

room temperature for 15 minutes. Lysates were adjusted to

300 mM NaCl, extracted with 5 ml Tris-saturated phenol and 5

ml CHCl3/isoamyl alcohol (IAA) (24:1, v/v) and centrifuged

at 3,000 rpm in an IEC rotor at room temperature. The

aqueous phase was then extracted with 5 ml CHCl3/IAA and

centrifuged at 3,000 rpm in an IEC rotor at room

temperature. The phenol/chloroform and chloroform

extractions were repeated as described above. Nucleic acids

were precipitated from the aqueous phase with 2.5 volumes of

95% ethanol at -20C in the presence of 53 mM potassium

acetate.

Nucleic acids were recovered from the ethanol

suspension by centrifugation at 12,000 x g and resuspended

in TCM buffer (10 mM Tris-Hcl, pH 7.4, CaCl2, 10 mM MgCl2)

for digestion with DNase I. The DNase I was dissolved in 20

mM Tris-HCl pH 7.4, 10 mM CaCl2 at 1 mg/ml, preincubated at

370C for 20 minutes and treated with 0.1 volume of

proteinase K [1 mg/ml] at 370C for 2 hours to remove

ribonuclease activity (Tullis and Rubin, 1980). The nucleic









42

acid samples were then digested with 0.1 mg/ml of proteinase

K treated DNase I for 20-30 minutes at 370C. Following DNase

I digestion, 0.05 volumes of 5 M NaCl and 0.25 volumes of

10% SDS were added to each sample and the RNA solutions were

repeatedly extracted with phenol and chloroform until the

interface between the aqueous and organic phases was clean.

RNA was precipitated from the aqueous phase with 2.5 volumes

of 95% ethanol, at -200C overnight. The RNA was collected at

12,000 rpm in a JA20 rotor at 40C for 30 minutes. The pellet

was resuspended in double distilled water, quantitated by

optical density at 260 nm and stored at -200C. Typical

yields of RNA were 400 gg per 90% confluent 10 cm plate of

HeLa S3 cells and 1 mg per 5 x 107 HeLa S3 cells with an

O.D.260/O.D.28g of 1.8 to 2.0.



Isolation of Nonmembrane-bound, Membrane-bound and Total
Polysomal RNA

The subcellular fractionation scheme used in these

studies is a modified procedure described by Venkatesan and

Steele (1972) and is outlined in Figure 2-2. Exponentially

growing HeLa cells (5 x 108 cells) were collected by

centrifugation at 1,500 rpm in an IEC rotor at 370C for 5

minutes. All subsequent procedures were carried out at 4C.

Cells were washed several times in PBS and resuspended in 25

ml RSB (10 mM Tris-HC1 pH 7.4, 10 mM NaC1, 2.5 mM MgCl2).

After incubation on ice for 20 minutes, the cells were











disrupted with 15 strokes in a tight fitting Dounce

homogenizer. The homogenate was then centrifuged at 2,000 x

g for 5 minutes. The supernatant supernatantt A) was removed

and maintained in an ice bath while the nuclear pellet

(pellet A) was resuspended in 9 ml TMN (20 mM Tris-HCl pH

7.5, 5 mM MgCl2, 25 mM NaC1, 3 mM dithiothreitol). The

nuclear suspension (pellet A) was adjusted to 1% Triton X-

100 and 1% NaDOC, vortexed briefly and centrifuged at 500 x

g for 10 minutes. The supernatant supernatantt C) from this

centrifugation step was removed and saved on ice for the

isolation of nucleus-associated rough endoplasmic reticulum

(NRER) polysomes. The supernatant from the centrifugation

step following homogenization supernatantt A) was

centrifuged at 12,000 rpm, 40C for 10 minutes in a Beckman

JA20 rotor. The resulting supernatant supernatantt B) was

removed onto ice for subsequent isolation of light

endoplasmic reticulum (LER)-associated polysomal RNA and

nonmembrane-bound polysomal RNA. The pellet (pellet B) was

resuspended in 9 ml TMN, adjusted to 1% Triton X-100 and 1%

NaDOC and centrifuged at 10,000 rpm, 40C for 10 minutes in a

JA20 rotor. The supernatant supernatantt D) was removed and

stored on ice for the isolation of mitochondria-associated

rough endoplasmic reticulum (MRER) polysomal RNA.

Exponentially growing HeLa cells (2.5 x 108 cells) were

collected by centrifugation for the isolation of total

polysomal RNA. The cell pellet was washed in PBS and



















Figure 2-2. Outline of procedure for the isolation of
subcellular polysomal fractions.
















HeLa Cells


homogenize
centrifuge


in RSB
2,000xg, 40C, 5'


supernatant A


centrifuge 11,400xg
10'



supernatant B pellet B

Triton/NaDOC
centrifuge 7,800x
10'
I 1


Triton/NaDOC
centrifuge 500xg
10'


discard pellet


discard pellet





centrifuge
200,000xg, 2 hrs
2M sucrose


Band at
interface


pellet


supernatant D super





centrifuge
200,000xg, 2 hrs
2M sucrose


pellet


natant C


centrifuge
200,000xg, 2 hrs
2M sucrose


pellet


LER Nonmembrane-bound


pellet A


NRER


MRER











resuspended in 15 ml TMN. The sample was adjusted to 1%

Triton X-100 and 1% NaDOC, disrupted with several strokes in

a tight fitting Dounce homogenizer and centrifuged at 12,000

x g at 40C for 10 minutes in a JA20 rotor. The supernatant

was removed and saved at 0C.

Each of the fractions (NRER, MRER, LER, nonmembrane-

bound, and total polysomal RNA) were layered onto a 3 ml 2 M

sucrose pad (prepared in RSB containing 100 uM spermidine)

and centrifuged at 40C for 2 hours at 48,000 x g in a

Beckman Type 50.2 Ti rotor (200,000 x g) without the brake.

The LER formed a cloudy band at the top of the sucrose pad

and was carefully removed by aspiration. The remaining

fractions were pelleted under these conditions and gently

resuspended in 3 ml TMN buffer. The samples were then

extracted extensively with phenol and chloroform as

described above for the isolation of total cellular RNA. The

RNA was precipitated from solution with 3 volumes of 95%

ethanol at -200C in the presence of 53 mM potassium acetate.

The RNA was digested with DNase I, extracted with phenol and

chloroform, and ethanol precipitated as described above for

the preparation of total cellular RNA. Typical yields were

1.5-2.0 mg MRER and NRER RNA, 2-3 mg nonmembrane-bound RNA,

1 mg LER RNA and 4-6 mg total polysomal RNA per liter of

exponentially growing HeLa cells.











Isolation of Cytoskeleton and Soluble Phase RNA

The cytoskeleton and soluble fractions were prepared as

previously described by Penman and coworkers (Cervera et

al., 1981). Exponentially growing HeLa cells (5 x 108 cells)

were harvested and washed in cold PBS. The cell pellet was

resuspended in 2 ml extraction buffer (10 mM Pipes pH 6.8,

100 mM KC1, 2.5 mM MgCl2, 0.3 M sucrose, 1 mM

phenylmethylsulfonyl fluoride) and Triton X-100 was

immediately added to a final concentration of 0.5%. The

sample was incubated at 0C for 3 minutes and then

centrifuged at 700 rpm in an IEC centrifuge for 3 minutes at

40C. The supernatant (SOL) was removed and stored on ice.

The pellet was resuspended in 8 ml RSB containing 1 mM

phenylmethylsulfonyl fluoride, 1% Tween 40 and 0.5% NaDOC.

The nuclei were stripped of cytoplasmic tags with 15 strokes

in a precision-bore stainless steel homogenizer with a

clearance of 0.002 inches and were pelleted at 2,000 rpm for

3 minutes at 40C in an IEC rotor. The supernatant (CSK) was

removed and saved at 0C for subsequent RNA isolation. Both

the CSK and SOL fractions were extracted in the presence of

1% SDS and 0.3 M NaCl with phenol and chloroform. Nucleic

acids were precipitated from the aqueous phase with 3

volumes of 95% ethanol at -200C in the presence of 53 mM

potassium acetate. The RNA was digested with DNase I, phenol

and chloroform extracted, and ethanol precipitated as

described above for the isolation of total cellular RNA.











Typical yields of RNA were 250-300 Mg from CSK and 100-150

Ag from SOL per 2.5 x 107 HeLa cells (50 ml exponentially

growing suspension culture).



Two-Dimensional Gel Electrophoresis and
Immunoblotting Analysis

The cytoskeleton and soluble fractions were isolated as

described above and each sample was dialyzed in 2 mM Tris-

HC1 pH 7.5 and treated with DNase I (20 gg/ml) and RNase A

(20 Ag/ml) in the presence of 0.5 mM MgC12 at 40C for 1

hour. The samples were dialyzed against 2 mM Tris-HCl pH 7.5

and then lyophilized. Two-dimensional gel electrophoresis

and immunoblotting analyses were performed by Dr. Warren

Schmidt (Department of Pathology, Vanderbilt University,

Nashville, TN). The protein preparations were focused in the

first dimension using equilibrium conditions as described by

O' Farrell (1975) with pH 3 to 10 or 5 to 7 ampholines. The

second dimension was electrophoresed on slab gels containing

a 3% stacking gel and 7.5% resolving gel as described by

Laemmli (1970). The proteins were then either stained with

Coomassie Brilliant Blue or electrotransferred to

nitrocellulose filters as described by Towbin et al. (1979).

Where indicated, immunotransfer analyses (Glass et al.,

1981) were performed with commercially available or other

(Schmidt et al., 1982) antibodies to actin, tubulin and

prekeratin.










Northern Blot Analysis

Aqarose-Formaldehyde RNA Denaturing Gel Electrophoresis

RNA samples were resolved electrophoretically in 1.5%

agarose-6% (w/v) formaldehyde horizontal slab gels and

running buffer (20 mM MOPS (morpholinepropane-sulfonic acid)

pH 7.0, 5 mM sodium acetate, 1 mM EDTA, 3.7% (w/v)

formaldehyde) as described by Rave et al. (1979). Routinely

10 gg RNA per sample were dried in a Savant Speed Vac

concentrator and resuspended in 3.2 ul double distilled

water (ddH0O), 5 ul formamide, 1 ul 10x MOPS (200 mM MOPS pH

7.0, 50 mM sodium acetate, 10 mM EDTA), 0.8 ul 37% (w/v)

formaldehyde and 2 ul sample buffer (0.5 ml formamide, 0.1

ml 10x MOPS, 80 ul 37% (w/v) formaldehyde and 0.32 ml 0.2%

bromophenol blue-90% glycerol). The samples were then heated

at 600C for 10 minutes, quick chilled to 0C and loaded onto

the gels. The 1.5% agarose-6% formaldehyde gels were

prepared by melting 3 g electrophoresis grade agarose in 144

ml ddHO2 plus 20 ml 10x MOPS buffer in a microwave oven

(Whirlpool) at full power for 3-4 minutes. The agarose

solution was cooled to approximately 650C (warm to the

touch) while stirring and then adjusted to 6% formaldehyde

with 36 ml 37% (w/v) formaldehyde. For minigels, 35 ml

agarose-formaldehyde solution was cast in a plexiglass gel

tray (10 cm x 6.3 cm) with an 8 well comb (1 mm x 5 mm per

well). The gel was electrophoresed at 40-45 milliamps until

the bromophenol blue migrated 9 cm (approximately 1.5









50

hours). For full size RNA gels, 200 ml agarose-formaldehyde

solution was cast in a plexiglass gel tray (24.5 cm x 19.8

cm) with 2: 20 well combs (1 mm x 6 mm per well). The gel

was electrophoresed at 100-120 volts until the bromophenol

blue migrated 11 cm (approximately 4 hours). The

formaldehyde gels were stained for 20 minutes in ethidium

bromide solution (1 Ag/ml ethidium bromide, 0.1 M ammonium

acetate) and destined for at least 2 hours in double

distilled water. RNA samples were then visualized by

fluorescence under ultraviolet light using a short

wavelength transilluminator (UVP, inc.; San Gabriel, CA).

The gels were usually photographed at f4.5 for 1/4 second

with Type 57 high speed Polaroid film and a Polaroid model

545 camera mounted on a stand above the transilluminator.



Transfer of RNA from Aqarose-Formaldehyde Gels to
Hybridization Filters

RNA samples were transferred to nitrocellulose (0.45

um) or nylon membrane hybridization filters in 20x SSC as

previously described by Thomas (1975). The gels were soaked

in 20x SSC (3 M NaCl, 0.3 M sodium citrate pH 7.0) for 20

minutes and were not stained with ethidium bromide. The

hybridization filters were uniformly wetted in ddH2O and

then 20x SSC. The gel was placed with the wells facing

downward onto a sheet of 3 MM paper (Whatman) and sponge

which had been soaked in 20x SSC. Strips of used X-ray films









51

were placed on areas of the sponge that were not covered by

the gel so as to direct the flow of 20x SSC only through the

gel. The filter was directly layered onto the gel and any

air bubbles that were trapped between the gel and filter

were carefully removed. The transfer set-up was completed by

placing two sheets of 3 MM paper followed by 2-3 boxes of

facial tissues (Kleenex brand) on top of the hybridization

filter. The buffer chamber was filled with 20x SSC and

transfer was carried out for 36 hours (note: complete

transfer is usually achieved within 14 hours). After

transfer, the filters were rinsed in 2x SSC for 10 minutes

to remove particles of agarose, air dried and baked in vacuo

at 800C for 2 hours. The filters were stored in sealed

hybridization bags at 40C.



Hybridization of Filters-Immobilized RNA (Northern Blot)

Filters (nylon or nitrocellulose) were prehybridized

for 3-4 hours in 0.5 ml hybridization buffer per cm2 filter

area at 430C for the detection of chorionic gonadotropin

alpha, HLA-B7 and c-fos mRNA and at 480C for histone mRNA.

Hybridization buffer consisted of 50% formamide, 5x SSC, 10x

Denhardt's (100x Denhardt's: 2% (w/v) ficoll 400, 2% (w/v)

polyvinylpyrrolidone), 1% SDS, 50 mM sodium phosphate pH

7.0, 20 gg/ml BSA, and 250 gg/ml E. coli DNA. The filters

were hybridized at the appropriate temperature for 36 hours

in 0.1 ml hybridization buffer per cm2 filter containing











1xl10 cpm/ml of each thermally denatured, 3P-radiolabeled

DNA probe. After hybridization, the filters were washed with

5x SSC, ix Denhardt's (0.5-1 ml/cm2 filter) for 10-15

minutes at room temperature to remove unbound probe. The

filters were then washed, in order, with the following

solutions (0.5 ml/cm2 filter area): 1) 5x SSC, Ix

Denhardt's; 2) 2x SSC, 0.1% SDS; 3) Ix SSC, 0.1% SDS; and 4)

0.1x SSC, 0.1% SDS at 600-650C for 30 minutes per wash.

After washing the filters were briefly air dried and then

exposed to preflashed Kodak XAR5 x-ray film at -70C with a

Cronex Lightning Plus intensifying screen (Dupont) for

varying lengths of time. Hybridization was quantitated by

scanning laser densitometric analysis of multiple exposures

of autoradiographs that were within the linear range of the

film.



S1 Nuclease Protection Analysis

Sl nuclease protection analysis was performed

essentially as described by Berk and Sharp (1977) and was

used in the work presented here to detect simultaneously

both endogenous H3 histone mRNA and signal peptide-histone

fusion mRNA in the same RNA sample. The Sl probe was

prepared by digesting 30 pg pSPH3E1, pSPH3Elalpha, or

pSPH3E1ATG- DNA with 30 units of SmaI at 300C for 3 hours.

The 450 bp SmaI fragment was isolated from a 0.8% agarose

minigel by the freeze-squeeze method. The solution









53

containing the SmaI fragment was butanol extracted to reduce

the volume and to remove the ethidium bromide. The fragment

was then extracted with phenol and chloroform and

precipitated in 2.5 volumes 95% ethanol at -200C overnight.

The DNA was collected by centrifugation at 12,000 x g at 40C

for 15"minutes and resuspended in 5 ul 10 mM Tris-HCl pH

8.0, 5 ul 10x CIP buffer (0.5 M Tris-HCl pH 8.0, 10 mM

MgCl2, 1 mM ZnCl2, 10 mM spermidine) and 38 ul ddH20. The

blunt-ended DNA was dephosphorylated with 0.12 units calf

intestinal alkaline phosphatase (CIP) at 370C for 15 minutes

and then at 550C for 15 minutes. A second aliquot of CIP was

added and the incubations at 370 and 550C were repeated. The

reaction was terminated with 40 ul ddH20, 10 ul 10x STE (100

mM Tris-HC1 pH 8.0, 1 M NaCI, 10 mM EDTA) and 5 ul 10% SDS

and heated at 680C for 15 minutes. The sample was extracted

with phenol and chloroform, adjusted to 0.25 M sodium

acetate and precipitated with 2.5 volumes of 95% ethanol at

-20C overnight. The dephosphorylated SmaI fragment was

pelleted at 12,000 x g, 40C for 15 minutes, vacuum dried and

resuspended in 12 ul Ix kinase buffer (66 mM Tris-HCl pH

9.5, 10 mM MgC12, 10 mM beta-mercaptoethanol, 2 mM

spermidine). The Sl fragment was phosphorylated with 50 uCi

[gamma-32 P]ATP and 1 ul polynucleotide kinase [30 units/ul]

at 370C for 30 minutes. The reaction was terminated with 0.5

ul 100 mM EDTA at 700C for 10 minutes. The DNA solution was

diluted with 1 ml elutip low salt buffer (20 mM Tris-HCl pH

7.4, 0.2 M NaC1, 1 mM EDTA) and passed over a Schleicher and









54

Schuell Elutip column. The column was washed with 5 ml low

salt buffer to remove unincorporated radiolabeled

ribonucleotide and the DNA fragment was eluted in 400 ul

high salt buffer (Tris-HC1 pH 7.4, 1 M NaCl, 1 mM EDTA). The

radiolabeled Sl probe was usually used immediately and any

excess was stored in high salt buffer at 4C.

Routinely, 10-25 gg RNA (30 ul) and 10 ul S1 probe were

co-precipitated with 100 ul 95% ethanol at -200C overnight.

The sample was centrifuged at 12,000 x g, 40C for 15-30

minutes. The pellet was vacuum dried and resuspended in 10

ul 5x hybridization buffer (0.2 M Pipes pH 6.4, 2 M NaC1, 5

mM EDTA) and 40 ul recrystallized formamide. The sample was

heated at 900C for 10 minutes and then incubated at 550C for

3 hours. Single-stranded nucleic acids were digested with Sl

nuclease by adding 400 ul ice cold Sl nuclease buffer (0.03

M sodium acetate pH 4.6, 0.25 M NaC1, 1 mM ZnSO4) and 900

units S1 nuclease at 370C for 30 minutes. The Sl digestion

reaction was terminated by organic extraction with 0.5

volume phenol and 0.5 volume chloroform and then 0.5 volume

chloroform alone. Nucleic acids were precipitated from the

aqueous phase with 1 ml 95% ethanol at -200C overnight.

The S1 nuclease digested samples were collected at

12,000 x g, 40C for 30 minutes. The pellets were vacuum

dried and resuspended in 5 ul loading buffer (80% formamide,

lx TBE [10x TBE: 500 mM Tris-HCl pH 8.3, 500 mM boric acid,

10 mM EDTA], 0.1% bromophenol blue, 0.1% xylene cyanol). The









55

samples were denatured by heating at 1000C for 2 minutes and

then quickly chilled at 0C to prevent reannealing prior to

loading onto the gel. Six percent polyacrylamide-8.3 M urea

(cross linked at 20:1 acrylamide:bis-acrylamide) gels with

dimensions of either 16 cm x 18 cm x 1.5 mm (1.5 mm x 6 mm

15 well comb or 1.5 mm x 4 mm 20 well comb) or 32.5 cm x 44

cm (0.5 mm x 8 mm 25 well comb) were prepared and pre-

electrophoresed for 30 minutes. The samples were loaded with

drawn-out capillary tubes or a Rainin Pipetteman (0-20 ul

capacity) using flat pipette tips. The large gels were

electrophoresed at 50 watts constant power (-1200 V, -45 A)

for 3-4 hours and the small gels were electrophoresed at 300

V (25-35 A) for 2-3 hours. The gels were placed on 2 sheets

of 3 MM Whatman paper, dried under vacuum at 800C for 1-2

hours and exposed to pre-flashed XAR5 x-ray film with a

Cronex Lightning Plus intensifying screen at -700C for

varying lengths of time. Sl analysis was quantitated by

scanning laser densitometry of multiple autoradiographic

exposures that were within the linear range of the film.



In Vitro Nuclear Run-On Transcription Analysis

The in vitro nuclear run-on transcription assays were

performed by Anna Ramsey-Ewing as described by Baumbach et

al. (1987). Cells were harvested by centrifugation and the

cell pellet washed twice in cold isotonic buffer (125 mM

KC1, 30 mM Tris-HCl pH 7.9, 5 mM MgCl2, 10 mM beta-









56
mercaptoethanol). Cells were disrupted by homogenization (6-

12 strokes) with a Dounce homogenizer, Wheaton type A

pestle. After >90% of the cells had been lysed, nuclei were

pelleted by centrifugation at 2000 rpm for 10 minutes in an

IEC centrifuge at 40C and resuspended in nuclei storage

buffer containing 40% glycerol (50 mM Tris-HCl pH 8.3, 5 mM

MgCl2, 0.1 mM EDTA). Nuclei were aliquoted and either snap

frozen in liquid nitrogen or used fresh in the in vitro

transcription reactions. Reactions typically contained 107

nuclei, 100 uCi alpha-32P-UTP (3000 Ci/mmole), 1 mM ATP,

0.25 mM GTP and CTP in a final volume of 130 ul and were

incubated for 30 minutes with intermittent shaking at 300C.

Radiolabeled RNAs were isolated by treatment of nuclei with

DNase I (100 Ag/ml) in the presence of 0.6 M NaC1, 50 mM

Tris-HCl pH 7.5, 20 mM MgCl2 for 15 minutes at room

temperature. The mixture was then incubated with proteinase

K (200 Ag/ml) for 30-60 minutes at 370C in the presence of

150 mM NaC1, 12.5 mM EDTA, 100 mM Tris-HCl pH 7.5 and 20 mM

MgC12. Sodium acetate (pH 5.5) was added to 0.2 M and

nucleic acids extracted several times by the hot phenol

method (Clayton and Darnell, Jr., 1983; Sherrer and Darnell,

1962). To the aqueous solution of 32P-labeled RNAs, 150 Ag

of yeast RNA and 2.5 volumes of 95% ethanol were added.

Precipitation was overnight at -200C. Radiolabeled

transcripts were resuspended in 10 mM Tris-HCl pH 8.0, 1 mM

EDTA and an aliquot of each sample was precipitated with 150









57

Mg yeast RNA and cold 10% TCA. TCA-precipitable counts were

determined by liquid scintillation spectrometry.

DNA excess hybridization has also been described by

Baumbach et al. (1987). In a preliminary experiment we

determined that 2 pg of H4 histone insert in pFO002 was at

least a two-fold DNA excess when hybridized to in vitro

nuclear run-on transcripts. Southern blots containing

restriction endonuclease digested DNA fragments or slot

blots containing linearized plasmid DNAs were prepared.

Southern blots on nitrocellulose or slot blots on Nylon were

prehybridized in 1 M NaCl, 20 mM Tris-HCl pH 7.4, 2 mM EDTA,

0.1% SDS, 5x Denhardt's, 250 Ag/ml E. coli DNA and 12.5 mM

sodium pyrophosphate at 650C for at least 6 hours.

Hybridizations were conducted at 650C for 72 hours in 1 M

NaCl, 20 mM Tris-HCl pH 7.4, 2 mM EDTA, 0.1% SDS, 2.5x

Denhardt's, 250 Ag/ml E. coli DNA with 32P-labeled

transcripts at 5x105-1X106 TCA-precipitable counts per ml of

hybridization solution. Blots were washed at 650C for 15

minutes in fresh prehybridization solution [0.13-0.14

ml/cm2], 1 hour in 2x SSC/0.1% SDS [0.5 ml/cm2], overnight

in 2x SSC/0.1% SDS [0.5 ml/cm2] and 1 hour in 0.2x SSC/0.1%

SDS [0.5 ml/cm2]. Air dried filters were autoradiographed

with XAR-5 or Cronex film and Cronex lightning Plus screens

at -70C for varying periods of time and developed with an

X-O-MAT X-ray film processor.











Radiolabeling DNA for Northern Blot Analysis

DNA used for Northern blot analysis was uniformally

labeled by nick translation as described by Maniatis et al.

(1982b) or by the random priming method of Roberts and

Wilson (1985). Radiolabeled probes prepared in this manner

are also suitable for other purposes such as Southern blot

analysis.



Random Priming Method

This protocol is very efficient for labeling small to

intermediate-sized DNA fragments (100-2000 bp in length) and

is the preferred method for work with purified DNA inserts.

Supercoiled plasmid DNA is labeled much less efficiently by

the random priming method and should be fragmented by

sonication prior to use. Routinely, specific activities of

1x109 cpm/gg DNA are achieved using as little as 20-100 ng

DNA. Material for oligolabeling DNA fragments was purchased

from Bethesda Research Laboratories as a kit.

The isotope (50 uCi [a-32P] dCTP) was vacuum dried and

resuspended in 10 ul of 2.5x reaction buffer (500 mM Hepes

pH 6.6, 12.5 mM MgCl2, 25 mM beta-mercaptoethanol, 125 mM

Tris-HCl pH 8, 1 mg/ml BSA, 7.5 mg/ml synthetic DNA

oligodeoxynucleotide primers, 50 uM each dGTP, dATP, dTTP).

The DNA fragment (100 ng) in 14 ul ddH2O was heated at 1000C

for 2 minutes and immediately chilled on ice. The denatured

DNA was added to the 1.5 ml microcentrifuge tube containing











the isotope and buffer and the priming reaction was

initiated with 5 units Klenow enzyme [5 U/ul]. The sample

was incubated at room temperature for at least 2 hours

(usually overnight). The reaction was terminated with 1 ul

250 mM EDTA and 175 ul TE buffer (10 mM Tris-HCl pH 8, 1 mM

EDTA) and extracted once with 200 ul chloroform/isoamyl

alcohol. Unincorporated radiolabeled deoxynucleotide was

removed from the DNA by chromatography over a (10 mm x 100

mm) Biogel A-15m (Bio Rad) column in TE buffer. Fourteen

fractions were collected; 800 ul in the first fraction and

200 ul in each subsequent fraction. Fractions containing

radiolabeled DNA were determined by Cerenkov counting in the

tritium channel with a Beckman LS-230 scintillation counter.

The peak fractions were pooled and accurately quantitated by

counting 2 ul DNA in 5 ml Triton-toluene based scintillation

cocktail (2000 ml toluene, 1066 ml Triton X-100, 134 ml

liquifluor) in the 32P channel. The samples were immediately

used for hybridization studies and excess probe was stored

at 4C.



Nick-Translation Method

Nick-translation was the common procedure for

uniformally labeling DNA fragments or supercoiled DNA before

the development of the random primer method. A disadvantage

to using the nick-translation procedure is that the system

must be optimized for efficiency for each lot of E. coli DNA









60
polymerase I enzyme. In addition, the specific activity (1-2

x 108 cpm/Ag DNA) achieved with the nick-translation

procedure is much lower than the random primer method (1 x

109 cpm/Ag DNA).

DNase I [1 mg/ml in 10 mM HCl] was activated by

incubating 1 ul of the enzyme in 9 ul DNase buffer (10 mM

Tris-HCl pH 7.5, 5 mM MgCl2, 1 mg/ml BSA) at 0C for 2

hours. The isotope (40-80 uCi [a-32p] dCTP) was dried in a

Savant Speed Vac centrifuge. The following components were

added, in order, to the isotope : 1) 12.25 ul ddHO2; 2) 2.5

ul 10x nick-translation buffer (500 mM Tris-HC1 pH 7.8, 100

mM beta-mercaptoethanol, 50 mM MgCl2); 3) 5 ul 10x dNTPs

(100 uM of each dATP, dGTP, dTTP); 4) 1.25 ul BSA [1 mg/ml];

5) 2.5 ul DNA [100 ng/ul]; and 6) 0.5 ul DNA polymerase I [5

U/ul]. The activated DNase I was diluted with 990 ul ice

cold ddH2O (1:1,000 dilution; 1 ng/ul final concentration)

and 1 ul of diluted enzyme was added to the nick-translation

mix. The sample was incubated at 140C for 45 minutes and the

reaction was then terminated with 20 ul 250 mM EDTA and 155

ul TE buffer. The sample was extracted once with 200 ul

chloroform and unincorporated radiolabeled

deoxyribonucleotides were removed by chromatography over a

Biogel A-15m column. The DNA was quantitated and immediately

used for hybridization.











Site-Directed Mutagenesis

In the past, researchers had to rely on nature to

produce the proper mutations to study how the normal

biological processes function. Zoller and Smith (1983) have

designed a powerful method for selectively introducing

single nucleotide mutations into specific sites of any

cloned, sequenced region of DNA (site-directed mutagenesis).

In addition, this procedure can be used to selectively

create deletion or insertion mutations in DNA fragments of

known sequence. The general approach of site-directed

mutagenesis is to incorporate an oligonucleotide containing

the appropriate mutation into the complementary strand of

recombinant M13 DNA. The heteroduplex is transfected into

bacteria and the resulting mutated bacteriophage are

isolated by various selection processes.

As discussed in Chapter 4, the signal peptide-histone

fusion mRNA is associated with membrane-bound polysomes and

is stable during inhibition of DNA synthesis. It was unclear

whether the uncoupling of the histone fusion mRNA stability

from DNA replication was due to the altered subcellular

location or to a change in mRNA structure, due to the

nucleotide sequence coding for the signal peptide. To

address these possibilities, we have introduced point

mutations into the signal peptide coding region with the

intention to inactivate signal peptide function. In doing

so, mRNA structure is perturbed as little as possible and











the mutated histone fusion mRNA, like endogenous histone

mRNA, associates with nonmembrane-bound polysomes.


Subcloning of the Signal Peptide-Histone Fusion Gene
into M13 DNA

The signal peptide-histone fusion gene (SPH3) was

subcloned into M13mpl8 RF DNA. SPH3 DNA was digested with

EcoRI and HindIII and the 1100 bp fragment was isolated from

a 0.8% agarose gel. The M13mpl8 DNA was digested with EcoRI

and HindIII and the 7196 bp fragment was isolated from a

0.8% agarose gel. The 1100 bp SPH3 DNA fragment (-50 ng) and

7196 bp M13 DNA fragment (-50 ng) were ligated in 10 ul Ix

ligation buffer (50 mM Tris pH 7.6, 10 mM MgCl2, 20 mM

dithiothreitol, 1 mM ATP and 50 .g/ml BSA) with 0.004 U T4

DNA ligase (NEB) at 220C overnight. The ligation mix was

transfected into competent JM101 bacteria, white plaques

were isolated, and M13/SPH3 recombinants were screened by

restriction endonuclease digestion analysis. Single-stranded

M13/SPH3 DNA was isolated from bacteriophage particles as

described above.



Synthesis and Processing of Oligonucleotides

The signal peptide encoded by the SPH3 histone fusion

gene was mutated by substituting leucine at amino acid

position 10 and proline at position 12 with histidine

residues. It was predicted that the introduction of two

positively charged amino acids into the hydrophobic region











of the signal peptide would disrupt its recognition by the

signal peptide recognition particle and would therefore,

prevent the translocation of the mutated histone fusion mRNA

to membrane-bound polysomes.

An oligonucleotide (SPa) containing two point mutations

was synthesized by Tom Doyle in the Department of

Microbiology and Immunology, University of Florida,

Gainesville, Florida. The oligonucleotide was composed of

the following sequence: 5' CAAAAAAGTGAATATGGGCGA 3', with

the mutated nucleotides underlined. The oligonucleotide was

received in -2.5 ml 37% ammonium hydroxide and was heated at

550C for 6 hours to remove the protecting amine group. The

sample was dried under a gentle stream of air in the fume

hood and resuspended in 1 ml ddH20. The sample was extracted

two times with 600 ul n-butanol and the oligonucleotide was

precipitated from the aqueous phase by adding 200 ul 0.5 M

ammonium acetate and 2 ml 95% ethanol and chilling at -700C

for 30 minutes. The sample was centrifuged at 12,000 x g at

40C for 30 minutes, washed in 80% ethanol and resuspended in

200 ul ddH20. The yield of oligonucleotide was estimated

spectrophotometrically and was found to be 2.64 mg

(A260/A80 of 2.0).

A second oligonucleotide (ATG-) was designed to destroy

the signal peptide ATG translation start codon of the SPH3

fusion gene and thereby completely block the synthesis of

any signal peptide. The SPH3ATG- mRNA retains the histone











translation start codon in its proper context, although

further downstream from the 5' terminus than endogenous H3

histone mRNA, and should therefore be translated on

nonmembrane-bound polysomes. In addition, SPH3ATG- mRNA

differs from SPH3 wild type signal peptide-histone fusion

mRNA at only one nucleotide. The sequence of the ATG-

oligonucleotide is as follows: 5' AATACTCAAACTCTTCC 3', and

was prepared and processed as described for the SPa

oligonucleotide. The yield of ATG- oligonucleotide was only

445 Ag (A260/A280 of 1.41).



Optimization of Conditions for Site-Directed Mutagenesis

It was necessary to optimize the conditions for the

hybridization of the oligonucleotide to M13 template so that

the synthesis of the second strand was primed by the

oligonucleotide from the correct site. Proper priming by the

oligonucleotide was tested by primer extension analysis.

After limited primer extension, the samples were digested

with a restriction endonuclease that recognized a site

downstream from the expected priming site. The number and

intensity of the fragments produced by the digestion

indicated the number of priming sites and the location of

the preferred site.

The oligonucleotide (20 pmol) was phosphorylated with 5

units T4 polynucleotide kinase in 100 mM Tris-HCl pH 8.0, 10

mM MgC12, 5 mM dithiothreitol, 20 uCi [gamma-32P] ATP at 370C











for 45 minutes. The reaction was terminated by heating at

650C for 10 minutes, and unincorporated radiolabeled

nucleotides were removed by chromatography through a

Sephadex G-25 (Pharmacia) column (10 mm x 100 mm) in 50 mM

ammonium bicarbonate (pH 7.8).

The phosphorylated oligonucleotide (10-30 molar excess

over template DNA) was added to 0.5 pmol of M13 recombinant

DNA and the volume was adjusted to 10 ul with ddHO2 and 1 ul

solution A (200 mM Tris-HCl pH 7.5, 100 mM MgCl2, 500 mM

NaCl, 10 mM dithiothreitol). The solution was heated at 550C

for 5 minutes and then cooled for 5 minutes at room

temperature. The primer extension reaction also contained 1

ul of 2.5 mM dCTP, 1 ul 2.5 mM dATP, 1 ul 2.5 mM dTTP, 1 ul

2.5 mM dGTP, 0.5 ul solution A and 1 unit Klenow enzyme. The

reaction proceeded at room temperature for 5 minutes and was

terminated by heating at 650C for 10 minutes. The sample was

diluted with 2 ul 10x PstI buffer (100 mM Tris-HCl pH 7.5, 1

M NaCI, 100 mM MgCl2, 1 mg/ml BSA) and 7 ul ddHO2 and then

digested with 1 ul PstI restriction endonuclease [5 U/ul] at

370C for 1 hour. The sample was heated at 1000C for 2

minutes, quickly cooled to 0C and diluted with 20 ul S1

loading buffer. The PstI digested samples (10 ul) were

resolved electrophoretically in 5% acrylamide (20:1;

acrylamide:bis), 7 M urea at 300 volts (-35 mA) for 1.5

hours. The gels were dried under vacuum at 800C for 1 hour

and analyzed by autoradiography. Primer extension analysis









66

revealed a single major band at -80 bp for each molar ratio

of oligonucleotide to template DNA studied. This was the

expected size fragment and suggested that under these

conditions, the oligonucleotide was priming the synthesis of

the second strand from the proper site. If several

predominant bands had been detected (i.e. multiple priming

sites), more stringent annealing conditions would have been

required.



Synthesis of the Mutated Strand

The oligonucleotide (1.3 Mg) was phosphorylated in 100

mM Tris-HCl pH 8, 10 mM MgCl2, 5 mM dithiothreitol, 10 uM

ATP with 15 units T4 polynucleotide kinase at 370C for 45

minutes. The reaction was terminated by heating at 650C for

10 minutes.

The phosphorylated oligonucleotide (13.3 pmole) was

annealed to M13/SPH3 recombinant DNA in Ix solution A by

incubating the sample at 550C for 5 minutes and then at room

temperature for 5 minutes. After hybridization, the DNA was

diluted with 10 ul solution C (1 ul solution B [0.2 M Tris-

HC1 pH 7.5, 0.1 M MgCl2, 0.1 M dithiothreitol], 1 ul 10 mM

dCTP, 1 ul 10 mM dTTP, 1 ul 10 mM dGTP, 0.5 ul 0.1 mM dATP,

1 ul 10 mM ATP, 1.5 ul [gamma-32P] dATP, 1.5 ul T4 DNA ligase

[1 U/ul], 2 ul ddH20) and the synthesis of the second strand

was initiated with 0.5 ul Klenow enzyme [100 U/16.7 ul] at

room temperature. After 5 minutes, 1 ul 10 mM dATP was added











and the incubation was continued at room temperature

overnight.



Enrichment for Covalently Closed Double-Stranded DNA

The in vitro synthesis of the entire second strand is

very inefficient due primarily to the large size of the M13

genome. The percent of covalently closed, double-stranded

DNA molecules is therefore, very small in comparison to the

partially double-stranded, incomplete DNA molecules. These

incomplete DNA molecules, when transfected into bacteria,

give rise to wild type plaque formation which complicates

the selection process for the mutated clones. Enrichment for

covalently closed, double-stranded DNA molecules by alkaline

sucrose gradient centrifugation greatly facilitates the

selection process.

The DNA was precipitated from the overnight extension

and ligation reaction by adding 30 ul ddH20, 50 ul 1.6 M

NaCl-13% PEG at 0C for 15 minutes. The DNA was pelleted at

12,000 x g at room temperature for 15 minutes. The pellet

was gently washed in 100 ul cold 0.8 M NaCl-6.5% PEG,

centrifuged at 12,000 x g for 30 seconds and resuspended in

180 ul TE buffer. The sample was denatured with 20 ul 2 N

sodium hydroxide and layered onto a 5%-20% alkaline sucrose

step gradient. The gradient was prepared by adding to a 0.5

in x 2 in centrifuge tube nitrocellulosee or polyallomer; do

not use Beckman Ultra-clear tubes which are incompatible











with the alkali conditions), in order, the following

solutions each containing 0.2 N NaOH, 1 M NaCl, 2 mM EDTA:

1) 1 ml 20% sucrose; 2) 1 ml 17.5% sucrose; 3) 1 ml 15%

sucrose; 4) 1 ml 10% sucrose; 5) 5% sucrose. The tube was

incubated at 40C for 2 hours. The samples were centrifuged

through the alkaline sucrose gradients in a Beckman SW 50.1

rotor at 37,000 rpm for 2 hours at 40C without the brake.

The samples were recovered from the gradients by puncturing

the bottom of the tube with a 21 gauge needle and collecting

7 drops per fraction in 1.5 ml microcentrifuge tubes

(usually -60 fractions). Fractions containing covalently

closed, double-stranded DNA were determined by Cherenkov

scintillation counting in the tritium channel, pooled and

neutralized with 1M Tris-citrate pH 5.



Isolation of Mutated M13 Recombinant Bacteriophage

The DNA was transfected into E. coli JM101 competent

cells, plated onto YT agar and incubated overnight at 370C.

Individual plaques were picked and bacteriophage were

subsequently prepared. Bacteriophage suspensions (2 ul) were

spotted onto a 0.45 micron nitrocellulose filter and

prehybridized in 6x SSC, 10x Denhardt's, 0.2% SDS (10 ml/100

cm2) at 670C for 1 hour. The filter was rinsed in 50 ml 6x

SSC for 1 minute at 210C and hybridized in 10 ml 6x SSC, 10x

Denhardt's containing 1x106 cpm radiolabeled oligonucleotide

at 210C for 1 hour. The filter was washed three times in 50

ml 6x SSC at 210C for a total of 10 minutes. The filter was











briefly dried under a heat lamp and placed against pre-

flashed XAR5 x-ray film without a screen for 1 hour at room

temperature. Under such low stringency each phage spot was

positive, including the negative control M13/SPH3 wild type

bacteriophage. The filter was re-washed at 350C in 6x SSC

for 5 minutes, air dried and exposed to pre-flashed XAR5

film for 1 hour. At the higher wash temperature many of the

phage spots were less intense whereas several remained

unchanged. When the filter was washed at 480C in 6x SSC for

5 minutes the negative control and nearly all the phage

spots completely disappeared. The bacteriophage (-5%) that

remained positive after the 480C wash were sequenced to

confirm the presence of the mutations.



DNA Sequencing

The site directed mutations in M13/SPH3 recombinant

DNAs were confirmed by Sanger's dideoxynucleotide chain

terminator DNA sequencing procedure using the "SequenaseTM

DNA Sequencing Kit" from United States Biochemical

Corporation, Cleveland, Ohio. The Sequenase enzyme is

derived from bacteriophage T7 DNA polymerase and has been

modified to efficiently use dideoxyribonucleotides, alpha-

thio deoxyribonucleotides and other nucleotide analogs that

are commonly used for sequencing. In addition, the Sequenase

enzyme expresses high processivity and low 3'-5' exonuclease

activity.










Annealing Template and Primer

The primer for synthesizing the complementary strand of

wild type and mutated M13/SPH3 recombinant DNA is a 20 bp

oligonucleotide which hybridizes to a site adjacent to the

polylinker. The annealing reaction was prepared by adding 1

ul priier (0.5 pmol), 7 ul single-stranded M13/SPH3

recombinant DNA (0.5-1.0 pmol) and 2 ul 5X sequencing buffer

(200 mM Tris-HCl pH 7.5, 100 mM MgCl2, 250 mM NaC1). The

sample was heated at 650C for 2 minutes and slowly cooled to

room temperature.



Labeling Reaction

The complementary strand was synthesized by incubating

the annealed sample with 1 ul 100 mM dithiothreitol, 2 ul

labeling nucleotide mix (1.5 uM dGTP, 1.5 uM dCTP, 1.5 uM

dTTP), 0.5 ul [a- S] dATP [10 uCi/ul] and 2 ul Sequenase

[1.5 U/ul] at room temperature for 5 minutes. Elongation was

terminated by mixing 3.5 ul aliquots of the labeling

reaction with 2.5 ul of one of the dideoxy nucleotide

solution (ddATP, ddCTP, ddGTP, ddTTP). The dideoxy

nucleotide solutions were 80 uM dATP, 80 uM dCTP, 80 uM

dGTP, 80 uM dTTP, 50 mM NaCl and 8 uM of the appropriate

dideoxyribonucleotide. The sequencing samples were diluted

with 4 ul stop solution (95% formamide, 20 mM EDTA, 0.05%

bromophenol blue and 0.05% xylene cyanol), heated at 750C

for 2 minutes and loaded onto a 6% acrylamide-8.3 M urea gel









71

(32.5 cm x 44 cm) with a 60 well shark toothed comb. The gel

was run at 75 W constant, ~2500 V, -40 mA for 3-5 hours

(Xylene Cyanol dye was 15 cm from the bottom of gel). The

gel was soaked in 10% acetic acid-12% methanol in the fume

hood for 1 hour, placed onto 2 sheets of 3 MM Whatman paper,

and dried under vacuum at 800C for 45 minutes. The gel was

exposed to preflashed XAR5 x-ray film at room temperature

for 12-18 hours.



Transfection of DNA into Cells

Introduction of Plasmid DNA into Bacterial Cells

An E. coli overnight cell culture was diluted 1:100

into 100 ml LB medium and incubated at 370C until the

optical density was approximately 0.37 at 590 nm. The cells

were chilled at 0C for 10 minutes and then centrifuged at

5,000 rpm, 40C for 5 minutes in a Beckman JA20 rotor. The

pellet was resuspended in 20 ml of cold CaC12 buffer (60 mM

CaCI2, 10 mM Pipes pH 7.0, 15% glycerol) and incubated at

0C for 30 minutes. The cells were collected by

centrifugation at 2,500 rpm for 5 minutes at 40C in a JA20

rotor, resuspended in 2.6 ml CaC12 buffer and divided into

100-200 ul aliquots. The competent cells were either used

immediately for transformation or stored in microcentrifuge

tubes at -80C.

For transformation, 100 ul competent E. coli cells were

mixed with 10 ul plasmid DNA (-10 ng) and incubated at OC

for 10 minutes. The cells were then heat shocked at 370C for











5 minutes, diluted with 0.9 ml prewarmed LB medium and

incubated at 370C for 1 hour. Different volumes of the

transfected bacterial cell culture (usually 10-200 ul) were

plated onto the appropriate medium dictated by the genotype

of the host and the introduced plasmid DNA.



Introduction of DNA into HeLa Cells

HeLa monolayer cell cultures were transfected with

plasmid DNA, according to Gorman et al. (1982), in a calcium

phosphate/DNA complex prepared as described by Graham and

van der Eb (1973). HeLa S3 cells growing exponentially in

suspension culture were plated into 20 ml completed EMEM

(EMEM supplemented with 5% fetal calf serum, 5% horse serum,

100 U penicillin, 100 gg/ml streptomycin and 1 mM glutamine)

at 3x106 cells per 10 cm tissue culture dish (Corning or

Falcon) and incubated overnight at 370C in 5% CO2. The

monolayer cell culture was refed with 10 ml completed EMEM

and incubated at 370C in 5% CO2 for 4 hours. The DNA/calcium

phosphate precipitate was prepared by adding 20 gg DNA in

500 ul 250 mM CaCl2 dropwise to 500 ul 2x Hebs buffer (50 mM

Hepes, 280 mM NaC1, 1.5 mM Na2HPO4, pH 7.120.05) while

vortexing the solution. The DNA/CaC12 precipitate was added

dropwise to the monolayer cell culture and incubation was

continued at 37C in 5% CO2 for 4 hours. The HeLa cells were

glycerol shocked by replacing the medium with 2 ml completed

EMEM containing 15% glycerol for 1 minute. The transfected










cells were then washed with 10 ml EMEM, refed with 20 ml

completed EMEM and incubated at 370C, 5% CO2.



Establishment of Polyclonal and Monoclonal Stably
Transformed Cell Lines

Polyclonal and monoclonal cell lines were isolated by

growing HeLa cell cultures that were co-transfected with

pSV2neo and the plasmid DNA of interest in medium containing

the antibiotic G418. The plasmid pSV2-neo, constructed by

Southern and Berg (1982), carries the bacterial

aminoglycoside phosphotransferase 3' (II) gene under the

control of the SV40 early promoter and confers resistance to

the aminoglycoside antibiotic G418. The rationale behind the

co-transfection method is that HeLa cells that are competent

to take up pSV2-neo DNA are equally competent to take up the

plasmid DNA of interest, and therefore, a certain percentage

of the G418 resistant cells contain both species of DNA.

Ideally, the selectable marker should be carried by the

vector that contains the gene which is to be studied.

HeLa cells in monolayer culture were co-transfected

with the plasmid DNA to be studied and pSV2-neo DNA (molar

ratio of 20:1, respectively) by the calcium phosphate

precipitation method as described above. After 36-48 hours

post-glycerol shock, the transfected cells were resuspended

in 5 ml Puck saline A [0.04% (w/v) KC1, 0.8% (w/v) NaC1,

0.035% (w/v) NaHCO3, 0.1% (w/v) glucose] containing 0.02%









74

(w/v) EDTA and 0.5 ml of the cell suspension was seeded into

20 ml completed EMEM containing 500 gg/ml biologically

active Geneticin (G418, Gibco) in 10 cm tissue culture

dishes. The cells were incubated at 370C, 5% CO, and refed

every 3-4 days with completed EMEM containing G418

antibiotic. Within 2-3 weeks, the majority of the cells had

died and individual colonies of G418 resistant HeLa cells

had formed. Polyclonal cell lines were established by

combining the G418 resistant colonies from a 10 cm dish as a

single, heterogeneous cell population (usually 20-40

colonies). For the establishment of monoclonal cell lines,

individual colonies were picked from the dish with sterile

pasteur pipettes and cultured as separate cell lines.















CHAPTER 3

SUBCELLULAR LOCATION OF HISTONE mRNAs ON CYTOSKELETON
ASSOCIATED NONMEMBRANE-BOUND POLYSOMES IN HELA S3 CELLS


INTRODUCTION

Histone biosynthesis occurs primarily during the S

phase of the cell cycle and is temporally as well as

functionally coupled with DNA replication (Allfrey et al.,

1963; Prescott, 1966; Spaulding et al., 1966; Robbins and

Borun, 1967; Stein and Borun, 1972). Early studies using in

vitro translation analysis of polysomal RNA isolated from S

phase HeLa cells provided an initial indication that histone

protein synthesis occurs on non-membrane-bound, light

polysomes (Jacobs-Lorena et al., 1972; Liautard and

Jeanteur, 1979; Gallwitz and Breindl, 1972; Borun et al.,

1967). In these studies it was demonstrated that

translatable, polysome-associated histone mRNAs have a

sedimentation coefficient of 8-10S and fluctuate in a cell

cycle-dependent manner. Histone mRNA was not detected by in

vitro translation studies on polyribosomes isolated from

HeLa cells in the G1 phase of the cell cycle (Borun et al.,

1967; Borun et al., 1975; Pederson and Robbins, 1970).

Consistent with this observation, the inhibition of DNA

replication by hydroxyurea or cytosine arabinoside treatment











resulted in the rapid disappearance of histone mRNAs from

these polysomes (Borun et al., 1967; Breindl and Gallwitz,

1974; Gallwitz, 1975).

The light polysomes on which translatable histone mRNAs

reside are presumably nonmembrane-bound ("free") polysomes

as predicted by the signal hypothesis (Blobel and

Dobberstein, 1975). Several lines of evidence suggest that

nonmembrane-bound polysomes are not freely diffusible within

the cytoplasm but are attached to the cytoskeleton. Penman

and co-workers (Lenk et al., 1977) have demonstrated that

nearly 100% of total cytoplasmic polysomes co-fractionate

with the cytoskeletal framework. In addition, vesicular

stomatitis virus encoded nonmembrane-bound and membrane-

bound polysomal mRNAs are associated with the cytoskeleton

in virally infected HeLa cells (Cervera et al., 1983).

Taken together, these observations provide a basis for

the possibility that histone mRNA-containing polysomes

and/or their subcellular location may play a role in the

regulation of histone gene expression. Such reasoning is

consistent with the appearance of translatable histone mRNAs

on polysomes within minutes following transcription (Borun

et al., 1967), the equally rapid transfer of newly

synthesized histones into the nucleus where they complex

with DNA (Robbins and Borun, 1967; Spaulding et al., 1966),

and the degradation of histone mRNAs and cessation of

histone protein synthesis in parallel with the inhibition of











DNA replication (Spaulding et al., 1966; Robbins and Borun,

1967; Borun et al., 1967; Baumbach et al., 1984). To address

the relationship between the localization of histone mRNAs

and the control of cellular histone mRNA levels, we have

used cloned human histone genes to examine the distribution

of histone mRNAs on nonmembrane-bound and membrane-bound

polysomes and the association of histone mRNA-containing

polysomes with the cytoskeleton. As predicted from previous

results of in vitro translation analysis and cDNA

hybridization studies, histone mRNAs were found to reside

predominantly on nonmembrane-bound polysomes. Both the

nonmembrane-bound polysomes containing histone mRNAs and the

membrane-bound polysomes containing the cell surface class I

HLA-B7 antigen mRNAs were found associated with the

cytoskeleton.



RESULTS

Localization of Histone mRNA and HLA-B7 mRNA in Subcellular
Fractions

To determine the intracellular locations of histone

mRNA and HLA-B7 mRNA, HeLa cells were first fractionated

into nonmembrane-bound and membrane-bound polysomal

components by the procedure outlined in Fig. 2-2. The

membrane-bound polysomes are represented by the mitochondria

associated rough endoplasmic reticulum (MRER) and the

nucleus associated rough endoplasmic reticulum (NRER)










fractions. The light endoplasmic reticulum (LER), which is

deficient in polysomes by definition, was also isolated

following this subcellular fractionation scheme. Histone and

HLA-B7 mRNA content in each of the subcellular fractions was

determined by Northern blot analysis. Briefly, RNA was

extracted from the samples, quantitated and electrophoresed

in agarose under denaturing conditions. The RNA was then

transferred to nitrocellulose filters for hybridization with
P-labeled (nick-translated) pFOO08A (H4 histone), pFF435C

(H3 histone) and pDP001 (cDNA clone of HLA-B7 class I

antigen) DNA probes. A typical screening of subcellular

fractions with these probes is represented by the

autoradiograph in Figure 3-1, and Table 3-1 summarizes the

numerical values derived from densitometric analysis of the

Northern blot. The HLA-B7 mRNA and histone mRNA species are

each present in total cellular RNA and total polysomal RNA

which were included as positive controls. Non-membrane-bound

polysomal RNA contains approximately 55% and 51% of the

total H4 and H3 histone mRNA, respectively. Surprisingly,

the LER fraction contains approximately 38% of the H4

histone mRNA and 41% of the H3 histone mRNA, which may be a

consequence of the method by which the fraction is isolated.

The LER and non-membrane-bound polysomal sample is

centrifuged through a 2 M sucrose pad under conditions where

the LER forms a loose band on top of the pad, while the

nonmembrane-bound polysomes are pelleted. The LER band is











removed by aspiration and occasionally material is removed

from within the sucrose pad. This material may be the source

of histone mRNA in the LER fraction, which could decrease

the true percentage of total histone message found on the

nonmembrane-bound polysomes. The HLA-B7 mRNA was located

almost exclusively on membrane-bound polysomes. Nearly 90%

of the HLA-B7 mRNA was collectively found on the NRER and

MRER, while less than 11% was measured in the LER and

nonmembrane-bound polysomal RNA fractions.



Histone mRNA and HLA-B7 mRNA Distribution Between Csk and
Sol Fractions

Existing within the cytoplasm of eukaryotic cells is an

intricate network of protein filaments, which is referred to

as the cytoskeleton. This structure can be prepared from

HeLa cells treated with non-ionic detergent as described by

Penman and coworkers (Cervera et al., 1981). We used this

procedure for the isolation of the cytoskeleton (Csk) and

soluble (Sol) fractions which were subsequently assayed for

mRNA content by Northern blot analysis.

Log phase, suspension grown HeLa S3 cells were

fractionated into a Csk phase and Sol phase as described in

Chapter 2. Although this fractionation procedure has been

characterized previously (Lenk and Penman, 1979), we

examined the protein heterogeneity of each fraction by two-

dimensional gel electrophoresis and immunoblotting

techniques in collaboration with Dr. Warren Schmidt. As seen









Figure 3-1. Northern blot analysis of H3 histone. H4 histone
and HLA-B7 mRNA in subcellular fractions.

The cytoskeleton and soluble fractions, as well as the
nonmembrane-bound and membrane-bound polysomal RNA
fractions, were isolated from HeLa cells by the procedures
described in Chapter 2 (Cervera et al., 1981; Venkatesan and
Steele, 1972). In addition, total polysomal RNA and total
cellular RNA were isolated (Plumb et al., 1983, Venkatesan
and Steele, 1972). RNA was extracted from each subcellular
fraction and 50 yg of RNA from each preparation was resolved
by 1.5% agarose, 6% formaldehyde gel electrophoresis. The
RNA was transferred to nitrocellulose and hybridized to
radiolabeled pFF435C (H3 histone and pFO108A (H4 histone)
probes, followed by hybridization to P-labeled (nick-
translated) pDP001 (cDNA of HLA-B7) probe. The hybridized
filters were then exposed to preflashed XAR5 x-ray film at -
700C for 72 hours. From left to right, lanes: 1) total
cellular RNA; 2) Csk RNA; 3) Sol phase RNA; 4) NRER; 5)
MRER; 6) nonmembrane-bound polysomal RNA; 7) LER; and 8)
total polysomal RNA.























H LA-B7








1 2 3 4 5 6 7 8









82

Table 3-1. Quantity of histone and HLA-B7 mRNAs in the
subcellular fractions.

H3 H4 HLA-B7
Cytoskeleton 90% 84% 97%
Soluble 10% 16% 3%

NRER 3% 3% 29%
MRER 6% 5% 60%
Free 51% 54% 5%
LER 40% 38% 6%

The distribution of histone and HLA-B7 mRNAs in the
NRER, MRER, LER and nonmembrane-bound polysomal fractions as
well as the cytoskeleton and soluble fractions was
determined by Northern blot analysis. The values have been
normalized with respect to the total yield of RNA within
each fraction.











in Figure 3-2, the cytoskeleton fraction clearly contained

actin as well as the major HeLa cytokeratins (Franke et al.,

1981). In addition, vimentin appeared to be enriched in the

cytoskeleton, which was confirmed by immunoblot analysis.

Extraction of the cytoskeleton required a cold, calcium

containing buffer and consequently the tubulins were

solubilized and therefore absent in the Csk fraction. The

tubulin proteins were clearly contained in the soluble

fraction as determined by immunoblot analysis. Additional

support for separation of cytoskeleton and soluble phases

comes from the observation that the bulk of the tRNA species

is in the soluble phase which is in agreement with

previously reported values (data not shown; Cervera et al.

1983). Although the HLA-B7 messenger RNA is found on

membrane-bound polysomes and the histone mRNA on

nonmembrane-bound polysomes, both species of messenger RNAs

are associated almost exclusively with the cytoskeletal

fraction. Nearly 97% of the HLA-B7 mRNA and approximately

85-90% of the H3 and H4 histone messages are retained in the

Csk (Fig. 3-1; Table 3-1).



The Effects of Metabolic Inhibitors on the Subcellular
Localization of Histone and HLA-B7 mRNA

Detectable levels of histone mRNA were observed in each

of the subcellular samples, including the Sol fraction and

the membrane-bound polysomal fractions. However, the HLA-B7









Figure 3-2. Two-dimensional gel electrophoresis of
cvtoskeleton and soluble associated proteins.

HeLa cells were fractionated into a cytoskeleton and
soluble phase, protein was extracted from each phase and
subsequently analyzed by two-dimensional gel electrophoresis
as described in Chapter 2. Proteins of the soluble fraction
(Sol), A, were separated using wide range (pH 3-7)
ampholines while cytoskeletal proteins (Csk), B, were
separated with narrow range (pH 5-7) ampholines to more
clearly identify cytoskeletal species. V: vimentin, a:
actin, and t: tubulins. Unlabeled arrows in B depict HeLa
cytokeratins of 54 Kd (pI 6.0); 52.5 Kd (pI 6.1); 46 Kd (pI
5.1); and 45 Kd (pI 5.7), which were spots highly reactive
with commercially prepared prekeratin or stratum corneum
antisera (Franke et al., 1981).









pH
7.0 6.4
T v


5.2 4.1
T T


5.4 5.0 4.4
T T V


A)








43K-


43K-


I A


CSK


SOL ...









86

mRNA was located almost exclusively in the cytoskeleton and

membrane-bound polysome fractions. The widespread

subcellular localization of histone mRNA may be due to non-

specific trapping of nonmembrane-bound polysomes in each of

the fractions during the isolation procedure. Alternatively,

this observation may reflect the presence of mRNAs encoding

variant histone proteins that could be targeted to the

different cytoplasmic compartments. Inhibition of DNA

synthesis with hydroxyurea results in a rapid

destabilization of cell cycle dependent histone mRNAs,

whereas cell cycle independent histone mRNAs remain stable

in the presence of this drug (Baumbach et al., 1984; Heintz

et al., 1983; DeLisle et al., 1983; Graves and Marzluff,

1984; Plumb et al., 1983). Conversely, inhibition of protein

synthesis with cycloheximide results in a 2-3 fold increase

in cytoplasmic levels of histone mRNAs (Butler and Mueller,

1973; Baumbach et al., 1984). Differential sensitivity of

histone mRNAs in the subcellular fractions to hydroxyurea or

cycloheximide treatment would favor the latter suggestion

that cell cycle dependent and independent histone gene

transcripts may selectively reside in specific cytoplasmic

compartments.

Treatment of exponentially growing HeLa cells with

hydroxyurea, as described in Chapter 2, resulted in a

complete and selective destabilization of histone mRNA in

the LER, nonmembrane-bound and membrane-bound polysomal











fractions (Figure 3-3). This is in sharp contrast to the

HLA-B7 mRNA which appeared to be fully stable under these

conditions. Cycloheximide treatment had a stabilizing effect

on the levels of histone mRNA in each subcellular fraction

(Figure 3-3). When compared with control HeLa cells,

cycloheximide treatment resulted in a 2.1 fold (H4) and a

2.4 fold (H3) increase in the levels of histone mRNA in the

nonmembrane-bound polysomal fraction. Similar increases were

also seen for levels of histone mRNAs in the LER and

membrane-bound polysomal fractions. Inhibition of protein

synthesis with cycloheximide followed by DNA synthesis

inhibition with hydroxyurea resulted in the stabilization of

histone mRNA in each subcellular fraction, whereas the

stability of HLA-B7 mRNA remained unaffected (Figure 3-3).

Similar results were obtained when these inhibitor studies

were expanded to include the cytoskeleton and soluble

fractions (Figure 3-4). Cycloheximide treatment resulted in

a 2-3 fold increase in the representation of H3 and H4

histone mRNA over untreated cultures in both the

cytoskeleton and soluble fractions. HLA-B7 mRNA appeared to

be only moderately stabilized in the presence of this drug.

When cells were treated with hydroxyurea, histone mRNA

levels were greatly reduced in both the cytoskeleton and

soluble fractions whereas the HLA-B7 mRNA was found in

comparable levels to those present in the untreated samples.

Consistent with previous data, cotreatment of HeLa cells









Figure 3-3. Northern blot analysis of H3 histone. H4 histone
and HLA-B7 mRNAs isolated from subcellular fractions of
cells treated with metabolic inhibitors.

Exponentially growing HeLa cells were incubated in the
presence of metabolic inhibitors and fractionated into
nonmembrane-bound and membrane-bound polysomes as described
in Chapter 2. RNA was extracted from each fraction and
assayed for H3 histone mRNA, H4 histone mRNA and HLA-B7 mRNA
content by Northern blot analysis as described in Figure 3-1
and Chapter 2. Upper panel from left to right, lanes: 1)
control (untreated), nonmembrane-bound polysomes; 2)
control, NRER; 3) control, MRER; 4) control, LER (omitted);
5) cycloheximide (Cy), nonmembrane-bound polysomes; 6) Cy,
NRER; 7) Cy, MRER; and 8) Cy, LER. Lower panel from left to
right, lanes: 1) hydroxyurea (HU), nonmembrane-bound
polysomes; 2) Hu, NRER; 3) Hu, MRER; 4) Hu, LER; 5) CY/HU,
nonmembrane-bound polysomes; 6) Cy/Hu, NRER; 7) Cy/Hu, MRER;
and 8) Cy/Hu, LER.




Full Text
172
In addition, CD at concentrations equal to or greater than
10 jLtg/ml releases poly-A+ RNA from the cytoskeleton in a
dose dependent manner (Ornelles et al.. 1986). The release
of poly-A+ RNA from the cytoskeleton occurs in a dose
dependent manner and results in a stoichiometric inhibition
of protein synthesis.
To study immediate gene responses to cytochalasin D,
exponentially growing HeLa S3 cells were treated with 10
lig/ml of the drug for only 15 minutes. From each cell
culture total cellular RNA for Northern blot analysis and
nuclei for in vitro nuclear run-on transcription analysis
were isolated as described in Chapter 2. As shown in Figure
7-1, Northern blot analysis reveals a rapid and extensive
accumulation of c-fos mRNA in CD treated cells (lane 2). The
response of the c-fos gene is in sharp contrast to hCGar mRNA
steady state levels which are unaffected by the drug
treatment (Fig. 7-1). Quantitation by scanning laser
densitometry of several exposures of autoradiograms obtained
from four separate experiments indicates an average increase
of 20 fold in c-fos mRNA levels (Fig. 7-2). The
concentration of cytochalasin D used in these experiments
inhibits protein synthesis by 50 percent (Ornelles et al..
1986) It is well established that inhibition of protein
synthesis leads to an increase in c-fos mRNA presumably due
to increased mRNA stability (Muller et al., 1984; Mitchell
et al. 1985? Andrews et al.. 1987? Rahmsdorf et al.. 1987).


4-8 Effects of hydroxyurea treatment on SPH3E1ATG"
mRNA 121
4-9 Effects of hydroxyurea treatment on SPH3Elalpha
mRNA 123
4-10 Quantitation of SPH3E1, SPH3Elalpha, SPH3E1ATG'
mRNA during hydroxyurea treatment 125
5-1 Northern blot analysis of the cytoskeleton and
soluble phase distribution of H3 histone, H4 histone
and HLA-B7 mRNAs in cytochalasin D treated cells.... 135
5-2 Northern blot analysis of the cytoskeleton and
soluble phase distribution of c-fos mRNA in
cytochalasin D, puromycin, CD/puro, and
CD/cycloheximide treated cells 140
5-3 Northern blot analysis of the cytoskeleton and
soluble phase distribution of nonmembrane-bound
and membrane bound polysomal mRNAs in cytochalasin D,
puromycin, CD/puro and CD/cycloheximide treated
cells 145
6-1 Cytochalasin D and puromycin cotreatment does not
release SPH3E1 or SPH3E1ATG mRNA from the
cytoskeleton 157
6-2 The wild type signal peptide-histone fusion mRNA is
not released from the cytoskeleton in CD and
puromycin cotreated polyclonal HeLa cell cultures... 161
6-3 Endogenous membrane-bound polysomal mRNAs are
released from the cytoskeleton during cotreatment
with CD and puromycin in SPH3E1 and SPH3E1ATG"
expressing cell cultures 163
7-1 Effects of cytochalasin D and puromycin treatment on
steady state levels of c-fos and chorionic
gonadotropin mRNA 175
7-2 Densitometric analysis of steady state levels of
c-fos mRNA during cytochalasin D and puromycin
treatments 177
vii


211
Huxley, H. (1973) Nature 243, 445.
Isaacs, W., and Fulton, A. (1987) Proc. Natl. Acad. Sci.
U.S.A. 84/ 6174.
Izant, J., and Lazarides, E. (1977) Proc. Natl. Acad. Sci.
U.S.A. 74, 1450.
Jacobs-Lorena, M., Baglioni, C., and Borun, T. (1972) Proc.
Natl. Acad. Sci. U.S.A. 69, 2095.
Jeffery, W. (1984) Dev. Biol. 103, 482.
Kaiser, C., and Botstein, D. (1986) Mol. Cell. Biol. 6,
2382.
Kaiser, C., Preuss, D., Grisafi, P., and Botstein, D. (1987)
Science 235. 312.
Kane, R. (1975) J. Cell Biol. 66, 305.
Kedes, L. (1979) Annu. Rev. Biochem. 4_8, 837.
Laeminli, U. (1970) Nature 227. 680.
Lawrence, J., and Singer, R. (1986) Cell 45, 407.
Lawrence, J., Singer, R., Villnave, C., Stein, J., and
Stein, G. (1988) Proc. Natl. Acad. Sci. U.S.A. 85, 463.
Lazarides, E. (1980) Nature 283. 249.
Lazarides, E., and Hubbard, B. (1976) Proc. Natl. Acad. Sci.
U.S.A. 73, 4344.
Lemieux, R., and Beaud, G. (1982) Eur. J. Bioch. 129, 273.
Lenk, R., and Penman, S. (1979) Cell 6, 289.
Lenk, R., Ransom, L., Kaufmann, Y., and Penman, S. (1977)
Cell 10, 67.
Levine, B., Chodchoy, N., Marzluff, W., and Skoultchi, A.
(1987) Proc. Natl. Acad. Sci. U.S.A. M/ 6189.
Li, W., Reddy, R., Henning, D., Epstein, P., and Busch, H.
(1982) J. Biol. Chem. 257, 5136.
Liautard, J., and Jeanteur, P. (1979) Nucl. Acids Res. 7
135.


CHAPTER 3
SUBCELLULAR LOCATION OF HISTONE mRNAs ON CYTOSKELETON
ASSOCIATED NONMEMBRANE-BOUND POLYSOMES IN HELA S3 CELLS
INTRODUCTION
Histone biosynthesis occurs primarily during the S
phase of the cell cycle and is temporally as well as
functionally coupled with DNA replication (Allfrey et al..
1963; Prescott, 1966; Spaulding et al. 1966; Robbins and
Borun, 1967; Stein and Borun, 1972). Early studies using in
vitro translation analysis of polysomal RNA isolated from S
phase HeLa cells provided an initial indication that histone
protein synthesis occurs on non-membrane-bound, light
polysomes (Jacobs-Lorena et al.. 1972; Liautard and
Jeanteur, 1979; Gallwitz and Breindl, 1972; Borun et al.
1967). In these studies it was demonstrated that
translatable, polysome-associated histone mRNAs have a
sedimentation coefficient of 8-10S and fluctuate in a cell
cycle-dependent manner. Histone mRNA was not detected by in
vitro translation studies on polyribosomes isolated from
HeLa cells in the G1 phase of the cell cycle (Borun et al..
1967; Borun et al.. 1975; Pederson and Robbins, 1970).
Consistent with this observation, the inhibition of DNA
replication by hydroxyurea or cytosine arabinoside treatment
75


mug. I would particularly like to thank the following
individuals for their tremendous inspiration: April, Emilie,
Ashley, Danny, Shannon, John, Aubrey, Stephan, Alan, Brooke,
Kate, Sarah and Nikki.
I would like to thank the members of the Department of
Immunology and Medical Microbiology. I will always remember
the football games in which Dr. Ed Siden led our team to an
undefeated record against the Pathology team starring Dr.
Ward Wakeland, and the late night ,,Nerf-ball" baseball games
in the hallway with Mark Labow, Paul Hermonat, Brian Masters
and Dick Wright.
Lastly, I would like to thank my family. The Lord has
truly blessed me with a loving and supportive family and I
fully appreciate all that they have done and sacrificed for
me during my studies.
iii


132
with the cytoskeleton (Jeffery, 1984; Stein et al.. 1977?
Adensik and Darnell, 1972; Zambetti et al.. 1985). The
distribution of histone mRNAs between the cytoskeletal and
the soluble fractions was determined by Northern blot
analysis before and after cytochalasin D treatment. As seen
in Figure 5-1, H3 and H4 cell cycle dependent histone mRNAs
are primarily associated with the cytoskeleton in untreated
HeLa cell cultures (lanes 1 and 2). Quantitation of the
autoradiograph in Figure 5-1 by scanning laser densitometry
revealed that >90% of the histone mRNAs were associated with
the cytoskeleton in control cell cultures (Table 5-1), which
is consistent with previously reported values (Zambetti et
al.. 1985). Upon addition of cytochalasin D, histone mRNAs
were released from the cytoskeleton into the soluble phase
(Fig. 5-1, lanes 3-8). As shown in Table 5-1, less than 35%
of the histone mRNAs remained associated with the
cytoskeleton in HeLa cell cultures treated with 10 g/ml
cytochalasin D. Treatment with 40 ^ig/xa. 1 cytochalasin D
resulted in the dissociation of >70% of the histone mRNAs
from the cytoskeleton. The extent to which cytochalasin D
releases non-polyadenylated histone mRNA from the
cytoskeleton is similar to that reported for total,
heterogeneous poly-A+ RNA (Ornelles et al. 1986).


BIOGRAPHICAL SKETCH
Gerard Paul Zambetti was born in Queens, New York, on
August 4, 1958. He attended Archbishop Molloy High School
(Queens, New York) and graduated in the spring of 1976. He
then attended the State University of New York at
Plattsburgh and graduated with a Bachelor of Science degree
in biochemistry and biophysics in the spring of 1980. He
then studied bacteriophage Lambda DNA replication and gene
regulation with Dr. Robert Shuster at Emory University
(Atlanta, Georgia) and graduated with a Master of Science
degree in biochemistry. He continued to research the
regulation of gene expression with Drs. Janet and Gary Stein
in the Department of Immunology and Medical Microbiology at
the University of Florida. In the fall of 1984, he married
Stacey Ann Chapman in Atlanta, Georgia, and on May 6, 1987,
they were blessed with a beautiful daughter, Shannon Kelly
Zambetti.
219


CHAPTER 2
GENERAL METHODS
Materials
[a-32P]dCTP (-3,000 Ci/mmol) [a-32P]dATP (-3,000
Ci/mmol), [gamma-32P]ATP (-3,000 Ci/mmol) and [a-35S]dATP
(>1,000 Ci/mmol) radionucleotides were purchased from
Amersham, Arlington Heights, IL; polyvinylsulfonic acid
(PVS), polyethylene glycol 8,000 (PEG), XAR-5 x-ray film
were obtained from Eastman Kodak Co., Rochester, NY;
ultrapure electrophoresis grade agarose and Zeta-probe nylon
membranes were from Bio Rad, Richmond, CA; formaldehyde
solution (37% w/w) was from Fisher Scientific, Fair Lawn,
NJ; nitrocellulose was from Schleicher and Schuell, Keene,
NH; RPI scintillator ppo-popop concentrated liquid
scintillator (Liquifluor) was from Research Products
International, Elk Grove Village, IL; formamide was from
Bethesda Reasearch laboratories (BRL), Gaithersburg, MD;
acrylamide, calf intestine alkaline phosphatase, nuclease
SI, Ej¡_ coli DNA polymerase I large fragment (Klenow enzyme)
and most restriction endonucleases were obtained from
Boehringer Mannheim Biochemicals (BMB), Indianapolis, IN; T4
DNA ligase was from New England Bio Labs (NEB), Beverly, MA;
27


36
standard protocol for the isolation of pUC8 related plasmid
DNAs is the same as described below except that the 500 ml
culture is inoculated directly and allowed to grow overnight
into the late log stage.
Starter cultures were prepared by inoculating 5 ml LB
medium"(Luria-Bertani medium: 1% Bacto tryptone, 0.5% Bacto
yeast extract, 1% NaCl) supplemented with 0.1% glucose with
25-50 ul of bacterial glycerol stock and incubated overnight
at 37C. The 5 ml starter culture was transferred to 500 ml
pre-warmed LB medium supplemented with glucose and incubated
at 37C until the optical density at 590 nm (O.D. 590) of the
culture reached 0.55 units. Plasmid DNA was amplified by
treating the culture with 200 ig/ml chloramphenicol
(prepared as a 20 mg/ml stock solution in 95% ethanol and
filter sterilized) at 37C for 12-16 hours. The bacteria
were centrifuged at 5,000 rpm, 4C for 10 minutes in a
Beckman JA-10 rotor and J2-21 centrifuge. The cell pellet
was resuspended in 10 ml solution I (50 mM glucose, 25 mM
Tris pH 8.0, 10 mM EDTA, 5 mg/ml lysozyme) and incubated at
room temperature for 5 minutes. The cells were lysed at 0C
with 20 ml solution II (1% SDS, 0.5 M NaOH) for 5 minutes.
The cell lysate was treated with 15 ml 5 M potassium
acetate, pH 4.8 and kept on ice for 5 minutes. The lysate
was transferred to a 50 ml polycarbonate tube and
centrifuged at 16,000 rpm for 20 minutes at 4C in Beckman
JA-20 rotor to pellet the bacterial DNA and cell debris. The


45
HeLa Cells
homogenize in RSB
centrifuge 2,000xg, 4C, 5'
I
supernatant A
pellet A
centrifuge ll,400xg
10'
I
supernatant B
pellet B
Triton/NaDOC
centrifuge 7,800x
10'
I I
Triton/NaDOC
centrifuge 500xg
10'
discard pellet
discard pellet supernatant D supernatant C
centrifuge
200,000xg, 2 hrs
2M sucrose
centrifuge
200,000xg, 2 hrs
2M sucrose
centrifuge
200,000xg, 2 hrs
2M sucrose
I 1
Band at pellet pellet pellet
interface
LER
Nonmembrane-bound
NRER
MRER


THE ROLE OF SUBCELLULAR COMPARTMENTS AND STRUCTURES
IN EUKARYOTIC GENE EXPRESSION
By
GERARD PAUL ZAMBETTI
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1989

ACKNOWLEDGEMENTS
I would like to acknowledge the dedication of Drs.
Janet and Gary Stein to their work. Their commitment to
science has provided us with a well-equipped lab and the
funds to pursue our research. I would also like to extend
special thanks to the Stein's for their constant support and
guidance during my graduate studies. In addition, I would
like to thank Drs. Bert Flanegan, Tom Rowe and Dick Moyer
for their helpful advice as members of my graduate
committee.
My experiences with my fellow coworkers in the Stein's
lab have been very interesting, rewarding and fulfilling. I
would like to acknowledge and thank Marie Giorgio, Linda
Green, Farhad Marashi, Brian O'Donnell and Mary Beth Kennedy
for their technical guidance and assistance, and Ken Wright
and Joe Teixeira for their help dealing with the computers.
I thank Pat Naples for her appreciation and pledge of care
for "the coffee mug." I would like to thank Dave Collart,
Tim Morris and Charles Stewart for their friendship and for
their advice on raising bees, fishing, and golfing,
respectively. I would like to thank Andre van Wijnen for his
friendship, scientific expertise on and off the "Foos" ball
field, and for keeping an eye on Pat Naples and my coffee
11

mug. I would particularly like to thank the following
individuals for their tremendous inspiration: April, Emilie,
Ashley, Danny, Shannon, John, Aubrey, Stephan, Alan, Brooke,
Kate, Sarah and Nikki.
I would like to thank the members of the Department of
Immunology and Medical Microbiology. I will always remember
the football games in which Dr. Ed Siden led our team to an
undefeated record against the Pathology team starring Dr.
Ward Wakeland, and the late night nNerf-ball" baseball games
in the hallway with Mark Labow, Paul Hermonat, Brian Masters
and Dick Wright.
Lastly, I would like to thank my family. The Lord has
truly blessed me with a loving and supportive family and I
fully appreciate all that they have done and sacrificed for
me during my studies.
iii

TABLE OF CONTENTS
PAGE
ACKNOWLEDGEMENTS ii
LIST Of1 FIGURES vi
LIST OF TABLES ix
KEY TO ABBREVIATIONS X
ABSTRACT xi
CHAPTERS
1) INTRODUCTION
General Background 1
Histone Genes and Proteins 2
Histone Gene Regulation: Transcriptional and Post-
Transcriptional Control 3
Subcellular Compartments 10
The Cytoskeleton: Microtubules, Intermediate
Filaments and Microfilaments 12
Cytoskeletal-mRNA Interactions 16
Nonmembrane-Bound and Membrane-Bound Polyribosomes:
The Signal Hypothesis 19
Overview of Project 22
2) GENERAL METHODS
Materials 27
Conditions for Enzyme Reactions 28
Mammalian Cell Culture 28
Plasmid DNA 3 0
DNA Isolation and Purification 35
Isolation of Mammalian RNA 40
Two-Dimensional Gel Electrophoresis and Immuno-
Blotting Analysis 48
Northern Blot Analysis 49
SI Nuclease Protection Analysis 52
In Vitro Nuclear Run-on Transcription Analysis 55
Radiolabeling DNA for Northern Blot Analysis 58
Site-Directed Mutagenesis 61
Transfection of DNA into Cells 71
Establishment of Polyclonal and Monoclonal Stably
Transformed Cell Lines 73
IV

3) SUBCELLULAR LOCATION OF HISTONE mRNAs ON CYTOSKELETON
ASSOCIATED NONMEMBRANE-BOUND POLYSOMES IN HELA S3
CELLS
Introduction 75
Results 77
Discussion 90
4) SUBCELLULAR LOCATION OF HISTONE mRNA PLAYS A ROLE IN
THE POSTTRANSCRIPTIONAL REGULATION OF HISTONE
GENE EXPRESSION
Introduction 96
Results 97
discussion 126
5) DIFFERENTIAL ASSOCIATION OF MEMBRANE-BOUND AND
NONMEMBRANE-BOUND POLYSOMES WITH CELL CYTOSTRUCTURE
Introduction 130
Results 131
Discussion 146
6) HETEROGENEOUS PATTERN FOR CYTOSKELETAL-mRNA
INTERACTIONS
Introduction 153
Results 154
Discussion 166
7) THE INFLUENCE OF THE CYTOSKELETON ON THE REGULATION
OF C-Fos GENE EXPRESSION
Introduction 168
Results 171
Discussion 190
8) SUMMARY AND FUTURE CONSIDERATIONS 196
REFERENCES 206
BIOGRAPHICAL SKETCH 219
v

LIST OF FIGURES
Figure Page
2-1 Outline of the cloning scheme for the construction of
the signal peptide-histone fusion gene 34
2-2 Outline of procedure for the isolation of subcellular
polysomal fractions 45
3-1 Northern blot analysis of H3 histone, H4 histone and
HLA-B7 mRNA in subcellular fractions 81
3-2 Two-dimensional gel electrophoresis of cytoskeleton
and soluble associated proteins 85
3-3 Northern blot analysis of H3 histone, H4 histone and
HLA-B7 mRNAs isolated from subcellular fractions of
cells treated with metabolic inhibitors 89
3-4 Northern blot analysis of H3 histone, H4 histone and
HLA-B7 mRNAs associated with the cytoskeleton and
soluble fractions from HeLa cells treated with
metabolic inhibitors 92
4-1 Schematic diagram of endogenous H3 histone mRNA ST519
and the beta-lactamase signal peptide-H3 histone
fusion mRNA 100
4-2 Expression of the signal peptide-histone fusion gene
in HeLa cells 103
4-3 The subcellular localization of the signal peptide-
histone fusion mRNA 106
4-4 The stability of the signal peptide-histone fusion
mRNA following inhibition of DNA synthesis 109
4-5 Schematic diagram of wild type and mutated signal
peptide-histone fusion mRNAs 112
4-6 Distribution of SPH3E1, SPH3Elalpha, and SPH3E1ATG'
mRNAs in nonmembrane-bound and membrane-bound
polysomal fractions 115
4-7 Effects of hydroxyurea treatment on SPH3E1 mRNA 119
vi

4-8 Effects of hydroxyurea treatment on SPH3E1ATG
mRNA 121
4-9 Effects of hydroxyurea treatment on SPH3Elalpha
mRNA 123
4-10 Quantitation of SPH3E1, SPH3Elalpha, SPH3E1ATG'
mRNA during hydroxyurea treatment 125
5-1 Northern blot analysis of the cytoskeleton and
soluble phase distribution of H3 histone, H4 histone
and HLA-B7 mRNAs in cytochalasin D treated cells.... 135
5-2 Northern blot analysis of the cytoskeleton and
soluble phase distribution of c-fos mRNA in
cytochalasin D, puromycin, CD/puro, and
CD/cycloheximide treated cells 14 0
5-3 Northern blot analysis of the cytoskeleton and
soluble phase distribution of nonmembrane-bound
and membrane bound polysomal mRNAs in cytochalasin D,
puromycin, CD/puro and CD/cycloheximide treated
cells 145
6-1 Cytochalasin D and puromycin cotreatment does not
release SPH3E1 or SPH3E1ATG’ mRNA from the
cytoskeleton 157
6-2 The wild type signal peptide-histone fusion mRNA is
not released from the cytoskeleton in CD and
puromycin cotreated polyclonal HeLa cell cultures... 161
6-3 Endogenous membrane-bound polysomal mRNAs are
released from the cytoskeleton during cotreatment
with CD and puromycin in SPH3E1 and SPH3E1ATG"
expressing cell cultures 163
7-1 Effects of cytochalasin D and puromycin treatment on
steady state levels of c-fos and chorionic
gonadotropin mRNA 175
7-2 Densitometric analysis of steady state levels of
c-fos mRNA during cytochalasin D and puromycin
treatments 177
Vll

7-3 Time course of c-fos mRNA accumulation during
cytochalasin D and puromycin treatments in
exponentially growing HeLa cells 181
7-4 Effect of cytochalasin D on transcription of c-fos
in HeLa cells 184
7-5 Effect of protein synthesis inhibition on
cytochalasin D induction of c-fos transcription in
HeLa cells 188
viii

LIST OF TABLES
Table Page
3-1 Quantity of histone and HLA-B7 mRNAs in the
subcellular fractions 82
4-1 Quántitation of SPH3E1, SPH3Elalpha and SPH3E1ATG*
mRNA in nonmembrane-bound and membrane-bound
polysome fractions 116
5-1 The percent of cytoskeleton and soluble phase
associated mRNAs in cells treated with
cytochalasin D 136
5-2 Percent distribution of nonmembrane-bound and
membrane bound polysomal RNAs in the Csk and Sol
fractions from CD, Puro, CD/puromycin and
CD/cycloheximide treated cells 141
6-1 The percent of cytoskeleton associated mRNAs
isolated from cytochalasin D, puromycin and
CD/puromycin treated cells 158
IX

KEY TO ABBREVIATIONS
BSA
Bovine Serum Albumin
cpm
Counts per minute
dCTÍ>
Deoxycytidine-5'-triphosphate
ddCTP
Dideoxycytidine-5'-triphosphate
DNA
Deoxyribonucleic acid
DNase
Deoxyribonuclease
DTT
Dithiothreitol
EDTA
Ethylenediaminetetraacetic acid
Hepes
N-2 hydroxyethyl piperazine-N'-2
ethanesulfonic acid
HU
Hydroxyurea
¡iCi
Microcurie
Mg
Microgram
Ml
Microliter
/im
Micrometer
mm
Micromolar
OD
Optical density
PVS
Polyvinyl sulfate
RNA
Ribonucleic acid
rpm
Revolutions per minute
SDS
Sodium dodecylsulfate
SSC
Standard saline citrate buffer
Tris
(hydroxymethyl)aminomethane
x

Abstract of Dissertation Presented to the Graduate School of the
University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
THE ROLE OF SUBCELLULAR COMPARTMENTS AND STRUCTURES
IN EUKARYOTIC GENE EXPRESSION
By
Gerard Paul Zambetti
December 1989
Chairperson: Janet Stein
Major Department: Immunology and Medical Microbiology
Histone mRNAs naturally reside on nonmembrane-bound
polysomes that are associated with the cytoskeleton. To
study whether the subcellular location plays a role in the
coupling of histone mRNA stability to DNA synthesis, a
signal peptide-histone fusion gene was constructed that
encodes a histone mRNA that is directed to membrane-bound
polysomes. The signal peptide-histone fusion mRNA when
localized in the foreign subcellular compartment is no
longer destabilized when DNA synthesis is inhibited. A
single nucleotide substitution in the signal peptide-histone
fusion gene by site-directed mutagenesis resulted in the
localization of the histone fusion mRNA on nonmembrane-bound
polysomes and restored the coupling of mRNA stability with
xi

DNA synthesis. These results indicate that the subcellular
location of histone mRNA plays an important role in the
posttranscriptional regulation of histone gene expression.
We have determined that the association of histone mRNA
with the cytoskeleton is dependent on the integrity of actin
filaments. Disruption of the microfilaments with
cytochalasin D efficiently released histone mRNA from the
cytoskeleton into the soluble phase. In contrast, membrane-
bound polysomal mRNAs remained attached to the cytoskeleton
during cytochalasin D treatment. Subsequent studies have
demonstrated that membrane-bound polysomal mRNAs are
attached to the cytoskeleton at two different sites: 1) a
cytochalasin D sensitive site and 2) a puromycin sensitive
site.
Disruption of the cytoskeleton with cytochalasin D
resulted in a rapid and marked stimulation in transcription
of the c-fos proto-oncogene and an increase in c-fos mRNA
cellular levels. Transcription of several other genes was
unaffected by the cytochalasin D treatment. These results
suggest that the nucleus can respond to the structural
organization of the cytoskeleton and selectively adjust the
regulation of gene expression accordingly.
xii

CHAPTER 1
INTRODUCTION
General Background
Histones are small basic proteins that are involved in
the packaging of eukaryotic DNA into chromatin. It is
estimated that more than 2 meters of DNA are contained
within the limited area of the nucleus. The condensation of
the DNA to such a great extent is due primarily to its
interactions with the histone proteins. In addition to their
structural properties histones play a functional role in
controlling gene transcription. The way in which the histone
proteins are arranged along a region of DNA appears to
affect the transcriptional properties of that DNA domain.
The complex nature of the regulation of histone protein
synthesis at the transcriptional, posttranscriptional, and
posttranslational levels makes these genes an important
model for studying eukaryotic gene expression. However, one
of the most important reasons for studying the regulation of
histone gene expression is that the synthesis of the
majority of histone proteins (cell cycle dependent histones)
occurs during DNA replication. Elucidation of the mechanisms
coupling histone protein synthesis with DNA synthesis may
1

2
yield a better understanding of the processes controlling
proliferation, differentiation, cell aging and neoplasia.
Histone Genes and Proteins
There are five major classes of histone proteins (H4,
H2B, HÍA, H3, and HI) ranging in molecular weight from
11,000 to 21,000 Daltons. They are basic proteins that have
been highly conserved throughout evolution. For example,
there are only two amino acids out of 102 which differ
between histone H4 from pea seedlings and calf thymus, which
suggests a critical role for these proteins in cell survival
(DeLange et al.. 1969). The histone proteins are encoded by
a family of moderately repeated genes (Wilson and Melli,
1977; Kedes, 1979). In humans, the histone genes are
arranged in clusters that are not organized in simple tandem
repeats as found in the sea urchin and other lower
eukaryotic genomes (Heintz et al. . 1981; Sierra et al..
1982; Carozzi et al.. 1984). The clusters are localized on
chromosomes 1, 6, and 12 and are interspersed with
repetitive elements (Green et al.. 1984b; Collart et al.,
1985). It has been speculated that these repetitive elements
may play a role in recombination and gene duplication during
the evolution of this multi-gene family. The general
structure of the human histone genes is less complex than
many other eukaryotic genes since the histone protein coding
region is not interrupted by introns and the encoded mRNAs

3
are not polyadenylated (Adensik and Darnell, 1972). However,
the genes coding for histone proteins that are synthesized
in the absence of DNA synthesis (cell cycle independent
histones) generally contain introns and code for mRNAs that
are poly-adenylated (Engel et al.. 1982; Harvey et al..
1983; 6rush et al.. 1985; Wells and Kedes, 1985). Subsequent
discussions will generalize what is true only for the cell
cycle dependent histones unless otherwise indicated.
Histone Gene Regulation: Transcriptional and Post-
Transcriptional Control
The stringent requirement to package newly replicated
DNA into chromatin, the invariable 1:1 molar ratio of
histone protein to DNA and the inability of most eukaryotic
cells to store unbound histone proteins implies a temporal
and functional coupling of histone protein synthesis with
DNA replication (Prescott, 1966; Allfrey et al♦. 1963).
Support for this hypothesis first came from studies on
synchronized cell cultures of Euplotes eurvstomus in which
the synthesis of histone protein was found to occur only
during the S phase of the cell cycle, the time when DNA
replication takes place (Prescott, 1966). Furthermore, the
quantity of histone protein within the cell doubled during S
phase. The temporal coupling of histone protein synthesis
with DNA synthesis has since been confirmed and well
documented by a variety of direct experimental approaches

4
(Robbins and Borun, 1967; Spaulding et al.. 1966; Stein and
Borun, 1972). In vitro translation studies using polysomal
RNA isolated from cells at various stages of the cell cycle
indicated that the histone mRNA was present exclusively
during S phase (Spalding et al.. 1966; Gallwitz and Mueller,
1969; Jacobs-Lorena et al.. 1972; Gallwitz and Briendl,
1972; Perry and Kelley, 1973; Borun et al.. 1975).
Quantitation of histone mRNA content during the cell cycle
by Northern blot analysis, using radiolabeled histone cDNA
and genomic clones, demonstrated that histone mRNA steady
levels closely parallel the rates of histone protein
synthesis as well as the rates of DNA replication (Heintz et
al.. 1983; Plumb et al.. 1983a; Baumbach et al.. 1984). The
timing of histone protein synthesis during DNA synthesis is
not coincidental and represents a functional relationship
between these two processes. This is supported by the
following observations: 1) there is a coordinate induction
of histone protein synthesis and DNA synthesis during
stimulation of quiescent cells to proliferate (DeLisle et
al.. 1983; Green et al.. 1984a); 2) inhibition of DNA
synthesis with compounds such as hydroxyurea or cytosine
arabinoside results in a loss of histone mRNA with a
concomitant decrease in histone protein synthesis (Breindl
and Gallwitz, 1973; Butler and Mueller, 1973; Borun et al..
1975; Baumbach et al.. 1984) and 3) terminal differentiation
is marked by a down regulation in histone protein synthesis

5
in conjunction with the decrease in proliferative activity
(Bird et al.. 1985; Challoner et al.. 1989; Stein et al..
1989) .
The expression of histone genes during the cell cycle
is regulated at both the transcriptional and
posttránscriptional levels. Analysis by hybrid selection of
RNA radiolabeled in synchronized HeLa cells demonstrates a
burst of 3- to 5- fold in transcription of the histone genes
within the first two hours of S phase, followed by a
reduction to a basal level that remains constant until the
following S phase (Plumb et al.. 1983a). Similar results
were obtained when histone gene transcription during the
cell cycle was measured by nuclear run-on analysis (Sive et
al.. 1984; Baumbach et al.. 1987). Steady state levels of
histone mRNA, however, were elevated 15-20 fold during S
phase and closely paralleled the rates of DNA synthesis
(Plumb et al.. 1983a; Heintz et al.. 1983; Baumbach et al..
1984) . Maximal accumulation of histone mRNA occurred at 5
hours into S phase and returned to a basal level as DNA
synthesis ended. The timing and extent of the increase in
histone transcription can not account for the marked
increase in histone mRNA steady state levels and must be a
consequence of a posttranscriptional control mechanism.
Posttranscriptional regulation of histone gene expression is
further supported by DNA synthesis inhibitor studies. The
estimated half life of histone mRNA in exponentially growing

6
HeLa cells is 45-60 minutes (Sittman et al., 1983; Plumb et
al.. 1983). Hydroxyurea, or any other compound or treatment
that inhibits DNA synthesis studied to date, results in a
rapid and selective destabilization of histone mRNA (half
life of 10-15 minutes) with little or no effect on histone
gene transcription (Sittman et al.. 1983; Plumb et al.,
1983; Baumbach et al.. 1987).
A long-standing observation concerning histone mRNA
turnover during inhibition of DNA synthesis is that the
mechanism is completely dependent on protein synthesis
(Butler and Mueller, 1973; Gallwitz, 1975; Stahl and
Gallwitz, 1977; Stimac et al.. 1983; Baumbach et al.. 1984;
Helms et al.. 1984). Pretreatment of cells with inhibitors
of protein synthesis such as high salt, cycloheximide or
puromycin followed by the addition of a DNA synthesis
inhibitor prevents the destabilization of histone mRNAs and
actually results in an increase in their steady state levels
due to the continued transcription of the histone genes.
This observation led Butler and Mueller and others to
hypothesize that the histone proteins can autoregulate their
own synthesis through a feedback loop operative at the
posttranscriptional level (Butler and Mueller, 1973; Stein
and Stein, 1984; Wu and Bonner, 1985). The model predicts
that at the natural end of S phase or when DNA synthesis is
interrupted with inhibitors a transient accumulation of
unbound histone protein may arise due to a loss in high-

7
affinity binding sites for the newly synthesized histone
proteins. Unbound histone protein may then interact with
histone mRNA-containing polysomes resulting in a
conformational change in the mRNA that is more accessible to
ribonuclease. Growing support for this model comes from a
cell-free mRNA decay system in which the destabilization of
histone mRNA is accelerated when incubated in the presence
of histone proteins (Peltz and Ross, 1987). This result
appears to be specific for histone mRNA since the rates of
decay for gamma globin mRNA, c-myc mRNA, total poly(A)" RNA
or total poly(A)+ RNA are unaffected by the presence of
histone proteins.
Gene transfer studies demonstrated that the minimal
seguence necessary for coupling histone mRNA stability with
DNA replication lies within the 3' terminus of the mRNA
(Stauber et al.. 1986; Levine et al.. 1987; Pandey and
Marzluff, 1987). Removal of only 60 nucleotides from the 3'
end of a murine H3 histone mRNA releases the mutant mRNA
from destabilization during inhibition of DNA synthesis
(Levine et al.. 1987). Conversely, the last 30 nucleotides
of histone H3 mRNA when fused to the 3' end of human gamma
globin mRNA are sufficient to confer instability to the
chimeric mRNA when DNA synthesis is inhibited, whereas wild
type gamma globin mRNA remains stable (Pandey and Marzluff,
1987). Contained within the minimal target seguence for
destabilization is a stem-loop structure which is common to

8
all known cell cycle dependent histone mRNAs (Hentschel and
Birnstiel, 1981). The presence of the stem-loop structure is
not sufficient by itself to couple mRNA stability with DNA
replication. The target sequences must be present in the
proper position at the 3' terminus of the mRNA to be
operative. Aberrant histone mRNA transcripts containing the
target sequences at an internal position are fully stable
during the inhibition of DNA synthesis (Alterman et al..
1985; Luscher et al.. 1985; Pandey and Marzluff, 1987;
Levine et al.. 1987). Furthermore, moving the stem-loop
structure 380 nucleotides downstream from the translation
stop codon abolishes the coupling of histone mRNA stability
with DNA replication despite the retention of the stem loop
structure at the 3'terminus (Graves et al.. 1987).
An additional constraint on the histone
posttranscriptional regulatory mechanism is that translation
must proceed to within close proximity of the natural
termination codon for destabilization to occur during
inhibition of DNA synthesis (Graves et al.. 1987). Frame-
shift mutations in histone mRNA resulting in the premature
termination of translation more than 300 nucleotides from
the 3' terminus stabilize the messenger RNA when DNA
synthesis is interrupted (Graves et al.. 1987). Other frame-
shift mutants that allow translation to continue past the
natural UGA stop codon into the 3' untranslated region are
also stable when DNA synthesis is inhibited (Capasso et al..

9
1987). Correcting for translation read through in these
mutants by inserting stop codons in all three reading frames
at the natural translation termination site restores the
coupling of mRNA stability to DNA replication.
Detailed studies on deletion mutants suggest that the
histone nascent polypeptide is not involved in the
posttranscriptional regulatory process (Capasso et al..
1987). Systematic removal of approximately 100 nucleotide
blocks spanning nearly the entire protein coding region had
no effect on the destabilization of these mutant histone
mRNAs when DNA synthesis was inhibited. Consistent with this
result, the deletion of the ATG translation start codon
which prevents the synthesis of histone protein did not
prevent the destabilization of the mRNA during inhibition of
DNA replication. Furthermore, as discussed above, the
minimal target sequences containing the stem-loop structure
confer instability to gamma globin mRNA when DNA synthesis
is interrupted (Pandey and Marzluff, 1987). The globin-
histone chimeric mRNA, although regulated as wild type
histone mRNA, does not encode any histone protein sequences.
Destabilization of histone mRNA appears to occur in a
stepwise manner initiating at the 3' terminus (Ross et al..
1986). The first detectable histone mRNA degradation product
lacks 5 nucleotides from the 3* terminus (Ross and Kobs,
1986; Ross et al.. 1986). This degradation product is
detectable in the in vitro mRNA decay system as well as in

10
whole cells (Ross et al., 1986). It is unclear whether this
product arises from a single endonucleolytic scission or
from multiple cuts removing one or more of the nucleotides
at a time. Degradation continues rapidly in a 3' to 5'
direction carried out by an exonuclease which is part of the
ribosome complex (Ross et al.. 1987). These results are
consistent with the requirement for translation to proceed
near the 3' terminus for the destabilization of the histone
mRNA during inhibition of DNA synthesis.
Subcellular Compartments
The appearance of histone mRNA on polysomes within
minutes after transcription and the equally rapid transfer
of the newly synthesized histone proteins into the nucleus
where they complex with DNA imply that histone protein
synthesis may be compartmentalized in a region located just
outside the nucleus (Spaulding et al.. 1966; Borun et al..
1967; Robbins and Borun, 1967). The asymmetric distribution
of eukaryotic mRNAs in the cytoplasm supports this
possibility. Morphological studies initially demonstrated
that polyribosomes are concentrated in areas surrounding the
nucleus in mouse 3T3 cells (Fulton et al.. 1980). Later,
localization of specific mRNAs by in situ hybridization
techniques provided direct evidence that mRNAs are
differentially distributed throughout the cell. Tubulin mRNA
is preferentially located in the periphery of the cytoplasm

11
whereas vimentin mRNA is concentrated in perinuclear regions
in chick myoblast and fibroblast cells (Lawrence and Singer,
1986). Furthermore, actin mRNA is concentrated in the
lamellipodia, structures involved in cell locomotion
(Lawrence and Singer, 1986). The requirement for actin
protein in the lamellipodia during cell movement (Wang,
1984) suggests a functional relationship between where the
mRNA is localized and where the encoded protein is needed
(Lawrence and Singer, 1986).
Localization of mRNAs in distinct areas of the
cytoplasm may occur due to an association of the mRNA with
an underlying structure. This structure could serve as an
anchor to concentrate mRNAs in particular regions of the
cytoplasm. Consistent with this reasoning is the existence
of a complex, proteinaceous structure in the cytoplasm of
most eukaryotic cells that is referred to as the
cytoskeleton. The cytoskeleton plays a role in cell shape,
cell motility and the intracellular transport of
macromolecules and organelles (Clarke and Spudich, 1977;
Goldman et al.. 1978; and reviewed in Nielsen et al.. 1983).
In addition, there is growing support for the hypothesis
that the cytoskeleton plays a key role in regulating the
translation process (Lenk et al.. 1977; Lenk and Penman,
1979; Fulton et al.. 1980; Cervera et al.. 1981; van Verooij
et al.. 1981; Nielsen et al.. 1983; Howe and Hershey, 1984).

12
The Cvtoskeleton: Microtubules. Intermediate Filaments
and Microfilaments
The cytoskeleton structure is an interconnected system
of three major classes of protein fibers, namely the
microtubules, intermediate filaments and microfilaments.
Each system represents a heterogenous population of
proteins. The microtubules, for example, are composed of 13
protofilaments of the well characterized alpha and beta
tubulin proteins (Amos and Klug, 1974). These filaments are
25 nm in diameter and appear to contain a hollow core.
Several microtubule associated proteins (MAPs) have also
been identified which appear to facilitate the assembly of
the microtubules and promote their interactions with other
cellular organelles (Sloboda et al.. 1976, Bulinski and
Borisy, 1980; Bloom et al.. 1984; Binder et al.. 1985). It
has long been recognized that in mitotic cells the
microtubules reorganize into the mitotic spindle including
the astral, kinetochore and continuous fibers which mediate
the segregation of the chromosomes (Weber et al.. 1975).
During interphase the role of the microtubules is less clear
and somewhat controversial. Studies on agents that cause the
depolymerization of the microtubules, such as colchicine and
colcemid, suggest that the microtubules play a role in cell
shape and the intracellular transport of macromolecules and
organelles (Hayden and Allen, 1984; Vale et al.. 1985).

13
There are at least five major classes of protein that
constitute the intermediate filaments: 1) desmin, 2)
vimentin, 3) keratin, 4) neurofilaments and 5) glial
filaments. Biochemical and immunological studies
demonstrated that these proteins are generally expressed in
a cell"type specific manner (for review see Lazarides,
1980). Desmin is an acidic protein with a molecular weight
of 53,000 Daltons and is predominantly found in intermediate
filaments from cardiac, smooth and skeletal muscle cells
(Lazarides and Hubbard, 1976; Hubbard and Lazarides, 1979).
The localization of desmin at the Z line in muscle suggests
a role for this protein in the coordination of the
contractile actions of the muscle fiber (Lazarides and
Hubbard, 1976; Izant and Lazarides, 1977).
Vimentin is found principally in intermediate filaments
from cells of mesenchymal origin. Immunological and
biochemical characterization of vimentin indicated that the
protein is 52,000 Daltons and widely conserved throughout
evolution (Brown et al.. 1976; Bennett et al.. 1978; Franke
et al.. 1979). A characteristic feature of the vimentin
intermediate filaments is their sensitivity to colchicine
which erroneously led researchers to believe initially that
they were disaggregation products of the microtubules
(Goldman and Knipe, 1973; Holltrop et al.. 1974; Blose and
Chacko, 1976). The collapse of vimentin into filamentous
bundles around the nucleus during colchicine treatment

14
suggests that these filaments are in close association with
the microtubules.
The keratin filaments, also known as tonofilaments, are
predominantly found in epithelial cells. The keratin
proteins are a family of approximately 30 distinct
polypeptides ranging in molecular weight from 40,000-65,000
Daltons (Steinert, 1975; Steinert and Gullino, 1976; Sun and
Green, 1978; Moll et al.. 1982). The keratins can be clearly
divided into acidic and basic groups based on two-
dimensional electrophoretic analysis. Although only a few
keratin subtypes are expressed in any one epithelial cell,
there is always a coordinated synthesis of at least one
class each of the acidic and basic keratin subunits (Moll et
al.. 1982; Woodcock-Mitchell et al.. 1982; Fuchs and
Marchuk, 1983). The combination of keratin subtypes
expressed by an epithelial cell is characteristic of the
tissue origin as well as the differentiation state of the
cell (Roop et al.. 1983; Sun et al.. 1983). In addition,
some epithelially derived tumors express different keratin
subtypes than those found in the normal counterpart and
these differences may be useful in diagnostic tests
(Summerhayes et al.. 1981; Osborn, 1983).
Little is known about the function and organization of
the intermediate filaments from neuronal and glial cells.
The major cytoskeletal element of neuronal axons and
dendrites are the neuronal filaments which are generally

15
composed of three large polypeptides with molecular weights
of 200,000, 150,000, and 68,000 Daltons (Hoffman and Lasek,
1975; Delacourte et al.. 1980). Glial filaments are
specifically found in astrocytes and other types of glial
cells and consist of a single protein, the glial fibrillary
acidic" protein (GFA), with a molecular weight of 51,000
Daltons (Schachner et al.. 1977; Goldman et al.. 1978).
Although the intermediate filament proteins from the
different cell types are heterogeneous in their chemical
composition, they assemble into similar morphological
structures consisting of filaments that are 10 nm in
diameter that are highly resistant to extraction with
detergents. It has been proposed that the intermediate
filaments contain a constant and variable region, analogous
to the structure of antibodies. The constant domain would
conserve the morphology whereas the variable domain would
confer functional specificity (Lazarides, 1980).
The microfilaments are the thinnest of the cytoskeletal
filament systems, with an average diameter of 6 nm, and are
composed primarily of actin protein. It has been estimated
that approximately 5-10% of total cellular protein is actin
which exists in an equilibrium between monomeric and
filamentous forms (Hartwig and Stossel, 1975).
Microfilaments from non-muscle cells contain two predominant
species of actin protein, namely beta- and gamma-actin. The
non-muscle actin proteins are quite similar to muscle

16
specific alpha-actin with respect to molecular weight
(42,000 Daltons), amino acid sequence and polymeric
structure (Tilney and Detmers, 1975? Hartwig and Stossel,
1975; Kane, 1975; Garrels and Gibson, 1976; Whalen et al..
1976; Rubenstein and Spudich, 1977). These results suggest
that beta- and gamma-actin protein in non-muscle cells may
play a role in contraction processes such as cell motility ,
chromosome movement and exocytosis (Berl et al.. 1973;
Huxley, 1973; Sanger, 1975).
Cvtoskeletal-mRNA Interactions
Early studies using stereo-electron microscopic
analysis of whole cells indicated that polysomes are
associated with the cytoskeleton structure (Wolosewick and
Porter, 1977). Closer examination of the cell interior, by
removing membranes and soluble cytoplasmic proteins with
detergent, reveals that many of the ultrastructural features
and nearly all the polysomes are retained on the
cytoskeletal framework (Lenk et al.. 1977). Furthermore, the
non-random topological distribution of the polysomes is
preserved in the cytoskeleton preparation (Fulton et al..
1980). The association of the polysomes with the
cytoskeleton does not appear to be due to trapping since
monosomes and independent ribosomal subunits are efficiently
removed by the detergent extraction (Lenk et al.. 1977;
Cervera et al.. 1981; Ornelles et al. . 1986). These results

17
indicate that actively translated mRNAs are associated with
the cytoskeleton and this association may be requisite for
protein synthesis. Consistent with this possibility, the
inhibition of host protein synthesis coincides with the
release of host mRNA from the cytoskeletal structure and the
attachment of viral mRNA to the cytoskeleton in poliovirus
infected and adenovirus infected cells (Cervera et al..
1981; Lemieux and Beaud, 1982). During development,
maternally inherited mRNAs participate in protein synthesis
only at the time they become associated with the
cytoskeleton in sea urchin oocytes (Moon et al.. 1983).
The region of the polysome involved in its attachment
to the cytoskeleton is not known. Previous results indicate
that the association of polysomal mRNA with the cytoskeleton
is through the mRNA or mRNP particle and not through the
ribosome (Lenk et al.. 1977; Lenk and Penman, 1979; Cervera
et al.. 1981; Lemieux and Beaud, 1982). Disruption of
polysomes by heat shock, high salt, sodium fluoride or
pactamycin treatment releases the ribosomal subunits but not
the mRNA from the cytoskeleton into the soluble phase (Lenk
et al.. 1977; Cervera et al.. 1981; van Venrooij et al..
1981; Howe and Hershey, 1984). Immunofluorescence studies
indicate that the cap-binding protein is associated with the
cytoskeleton suggesting that the attachment of eukaryotic
mRNAs to the cytoskeleton may occur via the 51 mRNA cap
structure (Zumbe et al.. 1982). Earlier studies suggest that

18
the association of eukaryotic mRNAs with the cytoskeleton
may occur through the poly-A tract (Milcarek and Penman,
1974). However, the association of uncapped poliovirus mRNAs
(Bonneau et al.. 1985) and poly-A" histone mRNAs (Jeffery,
1984; Bonneau et al.. 1985) indicates that the attachment
site might also exist in an internal region contained within
the 5' and 3' ends of the mRNA. Isaacs and Fulton (1987)
have presented evidence suggesting that myosin heavy chain
in embryonic chicken muscle cells is co-translationally
assembled into the cytoskeleton. The myosin nascent
polypeptide would therefore serve as an additional
cytoskeleton attachment site for the polysomal myosin mRNA.
Based on these results it appears that several mechanisms
may operate in the attachment of a specific polysomal mRNA
or mRNP to the cytoskeleton and these attachment sites may
vary depending on the species of mRNA.
The component of the cytoskeleton to which the mRNA is
attached is not known although the data accumulated to date
suggest that the microfilaments may be involved in the mRNA-
cytoskeleton association. Removal of tubulin from the
cytoskeleton by using cold calcium-containing buffers during
the isolation procedure has no effect on the attachment of
mRNA with the cytoskeleton; this result suggests that the
microtubules do not play a role in the association of mRNA
with the cytoskeleton (Lenk et al.. 1977; Howe and Hershey,
1984; Ornelles et al.. 1986). Disruption of vimentin

19
filaments with heat shock or indirectly with colcemid has no
effect on protein synthesis as determined by two-dimensional
electrophoresis indicating that translation is not dependent
on the integrity of the intermediate filaments (Welch and
Feramisco, 1985). Other reports suggest that the attachment
of mRNÁ to the cytoskeleton is mediated by the
microfilaments (Lenk et al. . 1977; Howe and Hershey, 1984;
Ornelles et al.. 1986) . Treatment with cytochalasins, either
cytochalasin B or D, to disrupt the actin-containing
microfilaments results in the release of poly-A+ mRNA from
the cytoskeleton into the soluble phase. At least in the
case of cytochalasin D treatment, the release of poly-A+
mRNA from the cytoskeleton occurs in a dose dependent manner
and stoichiometrically reflects the degree of inhibition of
protein synthesis (Ornelles et al.. 1986). These results
indicate that intact microfilaments are required for the
association of mRNA to the cytoskeleton; however, it can not
be concluded that the actin fibers directly bind the mRNA.
Nonmembrane-Bound and Membrane-Bound Polyribosomes;
The Signal Hypothesis
As described above, nearly 100 percent of the actively
translated polysomes are associated with the cytoskeleton.
Studies on vesicular stomatitis virus infected HeLa cells
demonstrated that membrane-bound polysomal mRNA coding for
the viral G glycoprotein and "free" polysomal mRNA coding

20
for the polymerase, nucleocapsid and matrix proteins are all
attached to the cytoskeleton during translation (Cervera et
al.. 1981). The association of free polysomes with the
cytoskeleton has also been demonstrated by electron
microscopy (Wolosewick and Porter, 1979). Free polysomes are
generally the class of polysomes involved in the synthesis
of intracellular proteins whereas membrane-bound polysomes
are generally involved in the synthesis of exported proteins
that are destined for the cell surface or secretion (for
review see Emr et al.. 1980). The association of free
polysomal mRNAs with the cytoskeleton suggests that the
polysome is anchored to a scaffold, albeit a dynamic
structure, and not freely diffusible within the cytoplasm.
The term "free polysome" is operationally defined and is a
misnomer when viewed in light of the biochemical and
microscopic studies on the cytoskeleton. The proper
reference to free polysomes is nonmembrane-bound polysomes.
The mechanism by which mRNAs assemble into membrane-
bound polysomes has been described in the signal hypothesis
(Blobel and Dobberstein, 1975). Many types of signal
peptides exist that target proteins to various subcellular
membrane compartments such as the endoplasmic reticulum,
mitochondria and chloroplasts in eukaryotes as well as the
cell membrane in prokaryotes (Date et al.. 1980; Schatz and
Butow, 1983; Muller and Blobel, 1984; Cline, 1986). The work
presented here is focused on events that occur at the

21
endoplasmic reticulum and on what is generally known for the
synthesis of proteins that are localized on the cell surface
or secreted from the cell. Most cell surface or secreted
proteins are initially synthesized with a segment of amino
acid residues at the N-terminus that is not usually found in
the mature protein. This segment is commonly referred to as
a signal peptide. Although signal peptides are not highly
conserved in sequence, they are fairly conserved in
structure and based on the compilation of 277 signal
peptides ranging from bacteria to higher eukaryotes, a
consensus structure has been deduced (Watson, 1984). In
general, signal peptides are 25-30 amino acids in length.
They contain a charged amino acid within the first five
residues, which is followed by a stretch of at least nine
hydrophobic amino acids, and terminate at a small uncharged
amino acid such as alanine, serine or glycine. It is
believed that the hydrophobic domain of the signal peptide
facilitates the transfer of the protein across the lipid
environment of specific intracellular membranes whereas the
small, uncharged amino acid is thought to direct the
protease to cleave the peptide from the mature protein
(Hortin and Boime, 1981).
Presumably, translation of mRNA encoding exported
proteins is initiated on nonmembrane-bound polysomes.
Following the synthesis of approximately 70 amino acids, the
point at which the entire signal peptide has emerged from

22
the ribosome, the signal recognition complex (SRP) binds to
the signal peptide and arrests translation (Walter and
Blobel, 1981). The SRP is an 11S ribonucleic acid-protein
complex composed of six non-identical polypeptides and one
small cytoplasmic RNA molecule (Walter and Blobel, 1980;
Walter'and Blobel, 1982; Walter and Blobel, 1983). The RNA
molecule of SRP is 300 ribonucleotides in length and shares
approximately 80% homology with a middle repetitive human
DNA family (Ullu et al.. 1982; Li et al.. 1982). With the
SRP bound to the signal peptide, the complex can then
associate with the membrane of the endoplasmic reticulum
through the interaction with the SRP receptor which is also
known as the docking protein (Gilmore et al.. 1982a; Gilmore
et al.. 1982b; Meyer et al.. 1982a; Meyer et al.. 1982b).
The SRP then dissociates from the polysome with the
resultant release of the translational block. As translation
resumes the protein is co-translocated from the cytoplasm
into the endoplasmic reticulum. During translocation, a
signal peptidase which is located in the lumen of the
endoplasmic reticulum removes the signal peptide from the
protein (Maurer and Mckean, 1978; Walter et al.. 1984).
Overview of Project
The overall aim of this work is to better define the
mechanism(s) involved in coupling histone protein synthesis
with DNA replication. Specifically, we have focused our

23
attention on the posttranscriptional regulation of histone
gene expression. The evidence cited above suggests that the
cytoplasm of eukaryotic cells is a highly organized region
which contains subcellular compartments that concentrate
translation factors as well as specific mRNAs (Fulton et
al.. 1980; Howe and Hershey, 1984; Lawrence and Singer,
1986; Ornelles et al.. 1986). Traditionally, the idea of
subcellular compartments refers to organelles such as the
nucleus, endoplasmic reticulum, and mitochondria. For
example, a popular biochemical definition of
compartmentation is "a simple regulatory mechanism in which
the enzymes for catabolism and biosynthesis are physically
separated by membrane-bound cell organelles" (Rawn, 1983) . A
broader definition of subcellular compartmentation includes
any specialized region of the cell regardless of the
presence of a membrane and may be extended to such
components and regions as 1) nonmembrane-bound and membrane-
bound polysomes, 2) cytoskeleton and soluble phase, and 3)
spatial location (i.e., near the nucleus versus near the
plasma membrane).
Histone mRNAs and the factors that are involved in
their selective degradation during the inhibition of DNA
synthesis may be localized in a particular subcellular
compartment. Consistent with this reasoning, the subcellular
compartment may influence the posttranscriptional control of
histone gene expression. Based on this possibility, we have

24
specifically asked the question: Does the subcellular
location of histone mRNA play a role in the destabilization
of histone mRNA during inhibition of DNA synthesis? To
address this question, we have applied a combination of
molecular and cellular approaches first, to determine the
natural location of histone mRNA and second, to alter the
subcellular compartment to study whether the foreign
location affects histone mRNA stability. The rationale for
this approach is that directing histone mRNA to an altered
subcellular location may physically remove the message from
the factors that are involved in the selective degradation
of histone mRNA during inhibition of DNA synthesis.
Our initial studies demonstrated that histone mRNAs are
translated predominantly on nonmembrane-bound polysomes that
are associated with the cytoskeleton. The subcellular
location of the mRNA was altered biologically by
constructing a histone fusion gene that contains sequences
coding for the signal peptide of the beta-lactamase gene of
E. coli pBR322 plasmid DNA. The encoded fusion histone mRNA
was directed to membrane-bound polysomes and was not
destabilized during inhibition of DNA synthesis.
Inactivation of signal peptide function by a single
nucleotide substitution in the ATG translation start codon
of the signal peptide resulted in the association of the
mutated fusion mRNA with nonmembrane-bound polysomes and
restored the coupling of the mRNA stability with DNA

25
replication. These results indicated that the subcellular
location plays a role in the posttranscriptional control of
histone gene regulation.
During the course of studying histone mRNA subcellular
localization it was necessary to examine histone mRNA-
cytoskéletal interactions in detail. Previously, Ornelles et
al♦ (1986) reported that total poly-A+ mRNA, as a
heterogenous population, is released from the cytoskeleton
in HeLa cells by cytochalasin D treatment. We demonstrated
by Northern blot analysis that cytochalasin D treatment of
HeLa cells readily releases histone mRNA, which is not poly-
adenylated, from the cytoskeleton into the soluble phase.
Surprisingly, mRNA coding for the cell surface HLA-B7 class
I antigen remained associated with the cytoskeleton during
cytochalasin D treatment. Subsequent studies demonstrated
that HLA-B7 mRNA expresses at least two distinct
cytoskeletal attachment sites: a cytochalasin D sensitive
site and a puromycin sensitive site. The puromycin sensitive
site appears to be a feature of membrane-bound polysomal
mRNAs and is most likely the nascent polypeptide and/or
ribosome associated with the protein substructure of the
endoplasmic reticulum.
Cytochalasin D and puromycin cotreatment, conditions
that efficiently release HLA-B7 membrane-bound polysomal
mRNA and histone nonmembrane-bound polysomal mRNA from the
cytoskeleton into the soluble phase, did not dissociate the

26
signal peptide-histone fusion mRNA from the cytoskeleton.
The nucleotide sequences of the signal peptide-histone
fusion mRNA encoding the signal peptide contain a
cytochalasin D and puromycin insensitive cytoskeletal
attachment site which is not expressed by histone mRNA or
HLA-B7"mRNA. These results provide evidence for the
heterogeneic pattern for cytoskeleton-mRNA interactions.
Lastly, we investigated the role of the cytoskeleton in
the regulation of gene expression by studying the effects of
disrupting the microfilaments with cytochalasin D on the
transcription and steady state mRNA levels of several genes.
We determined that the transcriptional and
posttranscriptional regulation of the c-fos proto-oncogene
is quite sensitive to the perturbation of the cytoskeleton
structure. Transcription of the c-fos gene is selectively
stimulated with a concomitant increase in c-fos mRNA steady
state levels within minutes of the addition of cytochalasin
D. Consistent with previous results, the transcription of
beta-actin was also elevated in cytochalasin D treated
cells. These findings suggests that the cell monitors the
organization of the cytoskeleton and selectively regulates
gene expression accordingly.

CHAPTER 2
GENERAL METHODS
Materials
[a-32P]dCTP (-3,000 Ci/mmol) , [a-32P]dATP (-3,000
Ci/mmol), [gamma-j2P] ATP (-3,000 Ci/mmol) and [a-j5S]dATP
(>1,000 Ci/mmol) radionucleotides were purchased from
Amersham, Arlington Heights, IL; polyvinylsulfonic acid
(PVS), polyethylene glycol 8,000 (PEG), XAR-5 x-ray film
were obtained from Eastman Kodak Co., Rochester, NY;
ultrapure electrophoresis grade agarose and Zeta-probe nylon
membranes were from Bio Rad, Richmond, CA; formaldehyde
solution (37% w/w) was from Fisher Scientific, Fair Lawn,
NJ; nitrocellulose was from Schleicher and Schuell, Keene,
NH; RPI scintillator ppo-popop concentrated liquid
scintillator (Liquifluor) was from Research Products
International, Elk Grove Village, IL; formamide was from
Bethesda Reasearch laboratories (BRL), Gaithersburg, MD;
acrylamide, calf intestine alkaline phosphatase, nuclease
SI, EL. coli DNA polymerase I large fragment (Klenow enzyme)
and most restriction endonucleases were obtained from
Boehringer Mannheim Biochemicals (BMB), Indianapolis, IN; T4
DNA ligase was from New England Bio Labs (NEB), Beverly, MA;
27

28
T4 polynucleotide kinase (cloned) was from United States
Biochemical Corp. (USB), Cleveland, OH; cytochalasin D,
polyoxyethylene sorbitan monopalmitate (Tween 40), sodium
deoxycholate (NaDOC), sodium dodecyl sulfate (SDS), Triton
X-100, and DNase I (DN-EP) were purchased from Sigma, St.
Louis,"MO; lysozyme was purchased from Worthington
Biochemicals, Freehold, N.J.; fetal bovine, calf bovine and
horse sera were obtained from Gibco Laboratories, Grand
Island, NY or Flow Laboratories, McLean, VA; cycloheximide
was kindly provided by Upjohn, Kalamazoo, Michigan.
Conditions for Enzyme Reactions
The buffers and incubation conditions for each of the
restriction endonucleases and DNA modifying enzymes used in
this study were as described by the specific recommendations
of the manufacturers unless otherwise indicated.
Mammalian Cell Culture
HeLa Monolayer Cell Cultures
HeLa S3 suspension cells (human cervical carcinoma cell
line; ATCC ccl 2.2) were seeded into EMEM (Eagles-modified
minimum essential medium) containing 5% fetal calf serum, 5%
horse serum, 100 U/ml penicillin, 100 nq/ml streptomycin and
1 mM glutamine at 3x10° cells per 10 cm tissue culture dish
and incubated at 37°C under 5% C02. Cells were maintained at
sub-confluent densities by splitting the cultures 1:10 into
fresh medium every 3 to 5 days.

29
HeLa Suspension Cell Cultures
HeLa S3 cells were grown and maintained in suspension
at 3-6xl05 cells/ml in SMEM (Joklik-modified minimum
essential medium) supplemented with 7% calf serum, 100 U/ml
penicillin, 100 ¡iq/ml streptomycin and 1 mM glutamine in a
warm room at 37°C. Monoclonal and polyclonal HeLa S3 cell
lines, which were isolated during G418 selection of
transfected monolayer cultures, were maintained in SMEM
containing 5% fetal calf serum, 5% horse serum, 100 U/ml
penicillin, 100 jLtg/ml streptomycin and 1 mM glutamine at 3-
6xl05 cells/ml.
WI-38 Monolayer Cell Cultures
WI-38 cells (normal, human diploid fibroblast-like
cells; ATCC ccl 75) at passage 28, were grown as monolayers
in EMEM containing 10% fetal calf serum, 100 U/ml
penicillin, 100 ¿¿g/ml streptomycin and 1 mM glutamine at
37°C under 5% C02. Cell cultures were made quiescent by
maintaining cultures in EMEM containing 0.5% fetal calf
serum, 100 U/ml penicillin, 100 /itg/ml streptomycin and 1 mM
glutamine at 37°C under 5% C02 for 48 hours.
Metabolic Inhibitor Treatments
Inhibition of DNA synthesis was usually performed with
1-5 mM hydroxyurea for 30 to 60 minutes. The drug was always
prepared fresh in IX spinner salts (Gibco) as a 100 mM stock
solution.

30
Protein synthesis was inhibited with 10 /¿g/ml
cycloheximide for varying lengths of time. Cycloheximide
stock solutions [500 ¿¿g/ml] were prepared fresh in IX
spinner salts. Protein synthesis was also inhibited with 0.4
mM puromycin for varying lengths of time. Puromycin stock
solutions [54.4 mg/ml] were prepared in dimethyl sulfoxide
(DMSO) and stored at -20°C.
To study the dependence of protein synthesis on the
destabilization of histone mRNA during the inhibition of DNA
synthesis, we pre-treated cells with 10 ng/ml cycloheximide
for 5 minutes and then added 1 mM hydroxyurea for an
additional 30 minutes.
Disruption of the cytoskeleton was generally achieved
by treating the cells with 5 to 60 /xg/ml cytochalasin D for
15-20 minutes at 37°C. The cytochalasin D stock solutions
[5-10 mg/ml] were prepared in DMSO and stored at -20°C.
Plasmid DNA
The human histone genes used in this study were
previously isolated by Sierra et al. from a lambda Ch4A
bacteriophage library (1982). The histone genes were
subsequently subcloned into Escherichia coli pBR322 plasmid
DNA and propagated in E^. coli strain HB101 bacterium (Sierra
et al.. 1982; Plumb et al.. 1983? Stein et al.. 1984).
Recombinant plasmids ST519 and FF435C each contain a
distinct genomic DNA fragment encoding a human H3 histone

31
gene. The recombinant plasmid F0108A contains an entire
human H4 histone gene which includes the promoter, coding
and 3' non-transcribed regions.
An HLA-B7 cDNA clone, pDPOOl, was generously supplied
by Dr. Sherman Weissman, Yale University, New Haven,
Connecticut (Sood et al.. 1981). The pDPOOl plasmid DNA was
constructed from the human cell line RPMI 4265 and contains
a 1400 bp insert coding for HLA-B7 in the PstI site of
pBR322 plasmid DNA. Therefore, the plasmid confers
tetracycline resistance and not ampicillin resistance to E.
coli HB101 bacteria.
The phCGa plasmid DNA was generously provided by Dr.
John Fiddes, California Biotechnology Inc., Mountain View,
California and Dr. John Nilson, Case Western Reserve
University, Cleveland, Ohio. The phCGa plasmid is a cDNA
clone containing a 621 bp fragment coding for the alpha
subunit of human chorionic gonadotropin inserted into the
Hindlll site of pBR322 and confers ampicillin resistance
(Fiddes and Goodman, 1979).
The beta-lactamase signal peptide-globin fusion gene,
pSP125e, was kindly provided by Dr. Vishwanath Lingappa,
University of California at San Francisco. The pSP125e clone
contains the signal peptide from the beta-lactamase gene of
E. coli plasmid pBR322 fused in frame to the coding sequence
for chimpanzee alpha globin. This fusion gene was subcloned
into the pSP64 vector (Promega Biotech) as a Hindlll-PstI

32
fragment and is under control of the SP6 promoter (Simon et
al.. 1987).
To study the regulation of histone mRNA localized on
membrane-bound polysomes, we constructed a beta-lactamase
signal peptide-histone fusion gene as outlined in Figure 2-
1. The"first intermediate, pSPpSTdeltaHH/E, involved the
placement of the coli pBR322 beta-lactamase signal
peptide-globin fusion gene from pSP125e under
transcriptional control of the cell cycle-dependent H3
histone regulatory region of pST519deltaH. The 840 bp
EcoRI/Hindlll fragment from pST519deltaH (Marashi et al..
1986) was isolated electrophoretically and blunt-ended using
the Klenow fragment of DNA polymerase I. The signal peptide-
globin fusion construct was digested with Bglll restriction
endonuclease, which cuts once in the untranslated leader
region of the fusion gene, blunt-ended with Klenow fragment
and dephosphorylated using calf intestinal alkaline
phosphatase. The 840 bp blunt-ended fragment from pST519 was
ligated into the Bglll blunt-ended site of pSP125e using T4
DNA ligase. The construction of the second intermediate,
pSPpSTdeltaHH/EpST, involved the substitution of the
chimpanzee globin coding sequences in pSPpSTdeltaHH/E with
the H3 histone coding sequences from pST519. The
pSPpSTdeltaHH/E DNA was digested with Ncol, which cuts at
the signal peptide-globin coding region junction, and
treated with calf intestinal phosphatase. The pST519 DNA was

Figure 2-1. Outline of the cloning scheme for the
construction of the signal peptide-histone fusion gene.
Restriction endonuclease sites are as follows: B,
BglII; E, EcoRI; H, HindIII; N, Ncol; P, PstI; S, Smal.
Abbreviations: S.P., E_¡_ coli pBR322 beta-lactamase signal
peptide coding sequences; H3REG, transcriptional H3 histone
regulatory region including CAAT and TATAA consensus
sequences; CAP, histone H3 mRNA transcription start site; H3
struct, sequences coding for H3 histone protein; ATG,
translation start codon; TAA, translation stop codon; SP6,
bacteriophage SP6 promoter.

34

35
digested with Ncol, which cuts at the ATG translation start
codon, and ligated to Ncol/dephosphorylated pSPpSTdeltaHH/E
DNA using T4 DNA ligase. The signal peptide-histone fusion
gene was subcloned into pUC8 plasmid DNA by ligating the 1.1
kbp Hpal fragment from pSPpSTdeltaHH/EpST into the Hindi
site of pUC8 to form pSPH3. The signal peptide-histone
fusion gene in pSPH3 was transcriptionally enhanced by
placing the SV40 viral enhancer element from pSVE108A
(Shiels et al.. 1985) into the EcoRI site in the polylinker
of pSPH3 to form pSPH3El (containing one enhancer element)
and pSPH3E2 (containing 2-3 enhancer elements).
DNA Isolation and Purification
Plasmid DNA Preparations
Plasmid DNAs used in the experiments described here
were harbored in Escherichia coli strain HB101 bacteria (F~,
recA 13) unless otherwise indicated. Bacteria containing the
appropriate plasmid were grown in medium containing
antibiotics (usually 25-50 ng/ml ampicillin or tetracycline)
as specified by the genotype of the plasmid or bacterium.
The following procedure is the routine method for
isolating pBR322 derived plasmid DNAs and includes an
amplification step with chloramphenicol for higher yields of
DNA (Maniatis et al.. 1982a). Plasmids that are derived from
pUC8 DNA are also prepared as follows; however, due to their
high copy number the amplification step is unnecessary. The

36
standard protocol for the isolation of pUC8 related plasmid
DNAs is the same as described below except that the 500 ml
culture is inoculated directly and allowed to grow overnight
into the late log stage.
Starter cultures were prepared by inoculating 5 ml LB
medium"(Luria-Bertani medium: 1% Bacto tryptone, 0.5% Bacto
yeast extract, 1% NaCl) supplemented with 0.1% glucose with
25-50 ul of bacterial glycerol stock and incubated overnight
at 37°C. The 5 ml starter culture was transferred to 500 ml
pre-warmed LB medium supplemented with glucose and incubated
at 37°C until the optical density at 590 nm (O.D.5go) of the
culture reached 0.55 units. Plasmid DNA was amplified by
treating the culture with 200 ng/ml chloramphenicol
(prepared as a 20 mg/ml stock solution in 95% ethanol and
filter sterilized) at 37°C for 12-16 hours. The bacteria
were centrifuged at 5,000 rpm, 4°C for 10 minutes in a
Beckman JA-10 rotor and J2-21 centrifuge. The cell pellet
was resuspended in 10 ml solution I (50 mM glucose, 25 mM
Tris pH 8.0, 10 mM EDTA, 5 mg/ml lysozyme) and incubated at
room temperature for 5 minutes. The cells were lysed at 0°C
with 20 ml solution II (1% SDS, 0.5 M NaOH) for 5 minutes.
The cell lysate was treated with 15 ml 5 M potassium
acetate, pH 4.8 and kept on ice for 5 minutes. The lysate
was transferred to a 50 ml polycarbonate tube and
centrifuged at 16,000 rpm for 20 minutes at 4°C in Beckman
JA-20 rotor to pellet the bacterial DNA and cell debris. The

37
supernatant was transferred to a 150 ml Corex tube (Corning)
and nucleic acids were precipitated from solution with 30 ml
isopropanol at room temperature for 15 minutes. The DNA was
pelleted by centrifugation at 5,000 rpm in Beckman JS-7.5
rotor for 30 minutes at 20°C. The pellet was briefly dried
under vacuum and resuspended in 8 ml 10:1 TE buffer (10 mM
Tris pH 8.0, 1 mM EDTA). The DNA solution was adjusted to a
final concentration of 1 mg/ml of CsCl (refractive index of
1.3860) and 600 fig/ra 1 ethidium bromide. The DNA solution was
then transferred to a 16 x 76 mm Quick-Seal tube (Beckman)
and centrifuged to equilibrium at 45,000 rpm in a Beckman
Type-50 Ti rotor and Beckman L5-65 ultracentrifuge for 36-48
hours at 20°C. The supercoiled plasmid DNA was removed from
the cesium chloride gradient by puncturing the side of the
tube just below the DNA band with a syringe containing a 22
gauge needle. The DNA was extracted repeatedly with 1-
butanol to remove the ethidium bromide and to reduce the
volume of the solution. Cesium chloride was removed by
dialysis against 10:1 TE buffer at 4°C. The DNA was
quantitated by absorbance at 260 nm and stored at -20°C.
Routinely, the yield for pBR322 derived plasmids was 500-800
í¿g DNA per 500 ml amplified culture.
M13 Bacteriophage DNA Preparations
M13 is a filamentous DNA bacteriophage that is specific
for male E^ coli bacteria. M13 phage DNA is useful for site

38
directed mutagenesis and dideoxy DNA sequencing reactions
since it can be readily isolated as a double-stranded
covalently closed circular DNA replicative form (RF) from
the infected bacteria and as a circular, single-stranded DNA
genome (SS) from the mature phage particle. In addition, M13
phage DNA has been altered genetically to facilitate cloning
and subsequent manipulations. One such derivative, M13mpl8,
is small in size, maintained in high copy number
(approximately 200 RF molecules per cell), contains a
convenient polylinker for the insertion of foreign DNA (up
to 25 kb), and expresses beta-galactosidase which is
inactivated by the insertion of DNA into the polylinker
site.
Single-stranded templates. Single-stranded M13 DNA
templates were prepared from individual "plaques" which
formed on a lawn of EL. coli strain JM103 bacteria
(delta(lac-proAB). supE / F'[traD36, proA+. proB^. laclq,
lacZdeltaM151) that were infected with M13 bacteriophage or
transfected with M13 double stranded, covalently closed
circular DNA. A single plaque was removed from the agar
plate with a sterile pasteur pipette and placed in 2 ml YT
medium (0.8% Bacto-tryptone, 0.5% Bacto-yeast extract, 0.5%
NaCl). Two drops of an overnight JM103 bacterial culture
were added to the phage and incubated at 37°C with agitation
for 6 hours. Longer incubation times result in deletions

39
within the phage DNA. The cell culture (1.5 ml) was
transferred to a microcentrifuge tube and spun in an
Eppendorf centrifuge at 12,000 x g for 10 minutes at 4°C.
The supernatant (1.2 ml) was transferred to a sterile
microcentrifuge tube, 300 ul of 20% PEG, 2.5 M NaCl were
added and the sample was incubated at room temperature for
15 minutes. The remaining 200-300 ul of supernatant were
removed and stored at -2 0°C as a phage stock. The phage
suspension was centrifuged at 12,000 x g for 5 minutes at
room temperature and the supernatant was decanted. The
pellet was resuspended in 150 ul TES buffer (20 mM Tris-HCl
pH 7.5, 10 mM NaCl, 0.1 mM Na2EDTA). The sample was
extracted with phenol as follows: 1) added 50 ul phenol and
vortexed for 2 seconds; 2) incubated at room temperature for
5 minutes; 3) vortexed for 2 seconds; and centrifuged at
12,000 x g for 4 minutes at room temperature. DNA was
precipitated from 130 ul of aqueous phase with 5 ul 3 M
sodium acetate and 340 ul 95% ethanol on crushed dry ice for
15 minutes and collected by centrifugation at 12,000 x g,
for 15 minutes at 4°C. The supernatant was discarded and the
pellet was washed in 70% ethanol at room temperature. The
sample was centrifuged at 12,000 x g, room temperature for 5
minutes. The DNA pellet was dried under vacuum, resuspended
in 80 ul TES buffer and stored at -20°C.

40
Double-stranded templates. E. coli strain JM101
(delta(lac-proAB) . supE / F' [groA+, proB*. laclq,
lacZdeltaM15]) bacteria were grown overnight in 5 ml YT
medium at 37°C with agitation. The overnight culture (1 ml)
was transferred to 100 ml prewarmed YT medium and incubated
at 37°C until the O.D. 590 reached 0.4 units. To assure
exponential growth the culture was again diluted 1:100 into
500 ml prewarmed YT medium and grown to an O.D.5go of 0.4
units. The culture was infected with 500 ul of M13
bacteriophage supernatant which had been prepared from an
isolated plaque as described above for the preparation of
M13 single-stranded DNA. The infected cells were incubated
at 37°C with agitation for no more than 8 hours. The cells
were harvested and RF M13 DNA was isolated as described
above for the preparation of plasmid DNA. Routinely, the
yield was 300 ^g of supercoiled M13 DNA per 500 ml culture.
Isolation of Mammalian RNA
Isolation of Total Cellular RNA
The following procedure for the isolation of total
cellular RNA was developed by Dr Mark Plumb in the Stein's
laboratory primarily for work with HeLa cells. However, this
procedure has been successfully used without modification
for the isolation of total cellular RNA from human HL-60
cells, mouse 3T3-L1 cells, rat osteoblast primary cell
cultures, and Trypanosome cruzii cells.

41
Exponentially growing HeLa cells (5 x 107 cells) were
harvested by centrifugation at 1,500 rpm for 5 minutes at
37°C in an IEC rotor. The cells were washed in phosphate-
buffered saline (PBS; 150 mM NaCl, 10 mM sodium phosphate pH
6.8) and resuspended in 3 ml lysis buffer (2 mM Tris-HCl pH
7.4, l"mM EDTA) containing 5 nq/ml PVS. Cells were lysed in
the presence of 2.4% SDS and 88 /¿g/ml of proteinase K at
room temperature for 15 minutes. Lysates were adjusted to
300 mM NaCl, extracted with 5 ml Tris-saturated phenol and 5
ml CHClj/isoamyl alcohol (IAA) (24:1, v/v) and centrifuged
at 3,000 rpm in an IEC rotor at room temperature. The
aqueous phase was then extracted with 5 ml CHC13/IAA and
centrifuged at 3,000 rpm in an IEC rotor at room
temperature. The phenol/chloroform and chloroform
extractions were repeated as described above. Nucleic acids
were precipitated from the aqueous phase with 2.5 volumes of
95% ethanol at -20°C in the presence of 53 mM potassium
acetate.
Nucleic acids were recovered from the ethanol
suspension by centrifugation at 12,000 x g and resuspended
in TCM buffer (10 mM Tris-Hcl, pH 7.4, CaCl2, 10 mM MgCl2)
for digestion with DNase I. The DNase I was dissolved in 20
mM Tris-HCl pH 7.4, 10 mM CaCl2 at 1 mg/ml, preincubated at
37°C for 20 minutes and treated with 0.1 volume of
proteinase K [1 mg/ml] at 37°C for 2 hours to remove
ribonuclease activity (Tullis and Rubin, 1980). The nucleic

42
acid samples were then digested with 0.1 mg/ml of proteinase
K treated DNase I for 20-30 minutes at 37°C. Following DNase
I digestion, 0.05 volumes of 5 M NaCl and 0.25 volumes of
10% SDS were added to each sample and the RNA solutions were
repeatedly extracted with phenol and chloroform until the
interface between the aqueous and organic phases was clean.
RNA was precipitated from the aqueous phase with 2.5 volumes
of 95% ethanol, at -20°C overnight. The RNA was collected at
12,000 rpm in a JA20 rotor at 4°C for 30 minutes. The pellet
was resuspended in double distilled water, quantitated by
optical density at 260 nm and stored at -20°C. Typical
yields of RNA were 400 /¿g per 90% confluent 10 cm plate of
HeLa S3 cells and 1 mg per 5 x 107 HeLa S3 cells with an
O.D. 260/O.D. 280 of 1.8 to 2.0.
Isolation of Nonmembrane-bound, Membrane-bound and Total
Polvsomal RNA
The subcellular fractionation scheme used in these
studies is a modified procedure described by Venkatesan and
Steele (1972) and is outlined in Figure 2-2. Exponentially
growing HeLa cells (5 x 108 cells) were collected by
centrifugation at 1,500 rpm in an IEC rotor at 37°C for 5
minutes. All subsequent procedures were carried out at 4°C.
Cells were washed several times in PBS and resuspended in 25
ml RSB (10 mM Tris-HCl pH 7.4, 10 mM NaCl, 2.5 mM MgCl2) .
After incubation on ice for 20 minutes, the cells were

43
disrupted with 15 strokes in a tight fitting Dounce
homogenizer. The homogenate was then centrifuged at 2,000 x
g for 5 minutes. The supernatant (supernatant A) was removed
and maintained in an ice bath while the nuclear pellet
(pellet A) was resuspended in 9 ml TMN (20 mM Tris-HCl pH
7.5, 5"mM MgCl2, 25 mM NaCl, 3 mM dithiothreitol) . The
nuclear suspension (pellet A) was adjusted to 1% Triton X-
100 and 1% NaDOC, vortexed briefly and centrifuged at 500 x
g for 10 minutes. The supernatant (supernatant C) from this
centrifugation step was removed and saved on ice for the
isolation of nucleus-associated rough endoplasmic reticulum
(NRER) polysomes. The supernatant from the centrifugation
step following homogenization (supernatant A) was
centrifuged at 12,000 rpm, 4°C for 10 minutes in a Beckman
JA20 rotor. The resulting supernatant (supernatant B) was
removed onto ice for subsequent isolation of light
endoplasmic reticulum (LER)-associated polysomal RNA and
nonmembrane-bound polysomal RNA. The pellet (pellet B) was
resuspended in 9 ml TMN, adjusted to 1% Triton X-100 and 1%
NaDOC and centrifuged at 10,000 rpm, 4°C for 10 minutes in a
JA20 rotor. The supernatant (supernatant D) was removed and
stored on ice for the isolation of mitochondria-associated
rough endoplasmic reticulum (MRER) polysomal RNA.
Exponentially growing HeLa cells (2.5 x 108 cells) were
collected by centrifugation for the isolation of total
polysomal RNA. The cell pellet was washed in PBS and

Figure 2-2. Outline of procedure for the isolation of
subcellular polvsomal fractions.

45
HeLa Cells
homogenize in RSB
centrifuge 2,000xg, 4°C, 5'
I
supernatant A
pellet A
centrifuge ll,400xg
10'
r
supernatant B pellet B
Triton/NaDOC
centrifuge 500xg
10'
discard pellet
Triton/NaDOC
centrifuge 7,800x
10'
I I
discard pellet supernatant D supernatant C
centrifuge
200,000xg, 2 hrs
2M sucrose
centrifuge
200,000xg, 2 hrs
2M sucrose
centrifuge
200,000xg, 2 hrs
2M sucrose
Band at pellet pellet pellet
interface
LER
Nonmembrane-bound
NRER
MRER

46
resuspended in 15 ml TMN. The sample was adjusted to 1%
Triton X-100 and 1% NaDOC, disrupted with several strokes in
a tight fitting Dounce homogenizer and centrifuged at 12,000
x g at 4°C for 10 minutes in a JA20 rotor. The supernatant
was removed and saved at 0°C.
Each of the fractions (NRER, MRER, LER, nonmembrane-
bound, and total polysomal RNA) were layered onto a 3 ml 2 M
sucrose pad (prepared in RSB containing 100 uM spermidine)
and centrifuged at 4°C for 2 hours at 48,000 x g in a
Beckman Type 50.2 Ti rotor (200,000 x g) without the brake.
The LER formed a cloudy band at the top of the sucrose pad
and was carefully removed by aspiration. The remaining
fractions were pelleted under these conditions and gently
resuspended in 3 ml TMN buffer. The samples were then
extracted extensively with phenol and chloroform as
described above for the isolation of total cellular RNA. The
RNA was precipitated from solution with 3 volumes of 95%
ethanol at -20°C in the presence of 53 mM potassium acetate.
The RNA was digested with DNase I, extracted with phenol and
chloroform, and ethanol precipitated as described above for
the preparation of total cellular RNA. Typical yields were
1.5-2.0 mg MRER and NRER RNA, 2-3 mg nonmembrane-bound RNA,
1 mg LER RNA and 4-6 mg total polysomal RNA per liter of
exponentially growing HeLa cells.

47
Isolation of Cvtoskeleton and Soluble Phase RNA
The cytoskeleton and soluble fractions were prepared as
previously described by Penman and coworkers (Cervera et
al.. 1981). Exponentially growing HeLa cells (5 x 108 cells)
were harvested and washed in cold PBS. The cell pellet was
resuspended in 2 ml extraction buffer (10 mM Pipes pH 6.8,
100 mM KC1, 2.5 mM MgCl2, 0.3 M sucrose, 1 mM
phenylmethylsulfonyl fluoride) and Triton X-100 was
immediately added to a final concentration of 0.5%. The
sample was incubated at 0°C for 3 minutes and then
centrifuged at 700 rpm in an IEC centrifuge for 3 minutes at
4°C. The supernatant (SOL) was removed and stored on ice.
The pellet was resuspended in 8 ml RSB containing 1 mM
phenylmethylsulfonyl fluoride, 1% Tween 40 and 0.5% NaDOC.
The nuclei were stripped of cytoplasmic tags with 15 strokes
in a precision-bore stainless steel homogenizer with a
clearance of 0.002 inches and were pelleted at 2,000 rpm for
3 minutes at 4°C in an IEC rotor. The supernatant (CSK) was
removed and saved at 0°C for subsequent RNA isolation. Both
the CSK and SOL fractions were extracted in the presence of
1% SDS and 0.3 M NaCl with phenol and chloroform. Nucleic
acids were precipitated from the aqueous phase with 3
volumes of 95% ethanol at -20°C in the presence of 53 mM
potassium acetate. The RNA was digested with DNase I, phenol
and chloroform extracted, and ethanol precipitated as
described above for the isolation of total cellular RNA.

48
Typical yields of RNA were 250-300 ¿xg from CSK and 100-150
Mg from SOL per 2.5 x 107 HeLa cells (50 ml exponentially
growing suspension culture).
Two-Dimensional Gel Electrophoresis and
Immunoblotting Analysis
The cytoskeleton and soluble fractions were isolated as
described above and each sample was dialyzed in 2 mM Tris-
HC1 pH 7.5 and treated with DNase I (20 iiq/ml) and RNase A
(20 iiq/ml) in the presence of 0.5 mM MgCl2 at 4°C for 1
hour. The samples were dialyzed against 2 mM Tris-HCl pH 7.5
and then lyophilized. Two-dimensional gel electrophoresis
and immunoblotting analyses were performed by Dr. Warren
Schmidt (Department of Pathology, Vanderbilt University,
Nashville, TN). The protein preparations were focused in the
first dimension using equilibrium conditions as described by
O' Farrell (1975) with pH 3 to 10 or 5 to 7 ampholines. The
second dimension was electrophoresed on slab gels containing
a 3% stacking gel and 7.5% resolving gel as described by
Laemmli (1970). The proteins were then either stained with
Coomassie Brilliant Blue or electrotransferred to
nitrocellulose filters as described by Towbin et al. (1979).
Where indicated, immunotransfer analyses (Glass et al..
1981) were performed with commercially available or other
(Schmidt et al.. 1982) antibodies to actin, tubulin and
prekeratin.

49
Northern Blot Analysis
Agarose-Formaldehyde RNA Denaturing Gel Electrophoresis
RNA samples were resolved electrophoretically in 1.5%
agarose-6% (w/v) formaldehyde horizontal slab gels and
running buffer (20 mM MOPS (morpholinepropane-sulfonic acid)
pH 7.0", 5 mM sodium acetate, 1 mM EDTA, 3.7% (w/v)
formaldehyde) as described by Rave et al. (1979). Routinely
10 nq RNA per sample were dried in a Savant Speed Vac
concentrator and resuspended in 3.2 ul double distilled
water (ddH20) , 5 ul formamide, 1 ul lOx MOPS (200 mM MOPS pH
7.0, 50 mM sodium acetate, 10 mM EDTA), 0.8 ul 37% (w/v)
formaldehyde and 2 ul sample buffer (0.5 ml formamide, 0.1
ml lOx MOPS, 80 ul 37% (w/v) formaldehyde and 0.32 ml 0.2%
bromophenol blue-90% glycerol). The samples were then heated
at 60°C for 10 minutes, quick chilled to 0°C and loaded onto
the gels. The 1.5% agarose-6% formaldehyde gels were
prepared by melting 3 g electrophoresis grade agarose in 144
ml ddH20 plus 20 ml lOx MOPS buffer in a microwave oven
(Whirlpool) at full power for 3-4 minutes. The agarose
solution was cooled to approximately 65°C (warm to the
touch) while stirring and then adjusted to 6% formaldehyde
with 36 ml 37% (w/v) formaldehyde. For minigels, 35 ml
agarose-formaldehyde solution was cast in a plexiglass gel
tray (10 cm x 6.3 cm) with an 8 well comb (1 mm x 5 mm per
well). The gel was electrophoresed at 40-45 milliamps until
the bromophenol blue migrated 9 cm (approximately 1.5

50
hours). For full size RNA gels, 200 ml agarose-formaldehyde
solution was cast in a plexiglass gel tray (24.5 cm x 19.8
cm) with 2: 20 well combs (1 mm x 6 mm per well). The gel
was electrophoresed at 100-120 volts until the bromophenol
blue migrated 11 cm (approximately 4 hours). The
formaldehyde gels were stained for 20 minutes in ethidium
bromide solution (1 nq/m1 ethidium bromide, 0.1 M ammonium
acetate) and destained for at least 2 hours in double
distilled water. RNA samples were then visualized by
fluorescence under ultraviolet light using a short
wavelength transilluminator (UVP, inc.; San Gabriel, CA).
The gels were usually photographed at f4.5 for 1/4 second
with Type 57 high speed Polaroid film and a Polaroid model
545 camera mounted on a stand above the transilluminator.
Transfer of RNA from Aqarose-Formaldehvde Gels to
Hybridization Filters
RNA samples were transferred to nitrocellulose (0.45
urn) or nylon membrane hybridization filters in 20x SSC as
previously described by Thomas (1975). The gels were soaked
in 20x SSC (3 M NaCl, 0.3 M sodium citrate pH 7.0) for 20
minutes and were not stained with ethidium bromide. The
hybridization filters were uniformly wetted in ddH20 and
then 2Ox SSC. The gel was placed with the wells facing
downward onto a sheet of 3 MM paper (Whatman) and sponge
which had been soaked in 2Ox SSC. Strips of used X-ray films

51
were placed on areas of the sponge that were not covered by
the gel so as to direct the flow of 2Ox SSC only through the
gel. The filter was directly layered onto the gel and any
air bubbles that were trapped between the gel and filter
were carefully removed. The transfer set-up was completed by
placing two sheets of 3 MM paper followed by 2-3 boxes of
facial tissues (Kleenex brand) on top of the hybridization
filter. The buffer chamber was filled with 20x SSC and
transfer was carried out for 36 hours (note: complete
transfer is usually achieved within 14 hours). After
transfer, the filters were rinsed in 2x SSC for 10 minutes
to remove particles of agarose, air dried and baked in vacuo
at 80°C for 2 hours. The filters were stored in sealed
hybridization bags at 4°C.
Hybridization of Filters-Immobilized RNA (Northern Blot)
Filters (nylon or nitrocellulose) were prehybridized
for 3-4 hours in 0.5 ml hybridization buffer per cm2 filter
area at 43°C for the detection of chorionic gonadotropin
alpha, HLA-B7 and c-fos mRNA and at 48°C for histone mRNA.
Hybridization buffer consisted of 50% formamide, 5x SSC, lOx
Denhardt's (lOOx Denhardt's: 2% (w/v) ficoll 400, 2% (w/v)
polyvinylpyrrolidone), 1% SDS, 50 mM sodium phosphate pH
7.0, 20 /zg/ml BSA, and 250 /xg/ml E_¡_ coli DNA. The filters
were hybridized at the appropriate temperature for 36 hours
in 0.1 ml hybridization buffer per cm2 filter containing

52
lxlO6 cpm/ml of each thermally denatured, j2P-radiolabeled
DNA probe. After hybridization, the filters were washed with
5x SSC, lx Denhardt's (0.5-1 ml/cm2 filter) for 10-15
minutes at room temperature to remove unbound probe. The
filters were then washed, in order, with the following
solutions (0.5 ml/cm2 filter area): 1) 5x SSC, lx
Denhardt's; 2) 2x SSC, 0.1% SDS; 3) lx SSC, 0.1% SDS; and 4)
O.lx SSC, 0.1% SDS at 60°-65°C for 30 minutes per wash.
After washing the filters were briefly air dried and then
exposed to preflashed Kodak XAR5 x-ray film at -70°C with a
Cronex Lightning Plus intensifying screen (Dupont) for
varying lengths of time. Hybridization was quantitated by
scanning laser densitometric analysis of multiple exposures
of autoradiographs that were within the linear range of the
film.
SI Nuclease Protection Analysis
SI nuclease protection analysis was performed
essentially as described by Berk and Sharp (1977) and was
used in the work presented here to detect simultaneously
both endogenous H3 histone mRNA and signal peptide-histone
fusion mRNA in the same RNA sample. The SI probe was
prepared by digesting 30 jug pSPH3El, pSPH3Elalpha, or
PSPH3E1ATG" DNA with 3 0 units of Smal at 3 0°C for 3 hours.
The 450 bp Smal fragment was isolated from a 0.8% agarose
minigel by the freeze-squeeze method. The solution

53
containing the Sinai fragment was butanol extracted to reduce
the volume and to remove the ethidium bromide. The fragment
was then extracted with phenol and chloroform and
precipitated in 2.5 volumes 95% ethanol at -20°C overnight.
The DNA was collected by centrifugation at 12,000 x g at 4°C
for 15 "minutes and resuspended in 5 ul 10 mM Tris-HCl pH
8.0, 5 ul lOx CIP buffer (0.5 M Tris-HCl pH 8.0, 10 mM
MgCl2, 1 mM ZnCl2, 10 mM spermidine) and 3 8 ul ddH,0. The
blunt-ended DNA was dephosphorylated with 0.12 units calf
intestinal alkaline phosphatase (CIP) at 37°C for 15 minutes
and then at 55°C for 15 minutes. A second aliquot of CIP was
added and the incubations at 37° and 55°C were repeated. The
reaction was terminated with 40 ul ddH20, 10 ul lOx STE (100
mM Tris-HCl pH 8.0, 1 M NaCl, 10 mM EDTA) and 5 ul 10% SDS
and heated at 68°C for 15 minutes. The sample was extracted
with phenol and chloroform, adjusted to 0.25 M sodium
acetate and precipitated with 2.5 volumes of 95% ethanol at
-20°C overnight. The dephosphorylated Smal fragment was
pelleted at 12,000 x g, 4°C for 15 minutes, vacuum dried and
resuspended in 12 ul lx kinase buffer (66 mM Tris-HCl pH
9.5, 10 mM MgCl2, 10 mM beta-mercaptoethanol, 2 mM
spermidine). The SI fragment was phosphorylated with 50 uCi
[gamma-32P]ATP and 1 ul polynucleotide kinase [30 units/ul]
at 37°C for 30 minutes. The reaction was terminated with 0.5
ul 100 mM EDTA at 70°C for 10 minutes. The DNA solution was
diluted with 1 ml elutip low salt buffer (20 mM Tris-HCl pH
7.4, 0.2 M NaCl, 1 mM EDTA) and passed over a Schleicher and

54
Schuell Elutip column. The column was washed with 5 ml low
salt buffer to remove unincorporated radiolabeled
ribonucleotide and the DNA fragment was eluted in 400 ul
high salt buffer (Tris-HCl pH 7.4, 1 M NaCl, 1 mM EDTA). The
radiolabeled SI probe was usually used immediately and any
excess"was stored in high salt buffer at 4°C.
Routinely, 10-25 /xg RNA (30 ul) and 10 ul SI probe were
co-precipitated with 100 ul 95% ethanol at -20°C overnight.
The sample was centrifuged at 12,000 x g, 4°C for 15-30
minutes. The pellet was vacuum dried and resuspended in 10
ul 5x hybridization buffer (0.2 M Pipes pH 6.4, 2 M NaCl, 5
mM EDTA) and 40 ul recrystallized formamide. The sample was
heated at 90°C for 10 minutes and then incubated at 55°C for
3 hours. Single-stranded nucleic acids were digested with SI
nuclease by adding 400 ul ice cold SI nuclease buffer (0.03
M sodium acetate pH 4.6, 0.25 M NaCl, 1 mM ZnSOJ and 900
units SI nuclease at 37°C for 30 minutes. The SI digestion
reaction was terminated by organic extraction with 0.5
volume phenol and 0.5 volume chloroform and then 0.5 volume
chloroform alone. Nucleic acids were precipitated from the
aqueous phase with 1 ml 95% ethanol at -20°C overnight.
The SI nuclease digested samples were collected at
12,000 x g, 4°C for 30 minutes. The pellets were vacuum
dried and resuspended in 5 ul loading buffer (80% formamide,
lx TBE [lOx TBE: 500 mM Tris-HCl pH 8.3, 500 mM boric acid,
10 mM EDTA], 0.1% bromophenol blue, 0.1% xylene cyanol). The

55
samples were denatured by heating at 100°C for 2 minutes and
then quickly chilled at 0°C to prevent reannealing prior to
loading onto the gel. Six percent polyacrylamide-8.3 M urea
(cross linked at 20:1 acrylamide:bis-acrylamide) gels with
dimensions of either 16 cm x 18 cm x 1.5 mm (1.5 mm x 6 mm
15 well comb or 1.5 mm x 4 mm 20 well comb) or 32.5 cm x 44
cm (0.5 mm x 8 mm 25 well comb) were prepared and pre-
electrophoresed for 30 minutes. The samples were loaded with
drawn-out capillary tubes or a Rainin Pipetteman (0-20 ul
capacity) using flat pipette tips. The large gels were
electrophoresed at 50 watts constant power (-1200 V, -45 A)
for 3-4 hours and the small gels were electrophoresed at 300
V (25-35 A) for 2-3 hours. The gels were placed on 2 sheets
of 3 MM Whatman paper, dried under vacuum at 80°C for 1-2
hours and exposed to pre-flashed XAR5 x-ray film with a
Cronex Lightning Plus intensifying screen at -70°C for
varying lengths of time. SI analysis was quantitated by
scanning laser densitometry of multiple autoradiographic
exposures that were within the linear range of the film.
In Vitro Nuclear Run-On Transcription Analysis
The in vitro nuclear run-on transcription assays were
performed by Anna Ramsey-Ewing as described by Baumbach et
al. (1987). Cells were harvested by centrifugation and the
cell pellet washed twice in cold isotonic buffer (125 mM
KC1, 30 mM Tris-HCl pH 7.9, 5 mM MgCl2, 10 mM beta-

56
mercaptoethanol). Cells were disrupted by homogenization (6-
12 strokes) with a Dounce homogenizer, Wheaton type A
pestle. After >90% of the cells had been lysed, nuclei were
pelleted by centrifugation at 2000 rpm for 10 minutes in an
IEC centrifuge at 4°C and resuspended in nuclei storage
buffer"containing 40% glycerol (50 mM Tris-HCl pH 8.3, 5 mH
MgCl2, 0.1 mM EDTA). Nuclei were aliquoted and either snap
frozen in liquid nitrogen or used fresh in the ¿n vitro
transcription reactions. Reactions typically contained 107
nuclei, 100 uCi alpha-2P-UTP (3000 Ci/mmole), 1 mM ATP,
0.25 mM GTP and CTP in a final volume of 130 ul and were
incubated for 30 minutes with intermittent shaking at 30°C.
Radiolabeled RNAs were isolated by treatment of nuclei with
DNase I (100 ¿xg/ml) in the presence of 0.6 M NaCl, 50 mM
Tris-HCl pH 7.5, 2 0 mM MgCl2 for 15 minutes at room
temperature. The mixture was then incubated with proteinase
K (200 nq/ml) for 30-60 minutes at 37°C in the presence of
150 mM NaCl, 12.5 mM EDTA, 100 mM Tris-HCl pH 7.5 and 20 mM
MgCl2. Sodium acetate (pH 5.5) was added to 0.2 M and
nucleic acids extracted several times by the hot phenol
method (Clayton and Darnell, Jr., 1983; Sherrer and Darnell,
1962) . To the aqueous solution of j2P-labeled RNAs, 150 ¿ig
of yeast RNA and 2.5 volumes of 95% ethanol were added.
Precipitation was overnight at -20°C. Radiolabeled
transcripts were resuspended in 10 mM Tris-HCl pH 8.0, 1 mM
EDTA and an aliquot of each sample was precipitated with 150

57
/Ltg yeast RNA and cold 10% TCA. TCA-precipitable counts were
determined by liquid scintillation spectrometry.
DNA excess hybridization has also been described by
Baumbach et al. (1987). In a preliminary experiment we
determined that 2 nq of H4 histone insert in pF0002 was at
least á two-fold DNA excess when hybridized to in vitro
nuclear run-on transcripts. Southern blots containing
restriction endonuclease digested DNA fragments or slot
blots containing linearized plasmid DNAs were prepared.
Southern blots on nitrocellulose or slot blots on Nylon were
prehybridized in 1 M NaCl, 20 mM Tris-HCl pH 7.4, 2 mM EDTA,
0.1% SDS, 5x Denhardt's, 250 /ng/ml coli DNA and 12.5 mM
sodium pyrophosphate at 65°C for at least 6 hours.
Hybridizations were conducted at 65°C for 72 hours in 1 M
NaCl, 20 mM Tris-HCl pH 7.4, 2 mM EDTA, 0.1% SDS, 2.5x
Denhardt's, 250 ^g/ml E_¡_ coli DNA with 32P-labeled
transcripts at 5x105-1X10° TCA-precipitable counts per ml of
hybridization solution. Blots were washed at 65°C for 15
minutes in fresh prehybridization solution [0.13-0.14
ml/cm2], 1 hour in 2x SSC/0.1% SDS [0.5 ml/cm2], overnight
in 2x SSC/0.1% SDS [0.5 ml/cm2] and 1 hour in 0.2x SSC/0.1%
SDS [0.5 ml/cm2]. Air dried filters were autoradiographed
with XAR-5 or Cronex film and Cronex lightning Plus screens
at -70°C for varying periods of time and developed with an
X-O-MAT X-ray film processor.

58
Radiolabelinq DNA for Northern Blot Analysis
DNA used for Northern blot analysis was uniformally
labeled by nick translation as described by Maniatis et al.
(1982b) or by the random priming method of Roberts and
Wilson (1985). Radiolabeled probes prepared in this manner
are also suitable for other purposes such as Southern blot
analysis.
Random Priming Method
This protocol is very efficient for labeling small to
intermediate-sized DNA fragments (100-2000 bp in length) and
is the preferred method for work with purified DNA inserts.
Supercoiled plasmid DNA is labeled much less efficiently by
the random priming method and should be fragmented by
sonication prior to use. Routinely, specific activities of
lxlO9 cpm/ng DNA are achieved using as little as 20-100 ng
DNA. Material for oligolabeling DNA fragments was purchased
from Bethesda Research Laboratories as a kit.
The isotope (50 uCi [a-32P] dCTP) was vacuum dried and
resuspended in 10 ul of 2.5x reaction buffer (500 mM Hepes
pH 6.6, 12.5 mM MgCl2, 25 mM beta-mercaptoethanol, 125 mM
Tris-HCl pH 8, 1 mg/ml BSA, 7.5 mg/ml synthetic DNA
oligodeoxynucleotide primers, 50 uM each dGTP, dATP, dTTP).
The DNA fragment (100 ng) in 14 ul ddH20 was heated at 100°C
for 2 minutes and immediately chilled on ice. The denatured
DNA was added to the 1.5 ml microcentrifuge tube containing

59
the isotope and buffer and the priming reaction was
initiated with 5 units Klenow enzyme [5 U/ul]. The sample
was incubated at room temperature for at least 2 hours
(usually overnight). The reaction was terminated with 1 ul
250 mM EDTA and 175 ul TE buffer (10 mM Tris-HCl pH 8, 1 mM
EDTA) and extracted once with 200 ul chloroform/isoamyl
alcohol. Unincorporated radiolabeled deoxynucleotide was
removed from the DNA by chromatography over a (10 mm x 100
mm) Biogel A-15m (Bio Rad) column in TE buffer. Fourteen
fractions were collected; 800 ul in the first fraction and
200 ul in each subsequent fraction. Fractions containing
radiolabeled DNA were determined by Cerenkov counting in the
tritium channel with a Beckman LS-230 scintillation counter.
The peak fractions were pooled and accurately quantitated by
counting 2 ul DNA in 5 ml Triton-toluene based scintillation
cocktail (2000 ml toluene, 1066 ml Triton X-100, 134 ml
liquifluor) in the '2P channel. The samples were immediately
used for hybridization studies and excess probe was stored
at 4°C.
Nick-Translation Method
Nick-translation was the common procedure for
uniformally labeling DNA fragments or supercoiled DNA before
the development of the random primer method. A disadvantage
to using the nick-translation procedure is that the system
must be optimized for efficiency for each lot of E^_ coli DNA

60
polymerase I enzyme. In addition, the specific activity (1-2
x 103 cpm/nq DNA) achieved with the nick-translation
procedure is much lower than the random primer method (1 x
103 cpm//ig DNA) .
DNase I [1 mg/ml in 10 mM HC1] was activated by
incubating 1 ul of the enzyme in 9 ul DNase buffer (10 mM
Tris-HCl pH 7.5, 5 mM MgCl2, 1 mg/ml BSA) at 0°C for 2
hours. The isotope (40-80 uCi [a-“2P] dCTP) was dried in a
Savant Speed Vac centrifuge. The following components were
added, in order, to the isotope : 1) 12.25 ul ddH20; 2) 2.5
ul lOx nick-translation buffer (500 mM Tris-HCl pH 7.8, 100
mM beta-mercaptoethanol, 50 mM MgCl2) ; 3) 5 ul lOx dNTPs
(100 uM of each dATP, dGTP, dTTP); 4) 1.25 ul BSA [1 mg/ml];
5) 2.5 ul DNA [100 ng/ul]; and 6) 0.5 ul DNA polymerase I [5
U/ul]. The activated DNase I was diluted with 990 ul ice
cold ddH20 (1:1,000 dilution; 1 ng/ul final concentration)
and 1 ul of diluted enzyme was added to the nick-translation
mix. The sample was incubated at 14°C for 45 minutes and the
reaction was then terminated with 20 ul 250 mM EDTA and 155
ul TE buffer. The sample was extracted once with 200 ul
chloroform and unincorporated radiolabeled
deoxyribonucleotides were removed by chromatography over a
Biogel A-15m column. The DNA was quantitated and immediately
used for hybridization.

Site-Directed Mutagenesis
In the past, researchers had to rely on nature to
61
produce the proper mutations to study how the normal
biological processes function. Zoller and Smith (1983) have
designed a powerful method for selectively introducing
single nucleotide mutations into specific sites of any
cloned, seguenced region of DNA (site-directed mutagenesis).
In addition, this procedure can be used to selectively
create deletion or insertion mutations in DNA fragments of
known sequence. The general approach of site-directed
mutagenesis is to incorporate an oligonucleotide containing
the appropriate mutation into the complementary strand of
recombinant M13 DNA. The heteroduplex is transfected into
bacteria and the resulting mutated bacteriophage are
isolated by various selection processes.
As discussed in Chapter 4, the signal peptide-histone
fusion mRNA is associated with membrane-bound polysomes and
is stable during inhibition of DNA synthesis. It was unclear
whether the uncoupling of the histone fusion mRNA stability
from DNA replication was due to the altered subcellular
location or to a change in mRNA structure, due to the
nucleotide sequence coding for the signal peptide. To
address these possibilities, we have introduced point
mutations into the signal peptide coding region with the
intention to inactivate signal peptide function. In doing
so, mRNA structure is perturbed as little as possible and

62
the mutated histone fusion mRNA, like endogenous histone
mRNA, associates with nonmembrane-bound polysomes.
Subcloning of the Signal Peptide-Histone Fusion Gene
into Ml3 DNA
The signal peptide-histone fusion gene (SPH3) was
subcloned into M13mpl8 RF DNA. SPH3 DNA was digested with
EcoRI and Hindlll and the 1100 bp fragment was isolated from
a 0.8% agarose gel. The M13mpl8 DNA was digested with EcoRI
and Hindlll and the 7196 bp fragment was isolated from a
0.8% agarose gel. The 1100 bp SPH3 DNA fragment (~50 ng) and
7196 bp M13 DNA fragment (-50 ng) were ligated in 10 ul lx
ligation buffer (50 mM Tris pH 7.6, 10 mM MgCl2, 20 mM
dithiothreitol, 1 mM ATP and 50 /¿g/ml BSA) with 0.004 U T4
DNA ligase (NEB) at 22°C overnight. The ligation mix was
transfected into competent JM101 bacteria, white plaques
were isolated, and M13/SPH3 recombinants were screened by
restriction endonuclease digestion analysis. Single-stranded
M13/SPH3 DNA was isolated from bacteriophage particles as
described above.
Synthesis and Processing of Oligonucleotides
The signal peptide encoded by the SPH3 histone fusion
gene was mutated by substituting leucine at amino acid
position 10 and proline at position 12 with histidine
residues. It was predicted that the introduction of two
positively charged amino acids into the hydrophobic region

63
of the signal peptide would disrupt its recognition by the
signal peptide recognition particle and would therefore,
prevent the translocation of the mutated histone fusion mRNA
to membrane-bound polysomes.
An oligonucleotide (SPa) containing two point mutations
was synthesized by Tom Doyle in the Department of
Microbiology and Immunology, University of Florida,
Gainesville, Florida. The oligonucleotide was composed of
the following sequence: 5' CAAAAAAGTGAATATGGGCGA 3', with
the mutated nucleotides underlined. The oligonucleotide was
received in -2.5 ml 37% ammonium hydroxide and was heated at
55°C for 6 hours to remove the protecting amine group. The
sample was dried under a gentle stream of air in the fume
hood and resuspended in 1 ml ddH20. The sample was extracted
two times with 600 ul n-butanol and the oligonucleotide was
precipitated from the aqueous phase by adding 200 ul 0.5 M
ammonium acetate and 2 ml 95% ethanol and chilling at -70°C
for 30 minutes. The sample was centrifuged at 12,000 x g at
4°C for 30 minutes, washed in 80% ethanol and resuspended in
200 ul ddH20. The yield of oligonucleotide was estimated
spectrophotometrically and was found to be 2.64 mg
(^26o/^280 2.0).
A second oligonucleotide (ATG) was designed to destroy
the signal peptide ATG translation start codon of the SPH3
fusion gene and thereby completely block the synthesis of
any signal peptide. The SPH3ATG mRNA retains the histone

64
translation start codon in its proper context, although
further downstream from the 5' terminus than endogenous H3
histone mRNA, and should therefore be translated on
nonmembrane-bound polysomes. In addition, SPH3ATG’ mRNA
differs from SPH3 wild type signal peptide-histone fusion
mRNA at only one nucleotide. The sequence of the ATG"
oligonucleotide is as follows: 5' AATACTCAAACTCTTCC 3', and
was prepared and processed as described for the SPa
oligonucleotide. The yield of ATG" oligonucleotide was only
445 |¿g (A260/A280 1.41) .
Optimization of Conditions for Site-Directed Mutagenesis
It was necessary to optimize the conditions for the
hybridization of the oligonucleotide to M13 template so that
the synthesis of the second strand was primed by the
oligonucleotide from the correct site. Proper priming by the
oligonucleotide was tested by primer extension analysis.
After limited primer extension, the samples were digested
with a restriction endonuclease that recognized a site
downstream from the expected priming site. The number and
intensity of the fragments produced by the digestion
indicated the number of priming sites and the location of
the preferred site.
The oligonucleotide (20 pmol) was phosphorylated with 5
units T4 polynucleotide kinase in 100 mM Tris-HCl pH 8.0, 10
mM MgCl2, 5 mM dithiothreitol, 20 uCi [gamma-^P] ATP at 37°C

65
for 45 minutes. The reaction was terminated by heating at
65°C for 10 minutes, and unincorporated radiolabeled
nucleotides were removed by chromatography through a
Sephadex G-25 (Pharmacia) column (10 mm x 100 mm) in 50 mM
ammonium bicarbonate (pH 7.8).
The phosphorylated oligonucleotide (10-30 molar excess
over template DNA) was added to 0.5 pmol of M13 recombinant
DNA and the volume was adjusted to 10 ul with ddH20 and 1 ul
solution A (200 mM Tris-HCl pH 7.5, 100 mM MgCl2, 500 mM
NaCl, 10 mM dithiothreitol). The solution was heated at 55°C
for 5 minutes and then cooled for 5 minutes at room
temperature. The primer extension reaction also contained 1
ul of 2.5 mM dCTP, 1 ul 2.5 mM dATP, 1 ul 2.5 mM dTTP, 1 ul
2.5 mM dGTP, 0.5 ul solution A and 1 unit Klenow enzyme. The
reaction proceeded at room temperature for 5 minutes and was
terminated by heating at 65°C for 10 minutes. The sample was
diluted with 2 ul lOx PstI buffer (100 mM Tris-HCl pH 7.5, 1
M NaCl, 100 mM MgCl2, 1 mg/ml BSA) and 7 ul ddH20 and then
digested with 1 ul PstI restriction endonuclease [5 U/ul] at
37°C for 1 hour. The sample was heated at 100°C for 2
minutes, quickly cooled to 0°C and diluted with 20 ul SI
loading buffer. The PstI digested samples (10 ul) were
resolved electrophoretically in 5% acrylamide (20:1;
acrylamide:bis), 7 M urea at 300 volts (-35 mA) for 1.5
hours. The gels were dried under vacuum at 80°C for 1 hour
and analyzed by autoradiography. Primer extension analysis

66
revealed a single major band at -80 bp for each molar ratio
of oligonucleotide to template DNA studied. This was the
expected size fragment and suggested that under these
conditions, the oligonucleotide was priming the synthesis of
the second strand from the proper site. If several
predominant bands had been detected (i.e. multiple priming
sites), more stringent annealing conditions would have been
required.
Synthesis of the Mutated Strand
The oligonucleotide (1.3 was phosphorylated in 100
mM Tris-HCl pH 8, 10 mM MgCl2, 5 mM dithiothreitol, 10 uM
ATP with 15 units T4 polynucleotide kinase at 37°C for 45
minutes. The reaction was terminated by heating at 65°C for
10 minutes.
The phosphorylated oligonucleotide (13.3 pmole) was
annealed to M13/SPH3 recombinant DNA in lx solution A by
incubating the sample at 55°C for 5 minutes and then at room
temperature for 5 minutes. After hybridization, the DNA was
diluted with 10 ul solution C (1 ul solution B [0.2 M Tris-
HCl pH 7.5, 0.1 M MgCl2, 0.1 M dithiothreitol], 1 ul 10 mM
dCTP, 1 ul 10 mM dTTP, 1 ul 10 mM dGTP, 0.5 ul 0.1 mM dATP,
1 ul 10 mM ATP, 1.5 ul [gamma-32P] dATP, 1.5 ul T4 DNA ligase
[1 U/ul], 2 ul ddH20) and the synthesis of the second strand
was initiated with 0.5 ul Klenow enzyme [100 U/16.7 ul] at
room temperature. After 5 minutes, 1 ul 10 mM dATP was added

67
and the incubation was continued at room temperature
overnight.
Enrichment for Covalently Closed Double-Stranded DNA
The in vitro synthesis of the entire second strand is
very inefficient due primarily to the large size of the M13
genome. The percent of covalently closed, double-stranded
DNA molecules is therefore, very small in comparison to the
partially double-stranded, incomplete DNA molecules. These
incomplete DNA molecules, when transfected into bacteria,
give rise to wild type plague formation which complicates
the selection process for the mutated clones. Enrichment for
covalently closed, double-stranded DNA molecules by alkaline
sucrose gradient centrifugation greatly facilitates the
selection process.
The DNA was precipitated from the overnight extension
and ligation reaction by adding 30 ul ddH20, 50 ul 1.6 M
NaCl-13% PEG at 0cC for 15 minutes. The DNA was pelleted at
12,000 x g at room temperature for 15 minutes. The pellet
was gently washed in 100 ul cold 0.8 M NaCl-6.5% PEG,
centrifuged at 12,000 x g for 30 seconds and resuspended in
180 ul TE buffer. The sample was denatured with 20 ul 2 N
sodium hydroxide and layered onto a 5%-20% alkaline sucrose
step gradient. The gradient was prepared by adding to a 0.5
in x 2 in centrifuge tube (nitrocellulose or polyallomer; do
not use Beckman Ultra-clear tubes which are incompatible

68
with the alkali conditions), in order, the following
solutions each containing 0.2 N NaOH, 1 M NaCl, 2 mM EDTA:
1) 1 ml 20% sucrose; 2) 1 ml 17.5% sucrose; 3) 1 ml 15%
sucrose; 4) 1 ml 10% sucrose; 5) 5% sucrose. The tube was
incubated at 4°C for 2 hours. The samples were centrifuged
through the alkaline sucrose gradients in a Beckman SW 50.1
rotor at 37,000 rpm for 2 hours at 4°C without the brake.
The samples were recovered from the gradients by puncturing
the bottom of the tube with a 21 gauge needle and collecting
7 drops per fraction in 1.5 ml microcentrifuge tubes
(usually ~60 fractions). Fractions containing covalently
closed, double-stranded DNA were determined by Cherenkov
scintillation counting in the tritium channel, pooled and
neutralized with 1M Tris-citrate pH 5.
Isolation of Mutated M13 Recombinant Bacteriophage
The DNA was transfected into E^. coli JM101 competent
cells, plated onto YT agar and incubated overnight at 37°C.
Individual plaques were picked and bacteriophage were
subsequently prepared. Bacteriophage suspensions (2 ul) were
spotted onto a 0.45 micron nitrocellulose filter and
prehybridized in 6x SSC, lOx Denhardt's, 0.2% SDS (10 ml/100
cm2) at 67°C for 1 hour. The filter was rinsed in 50 ml 6x
SSC for 1 minute at 21°C and hybridized in 10 ml 6x SSC, lOx
Denhardt's containing 1x10° cpm radiolabeled oligonucleotide
at 21°C for 1 hour. The filter was washed three times in 50
ml 6x SSC at 21°C for a total of 10 minutes. The filter was

69
briefly dried under a heat lamp and placed against pre¬
flashed XAR5 x-ray film without a screen for 1 hour at room
temperature. Under such low stringency each phage spot was
positive, including the negative control M13/SPH3 wild type
bacteriophage. The filter was re-washed at 35°C in 6x SSC
for 5 minutes, air dried and exposed to pre-flashed XAR5
film for 1 hour. At the higher wash temperature many of the
phage spots were less intense whereas several remained
unchanged. When the filter was washed at 48°C in 6x SSC for
5 minutes the negative control and nearly all the phage
spots completely disappeared. The bacteriophage (~5%) that
remained positive after the 48°C wash were sequenced to
confirm the presence of the mutations.
DNA Sequencing
The site directed mutations in M13/SPH3 recombinant
DNAs were confirmed by Sanger's dideoxynucleotide chain
• . TU
terminator DNA sequencing procedure using the "Sequenase
DNA Sequencing Kit" from United States Biochemical
Corporation, Cleveland, Ohio. The Sequenase enzyme is
derived from bacteriophage T7 DNA polymerase and has been
modified to efficiently use dideoxyribonucleotides, alpha-
thio deoxyribonucleotides and other nucleotide analogs that
are commonly used for sequencing. In addition, the Sequenase
enzyme expresses high processivity and low 3'-5' exonuclease
activity.

70
Annealing Template and Primer
The primer for synthesizing the complementary strand of
wild type and mutated M13/SPH3 recombinant DNA is a 20 bp
oligonucleotide which hybridizes to a site adjacent to the
polylinker. The annealing reaction was prepared by adding 1
ul primer (0.5 pmol), 7 ul single-stranded M13/SPH3
recombinant DNA (0.5-1.0 pmol) and 2 ul 5X sequencing buffer
(200 mM Tris-HCl pH 7.5, 100 mM MgCl2, 250 mM NaCl) . The
sample was heated at 65°C for 2 minutes and slowly cooled to
room temperature.
Labeling Reaction
The complementary strand was synthesized by incubating
the annealed sample with 1 ul 100 mM dithiothreitol, 2 ul
labeling nucleotide mix (1.5 uM dGTP, 1.5 uM dCTP, 1.5 uM
dTTP) , 0.5 ul [a-35S] dATP [10 uCi/ul] and 2 ul Sequenase
[1.5 U/ul] at room temperature for 5 minutes. Elongation was
terminated by mixing 3.5 ul aliquots of the labeling
reaction with 2.5 ul of one of the dideoxy nucleotide
solution (ddATP, ddCTP, ddGTP, ddTTP). The dideoxy
nucleotide solutions were 80 uM dATP, 80 uM dCTP, 80 uM
dGTP, 80 uM dTTP, 50 mM NaCl and 8 uM of the appropriate
dideoxyribonucleotide. The sequencing samples were diluted
with 4 ul stop solution (95% formamide, 20 mM EDTA, 0.05%
bromophenol blue and 0.05% xylene cyanol), heated at 75°C
for 2 minutes and loaded onto a 6% acrylamide-8.3 M urea gel

71
(32.5 cm x 44 cm) with a 60 well shark toothed comb. The gel
was run at 75 W constant, -2500 V, -40 mA for 3-5 hours
(Xylene Cyanol dye was 15 cm from the bottom of gel). The
gel was soaked in 10% acetic acid-12% methanol in the fume
hood for 1 hour, placed onto 2 sheets of 3 MM Whatman paper,
and dried under vacuum at 80°C for 45 minutes. The gel was
exposed to preflashed XAR5 x-ray film at room temperature
for 12-18 hours.
Transfection of DNA into Cells
Introduction of Plasmid DNA into Bacterial Cells
An L coli overnight cell culture was diluted 1:100
into 100 ml LB medium and incubated at 37°C until the
optical density was approximately 0.37 at 590 nm. The cells
were chilled at 0°C for 10 minutes and then centrifuged at
5,000 rpm, 4°C for 5 minutes in a Beckman JA20 rotor. The
pellet was resuspended in 20 ml of cold CaCl2 buffer (60 mM
CaCl2, 10 mM Pipes pH 7.0, 15% glycerol) and incubated at
0°C for 30 minutes. The cells were collected by
centrifugation at 2,500 rpm for 5 minutes at 4°C in a JA20
rotor, resuspended in 2.6 ml CaCl2 buffer and divided into
100-200 ul aliquots. The competent cells were either used
immediately for transformation or stored in microcentrifuge
tubes at -80°C.
For transformation, 100 ul competent EL_ coli cells were
mixed with 10 ul plasmid DNA (-10 ng) and incubated at 0°C
for 10 minutes. The cells were then heat shocked at 37°C for

72
5 minutes, diluted with 0.9 ml prewarmed LB medium and
incubated at 37°C for 1 hour. Different volumes of the
transfected bacterial cell culture (usually 10-200 ul) were
plated onto the appropriate medium dictated by the genotype
of the host and the introduced plasmid DNA.
Introduction of DNA into HeLa Cells
HeLa monolayer cell cultures were transfected with
plasmid DNA, according to Gorman et al. (1982), in a calcium
phosphate/DNA complex prepared as described by Graham and
van der Eb (1973). HeLa S3 cells growing exponentially in
suspension culture were plated into 20 ml completed EMEM
(EMEM supplemented with 5% fetal calf serum, 5% horse serum,
100 U penicillin, 100 /¿g/ml streptomycin and 1 mM glutamine)
c
at 3x10 cells per 10 cm tissue culture dish (Corning or
Falcon) and incubated overnight at 37°C in 5% C02. The
monolayer cell culture was refed with 10 ml completed EMEM
and incubated at 37°C in 5% C02 for 4 hours. The DNA/calcium
phosphate precipitate was prepared by adding 2 0 fig DNA in
500 ul 250 mM CaCl2 dropwise to 500 ul 2x Hebs buffer (50 mM
Hepes, 280 mM NaCl, 1.5 mM Na2HP04, pH 7.12±0.05) while
vortexing the solution. The DNA/CaCl2 precipitate was added
dropwise to the monolayer cell culture and incubation was
continued at 37°C in 5% C02 for 4 hours. The HeLa cells were
glycerol shocked by replacing the medium with 2 ml completed
EMEM containing 15% glycerol for 1 minute. The transfected

73
cells were then washed with 10 ml EMEM, refed with 20 ml
completed EMEM and incubated at 37°C, 5% C02.
Establishment of Polyclonal and Monoclonal Stably
Transformed Cell Lines
Polyclonal and monoclonal cell lines were isolated by
growing HeLa cell cultures that were co-transfected with
pSV2neo and the plasmid DNA of interest in medium containing
the antibiotic G418. The plasmid pSV2-neo, constructed by
Southern and Berg (1982), carries the bacterial
aminoglycoside phosphotransferase 3' (II) gene under the
control of the SV40 early promoter and confers resistance to
the aminoglycoside antibiotic G418. The rationale behind the
co-transfection method is that HeLa cells that are competent
to take up pSV2-neo DNA are egually competent to take up the
plasmid DNA of interest, and therefore, a certain percentage
of the G418 resistant cells contain both species of DNA.
Ideally, the selectable marker should be carried by the
vector that contains the gene which is to be studied.
HeLa cells in monolayer culture were co-transfected
with the plasmid DNA to be studied and pSV2-neo DNA (molar
ratio of 20:1, respectively) by the calcium phosphate
precipitation method as described above. After 36-48 hours
post-glycerol shock, the transfected cells were resuspended
in 5 ml Puck saline A [0.04% (w/v) KC1, 0.8% (w/v) NaCl,
0.035% (w/v) NaHC03, 0.1% (w/v) glucose] containing 0.02%

74
(w/v) EDTA and 0.5 ml of the cell suspension was seeded into
20 ml completed EMEM containing 500 /¿g/ml biologically
active Geneticin (G418, Gibco) in 10 cm tissue culture
dishes. The cells were incubated at 37°C, 5% C02 and refed
every 3-4 days with completed EMEM containing G418
antibiotic. Within 2-3 weeks, the majority of the cells had
died and individual colonies of G418 resistant HeLa cells
had formed. Polyclonal cell lines were established by
combining the G418 resistant colonies from a 10 cm dish as a
single, heterogeneous cell population (usually 20-40
colonies). For the establishment of monoclonal cell lines,
individual colonies were picked from the dish with sterile
pasteur pipettes and cultured as separate cell lines.

CHAPTER 3
SUBCELLULAR LOCATION OF HISTONE mRNAs ON CYTOSKELETON
ASSOCIATED NONMEMBRANE-BOUND POLYSOMES IN HELA S3 CELLS
INTRODUCTION
Histone biosynthesis occurs primarily during the S
phase of the cell cycle and is temporally as well as
functionally coupled with DNA replication (Allfrey et al..
1963; Prescott, 1966; Spaulding et al.. 1966; Robbins and
Borun, 1967; Stein and Borun, 1972). Early studies using in
vitro translation analysis of polysomal RNA isolated from S
phase HeLa cells provided an initial indication that histone
protein synthesis occurs on non-membrane-bound, light
polysomes (Jacobs-Lorena et al.. 1972; Liautard and
Jeanteur, 1979; Gallwitz and Breindl, 1972; Borun et al♦.
1967). In these studies it was demonstrated that
translatable, polysome-associated histone mRNAs have a
sedimentation coefficient of 8-10S and fluctuate in a cell
cycle-dependent manner. Histone mRNA was not detected by in
vitro translation studies on polyribosomes isolated from
HeLa cells in the G1 phase of the cell cycle (Borun et al..
1967; Borun et al.. 1975; Pederson and Robbins, 1970).
Consistent with this observation, the inhibition of DNA
replication by hydroxyurea or cytosine arabinoside treatment
75

76
resulted in the rapid disappearance of histone mRNAs from
these polysomes (Borun et al.. 1967; Breindl and Gallwitz,
1974; Gallwitz, 1975).
The light polysomes on which translatable histone mRNAs
reside are presumably nonmembrane-bound ("free") polysomes
as predicted by the signal hypothesis (Blobel and
Dobberstein, 1975). Several lines of evidence suggest that
nonmembrane-bound polysomes are not freely diffusible within
the cytoplasm but are attached to the cytoskeleton. Penman
and co-workers (Lenk et al.. 1977) have demonstrated that
nearly 100% of total cytoplasmic polysomes co-fractionate
with the cytoskeletal framework. In addition, vesicular
stomatitis virus encoded nonmembrane-bound and membrane-
bound polysomal mRNAs are associated with the cytoskeleton
in virally infected HeLa cells (Cervera et al.. 1983).
Taken together, these observations provide a basis for
the possibility that histone mRNA-containing polysomes
and/or their subcellular location may play a role in the
regulation of histone gene expression. Such reasoning is
consistent with the appearance of translatable histone mRNAs
on polysomes within minutes following transcription (Borun
et al. . 1967), the equally rapid transfer of newly
synthesized histones into the nucleus where they complex
with DNA (Robbins and Borun, 1967; Spaulding et al.. 1966),
and the degradation of histone mRNAs and cessation of
histone protein synthesis in parallel with the inhibition of

77
DNA replication (Spaulding et al.. 1966; Robbins and Borun,
1967; Borun et al.. 1967; Baumbach et al.. 1984). To address
the relationship between the localization of histone mRNAs
and the control of cellular histone mRNA levels, we have
used cloned human histone genes to examine the distribution
of histone mRNAs on nonmembrane-bound and membrane-bound
polysomes and the association of histone mRNA-containing
polysomes with the cytoskeleton. As predicted from previous
results of in vitro translation analysis and cDNA
hybridization studies, histone mRNAs were found to reside
predominantly on nonmembrane-bound polysomes. Both the
nonmembrane-bound polysomes containing histone mRNAs and the
membrane-bound polysomes containing the cell surface class I
HLA-B7 antigen mRNAs were found associated with the
cytoskeleton.
RESULTS
Localization of Histone mRNA and HLA-B7 mRNA in Subcellular
Fractions
To determine the intracellular locations of histone
mRNA and HLA-B7 mRNA, HeLa cells were first fractionated
into nonmembrane-bound and membrane-bound polysomal
components by the procedure outlined in Fig. 2-2. The
membrane-bound polysomes are represented by the mitochondria
associated rough endoplasmic reticulum (MRER) and the
nucleus associated rough endoplasmic reticulum (NRER)

78
fractions. The light endoplasmic reticulum (LER), which is
deficient in polysomes by definition, was also isolated
following this subcellular fractionation scheme. Histone and
HLA-B7 mRNA content in each of the subcellular fractions was
determined by Northern blot analysis. Briefly, RNA was
extracted from the samples, quantitated and electrophoresed
in agarose under denaturing conditions. The RNA was then
transferred to nitrocellulose filters for hybridization with
3?P-labeled (nick-translated) pFO108A (H4 histone), pFF435C
(H3 histone) and pDPOOl (cDNA clone of HLA-B7 class I
antigen) DNA probes. A typical screening of subcellular
fractions with these probes is represented by the
autoradiograph in Figure 3-1, and Table 3-1 summarizes the
numerical values derived from densitometric analysis of the
Northern blot. The HLA-B7 mRNA and histone mRNA species are
each present in total cellular RNA and total polysomal RNA
which were included as positive controls. Non-membrane-bound
polysomal RNA contains approximately 55% and 51% of the
total H4 and H3 histone mRNA, respectively. Surprisingly,
the LER fraction contains approximately 38% of the H4
histone mRNA and 41% of the H3 histone mRNA, which may be a
consequence of the method by which the fraction is isolated.
The LER and non-membrane-bound polysomal sample is
centrifuged through a 2 M sucrose pad under conditions where
the LER forms a loose band on top of the pad, while the
nonmembrane-bound polysomes are pelleted. The LER band is

79
removed by aspiration and occasionally material is removed
from within the sucrose pad. This material may be the source
of histone mRNA in the LER fraction, which could decrease
the true percentage of total histone message found on the
nonmembrane-bound polysomes. The HLA-B7 mRNA was located
almost exclusively on membrane-bound polysomes. Nearly 90%
of the HLA-B7 mRNA was collectively found on the NRER and
MRER, while less than 11% was measured in the LER and
nonmembrane-bound polysomal RNA fractions.
Histone mRNA and HLA-B7 mRNA Distribution Between Csk and
Sol Fractions
Existing within the cytoplasm of eukaryotic cells is an
intricate network of protein filaments, which is referred to
as the cytoskeleton. This structure can be prepared from
HeLa cells treated with non-ionic detergent as described by
Penman and coworkers (Cervera et alâ– â–  1981). We used this
procedure for the isolation of the cytoskeleton (Csk) and
soluble (Sol) fractions which were subsequently assayed for
mRNA content by Northern blot analysis.
Log phase, suspension grown HeLa S3 cells were
fractionated into a Csk phase and Sol phase as described in
Chapter 2. Although this fractionation procedure has been
characterized previously (Lenk and Penman, 1979), we
examined the protein heterogeneity of each fraction by two-
dimensional gel electrophoresis and immunoblotting
techniques in collaboration with Dr. Warren Schmidt. As seen

Figure 3-1. Northern blot analysis of H3 histone. H4 histone
and HLA-B7 mRNA in subcellular fractions.
The cytoskeleton and soluble fractions, as well as the
nonmembrane-bound and membrane-bound polysomal RNA
fractions, were isolated from HeLa cells by the procedures
described in Chapter 2 (Cervera et alâ– . 1981; Venkatesan and
Steele, 1972). In addition, total polysomal RNA and total
cellular RNA were isolated (Plumb et al. â–  1983, Venkatesan
and Steele, 1972). RNA was extracted from each subcellular
fraction and 50 Mg of RNA from each preparation was resolved
by 1.5% agarose, 6% formaldehyde gel electrophoresis. The
RNA was transferred to nitrocellulose and hybridized to
radiolabeled pFF435C (H3 histone and PFO108A (H4 histone)
probes, followed by hybridization to 32P-labeled (nick-
translated) pDPOOl (cDNA of HLA-B7) probe. The hybridized
filters were then exposed to preflashed XAR5 x-ray film at -
70°C for 72 hours. From left to right, lanes: 1) total
cellular RNA; 2) Csk RNA; 3) Sol phase RNA; 4) NRER; 5)
MRER; 6) nonmembrane-bound polysomal RNA; 7) LER; and 8)
total polysomal RNA.

81
* * ~
a
1 2 3 4 5
&
6 7 8
HLA-B7
H 3
H 4

Table 3-1. Quantity of histone and HLA-B7 mRNAs in the
subcellular fractions.
82
H3
H4
HLA-B7
Cytoskeleton
90%
84%
97%
Soluble
10%
16%
3%
NRER
3%
3%
29%
MRER
6%
5%
60%
Free
51%
54%
5%
LER
40%
38%
6%
The distribution of histone and HLA-B7 mRNAs in the
NRER, MRER, LER and nonmembrane-bound polysomal fractions as
well as the cytoskeleton and soluble fractions was
determined by Northern blot analysis. The values have been
normalized with respect to the total yield of RNA within
each fraction.

83
in Figure 3-2, the cytoskeleton fraction clearly contained
actin as well as the major HeLa cytokeratins (Franke et al..
1981). In addition, vimentin appeared to be enriched in the
cytoskeleton, which was confirmed by immunoblot analysis.
Extraction of the cytoskeleton required a cold, calcium
containing buffer and consequently the tubulins were
solubilized and therefore absent in the Csk fraction. The
tubulin proteins were clearly contained in the soluble
fraction as determined by immunoblot analysis. Additional
support for separation of cytoskeleton and soluble phases
comes from the observation that the bulk of the tRNA species
is in the soluble phase which is in agreement with
previously reported values (data not shown; Cervera et alâ– 
1983). Although the HLA-B7 messenger RNA is found on
membrane-bound polysomes and the histone mRNA on
nonmembrane-bound polysomes, both species of messenger RNAs
are associated almost exclusively with the cytoskeletal
fraction. Nearly 97% of the HLA-B7 mRNA and approximately
85-90% of the H3 and H4 histone messages are retained in the
Csk (Fig. 3-1; Table 3-1).
The Effects of Metabolic Inhibitors on the Subcellular
Localization of Histone and HLA-B7 mRNA
Detectable levels of histone mRNA were observed in each
of the subcellular samples, including the Sol fraction and
the membrane-bound polysomal fractions. However, the HLA-B7

Figure 3-2. Two-dimensional gel electrophoresis of
cvtoskeleton and soluble associated proteins.
HeLa cells were fractionated into a cytoskeleton and
soluble phase, protein was extracted from each phase and
subsequently analyzed by two-dimensional gel electrophoresis
as described in Chapter 2. Proteins of the soluble fraction
(Sol), A, were separated using wide range (pH 3-7)
ampholines while cytoskeletal proteins (Csk) , B, were
separated with narrow range (pH 5-7) ampholines to more
clearly identify cytoskeletal species. V: vimentin, a:
actin, and t: tubulins. Unlabeled arrows in B depict HeLa
cytokeratins of 54 Kd (pi 6.0); 52.5 Kd (pi 6.1); 46 Kd (pi
5.1); and 45 Kd (pi 5.7), which were spots highly reactive
with commercially prepared prekeratin or stratum corneum
antisera (Franke et alâ– . 1981).

7.0
T
5.2
T
4.1
T
85
A)
PH
6.4
l
SOL
B)
6.0 5.4 5.0 4.4
T T T T
43K—
CSK

86
mRNA was located almost exclusively in the cytoskeleton and
membrane-bound polysome fractions. The widespread
subcellular localization of histone mRNA may be due to non¬
specific trapping of nonmembrane-bound polysomes in each of
the fractions during the isolation procedure. Alternatively,
this observation may reflect the presence of mRNAs encoding
variant histone proteins that could be targeted to the
different cytoplasmic compartments. Inhibition of DNA
synthesis with hydroxyurea results in a rapid
destabilization of cell cycle dependent histone mRNAs,
whereas cell cycle independent histone mRNAs remain stable
in the presence of this drug (Baumbach et al.. 1984; Heintz
et al. . 1983; DeLisle et al.â–  1983; Graves and Marzluff,
1984; Plumb et al.. 1983). Conversely, inhibition of protein
synthesis with cycloheximide results in a 2-3 fold increase
in cytoplasmic levels of histone mRNAs (Butler and Mueller,
1973; Baumbach et alâ– . 1984). Differential sensitivity of
histone mRNAs in the subcellular fractions to hydroxyurea or
cycloheximide treatment would favor the latter suggestion
that cell cycle dependent and independent histone gene
transcripts may selectively reside in specific cytoplasmic
compartments.
Treatment of exponentially growing HeLa cells with
hydroxyurea, as described in Chapter 2, resulted in a
complete and selective destabilization of histone mRNA in
the LER, nonmembrane-bound and membrane-bound polysomal

87
fractions (Figure 3-3). This is in sharp contrast to the
HLA-B7 mRNA which appeared to be fully stable under these
conditions. Cycloheximide treatment had a stabilizing effect
on the levels of histone mRNA in each subcellular fraction
(Figure 3-3). When compared with control HeLa cells,
cycloheximide treatment resulted in a 2.1 fold (H4) and a
2.4 fold (H3) increase in the levels of histone mRNA in the
nonmembrane-bound polysomal fraction. Similar increases were
also seen for levels of histone mRNAs in the LER and
membrane-bound polysomal fractions. Inhibition of protein
synthesis with cycloheximide followed by DNA synthesis
inhibition with hydroxyurea resulted in the stabilization of
histone mRNA in each subcellular fraction, whereas the
stability of HLA-B7 mRNA remained unaffected (Figure 3-3).
Similar results were obtained when these inhibitor studies
were expanded to include the cytoskeleton and soluble
fractions (Figure 3-4). Cycloheximide treatment resulted in
a 2-3 fold increase in the representation of H3 and H4
histone mRNA over untreated cultures in both the
cytoskeleton and soluble fractions. HLA-B7 mRNA appeared to
be only moderately stabilized in the presence of this drug.
When cells were treated with hydroxyurea, histone mRNA
levels were greatly reduced in both the cytoskeleton and
soluble fractions whereas the HLA-B7 mRNA was found in
comparable levels to those present in the untreated samples.
Consistent with previous data, cotreatment of HeLa cells

Figure 3-3. Northern blot analysis of H3 histone. H4 histone
and HLA-B7 mRNAs isolated from subcellular fractions of
ceils treated with metabolic inhibitors.
Exponentially growing HeLa cells were incubated in the
presence of metabolic inhibitors and fractionated into
nonmembrane-bound and membrane-bound polysomes as described
in Chapter 2. RNA was extracted from each fraction and
assayed for H3 histone rnRNA, H4 histone mRNA and HLA-B7 mRNA
content by Northern blot analysis as described in Figure 3-1
and Chapter 2. Upper panel from left to right, lanes: 1)
control (untreated), nonmembrane-bound polysomes; 2)
control, NRER; 3) control, MRER; 4) control, LER (omitted);
5) cycloheximide (Cy), nonmembrane-bound polysomes; 6) Cy,
NRER; 7) Cy, MRER; and 8) Cy, LER. Lower panel from left to
right, lanes: 1) hydroxyurea (HU), nonmembrane-bound
polysomes; 2) Hu, NRER; 3) Hu, MRER; 4) Hu, LER; 5) CY/HU,
nonmembrane-bound polysomes; 6) Cy/Hu, NRER; 7) Cy/Hu, MRER;
and 8) Cy/Hu, LER.

89
DC
Ui
CONTROL CYCLO
a. a.
ui m ui
g K *
IL Z S
cc
UI
cc
UI
DC
UI
UI
£ Z S
DC
UI
DC
HLA-B7
H 3
H 4
HYDROXY
CYC/HYDROXY

90
with cycloheximide and hydroxyurea had a stabilizing effect
on histone mRNA levels and no apparent effect on HLA-B7 mRNA
stability in both subcellular fractions. There were no
obvious differences in the stability of histone mRNAs
localized in the various subcellular compartments when
treated with hydroxyurea or cycloheximide. It is therefore
reasonable to conclude that the histone mRNA isolated in the
LER and membrane-bound fractions is a consequence of the
subcellular fractionation procedure. Our results, taken
together with previous findings (Jacobs-Lorena et al.. 1972;
Liautard and Jeanteur, 1979; Gallwitz and Breindl, 1972;
Borun et al.. 1975; Stein et alâ– . 1975) suggest that histone
mRNAs, during normal cellular processing, are targeted to
the nonmembrane-bound polysomes, the common site of histone
protein synthesis.
DISCUSSION
Our objectives in pursuing these studies were two-fold.
The first was to confirm that histone proteins are
synthesized on nonmembrane-bound polysomes. The second was
to determine if the histone mRNA-containing polysomes are
associated with the cytoskeleton. Our rationale for this
approach was to begin addressing the questions whether the
subcellular location of histone containing polysomes
contributes to: a) the ability of histone mRNAs to be
translated immediately following transcription, b) the rapid

Figure 3-4. Northern blot analysis of H3 histone. H4 histone
and HLA-B7 mRNAs associated with the cvtoskeleton and
soluble fractions from HeLa cell treated with metabolic
inhibitors.
Exponentially growing HeLa cells were incubated in the
presence of metabolic inhibitors and fractionated into
cytoskeleton and soluble fractions as described in Chapter
2. RNA was extracted and assayed for H3 histone, H4 histone
and HLA-B7 mRNA content by Northern blot analysis as
described in Figure 3-1. From left to right, lanes: 1)
control, Csk; 2) control, Sol; 3) cycloheximide (Cy), Csk;
4) Cy,"Sol; 5) hydroxyurea (HU), Csk; 6) HU, Sol; 7) Cy/HU,
Csk; and 8) Cy/HU, Sol.

II
*> w
HLA-B7
(
r
(
(
CSK
SOL
CSK
SOL
CSK
SOL
CSK
SOL
UD

93
shuttling of newly synthesized histone polypeptides into the
nucleus and c) histone mRNA stability.
In agreement with long-standing in vitro translation
and cDNA hybridization results (Jacobs-Lorena et al. . 1972;
Liautard and Jeanteur, 1979; Gallwitz and Briendl, 1972;
Borun ét al.. 1975; Stein et al.. 1975), we find that
histone mRNAs reside predominantly on nonmembrane-bound
polysomes. While we cannot dismiss the possibility that the
amounts of histone mRNAs detected in membrane-bound
polysomal subcellular fractions are biologically significant
(see Figures 3-1 and 3-3), we feel that this is a
consequence of the isolation procedure which operationally
defines the fraction obtained. The absence of detectable
HLA-B7 mRNAs on nonmembrane-bound polysomes suggests that
while mRNAs encoding intracellular proteins may be trapped
by the endoplasmic reticulum during this subcellular
fractionation technique, there is little, if any, stripping
of membrane-bound polysomes encoding extracellular
polypeptides, such as the HLA-B7 class I antigen. Within
this context, the question arises whether the transcripts
from histone genes that are not expressed in a cell cycle-
dependent manner (approximately 8-10%) are represented on
membrane-bound polysomes. However, this is not a likely
possibility because inhibition of DNA replication with
hydroxyurea brings about a comparable degradation of histone
mRNAs in all the subcellular fractions examined.

94
The association of histone mRNAs with the cytoskeleton
(Figure 3-1) is consistent with studies carried out by
Penman and co-workers in which they examined the subcellular
localization of vesicular stomatitis virus (VSV) mRNAs in
HeLa cells following viral infection (Cervera et alâ– â–  1981).
These investigators observed that in VSV-infected HeLa cells
both the viral mRNA coding for the G viral glycoprotein and
the nonmembrane-bound polysomal viral mRNAs were associated
with the cytoskeleton. While the specific nature of the
association of polysomes with the cytoskeleton remains to be
definitively established, the ability to release ribosomes
and retain mRNAs on the cytoskeleton suggests that there may
be a direct attachment of mRNAs to the cytoskeleton (Lenk et
al. . 1977; Lenk and Penman, 1979; Lemieux and Beaud, 1982).
It is therefore reasonable to speculate that the association
of mRNAs with the cytoskeleton plays a functional role in
protein synthesis and/or in the regulation of mRNA
stability. This is a particularly attractive possibility
because the association of histone mRNAs with the
cytoskeleton could permit the compartmentalization of
histone mRNA-containing polysomes in a particular region of
the cytoplasm. The potential for compartmentalization
afforded by the cytoskeleton may also be related to the
rapid and selective destabilization of histone mRNAs at the
natural end of S phase or following inhibition of DNA
synthesis.

95
As with the protocol utilized to separate nonmembrane-
bound and membrane-bound polysomes, the cytoskeleton
preparation obtained must be operationally defined. While
this does not necessarily detract from the biological
relevance, differences in cytoskeleton preparations are
undoubtedly reflected in the biochemical composition of the
isolated complex. Additional characterization of the
cytoskeleton preparation with respect to possible variations
in regions of the complex may provide insight into
mechanisms by which specific mRNAs may be localized and
their functional properties selectively modulated.

CHAPTER 4
SUBCELLULAR LOCATION OF HISTONE mRNA PLAYS A ROLE IN THE
POSTTRANSCRIPTIONAL REGULATION OF HISTONE GENE EXPRESSION
Introduction
The association of histone mRNA-containing nonmembrane-
bound polysomes with the cytoskeleton may provide a
structural basis for the localization of the mRNA in
specific regions of the cytoplasm. Consistent with this
reasoning, the factors that mediate the rapid and selective
destabilization of histone mRNA during inhibition of DNA
synthesis may be co-localized with the message in the same
subcellular compartment. Therefore, the subcellular
localization of histone mRNA may play an important role in
its posttranscriptional regulation.
In this chapter, we describe a biological approach,
using a series of signal peptide-histone fusion genes, to
study whether the subcellular location of histone mRNA-
containing polysomes is functionally related to the coupling
of histone mRNA stability with DNA replication. In order to
study the influence of subcellular location on histone mRNA
stability, we constructed a signal peptide-histone gene
which was designed to target the encoded fusion message to
96

97
membrane-bound polysomes. The ability of the cell to
selectively recognize and destabilize histone mRNA in a
foreign subcellular compartment could then be examined.
Based on the data presented here, we propose that the
subcellular location of histone mRNA, with respect to the
class of polysomes in which they are associated with, plays
a significant role in the coupling of histone mRNA stability
with DNA replication.
Results
The construction of the signal peptide-histone fusion
gene is described in Chapter 2 and outlined in Figure 2-1.
The upstream flanking region of the chimeric gene contains
210 base pairs of the 5' regulatory sequences of the cell
cycle-dependent human H3 histone gene pST519 including the
TATAA and CCAAT consensus sequences. This region is followed
by the H3 mRNA cap site and sequences encoding the initial
20 nucleotides of the non-translated H3 histone leader. The
fusion gene is, therefore, under transcriptional control of
the H3 histone gene promoter, and a segment of the H3 leader
which has been implicated in the coupling of histone mRNA
levels to DNA synthesis is present (Morris et alâ– â–  1986).
The H3 histone leader segment is fused to the untranslated
leader sequences of the beta-lactamase signal peptide
including the ATG translation start codon. The H3 histone
structural gene is fused in frame to the signal peptide

98
coding sequences and extends approximately 300 base pairs
beyond the TAA translation stop codon, including a region
which has been implicated in histone mRNA destabilization
when DNA synthesis is inhibited (Luscher et al.â–  1985;
Pandey and Marzluff, 1987; Graves et alâ–  . 1987). An SV40
enhancer element was incorporated into the upstream EcoRI
site in the recombinant plasmid to increase cellular levels
of the fusion transcript. The mRNA encoded by the fusion
gene consists of the first 20 nucleotides of the H3 histone
5' leader sequences followed by the untranslated leader
sequence of the beta-lactamase signal peptide, the entire
beta-lactamase signal peptide coding sequences (including
the translation start codon), the entire H3 histone coding
region which is fused in frame to the signal peptide coding
region, and the H3 histone 3' untranslated region (Figure 4-
1). The junction between the signal peptide coding region
and the H3 histone coding region has been sequenced by
Sanger's dideoxy method and the reading frame has been
conserved. All sequences required for the synthesis and
processing of the chimeric mRNA, as well as for translation
of the fusion protein, are present. The signal peptide-H3
histone fusion gene with a single SV40 enhancer element is
designated pSPH3El and an identical fusion construct
containing multiple SV40 enhancer elements in the EcoRI site
is designated pSPH3E2.

Figure 4-1. Schematic diagram of endogenous H3 histone mRNA
ST519 and the beta-lactamase signal peptide-H3 histone
fusion mRNA.
The signal peptide-histone fusion mRNA is identical to
the wild type histone mRNA with the exception of the signal
peptide encoded sequences that have been inserted in frame
into the histone gene. The 5' mRNA sequences and the 3' mRNA
sequences that have been implicated in coupling histone mRNA
stability to DNA synthesis are retained in place in the
signal peptide-histone fusion mRNA (Morris et alâ– . 1986;
Pandey and Marzluff, 1987). Boxed areas represent histone H3
mRNA sequences, single lined areas represent beta-lactamase
signal peptide derived sequences.

100
CAP ATG TAA
' i i
CAP
I
5' [
atg
TAA
i
3'
pST5l9
ATG
pSPH3E1

101
To test for expression of the signal peptide-histone
fusion gene in human cells, we transfected pSPH3El and
pSPH3E2 DNA into HeLa cell monolayers by the calcium
phosphate precipitation method as described in Chapter 2
(Gorman et alâ– â–  1982; Graham and van der Eb, 1973). Forty-
six hours post-transfection cells were harvested and total
cellular RNA was isolated. The RNA was subjected to SI
nuclease protection analysis by a modification of Berk and
Sharp (1978) . The probe used in the SI nuclease assays was
the Smal fragment from pSPH3El which was radiolabeled at its
5' termini as described in Chapter 2. The probe is
complimentary to endogenous HeLa histone H3 mRNA from the 5'
end-labeled Smal site within the protein coding region to
the signal peptide-histone fusion junction and therefore
protects a 130 nucleotide region of the H3 mRNA from SI
nuclease digestion. As seen in Figure 4-2, when total
cellular RNA from two independent transfections of HeLa
cells with pSPH3El or pSPH3E2 DNA was analyzed by SI
nuclease digestion, a 130 nucleotide fragment corresponding
to endogenous H3 histone mRNA was protected as well as an
approximately 280 nucleotide fragment corresponding to the
signal peptide-histone fusion mRNA species (lanes 3-6) . In
HeLa cells transfected with salmon sperm DNA, only the 130
nucleotide fragment was detected (lanes 1 and 2). These data
demonstrate that the signal peptide-histone fusion gene is
capable of expression in HeLa cells and that the mRNA

Figure 4-2. Expression of the signal peptide-histone fusion
gene in HeLa cells.
The expression of the signal peptide-histone fusion
gene was tested in a short term transient assay and analyzed
by an SI nuclease protection assay. HeLa cell monolayers
were transfected with 20 ng DNA according to Gorman et al.
(1982), in a calcium phosphate/DNA complex prepared as
described by Graham and van der Eb (1973). The transfected
cells were incubated at 37°C, 5% C02 for 46 hours following
transfection. The cells were then harvested, total cellular
RNA was isolated and analyzed by SI nuclease protection
assay (200 ng RNA per sample) as described in Chapter 2.
Lanes 1 and 2, HeLa cells transfected with salmon sperm DNA;
lanes 3 and 4, HeLa cells transfected with pSPH3El DNA; and
lanes 5 and 6, HeLa cells transfected with pSPH3E2 DNA.
Marker lanes (M) are radiolabeled Hinfl digests of pBR322
DNA. Each lane represents RNA isolated from an independently
transfected cell culture.

103
M
<
Z
o

(/)
1
M
HYBRID H3 mRNA
ENDOGENOUS H3mRNA

104
transcribed from this gene is sufficiently stable to be
detected in a short-term transient transfection assay.
To determine the class of polysomes with which the
signal peptide-histone fusion mRNAs are associated, HeLa
cells were transfected with pSPH3El and 46 hours later
nonmembrane-bound and membrane-bound polysomes were isolated
as described in Chapter 2. RNA was isolated from both
classes of polysomes and analyzed by SI nuclease protection
assay using the Smal fragment of the signal peptide-histone
fusion gene (32P-5'-labeled) . As seen in Figure 4-3,
approximately 30% of the signal peptide-histone fusion mRNA
was found in the nonmembrane-bound polysome fraction and
approximately 70% was associated with the membrane-bound
polysomes (lanes 1 and 2 respectively). This is in contrast
to endogenous H3 histone mRNA which is represented by
approximately 90% in the nonmembrane-bound polysome and 10%
in the membrane-bound polysome fraction. The percent
distribution of the histone chimeric mRNA within the
polysomal RNA fractions was determined by densitometric
analysis of the autoradiograms and the values corrected for
the total yield of RNA in each subcellular fraction. While
the extent to which the signal peptide-histone fusion mRNA
is represented on the membrane-bound polysomes varied from
experiment to experiment, we always observed 60-90% of the
fusion mRNA in the membrane-bound polysome fraction (n=6).

Figure 4-3. The subcellular localization of the signal
peptide-histone fusion mRNA.
HeLa cell monolayers were transfected with pSPH3El and
cultured as described in Chapter 2. Forty-six hours after
transfection, cell cultures were harvested and nonmembrane-
bound and membrane-bound polysomes were isolated. The RNAs
from these fractions were analyzed by the SI nuclease
protection assay (5 fjg RNA per sample) . The distribution of
mRNA species within the subcellular fractions was
quantitated by scanning laser densitometric analysis of the
autoradiograms. Lanes: 1) pBR322 Hinfl molecular weight
marker; 2) nonmembrane-bound polysomal RNA; and 3) membrane-
bound polysomal RNA.

106
LU
LU CQ
DC 5
LL ¿
1634
512
344
220
FUSION mRNA

107
The ability to direct a cell cycle-dependent histone
mRNA to membrane-bound polysomes provided the possibility to
address the involvement of subcellular location in the
coupling of histone mRNA stability with DNA replication. As
seen in Figure 4-4 (lanes 1-4), endogenous histone mRNA
levels"were reduced by greater than 90%, as determined by
densitometric analysis of the autoradiogram, in HeLa cells
following inhibition of DNA replication by treatment with 1
mM hydroxyurea for 60 minutes. In contrast, inhibition of
DNA replication by hydroxyurea treatment does not result in
a reduction of fusion message levels on membrane-bound
polysomes (lane 4). The finding that the signal peptide-
histone fusion mRNA is stable following DNA synthesis
inhibition is further supported by the SI nuclease analysis
of total cellular RNA (Figure 4-4). The endogenous histone
mRNA levels in hydroxyurea treated HeLa cells were reduced
by approximately 95% compared with those in untreated
HeLa cells (lanes 5-8) (with each lane representing an
independent transfected cell culture). In contrast, only a
4% reduction in fusion mRNA levels was measured (lanes 5-8).
These results demonstrate that the incorporation of the E.
coli beta-lactamase signal peptide into a human H3 histone
gene is sufficient to target the encoded fusion mRNA to the
membrane-bound polysomes and to confer stability to the
fusion message when DNA synthesis is inhibited.

Figure 4-4. The stability of the signal peptide-histone
fusion mRNA following inhibition of DNA synthesis.
HeLa cell monolayers were transfected with pSPH3El DNA
and cultured as described in Chapter 2. Forty-six hours
after transfection, half of the transfected HeLa cell
cultures were treated with 1 mM hydroxyurea in completed
medium for 1 hour at 37°C, 5% C02. Cell cultures were
harvested and nonmembrane-bound and membrane-bound polysomes
were isolated. Total cellular RNA from both the control and
hydroxyurea-treated samples was also isolated. The RNA was
analyzed by the SI nuclease protection assay as described in
Chapter 2, using 5 /¿g RNA per sample. The distribution of
mRNA species within the subcellular fractions was
quantitated by densitometric analysis of the autoradiograms.
Lanes: 1) control (untreated cells) , nonmembrane-bound
polysomes; 2) control, membrane-bound polysomes; 3)
hydroxyurea, nonmembrane-bound polysomes; 4) hydroxyurea,
membrane-bound polysomes; 5) and 6) control, total cellular
RNA; and 7) and 8) hydroxyurea, total cellular RNA. Lanes 5-
8 represent RNA isolated from independently transfected cell
cultures.

109
M 1
396
344
298
154
130
}ÍUf?ff?
* •
I » I | » I I I
HYBRID H3mRNA
ENDOGENOUS H3 mRNA

110
It was possible that the stability of the histone
fusion mRNA during inhibition of DNA synthesis was due to
the targeting of the mRNA to a foreign subcellular
compartment, namely the membrane-bound polysomes.
Alternatively, the nucleotide sequences coding for the
signal"peptide may have disrupted the histone mRNA structure
in such a way that it was no longer recognized by the
factors involved in the destabilization of histone mRNA. To
test these possibilities, the signal peptide was mutated in
order to retain the fusion mRNA, without significantly
changing the mRNA structure, on nonmembrane-bound polysomes.
If the class of polysomes plays a role in the coupling of
histone mRNA stability with DNA replication then the mutated
fusion mRNAs that are retained on nonmembrane-bound
polysomes should be efficiently destabilized during
inhibition of DNA synthesis. Conversely, if the mRNA
structure has been disrupted by the nucleotide sequences
encoding the signal peptide then the mutated histone fusion
mRNAs localized on nonmembrane-bound polysomes should be
stable during inhibition DNA synthesis.
The signal peptide-histone fusion gene was mutated by
site directed mutagenesis (Zoller and Smith, 1983) as
described in Chapter 2 and the corresponding mRNAs are
schematically diagrammed in Figure 4-5. The first mutant,
pSPH3Elalpha, contains two point mutations in the region
coding for the hydrophobic domain of the signal peptide,

Figure 4-5. Schematic diagram of wild type and mutated
signal peptide-histone fusion mRNAs.
The signal peptide-histone fusion gene (SPH3E1) was
mutated by site directed mutagenesis (Zoller and Smith,
1983) as described in Chapter 2. Abbreviations: H3,
endogenous H3 histone mRNA; SPH3E1, signal peptide-histone
fusion mRNA; SPH3Elalpha, signal peptide-histone fusion mRNA
mutated in the hydrophobic region of the signal peptide (leu
at position 10 and pro at position 12 substituted with his);
SPH3E1ATG , signal peptide-histone fusion mRNA mutated in
the signal peptide translation start codon; CAP, 5' cap
structure; ATG, predicted translation start codon; atg,
internal translation start codon for histone H3 protein;
TAA, translation stop codon.

112
50 nt
CAP ATG atg Sma TAA
□— 1 ' I—n
SPH3E1: ...GGAAGAGU AU6 A6U AUU CAA CAU UUC CGU GUC GCC CUU AUU CCC UUU UUU GCG GCA UUU UGC CUU CCU GUU UUU GCC AUG GCU ...
Met Ser lie Gin H1s Phe Arg Val Al a IH1si 11e|hísIPhe Phe Ala Ala Phe Cys Leu Pro Val Phe Ala Met Ala
Alpha : ...GGAAGAGU AUG AGU AUU CAA CAU UUC CGU GUC GCC[CaUjAUU[CaCjUUU UUU GCG GCA UUU UGC CUU CCU GUU UUU GCC AUG GCU .. .
Met
AGU AUU CAA CAU UUC CGU GUC GCC CUU AUU CCC UUU UUU GCG GCA UUU UGC CUU CCU GUU UUU GCC AUG GCU ...

113
which results in the substitution of leucine at amino acid
position 10 (GAA->GTA) and proline at position 12 (GGG-*GTG)
with positively charged histidine residues. The second
mutant, pSPH3ElATG”, contains a single nucleotide
substitution which destroys the ATG translation initiation
codon of the signal peptide (ATG-+TTG) . This mutation should
result in the initiation of translation at the natural ATG
codon of the histone coding region without the synthesis of
the signal peptide.
The distribution of the SPH3E1, SPH3Elalpha and
SPH3E1ATG" signal peptide-histone fusion mRNAs in
nonmembrane-bound and membrane-bound polysome fractions from
transfected HeLa monolayer cell cultures was determined by
SI nuclease protection analysis. As seen in Figure 4-6, the
SPH3E1ATG mRNA partitioned into the nonmembrane-bound
polysome fraction to the same extent as endogenous histone
H3 mRNA. Approximately 82% of the SPH3E1ATG mRNA and 83%
endogenous histone mRNA was found associated with the
nonmembrane-bound polysomes (Table 4-1). In contrast, SPH3E1
mRNA was predominantly associated with membrane-bound
polysomes (Fig. 4-6). Approximately 68% of SPH3E1 mRNA was
localized in the membrane-bound polysomal RNA fraction
(Table 4-1). The SPH3Elalpha mRNA displayed a more
intermediate association with membrane-bound polysomes;
approximately 40% of the SPH3Elalpha mRNA was localized in
the membrane-bound polysome fraction and 60% in the
nonmembrane-bound polysomal fraction (Fig. 4-6 and Table 4-

Figure 4-6. Distribution of SPH3E1. SPH3Elalpha. and
SPH3E1ATG mRNAs in nonmembrane-bound and membrane-bound
polvsomal fractions.
HeLa cell monolayers were transfected with the signal
peptide-histone fusion genes by the calcium phosphate
precipitation method and 46 hours posttransfection
nonmembrane-bound and membrane-bound polysomes were
isolated. RNA (5 ng) from each subcellular fraction was
assayed for signal peptide-histone fusion mRNA and
endogenous H3 histone mRNA content by SI nuclease protection
analysis as described in Chapter 2. Abbreviations: SPH3E1,
wild type signal peptide-histone fusion mRNA; alpha,
hydrophobic mutant of SPH3E1; ATG, signal peptide
translation initiation codon mutant of SPH3E1. Lane 1,
nonmembrane-bound polysomal RNA; Lane 2, membrane-bound
polysomal RNA.

;15
S P H 3
1 2
ALPHA
1 2
A TG
1 2
probe
fusion
H 3

116
Table 4-1. Quantitation of SPH3E1. SPH3Elalpha and
SPH3E1ATG' mRNA in nonmembrane-bound and membrane-bound
polysome fractions.
Nonmembrane-bound Membrane-bound
SPH3E1 mRNA
32%
68%
SPH3Elalpha mRNA
60%
40%
SPH3E1ATG' mRNA
82%
18%
Endogenous H3 mRNA
83%
17%
The distribution of SPH3E1, SPH3Elalpha, SPH3E1ATG ,
and endogenous H3 histone mRNAs in nonmembrane-bound and
membrane-bound polysome fractions was quantitated by
densitometric analysis of the SI nuclease protection assay
presented in Figure 4-6. The densitometric values obtained
represent the quantity of mRNA in 5 /jg RNA. The values were
then adjusted to reflect the total yield of RNA recovered in
each fraction.

117
1). These results indicate that the mutation of the ATG
translation start codon of the signal peptide results in the
synthesis of a signal peptide-histone fusion mRNA
(SPH3E1ATG ) that is associated with nonmembrane-bound
polysomes. In addition, the incorporation of two histidine
amino ácid residues in the hydrophobic domain of the signal
peptide partially blocks the translocation of the mRNA from
nonmembrane-bound polysomes to membrane-bound polysomes.
To study the stability of the mutated signal peptide-
histone fusion mRNAs during inhibition of DNA synthesis, the
genes (SPH3E1, SPH3Elalpha, and SPH3E1ATG") were transfected
into HeLa cell monolayers by the calcium phosphate
precipitation method as described in Chapter 2. Forty-six
hours posttransfection, the cells were treated with 1 mM
hydroxyurea and samples were taken at 20 minute time
intervals. Total cellular RNA was isolated and assayed for
signal peptide-histone fusion mRNA and endogenous H3 histone
mRNA content by SI nuclease protection analysis as described
in Chapter 2. As seen in Figure 4-7 and summarized in Figure
4-10, inhibition of DNA synthesis for one hour resulted in
the destabilization of only 46% of SPH3E1 mRNA as compared
to the control cell culture (lanes 1,2 and 7,8). In sharp
contrast, SPH3E1ATG" mRNA was destabilized by 94%, which is
to the same extent as measured for endogenous H3 histone
mRNA (Figure 4-8 lanes 1,2 and 7,8; Fig. 4-10). Inhibition
of DNA synthesis for one hour with 1 mM hydroxyurea

Figure 4-7. Effects of hydroxyurea treatment on SPH3E1 niRNA.
HeLa cell monolayers were transfected with pSPH3El DNA
and 46 hours post-transfection were treated with 1 mM
hydroxyurea (HU). Cells were harvested at 20 minute
intervals and total cellular RNA was prepared. The RNA (10
Hq) was subjected to SI nuclease protection analysis to
quantitate the levels of wild type signal peptide-histone
fusion mRNA and endogenous H3 histone laRNA. Each lane
represents an individually transfected cell culture. MW is
radiolabeled Hpa II digest of pBR322 DNA. Lanes 1 and 2,
control; lanes 3 and 4, 1 mM HU 20 minutes; lanes 5 and 6, 1
mM HU 40 minutes; lanes 7 and 8, 1 mM HU 60 minutes.


Figure 4-8. Effects of hydroxyurea treatment on SPH3E1ATG
mRNA.
HeLa cell monolayers were transfected with pSPH3ElATG’
DNA and 46 hours post-transfection were treated with 1 mM
hydroxyurea (HU). Cells were harvested at 20 minute
intervals and total cellular RNA was prepared. The RNA (10
fig) was subjected to SI nuclease protection analysis to
quantitate the levels of SPH3E1ATG’ mRNA and endogenous H3
histone mRNA. Each lane represents an individually
transfected cell culture. MW is radiolabeled Hpa II digest
of pBR322 DNA. Lanes 1 and 2, control; lanes 3 and 4, 1 mM
HU 20 minutes; lanes 5 and 6, 1 mM HU 40 minutes; lanes 7
and 8, 1 mM HU 60 minutes.

SPH3E1ATG-
121

Figure 4-9. Effects of hydroxyurea treatment on SPH3Elalpha
mRNA.
HeLa cell monolayers were transfected with pSPH3Elalpha
DNA and 46 hours post-transfection were treated with 1 mM
hydroxyurea (HU). Cells were harvested at 20 minute
intervals and total cellular RNA was prepared. The RNA (10
Mg) was subjected to SI nuclease protection analysis to
quantitate the levels of the SPH3Elalpha mRNA and endogenous
H3 histone fusion mRNA. Each lane represents an individually
transfected cell culture. MW is radiolabeled Hpa II digest
of pBR322 DNA. Lanes 1 and 2, control; lanes 3 and 4, 1 mM
HU 20 minutes; lanes 5 and 6, 1 mM Hu 40 minutes; lanes 7
and 8, 1 mM HU 60 minutes.

S P H 3E1 alpha
123

Figure 4-10. Quantitation of SPH3E1, SPH3Elaloha. SPH3E1ATG
mRNA during hydroxyurea treatment.
The cellular levels of the signal peptide-histone
fusion mRNAs and endogenous H3 histone mRNA during
hydroxyurea treatment were determined by densitometric
analysis of the SI nuclease protection assays presented in
Figures 4-7 to 4-9. Values are presented as percentage of
control cells (•, SPH3E1 mRNASPH3Elalpha mRNA;■,
SPH3E1ATG’ mRNA;0, endogenous H3 histone mRNA).

Time (HU)
% Decrease
ro
cn
100

126
resulted in the destabilization of approximately 70% of
SPH3Elalpha mRNA, a value which is intermediate to that
observed for SPH3E1 mRNA and SPH3E1ATG mRNA (Figure 4-9
lanes 1,2 and 7,8; Fig. 4-10). The degree to which the
signal peptide-histone fusion mRNAs are degraded during
inhibition of DNA synthesis is correlated with the extent to
which the mRNA is associated with nonmembrane-bound
polysomes. These results suggest that the subcellular
localization of the histone mRNA plays a significant role in
the coupling of its stability to DNA replication.
Discussion
To our knowledge, this represents the first
demonstration that the addition of a signal peptide coding
sequence to an mRNA that is normally translated on
nonmembrane-bound polysomes results in the targeting of the
chimeric message to membrane-bound polysomes in intact
mammalian cells. The recognition of a prokaryotic signal
peptide in a eukaryotic cell is supported by both in vitro
and in vivo studies (Lingappa et al â–  â–  1984; Talmadge et al â–  .
1980a; Talmadge et al.â–  1980b). Saccharomvces cerevisiae has
been shown to express the Eh. coli pBR325 beta-lactamase gene
in vivo and to process the precursor protein to the
enzymatically active, mature protein (Roggenkamp et al.â– 
1981). The processing of the bacterial pre-protein into the
mature species has also been demonstrated in vitro using a

127
crude yeast extract (Roggenkamp et al.â–  1981). In addition,
it has been reported by Weidmann et alâ–  (1984) that
E. coli plasmid pBR322 beta-lactamase mRNA that is
synthesized and capped in vitro. when microinjected into
Xenopus oocytes, is translated into protein that is
ultima£ely secreted from the cell. These results suggest a
common mechanism of signal peptide recognition among
prokaryotic and eukaryotic organisms.
In an attempt to prevent the translocation of the
signal peptide-histone fusion mRNA to membrane-bound
polysomes, we initially disrupted the hydrophobic region of
the signal peptide by site directed mutagenesis. This
approach was based on previous studies demonstrating that
bacterial secretory proteins that were mutated in the
hydrophobic domain, either by insertion of charged amino
acids or by deletion, accumulate in the cytoplasm of the
bacterium (Emr et alâ– â–  1978; Bedouelle et alâ– . 1980; Emr and
Bassford, Jr., 1982). The incorporation of positively
charged histidine residues in the hydrophobic domain of the
signal peptide (SPH3Elalpha) however, resulted in only a
partial block in the association of mutated signal peptide-
histone fusion mRNA with membrane-bound polysomes (Figure 4-
6). This result is not surprising in light of more recent
studies on protein export in eukaryotes, which demonstrate
that the relationship between the primary sequence of the
signal peptide and its ability to function as a secretory

128
signal is quite variable and can withstand substantial
insertion, substitution or deletion mutations (Kaiser and
Botstein, 1986; Kaiser et alâ– â–  1987; Randall and Hardy,
1989). Subsequently, the signal peptide ATG translation
start codon of SPH3E1 was altered by site directed
mutagenesis which completely inactivated signal peptide
function (Figure 4-6).
The relative stability of signal peptide-histone fusion
mRNA (SPH3E1) during inhibition of DNA synthesis appears to
be a result of the subcellular location of the mRNA.
Destabilization of SPH3E1ATG" mRNA, which differs from
SPH3E1 mRNA by a single nucleotide, with essentially the
same kinetics and to the same extent as endogenous H3
histone mRNA indicates that SPH3E1 mRNA expresses the proper
structural confirmation for recognition by the factors that
mediate the selective destabilization of histone mRNA during
inhibition of DNA synthesis.
The difference in SPH3E1 and SPH3E1ATG mRNA stability
during inhibition of DNA synthesis may reflect a
perturbation of protein structure due to the signal peptide
at the N-terminus of SPH3E1. This possibility does not
appear likely based on the observations that histone amino
acid sequences are not required to target the mRNA for
destabilization (Pandey and Marzluff, 1987) and SPH3Elalpha
mRNA, which expresses the mutated form of the signal
peptide, is moderately destabilized during inhibition of DNA
synthesis (Figures 4-8 and 4-10).

129
These results indicate that the stability of SPH3E1
mRNA during inhibition of DNA replication is functionally
related to the change in the subcellular location of the
mRNA. The differential stability of SPH3E1 mRNA and
SPH3E1ATG’ mRNA may be due to qualitative differences in the
composition of membrane-bound and nonmembrane-bound
polysomes; however, there is no previously reported evidence
to support this possibility. Alternatively, the presence of
the fusion mRNA on membrane-bound polysomes rather than
nonmembrane-bound polysomes, where histone mRNA normally
resides, may physically separate the message from the
factors that are involved in the selective destabilization
of histone mRNA during inhibition of DNA synthesis. These
results are consistent with the hypothesis that the
subcellular location of histone mRNA plays an important role
in the posttranscriptional regulation of histone gene
expression.

CHAPTER 5
DIFFERENTIAL ASSOCIATION OF MEMBRANE-BOUND AND
NONMEMBRANE-BOUND POLYSOMES WITH CELL CYTOSTRUCTURE
Introduction
Localization of nearly all actively translated
polyribosomes on the cytoskeletal structure indicates that
polysomal mRNAs are not freely diffusible throughout the
cytoplasm (Lenk et alâ– . 1977; Cervera et alâ– â–  1981; Ornelles
et al.. 1986). This observation is further supported by
subcellular fractionation studies that demonstrate the
association of both "free" (nonmembrane-bound) and membrane-
bound polysomal mRNAs with the cytoskeleton (Cervera et alâ– â– 
1981; Jeffery, 1984; Bonneau et al.â–  1985).
Previous results indicate that the association of
polysomes with the cytoskeleton is mediated in part by the
mRNA, as treatments which dissociate the ribosomes from the
mRNA, such as heat shock and high salt, release the
ribosomal subunits into the soluble phase while the mRNA
remains attached to the cytoskeleton (Lenk et al.. 1977;
Cervera et al.â–  1981; van Venrooij et alâ– . 1981; Howe and
Hershey, 1984). However, the region of the mRNA and the
cytoskeletal elements to which it is attached are not known.
130

131
We present evidence here to support a model for
multiple sites of attachment of eukaryotic mRNA to the
cytoskeleton. Using cytochalasin D for the disruption of
microfilaments, we have observed a differential release of
specific mRNAs from the cytoskeleton. Further
characterization of the association of mRNA with the
cytoskeleton revealed an additional attachment site that is
insensitive to cytochalasin D treatment and present only in
membrane-bound polysomal mRNAs. This second attachment site
is puromycin sensitive and appears to involve the
association of the nascent polypeptide and/or ribosome with
the remnant protein structure of the endoplasmic reticulum.
Results
Cytochalasin D Releases Poly A" RNA from the Cytoskeleton
Disruption of actin fibers by cytochalasin D treatment
results in the release of cytoplasmic poly A+ RNA from the
cytoskeleton in a dose dependent manner (Ornelles et al..
1986). To further understand the association of eukaryotic
mRNAs with the cytoskeleton, we have studied the effects of
cytochalasin D treatment on the release of specific mRNAs,
including those that are non-polyadenylated. In addition we
have examined and compared the cytoskeletal association of
nonmembrane-bound and membrane-bound polysomal mRNAs.
Cell cycle dependent histone genes code for mRNAs that
are capped, non-polyadenylated and predominantly associated

132
with the cytoskeleton (Jeffery, 1984; Stein et al.â–  1977;
Adensik and Darnell, 1972; Zambetti et al.. 1985). The
distribution of histone inRNAs between the cytoskeletal and
the soluble fractions was determined by Northern blot
analysis before and after cytochalasin D treatment. As seen
in Figure 5-1, H3 and H4 cell cycle dependent histone mRNAs
are primarily associated with the cytoskeleton in untreated
HeLa cell cultures (lanes 1 and 2). Quantitation of the
autoradiograph in Figure 5-1 by scanning laser densitometry
revealed that >90% of the histone mRNAs were associated with
the cytoskeleton in control cell cultures (Table 5-1), which
is consistent with previously reported values (Zambetti et
al.. 1985). Upon addition of cytochalasin D, histone mRNAs
were released from the cytoskeleton into the soluble phase
(Fig. 5-1, lanes 3-8). As shown in Table 5-1, less than 35%
of the histone mRNAs remained associated with the
cytoskeleton in HeLa cell cultures treated with 10 ng/ml
cytochalasin D. Treatment with 40 nq/ml cytochalasin D
resulted in the dissociation of >70% of the histone mRNAs
from the cytoskeleton. The extent to which cytochalasin D
releases non-polyadenylated histone mRNA from the
cytoskeleton is similar to that reported for total,
heterogeneous poly-A+ RNA (Ornelles et alâ–  â–  1986).

133
Membrane-Bound Polvsomal RNA is not Released from the
Cytoskeleton bv Cvtochalasin D Treatment
HLA-B7 is a class X histocompatibility antigen that is
located on the cell surface (Robb et al.. 1978). Previously,
HLA-B7 mRNA was shown to be associated with the cytoskeleton
and translated into protein on membrane-bound polysomes
(Zambetti et al.â–  1985). As seen in Figure 5-1, HLA-B7 mRNA
was almost exclusively associated with the cytoskeleton in
control cells (lanes 1 and 2) and remained largely
associated with the cytoskeleton in cytochalasin D treated
cells (lanes 3-8). Approximately 98% of the HLA-B7 mRNA was
associated with the cytoskeleton in untreated cell cultures
(Table 5-1). In contrast to histone mRNA and total poly-A+
RNA, only 11% of the HLA-B7 mRNA was released from the
cytoskeleton in cell cultures treated with 10 /xg/ml
cytochalasin D (Table 5-1).
The retention of HLA-B7 mRNA on the cytoskeleton during
cytochalasin D treatment may be a consequence of translating
the mRNA on membrane-bound polysomes. To examine this
possibility, the distribution of chorionic gonadotropin
alpha mRNA within the cytoskeleton and soluble fractions was
determined in cytochalasin D treated cells. Chorionic
gonadotropin is a secreted, heterodimeric protein composed
of alpha and beta subunits (Milsted et al.. 1985). The alpha
subunit is expressed in exponentially growing HeLa cells
(Milsted et al.â–  1985), and Northern blot analysis of

Figure 5-1. Northern blot analysis of the cvtoskeleton and
soluble phase distribution of H3 histone. H4 histone and
HLA-B7 mRNAs in cvtochalasin D treated cells.
HeLa cells were treated with the indicated
concentration of cytochalasin D for 20 minutes and
cytoskeleton and soluble phase RNAs were isolated. Equal
quantities of RNA per sample (10 [iq/lane) were analyzed by
Northern blot analysis as described in Chapter 2. Lanes: 1)
Control/csk; 2) Control/sol; 3) CD [5 jug/ml]/csk; 4) CD [5
jig/ml]/sol; 5) CD [10 /ig/ml]/csk; 6) CD [10 iig/ml]/sol; 7)
CD [40 ng/ml]/csk; 8) CD [40 Mg/®l]/sol.

135
1 2345678
HLA-B7
H3
H 4

Table 5-1. The percent of cytoskeleton and soluble phase
associated mRNAs in cells treated with cvtochalasin D.
136
H3
H4
HLA-B7
hCG
Control
Csk
93%
96%
>98%
90%
Sol
7%
4%
< 2%
10%
CD [5 pg/ml]
Csk
24%
20%
79%
69%
Sol
76%
80%
21%
31%
CD [10 pg/ml]
Csk
33%
31%
87%
77%
Sol
67%
69%
13%
23%
CD [40 Mg/ml]
Csk
26%
20%
64%
41%
Sol
74%
80%
36%
59%
The densitometric
results
from Figure
5.1 (hCG
data not
shown) were normalized for the yield of RNA from the
cytoskeleton and soluble fractions of each cell sample
(note: equal quantities of cytoskeleton and soluble RNAs
were analyzed which does not take into consideration the
unequal distribution of RNA within these fractions or the
changes that occur during cytochalasin D treatment). The
values are presented as percent distribution of mRNA between
the cytoskeleton and soluble fractions.

137
subcellular fractions demonstrated that approximately 75% of
the chorionic gonadotropin alpha (hCGa) mRNA was associated
with membrane-bound polysomes (autoradiograph not shown).
Analogous to HLA-B7 mRNA, hCGa mRNA was not fully released
from the cytoskeleton in cytochalasin D treated cells;
greater than 90% of the gonadotropin mRNA was associated
with the cytoskeleton in control cells and approximately 77%
remained associated with the cytoskeleton in cells treated
with 10 ng/ml cytochalasin D (autoradiograph not shown; data
summarized in Table 5-1).
Nonmembrane-Bound Polvsomal mRNAs are Released from the
Cytoskeleton by Cytochalasin D Treatment
Previous results have demonstrated that histone mRNAs
are predominantly translated on cytoskeleton associated,
nonmembrane-bound polysomes (Zambetti et al.â–  1985) and as
described above, are released from the cytoskeleton into the
soluble phase by cytochalasin D treatment. As an additional
control for studying the release of nonmembrane-bound
polysomal RNAs from the cytoskeleton by cytochalasin D
treatment, we analyzed the distribution of c-fos mRNA
between the cytoskeleton and soluble fractions from control
and cytochalasin D treated cell cultures. The c-fos protein
is localized within the nucleus (Curran et al. . 1984) and
therefore, according to the signal hypothesis, should be
synthesized on nonmembrane-bound polysomes (Blobel and

138
Dobberstein, 1975). Subcellular fractionation and Northern
blot analysis revealed that c-fos mRNA is predominantly
associated with nonmembrane-bound polysomes (data not
shown), and that approximately 80% of the c-fos mRNA is
associated with the cytoskeleton (Figure 5-2, lanes 1 and 2;
Table 5-2). Cytochalasin D as well as puromycin,
CD/puromycin and CD/cycloheximide dramatically increased c-
fos mRNA levels, and the newly synthesized c-fos mRNA
partitioned into both the cytoskeletal and soluble fractions
in a manner that reflects the drug treatment (Figure 5-2,
lanes 3-10; Table 5-2). Less than 30% of the c-fos mRNA
remained associated with the cytoskeleton in cytochalasin D
[10 /¿g/ml] treated cells (Table 5-2).
Membrane-Bound Polvsomal mRNAs Express Multiple Cytoskeleton
Attachment Sites
Retention of membrane-bound polysomal mRNAs on the
cytoskeleton in cytochalasin D treated cells suggests that
this class of polysomal mRNA may contain additional
cytoskeleton attachment sites. Cell surface proteins, such
as HLA-B7 and chorionic gonadotropin, are generally
synthesized on membrane-bound polysomes that are complexed
with the endoplasmic reticulum (for review see Emr et alâ– .
1980). The nature of the interaction between the polysomes
and the endoplasmic reticulum appears to be mediated in part
by the nascent polypeptide that is inserted into the

Figure 5-2. Northern blot analysis of the cvtoskeleton and
soluble phase distribution of c-fos mRNA in cvtochalasin D.
puromvcin. CD/puro. and CD/cvcloheximide treated cells.
HeLa cells were cultured in the presence of the drugs
and fractionated into cytoskeleton and soluble phases as
described in the Chapter 2. Equal quantities of RNA (10
¿tg/sample) from each fraction were assayed for c-fos mRNA
content by Northern blot analysis. Lanes: 1) Control/csk; 2)
Control/sol; 3) CD/csk; 4) CD/sol; 5) Puro/csk; 6) Puro/sol;
7) CD/puro/csk; 8) CD/puro/sol; 9) CD/cyclo/csk; 10)
CD/cyclo/sol.

140
1 23456789 10
c-Fos

141
Table 5-2. Percent distribution of nonmembrane-bound and
membrane-bound polvsomal RNAs in the Csk and Sol fractions
from CD. Puro, CD/puromvcin and CD/cvcloheximide treated
cells.
H3/H4
HLA-B7
hCG
fos
Control
CSK
87%
(±5)
97% ( + 1)
96%
(±D
80%
SOL
13%
3%
4%
20%
CD
CSK
18%
(±4)
75% (±6)
72%
(±6)
29%
SOL
82%
25%
28%
71%
Puro
CSK
79%
(±7)
79% (±9)
85%
(±5)
83%
SOL
21%
21%
15%
17%
CD/Puro
CSK
18%
(±4)
25% (±3)
26%
(±3)
20%
SOL
82%
75%
74%
80%
CD/Cyclo
CSK
19%
(±8)
81% (±2)
52%
(±10)
40%
SOL
81%
19%
48%
60%
The Csk and Sol distribution of c-fos mRNA during CD,
puro, CD/puro and CD/cyclo treatment was quantitated by
normalizing the densitometric results from the autoradiogram
presented in Figure 5-2 for the yield of RNA in each
fraction. The Csk and Sol distribution of nonmembrane-bound
and membrane-bound polysomal mRNAs was quantitated directly
from the densitometric analysis. The numbers in brackets
indicate standard error.

142
membrane of the endoplasmic reticulum. In addition, the
nascent polypeptide is complexed with the proteins involved
in the translocation of the exported protein from the
cytoplasm into the lumen of the endoplasmic reticulum.
During the isolation of the cytoskeleton nearly all the
lipid components but not the protein substructure of the
endoplasmic reticulum membrane are removed by detergent
extraction (Lenk et al.. 1977). The cytoskeleton attachment
site associated with membrane-bound polysomes, which is
insensitive to cytochalasin D, may be the nascent
polypeptide complexed with the remnant protein structure of
the endoplasmic reticulum.
To test the possibility that the nascent polypeptide
serves as an additional anchor for the attachment of mRNA to
the cytoskeleton, we co-treated HeLa cells with cytochalasin
D and puromycin. Puromycin inhibits protein synthesis by
interrupting polypeptide chain formation which results in
the release of the nascent polypeptide and ribosomal
subunits from the mRNA. The distribution of membrane-bound
and nonmembrane-bound polysomal RNA within the cytoskeleton
and soluble fractions was then determined by Northern blot
analysis.
As seen in Figure 5-3, treatment of HeLa cells with
cytochalasin D and puromycin released membrane-bound
polysomal mRNAs from the cytoskeleton (lanes 7 and 8).
Approximately 25% of HLA-B7 mRNA and 20% of hCGa mRNA

143
remained associated with the cytoskeleton in cytochalasin D
treated cells that were subsequently treated with puromycin
(Table 5-2). In contrast, more than 75% of the HLA-B7 mRNA
and 72% of the hCGa mRNA remained associated with the
cytoskeleton in HeLa cells treated with cytochalasin D alone
(Table'5-2). The release of membrane-bound polysomal mRNAs
from the cytoskeleton in cytochalasin D and puromycin
treated HeLa cells was not a direct result of the inhibition
of protein synthesis. Cytochalasin D treatment followed by
the inhibition of protein synthesis with cycloheximide, a
compound that preserves the polysomal structure, failed to
release HLA-B7 mRNA from the cytoskeleton (Figure 5-3, lanes
9 and 10). Greater than 80% of the HLA-B7 mRNA was
associated with the cytoskeleton in cytochalasin D and
cycloheximide treated cells (Table 5-2). The results were
not as dramatic for the hCGa mRNA (Figure 5-3, lanes 9 and
10). Approximately 52% of hCGa mRNA was associated with the
cytoskeleton in cells treated with both cytochalasin D and
cycloheximide (Table 5-2).
The release of HLA-B7 mRNA and hCGa mRNA from the
cytoskeleton in cytochalasin D and puromycin treated HeLa
cells was not due solely to the inhibition of protein
synthesis by puromycin. As previously reported, the
dissociation of polysomes with high salt or heat shock
resulted in the release of the ribosomal subunits into the
soluble phase without affecting the attachment of poly A+

Figure 5-3. Northern blot analysis of the cytoskeleton and
soluble phase distribution of nonmembrane-bound and
membrane-bound polvsomal mRNAs in cvtochalasin D. puromvcin.
CD/puro and CD/cvcloheximide treated cells.
HeLa cells were cultured in the presence of the drugs
and fractionated into cytoskeleton and soluble phases as
described in Chapter 2. Equal volumes of cytoskeleton and
soluble phase RNA from each cell culture were assayed for H3
histone, H4 histone, HLA-B7 and chorionic gonadotropin mRNAs
by Northern blot analysis. Upper panel, probed for HLA-B7
and chorionic gonadotropin mRNA, and lower panel, probed for
H3 and'H4 histone mRNA. Upper and lower panels represent the
same Northern filter. Lanes: 1) Control/csk; 2) Control/sol;
3) CD/csk; 4) CD/sol; 5) Puro/csk; 6) Puro/sol; 7)
CD/puro/csk; 8) CD/puro/sol; 9) CD/cyclo/csk; 10)
CD/cyclo/sol.

145
1 2 345 6789 10
H3
H4

146
mRNA to the cytoskeleton (Cervera et al.. 1981; Howe and
Hershey, 1984). Consistent with this finding, 79% of HLA-B7
mRNA and 85% of hCGa mRNA remained attached to the
cytoskeleton in puromycin treated HeLa cells (Table 5-2). In
addition, 79% of the H3 and H4 histone poly A" mRNA and 83%
of the'c-fos mRNA remained associated with the cytoskeleton
in HeLa cells treated with puromycin (Table 5-2).
Discussion
There is growing support for the hypothesis that the
cytoskeleton of eukaryotes is more than a structural
scaffold. Evidence is accumulating to suggest that the
cytoskeleton is a dynamic structure that plays an integral
role in the process of protein synthesis. In addition, the
cytoskeleton may influence the posttranscriptional
regulation of eukaryotic genes by compartmentalizing mRNAs
within the cytoplasmic space. It is therefore of importance
to study the interactions of eukaryotic mRNAs with the
cytoskeleton.
The experiments presented here were designed to examine
the association of specific mRNAs with the cytoskeleton. As
described previously and confirmed in this study, histone
mRNA and HLA-B7 mRNA were found predominantly associated
with the cytoskeleton (Zambetti et al.. 1985). In addition,
hCGa and c-fos mRNAs were also localized in the cytoskeleton
fraction. Recently it was shown that the association of poly

147
A+ RNA with the cytoskeleton could be disrupted by
perturbing the microfilaments with cytochalasin D in HeLa
cells (Ornelles et alâ– â–  1986). Consistent with these
results, cytochalasin D treatment readily releases c-fos
mRNA and poly-A’ histone mRNA from the cytoskeleton into the
soluble phase. Surprisingly, HLA-B7 mRNA and hCGa mRNA were
only partially released from the cytoskeleton in
cytochalasin D treated HeLa cells.
The inefficiency of cytochalasin D in releasing HLA-B7
and hCGa mRNA from the cytoskeleton appears to be a result
of the association of these mRNAs with membrane-bound
polysomes. Dissociation of the polysomes with puromycin in
the presence of cytochalasin D releases HLA-B7 and hCGa mRNA
from the cytoskeleton. In contrast, cycloheximide, an
inhibitor of protein synthesis that preserves polysomal
structure, fails to release HLA-B7 mRNA from the
cytoskeleton in cytochalasin D treated cells. These results
suggest that membrane-bound polysomal mRNAs are attached to
the cytoskeleton in at least two different ways. The first
cytoskeleton attachment site is through the mRNA/mRNP itself
and is sensitive to cytochalasin D treatment. This site is
most likely common to all cytoskeletal associated mRNAs in
HeLa cells and has been previously observed for poly A+ RNA
by Ornelles et al. (1986). The second cytoskeletal
attachment site which is characteristic of membrane-bound
polysomes appears to be mediated by the nascent polypeptide

148
and/or ribosome that is complexed with the endoplasmic
reticulum. The possibility that the nascent polypeptide
serves as an additional cytoskeleton attachment site for
membrane-bound polysomes is supported by the observation
that nascent polypeptides undergoing translocation into the
lumen of the endoplasmic reticulum are resistant to protease
digestion following detergent extraction (Connolly et al..
1989). The protection of the nascent polypeptides from
protease digestion, in the absence of membranes, appears to
be due to the association of the polypeptide with the
proteins of the translocation apparatus of the endoplasmic
reticulum.
To release membrane-bound polysomal mRNAs efficiently
from the cytoskeleton, both cytoskeletal attachment sites
must be disrupted. Dissociation of the polysomes with
puromycin alone fails to release membrane-bound polysomal
mRNA from the cytoskeleton. In this case, the mRNA remains
associated with the cytoskeleton through the cytochalasin D
sensitive site even though the polysomes are no longer
anchored to the endoplasmic reticulum. Consistent with this
reasoning, cytochalasin D treatment alone results in only a
partial release of membrane-bound polysomal mRNA from the
cytoskeleton. The partial release of membrane-bound
polysomal mRNA from the cytoskeleton in cytochalasin D
treated cells is most likely a result of the release of mRNA
from the cytochalasin D sensitive cytoskeleton attachment

149
site coupled with a partial "run off" of the polysomes from
the mRNA, which in effect dissociates the nascent
polypeptide from the mRNA.
In addition to the differential association of
membrane-bound and nonmembrane-bound polysomes with the
cytoskeleton, there also appears to be differences in the
cytoskeletal interaction of individual species of mRNA. As
seen in Table 5-2, hCGa mRNA and HLA-B7 mRNA are released
from the cytoskeleton to the same extent during CD and
puromycin cotreatment, which is consistent with the
mechanism for the association of membrane-bound polysomes
with the cytoskeleton as described above. However, CD and
cycloheximide cotreatment releases a larger percentage of
hCGa mRNA from the cytoskeleton than CD treatment alone. In
contrast, HLA-B7 mRNA is associated with the cytoskeleton to
the same extent or greater in cytochalasin D and
cycloheximide co-treated cells than HeLa cells treated with
CD alone. Furthermore, CD and cycloheximide cotreatment
readily dissociates histone mRNA from the cytoskeleton
whereas c-fos mRNA is only partially affected. These results
suggest that multiple mechanisms are operative for the
association of individual mRNAs and/or polysomes with the
cytoskeleton and may reflect differences in the structural
environment of the cell.
The possibility that the structural environment of the
cell may affect cytoskeletal-mRNA interactions is supported

150
by the apparent cell type specific differences that are
observed for the components of the cytoskeleton that are
involved in the attachment of mRNA and polysomes. Adams et
al. (1983) found that polyribosomes in rat liver cells are
associated with the microfilaments and are released from the
cytoskeleton by deoxyribonuclease I (DNase I), an enzyme
that can depolymerize actin. Pretreatment of the rat liver
cells with phalloidin, a compound that stabilizes actin
filaments, abrogates the DNase I effects on the
polyribosome-cytoskeleton interactions (Adams et al.â–  1983).
Consistent with this result, poly A+ RNA, translation
factors and nearly 100% of the polysomes are preferentially
associated with the cytoskeleton in HeLa cells and can be
released into the soluble phase by the disruption of the
microfilaments with cytochalasin B or cytochalasin D (Lenk
et al.. 1977; Howe and Hershey, 1984; Ornelles et al..
1986). In contrast, Jeffery (1984) reported that the
association of actin mRNA and histone mRNA with the
cytoskeleton in ascidian eggs is not affected by
cytochalasin B treatment, which suggests that the
association is independent of the microfilaments.
Furthermore, disruption of the microfilaments by
cytochalasin B treatment does not release poly A+ histone H4
mRNA from the cytoskeleton in L6 rat myoblast cells (Bagchi
et al.. 1987).

151
The association of mRNA with the cytoskeleton also
appears to change during differentiation in certain cell
types (Toh et al.. 1980). Microscopic analysis of fetal rat
lung fibroblast cells indicates that polyribosomes are
aligned in linear arrays along the actin-like stress fibers
(Toh e€ al■■ 1980). These polysomes are difficult to
visualize after disruption of the microfilaments by
cytochalasin B treatment and are presumed to be released
from the cytoskeleton into the soluble phase. Polysomes in
the more differentiated fetal lung fibroblasts are
associated with the endoplasmic reticulum and are no longer
affected by cytochalasin B treatment (Toh et al.. 1980).
These results are consistent with the possibility that the
association of polysomes with the cytoskeleton changes
during differentiation.
In conclusion, the results presented here provide
evidence for heterogeneity in the interactions of mRNA with
the cytoskeleton in HeLa cells. Membrane-bound and non¬
membrane-bound polysomes are differentially associated with
the cytoskeleton. Studies on the association of mRNA with
the cytoskeleton must take into consideration these
differences when interpreting which elements are involved in
the cytoskeletal attachment of mRNA. Failure to release mRNA
from the cytoskeleton with cytochalasins may only suggest
that additional cytoskeletal attachment sites are involved.
The differential association of mRNA with the cytoskeleton

152
may play an important part in the subcellular localization
of messenger RNA, which may ultimately affect
posttranscriptional regulation. At this time, we can not
rule out the possibility that mRNAs localized in different
regions of the cytoplasm are associated with the
cytoskeleton by different mechanisms. Additional studies are
currently underway to define further the association of mRNA
with the cytoskeleton and what influence this structure has
on the regulation of eukaryotic gene expression.

CHAPTER 6
HETEROGENEOUS PATTERN FOR CYTOSKELETAL-mRNA INTERACTIONS
Introduction
In the preceding chapter we presented evidence
indicating that eukaryotic mRNAs are associated with the
cytoskeleton in a heterogeneous manner. Disruption of the
microfilaments by cytochalasin D treatment readily releases
nonmembrane-bound polysomes from the cytoskeleton into the
soluble phase, whereas membrane-bound polysomes remain
attached to the cytoskeleton. The cytoskeletal association
of membrane-bound polysomal mRNAs occurs through at least
two distinct sites: a cytochalasin D sensitive site and a
puromycin sensitive site, which is the interaction of the
nascent polypeptide/polysome with the protein substructure
of the endoplasmic reticulum (Chapter 5). To release
membrane-bound polysomes from the cytoskeleton, both
attachment sites must be disrupted as seen during
cytochalasin D and puromycin cotreatment. Previously, we
have demonstrated that a signal peptide-histone fusion mRNA
(SPH3E1) is targeted from nonmembrane-bound polysomes, the
natural site of histone protein synthesis, to membrane-bound
polysomes (Chapter 4). Relocating the histone mRNA within
153

154
the cell enabled us to examine mRNA-cytoskeleton
interactions in more detail. The results presented here
indicate that the signal peptide-histone fusion mRNA is
associated with the cytoskeleton in a manner unlike that
reported for membrane-bound and nonmembrane-bound polysomes.
These results are consistent with the existence of a
cytoskeletal attachment element within the mRNP or mRNA
primary sequence and provide additional evidence for
heterogeneity in mRNA-cytoskeleton interactions.
Results
To investigate further the mechanism of mRNA attachment
to the cytoskeleton, we have studied and compared the
cytoskeletal association of endogenous nonmembrane-bound
polysomal histone mRNA with that of membrane-bound
polysomal, signal peptide-histone chimeric mRNA. In Chapter
4, we have described the construction of a signal peptide-
histone chimeric gene (SPH3E1) that when transfected into
HeLa cells is transcribed and the resulting chimeric histone
mRNAs are targeted to membrane-bound polysomes (Zambetti et
al.. 1986). Although the signal peptide fused to the histone
gene is derived from the beta-lactamase gene of the Eh. coli
plasmid pBR322, it is functional in HeLa cells (Chapter 4).
Figure 4-5 schematically diagrams endogenous H3 histone and
the signal peptide-histone (SPH3E1) mRNAs.

155
Monoclonal HeLa cell lines expressing the chimeric gene
were fractionated into cytoskeleton and soluble phases as
described by Cervera et al. (1981). Subsequently, RNA from
each fraction was isolated and SPH3E1 chimeric mRNA and
endogenous H3 histone mRNA content was determined by SI
nuclease protection analysis (Chapter 2). As seen in Figure
6-1 and Table 6-1, both endogenous histone and SPH3E1 mRNAs
are predominantly associated with the cytoskeleton in
control cells (92% and 93%, respectively). Cytochalasin D
treatment [10 Mg/ml] brought about a limited release of
SPH3E1 mRNA from the cytoskeleton as expected for a
membrane-bound polysomal mRNA. Only 15% of SPH3E1 mRNA
compared to 40% of endogenous histone mRNA was dissociated
from the cytoskeleton in CD treated cells. As described
above, dissociation of the ribosomes by puromycin during
cytochalasin D treatment is necessary for the efficient
release of membrane-bound polysomal mRNAs from the
cytoskeleton. Surprisingly, CD and puromycin cotreatment
also failed to release the SPH3E1 mRNA from the
cytoskeleton. Greater than 70% of the signal peptide-histone
chimeric mRNA and less than 26% of endogenous H3 histone
mRNA remained associated with the cytoskeleton after CD and
puromycin cotreatment. The selective retention of the
chimeric mRNA on the cytoskeleton during CD and puromycin
cotreatment is demonstrated by the marked increase in the
intensity of the signal obtained for the same quantity of

Figure 6-1. Cvtochalasin D and puromycin cotreatment does
not release SPH3E1 or SPH3E1ATG~ mRNA from the cvtoskeletonâ– 
Clonal cell lines expressing SPH3E1 or SPH3E1ATG"
mRNA were treated with 10 ng/m1 cytochalasin D, 0.4 mM
puromycin or 10 fjg/ml cytochalasin D and 0.4 mM puromycin.
The distribution of endogenous histone and fusion mRNAs
within the cytoskeleton and soluble phase fractions was
determined by SI nuclease protection analysis as described
in Chapter 2. Analysis of a) SPH3E1 and SPH3E1ATG’ mRNAs and
b) endogenous H3 histone mRNA. Lanes 1-8 represent RNA
samples isolated from SPHE1 expressing cells and lanes 9-16
represent RNA samples isolated from SPH3E1ATG cell
cultures. Lanes 1 and 9, control-csk RNA; lanes 2 and 10,
control-sol RNA; lanes 3 and 11, CD-csk RNA; lanes 4 and 12,
CD-sol RNA; lanes 5 and 13, puro-csk RNA; lanes 6 and 14,
puro-sol RNA; lanes 7 and 15, CD/puro-csk RNA; lanes 8 and
16, CD/puro-sol RNA.

157
I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

158
Table 6-1. The percent of cvtoskeleton associated mRNAs
isolated from cvtochalasin D. puromvcin and CD/puromvcin
treated cells.
Control
CD
Puro
CD/Puro
A)
SPH3E1
>95%
93%
—
62%
Endogenous
H3
81%
31%
—
27%
B)
SPH3E1
93%
79%
94%
72%
hCG alpha
98%
81%
95%
18%
HLA-B7
99%
84%
93%
28%
Endogenous
H3
92%
46%
90%
26%
C)
SPH3E1ATG
>90%
87%
>89%
77%
hCG alpha
98%
81%
89%
29%
HLA-B7
97%
89%
88%
35%
Endogenous
H3
78%
56%
79%
24%
The densitometric results from autoradiographs
represented in Figures 6-2 to 6-4 (A to C, respectively)
were corrected for the total yield of RNA from each fraction
(note: equal quantities of cytoskeleton and soluble RNAs
were analyzed in SI and northern assays which does not take
into consideration the unequal distribution of RNA within
these fractions or the changes that occur during CD and/or
puromycin treatment).

159
RNA in this sample compared with the control cytoskeleton
sample (Figure 6-1, lane 1 compared to lane 7). The
inability of CD and puromycin to release SPH3E1 chimeric
mRNA efficiently from the cytoskeleton is not unique to the
monoclonal cell line and is also observed in a polyclonal
cell population (Figure 6-2, Table 6-1). In this case, 62%
of the endogenous H3 histone mRNA and less than 5% of the
SPH3E1 mRNA is released from the cytoskeleton during
cytochalasin D treatment. In CD and puromycin co-treated
polyclonal cell cultures approximately 27% of the endogenous
histone mRNA and 62% of the SPH3E1 mRNA remain associated
with the cytoskeleton.
To investigate whether the failure to release the
signal peptide-histone mRNA from the cytoskeleton with
cytochalasin D and puromycin was related to the efficiency
of the drug treatments, we studied the association of
endogenous membrane-bound polysomal mRNAs with the
cytoskeleton. The mRNAs coding for HLA-B7, a class I cell
surface histocompatibility antigen and chorionic
gonadotropin alpha (hCGa), a secreted protein, are
translated on membrane-bound polysomes and well represented
in HeLa cells. The distribution of HLA-B7 and hCGa mRNA in
the cytoskeleton and soluble samples used in Figure 6-1 was
determined by Northern blot analysis (Figure 6-3 and Table
6-1). As expected for membrane-bound polysomal mRNAs, HLA-B7
and hCGa mRNAs were not released by cytochalasin D treatment

Figure 6-2. The wild type signal peptide-histone fusion itRNA
is not released from the cvtoskeleton in CD and puromvcin
co-treated polyclonal HeLa cell cultures.
Polyclonal cell cultures expressing the SPH3E1 fusion
gene were treated with cytochalasin D and puromycin for 20
minutes. Cytoskeleton and soluble RNAs were isolated and
subjected to SI nuclease protection analysis as described in
Chapter 2. Lane 1, control-csk; lane 2, control-sol; lane 3,
CD [10 ¿ig/ml]-csk; lane 4, CD [10 Mg/ml ]-sol; lane 5, CD [30
Mg/ml]-csk; lane 6, CD [30 fig/ml ]-sol; lane 7, CD [10 fig/nl]
and puro [0.4 mM]-csk; and lane 8, CD [10 /ug/ml] and puro
[0.4 mM]-sol. MW is radiolabeled Hpa II digest of pBR322.

•I • 99
161
MW 12345678
SPH3E1
excess probe
Endogenous H3

Figure 6-3. Endogenous membrane-bound polvsomal mRNAs are
released from the cvtoskeleton during cotreatment with CD
and puromvcin in SPH3E1 and SPH3E1ATG' expressing cell
cultures.
Cytoskeleton and soluble RNAs (10 ;jg/sample) were
assayed by Northern blot analysis for HLA-B7 mRNA and
chorionic gonadotropin mRNA content as described in Chapter
2. Lane 1, control-csk; lane 2, control-sol; lane 3, CD-csk;
lane 4, CD-sol; lane 5, puro-csk; lane 6, puro-sol; lane 7,
CD/puro-csk; and lane 8, CD/puro-sol.

SPHE1
1 2 3 4 5
6
7 8
1 2
SPH3E1ATG”
3 4 5 6 7 8
HLA-
hCG

164
alone. Over 80% of these mRNAs were associated with the
cytoskeleton after CD treatment. In contrast, less than 28%
of the hCGa and HLA-B7 mRNAs remained associated with the
cytoskeleton after CD and puromycin cotreatment indicating
that the drug treatments were effective.
The association of the signal peptide-histone mRNA with
the cytoskeleton in CD and puromycin co-treated cells
suggests the existence of a cytoskeleton attachment site
that is distinct from the site associated with endogenous
histone mRNA and from other membrane-bound polysomes. These
results indicate that the site is independent of the
ribosomes and the nascent polypeptide and maybe a property
of the nucleotide sequence and/or proteins that interact
with the mRNA itself. Alternatively, the SPH3E1 chimeric
mRNA may be efficiently translated and therefore support re¬
initiation of translation even in the presence of
cytochalasin D and puromycin (note: endogenous histone mRNAs
are efficiently translated; Stahl and Gallwitz, 1977). This
would result in the synthesis of a portion of the signal
peptide which could then serve as an anchor to the
cytoskeletal structure as if the cells were treated with
cytochalasin D alone.
To address this possibility, we have studied the
cytoskeletal association of a mutated signal peptide-histone
chimeric mRNA (SPH3E1ATG ) . Using site directed mutagenesis
(Zoller and Smith, 1983) the ATG translation start codon for

165
the signal peptide of SPH3E1 was changed to TTG (Figure 4-
5). Translation should therefore bypass this altered start
codon and initiate at the normal ATG codon in the histone
coding region, resulting in the synthesis of histone protein
without the signal peptide sequence. Initial studies using a
HeLa monoclonal cell line expressing the SPH3E1ATG" gene
demonstrated the localization of SPH3E1ATG" mRNA on
nonmembrane-bound polysomes, which indicates that in vivo
the signal peptide was not synthesized (Fig. 4-6) . As seen
in Figure 6-1, greater than 90% of SPH3E1ATG" mRNA is
associated with the cytoskeleton in the control cell
culture. Cytochalasin D and puromycin cotreatment
effectively released hCGa and HLA-B7 mRNA from the
cytoskeleton with little or no effect on the association of
the SPH3E1ATG mRNA with the cytoskeleton (Figures 6-1 and
6-3, Table 6-1). Approximately 77% of SPH3E1ATG" mRNA and
less than 35% of HLA and hCGa mRNA remained attached to the
cytoskeleton under these conditions. Regardless of whether
the SPH3E1ATG' mRNA could be translated in the presence of
cytochalasin D and puromycin, no signal peptide would be
synthesized to anchor the message to the cytoskeleton. This
result further supports the proposal that the nucleotide
sequences coding for the signal peptide express a
cytoskeleton attachment site which is distinct from the
cytoskeletal attachment site of endogenous histone mRNA.

166
Discussion
The data presented in this chapter demonstrate that the
signal peptide-histone fusion mRNAs (SPH3E1 and SPH3E1ATG")
express a cytoskeleton attachment site that is not
associated with histone mRNA, HLA-B7 mRNA or heterogeneous
poly-A^ mRNA (Chapter 5; Ornelles et al.â–  1986). It is not
known whether the cytoskeleton attachment site unique to
SPH3E1 and SPH3E1ATG" mRNA is due directly to the primary
nucleotide sequences coding for the signal peptide, an
alteration in the overall conformation of the mRNA or the
proteins associated with the mRNP.
The elements of the cytoskeleton that are involved in
the attachment of SPH3E1 mRNA and SPH3E1ATG" during
cytochalasin D and puromycin cotreatment have not been
identified. Association of the signal peptide-histone fusion
mRNAs with the cytoskeleton during cytochalasin D and
puromycin cotreatment indicates that the attachment site is
independent of intact microfilaments. In addition, the
cytochalasin D and puromycin insensitive cytoskeleton
attachment site does not directly involve the microtubules
since this cytoskeletal component is removed during the
fractionation procedure (Lenk et alâ– . 1977). The signal
peptide-histone fusion mRNAs may be associated with the
intermediate filaments of the cytoskeleton. Previous studies
on other cell types indicate that mRNPs are associated with
the cytoskeleton through the intermediate filaments

167
(Jeffery, 1984; Bagchi et alâ– . 1987). Furthermore, prosomes,
which are small ribonucleoprotein complexes that are
associated with translationally inactive mRNAs, are attached
to the cytokeratin-intermediate filaments in HeLa cells
(Grossi de Sa et alâ– . 1988). The role of the intermediate
filaments in the attachment of the chimeric mRNAs to the
cytoskeleton remains to be determined.
In Chapter 5, we demonstrated that the signal peptide-
histone fusion mRNA (SPH3E1) is relatively stable, as
compared to endogenous H3 histone mRNA, during inhibition of
DNA synthesis. The uncoupling of signal peptide-histone
fusion mRNA stability from DNA synthesis does not appear to
be a consequence of its association with the cytoskeleton
through the cytochalasin D and puromycin insensitive site.
The mutated signal peptide-histone fusion mRNA (SPH3E1ATG')
is efficiently destabilized during inhibition of DNA
synthesis and most likely expresses the same CD and
puromycin insensitive cytoskeletal attachment site as SPH3E1
mRNA.
In summary, a cytoskeletal attachment site has been
detected which is associated with a signal peptide-histone
chimeric mRNA. This site appears to be a direct or indirect
result of the nucleotide sequences coding for the signal
peptide. Studies on deletion and point mutants of the signal
peptide-histone chimeric mRNAs should prove useful in
providing insight into the elements involved in the
attachment of the mRNA to the cytoskeleton.

CHAPTER 7
THE INFLUENCE OF THE CYTOSKELETON ON THE REGULATION
OF C-Fos GENE EXPRESSION
Introduction
Modifications in gene expression is central to
proliferation and differentiaton of eukaryotic cells. These
processes are tightly regulated and dependent in part on the
precise expression of proto-oncogenes (for review see
Ohlsson and Pfeifer-Ohlsson, 1987; Bishop, 1983). The
proliferative response of quiescent fibroblasts to serum or
growth factor stimulation is accompanied by a rapid and
transient induction of the c-fos proto-oncogene (Cochran et
al.. 1984; Greenberg et al.. 1984; Curran et al.. 1985).
Within minutes of the addition of serum or growth factors to
quiescent fibroblasts the c-fos gene is transcribed and c-
fos cytoplasmic mRNA levels rapidly increase. Generally, the
expression of the c-fos gene during serum/growth factor
induction is maximal after 30 minutes and returns to pre¬
stimulatory levels within 1-2 hours. While the exact
function of the c-fos protein is unknown, recent evidence
suggests that this nuclear protein is part of a DNA binding
168

169
complex that may influence the transcription of specific
genes (Curran et al.. 1984; Sytoyama et al.. 1986; Distel et
alâ– . 1987; Rauscher et al.â–  1988).
Signal transduction models describing the induction of
c-fos gene expression due to the binding of growth factor(s)
at the plasma membrane must include a mechanism for relaying
the stimulatory signal from the cell surface, through the
cytoplasm and into the nucleus. Existing within the
cytoplasm of the cell is an extensive and intricate protein
scaffold referred to as the cytoskeleton (for review see
Nielsen et al.â–  1983). In addition to its structural
properties the cytoskeleton also appears to function in cell
shape, cell motility, the intracellular transport of
macromolecules and translation of mRNA into protein.
Furthermore, the cytoskeleton may influence the regulation
of gene expression. Suspension of anchorage dependent mouse
3T6 fibroblasts induces a morphological change in cell shape
and the cytoskeleton with a concomitant reduction in DNA,
RNA and protein synthesis (Benecke et al.. 1978; Farmer et
al.â–  1983). Reattachment of these cells to a solid support
restores cell shape and cytoskeleton structure as well as
DNA, RNA and protein synthesis. In addition, disruption of
the microfilaments of the cytoskeleton with cytochalasin D
in HEp-2 cells results in the enhanced expression of beta
and gamma actin protein synthesis (Tannenbaum and Godman,
1983). Removal of the drug restores the structure of the
cytoskeleton and returns actin synthesis to normal levels.

170
The cytoskeleton also appears to play a role in the
transition between proliferation and the expression of
differentiated phenotypes. For example, normal B cells can
be induced to proliferate by co-stimulation with anti¬
immunoglobulin and B cell growth factor (Yoshizaki et al..
1982) . Cytochalasin D (CD) can substitute for B cell growth
factor and provide the signal, presumably through the
perturbation of the cytoskeleton, to induce proliferation of
anti-immunoglobulin treated normal B cells (Rothstein, 1985;
Rothstein, 1986) . Also, promyelocytic leukemia HL60 cells
cultured in the presence of tumor-promoting phorbol esters
undergo rearrangement of the cytoskeleton and differentiate
into non-dividing macrophage-like cells (Rovera et al..
1979; Huberman et al.â–  1979). Although the effects of
phorbol esters on cultured cells are many, it is possible
that the alteration in the cytoskeletal structure may be
involved in the progression of the differentiation pathway
(Diamond et al.â–  1980; Penman et al.. 1983).
To define further the role of the cytoskeleton in gene
regulation, we have studied the effects of disrupting the
cytoskeleton with cytochalasin D on the regulation of a
series of genes. We report here that cytochalasin D
treatment of exponentially growing HeLa cells rapidly
induces the transcription of the c-fos gene. This increase
in transcription is accompanied by an increase in c-fos mRNA
levels. The effect of cytochalasin D on c-fos gene

171
expression does not appear to be a general phenomenon since
the transcription and steady state levels of several other
active and inactive genes are unaffected by the drug
treatment. These results imply that the cell monitors the
organization and/or composition of the cytoskeleton and
adjust the regulation of c-fos gene expression accordingly.
This finding is consistent with signal transduction models
proposing that the cytoskeleton may be involved in relaying
stimuli from the plasma membrane to the nucleus.
Results
Cvtochalasin D Increases c-Fos mRNA Levels
Previous reports have suggested that the cytoskeleton
plays a role in signal transduction and the regulation of
gene expression. To further investigate this functional
property of the cytoskeleton, we have disrupted the
microfilaments of HeLa S3 cells with cytochalasin D and
measured the effects on transcription and mRNA cellular
levels of a series of genes. Within minutes of the addition
of cytochalasin D, cell morphology is drastically altered as
the cytoskeleton rearranges without any apparent
quantitative or qualitative loss of protein (Miranda et alâ– â– 
1974a; Miranda et alâ– â–  1974b; Schliwa, 1982; Ornelles et
al â–  . 1986) . It has been proposed that cytochalasin D acts by
binding to actin, causing the capping or fragmentation of
the microfilaments (Selden et al.. 1980; Lin et al.. 1980).

172
In addition, CD at concentrations equal to or greater than
10 ng/ml releases poly-A+ RNA from the cytoskeleton in a
dose dependent manner (Ornelles et al.. 1986). The release
of poly-A+ RNA from the cytoskeleton occurs in a dose
dependent manner and results in a stoichiometric inhibition
of protein synthesis.
To study immediate gene responses to cytochalasin D,
exponentially growing HeLa S3 cells were treated with 10
ng/ml of the drug for only 15 minutes. From each cell
culture total cellular RNA for Northern blot analysis and
nuclei for in vitro nuclear run-on transcription analysis
were isolated as described in Chapter 2. As shown in Figure
7-1, Northern blot analysis reveals a rapid and extensive
accumulation of c-fos mRNA in CD treated cells (lane 2). The
response of the c-fos gene is in sharp contrast to hCGa mRNA
steady state levels which are unaffected by the drug
treatment (Fig. 7-1). Quantitation by scanning laser
densitometry of several exposures of autoradiograms obtained
from four separate experiments indicates an average increase
of 20 fold in c-fos mRNA levels (Fig. 7-2). The
concentration of cytochalasin D used in these experiments
inhibits protein synthesis by 50 percent (Ornelles et al.â– 
1986) . It is well established that inhibition of protein
synthesis leads to an increase in c-fos mRNA presumably due
to increased mRNA stability (Muller et al.. 1984; Mitchell
et al. . 1985; Andrews et alâ– â–  1987; Rahmsdorf et al.â–  1987).

173
To address the possibility that the stimulation of c-fos
mRNA levels was due to the partial inhibition of protein
synthesis, we investigated the effect of puromycin at a
concentration of 0.4 mM, which inhibits protein synthesis by
more than 90 percent (Helms et al.. 1984), on c-fos gene
expression. Cells cultured in the presence of puromycin for
15 minutes exhibited 9 fold higher levels of c-fos mRNA
compared to untreated cell cultures (Fig. 7-1, lane 3; and
Figure 8-2) . However, cotreatment with CD and puromycin for
only 15 minutes resulted in a 49 fold increase in c-fos mRNA
levels (Fig. 7-1, lane 4 and Fig. 7-2) . The superinduction
of c-fos mRNA cellular levels by cytochalasin D and
puromycin cotreatment suggests that the stimulation by CD
alone is not due solely to its effect on protein synthesis
and may be the result of the perturbation of the
cytoskeleton.
Kinetics of c-Fos mRNA Accumulation During Cytochalasin D
Treatment
Treatment of quiescent BALB/C-3T3 mouse fibroblast
cells with growth factors or serum results in a rapid and
short lived increase in c-fos mRNA levels (Cochran et al.â– 
1984; Greenberg and Ziff, 1984; Muller et al.â–  1984).
Generally, cellular levels of c-fos mRNA are maximal after
30 minutes of serum/growth factor treatment and return to
pre-stimulatory levels within an hour. In contrast to serum

Figure 7-1. Effects of cvtochalasin D and puromvcin
treatment on steady state levels of c-fos and chorionic
gonadotropin mRNA.
Exponentially growing HeLa cells were treated with
cytochalasin D [10 /¿g/ml]» puromycin [0.4 mM] , or
cytochalasin D [10 ng/ml] and puromycin [0.4 mM] together
for 15 minutes. The cells were harvested and total cellular
RNA was isolated. Chorionic gonadotropin (hCGa) and c-fos
mRNA levels were determined by Northern blot analysis as
described in Chapter 2. Each lane represents 10 /¿g total
cellular mRNA. Lanes: 1) control; 2) cytochalasin D for 15
min.; 3) puromycin for 15 min.; and 4) CD and puromycin for
15 minutes.

175
12 3 4
c-f os
mm mm hcg

Figure 7-2. Densitometric anlavsis of steady state levels of
c-fos mRNA during cvtochalasin D and puromvcin treatments.
Steady state levels of c-fos mRNA during cytochalasin D
and puromycin treatments were determined by densitometric
analysis of the Northern filters. The values presented
represent the average of the minimal fold increase in c-fos
mRNA levels during the drug treatments as compared to the
control sample. Brackets indicate standard error from three
independent experiments.

fos Induction
177

178
and growth factor stimulation, inhibition of protein
synthesis produces a larger accumulation of c-fos mRNA
levels which persists for longer periods of time, usually
several hours (Greenberg et al.. 1986; Rahmsdorf et alâ– .
1987; Andrews et al.â–  1987). To determine whether the c-fos
gene response to cytochalasin D treatment corresponds to
that observed for serum/growth factor stimulation or
inhibition of protein synthesis, we examined the kinetics of
c-fos accumulation during CD treatment. Northern blot
analysis of total cellular RNA isolated from cells treated
with 10 pg/ml cytochalasin D for various lengths of time
demonstrated that maximal induction of c-fos mRNA,
approaching a 40 fold increase over control cultures,
occured after 30 minutes and was followed by a sharp decline
to near pre-stimulatory levels within one hour (Fig. 7-3,
lanes 4-9). In contrast, inhibition of protein synthesis
with puromycin resulted in maximal accumulation of c-fos
mRNA levels after 90 minutes, nearly 130 fold higher than in
untreated cells, and the mRNA levels remained elevated for
at least two hours (Fig. 7-3, lanes 16-20). The accumulation
of c-fos mRNA in CD and puromycin co-treated cells followed
the same time course as observed for puromycin treatment and
reached higher levels than with the CD treatment and
puromycin treatment combined (Fig. 7-3, lanes 10-15).
Parallel studies were done on serum starved WI38 human
diploid fibroblast cells. Stimulation of quiescent cell

179
cultures with culture medium containing 10% fetal calf serum
resulted in an increase in c-fos mRNA levels for 60 minutes
followed by a sharp reduction to backround levels by 120
minutes. Cytochalasin D affected c-fos gene expression with
the same kinetics as serum; however, slightly higher levels
of c-fos mRNA were measured in CD treated cells. Puromycin
treatment or CD and puromycin cotreatment caused a
continuous rise in c-fos mRNA levels for two hours, which
was the last time point examined. These results indicate
that the effects of cytochalasin D on either exponentially
growing HeLa cells or guiescent diploid fibroblasts more
closely resemble the effects of serum and growth factor
stimulation than the inhibition of protein synthesis on the
expression of c-fos mRNA levels.
Cytochalasin D Stimulates the Transcription of the c-Fos
Gene
To assess the contribution of transcription to the
increase in c-fos mRNA levels during CD treatment, we
measured the transcriptional activity of the c-fos gene by
in vitro nuclear run-on transcription analysis as described
in Chapter 2. Treatment of HeLa cells with cytochalasin D
for 15 minutes resulted in a significant increase in
transcription of the c-fos gene. The data presented in Fig.
7-4 represent a twelve fold increase in c-fos transcription
in cytochalasin D treated cells compared to untreated,

Figure 7-3. Time course of c-fos mRNA accumulation during
cvtochalasin D and puromycin treatments in exponentially
growing HeLa cells.
HeLa cells were treated with cytochalasin D [10 fiq/ml],
puromycin [0.4 mM] or cytochalasin D and puromycin for the
indicated times. The cells were collected and total cellular
RNA was isolated. The levels of c-fos mRNA were determined
by Northern blot analysis as described in Chapter 2. Each
lane represents 10 fig total cellular RNA. Lanes: 1)
control/0 min.; 2) control/60 min.; 3) DMSO control/60 min.;
4) CD/15 min.; 5) CD/30 min.; 6) CD/45 min.; 7) CD/60 min.;
8) CD/90 min.; 9) CD/120 min.; 10) CDPuro/15 min.; 11)
CDPuro/30 min.; 12) CDPuro/45 min.; 13) CDPuro/60 min.; 14)
CDPuro/90 min.; 15) CDPuro/120 min.; 16) Puro/15 min.; 17)
Puro/45 min.; 18) Puro/60 min.; 19) Puro/90 min.; 20)
Puro/120 minutes.

181
c-FOS

182
control cells. In contrast, puromycin treatment for 15
minutes did not significantly affect the transcription of
the c-fos gene. Cotreatment of HeLa cells with puromycin and
cytochalasin D for 15 minutes also stimulated c-fos gene
transcription, which indicates that inhibition of protein
synthesis during CD treatment does not affect the induction
of c-fos transcription.
Tannenbaum and co-workers have demonstrated an increase
in actin transcription during CD treatment (1983). In light
of this observation, we measured the transcription of the
beta-actin gene in our studies as a positive control.
Consistent with previous findings, CD increased actin
transcription, but only 5 fold over control cells (Fig. 7-
4). Cotreatment with CD and puromycin for 15 minutes also
increased transcription of the actin gene 5 fold compared to
controls and presumably reflects the effects of the
cytochalasin D (Fig. 7-4) . The cytochalasin D stimulation of
c-fos and beta-actin transcription is unaffected by
inhibition of protein synthesis.
The elevation in c-fos and actin transcription during
cytochalasin D treatment may be part of a general phenomenon
in which the transcriptional activity of all genes becomes
increased. Alternatively, transcription of ribosomal RNA
genes by RNA polymerase I, which represents approximately
50% of the transcriptional activity measured by nuclear run-
on assays (Marzluff and Huang, 1985), may be inhibited by

Figure 7-4. Effect of cytochalasin D on transcription of c-
fos in HeLa cells.
Exponentially growing HeLa cells were treated with
cytochalasin D, puromycin, or CD and puromycin for 15
minutes. Radiolabeled RNAs were prepared by in vitro nuclear
run-on transcription and hybridized to Southern blots of
electrophoretically separated restriction enzyme digested
plasmid DNAs (107 cpm/filter) . Insert DNAs hybridizing to
transcripts from the corresponding genes are indicated (both
Ribo signals represent 18S ribosomal RNA transcripts).
Untreated, DMSO only; Cytochalasin D, CD [10 iig/ml];
Puromycin, puromycin [0.04 mM]; Puro/Cyto, cotreatment with
CD [10 ng/ml] and puromycin [0.04 mM] . Autoradiograms in
upper panel for each treatment represent a 28 hour exposure.
Autoradiograms in lower panel, indicating Fos and H4 histone
transcription, represent a 4 day exposure.

184
UNTREATED
RIBO
ACTIN
RIBO
FOS
H4
FOS
H4
CYTOCHALASIN D
PUROMYCIN
PURO/CYTO
RIBO
ACTIN
RIBO
FOS
H4
FOS
H4

185
cytochalasin D and thereby raise the effective concentration
of other transcripts. To address these possibilities, we
examined the transcription of the 18S rRNA gene as well as
other genes during CD treatment. As seen in Fig. 7-4, the
transcription of 18S ribosomal RNA and cell cycle dependent
histone H4 genes is not significantly affected by
cytochalasin D. In addition, CD has no effect on the
transcription of genes coding for HLA-B7 (Fig. 7-5), beta-
globin and a cell cycle independent H2B histone (data not
shown). These results suggest that the increase in
transcription of c-fos and beta-actin genes during
cytochalasin D treatment is a selective response and not an
effect on transcription in general.
Transcriptional Induction of c-Fos During CD Treatment does
not Require New Protein Synthesis
The increase in c-fos gene transcription during CD
treatment may be a secondary effect of the drug.
Cytochalasin D may induce the synthesis of a protein that in
turn regulates the transcriptional activity of the c-fos
gene. To test this possibility, HeLa cells were pre-treated
with puromycin for 15 minutes prior to treatment with
cytochalasin D. The cells were then incubated in the
presence of both drugs for an additional 15 minutes. Cell
cultures treated with puromycin alone were incubated for 30
minutes in the presence of the inhibitor while control cells

186
were incubated in an equivalent concentration of DMSO for 30
minutes. Treatment of HeLa cells with cytochalasin D for 15
minutes resulted in the expected stimulation of c-fos
transcription (Fig. 7-5). Pretreatment of HeLa cells with
puromycin did not block the CD induced stimulation of c-fos
transcription and in fact, markedly increased the
transcription of the c-fos gene (Fig. 7-5). These results
suggest that the transcription factors are present within
the cells prior to cytochalasin D treatment, which is
consistent with previous findings that the factors involved
in the increase in c-fos transcription during serum
stimulation of quiescent fibroblasts are pre-existent
(Sassone-Corsi and Verma, 1987; Greenberg et al.. 1986).
Cytochalasin D Releases c-Fos mRNA from the Cvtoskeleton
into the Soluble Phase
The marked increase in c-fos mRNA cellular levels
during cytochalasin D treatment is not completely accounted
for by the increase in c-fos transcription and may be partly
a result of enhanced mRNA stability. Previous results
indicated that cytochalasin D [>10 ng/ml] releases poly-A+
RNA from the cytoskeleton into the soluble phase in a dose
dependent manner (Ornelles et alâ– . 1986). The stabilization
of c-fos mRNA during cytochalasin D treatment may therefore
be a consequence of an alteration in mRNA subcellular
location. To begin addressing this possibility, we have

Figure 7-5. Effect of protein synthesis inhibition on
cvtochalasin D induction of c-fos transcription in HeLa
cells.
Exponentially growing HeLa cells were treated with
cytochalasin D, puromycin, or puromycin and CD for the
indicated times. Radiolabeled RNAs were prepared by in vitro
nuclear run-on transcription and hybridized to Southern
blots of electrophoretically separated restriction enzyme
digested plasmid DNAs (107 cpm/filter). Insert DNAs
hybridizing to transcripts from the corresponding genes are
indicated. Untreated, DMSO for 15 min.; Cytochalasin D, CD
[10 Mg/ml] for 15 min.; Puromycin, puromycin [0.04 mM] for
30 min.; Puro + Cyto, puromycin [0.04 mM] for 30 min. and CD
[10 Mg/ml] for the last 15 minutes.

188
UNTREATED CYTOCHALASIN D
RIBO-
ACTIN -
H 4 -
FOS -
HLA-B7-
PUROMYCIN
PURO+CYTO
RIBO-
ACTIN -
HA -
FOS -
HLA-B7-
t

189
examined the subcellular location of c-fos mRNA, with
respect to its association with the cytoskeletal structure,
in both control and cytochalasin D treated cells.
Northern blot analysis of cytoskeletal and soluble
phase RNAs from untreated HeLa cells demonstrated that
approximately 80% of c-fos mRNA is associated with the
cytoskeletal structure (Figure 5-2 and Table 5-2).
Cytochalasin D treatment efficiently disrupted this
association and released c-fos mRNA from the cytoskeleton
into the soluble phase. Greater than 70% of the c-fos mRNA
accumulated in the soluble phase in cytochalasin D treated
cells. Cytochalasin D and puromycin cotreatment also
preferentially increased c-fos mRNA levels in the soluble
phase. Nearly 80% of the c-fos mRNA was localized in the
soluble fraction isolated from CD and puromycin co-treated
cells. Consistent with the work of Ornelles et alâ–  (1986),
puromycin treatment alone, which also increased c-fos mRNA
levels, had essentially no effect on the cytoskeletal and
soluble phase distribution of the mRNA. Approximately 85% of
the c-fos mRNA remained associated with the cytoskeleton in
puromycin treated cells. These results indicate that the
majority of the c-fos mRNA synthesized during cytochalasin D
treatment is localized in an altered subcellular region.
Accumulation of c-fos mRNA in an unnatural subcellular
compartment may physically remove it from the regulatory
factors that are involved in its stability.

190
Discussion
Previous results suggest that the organization of the
cytoskeleton may influence gene expression and ultimately
cell growth and differentiation. We present evidence here
supporting the role of the cytoskeleton in regulating gene
expression. Specifically, disruption of the microfilaments
with cytochalasin D results in a rapid and dramatic
accumulation of c-fos mRNA. The induction of c-fos gene
expression during cytochalasin D treatment is accompanied by
an increase in transcription of the c-fos gene as directly
determined by in vitro nuclear run on analysis.
The effect of cytochalasin D treatment on transcription
is not limited to the c-fos gene. Consistent with previous
results actin transcription was also elevated in
cytochalasin D treated cells (Tannenbaum and Godman, 1983).
However, the stimulation of transcription of the c-fos and
actin genes was not a conseguence of increased transcription
in general. RNA polymerase II transcription of several
genes, such as histone H4 (cell cycle dependent), histone
H2B (cell cycle independent), and HLA-B7 histocompatability
antigen, was unaffected in cells treated with cytochalasin
D. In addition, cytochalasin D had no effect on
transcription of the 18S ribosomal RNA genes by RNA
polymerase I or transcription of the beta-globin gene, a
transcriptionally inactive gene in HeLa cells.

191
Regulation of c-fos gene expression is controlled at
both the transcriptional and posttranscriptional levels. A
variety of agents such as serum, growth factors and
differentiation factors transcriptionally induce c-fos gene
expression in a rapid and transient manner (Greenberg and
Ziff, 1*984) . In contrast, inhibition of protein synthesis
with cycloheximide or heat shock results in a slower but
more persistent accumulation of c-fos mRNA with minimal or
no effect on c-fos transcription (Andrews et alâ– â–  1987;
Rahmsdorf et alâ– . 1987). The increase in c-fos mRNA levels
during inhibition of protein synthesis appears to be largely
posttranscriptionally regulated through the stabilization of
the c-fos mRNA. Previous results have demonstrated that
cytochalasin D can inhibit protein synthesis in a dose
dependent manner by releasing poly-A* RNA from the
cytoskeleton into the soluble phase as mRNP particles
(Ornelles et alâ– . 1986). The concentration of cytochalasin D
used in this study was sufficient to inhibit protein
synthesis by 50 percent. However, our results suggest that
the induction of c-fos gene expression by cytochalasin D
treatment is not a result of the partial inhibition of
protein synthesis. Cytochalasin D and puromycin cotreatment
resulted in roughly an additive increase in c-fos gene
transcription, suggesting the effects of the two drugs on
transcription are due to different mechanisms. Consistent
with this reasoning, the cellular level of c-fos mRNA in

192
cytochalasin D and puromycin co-treated cells is greater
than the levels of c-fos mRNA in CD treated and puromycin
treated cells combined. Furthermore, the time course of c-
fos mRNA accumulation in cytochalasin D treated cells is
more rapid and shorter lived than that observed in puromycin
treated cells.
Stimulation of c-fos transcription by cytochalasin D
treatment can occur during inhibition of protein synthesis
and therefore appears to be dependent on factors that are
already present within the treated cells. Pretreatment with
puromycin for 15 minutes followed by the addition of
cytochalasin D for 15 minutes results in a marked increase
in c-fos transcription (Fig. 7-5). Furthermore, inhibition
of protein synthesis for 30 minutes with puromycin alone
results in a significant increase in c-fos transcription.
These results are consistent with previous reports
suggesting that the c-fos gene is negatively regulated by a
labile factor (Mitchell et alâ– . 1985; Greenberg et al..
1986; Sassone-Corsi and Verma, 1987).
The increase in the cellular levels of c-fos mRNA
during cytochalasin D treatment is greater than the measured
increase in the transcription of the gene and may be due to
the stabilization of c-fos mRNA under these conditions.
However, it is possible that during cytochalasin D treatment
the mechanism involved in the turnover of c-fos mRNA remains
operative despite the reduced levels of protein synthesis.

193
Previous studies on cytochalasin D treated HeLa cells
demonstrated the selective destabilization of histone mRNA
during inhibition of DNA synthesis (unpublished data). This
event is strictly dependent on protein synthesis and
demonstrates the competence of cytochalasin D treated cells
to carry out certain posttranscriptional processes. The
apparent stabilization of c-fos mRNA during cytochalasin D
treatment may therefore be a result of the release of the
mRNA from the cytoskeleton in monosome or mRNP form. Release
of c-fos mRNA from the cytoskeleton into the soluble phase
could physically separate the message from the factors
involved in its turnover. These regulatory factors may be
seguestered and concentrated by the cytoskeleton or an
integral component of the ribosome.
The mechanism of induction of c-fos gene expression by
cytochalasin D treatment is not known. Many of the inducers
of c-fos transcription studied to date appear to effect
protein kinase C and/or adenylate cyclase enzyme activities
(Verma and Sassone-Corsi, 1987). The induction of c-fos gene
expression by cytochalasin D may also occur as an indirect
effect on protein kinase C or adenylate cyclase levels
although this has not yet been determined. Alternatively,
the disruption of the cytoskeleton by cytochalasin D may
also perturb the nuclear matrix which may in turn affect
transcription. However, a model which includes the
alteration in transcription as a result of the perturbation

194
of the nuclear matrix must explain the selective activation
of c-fos and actin genes and therefore does not appear
likely. In contrast, stimulation of c-fos gene expression by
cytochalasin D may be a result of the alterations in the
cytoskeleton due to the disruption of the microfilaments.
Analogous to the work presented here, it has recently been
reported that the disruption of the microtubules with
colchicine, nocodazole, or vinblastine induces the
transcription of chloramphenicol acetyl transferase genes
containing promoter elements from the human c-fos or beta-
actin gene (Ng, 1989). Interestingly, sequence analysis
indicates that actin and c-fos genes contain similar
transcription consensus sequences, referred to as serum
response elements, located in their promoter regions (Mohun
and Garret, 1987). The coordinate induction of c-fos and
actin transcription during cytochalasin D or colchicine
treatment may be mediated by the interaction of the same or
similar trans acting factor(s) with the serum response
element.
Several reports have postulated that the cytoskeleton
is involved in signal transduction (Rothstein, 1985;
Rebillard et ai. . 1987; Blum and Wicha, 1988). For example,
Rebillard and co-workers (1987) reported that the
stimulation of quiescent Swiss 3T3 cells with epidermal
growth factor (EGF) and insulin results in the rapid
accumulation of c-fos mRNA. The increase in c-fos gene

195
expression during EGF and insulin cotreatment can be
inhibited by disruption of the cytoskeleton with
cytochalasin D. This result is consistent with the
hypothesis that the cytoskeleton may be involved in relaying
the EGF/insulin signal from the cell surface to the nucleus.
The results presented here describing the stimulation of c-
fos gene expression with cytochalasin D, although contrary
to Rebillard's results, are consistent with their
interpretation that the nucleus monitors the integrity of
the cytoskeleton and adjusts the expression of particular
genes accordingly. Rather than cytochalasin D inhibiting a
signal from the cell surface to the nucleus, the drug in our
hands may mimic the signal to stimulate c-fos gene
expression. Cell type, culture conditions and drug dosage
effects may explain the differences observed between our
results and those of Rebillard's et alâ–  (1987).
In summary, cytochalasin D stimulates c-fos gene
expression through increased transcription and mRNA
stability. The effect of cytochalasin D on c-fos gene
expression may be in response to the alterations in the
cytoskeleton and cell shape. Whether other elements of the
cytoskeleton such as the intermediate filaments can elicit
the same response remains to be determined. Taken together
with previous reports, the ability of cytochalasin D to
stimulate c-fos in guiescent fibroblasts with the same
kinetics and virtually to the same extent as serum supports
the role of the cytoskeleton in signal transduction.

CHAPTER 8
SUMMARY AND FUTURE CONSIDERATIONS
Tlie initial emphasis of the work presented in this
dissertation was focused on studying the posttranscriptional
regulation of histone gene expression. When this project was
undertaken little was known concerning the coupling of
histone mRNA stability to DNA replication. It was clear that
the destabilization of histone mRNA during inhibition of DNA
synthesis was selective and reguired ongoing protein
synthesis (Butler and Mueller, 1973; Gallwitz, 1975; Stahl
and Gallwitz, 1977; Stimac et alâ–  â–  1983; Bauxnbach et al. â– 
1984; Helms et al.. 1984). In addition, it was known that
the kinetics of histone gene expression were extremely rapid
which is evident by the appearance of translatable histone
mRNAs on polysomes within minutes following transcription
(Borun et al.. 1967) and the equally rapid transfer of newly
synthesized histone protein into the nucleus (Spaulding et
al.â–  1966; Robbins and Borun, 1967). At that time, several
research groups were providing evidence that suggested the
cell is organized in a highly ordered manner. Penman and co¬
workers, while studying the structural and functional
properties of the cytoskeleton, were developing the
196

197
perspective that cytostructure can form subcellular
compartments that are independent of membrane envelopes
(Fulton et al.. 1980; Cervera et al.. 1981). Furthermore,
Singer and co-workers, using in situ hybridization
techniques, were demonstrating the localization of specific
mRNAs in discrete areas of the cytoplasm (Lawrence and
Singer, 1986) .
Based on these observations and the information
available concerning histone gene expression, we proposed
that histone mRNAs and the factors that are involved in
their selective destabilization during inhibition of DNA
synthesis are co-compartmentalized in a perinuclear region.
This model predicted that the subcellular location of
histone mRNA plays a role in the coupling of histone mRNA
stability with DNA replication.
To address this question, we first determined the
normal subcellular location of histone mRNA. Subcellular
fractionation and Northern blot analysis demonstrated that
histone mRNAs are naturally located on nonmembrane-bound
polysomes that are associated with the cytoskeleton (Chapter
3). The subcellular localization of histone mRNA was then
altered by incorporating nucleotide sequences coding for a
signal peptide into an H3 histone mRNA (Chapter 4). The
modified histone gene, when transfected into HeLa cells,
expressed a signal peptide-histone fusion mRNA which was
directed to membrane-bound polysomes. The histone fusion

198
mRNA, in contrast to endogenous histone mRNA, was not
efficiently destabilized when DNA synthesis was interrupted
with hydroxyurea. Site directed mutagenesis analysis of the
histone fusion gene and subsequent in vivo studies indicated
that the uncoupling of the histone fusion mRNA stability
from DííA replication was due to the altered subcellular
location and was not due to the perturbation of mRNA
structure.
It has since been demonstrated by in situ hybridization
analysis that histone mRNAs are localized in "grape like
clusters" throughout the cytoplasm (Lawrence et alâ– . 1988).
In this regard, the initial proposal that histone mRNAs are
compartmentalized in a region close to the nucleus was
incorrect. The non-uniform distribution of histone mRNAs
within the cytoplasm, however, suggests that histone mRNAs
are compartmentalized. The stability of histone fusion mRNA
during inhibition of DNA synthesis, due to the association
of the mRNA with membrane-bound polysomes, is consistent
with the possibility that these compartments are involved in
the posttranscriptional regulation of histone gene
expression.
The proposal that histone protein synthesis is
autogenously regulated provides one possible explanation for
the role of the subcellular compartments in the
posttranscriptional regulation of histone gene expression.
As proposed by Butler and Mueller (1973) and several other

199
research groups (Stein and Stein, 1984; Wu and Bonner,
1985), a critical concentration of unbound histone protein
may be required for the destabilization of histone mRNA
during inhibition of DNA synthesis. Localization of histone
mRNA in clusters may create pockets within the cytoplasm in
which £he unbound histone protein could reach the necessary
concentration for the destabilization of histone mRNA to
occur.
The alteration in subcellular location of the histone
fusion mRNA has been demonstrated by subcellular
fractionation and SI nuclease protection analysis (Chapter
4). It remains to be determined by in situ hybridization
analysis whether the spatial distribution of the histone
fusion mRNA has been significantly altered compared with
that of endogenous histone mRNA. Furthermore, identification
and purification of the components involved in the
posttranscriptional regulation of histone gene expression
may provide the opportunity to monitor directly the
subcellular localization of these factors by
immunofluorescence, cell fractionation and biochemical
analysis.
Another aspect concerning the localization of histone
mRNA in the cytoplasm, which warrants additional
consideration, is whether a single class or multiple classes
of histone mRNAs are contained within a cluster. For
example, a particular cluster may contain only H3 histone

200
mRNA, H3 and H4 histone mRNA together, or any combination of
groupings of the different classes of histone mRNAs. The
pattern for these groupings may be important in the
transport of the newly synthesized histone proteins into the
nucleus and/or the coordinate regulation of histone protein
synthesis.
Although unrelated to the problems addressed here, it
would be interesting from the standpoint of studying the
process of protein secretion to determine the subcellular
localization of the histone fusion protein. Examination of
subcellular fractions isolated from HeLa cells which are
expressing the signal peptide-histone fusion gene, by
standard biochemical technigues, such as acid extraction and
triton-urea gel electrophoresis, would provide insight into
how the fusion protein is processed and where the fusion
protein is located.
In studying the subcellular location of histone mRNA,
it was determined that histone mRNA is predominantly
associated with the cytoskeleton and that this association
is dependent on the integrity of the microfilaments
(Chapters 3 and 5). Cytochalasin D treatment, which disrupts
the actin filaments, releases histone mRNA from the
cytoskeleton into the soluble phase. This result is
consistent with the observation of Penman and co-workers
that heterogeneous poly-A+ RNA is released from the
cytoskeleton in cytochalasin D treated HeLa cells (Ornelles
et alj_, 1986) .

201
The release of histone mRNA from the cytoskeleton by
cytochalasin D treatment presented the opportunity to study
the role of the cytoskeleton in the posttranscriptional
regulation of histone gene expression. Histone mRNA was
released into the soluble phase with moderate amounts of
cytochalasin D [10 ug/ml], since higher levels of CD
significantly inhibit protein synthesis. The cells were then
treated with hydroxyurea to inhibit DNA replication. These
treatments resulted in the destabilization of histone mRNA
in the soluble fraction to the same extent and with the same
kinetics as the destabilization of histone mRNA in the
cytoskeleton fraction. One interpretation of this result is
that the cytoskeleton is not required for histone mRNA
destabilization during inhibition of DNA synthesis. In
addition, Ornelles et al. (1986) reported that CD treatment
of HeLa cells releases poly-A+ RNA from the cytoskeleton
into the soluble phase in monosome or mRNP form, which
suggests that histone mRNA in the soluble phase does not
need to be translated to be destabilized during inhibition
of DNA synthesis. This result is in contrast to the genetic
studies that indicate histone mRNA must be translated to
within close proximity of the translation stop codon for
destabilization to occur (Graves et al â–  â–  1987; Capasso et
al.. 1987). The release of histone mRNA into the soluble
phase as a monosome or mRNP during cytochalasin D treatment
has not yet been determined and additional studies will be

202
necessary to clarify these interpretations of the
cytochalasin D-hydroxyurea experiments. For example, sucrose
density gradient analysis of soluble phase RNA from
cytochalasin D treated cells should demonstrate whether
histone mRNA is released from the cytoskeleton in the same
manner"as poly-A+ RNA. Furthermore, gel filtration studies
as described by Adams et alâ–  (1983) on the composition of
the histone mRNA in the soluble phase after cytochalasin D
treatment should also provide information on whether the
histone mRNA, although no longer anchored to the
cytoskeleton, may still be complexed with cytoskeletal
components. These cytoskeletal components may be necessary
for histone mRNA destabilization during inhibition of DNA
replication.
The distribution of HLA-B7 mRNA between the
cytoskeleton and soluble fractions during cytochalasin D and
hydroxyurea cotreatment was also analyzed as an internal
control for equal loading of the RNA. Surprisingly, in view
of the results of Ornelles et al. (1986), cytochalasin D
treatment failed to release HLA-B7 mRNA from the
cytoskeleton (Chapter 5). Subsequent studies demonstrated
that polysomal HLA-B7 mRNA, as well as other membrane-bound
polysomes, are released from the cytoskeleton during
cytochalasin D and puromycin cotreatment (Chapter 5).
Puromycin alone or cytochalasin D and cycloheximide
cotreatment does not release membrane-bound polysomes from

203
the cytoskeleton. These results suggest that one mechanism
for the attachment of membrane-bound polysomes to the
cytoskeleton is through the interaction of the polysome
complex with the endoplasmic reticulum. This interaction
appears to involve the nascent polypeptide; however other
possibilities may exist.
The relocation of signal peptide-histone fusion mRNA to
membrane-bound polysomes presented the opportunity to study
the cytoskeletal interactions of membrane-bound polysomal
mRNAs in more detail. Our model predicted that histone
fusion mRNA would remain attached to the cytoskeleton during
CD treatment and would be released into the soluble phase
during CD and puromycin cotreatment. As seen in Chapter 6,
the signal peptide-histone fusion mRNA remained attached to
the cytoskeleton during cytochalasin D treatment as well as
during cytochalasin D and puromycin cotreatment. The
retention of signal peptide-histone fusion mRNA on the
cytoskeleton under these conditions demonstrates that the
histone fusion mRNA is associated with the cytoskeleton in a
manner that is different from that observed for wild type
histone mRNA, HLA-B7 membrane-bound polysomal RNA and
heterogeneous poly-A+ RNA. Although it is well documented
that mRNAs are associated with the cytoskeleton, there are
no known cytoskeleton attachment sites identified to date.
Dissection of the signal peptide-histone fusion mRNA, by
deletion and site directed mutation analysis, should prove

204
useful in the identification of an mRNA-cytoskeleton
attachment site.
The role of the cytoskeleton structure in influencing
gene expression is supported by the observation that
cytochalasin D treatment resulted in a rapid and dramatic
increase in c-fos gene transcription with a concomitant
increase in c-fos mRNA cellular levels (Chapter 7).
Consistent with previous results (Tannenbaum and Godman,
1983), CD treatment also increased beta-actin gene
transcription. However, these modifications in gene
expression appear to be selective and not a general
phenomenon, since the transcription of several other genes
was unaffected by the drug treatment. The increase in c-fos
mRNA cellular levels was larger than the increase in c-fos
gene transcription and may be due to a stabilization of the
c-fos message under these conditions. Cell fractionation and
Northern blot analysis demonstrated that c-fos mRNA is
associated with the cytoskeleton in control cells and
accumulates in the soluble phase in cytochalasin D treated
cells (Chapter 5). Stabilization of c-fos mRNA may be due to
the altered subcellular location of the message during
cytochalasin D treatment. Additional studies are needed to
define the molecular basis for the role of the cytoskeletal
structure in the transcriptional and posttranscriptional
regulation of c-fos gene expression.

205
In summary, the work presented in this dissertation is
consistent with the hypothesis that subcellular compartments
and cell structure play an important role in the regulation
of gene expression. The influence of mRNA subcellular
location on the posttranscriptional control of histone gene
expression and possibly c-fos gene expression, indicates
that regulatory factors may be concentrated and sequestered
in specific regions or by specific structures. The
differential association of mRNA with the cytoskeleton may
be a mechanism for localization of mRNA in a particular
region of the cell, which in turn may affect the
posttranscriptional regulatory process. Lastly, the effect
of disrupting the microfilaments with cytochalasin D on c-
fos gene and beta-actin gene expression suggests that the
nucleus can transcriptionally respond in a selective manner
to the structural organization of the cytoskeleton.

REFERENCES
Adams, A., Fey, E. G., Pike, S., Taylorson, C., White, H.,
and Rabin, B. (1983) Biochem. J. 216â–  215.
Adesnik, M., and Darnell, J. (1972) J. Mol. Biol. 67, 397.
Allfrey, V., Littau, V., and Mirsky, A. (1963) Proc. Natl.
Acad. Sci. U.S.A. 49, 414.
Alternan, R., Sprecher, C., Graves, R., Marzluff, W., and
Skoultchi, A. (1985) Mol. Cell. Biol. 5, 2316.
Amos, L., and Klug, A. (1974) J. Cell Sci. 14, 523.
Andrews, G., Harding, M., Calvert, J., and Adamson, E.
(1987) Mol. Cell. Biol. 7, 117.
Bagchi, T., Larson, D., and Sells, B. (1987) Exp. Cell Res.
168â–  160.
Baumbach, L. , Stein, G., and Stein, J. (1987) Biochemistry
26, 6178.
Baumbach, L., Marashi, F., Plumb, M., Stein, G., and Stein,
J. (1984) Biochemistry 23., 1618.
Bedouelle, H., Bassford Jr., P., Fowler, A., Zabin, X.,
Beckwith, J., and Hofnung, M. (1980) Nature 285. 78.
Benecke, B.-J., Ben Ze'ev, A., and Penman, S. (1978) Cell
14, 931.
Bennet, G., Fellini, S., and Holtzer, H. (1978)
Differentiation 12., 71.
Berk, A., and Sharp, P. (1978) Proc. Natl. Acad. Sci. U.S.A.
75, 1274.
Berl, S., Puszkin, S., and Nicklas, W. (1973) Science 179,
441.
Binder, L., Frankfurter, A., and Rebhun, L. (1985) J. Cell
Biol. 101, 1371.
206

207
Bird, R., Jacobs, F., Stein, G., Stein, J., and Sells, B.
(1985) Biochem. Biophys. Acta 824â–  209.
Bishop, J.M. (1983) Annu. Rev. Biochem. 52, 301.
Blobel, G., and Dobberstein, B. (1975) J. Cell Biol. 67,
835.
Bloom, G.
331. „
, Luca, F.,
and
Vallee,
R. (1984) J.
Cell
Biol. 98
Blose, S.
, and Chacko
, s
. (1976)
J. Cell Biol.
. 70,
459.
Blum, J.,
and Wicha,
M.
(1988) J.
. Cell. Phys.
135,
13.
Bonneau, A., Darveau, A., and Sonenberg, N. (1985) J. Cell
Biol. 100, 1209.
Borun, T., Scharff, M., and Robbins, E. (1967) Proc. Natl.
Acad. Sci. U.S.A. 58, 1977.
Borun, T., Gabrielli, F., Ajiro, K., Zweidler, A., and
Baglioni, C. (1975) Cell 4., 59.
Briendl, M.,
381.
, and
Gallwitz,
D.
(1973)
Eur.
J.
Biochem.
22
Briendl, M. ,
91.
, and
Gallwitz,
D.
(1974)
Eur.
J.
Biochem.
45
Brown, S., Levinson, W., and Spudich, J. (1976) Supramolec.
Struct. 5, 119.
Brush, D., Dodgson, J., Choi, O.-R., Stevens, W., and Engel,
J. (1985) Mol. Cell. Biol. 5, 1307.
Bulinski, J., and Borisy, G. (1980) J. Cell Biol. 87, 802.
Butler, W., and Mueller, G. (1973) Biochem. Biophys. Acta
294. 481.
Capasso, O., Bleecker, G., and Heintz, N. (1987) EMBO J. 6,
1825.
Carozzi, N., Marashi, F., Plumb, M., Zimmerman, S.,
Zimmerman, A., Wells, J., Stein, G., and Stein, J. (1984)
Science 224â–  1115.
Cervera, M., Dreyfuss, G., and Penman, P. (1981) Cell 23â– 
113.

208
Challoner, P., Moss, S., and Groudine, M. (1989) Mol. Cell.
Biol. 9, 902.
Clarke, M., and Spudich, J. (1977) Annu. Rev. Biochem. 46â– 
797.
Clayton, D., and Darnell Jr., J. (1983) Mol. Cell. Biol. 3,
1552.
Cline,^K. (1986) J. Biol. Chem. 261. 14804.
Cochran, B., Zullo, J., Verma, I., and Stiles, C. (1984)
Science 226â–  1080.
Collart, D., Stein, G., and Stein, J. (1985) Mol. Cell.
Bioch. 62, 161.
Connolly, T., Collins, P., and Gilmore, R. (1989) J. Cell
Biol. 108, 299.
Curran, T., Miller, A., Zokas, L., and Verma, I. (1984) Cell
36, 259.
Curran, T., and Morgan, J. (1985) Science 229. 1265.
Date, T., Zwizinski, C., Ludmerer, S., and Wickner, W.
(1980) Proc. Natl. Acad. Sci. Ü.S.A. 22, 827.
Delange,
J. Biol.
R. , Fambrough, D., Smith, E.,
Chem. 244, 5669.
, and
Bonner, J
. (1969)
DeLisle,
A., Graves, R., Marzluff, W.,
. and
Johnson,
L.
(1983) Mol. Cell. Biol. 3, 1920.
Delacourte, A., Filleatreau, G., Boulteau, F., Biserte, G.,
and Schrevel, J. (1980) Biol. J. 191, 543.
Diamond, L., O'Brien, T., and Baird, W. (1980) Adv. Cancer
Res. 32, 1.
Distel, R., Ro, H.-S., Rosen, B., Groves, D., and
Spiegelman, B. (1987) Cell 49, 835.
Emr,
5852 .
S., and Bassford
Jr.,
P. (1982) J. Biol.
Chem.
257 .
Emr,
Acad.
S., Schwartz, M. ,
. Sci. U.S.A. 25,
, and
5802.
Silhavy, T. (1978)
Proc.
Natl
Emr, S., Hall, M., and Silhavy, T. (1980) J. Cell Biol. 86.
701.

209
Engel, J., Sugarman, B., and Dodgson, J. (1982) Nature 297.
434.
Farmer, S., Wan, K., Ben-Ze'ev, A., and Penman, S. (1983)
Mol. Cell. Biol. 3, 182.
Fiddes, J., and Goodman, H. (1979) Nature 281. 351.
Franke, W., Schmid, E., Osborn, M., and Weber, K. (1979) J.
Cell Biol. 81, 570.
Franke, W., Schiller, D., Moll, R., Winter, S., Schmid, E.,
Engelbrecht, I., Denk, H., Krepler, R., and Platzer, B.
(1981) J. Mol. Biol. 153. 933.
Fuchs, E., and Marchuk, D. (1983) Proc. Natl. Acad. Sci.
Ü.S.A. 80, 4321.
Fulton, A., Wan, K., and Penman, S. (1980) Cell 20, 849.
Gallwitz, D. (1975) Nature 257. 247.
Gallwitz, D., and Mueller, G. (1969) J. Biol. Chem. 244â– 
5947.
Gallwitz, D., and Briendl, M. (1972) Biochem. Biophys. Res.
Commun. 47., 1106.
Garrels, J., and Gibson, W. (1976) Cell 9, 793.
Gilmore, R., Blobel, G., and Walter, P. (1982a) J. Cell
Biol. 95, 463.
Gilmore, R., Walter, P., and Blobel, G. (1982b) J. Cell
Biol. 95, 470.
Glass, W., Briggs, R., and Hnilica, L. (1981) Science 211.
70.
Goldman, R., and Knipe, D. (1973) Cold Spring Harb. Symp.
Quant. Biol. 37., 523.
Goldman, J., Schaumberg, H., and Norton, W. (1978) J. Cell
Biol. 78, 426.
Gorman, C., Moffat, L., and Howard, B. (1982) Mol. Cell.
Biol. 2, 1044.
Graham, F., and van der Eb, A. (1973) Virology 52, 456.
Graves, R. , and Marzluff, W. (1984) Mol. Cell. Biol. 4., 351.

210
Graves, R., Pandey, N., Chodchoy, N., and Marzluff, W.
(1987) Cell 48, 615.
Green, L., Stein, G., and Stein, J. (1984a) Mol. Cell.
Biochem. .60, 123.
Green, L., van Antwerpen, R. , Stein, J., and Stein, G.
(1984b) Science 226â–  838.
Greenberg, M., and Ziff, E. (1984) Nature 311â–  433.
Greenberg, M., Hermanowski, A., and Ziff, E. (1986) Mol.
Cell. Biol. 6, 1050.
Grossi de Sa, M.-F., Martins de Sa, C., Harper, F., Link-
Coux, M., Huesca, M., and Scherrer, K. (1988) J. Cell Biol.
107, 1517.
Gunning, P., Ponte, P. , Okayama, H., Engel, J., Blau, H.,
and Redes, L. (1983) Mol. Cell. Biol. 3, 787.
Hartwig, J., and Stossel, T. (1975) J. Biol. Chem. 250â– 
5696.
Harvey, R., Whiting, J., Coles, L., Krieg, P., and Wells, J.
(1983) Proc. Natl. Acad. Sci. U.S.A. 80, 2819.
Hayden, J., and Allen, R. (1984) J. Cell Biol. 9£, 1785.
Heintz, N. , Zernik, M., and Roeder, R. (1981) Cell 24., 661.
Heintz, N., Sive, H., and Roeder, R. (1983) Mol. Cell. Biol.
3, 539.
Helms, S., Baumbach, L., Stein, G., and Stein, J. (1984)
FEBS Letters 168, 65.
Hentschel, C., and Birnstiel, M. (1981) Cell 25, 301.
Hoffman, P., and Lasek, R. (1975) J. Cell Biol. 66, 351.
Holltrop, M., Raisz, L., and Simmons, H. (1974) J. Cell
Biol. 60, 346.
Hortin, G., and Boime, I. (1981) Cell 24, 453.
Howe, J., and Hershey, J. (1984) Cell 37., 85.
Hubbard, B. , and Lazarides, E. (1979) J. Cell Biol. 8C), 166.
Huberman, E., and Callahan, M. (1979) Proc. Natl. Acad. Sci.
USA 76, 1293.

211
Huxley, H. (1973) Nature 243, 445.
Isaacs, W., and Fulton, A. (1987) Proc. Natl. Acad. Sci.
U.S.A. M, 6174.
Izant, J., and Lazarides, E. (1977) Proc. Natl. Acad. Sci.
U.S.A. 24, 1450.
Jacobs-Lorena, M., Baglioni, C., and Borun, T. (1972) Proc.
Natl. Acad. Sci. U.S.A. 69, 2095.
Jeffery, W. (1984) Dev. Biol. 103, 482.
Kaiser, C., and Botstein, D. (1986) Mol. Cell. Biol. 6,
2382.
Kaiser, C. , Preuss, D., Grisafi, P., and Botstein, D. (1987)
Science 235â–  312.
Kane, R. (1975) J. Cell Biol. 66, 305.
Kedes, L. (1979) Annu. Rev. Biochem. 4_8, 837.
Laemmli, U. (1970) Nature 222, 680.
Lawrence, J., and Singer, R. (1986) Cell ±5, 407.
Lawrence, J., Singer, R., Villnave, C., Stein, J., and
Stein, G. (1988) Proc. Natl. Acad. Sci. U.S.A. 85, 463.
Lazarides, E. (1980) Nature 283, 249.
Lazarides, E., and Hubbard, B. (1976) Proc. Natl. Acad. Sci.
U.S.A. 22, 4344.
Lemieux, R., and Beaud, G. (1982) Eur. J. Bioch. 129â–  273.
Lenk, R. , and Penman, S. (1979) Cell JL6, 289.
Lenk, R., Ransom, L., Kaufmann, Y., and Penman, S. (1977)
Cell 10, 67.
Levine, B., Chodchoy, N., Marzluff, W., and Skoultchi, A.
(1987) Proc. Natl. Acad. Sci. U.S.A. 84, 6189.
Li, W., Reddy, R., Henning, D., Epstein, P., and Busch, H.
(1982) J. Biol. Chem. 252, 5136.
Liautard, J., and Jeanteur, P. (1979) Nucl. Acids Res. 2,
135.

212
Lin, D., Tobin, K., Grumet, M., and Lin, S. (1980) J. Cell
Biol. 84» 455.
Lingappa, V., Chaidez, J., Yost, C., and Hedgpeth, J. (1984)
Proc. Natl. Acad. Sci. U.S.A. 8i, 456.
Luscher, B., Stauber, C., Schindler, R., and Schumperli, D.
(1985) Proc. Natl. Acad. Sci. U.S.A. 82, 4389.
Maniatis T., Fritsch, E., and Sambrook, J. (1982a) in
Molecular Cloning. (Cold Spring Harbor, New York) pp 86-90.
Maniatis T., Fritsch, E., and Sambrook, J. (1982b) in
Molecular Cloning. (Cold Spring Harbor, New York) pp 109-
112.
Marashi, F., Helms, S., Shiels, A., Silverstein, S.,
Greenspan, D., Stein, G., and Stein, J. (1986) Can. J.
Bioch. Cell Biol. M, 277.
Marzluff, W., and Huang, R. (1985) in Transcription and
Translation. Hames, B. and Higgins, S. Eds. (IRL Press,
Oxford) pp 89-129.
Maurer, R., and Mckean, D. (1978) J. Biol. Chem. 253. 6315.
Meyer, D., Krause, E., and Dobberstein, B. (1982a) Nature
297. 647.
Meyer, R., Louvard, D., and Dobberstein, B. (1982b) J. Cell
Biol. 92, 579.
Milcarek, C. , and Penman, S. (1974) J. Mol. Biol. 89, 327.
Milsted, A., Silver, B., Cox, R., and Nilson, J. (1985)
Endocrinology 117. 2033.
Miranda, A., Godman, G. , Deitch, A., and Tannenbaum, S.
(1974a) J. Cell Biol. 61, 481.
Miranda, A., Godman, G., and Tannenbaum, S. (1974b) J. Cell
Biol. 62, 406.
Mitchell, R., Zokas, L., Schreiber, R., and Verma, I. (1985)
Cell £0, 209.
Mohun, T., and Garret, N. (1987) Develop. 101â–  393.
Moll, R., Franke, W., Schiller, D., Geiger, B., and Krepler,
R. (1982) Cell 31, 11.

213
Moon, R., Nicosia, R., Olsen, C., Hille, M., and Jeffery, W.
(1983) Dev. Biol. 95, 447.
Morris, T., Marashi, F., Weber, L. , Hickey, E., Greenspan,
D., Bonner, J., Stein, J., and Stein, G. (1986) Proc. Natl.
Acad. Sci. U.S.A. 83, 981.
Muller, M., and Blobel, G. (1984) Proc. Natl. Acad. Sci.
U.S.A. 81, 7421.
Muller, R., Bravo, R., Burckhardt, J., and Curran, T. (1984)
Nature 312, 716.
Ng, S. (1989) J. Cell Biol. 102, 79a.
Nielsen, P., Goelz, S., and Trachsel, H. (1983) Cell Biol.
Int. Rep. 7, 245.
O'Farrell, P. (1975) J. Biol. Chem. 250. 4007.
Ohlsson, R., and Pfeifer-Ohlsson, S. (1987) Exp. Cell Res.
173. 1.
Ornelles, D., Fey, E., and Penman, S. (1986) Mol. Cell.
Biol. 6, 1650.
Osborn, M. (1983) J. Invest. Dermatol. 8J., 104s.
Pandey, N., and Marzluff, W. (1987) Mol. Cell. Biol. 7,
4557.
Pederson, T., and Robbins, E. (1970) J. Cell Biol. 4¿, 509.
Peltz, S., and Ross, J. (1987) Mol. Cell. Biol. 7, 4345.
Penman, S., Capeo, D., Fey, E., Chatterjee, P., Reiter, T.,
Ermisch, S., and Wan, K. (1983) in The modern cell biology
series: Spatial organization of eukaryotic cells. McIntosh,
J.R., Ed. (Alan R. Liss, New York) pp 385-415.
Perry, R. , and Kelley, D. (1973) J. Mol. biol. 22., 681.
Plumb, M., Stein, J., and Stein, G. (1983) Nucl. Acids Res.
11, 2391.
Prescott, D. (1966) J. Cell Biol. 31, 1.
Rahmsdorf, H., Schonthal, A., Angel, P., Litfin, M., Ruther,
U., and Herrlich, P. (1987) Nucl. Acids Res. 15, 1643.
Randall, L., and Hardy, S. (1989) Science 243â–  1156.

214
Rauscher III, F., Cohen, D., Curran, T., Bos, T., Vogt, P.,
Bohmann, D., Tjian, R., and Franza Jr., B. (1988) Science
240. 1010.
Rave, N., Crkvenjakov, R., and Boedtker, H. (1979) Nucl.
Acids Res. 6, 3559.
Rawn, J. (1983) in Biochemistry. Wasserman, M., and Nickol,
D., Eds. (Harper and Row, New York) p 528.
Rebillard, M-, Leibovitch, S., Jullien, M., Talha, S., and
Harel, L. (1987) Exp. Cell. Res. 172â–  432.
Robb, R., Terhorst, C., and Strominger, J. (1978) J. Biol.
Chem. 253, 5319.
Robbins, E., and Borun, T. (1967) Proc. Natl. Acad. Sci.
U.S.A. 52, 409.
Roberts, C., and Wilson, G. (1985) Focus 2, 16.
Roggenkamp, R., Kustermann-Kuhn, B., and Hollenberg, C.
(1981) Proc. Natl. Acad. Sci. U.S.A. 28, 4466.
Roop, D., Hawley-Nelson, P., Cheng, C., and Yuspa, S. (1983)
Proc. Natl. Acad. Sci. U.S.A. SO, 716.
Ross, J., and Kobs, G. (1986) J. Mol. Biol. 188â–  579.
Ross, J. Peltz, S. Kobs, G., and Brewer, G. (1986) Mol.
Cell. Biol. 6, 4362.
Ross, J., Kobs, G., Brewer, G., and Peltz, S. (1987) J.
Biol. Chem. 262. 9374.
Rovera, G., Santoli, D., and Damsky, C. (1979) Proc. Natl.
Acad. Sci. U.S.A. 21, 1779.
Rothstein, T. (1985) J. Immunology 135â–  106.
Rothstein, T. (1986) J. Immunology 136â–  813.
Rubenstein, P., and Spudich, J. (1977) Proc. Natl. Acad.
Sci. U.S.A. 21, 120.
Sanger, J. (1975) Proc. Natl. Acad. Sci. U.S.A. T2, 2451.
Sassone-Corsi, P., and Verma, I. (1987) Nature 326â–  507.
Schachner, M., Hedley-White, E., Hsu, D., Schoonmaker, G.,
and Bignami, A. (1977) J. Cell Biol. 15, 67.

215
Schatz, G. , and Butow, R. (1983) Cell 32., 316.
Scherrer, K., and Darnell, J. (1962) Biochem. Biophys. Res.
Comm. 2i 486.
Schliwa, M. (1982) J. Cell Biol. 92, 79.
Schmidt, W. , Pardue, R., Tutt, M. , Briggs, R., Brinkley, B.,
and Hnilica, L. (1982) Proc. Natl. Acad. Sci. U.S.A. 79.
3138.
Selden, L., Gershman, L., and Estes, J. (1980) Bioch.
Biophys. Res. Commun. 95, 1854.
Shiels, A., Marashi, F., Stein, J., and Stein, G. (1985)
Biochem. Biophys. Res. Comm. 127. 239.
Sierra, F., Lichtler, A., Marashi, F., Rickies, R., Van
Dyke, T., Clask, S., Wells, J., Stein, J., and Stein, G.
(1982) Proc. Natl. Acad. Sci. U.S.A. 79, 1795.
Simon, K., Perara, E., and Lingappa, V. (1987) J. Cell Biol.
104. 1165.
Sittman, D., Graves, R., and Marzluff, W. (1983) Proc. Natl.
Acad. Sci. U.S.A. 80, 1849.
Sive, H., Heintz, N., and Roeder, R. (1984) Mol. Cell. Biol.
4, 2723.
Sloboda, R. , Dentler, W., and Rosenbaum, J. (1976)
Biochemistry 15, 4497.
Sood, A., Pereira, D., and Weissman, S. (1981) Proc. Natl.
Acad. Sci. U.S.A. 78, 616.
Southern, P., and Berg, P. (1982) J. Mol. Appl. Gen. 1, 327.
Spaulding, J., Kajiwara, K., and Mueller, G. (1966) Proc.
Natl. Acad. Sci. U.S.A. 56, 1535.
Stahl, H. , and Gallwitz, D. (1977) Eur. J. Biochem. 72., 385.
Stauber, C., Luscher, B., Eckner, R., Lotscher, E., and
Shumperli, D. (1986) EMBO J. 5, 3297.
Stein, G., and Borun, T. (1972) J. Cell Biol. 52, 292.
Stein, G., and Stein, J. (1984) Mol. Cell. Biochem. 64, 105.
Stein, J., Stein, G., and McGuire, P. (1977) Biochemistry
16, 2207.

Stein, J., Thrall, C., Park, W., Mans, R., and Stein, G.
(1975) Science 189, 557.
216
Stein, G., Sierra, F., Stein, J., Plumb, M., Marashi, F.,
Carozzi, N., Prokopp, K., and Baumbach, L. (1984) in Histone
Genes. Stein, G., Stein J., and Marzluff, W. Eds. (John
Wiley and Sons, New York) pp 397-455.
Stein, G., Lian, J., Stein, J., Briggs, R. , Shalhoub, V.,
Wright, K., Pauli, 0., and van Wijnen, A. (1989) Proc. Natl.
Acad. Sci. U.S.A. 86, 1865.
Steinert, P. (1975) Biochem. J. 149, 39.
Steinert, P., and Gullino, M. (1976) Biochem. Biophys. Res.
Commun. 70, 221.
Stimac, E., Groppi, V., and Coffino, P. (1983) Biochem.
Biophys. Res. Commun. 114. 131.
Summerhayes, I.
(1981) J. Cell
,, Cheng, Y.-S.,
Biol. 90, 63.
Sun,
, T-T.
, and Chen,
L.
Sun, T-T., and
Green, H. (1978)
1 J.
Biol.
Chem. 253â– 
2053
Sun, T-T., Eichner, R., Nelson, W., Tseng, S.( and Weiss, R.
(1983) J. Invest. Dermatol. 81, 109s.
Sytoyama, C., Frunzio, R., Liau, G., Mudryj, M., and De
Crombrugghe, B. (1986) Proc. Natl. Acad. Sci. U.S.A. 83.
3213.
Talmadge, K., Stahl, S., and Gilbert, W. (1980a) Proc. Natl.
Acad. Sci. U.S.A. 77, 3369.
Talmadge, K., Kaufman, J., and Gilbert, W. (1980b) Proc.
Natl. Acad. Sci. U.S.A. 77, 3988.
Tannenbaum, J., and Godman, G. (1983) Mol. Cell. Biol. 3,
132.
Thomas, P. (1980) Proc. Natl. Acad. Sci. U.S.A. 72., 5201.
Tilney, L., and Detmers, P. (1975) J. Cell Biol. 66, 508.
Toh, B., Lolait, S., Mathy, J-P., and Baum, R. (1980) Eur.
J. Cell Biol. 22, 376.
Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl.
Acad. Sci. U.S.A. 26, 4350.
Tullis, R., and Rubin, H. (1980) Anal. Biochem. 107â–  260.

217
Ullu, E., Murphy, S., and Melli, M. (1982) Cell 29, 195.
Vale, R., Schnapp, B., Reese, T., and Sheetz, M. (1985) Cell
40, 449.
van Venrooij, W., Sillekens, P., van Eekelen, A., and
Reinders, R. (1981) Exp. Cell Res. 135. 79.
Venkatesan, N., and Steele, W. (1972) Biochem. Biophys. Acta
287 â–  52 6.
Verma, I., and Sassone-Corsi, P. (1987) Cell 51, 513.
Walter, P., and Blobel, G. (1980) Proc. Natl. Acad. Sci.
U.S.A. 77, 7112.
Walter, P., and Blobel, G. (1981) J. Cell Biol. 91, 557.
Walter, P., and Blobel, G. (1982) Nature 299â–  691.
Walter, P., and Blobel, G. (1983) Cell 34, 525.
Walter, P. , Gilmore, R. , and Blobel, G. (1984) Cell 38,, 5.
Wang, Y-L. (1984) J. Cell Biol. 99, 1478.
Watson, M. (1984) Nucl. Acids Res. 12, 5145.
Weber, K., Bibring, T., and Osborn, M. (1975) Exp. Cell Res.
95, 111.
Welch, W., and Feramisco, J. (1985) Mol. Cell. Biol. 5,
1571.
Wells, D., and Redes, L. (1985) Proc. Natl. Acad. Sci.
U.S.A. 82, 2834.
Whalen, R. , Butler-Browne, G., and Gros, F. (1976) Proc.
Natl. Acad. Sci. U.S.A. 73., 2018.
Wiedman, M., Huth, A., and Rapoport, T. (1984) Nature 309.
637.
Wilson, M. , and Melli, M. (1977) J. Mol. Biol. 110â–  511.
Wolosewick, J., and Porter, K. (1977) Am. J. Anat. 149â–  197.
Wolosewick, J. , and Porter, K. (1979) J. Cell Biol. 82, 114.
Woodcock-Mitchell, J., Eichner, R., Nelson, W., and Sun, T-
T. (1982) J. Cell Biol. 95, 580.

218
Wu, R., and Bonner, W. (1985) Mol. Cell. Biol. 5, 2959.
Yoshizaki, K., Nakagawa, T., Kaieda, T., Muraguchi, A.,
Yamamura, Y., and Kishimoto, T. (1982) J. Immunology 128â– 
1296.
Zambetti, G., Stein, J., and Stein, G. (1986) Proc. Natl.
Acad. Sci. U.S.A. 84, 2683.
Zambetti, G., Schmidt, W., Stein, G., and Stein, J. (1985)
J. Cell. Phys. 125, 345.
Zoller, M., and Smith, M. (1983) Methods Enzymol. 100. 468.
Zumbe, A., Stahli, C., and Trachsel, H. (1982) Proc. Natl.
Acad. Sci. U.S.A. 29, 2927.

BIOGRAPHICAL SKETCH
Gerard Paul Zambetti was born in Queens, New York, on
August 4, 1958. He attended Archbishop Molloy High School
(Queens, New York) and graduated in the spring of 1976. He
then attended the State University of New York at
Plattsburgh and graduated with a Bachelor of Science degree
in biochemistry and biophysics in the spring of 1980. He
then studied bacteriophage Lambda DNA replication and gene
regulation with Dr. Robert Shuster at Emory University
(Atlanta, Georgia) and graduated with a Master of Science
degree in biochemistry. He continued to research the
regulation of gene expression with Drs. Janet and Gary Stein
in the Department of Immunology and Medical Microbiology at
the University of Florida. In the fall of 1984, he married
Stacey Ann Chapman in Atlanta, Georgia, and on May 6, 1987,
they were blessed with a beautiful daughter, Shannon Kelly
Zambetti.
219

I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as a dissertation for the
degree of Doctor of Philosophy.
íanet L. Stein, Chair
'Professor of Immunology and
Medical Microbiology
I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as^aidisser^ation for the
degree of Doctor of Philosophy.
5áry S. ¡in
-''Professor of Biochemistry
and Molecular Biology
I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as a dissertation for the
degree of Doctor of Philosophy.
I
ümes B. Flanegan
(professor of Immunóiogy and
Medical Microbiology
I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as a dissertation for the
degree of Doctor of Philosophy.
-’fhomas C. Rowe
Assistant Professor of
Pharmacology and Therapeutics
I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as a dissertation for the
degree of Doctor of Philosophy.
Richard W. Moyer/
Professor of Immuholo y and
Medical Microbiology

This dissertation was submitted to the Graduate Faculty of the
College of Medicine and to the Graduate School and was accepted as
partial fulfillment of the requirements for the degree of Doctor
of Philosophy.
December 1989
Dean, College of Medicine
Dean, Graduate School

UNIVERSITY OF FLORIDA
3 1262 08557 0066



Figure 5-2. Northern blot analysis of the cvtoskeleton and
soluble phase distribution of c-fos mRNA in cytochalasin D.
puromvcin. CD/puro. and CD/cvcloheximide treated cells.
HeLa cells were cultured in the presence of the drugs
and fractionated into cytoskeleton and soluble phases as
described in the Chapter 2. Equal quantities of RNA (10
g/sample) from each fraction were assayed for c-fos mRNA
content by Northern blot analysis. Lanes: 1) Control/csk? 2)
Control/sol? 3) CD/csk? 4) CD/sol? 5) Puro/csk; 6) Puro/sol;
7) CD/puro/csk; 8) CD/puro/sol; 9) CD/cyclo/csk? 10)
CD/cyclo/sol.


165
the signal peptide of SPH3E1 was changed to TTG (Figure 4-
5). Translation should therefore bypass this altered start
codon and initiate at the normal ATG codon in the histone
coding region, resulting in the synthesis of histone protein
without the signal peptide sequence. Initial studies using a
HeLa monoclonal cell line expressing the SPH3E1ATG" gene
demonstrated the localization of SPH3E1ATG" mRNA on
nonmembrane-bound polysomes, which indicates that in vivo
the signal peptide was not synthesized (Fig. 4-6) As seen
in Figure 6-1, greater than 90% of SPH3E1ATG mRNA is
associated with the cytoskeleton in the control cell
culture. Cytochalasin D and puromycin cotreatment
effectively released hCGcr and HLA-B7 mRNA from the
cytoskeleton with little or no effect on the association of
the SPH3E1ATG mRNA with the cytoskeleton (Figures 6-1 and
6-3, Table 6-1). Approximately 77% of SPH3E1ATG" mRNA and
less than 35% of HLA and hCGa mRNA remained attached to the
cytoskeleton under these conditions. Regardless of whether
the SPH3E1ATG' mRNA could be translated in the presence of
cytochalasin D and puromycin, no signal peptide would be
synthesized to anchor the message to the cytoskeleton. This
result further supports the proposal that the nucleotide
sequences coding for the signal peptide express a
cytoskeleton attachment site which is distinct from the
cytoskeletal attachment site of endogenous histone mRNA.


37
supernatant was transferred to a 150 ml Corex tube (Corning)
and nucleic acids were precipitated from solution with 30 ml
isopropanol at room temperature for 15 minutes. The DNA was
pelleted by centrifugation at 5,000 rpm in Beckman JS-7.5
rotor for 30 minutes at 20C. The pellet was briefly dried
under vacuum and resuspended in 8 ml 10:1 TE buffer (10 mM
Tris pH 8.0, 1 mM EDTA). The DNA solution was adjusted to a
final concentration of 1 mg/ml of CsCl (refractive index of
1.3860) and 600 Mg/ml ethidium bromide. The DNA solution was
then transferred to a 16 x 76 mm Quick-Seal tube (Beckman)
and centrifuged to equilibrium at 45,000 rpm in a Beckman
Type-50 Ti rotor and Beckman L5-65 ultracentrifuge for 36-48
hours at 20C. The supercoiled plasmid DNA was removed from
the cesium chloride gradient by puncturing the side of the
tube just below the DNA band with a syringe containing a 22
gauge needle. The DNA was extracted repeatedly with 1-
butanol to remove the ethidium bromide and to reduce the
volume of the solution. Cesium chloride was removed by
dialysis against 10:1 TE buffer at 4C. The DNA was
quantitated by absorbance at 260 nm and stored at -20C.
Routinely, the yield for pBR322 derived plasmids was 500-800
jig DNA per 500 ml amplified culture.
M13 Bacteriophage DNA Preparations
M13 is a filamentous DNA bacteriophage that is specific
for male L. coli bacteria. M13 phage DNA is useful for site


74
(w/v) EDTA and 0.5 ml of the cell suspension was seeded into
20 ml completed EMEM containing 500 /xg/ml biologically
active Geneticin (G418, Gibco) in 10 cm tissue culture
dishes. The cells were incubated at 37C, 5% C02 and refed
every 3-4 days with completed EMEM containing G418
antibiotic. Within 2-3 weeks, the majority of the cells had
died and individual colonies of G418 resistant HeLa cells
had formed. Polyclonal cell lines were established by
combining the G418 resistant colonies from a 10 cm dish as a
single, heterogeneous cell population (usually 20-40
colonies). For the establishment of monoclonal cell lines,
individual colonies were picked from the dish with sterile
pasteur pipettes and cultured as separate cell lines.


41
Exponentially growing HeLa cells (5 x 107 cells) were
harvested by centrifugation at 1,500 rpm for 5 minutes at
37C in an IEC rotor. The cells were washed in phosphate-
buffered saline (PBS; 150 mM NaCl, 10 mM sodium phosphate pH
6.8) and resuspended in 3 ml lysis buffer (2 mM Tris-HCl pH
7.4, l"mM EDTA) containing 5 /Lig/ml PVS. Cells were lysed in
the presence of 2.4% SDS and 88 /xg/ml of proteinase K at
room temperature for 15 minutes. Lysates were adjusted to
300 mM NaCl, extracted with 5 ml Tris-saturated phenol and 5
ml CHCl3/isoamyl alcohol (IAA) (24:1, v/v) and centrifuged
at 3,000 rpm in an IEC rotor at room temperature. The
aqueous phase was then extracted with 5 ml CHC13/IAA and
centrifuged at 3,000 rpm in an IEC rotor at room
temperature. The phenol/chloroform and chloroform
extractions were repeated as described above. Nucleic acids
were precipitated from the aqueous phase with 2.5 volumes of
95% ethanol at -20C in the presence of 53 mM potassium
acetate.
Nucleic acids were recovered from the ethanol
suspension by centrifugation at 12,000 x g and resuspended
in TCM buffer (10 mM Tris-Hcl, pH 7.4, CaCl2, 10 mM MgCl2)
for digestion with DNase I. The DNase I was dissolved in 20
mM Tris-HCl pH 7.4, 10 mM CaCl2 at 1 mg/ml, preincubated at
37C for 20 minutes and treated with 0.1 volume of
proteinase K [1 mg/ml] at 37C for 2 hours to remove
ribonuclease activity (Tullis and Rubin, 1980). The nucleic


fos Induction
177
70
60
CDPuro
control


157
a)
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16


146
mRNA to the cytoskeleton (Cervera et al.. 1981; Howe and
Hershey, 1984). Consistent with this finding, 79% of HLA-B7
mRNA and 85% of hCGa mRNA remained attached to the
cytoskeleton in puromycin treated HeLa cells (Table 5-2). In
addition, 79% of the H3 and H4 histone poly A mRNA and 83%
of the"c-fos mRNA remained associated with the cytoskeleton
in HeLa cells treated with puromycin (Table 5-2).
Discussion
There is growing support for the hypothesis that the
cytoskeleton of eukaryotes is more than a structural
scaffold. Evidence is accumulating to suggest that the
cytoskeleton is a dynamic structure that plays an integral
role in the process of protein synthesis. In addition, the
cytoskeleton may influence the posttranscriptional
regulation of eukaryotic genes by compartmentalizing mRNAs
within the cytoplasmic space. It is therefore of importance
to study the interactions of eukaryotic mRNAs with the
cytoskeleton.
The experiments presented here were designed to examine
the association of specific mRNAs with the cytoskeleton. As
described previously and confirmed in this study, histone
mRNA and HLA-B7 mRNA were found predominantly associated
with the cytoskeleton (Zambetti et al.. 1985). In addition,
hCGa and c-fos mRNAs were also localized in the cytoskeleton
fraction. Recently it was shown that the association of poly


23
attention on the posttranscriptional regulation of histone
gene expression. The evidence cited above suggests that the
cytoplasm of eukaryotic cells is a highly organized region
which contains subcellular compartments that concentrate
translation factors as well as specific mRNAs (Fulton et
al.. 1980; Howe and Hershey, 1984; Lawrence and Singer,
1986; Ornelles et al.. 1986). Traditionally, the idea of
subcellular compartments refers to organelles such as the
nucleus, endoplasmic reticulum, and mitochondria. For
example, a popular biochemical definition of
compartmentation is "a simple regulatory mechanism in which
the enzymes for catabolism and biosynthesis are physically
separated by membrane-bound cell organelles" (Rawn, 1983). A
broader definition of subcellular compartmentation includes
any specialized region of the cell regardless of the
presence of a membrane and may be extended to such
components and regions as 1) nonmembrane-bound and membrane-
bound polysomes, 2) cytoskeleton and soluble phase, and 3)
spatial location (i.e., near the nucleus versus near the
plasma membrane).
Histone mRNAs and the factors that are involved in
their selective degradation during the inhibition of DNA
synthesis may be localized in a particular subcellular
compartment. Consistent with this reasoning, the subcellular
compartment may influence the posttranscriptional control of
histone gene expression. Based on this possibility, we have


4
(Robbins and Borun, 1967; Spaulding et al.. 1966; Stein and
Borun, 1972). In vitro translation studies using polysomal
RNA isolated from cells at various stages of the cell cycle
indicated that the histone mRNA was present exclusively
during S phase (Spalding et al.. 1966; Gallwitz and Mueller,
1969; Jacobs-Lorena et al.. 1972; Gallwitz and Briendl,
1972; Perry and Kelley, 1973; Borun et al.. 1975).
Quantitation of histone mRNA content during the cell cycle
by Northern blot analysis, using radiolabeled histone cDNA
and genomic clones, demonstrated that histone mRNA steady
levels closely parallel the rates of histone protein
synthesis as well as the rates of DNA replication (Heintz et
al.. 1983; Plumb et al.. 1983a; Baumbach et al. 1984). The
timing of histone protein synthesis during DNA synthesis is
not coincidental and represents a functional relationship
between these two processes. This is supported by the
following observations: 1) there is a coordinate induction
of histone protein synthesis and DNA synthesis during
stimulation of quiescent cells to proliferate (DeLisle et
al.. 1983; Green et al.. 1984a); 2) inhibition of DNA
synthesis with compounds such as hydroxyurea or cytosine
arabinoside results in a loss of histone mRNA with a
concomitant decrease in histone protein synthesis (Breindl
and Gallwitz, 1973; Butler and Mueller, 1973; Borun et al..
1975; Baumbach et al.. 1984) and 3) terminal differentiation
is marked by a down regulation in histone protein synthesis


31
gene. The recombinant plasmid F0108A contains an entire
human H4 histone gene which includes the promoter, coding
and 3' non-transcribed regions.
An HLA-B7 cDNA clone, pDPOOl, was generously supplied
by Dr. Sherman Weissman, Yale University, New Haven,
Connecticut (Sood et al.. 1981). The pDPOOl plasmid DNA was
constructed from the human cell line RPMI 4265 and contains
a 1400 bp insert coding for HLA-B7 in the PstI site of
pBR322 plasmid DNA. Therefore, the plasmid confers
tetracycline resistance and not ampicillin resistance to E.
coli HB101 bacteria.
The phCGa plasmid DNA was generously provided by Dr.
John Fiddes, California Biotechnology Inc., Mountain View,
California and Dr. John Nilson, Case Western Reserve
University, Cleveland, Ohio. The phCGa plasmid is a cDNA
clone containing a 621 bp fragment coding for the alpha
subunit of human chorionic gonadotropin inserted into the
Hindlll site of pBR322 and confers ampicillin resistance
(Fiddes and Goodman, 1979).
The beta-lactamase signal peptide-globin fusion gene,
pSP125e, was kindly provided by Dr. Vishwanath Lingappa,
University of California at San Francisco. The pSP125e clone
contains the signal peptide from the beta-lactamase gene of
E. coli plasmid pBR322 fused in frame to the coding sequence
for chimpanzee alpha globin. This fusion gene was subcloned
into the pSP64 vector (Promega Biotech) as a Hindlll-PstI


Abstract of Dissertation Presented to the Graduate School of the
University of Florida in Partial Fulfillment of the Requirements
for the Degree of Doctor of Philosophy
THE ROLE OF SUBCELLULAR COMPARTMENTS AND STRUCTURES
IN EUKARYOTIC GENE EXPRESSION
By
Gerard Paul Zambetti
December 1989
Chairperson: Janet Stein
Major Department: Immunology and Medical Microbiology
Histone mRNAs naturally reside on nonmembrane-bound
polysomes that are associated with the cytoskeleton. To
study whether the subcellular location plays a role in the
coupling of histone mRNA stability to DNA synthesis, a
signal peptide-histone fusion gene was constructed that
encodes a histone mRNA that is directed to membrane-bound
polysomes. The signal peptide-histone fusion mRNA when
localized in the foreign subcellular compartment is no
longer destabilized when DNA synthesis is inhibited. A
single nucleotide substitution in the signal peptide-histone
fusion gene by site-directed mutagenesis resulted in the
localization of the histone fusion mRNA on nonmembrane-bound
polysomes and restored the coupling of mRNA stability with
xi


216
Stein, J., Thrall, C., Park, W., Mans, R., and Stein, G.
(1975) Science 189, 557.
Stein, G., Sierra, F., Stein, J., Plumb, M., Marashi, F.,
Carozzi, N., Prokopp, K., and Baumbach, L. (1984) in Histone
Genes, Stein, G., Stein J., and Marzluff, W. Eds. (John
Wiley and Sons, New York) pp 397-455.
Stein, G., Lian, J., Stein, J., Briggs, R., Shalhoub, V. ,
Wright, K., Pauli, U., and van Wijnen, A. (1989) Proc. Natl.
Acad* Sci. U.S.A. 86, 1865.
Steinert, P. (1975) Biochem. J. 149, 39.
Steinert, P., and Gullino, M. (1976) Biochem. Biophys. Res.
Commun. 70, 221.
Stimac, E., Groppi, V., and Coffino, P. (1983) Biochem.
Biophys. Res. Commun. 114. 131.
Summerhayes, I., Cheng, Y.-S., Sun, T-T., and Chen, L.
(1981) J. Cell Biol. 90, 63.
Sun, T-T., and Green, H. (1978) J. Biol. Chem. 253. 2053.
Sun, T-T., Eichner, R., Nelson, W., Tseng, S., and Weiss, R.
(1983) J. Invest. Dermatol. 81, 109s.
Sytoyama, C., Frunzio, R., Liau, G., Mudryj, M., and De
Crombrugghe, B. (1986) Proc. Natl. Acad. Sci. U.S.A. 83.,
3213.
Talmadge, K., Stahl, S,, and Gilbert, W. (1980a) Proc. Natl.
Acad. Sci. U.S.A. 77, 3369.
Talmadge, K., Kaufman, J., and Gilbert, W. (1980b) Proc.
Natl. Acad. Sci. U.S.A. 77, 3988.
Tannenbaum, J., and Godman, G. (1983) Mol. Cell. Biol. 3,
132.
Thomas, P. (1980) Proc. Natl. Acad. Sci. U.S.A. 72, 5201.
Tilney, L., and Detmers, P. (1975) J. Cell Biol. 66, 508.
Toh, B., Lolait, S., Mathy, J-P., and Baum, R. (1980) Eur.
J. Cell Biol. 22, 376.
Towbin, H., Staehelin, T., and Gordon, J. (1979) Proc. Natl.
Acad. Sci. U.S.A. 76, 4350.
Tullis, R., and Rubin, H. (1980) Anal. Biochem. 107. 260.


152
may play an important part in the subcellular localization
of messenger RNA, which may ultimately affect
posttranscriptional regulation. At this time, we can not
rule out the possibility that mRNAs localized in different
regions of the cytoplasm are associated with the
cytoskeleton by different mechanisms. Additional studies are
currently underway to define further the association of mRNA
with the cytoskeleton and what influence this structure has
on the regulation of eukaryotic gene expression.


i 15
S P H 3
1 2
ALPHA ATG"
12 12
p robe
fu sion
H 3


148
and/or ribosome that is complexed with the endoplasmic
reticulum. The possibility that the nascent polypeptide
serves as an additional cytoskeleton attachment site for
membrane-bound polysomes is supported by the observation
that nascent polypeptides undergoing translocation into the
lumen of the endoplasmic reticulum are resistant to protease
digestion following detergent extraction (Connolly et al..
1989). The protection of the nascent polypeptides from
protease digestion, in the absence of membranes, appears to
be due to the association of the polypeptide with the
proteins of the translocation apparatus of the endoplasmic
reticulum.
To release membrane-bound polysomal mRNAs efficiently
from the cytoskeleton, both cytoskeletal attachment sites
must be disrupted. Dissociation of the polysomes with
puromycin alone fails to release membrane-bound polysomal
mRNA from the cytoskeleton. In this case, the mRNA remains
associated with the cytoskeleton through the cytochalasin D
sensitive site even though the polysomes are no longer
anchored to the endoplasmic reticulum. Consistent with this
reasoning, cytochalasin D treatment alone results in only a
partial release of membrane-bound polysomal mRNA from the
cytoskeleton. The partial release of membrane-bound
polysomal mRNA from the cytoskeleton in cytochalasin D
treated cells is most likely a result of the release of mRNA
from the cytochalasin D sensitive cytoskeleton attachment


54
Schuell Elutip column. The column was washed with 5 ml low
salt buffer to remove unincorporated radiolabeled
ribonucleotide and the DNA fragment was eluted in 400 ul
high salt buffer (Tris-HCl pH 7.4, 1 M NaCl, 1 mM EDTA). The
radiolabeled SI probe was usually used immediately and any
excess "was stored in high salt buffer at 4C.
Routinely, 10-25 nq RNA (30 ul) and 10 ul SI probe were
co-precipitated with 100 ul 95% ethanol at -20C overnight.
The sample was centrifuged at 12,000 x g, 4C for 15-30
minutes. The pellet was vacuum dried and resuspended in 10
ul 5x hybridization buffer (0.2 M Pipes pH 6.4, 2 M NaCl, 5
mM EDTA) and 40 ul recrystallized formamide. The sample was
heated at 90C for 10 minutes and then incubated at 55C for
3 hours. Single-stranded nucleic acids were digested with SI
nuclease by adding 400 ul ice cold SI nuclease buffer (0.03
M sodium acetate pH 4.6, 0.25 M NaCl, 1 mM ZnSOi) and 900
units SI nuclease at 37C for 30 minutes. The SI digestion
reaction was terminated by organic extraction with 0.5
volume phenol and 0.5 volume chloroform and then 0.5 volume
chloroform alone. Nucleic acids were precipitated from the
aqueous phase with 1 ml 95% ethanol at -20C overnight.
The SI nuclease digested samples were collected at
12,000 x g, 4C for 30 minutes. The pellets were vacuum
dried and resuspended in 5 ul loading buffer (80% formamide,
lx TBE [lOx TBE: 500 mM Tris-HCl pH 8.3, 500 mM boric acid,
10 mM EDTA], 0.1% bromophenol blue, 0.1% xylene cyanol). The


185
cytochalasin D and thereby raise the effective concentration
of other transcripts. To address these possibilities, we
examined the transcription of the 18S rRNA gene as well as
other genes during CD treatment. As seen in Fig. 7-4, the
transcription of 18S ribosomal RNA and cell cycle dependent
histone H4 genes is not significantly affected by
cytochalasin D. In addition, CD has no effect on the
transcription of genes coding for HLA-B7 (Fig. 7-5), beta-
globin and a cell cycle independent H2B histone (data not
shown). These results suggest that the increase in
transcription of c-fos and beta-actin genes during
cytochalasin D treatment is a selective response and not an
effect on transcription in general.
Transcriptional Induction of c-Fos During CD Treatment does
not Require New Protein Synthesis
The increase in c-fos gene transcription during CD
treatment may be a secondary effect of the drug.
Cytochalasin D may induce the synthesis of a protein that in
turn regulates the transcriptional activity of the c-fos
gene. To test this possibility, HeLa cells were pre-treated
with puromycin for 15 minutes prior to treatment with
cytochalasin D. The cells were then incubated in the
presence of both drugs for an additional 15 minutes. Cell
cultures treated with puromycin alone were incubated for 30
minutes in the presence of the inhibitor while control cells


199
research groups (Stein and Stein, 1984; Wu and Bonner,
1985), a critical concentration of unbound histone protein
may be required for the destabilization of histone mRNA
during inhibition of DNA synthesis. Localization of histone
mRNA in clusters may create pockets within the cytoplasm in
which £he unbound histone protein could reach the necessary
concentration for the destabilization of histone mRNA to
occur.
The alteration in subcellular location of the histone
fusion mRNA has been demonstrated by subcellular
fractionation and SI nuclease protection analysis (Chapter
4). It remains to be determined by in situ hybridization
analysis whether the spatial distribution of the histone
fusion mRNA has been significantly altered compared with
that of endogenous histone mRNA. Furthermore, identification
and purification of the components involved in the
posttranscriptional regulation of histone gene expression
may provide the opportunity to monitor directly the
subcellular localization of these factors by
immunofluorescence, cell fractionation and biochemical
analysis.
Another aspect concerning the localization of histone
mRNA in the cytoplasm, which warrants additional
consideration, is whether a single class or multiple classes
of histone mRNAs are contained within a cluster. For
example, a particular cluster may contain only H3 histone


103
<
<
III
z
z
CO
Q
a
X
a
v>
(/>
V)
a
1
2
3
eg
OJ
UJ
Ui
UJ
CO
CO
CO
r
X
X
a
a
CL
a
CO
a
a o. a
4 5 6
M
344
220
HYBRID H3 mRNA
130
ENDOGENOUS H3mRNA
75


167
(Jeffery, 1984; Bagchi et al., 1987), Furthermore, prosomes,
which are small ribonucleoprotein complexes that are
associated with translationally inactive mRNAs, are attached
to the cytokeratin-intermediate filaments in HeLa cells
(Grossi de Sa et al.. 1988). The role of the intermediate
filaments in the attachment of the chimeric mRNAs to the
cytoskeleton remains to be determined.
In Chapter 5, we demonstrated that the signal peptide-
histone fusion mRNA (SPH3E1) is relatively stable, as
compared to endogenous H3 histone mRNA, during inhibition of
DNA synthesis. The uncoupling of signal peptide-histone
fusion mRNA stability from DNA synthesis does not appear to
be a consequence of its association with the cytoskeleton
through the cytochalasin D and puromycin insensitive site.
The mutated signal peptide-histone fusion mRNA (SPH3E1ATG")
is efficiently destabilized during inhibition of DNA
synthesis and most likely expresses the same CD and
puromycin insensitive cytoskeletal attachment site as SPH3E1
mRNA.
In summary, a cytoskeletal attachment site has been
detected which is associated with a signal peptide-histone
chimeric mRNA. This site appears to be a direct or indirect
result of the nucleotide sequences coding for the signal
peptide. Studies on deletion and point mutants of the signal
peptide-histone chimeric mRNAs should prove useful in
providing insight into the elements involved in the
attachment of the mRNA to the cytoskeleton.


Figure 2-2. Outline of procedure for the isolation of
subcellular polvsomal fractions.


202
necessary to clarify these interpretations of the
cytochalasin D-hydroxyurea experiments. For example, sucrose
density gradient analysis of soluble phase RNA from
cytochalasin D treated cells should demonstrate whether
histone mRNA is released from the cytoskeleton in the same
manner as poly-A+ RNA. Furthermore, gel filtration studies
as described by Adams et al. (1983) on the composition of
the histone mRNA in the soluble phase after cytochalasin D
treatment should also provide information on whether the
histone mRNA, although no longer anchored to the
cytoskeleton, may still be complexed with cytoskeletal
components. These cytoskeletal components may be necessary
for histone mRNA destabilization during inhibition of DNA
replication.
The distribution of HLA-B7 mRNA between the
cytoskeleton and soluble fractions during cytochalasin D and
hydroxyurea cotreatment was also analyzed as an internal
control for equal loading of the RNA. Surprisingly, in view
of the results of Ornelles et al. (1986), cytochalasin D
treatment failed to release HLA-B7 mRNA from the
cytoskeleton (Chapter 5). Subsequent studies demonstrated
that polysomal HLA-B7 mRNA, as well as other membrane-bound
polysomes, are released from the cytoskeleton during
cytochalasin D and puromycin cotreatment (Chapter 5).
Puromycin alone or cytochalasin D and cycloheximide
cotreatment does not release membrane-bound polysomes from


170
The cytoskeleton also appears to play a role in the
transition between proliferation and the expression of
differentiated phenotypes. For example, normal B cells can
be induced to proliferate by co-stimulation with anti
immunoglobulin and B cell growth factor (Yoshizaki et al..
1982). "Cytochalasin D (CD) can substitute for B cell growth
factor and provide the signal, presumably through the
perturbation of the cytoskeleton, to induce proliferation of
anti-immunoglobulin treated normal B cells (Rothstein, 1985;
Rothstein, 1986). Also, promyelocytic leukemia HL60 cells
cultured in the presence of tumor-promoting phorbol esters
undergo rearrangement of the cytoskeleton and differentiate
into non-dividing macrophage-like cells (Rovera et al..
1979? Huberman et al.. 1979). Although the effects of
phorbol esters on cultured cells are many, it is possible
that the alteration in the cytoskeletal structure may be
involved in the progression of the differentiation pathway
(Diamond et al., 1980; Penman et al., 1983).
To define further the role of the cytoskeleton in gene
regulation, we have studied the effects of disrupting the
cytoskeleton with cytochalasin D on the regulation of a
series of genes. We report here that cytochalasin D
treatment of exponentially growing HeLa cells rapidly
induces the transcription of the c-fos gene. This increase
in transcription is accompanied by an increase in c-fos mRNA
levels. The effect of cytochalasin D on c-fos gene


Figure 3-2. Two-dimensional gel electrophoresis of
cvtoskeleton and soluble associated proteins.
HeLa cells were fractionated into a cytoskeleton and
soluble phase, protein was extracted from each phase and
subsequently analyzed by two-dimensional gel electrophoresis
as described in Chapter 2. Proteins of the soluble fraction
(Sol), A, were separated using wide range (pH 3-7)
ampholines while cytoskeletal proteins (Csk), B, were
separated with narrow range (pH 5-7) ampholines to more
clearly identify cytoskeletal species. V: vimentin, a:
actin, and t: tubulins. Unlabeled arrows in B depict HeLa
cytokeratins of 54 Kd (pi 6.0); 52.5 Kd (pi 6.1); 46 Kd (pi
5.1); and 45 Kd (pi 5.7), which were spots highly reactive
with commercially prepared prekeratin or stratum corneum
antisera (Franke et al.. 1981).


154
the cell enabled us to examine mRNA-cytoskeleton
interactions in more detail. The results presented here
indicate that the signal peptide-histone fusion mRNA is
associated with the cytoskeleton in a manner unlike that
reported for membrane-bound and nonmembrane-bound polysomes.
These results are consistent with the existence of a
cytoskeletal attachment element within the mRNP or mRNA
primary sequence and provide additional evidence for
heterogeneity in mRNA-cytoskeleton interactions.
Results
To investigate further the mechanism of mRNA attachment
to the cytoskeleton, we have studied and compared the
cytoskeletal association of endogenous nonmembrane-bound
polysomal histone mRNA with that of membrane-bound
polysomal, signal peptide-histone chimeric mRNA. In Chapter
4, we have described the construction of a signal peptide-
histone chimeric gene (SPH3E1) that when transfected into
HeLa cells is transcribed and the resulting chimeric histone
mRNAs are targeted to membrane-bound polysomes (Zambetti et
al.. 1986). Although the signal peptide fused to the histone
gene is derived from the beta-lactamase gene of the EL_ coli
plasmid pBR322, it is functional in HeLa cells (Chapter 4).
Figure 4-5 schematically diagrams endogenous H3 histone and
the signal peptide-histone (SPH3E1) mRNAs.


9
1987). Correcting for translation read through in these
mutants by inserting stop codons in all three reading frames
at the natural translation termination site restores the
coupling of mRNA stability to DNA replication.
Detailed studies on deletion mutants suggest that the
histone nascent polypeptide is not involved in the
posttranscriptional regulatory process (Capasso et al..
1987) Systematic removal of approximately 100 nucleotide
blocks spanning nearly the entire protein coding region had
no effect on the destabilization of these mutant histone
mRNAs when DNA synthesis was inhibited. Consistent with this
result, the deletion of the ATG translation start codon
which prevents the synthesis of histone protein did not
prevent the destabilization of the mRNA during inhibition of
DNA replication. Furthermore, as discussed above, the
minimal target sequences containing the stem-loop structure
confer instability to gamma globin mRNA when DNA synthesis
is interrupted (Pandey and Marzluff, 1987). The globin-
histone chimeric mRNA, although regulated as wild type
histone mRNA, does not encode any histone protein sequences.
Destabilization of histone mRNA appears to occur in a
stepwise manner initiating at the 3' terminus (Ross et al..
1986). The first detectable histone mRNA degradation product
lacks 5 nucleotides from the 3' terminus (Ross and Kobs,
1986; Ross et al. 1986). This degradation product is
detectable in the in vitro mRNA decay system as well as in


Figure 7-3. Time course of c-fos mRNA accumulation during
cvtochalasin D and puromvcin treatments in exponentially
growing HeLa cells.
HeLa cells were treated with cytochalasin D [10 nq/ml],
puromycin [0.4 mM] or cytochalasin D and puromycin for the
indicated times. The cells were collected and total cellular
RNA was isolated. The levels of c-fos mRNA were determined
by Northern blot analysis as described in Chapter 2. Each
lane represents 10 /g total cellular RNA. Lanes: 1)
control/0 min.? 2) control/60 min.? 3) DMSO control/60 min.;
4) CD/15 min.; 5) CD/30 min.; 6) CD/45 min.; 7) CD/60 min.;
8) CD/90 min.; 9) CD/120 min.; 10) CDPuro/15 min.; 11)
CDPuro/30 min.; 12) CDPuro/45 min.; 13) CDPuro/60 min.; 14)
CDPuro/90 min.; 15) CDPuro/120 min.; 16) Puro/15 min.; 17)
Puro/45 min.; 18) Puro/60 min.; 19) Puro/90 min.; 20)
Puro/120 minutes.


93
shuttling of newly synthesized histone polypeptides into the
nucleus and c) histone mRNA stability.
In agreement with long-standing in vitro translation
and cDNA hybridization results (Jacobs-Lorena et al.. 1972?
Liautard and Jeanteur, 1979; Gallwitz and Briendl, 1972?
Borun et al.. 1975; Stein et al.. 1975), we find that
histone mRNAs reside predominantly on nonmembrane-bound
polysomes. While we cannot dismiss the possibility that the
amounts of histone mRNAs detected in membrane-bound
polysomal subcellular fractions are biologically significant
(see Figures 3-1 and 3-3), we feel that this is a
consequence of the isolation procedure which operationally
defines the fraction obtained. The absence of detectable
HLA-B7 mRNAs on nonmembrane-bound polysomes suggests that
while mRNAs encoding intracellular proteins may be trapped
by the endoplasmic reticulum during this subcellular
fractionation technique, there is little, if any, stripping
of membrane-bound polysomes encoding extracellular
polypeptides, such as the HLA-B7 class I antigen. Within
this context, the question arises whether the transcripts
from histone genes that are not expressed in a cell cycle-
dependent manner (approximately 8-10%) are represented on
membrane-bound polysomes. However, this is not a likely
possibility because inhibition of DNA replication with
hydroxyurea brings about a comparable degradation of histone
mRNAs in all the subcellular fractions examined.


137
subcellular fractions demonstrated that approximately 75% of
the chorionic gonadotropin alpha (hCGa) mRNA was associated
with membrane-bound polysomes (autoradiograph not shown).
Analogous to HLA-B7 mRNA, hCGa mRNA was not fully released
from the cytoskeleton in cytochalasin D treated cells?
greater than 90% of the gonadotropin mRNA was associated
with the cytoskeleton in control cells and approximately 77%
remained associated with the cytoskeleton in cells treated
with 10 nq/TC 1 cytochalasin D (autoradiograph not shown; data
summarized in Table 5-1).
Nonmembrane-Bound Polvsomal mRNAs are Released from the
Cytoskeleton bv Cytochalasin D Treatment
Previous results have demonstrated that histone mRNAs
are predominantly translated on cytoskeleton associated,
nonmembrane-bound polysomes (Zambetti et al., 1985) and as
described above, are released from the cytoskeleton into the
soluble phase by cytochalasin D treatment. As an additional
control for studying the release of nonmembrane-bound
polysomal RNAs from the cytoskeleton by cytochalasin D
treatment, we analyzed the distribution of c-fos mRNA
between the cytoskeleton and soluble fractions from control
and cytochalasin D treated cell cultures. The c-fos protein
is localized within the nucleus (Curran et al.. 1984) and
therefore, according to the signal hypothesis, should be
synthesized on nonmembrane-bound polysomes (Blobel and


208
Challoner, P., Moss, S., and Groudine, M. (1989) Mol. Cell.
Biol. 9, 902.
Clarke, M., and Spudich, J. (1977) Annu. Rev. Biochem. 46.
797.
Clayton, D., and Darnell Jr., J. (1983) Mol. Cell. Biol. 3,
1552.
Cline,JK. (1986) J. Biol. Chem. 261, 14804.
Cochran, B., Zullo, J., Verma, I., and Stiles, C. (1984)
Science 226. 1080.
Collart, D., Stein, G., and Stein, J. (1985) Mol. Cell.
Bioch. 67. 161*
Connolly, T., Collins, P., and Gilmore, R. (1989) J. Cell
Biol. 108, 299.
Curran, T., Miller, A., Zokas, L., and Verma, I. (1984) Cell
36, 259.
Curran, T., and Morgan, J. (1985) Science 229. 1265.
Date, T., Zwizinski, C., Ludmerer, S., and Wickner, W.
(1980) Proc. Natl. Acad. Sci. U.S.A. 77, 827.
Delange, R., Fambrough, D., Smith, E. and Bonner, J. (1969)
J. Biol. Chem. 244, 5669.
DeLisle, A., Graves, R., Marzluff, W., and Johnson, L.
(1983) Mol. Cell. Biol. 3, 1920.
Delacourte, A., Filleatreau, G., Boulteau, F., Biserte, G.,
and Schrevel, J. (1980) Biol. J. 191, 543.
Diamond, L., O'Brien, T., and Baird, W. (1980) Adv. Cancer
Res. 32., 1.
Distel, R., Ro, H.-S., Rosen, B., Groves, D., and
Spiegelman, B. (1987) Cell 49, 835.
Emr, S., and Bassford Jr., P. (1982) J. Biol. Chem. 257,
5852.
Emr, S., Schwartz, M., and Silhavy, T. (1978) Proc. Natl.
Acad. Sci, U.S.A. T5, 5802.
Emr, S., Hall, M., and Silhavy, T. (1980) J. Cell Biol. 86.
701.


204
useful in the identification of an mRNA-cytoskeleton
attachment site.
The role of the cytoskeleton structure in influencing
gene expression is supported by the observation that
cytochalasin D treatment resulted in a rapid and dramatic
increase in c-fos gene transcription with a concomitant
increase in c-fos mRNA cellular levels (Chapter 7).
Consistent with previous results (Tannenbaum and Godman,
1983), CD treatment also increased beta-actin gene
transcription. However, these modifications in gene
expression appear to be selective and not a general
phenomenon, since the transcription of several other genes
was unaffected by the drug treatment. The increase in c-fos
mRNA cellular levels was larger than the increase in c-fos
gene transcription and may be due to a stabilization of the
c-fos message under these conditions. Cell fractionation and
Northern blot analysis demonstrated that c-fos mRNA is
associated with the cytoskeleton in control cells and
accumulates in the soluble phase in cytochalasin D treated
cells (Chapter 5). Stabilization of c-fos mRNA may be due to
the altered subcellular location of the message during
cytochalasin D treatment. Additional studies are needed to
define the molecular basis for the role of the cytoskeletal
structure in the transcriptional and posttranscriptional
regulation of c-fos gene expression.


13
There are at least five major classes of protein that
constitute the intermediate filaments: 1) desmin, 2)
vimentin, 3) keratin, 4) neurofilaments and 5) glial
filaments. Biochemical and immunological studies
demonstrated that these proteins are generally expressed in
a cell"type specific manner (for review see Lazarides,
1980). Desmin is an acidic protein with a molecular weight
of 53,000 Daltons and is predominantly found in intermediate
filaments from cardiac, smooth and skeletal muscle cells
(Lazarides and Hubbard, 1976; Hubbard and Lazarides, 1979).
The localization of desmin at the Z line in muscle suggests
a role for this protein in the coordination of the
contractile actions of the muscle fiber (Lazarides and
Hubbard, 1976; Izant and Lazarides, 1977).
Vimentin is found principally in intermediate filaments
from cells of mesenchymal origin. Immunological and
biochemical characterization of vimentin indicated that the
protein is 52,000 Daltons and widely conserved throughout
evolution (Brown et al.. 1976; Bennett et al.. 1978; Franke
et al.. 1979). A characteristic feature of the vimentin
intermediate filaments is their sensitivity to colchicine
which erroneously led researchers to believe initially that
they were disaggregation products of the microtubules
(Goldman and Knipe, 1973; Holltrop et al.. 1974; Blose and
Chacko, 1976). The collapse of vimentin into filamentous
bundles around the nucleus during colchicine treatment


8
all known cell cycle dependent histone mRNAs (Hentschel and
Birnstiel, 1981). The presence of the stem-loop structure is
not sufficient by itself to couple mRNA stability with DNA
replication. The target sequences must be present in the
proper position at the 3' terminus of the mRNA to be
operative. Aberrant histone mRNA transcripts containing the
target sequences at an internal position are fully stable
during the inhibition of DNA synthesis (Alterman et al..
1985; Luscher et al.. 1985; Pandey and Marzluff, 1987;
Levine et al. 1987). Furthermore, moving the stem-loop
structure 380 nucleotides downstream from the translation
stop codon abolishes the coupling of histone mRNA stability
with DNA replication despite the retention of the stem loop
structure at the 3'terminus (Graves et al.. 1987).
An additional constraint on the histone
posttranscriptional regulatory mechanism is that translation
must proceed to within close proximity of the natural
termination codon for destabilization to occur during
inhibition of DNA synthesis (Graves et al.. 1987). Frame-
shift mutations in histone mRNA resulting in the premature
termination of translation more than 300 nucleotides from
the 3' terminus stabilize the messenger RNA when DNA
synthesis is interrupted (Graves et al.. 1987). Other frame-
shift mutants that allow translation to continue past the
natural UGA stop codon into the 31 untranslated region are
also stable when DNA synthesis is inhibited (Capasso et al..


68
with the alkali conditions), in order, the following
solutions each containing 0.2 N NaOH, 1 M NaCl, 2 mM EDTA:
1) 1 ml 20% sucrose? 2) 1 ml 17.5% sucrose; 3) 1 ml 15%
sucrose; 4) 1 ml 10% sucrose; 5) 5% sucrose. The tube was
incubated at 4C for 2 hours. The samples were centrifuged
througt the alkaline sucrose gradients in a Beckman SW 50.1
rotor at 37,000 rpm for 2 hours at 4C without the brake.
The samples were recovered from the gradients by puncturing
the bottom of the tube with a 21 gauge needle and collecting
7 drops per fraction in 1.5 ml microcentrifuge tubes
(usually -60 fractions). Fractions containing covalently
closed, double-stranded DNA were determined by Cherenkov
scintillation counting in the tritium channel, pooled and
neutralized with 1M Tris-citrate pH 5.
Isolation of Mutated M13 Recombinant Bacteriophage
The DNA was transfected into E^_ coli JM101 competent
cells, plated onto YT agar and incubated overnight at 37C.
Individual plaques were picked and bacteriophage were
subsequently prepared. Bacteriophage suspensions (2 ul) were
spotted onto a 0.45 micron nitrocellulose filter and
prehybridized in 6x SSC, lOx Denhardt's, 0.2% SDS (10 ml/100
cm2) at 67C for 1 hour. The filter was rinsed in 50 ml 6x
SSC for 1 minute at 21C and hybridized in 10 ml 6x SSC, lOx
Denhardt's containing 1x10 cpm radiolabeled oligonucleotide
at 21C for 1 hour. The filter was washed three times in 50
ml 6x SSC at 21C for a total of 10 minutes. The filter was


53
containing the Smal fragment was butanol extracted to reduce
the volume and to remove the ethidium bromide. The fragment
was then extracted with phenol and chloroform and
precipitated in 2.5 volumes 95% ethanol at -20C overnight.
The DNA was collected by centrifugation at 12,000 x g at 4C
for 15"minutes and resuspended in 5 ul 10 mM Tris-HCl pH
8.0, 5 ul lOx CIP buffer (0.5 M Tris-HCl pH 8.0, 10 mM
MgCl2, 1 mM ZnCl2, 10 mM spermidine) and 38 ul ddH20. The
blunt-ended DNA was dephosphorylated with 0.12 units calf
intestinal alkaline phosphatase (CIP) at 37C for 15 minutes
and then at 55C for 15 minutes. A second aliquot of CIP was
added and the incubations at 37 and 55C were repeated. The
reaction was terminated with 4 0 ul ddH20, 10 ul lOx STE (100
mM Tris-HCl pH 8.0, 1 M NaCl, 10 mM EDTA) and 5 ul 10% SDS
and heated at 68C for 15 minutes. The sample was extracted
with phenol and chloroform, adjusted to 0.25 M sodium
acetate and precipitated with 2.5 volumes of 95% ethanol at
-20C overnight. The dephosphorylated Smal fragment was
pelleted at 12,000 x g, 4C for 15 minutes, vacuum dried and
resuspended in 12 ul lx kinase buffer (66 mM Tris-HCl pH
9.5, 10 mM MgCl2, 10 mM beta-mercaptoethanol, 2 mM
spermidine). The SI fragment was phosphorylated with 50 uCi
[gamma-32P]ATP and 1 ul polynucleotide kinase [30 units/ul]
at 37C for 30 minutes. The reaction was terminated with 0.5
ul 100 mM EDTA at 70C for 10 minutes. The DNA solution was
diluted with 1 ml elutip low salt buffer (20 mM Tris-HCl pH
7.4, 0.2 M NaCl, 1 mM EDTA) and passed over a Schleicher and


30
Protein synthesis was inhibited with 10 /ng/ml
cycloheximide for varying lengths of time. Cycloheximide
stock solutions [500 ig/ml] were prepared fresh in IX
spinner salts. Protein synthesis was also inhibited with 0.4
mM puromycin for varying lengths of time. Puromycin stock
solutions [54.4 mg/ml] were prepared in dimethyl sulfoxide
(DMSO) and stored at -2 0C.
To study the dependence of protein synthesis on the
destabilization of histone mRNA during the inhibition of DNA
synthesis, we pre-treated cells with 10 Mg/ml cycloheximide
for 5 minutes and then added 1 mM hydroxyurea for an
additional 30 minutes.
Disruption of the cytoskeleton was generally achieved
by treating the cells with 5 to 60 lig/ml cytochalasin D for
15-20 minutes at 37C. The cytochalasin D stock solutions
[5-10 mg/ml] were prepared in DMSO and stored at -20C.
Plasmid DNA
The human histone genes used in this study were
previously isolated by Sierra et al. from a lambda Ch4A
bacteriophage library (1982). The histone genes were
subseguently subcloned into Escherichia coli pBR322 plasmid
DNA and propagated in E_s_ coli strain HB101 bacterium (Sierra
et al.. 1982; Plumb et al.. 1983; Stein et al.. 1984).
Recombinant plasmids ST519 and FF435C each contain a
distinct genomic DNA fragment encoding a human H3 histone


Figure 7-1. Effects of cvtochalasin D and puromycin
treatment on steady state levels of c-fos and chorionic
gonadotropin mRNA.
Exponentially growing Hela cells were treated with
cytochalasin D [10 /g/ml], puromycin [0.4 mM], or
cytochalasin D [10 ig/ml] and puromycin [0.4 mM] together
for 15 minutes. The cells were harvested and total cellular
RNA was isolated. Chorionic gonadotropin (hCGa) and c-fos
mRNA levels were determined by Northern blot analysis as
described in Chapter 2. Each lane represents 10 ig total
cellular mRNA. Lanes: 1) control? 2) cytochalasin D for 15
min.; 3) puromycin for 15 min.? and 4) CD and puromycin for
15 minutes.


107
The ability to direct a cell cycle-dependent histone
mRNA to membrane-bound polysomes provided the possibility to
address the involvement of subcellular location in the
coupling of histone mRNA stability with DNA replication. As
seen in Figure 4-4 (lanes 1-4), endogenous histone mRNA
levels"were reduced by greater than 90%, as determined by
densitometric analysis of the autoradiogram, in HeLa cells
following inhibition of DNA replication by treatment with 1
mM hydroxyurea for 60 minutes. In contrast, inhibition of
DNA replication by hydroxyurea treatment does not result in
a reduction of fusion message levels on membrane-bound
polysomes (lane 4). The finding that the signal peptide-
histone fusion mRNA is stable following DNA synthesis
inhibition is further supported by the SI nuclease analysis
of total cellular RNA (Figure 4-4). The endogenous histone
mRNA levels in hydroxyurea treated HeLa cells were reduced
by approximately 95% compared with those in untreated
HeLa cells (lanes 5-8) (with each lane representing an
independent transfected cell culture). In contrast, only a
4% reduction in fusion mRNA levels was measured (lanes 5-8).
These results demonstrate that the incorporation of the E.
coli beta-lactamase signal peptide into a human H3 histone
gene is sufficient to target the encoded fusion mRNA to the
membrane-bound polysomes and to confer stability to the
fusion message when DNA synthesis is inhibited.


Figure 2-1. Outline of the cloning scheme for the
construction of the signal peptide-histone fusion gene.
Restriction endonuclease sites are as follows: B,
BglII; E, EcoRI; H, HindIII; N, Ncol; P, PstI; S, Smal.
Abbreviations: S.P., coli pBR322 beta-lactamase signal
peptide coding sequences; H3REG, transcriptional H3 histone
regulatory region including CAAT and TATAA consensus
sequences; CAP, histone H3 mRNA transcription start site; H3
struct, sequences coding for H3 histone protein; ATG,
translation start codon; TAA, translation stop codon; SP6,
bacteriophage SP6 promoter.


171
expression does not appear to be a general phenomenon since
the transcription and steady state levels of several other
active and inactive genes are unaffected by the drug
treatment. These results imply that the cell monitors the
organization and/or composition of the cytoskeleton and
adjust "the regulation of c-fos gene expression accordingly.
This finding is consistent with signal transduction models
proposing that the cytoskeleton may be involved in relaying
stimuli from the plasma membrane to the nucleus.
Results
Cvtochalasin D Increases c-Fos mRNA Levels
Previous reports have suggested that the cytoskeleton
plays a role in signal transduction and the regulation of
gene expression. To further investigate this functional
property of the cytoskeleton, we have disrupted the
microfilaments of HeLa S3 cells with cytochalasin D and
measured the effects on transcription and mRNA cellular
levels of a series of genes. Within minutes of the addition
of cytochalasin D, cell morphology is drastically altered as
the cytoskeleton rearranges without any apparent
quantitative or qualitative loss of protein (Miranda et al..
1974a; Miranda et al., 1974b; Schliwa, 1982; Ornelles et
al.. 1986). It has been proposed that cytochalasin D acts by
binding to actin, causing the capping or fragmentation of
the microfilaments (Selden et al.. 1980; Lin et al.. 1980).


UNIVERSITY OF FLORIDA
3 1262 08557 0066


Figure 3-1. Northern blot analysis of H3 histone, H4 histone
and HLA-B7 mRNA in subcellular fractions.
The cytoskeleton and soluble fractions, as well as the
nonmembrane-bound and membrane-bound polysomal RNA
fractions, were isolated from HeLa cells by the procedures
described in Chapter 2 (Cervera et al., 1981; Venkatesan and
Steele, 1972) In addition, total polysomal RNA and total
cellular RNA were isolated (Plumb et al.. 1983, Venkatesan
and Steele, 1972) RNA was extracted from each subcellular
fraction and 50 fsq of RNA from each preparation was resolved
by 1.5% agarose, 6% formaldehyde gel electrophoresis. The
RNA was transferred to nitrocellulose and hybridized to
radiolabeled pFF435C (H3 histone and pFO108A (H4 histone)
probes, followed by hybridization to j2P-labeled (nick-
translated) pDPOOl (cDNA of HLA-B7) probe. The hybridized
filters were then exposed to preflashed XAR5 x-ray film at -
70C for 72 hours. From left to right, lanes: 1) total
cellular RNA? 2) Csk RNA; 3) Sol phase RNA; 4) NRER? 5)
MRER; 6) nonmembrane-bound polysomal RNA? 7) LER; and 8)
total polysomal RNA.


Figure 4-10. Quantitation of SPH3E1. SPH3ElalPha. SPH3E1ATG
mRNA during hvdroxvurea treatment.
The cellular levels of the signal peptide-histone
fusion mRNAs and endogenous H3 histone mRNA during
hydroxyurea treatment were determined by densitometric
analysis of the SI nuclease protection assays presented in
Figures 4-7 to 4-9. Values are presented as percentage of
control cells (, SPH3E1 mRNASPH3Elalpha mRNA;,
SPH3E1ATG mRNA?0, endogenous H3 histone mRNA).


Figure 3-4. Northern blot analysis of H3 histone. H4 histone
and HLA-B7 mRNAs associated with the cytoskeleton and
soluble fractions from HeLa cell treated with metabolic
inhibitors.
Exponentially growing HeLa cells were incubated in the
presence of metabolic inhibitors and fractionated into
cytoskeleton and soluble fractions as described in Chapter
2. RNA was extracted and assayed for H3 histone, H4 histone
and HLA-B7 mRNA content by Northern blot analysis as
described in Figure 3-1. From left to right, lanes: 1)
control, Csk; 2) control, Sol; 3} cycloheximide (Cy), Csk;
4) Cy,"Sol; 5) hydroxyurea (HU), Csk; 6) HU, Sol; 7) Cy/HU,
Csk; and 8) Cy/HU, Sol.


Figure 3-3. Northern blot analysis of H3 histone. H4 histone
and HLA-B7 mRNAs isolated from subcellular fractions of
cells treated with metabolic inhibitors.
Exponentially growing HeLa cells were incubated in the
presence of metabolic inhibitors and fractionated into
nonmembrane-bound and membrane-bound polysomes as described
in Chapter 2. RNA was extracted from each fraction and
assayed for H3 histone mRNA, H4 histone mRNA and HLA-B7 mRNA
content by Northern blot analysis as described in Figure 3-1
and Chapter 2. Upper panel from left to right, lanes: 1)
control (untreated), nonmembrane-bound polysomes; 2)
contro, NRER; 3) control, MRER; 4) control, LER (omitted);
5) cycloheximide (Cy), nonmembrane-bound polysomes; 6) Cy,
NRER; 7) Cy, MRER; and 8) Cy, LER. Lower panel from left to
right, lanes: 1) hydroxyurea (HU), nonmembrane-bound
polysomes; 2) Hu, NRER; 3) Hu, MRER; 4) Hu, LER; 5) CY/HU,
nonmembrane-bound polysomes; 6) Cy/Hu, NRER; 7) Cy/Hu, MRER;
and 8) Cy/Hu, LER.


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID E7JEYIHGD_AM9AO5 INGEST_TIME 2012-12-07T22:58:22Z PACKAGE AA00012891_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


192
cytochalasin D and puromycin co-treated cells is greater
than the levels of c-fos mRNA in CD treated and puromycin
treated cells combined. Furthermore, the time course of c-
fos mRNA accumulation in cytochalasin D treated cells is
more rapid and shorter lived than that observed in puromycin
treated cells.
Stimulation of c-fos transcription by cytochalasin D
treatment can occur during inhibition of protein synthesis
and therefore appears to be dependent on factors that are
already present within the treated cells. Pretreatment with
puromycin for 15 minutes followed by the addition of
cytochalasin D for 15 minutes results in a marked increase
in c-fos transcription (Fig. 7-5). Furthermore, inhibition
of protein synthesis for 30 minutes with puromycin alone
results in a significant increase in c-fos transcription.
These results are consistent with previous reports
suggesting that the c-fos gene is negatively regulated by a
labile factor (Mitchell et al.. 1985? Greenberg et al..
1986? Sassone-Corsi and Verma, 1987).
The increase in the cellular levels of c-fos mRNA
during cytochalasin D treatment is greater than the measured
increase in the transcription of the gene and may be due to
the stabilization of c-fos mRNA under these conditions.
However, it is possible that during cytochalasin D treatment
the mechanism involved in the turnover of c-fos mRNA remains
operative despite the reduced levels of protein synthesis.


Figure 6-1. Cvtochalasin D and puromycin cotreatment does
not release SPH3E1 or SPH3E1ATG' mRNA from the cvtoskeleton.
Clonal cell lines expressing SPH3E1 or SPH3E1ATG"
mRNA were treated with 10 ig/ml cytochalasin D, 0.4 mM
puromycin or 10 y.q/ml cytochalasin D and 0.4 mM puromycin.
The distribution of endogenous histone and fusion mRNAs
within the cytoskeleton and soluble phase fractions was
determined by SI nuclease protection analysis as described
in Chapter 2. Analysis of a) SPH3E1 and SPH3E1ATG mRNAs and
b) endogenous H3 histone mRNA. Lanes 1-8 represent RNA
samples isolated from SPHE1 expressing cells and lanes 9-16
represent RNA samples isolated from SPH3E1ATG" cell
cultures. Lanes 1 and 9, control-csk RNA; lanes 2 and 10,
control-sol RNA; lanes 3 and 11, CD-csk RNA; lanes 4 and 12,
CD-sol RNA; lanes 5 and 13, puro-csk RNA; lanes 6 and 14,
puro-sol RNA; lanes 7 and 15, CD/puro-csk RNA; lanes 8 and
16, CD/puro-sol RNA.


S P H 3 E 1 alpha
3 4 5 6 7 8
123
MW 1 2
probe
fusion
H3


28
T4 polynucleotide kinase (cloned) was from United States
Biochemical Corp. (USB), Cleveland, OH; cytochalasin D,
polyoxyethylene sorbitan monopalmitate (Tween 40), sodium
deoxycholate (NaDOC), sodium dodecyl sulfate (SDS), Triton
X-100, and DNase I (DN-EP) were purchased from Sigma, St.
Louis,"MO; lysozyme was purchased from Worthington
Biochemicals, Freehold, N.J.; fetal bovine, calf bovine and
horse sera were obtained from Gibco Laboratories, Grand
Island, NY or Flow Laboratories, McLean, VA; cycloheximide
was kindly provided by Upjohn, Kalamazoo, Michigan.
Conditions for Enzyme Reactions
The buffers and incubation conditions for each of the
restriction endonucleases and DNA modifying enzymes used in
this study were as described by the specific recommendations
of the manufacturers unless otherwise indicated.
Mammalian Cell Culture
HeLa Monolayer Cell Cultures
HeLa S3 suspension cells (human cervical carcinoma cell
line; ATCC ccl 2.2) were seeded into EMEM (Eagles-modified
minimum essential medium) containing 5% fetal calf serum, 5%
horse serum, 100 U/ml penicillin, 100 /g/ml streptomycin and
1 mM glutamine at 3x10 cells per 10 cm tissue culture dish
and incubated at 37C under 5% C02. Cells were maintained at
sub-confluent densities by splitting the cultures 1:10 into
fresh medium every 3 to 5 days.


32
fragment and is under control of the SP6 promoter (Simon et
al.. 1987).
To study the regulation of histone mRNA localized on
membrane-bound polysomes, we constructed a beta-lactamase
signal peptide-histone fusion gene as outlined in Figure 2-
1. The"first intermediate, pSPpSTdeltaHH/E, involved the
placement of the E_j_ coli pBR3 22 beta-lactamase signal
peptide-globin fusion gene from pSP125e under
transcriptional control of the cell cycle-dependent H3
histone regulatory region of pST519deltaH. The 840 bp
EcoRI/Hindlll fragment from pST519deltaH (Marashi et al..
1986) was isolated electrophoretically and blunt-ended using
the Klenow fragment of DNA polymerase I. The signal peptide-
globin fusion construct was digested with Bglll restriction
endonuclease, which cuts once in the untranslated leader
region of the fusion gene, blunt-ended with Klenow fragment
and dephosphorylated using calf intestinal alkaline
phosphatase. The 840 bp blunt-ended fragment from pST519 was
ligated into the Bglll blunt-ended site of pSP125e using T4
DNA ligase. The construction of the second intermediate,
pSPpSTdeltaHH/EpST, involved the substitution of the
chimpanzee globin coding sequences in pSPpSTdeltaHH/E with
the H3 histone coding sequences from pST519. The
pSPpSTdeltaHH/E DNA was digested with Ncol, which cuts at
the signal peptide-globin coding region junction, and
treated with calf intestinal phosphatase. The pST519 DNA was


10
whole cells (Ross et al. 1986). It is unclear whether this
product arises from a single endonucleolytic scission or
from multiple cuts removing one or more of the nucleotides
at a time. Degradation continues rapidly in a 3' to 5'
direction carried out by an exonuclease which is part of the
ribosome complex (Ross et al.. 1987). These results are
consistent with the requirement for translation to proceed
near the 3' terminus for the destabilization of the histone
mRNA during inhibition of DNA synthesis.
Subcellular Compartments
The appearance of histone mRNA on polysomes within
minutes after transcription and the equally rapid transfer
of the newly synthesized histone proteins into the nucleus
where they complex with DNA imply that histone protein
synthesis may be compartmentalized in a region located just
outside the nucleus (Spaulding et al.. 1966; Borun et al..
1967; Robbins and Borun, 1967). The asymmetric distribution
of eukaryotic mRNAs in the cytoplasm supports this
possibility. Morphological studies initially demonstrated
that polyribosomes are concentrated in areas surrounding the
nucleus in mouse 3T3 cells (Fulton et al.. 1980). Later,
localization of specific mRNAs by in situ hybridization
techniques provided direct evidence that mRNAs are
differentially distributed throughout the cell. Tubulin mRNA
is preferentially located in the periphery of the cytoplasm


60
polymerase I enzyme. In addition, the specific activity (1-2
x 103 cpm//ng DNA) achieved with the nick-translation
procedure is much lower than the random primer method (1 x
103 cpm/jug DNA) .
DNase I [1 mg/ml in 10 mM HCl] was activated by
incubating 1 ul of the enzyme in 9 ul DNase buffer (10 mM
Tris-HCl pH 7.5, 5 mM MgCl2, 1 mg/ml BSA) at 0C for 2
hours. The isotope (40-80 uCi [a-j2P] dCTP) was dried in a
Savant Speed Vac centrifuge. The following components were
added, in order, to the isotope : 1) 12.25 ul ddH20; 2) 2.5
ul lOx nick-translation buffer (500 mM Tris-HCl pH 7.8, 100
mM beta-mercaptoethanol, 50 mM MgCl2) ; 3) 5 ul lOx dNTPs
(100 uM of each dATP, dGTP, dTTP); 4) 1.25 ul BSA [1 mg/ml];
5) 2.5 ul DNA (100 ng/ul]; and 6) 0.5 ul DNA polymerase I [5
U/ul]. The activated DNase I was diluted with 990 ul ice
cold ddH20 (1:1,000 dilution; 1 ng/ul final concentration)
and 1 ul of diluted enzyme was added to the nick-translation
mix. The sample was incubated at 14C for 45 minutes and the
reaction was then terminated with 20 ul 250 mM EDTA and 155
ul TE buffer. The sample was extracted once with 200 ul
chloroform and unincorporated radiolabeled
deoxyribonucleotides were removed by chromatography over a
Biogel A-15m column. The DNA was quantitated and immediately
used for hybridization.


DNA synthesis. These results indicate that the subcellular
location of histone mRNA plays an important role in the
posttranscriptional regulation of histone gene expression.
We have determined that the association of histone mRNA
with the cytoskeleton is dependent on the integrity of actin
filaments. Disruption of the microfilaments with
cytochalasin D efficiently released histone mRNA from the
cytoskeleton into the soluble phase. In contrast, membrane-
bound polysomal mRNAs remained attached to the cytoskeleton
during cytochalasin D treatment. Subseguent studies have
demonstrated that membrane-bound polysomal mRNAs are
attached to the cytoskeleton at two different sites: 1) a
cytochalasin D sensitive site and 2) a puromycin sensitive
site.
Disruption of the cytoskeleton with cytochalasin D
resulted in a rapid and marked stimulation in transcription
of the c-fos proto-oncogene and an increase in c-fos mRNA
cellular levels. Transcription of several other genes was
unaffected by the cytochalasin D treatment. These results
suggest that the nucleus can respond to the structural
organization of the cytoskeleton and selectively adjust the
regulation of gene expression accordingly.
xii


CHAPTER 4
SUBCELLULAR LOCATION OF HISTONE mRNA PLAYS A ROLE IN THE
POSTTRANSCRIPTIONAL REGULATION OF HISTONE GENE EXPRESSION
Introduction
The association of histone mRNA-containing nonmembrane-
bound polysomes with the cytoskeleton may provide a
structural basis for the localization of the mRNA in
specific regions of the cytoplasm* Consistent with this
reasoning, the factors that mediate the rapid and selective
destabilization of histone mRNA during inhibition of DNA
synthesis may be co-localized with the message in the same
subcellular compartment. Therefore, the subcellular
localization of histone mRNA may play an important role in
its posttranscriptional regulation.
In this chapter, we describe a biological approach,
using a series of signal peptide-histone fusion genes, to
study whether the subcellular location of histone mRNA-
containing polysomes is functionally related to the coupling
of histone mRNA stability with DNA replication. In order to
study the influence of subcellular location on histone mRNA
stability, we constructed a signal peptide-histone gene
which was designed to target the encoded fusion message to
96


193
Previous studies on cytochalasin D treated HeLa cells
demonstrated the selective destabilization of histone mRNA
during inhibition of DNA synthesis (unpublished data). This
event is strictly dependent on protein synthesis and
demonstrates the competence of cytochalasin D treated cells
to carry out certain posttranscriptional processes. The
apparent stabilization of c-fos mRNA during cytochalasin D
treatment may therefore be a result of the release of the
mRNA from the cytoskeleton in monosome or mRNP form. Release
of c-fos mRNA from the cytoskeleton into the soluble phase
could physically separate the message from the factors
involved in its turnover. These regulatory factors may be
sequestered and concentrated by the cytoskeleton or an
integral component of the ribosome.
The mechanism of induction of c-fos gene expression by
cytochalasin D treatment is not known. Many of the inducers
of c-fos transcription studied to date appear to effect
protein kinase C and/or adenylate cyclase enzyme activities
(Verma and Sassone-Corsi, 1987). The induction of c-fos gene
expression by cytochalasin D may also occur as an indirect
effect on protein kinase C or adenylate cyclase levels
although this has not yet been determined. Alternatively,
the disruption of the cytoskeleton by cytochalasin D may
also perturb the nuclear matrix which may in turn affect
transcription. However, a model which includes the
alteration in transcription as a result of the perturbation


39
within the phage DNA. The cell culture (1.5 ml) was
transferred to a microcentrifuge tube and spun in an
Eppendorf centrifuge at 12,000 x g for 10 minutes at 4C.
The supernatant (1.2 ml) was transferred to a sterile
microcentrifuge tube, 300 ul of 20% PEG, 2.5 M NaCl were
added and the sample was incubated at room temperature for
15 minutes. The remaining 200-300 ul of supernatant were
removed and stored at -20C as a phage stock. The phage
suspension was centrifuged at 12,000 x g for 5 minutes at
room temperature and the supernatant was decanted. The
pellet was resuspended in 150 ul TES buffer (20 mM Tris-HCl
pH 7.5, 10 mM NaCl, 0.1 mM Na2EDTA) The sample was
extracted with phenol as follows: 1) added 50 ul phenol and
vortexed for 2 seconds? 2) incubated at room temperature for
5 minutes? 3) vortexed for 2 seconds? and centrifuged at
12,000 x g for 4 minutes at room temperature. DNA was
precipitated from 130 ul of aqueous phase with 5 ul 3 M
sodium acetate and 340 ul 95% ethanol on crushed dry ice for
15 minutes and collected by centrifugation at 12,000 x g,
for 15 minutes at 4C. The supernatant was discarded and the
pellet was washed in 70% ethanol at room temperature. The
sample was centrifuged at 12,000 x g, room temperature for 5
minutes. The DNA pellet was dried under vacuum, resuspended
in 80 ul TES buffer and stored at -20C.


ACKNOWLE DGEMENTS
I would like to acknowledge the dedication of Drs.
Janet and Gary Stein to their work. Their commitment to
science has provided us with a well-equipped lab and the
funds to pursue our research. I would also like to extend
special thanks to the Stein's for their constant support and
guidance during my graduate studies. In addition, I would
like to thank Drs. Bert Flanegan, Tom Rowe and Dick Moyer
for their helpful advice as members of my graduate
committee.
My experiences with my fellow coworkers in the Stein's
lab have been very interesting, rewarding and fulfilling. I
would like to acknowledge and thank Marie Giorgio, Linda
Green, Farhad Marashi, Brian O'Donnell and Mary Beth Kennedy
for their technical guidance and assistance, and Ken Wright
and Joe Teixeira for their help dealing with the computers.
I thank Pat Naples for her appreciation and pledge of care
for "the coffee mug." I would like to thank Dave Collart,
Tim Morris and Charles Stewart for their friendship and for
their advice on raising bees, fishing, and golfing,
respectively. I would like to thank Andre van Wijnen for his
friendship, scientific expertise on and off the "Foos" ball
field, and for keeping an eye on Pat Naples and my coffee
11


188
UNTREATED
RIBO-
ACTIN -
H4 -
i
FOS -
HLA-B7-
CYTOCHALASIN D
PUROMYC IN
PURO+CYTO
RIBO
AC TIN -
HA -
FOS -
HLA-B7-


19
filaments with heat shock or indirectly with colcemid has no
effect on protein synthesis as determined by two-dimensional
electrophoresis indicating that translation is not dependent
on the integrity of the intermediate filaments (Welch and
Feramisco, 1985). Other reports suggest that the attachment
of mRNA to the cytoskeleton is mediated by the
microfilaments (Lenk et al.. 1977; Howe and Hershey, 1984;
Ornelles et al.. 1986) Treatment with cytochalasins, either
cytochalasin B or D, to disrupt the actin-containing
microfilaments results in the release of poly-A+ mRNA from
the cytoskeleton into the soluble phase. At least in the
case of cytochalasin D treatment, the release of poly-A+
mRNA from the cytoskeleton occurs in a dose dependent manner
and stoichiometrically reflects the degree of inhibition of
protein synthesis (Ornelles et al. 1986). These results
indicate that intact microfilaments are required for the
association of mRNA to the cytoskeleton; however, it can not
be concluded that the actin fibers directly bind the mRNA.
Nonmembrane-Bound and Membrane-Bound Polyribosomes;
The Signal Hypothesis
As described above, nearly 100 percent of the actively
translated polysomes are associated with the cytoskeleton.
Studies on vesicular stomatitis virus infected HeLa cells
demonstrated that membrane-bound polysomal mRNA coding for
the viral G glycoprotein and "free" polysomal mRNA coding


2
yield a better understanding of the processes controlling
proliferation, differentiation, cell aging and neoplasia.
Histone Genes and Proteins
There are five major classes of histone proteins (H4,
H2B, HiA, H3, and HI) ranging in molecular weight from
11,000 to 21,000 Daltons. They are basic proteins that have
been highly conserved throughout evolution. For example,
there are only two amino acids out of 102 which differ
between histone H4 from pea seedlings and calf thymus, which
suggests a critical role for these proteins in cell survival
(DeLange et al.. 1969). The histone proteins are encoded by
a family of moderately repeated genes (Wilson and Melli,
1977; Kedes, 1979). In humans, the histone genes are
arranged in clusters that are not organized in simple tandem
repeats as found in the sea urchin and other lower
eukaryotic genomes (Heintz et al.. 1981; Sierra et al.
1982; Carozzi et al.. 1984). The clusters are localized on
chromosomes 1, 6, and 12 and are interspersed with
repetitive elements (Green et al.. 1984b; Collart et al..
1985). It has been speculated that these repetitive elements
may play a role in recombination and gene duplication during
the evolution of this multi-gene family. The general
structure of the human histone genes is less complex than
many other eukaryotic genes since the histone protein coding
region is not interrupted by introns and the encoded mRNAs


63
of the signal peptide would disrupt its recognition by the
signal peptide recognition particle and would therefore,
prevent the translocation of the mutated histone fusion mRNA
to membrane-bound polysomes.
An oligonucleotide (SPa) containing two point mutations
was synthesized by Tom Doyle in the Department of
Microbiology and Immunology, University of Florida,
Gainesville, Florida. The oligonucleotide was composed of
the following seguence: 5' CAAAAAAGTGAATATGGGCGA 3', with
the mutated nucleotides underlined. The oligonucleotide was
received in ~2.5 ml 37% ammonium hydroxide and was heated at
55C for 6 hours to remove the protecting amine group. The
sample was dried under a gentle stream of air in the fume
hood and resuspended in 1 ml ddH20. The sample was extracted
two times with 600 ul n-butanol and the oligonucleotide was
precipitated from the aqueous phase by adding 200 ul 0.5 M
ammonium acetate and 2 ml 95% ethanol and chilling at -70C
for 30 minutes. The sample was centrifuged at 12,000 x g at
4C for 30 minutes, washed in 80% ethanol and resuspended in
200 ul ddH20. The yield of oligonucleotide was estimated
spectrophotometrically and was found to be 2.64 mg
(^26o/^280 2.0).
A second oligonucleotide (ATG) was designed to destroy
the signal peptide ATG translation start codon of the SPH3
fusion gene and thereby completely block the synthesis of
any signal peptide. The SPH3ATG" mRNA retains the histone


135
1 2345678
HLA-B7
H3
H4


67
and the incubation was continued at room temperature
overnight.
Enrichment for Covalently Closed Double-Stranded DNA
The in vitro synthesis of the entire second strand is
very inefficient due primarily to the large size of the M13
genome. The percent of covalently closed, double-stranded
DNA molecules is therefore, very small in comparison to the
partially double-stranded, incomplete DNA molecules. These
incomplete DNA molecules, when transfected into bacteria,
give rise to wild type plaque formation which complicates
the selection process for the mutated clones. Enrichment for
covalently closed, double-stranded DNA molecules by alkaline
sucrose gradient centrifugation greatly facilitates the
selection process.
The DNA was precipitated from the overnight extension
and ligation reaction by adding 30 ul ddH20, 50 ul 1.6 M
NaCl-13% PEG at 0cC for 15 minutes. The DNA was pelleted at
12,000 x g at room temperature for 15 minutes. The pellet
was gently washed in 100 ul cold 0.8 M NaCl-6.5% PEG,
centrifuged at 12,000 x g for 30 seconds and resuspended in
180 ul TE buffer. The sample was denatured with 20 ul 2 N
sodium hydroxide and layered onto a 5%-20% alkaline sucrose
step gradient. The gradient was prepared by adding to a 0.5
in x 2 in centrifuge tube (nitrocellulose or polyallomer; do
not use Beckman Ultra-clear tubes which are incompatible


189
examined the subcellular location of c-fos mRNA, with
respect to its association with the cytoskeletal structure,
in both control and cytochalasin D treated cells.
Northern blot analysis of cytoskeletal and soluble
phase RNAs from untreated HeLa cells demonstrated that
approximately 80% of c-fos mRNA is associated with the
cytoskeletal structure (Figure 5-2 and Table 5-2).
Cytochalasin D treatment efficiently disrupted this
association and released c-fos mRNA from the cytoskeleton
into the soluble phase. Greater than 70% of the c-fos mRNA
accumulated in the soluble phase in cytochalasin D treated
cells. Cytochalasin D and puromycin cotreatment also
preferentially increased c-fos mRNA levels in the soluble
phase. Nearly 80% of the c-fos mRNA was localized in the
soluble fraction isolated from CD and puromycin co-treated
cells. Consistent with the work of Ornelles et al. (1986),
puromycin treatment alone, which also increased c-fos mRNA
levels, had essentially no effect on the cytoskeletal and
soluble phase distribution of the mRNA. Approximately 85% of
the c-fos mRNA remained associated with the cytoskeleton in
puromycin treated cells. These results indicate that the
majority of the c-fos mRNA synthesized during cytochalasin D
treatment is localized in an altered subcellular region.
Accumulation of c-fos mRNA in an unnatural subcellular
compartment may physically remove it from the regulatory
factors that are involved in its stability.


207
Bird, R., Jacobs, F., Stein, G. Stein, J., and Sells, B.
(1985) Biochem. Biophys. Acta 824, 209.
Bishop, J.M. (1983) Annu. Rev. Biochem. 52, 301.
Blobel, G., and Dobberstein, B. (1975) J. Cell Biol. 67,
835.
Bloom, G. Luca, F. and Vallee, R. (1984) J. Cell Biol. 98
331.
Blose, S., and Chacko, S. (1976) J. Cell Biol. 70, 459.
Blum, J., and Wicha, M. (1988) J. Cell. Phys. 135, 13.
Bonneau, A., Darveau, A., and Sonenberg, N. (1985) J. Cell
Biol. 100, 1209.
Borun, T., Scharff, M., and Robbins, E. (1967) Proc. Natl.
Acad. Sci. U.S.A. 58, 1977.
Borun, T., Gabrielli, F., Ajiro, K., Zweidler, A., and
Baglioni, C. (1975) Cell 4, 59.
Briendl, M., and Gallwitz, D. (1973) Eur. J. Biochem. 32.
381.
Briendl, M., and Gallwitz, D. (1974) Eur. J. Biochem. 45.
91.
Brown, S., Levinson, W., and Spudich, J. (1976) Supramolec.
Struct. 5, 119.
Brush, D., Dodgson, J., Choi, O.-R., Stevens, W., and Engel,
J. (1985) Mol. Cell. Biol. 5, 1307.
Bulinski, J., and Borisy, G. (1980) J. Cell Biol. 87, 802.
Butler, W., and Mueller, G. (1973) Biochem. Biophys. Acta
294. 481.
Capasso, O., Bleecker, G., and Heintz, N, (1987) EMBO J. 6,
1825.
Carozzi, N., Marashi, F., Plumb, M., Zimmerman, S.,
Zimmerman, A., Wells, J., Stein, G., and Stein, J. (1984)
Science 224, 1115.
Cervera, M., Dreyfuss, G., and Penman, P. (1981) Cell 23,
113.


76
resulted in the rapid disappearance of histone mRNAs from
these polysomes (Borun et al.. 1967; Breindl and Gallwitz,
1974; Gallwitz, 1975).
The light polysomes on which translatable histone mRNAs
reside are presumably nonmembrane-bound ("free) polysomes
as predicted by the signal hypothesis (Blobel and
Dobberstein, 1975). Several lines of evidence suggest that
nonmembrane-bound polysomes are not freely diffusible within
the cytoplasm but are attached to the cytoskeleton. Penman
and co-workers (Lenk et al.. 1977) have demonstrated that
nearly 100% of total cytoplasmic polysomes co-fractionate
with the cytoskeletal framework. In addition, vesicular
stomatitis virus encoded nonmembrane-bound and membrane-
bound polysomal mRNAs are associated with the cytoskeleton
in virally infected HeLa cells (Cervera et al. 1983).
Taken together, these observations provide a basis for
the possibility that histone mRNA-containing polysomes
and/or their subcellular location may play a role in the
regulation of histone gene expression. Such reasoning is
consistent with the appearance of translatable histone mRNAs
on polysomes within minutes following transcription (Borun
et al. 1967) the equally rapid transfer of newly
synthesized histones into the nucleus where they complex
with DNA (Robbins and Borun, 1967; Spaulding et al. 1966),
and the degradation of histone mRNAs and cessation of
histone protein synthesis in parallel with the inhibition of


90
with cycloheximide and hydroxyurea had a stabilizing effect
on histone mRNA levels and no apparent effect on HLA-B7 mRNA
stability in both subcellular fractions. There were no
obvious differences in the stability of histone mRNAs
localized in the various subcellular compartments when
treated with hydroxyurea or cycloheximide. It is therefore
reasonable to conclude that the histone mRNA isolated in the
LER and membrane-bound fractions is a consequence of the
subcellular fractionation procedure. Our results, taken
together with previous findings (Jacobs-Lorena et al.. 1972;
Liautard and Jeanteur, 1979; Gallwitz and Breindl, 1972;
Borun et al., 1975; Stein et al.. 1975) suggest that histone
mRNAs, during normal cellular processing, are targeted to
the nonmembrane-bound polysomes, the common site of histone
protein synthesis.
DISCUSSION
Our objectives in pursuing these studies were two-fold.
The first was to confirm that histone proteins are
synthesized on nonmembrane-bound polysomes. The second was
to determine if the histone mRNA-containing polysomes are
associated with the cytoskeleton. Our rationale for this
approach was to begin addressing the questions whether the
subcellular location of histone containing polysomes
contributes to; a) the ability of histone mRNAs to be
translated immediately following transcription, b) the rapid


66
revealed a single major band at -80 bp for each molar ratio
of oligonucleotide to template DNA studied. This was the
expected size fragment and suggested that under these
conditions, the oligonucleotide was priming the synthesis of
the second strand from the proper site. If several
predominant bands had been detected (i.e. multiple priming
sites), more stringent annealing conditions would have been
required.
Synthesis of the Mutated Strand
The oligonucleotide (1.3 nq) was phosphorylated in 100
mM Tris-HCl pH 8, 10 mM MgCl2, 5 mM dithiothreitol, 10 uM
ATP with 15 units T4 polynucleotide kinase at 37C for 45
minutes. The reaction was terminated by heating at 65C for
10 minutes.
The phosphorylated oligonucleotide (13.3 pmole) was
annealed to M13/SPH3 recombinant DNA in lx solution A by
incubating the sample at 55C for 5 minutes and then at room
temperature for 5 minutes. After hybridization, the DNA was
diluted with 10 ul solution C (1 ul solution B [0.2 M Tris-
HCl pH 7.5, 0.1 M MgCl2, 0.1 M dithiothreitol], 1 ul 10 mM
dCTP, 1 ul 10 mM dTTP, 1 ul 10 mM dGTP, 0.5 ul 0.1 mM dATP,
1 ul 10 mM ATP, 1.5 ul [gamma-32P] dATP, 1.5 ul T4 DNA ligase
[1 U/ul], 2 ul ddH20) and the synthesis of the second strand
was initiated with 0.5 ul Klenow enzyme [100 U/16.7 ul] at
room temperature. After 5 minutes, 1 ul 10 mM dATP was added


175
1
2
3
4
c-f os
hCG


62
the mutated histone fusion mRNA, like endogenous histone
mRNA, associates with nonmembrane-bound polysomes.
Subcloninq of the Signal Peptide-Histone Fusion Gene
into M13 DNA
The signal peptide-histone fusion gene (SPH3) was
subcloned into M13mpl8 RF DNA. SPH3 DNA was digested with
EcoRI and Hindlll and the 1100 bp fragment was isolated from
a 0.8% agarose gel. The M13mpl8 DNA was digested with EcoRI
and Hindlll and the 7196 bp fragment was isolated from a
0.8% agarose gel. The 1100 bp SPH3 DNA fragment (-50 ng) and
7196 bp M13 DNA fragment (-50 ng) were ligated in 10 ul lx
ligation buffer (50 mM Tris pH 7.6, 10 mM MgCl2, 20 mM
dithiothreitol, 1 mM ATP and 50 nq/ral BSA) with 0.004 U T4
DNA ligase (NEB) at 22C overnight. The ligation mix was
transfected into competent JM101 bacteria, white plaques
were isolated, and M13/SPH3 recombinants were screened by
restriction endonuclease digestion analysis. Single-stranded
M13/SPH3 DNA was isolated from bacteriophage particles as
described above.
Synthesis and Processing of Oligonucleotides
The signal peptide encoded by the SPH3 histone fusion
gene was mutated by substituting leucine at amino acid
position 10 and proline at position 12 with histidine
residues. It was predicted that the introduction of two
positively charged amino acids into the hydrophobic region


KEY TO ABBREVIATIONS
BSA
Bovine Serum Albumin
cpm
Counts per minute
dCT>
Deoxycytidine-5'-triphosphate
ddCTP
Dideoxycytidine-5'-triphosphate
DNA
Deoxyribonucleic acid
DNase
Deoxyribonuclease
DTT
Dithiothreitol
EDTA
Ethylenediaminetetraacetic acid
Hepes
N-2 hydroxyethyl piperazine-N'-2
ethanesulfonic acid
HU
Hydroxyurea
juci
Microcurie
Mg
Microgram
Ml
Microliter
Mm
Micrometer
mm
Micromolar
OD
Optical density
PVS
Polyvinyl sulfate
RNA
Ribonucleic acid
rpm
Revolutions per minute
SDS
Sodium dodecylsulfate
SSC
Standard saline citrate buffer
Tris
(hydroxymethyl)aminomethane
x


166
Discussion
The data presented in this chapter demonstrate that the
signal peptide-histone fusion mRNAs (SPH3E1 and SPH3E1ATG")
express a cytoskeleton attachment site that is not
associated with histone mRNA, HLA-B7 mRNA or heterogeneous
poly-A^ mRNA (Chapter 5; Ornelles et al.. 1986). It is not
known whether the cytoskeleton attachment site unique to
SPH3E1 and SPH3E1ATG mRNA is due directly to the primary
nucleotide sequences coding for the signal peptide, an
alteration in the overall conformation of the mRNA or the
proteins associated with the mRNP.
The elements of the cytoskeleton that are involved in
the attachment of SPH3E1 mRNA and SPH3E1ATG" during
cytochalasin D and puromycin cotreatment have not been
identified. Association of the signal peptide-histone fusion
mRNAs with the cytoskeleton during cytochalasin D and
puromycin cotreatment indicates that the attachment site is
independent of intact microfilaments. In addition, the
cytochalasin D and puromycin insensitive cytoskeleton
attachment site does not directly involve the microtubules
since this cytoskeletal component is removed during the
fractionation procedure (Lenk et al.. 1977). The signal
peptide-histone fusion mRNAs may be associated with the
intermediate filaments of the cytoskeleton. Previous studies
on other cell types indicate that mRNPs are associated with
the cytoskeleton through the intermediate filaments


17
indicate that actively translated mRNAs are associated with
the cytoskeleton and this association may be requisite for
protein synthesis. Consistent with this possibility, the
inhibition of host protein synthesis coincides with the
release of host mRNA from the cytoskeletal structure and the
attachment of viral mRNA to the cytoskeleton in poliovirus
infected and adenovirus infected cells (Cervera et al..
1981; Lemieux and Beaud, 1982). During development,
maternally inherited mRNAs participate in protein synthesis
only at the time they become associated with the
cytoskeleton in sea urchin oocytes (Moon et al.. 1983).
The region of the polysome involved in its attachment
to the cytoskeleton is not known. Previous results indicate
that the association of polysomal mRNA with the cytoskeleton
is through the mRNA or mRNP particle and not through the
ribosome (Lenk et al.. 1977; Lenk and Penman, 1979; Cervera
et al.. 1981; Lemieux and Beaud, 1982). Disruption of
polysomes by heat shock, high salt, sodium fluoride or
pactamycin treatment releases the ribosomal subunits but not
the mRNA from the cytoskeleton into the soluble phase (Lenk
et al.. 1977; Cervera et al.. 1981; van Venrooij et al..
1981; Howe and Hershey, 1984). Immunofluorescence studies
indicate that the cap-binding protein is associated with the
cytoskeleton suggesting that the attachment of eukaryotic
mRNAs to the cytoskeleton may occur via the 5' mRNA cap
structure (Zumbe et al.. 1982). Earlier studies suggest that


138
Dobberstein, 1975). Subcellular fractionation and Northern
blot analysis revealed that c-fos mRNA is predominantly
associated with nonmembrane-bound polysomes (data not
shown), and that approximately 80% of the c-fos mRNA is
associated with the cytoskeleton (Figure 5-2, lanes 1 and 2;
Table $-2). Cytochalasin D as well as puromycin,
CD/puromycin and CD/cycloheximide dramatically increased c-
fos mRNA levels, and the newly synthesized c-fos mRNA
partitioned into both the cytoskeletal and soluble fractions
in a manner that reflects the drug treatment (Figure 5-2,
lanes 3-10; Table 5-2). Less than 30% of the c-fos mRNA
remained associated with the cytoskeleton in cytochalasin D
[10 fiq/ml] treated cells (Table 5-2).
Membrane-Bound Polvsomal mRNAs Express Multiple Cytoskeleton
Attachment Sites
Retention of membrane-bound polysomal mRNAs on the
cytoskeleton in cytochalasin D treated cells suggests that
this class of polysomal mRNA may contain additional
cytoskeleton attachment sites. Cell surface proteins, such
as HLA-B7 and chorionic gonadotropin, are generally
synthesized on membrane-bound polysomes that are complexed
with the endoplasmic reticulum (for review see Emr et al.,
1980). The nature of the interaction between the polysomes
and the endoplasmic reticulum appears to be mediated in part
by the nascent polypeptide that is inserted into the


158
Table 6-1. The percent of cvtoskeleton associated mRNAs
isolated from cvtochalasin D. puromycin and CD/puromycin
treated cells.
Control
CD
Puro
CD/Puro
A)
SPH3E1
>95%
93%
--
62%
Endogenous
H3
81%
31%

27%
B)
SPH3E1
93%
79%
94%
72%
hCG alpha
98%
81%
95%
18%
HLA-B7
99%
84%
93%
28%
Endogenous
H3
92%
46%
90%
26%
C)
SPH3E1ATG"
>90%
87%
>89%
77%
hCG alpha
98%
81%
89%
29%
HLA-B7
97%
89%
88%
35%
Endogenous
H3
78%
56%
79%
24%
The densitometric results from autoradiographs
represented in Figures 6-2 to 6-4 (A to C, respectively)
were corrected for the total yield of RNA from each fraction
(note: equal quantities of cytoskeleton and soluble RNAs
were analyzed in SI and northern assays which does not take
into consideration the unequal distribution of RNA within
these fractions or the changes that occur during CD and/or
puromycin treatment).


40
Double-stranded templates. E. coli strain JM101
(delta(lac-proAB) supE / F' fproA+. proB*. laclq.
lacZdeltaMIS]) bacteria were grown overnight in 5 ml YT
medium at 37C with agitation. The overnight culture (1 ml)
was transferred to 100 ml prewarmed YT medium and incubated
at 37C until the O.D. 590 reached 0.4 units. To assure
exponential growth the culture was again diluted 1:100 into
500 ml prewarmed YT medium and grown to an O.D. 590 of 0.4
units. The culture was infected with 500 ul of M13
bacteriophage supernatant which had been prepared from an
isolated plaque as described above for the preparation of
M13 single-stranded DNA. The infected cells were incubated
at 37C with agitation for no more than 8 hours. The cells
were harvested and RF M13 DNA was isolated as described
above for the preparation of plasmid DNA. Routinely, the
yield was 300 nq of supercoiled M13 DNA per 500 ml culture.
Isolation of Mammalian RNA
Isolation of Total Cellular RNA
The following procedure for the isolation of total
cellular RNA was developed by Dr Mark Plumb in the Stein's
laboratory primarily for work with HeLa cells. However, this
procedure has been successfully used without modification
for the isolation of total cellular RNA from human HL-60
cells, mouse 3T3-L1 cells, rat osteoblast primary cell
cultures, and Trypanosome cruzii cells.


TABLE OF CONTENTS
PAGE
ACKNOWLEDGEMENTS ii
LIST O FIGURES vi
LIST OF TABLES ix
KEY TO ABBREVIATIONS X
ABSTRACT xi
CHAPTERS
1) INTRODUCTION
General Background 1
Histone Genes and Proteins 2
Histone Gene Regulation: Transcriptional and Post-
Transcriptional Control 3
Subcellular Compartments 10
The Cytoskeleton: Microtubules, Intermediate
Filaments and Microfilaments 12
Cytoskeletal-mRNA Interactions 16
Nonmembrane-Bound and Membrane-Bound Polyribosomes:
The Signal Hypothesis 19
Overview of Project 22
2) GENERAL METHODS
Materials 27
Conditions for Enzyme Reactions 28
Mammalian Cell Culture 28
Plasmid DNA 3 0
DNA Isolation and Purification 35
Isolation of Mammalian RNA 40
Two-Dimensional Gel Electrophoresis and Immuno-
Blotting Analysis 48
Northern Blot Analysis 49
SI Nuclease Protection Analysis 52
In Vitro Nuclear Run-on Transcription Analysis 55
Radiolabeling DNA for Northern Blot Analysis 58
Site-Directed Mutagenesis 61
Transfection of DNA into Cells 71
Establishment of Polyclonal and Monoclonal Stably
Transformed Cell Lines 73
iv


214
Rauscher III, F. Cohen, D., Curran, T. Bos, T., Vogt, P.,
Bohmann, D., Tjian, R., and Franza Jr., B. (1988) Science
240. 1010.
Rave, N., Crkvenjakov, R., and Boedtker, H. (1979) Nucl.
Acids Res. 6, 3559.
Rawn, J. (1983) in Biochemistry. Wasserman, M., and Nickol,
D., Eds. (Harper and Row, New York) p 528.
Rebillard, M., Leibovitch, S., Jullien, M., Talha, S., and
Harel, L. (1987) Exp. Cell. Res. 172, 432.
Robb, R., Terhorst, C., and Strominger, J. (1978) J. Biol.
Chem. 253. 5319.
Robbins, E., and Borun, T. (1967) Proc. Natl. Acad. Sci.
U.S.A. 57, 409.
Roberts, C., and Wilson, G. (1985) Focus 7, 16.
Roggenkamp, R., Kustermann-Kuhn, B., and Hollenberg, C.
(1981) Proc. Natl. Acad. Sci. U.S.A. 78, 4466.
Roop, D., Hawley-Nelson, P., Cheng, C., and Yuspa, S. (1983)
Proc. Natl. Acad. Sci. U.S.A. 80, 716.
Ross, J., and Kobs, G. (1986) J. Mol. Biol. 188, 579.
Ross, J. Peltz, S. Kobs, G., and Brewer, G. (1986) Mol.
Cell. Biol. 6, 4362.
Ross, J., Kobs, G., Brewer, G., and Peltz, S. (1987) J.
Biol. Chem. 262, 9374.
Rovera, G,, Santoli, D., and Damsky, C. (1979) Proc. Natl.
Acad. Sci. U.S.A. 26, 1779.
Rothstein, T. (1985) J. Immunology 135. 106.
Rothstein, T. (1986) J. Immunology 136. 813.
Rubenstein, P., and Spudich, J. (1977) Proc. Natl. Acad.
Sci. U.S.A. 74, 120.
Sanger, J. (1975) Proc. Natl. Acad. Sci. U.S.A. 72, 2451.
Sassone-Corsi, P., and Verma, I. (1987) Nature 326. 507.
Schachner, M., Hedley-White, E., Hsu, D., Schoonmaker, G.,
and Bignami, A. (1977) J. Cell Biol. 75., 67.


42
acid samples were then digested with 0.1 mg/ml of proteinase
K treated DNase I for 20-30 minutes at 37C. Following DNase
I digestion, 0.05 volumes of 5 M NaCl and 0.25 volumes of
10% SDS were added to each sample and the RNA solutions were
repeatedly extracted with phenol and chloroform until the
interfce between the aqueous and organic phases was clean.
RNA was precipitated from the aqueous phase with 2.5 volumes
of 95% ethanol, at -20C overnight. The RNA was collected at
12,000 rpm in a JA20 rotor at 4C for 30 minutes. The pellet
was resuspended in double distilled water, quantitated by
optical density at 260 nm and stored at -20C. Typical
yields of RNA were 400 /xg per 90% confluent 10 cm plate of
HeLa S3 cells and 1 mg per 5 x 107 HeLa S3 cells with an
0 D 260/ O D. 280 of 1.8 to 2.0.
Isolation of Nonmembrane-bound. Membrane-bound and Total
Polvsomal RNA
The subcellular fractionation scheme used in these
studies is a modified procedure described by Venkatesan and
Steele (1972) and is outlined in Figure 2-2. Exponentially
growing HeLa cells (5 x 108 cells) were collected by
centrifugation at 1,500 rpm in an IEC rotor at 37C for 5
minutes. All subsequent procedures were carried out at 4C.
Cells were washed several times in PBS and resuspended in 25
ml RSB (10 mM Tris-HCl pH 7.4, 10 mM NaCl, 2.5 mM MgCl2) .
After incubation on ice for 20 minutes, the cells were


119
probe
fusion
H3


50
hours). For full size RNA gels, 200 ml agarose-formaldehyde
solution was cast in a plexiglass gel tray (24.5 cm x 19.8
cm) with 2: 20 well combs (1 mm x 6 mm per well). The gel
was electrophoresed at 100-120 volts until the bromophenol
blue migrated 11 cm (approximately 4 hours). The
formaldehyde gels were stained for 20 minutes in ethidium
bromide solution (1 /g/ml ethidium bromide, 0.1 M ammonium
acetate) and destained for at least 2 hours in double
distilled water. RNA samples were then visualized by
fluorescence under ultraviolet light using a short
wavelength transilluminator (UVP, inc.; San Gabriel, CA).
The gels were usually photographed at f4.5 for 1/4 second
with Type 57 high speed Polaroid film and a Polaroid model
545 camera mounted on a stand above the transilluminator.
Transfer of RNA from Aqarose-Formaldehvde Gels to
Hybridization Filters
RNA samples were transferred to nitrocellulose (0.45
urn) or nylon membrane hybridization filters in 20x SSC as
previously described by Thomas (1975). The gels were soaked
in 2Ox SSC (3 M NaCl, 0.3 M sodium citrate pH 7.0) for 20
minutes and were not stained with ethidium bromide. The
hybridization filters were uniformly wetted in ddH20 and
then 2Ox SSC. The gel was placed with the wells facing
downward onto a sheet of 3 MM paper (Whatman) and sponge
which had been soaked in 2Ox SSC. Strips of used X-ray films


14
suggests that these filaments are in close association with
the microtubules.
The keratin filaments, also known as tonofilaments, are
predominantly found in epithelial cells. The keratin
proteins are a family of approximately 30 distinct
polypeptides ranging in molecular weight from 40,000-65,000
Daltons (Steinert, 1975; Steinert and Gullino, 1976; Sun and
Green, 1978; Moll et al.. 1982). The keratins can be clearly
divided into acidic and basic groups based on two-
dimensional electrophoretic analysis. Although only a few
keratin subtypes are expressed in any one epithelial cell,
there is always a coordinated synthesis of at least one
class each of the acidic and basic keratin subunits (Moll et
al. 1982; Woodcock-Mitchell et al.. 1982; Fuchs and
Marchuk, 1983). The combination of keratin subtypes
expressed by an epithelial cell is characteristic of the
tissue origin as well as the differentiation state of the
cell (Roop et al.. 1983; Sun et al.. 1983). In addition,
some epithelially derived tumors express different keratin
subtypes than those found in the normal counterpart and
these differences may be useful in diagnostic tests
(Summerhayes et al.. 1981; Osborn, 1983).
Little is known about the function and organization of
the intermediate filaments from neuronal and glial cells.
The major cytoskeletal element of neuronal axons and
dendrites are the neuronal filaments which are generally


164
alone. Over 80% of these mRNAs were associated with the
cytoskeleton after CD treatment. In contrast, less than 28%
of the hCGa and HLA-B7 mRNAs remained associated with the
cytoskeleton after CD and puromycin cotreatment indicating
that the drug treatments were effective.
The association of the signal peptide-histone mRNA with
the cytoskeleton in CD and puromycin co-treated cells
suggests the existence of a cytoskeleton attachment site
that is distinct from the site associated with endogenous
histone mRNA and from other membrane-bound polysomes. These
results indicate that the site is independent of the
ribosomes and the nascent polypeptide and maybe a property
of the nucleotide sequence and/or proteins that interact
with the mRNA itself. Alternatively, the SPH3E1 chimeric
mRNA may be efficiently translated and therefore support re
initiation of translation even in the presence of
cytochalasin D and puromycin (note: endogenous histone mRNAs
are efficiently translated; Stahl and Gallwitz, 1977). This
would result in the synthesis of a portion of the signal
peptide which could then serve as an anchor to the
cytoskeletal structure as if the cells were treated with
cytochalasin D alone.
To address this possibility, we have studied the
cytoskeletal association of a mutated signal peptide-histone
chimeric mRNA (SPH3E1ATG ) Using site directed mutagenesis
(Zoller and Smith, 1983) the ATG translation start codon for


SPH3E1ATG-
121


149
site coupled with a partial "run off" of the polysomes from
the mRNA, which in effect dissociates the nascent
polypeptide from the mRNA.
In addition to the differential association of
membrane-bound and nonmembrane-bound polysomes with the
cytoskeleton, there also appears to be differences in the
cytoskeletal interaction of individual species of mRNA. As
seen in Table 5-2, hCGa mRNA and HLA-B7 mRNA are released
from the cytoskeleton to the same extent during CD and
puromycin cotreatment, which is consistent with the
mechanism for the association of membrane-bound polysomes
with the cytoskeleton as described above. However, CD and
cycloheximide cotreatment releases a larger percentage of
hCGa mRNA from the cytoskeleton than CD treatment alone. In
contrast, HLA-B7 mRNA is associated with the cytoskeleton to
the same extent or greater in cytochalasin D and
cycloheximide co-treated cells than HeLa cells treated with
CD alone. Furthermore, CD and cycloheximide cotreatment
readily dissociates histone mRNA from the cytoskeleton
whereas c-fos mRNA is only partially affected. These results
suggest that multiple mechanisms are operative for the
association of individual mRNAs and/or polysomes with the
cytoskeleton and may reflect differences in the structural
environment of the cell.
The possibility that the structural environment of the
cell may affect cytoskeletal-mRNA interactions is supported


64
translation start codon in its proper context, although
further downstream from the 5' terminus than endogenous H3
histone mRNA, and should therefore be translated on
nonmembrane-bound polysomes. In addition, SPH3ATG mRNA
differs from SPH3 wild type signal peptide-histone fusion
mRNA at only one nucleotide. The sequence of the ATG~
oligonucleotide is as follows: 5' AATACTCAAACTCTTCC 3', and
was prepared and processed as described for the SPa
oligonucleotide. The yield of ATG" oligonucleotide was only
445 Mg (^26o/^280 1.41) .
Optimization of Conditions for Site-Directed Mutagenesis
It was necessary to optimize the conditions for the
hybridization of the oligonucleotide to M13 template so that
the synthesis of the second strand was primed by the
oligonucleotide from the correct site. Proper priming by the
oligonucleotide was tested by primer extension analysis.
After limited primer extension, the samples were digested
with a restriction endonuclease that recognized a site
downstream from the expected priming site. The number and
intensity of the fragments produced by the digestion
indicated the number of priming sites and the location of
the preferred site.
The oligonucleotide (20 pmol) was phosphorylated with 5
units T4 polynucleotide kinase in 100 mM Tris-HCl pH 8.0, 10
mM MgCl2, 5 mM dithiothreitol, 20 uCi [gamma-32P] ATP at 37C


184
UNTREATED
RIBO
ACTIN
RIBO
FOS
H4
FOS
H4
RIBO
ACTIN
RIBO
FOS
H4
FOS
H4
CYTOCHALASIN D
PU RO/CYTO


109
512
396
344
298
220
154
130
HYBRID H3mRNA
ENDOGENOUS H3 mRNA


24
specifically asked the question: Does the subcellular
location of histone mRNA play a role in the destabilization
of histone mRNA during inhibition of DNA synthesis? To
address this question, we have applied a combination of
molecular and cellular approaches first, to determine the
natural location of histone mRNA and second, to alter the
subcellular compartment to study whether the foreign
location affects histone mRNA stability. The rationale for
this approach is that directing histone mRNA to an altered
subcellular location may physically remove the message from
the factors that are involved in the selective degradation
of histone mRNA during inhibition of DNA synthesis.
Our initial studies demonstrated that histone mRNAs are
translated predominantly on nonmembrane-bound polysomes that
are associated with the cytoskeleton. The subcellular
location of the mRNA was altered biologically by
constructing a histone fusion gene that contains sequences
coding for the signal peptide of the beta-lactamase gene of
E. coli pBR322 plasmid DNA. The encoded fusion histone mRNA
was directed to membrane-bound polysomes and was not
destabilized during inhibition of DNA synthesis.
Inactivation of signal peptide function by a single
nucleotide substitution in the ATG translation start codon
of the signal peptide resulted in the association of the
mutated fusion mRNA with nonmembrane-bound polysomes and
restored the coupling of the mRNA stability with DNA


143
remained associated with the cytoskeleton in cytochalasin D
treated cells that were subsequently treated with puromycin
(Table 5-2). In contrast, more than 75% of the HLA-B7 mRNA
and 72% of the hCGa mRNA remained associated with the
cytoskeleton in HeLa cells treated with cytochalasin D alone
(Table"5-2). The release of membrane-bound polysomal mRNAs
from the cytoskeleton in cytochalasin D and puromycin
treated HeLa cells was not a direct result of the inhibition
of protein synthesis. Cytochalasin D treatment followed by
the inhibition of protein synthesis with cycloheximide, a
compound that preserves the polysomal structure, failed to
release HLA-B7 mRNA from the cytoskeleton (Figure 5-3, lanes
9 and 10). Greater than 80% of the HLA-B7 mRNA was
associated with the cytoskeleton in cytochalasin D and
cycloheximide treated cells (Table 5-2). The results were
not as dramatic for the hCGa mRNA (Figure 5-3, lanes 9 and
10). Approximately 52% of hCGa mRNA was associated with the
cytoskeleton in cells treated with both cytochalasin D and
cycloheximide (Table 5-2).
The release of HLA-B7 mRNA and hCGa mRNA from the
cytoskeleton in cytochalasin D and puromycin treated HeLa
cells was not due solely to the inhibition of protein
synthesis by puromycin. As previously reported, the
dissociation of polysomes with high salt or heat shock
resulted in the release of the ribosomal subunits into the
soluble phase without affecting the attachment of poly A+


112
SPH3E1:
Alpha :
ATG' :
80 nt
CAP
ATG
$m
5
10
15
20
25
Gin
Hi s
Phe
Arg
Val
Ale
Leu
He
Pro
Phe
Phe
Ala
Ala
Phe
Cys
Leu
Pro
Val
Phe
Ala1
Met
Ala
CAA
CAU
UUC
CGU
GUC
GCC
CUU
AUU
CCC
UUU
UUU
GCG
GCA
UUU
UGC
CUU
CCU
GUU
UUU
GCC
AUG
GCU
Gin
His
Phe
Arg
Val
Ala
His
lie
His
Phe
Phe
Ala
Ala
Phe
Cys
Leu
Pro
Val
Phe
Ala
Met
Ala
CAA
CAU
UUC
CGU
GUC
GCC
CaU
AUU
CaC
UUU
UUU
GCG
GCA
UUU
UGC
CUU
CCU
GUU
UUU
GCC
AUG
GCU
GGAAGAGU uUG AGU AUU CAA CAU UC CGU GUC GCC CUU AUU
- - Het Ala
CCC UUU UUU GCG GCA UUU UGC CU CCU GUU UUU GCC AUG GCU


26
signal peptide-histone fusion mRNA from the cytoskeleton.
The nucleotide sequences of the signal peptide-histone
fusion mRNA encoding the signal peptide contain a
cytochalasin D and puromycin insensitive cytoskeletal
attachment site which is not expressed by histone mRNA or
HLA-B7"mRNA. These results provide evidence for the
heterogeneic pattern for cytoskeleton-mRNA interactions.
Lastly, we investigated the role of the cytoskeleton in
the regulation of gene expression by studying the effects of
disrupting the microfilaments with cytochalasin D on the
transcription and steady state mRNA levels of several genes.
We determined that the transcriptional and
posttranscriptional regulation of the c-fos proto-oncogene
is quite sensitive to the perturbation of the cytoskeleton
structure. Transcription of the c-fos gene is selectively
stimulated with a concomitant increase in c-fos mRNA steady
state levels within minutes of the addition of cytochalasin
D. Consistent with previous results, the transcription of
beta-actin was also elevated in cytochalasin D treated
cells. These findings suggests that the cell monitors the
organization of the cytoskeleton and selectively regulates
gene expression accordingly.


59
the isotope and buffer and the priming reaction was
initiated with 5 units Klenow enzyme [5 U/ul]. The sample
was incubated at room temperature for at least 2 hours
(usually overnight). The reaction was terminated with 1 ul
250 mM EDTA and 175 ul TE buffer (10 mM Tris-HCl pH 8, 1 mM
EDTA) and extracted once with 200 ul chloroform/isoamyl
alcohol. Unincorporated radiolabeled deoxynucleotide was
removed from the DNA by chromatography over a (10 mm x 100
mm) Biogel A-15m (Bio Rad) column in TE buffer. Fourteen
fractions were collected; 800 ul in the first fraction and
200 ul in each subsequent fraction. Fractions containing
radiolabeled DNA were determined by Cerenkov counting in the
tritium channel with a Beckman LS-230 scintillation counter.
The peak fractions were pooled and accurately quantitated by
counting 2 ul DNA in 5 ml Triton-toluene based scintillation
cocktail (2000 ml toluene, 1066 ml Triton X-100, 134 ml
liquifluor) in the j2P channel. The samples were immediately
used for hybridization studies and excess probe was stored
at 4C.
Nick-Translation Method
Nick-translation was the common procedure for
uniformally labeling DNA fragments or supercoiled DNA before
the development of the random primer method. A disadvantage
to using the nick-translation procedure is that the system
must be optimized for efficiency for each lot of E_¡_ coli DNA


Figure 7-5. Effect of protein synthesis inhibition on
cvtochalasin D induction of c-fos transcription in HeLa
cells.
Exponentially growing HeLa cells were treated with
cytochalasin D, puromycin, or puromycin and CD for the
indicated times. Radiolabeled RNAs were prepared by in vitro
nuclear run-on transcription and hybridized to Southern
blots of electrophoretically separated restriction enzyme
digested plasmid DNAs (107 cpm/filter). Insert DNAs
hybridizing to transcripts from the corresponding genes are
indicated. Untreated, DMSO for 15 min.; Cytochalasin D, CD
[10 Mg/ral] for 15 min.; Puromycin, puromycin [0.04 mM] for
30 min.; Puro + Cyto, puromycin [0.04 mM] for 30 min. and CD
[10 g/ml] for the last 15 minutes.


18
the association of eukaryotic mRNAs with the cytoskeleton
may occur through the poly-A tract (Milcarek and Penman,
1974). However, the association of uncapped poliovirus mRNAs
(Bonneau et al.. 1985) and poly-A" histone mRNAs (Jeffery,
1984; Bonneau et al.. 1985) indicates that the attachment
site might also exist in an internal region contained within
the 5' and 3* ends of the mRNA. Isaacs and Fulton (1987)
have presented evidence suggesting that myosin heavy chain
in embryonic chicken muscle cells is co-translationally
assembled into the cytoskeleton. The myosin nascent
polypeptide would therefore serve as an additional
cytoskeleton attachment site for the polysomal myosin mRNA.
Based on these results it appears that several mechanisms
may operate in the attachment of a specific polysomal mRNA
or mRNP to the cytoskeleton and these attachment sites may
vary depending on the species of mRNA.
The component of the cytoskeleton to which the mRNA is
attached is not known although the data accumulated to date
suggest that the microfilaments may be involved in the mRNA-
cytoskeleton association. Removal of tubulin from the
cytoskeleton by using cold calcium-containing buffers during
the isolation procedure has no effect on the attachment of
mRNA with the cytoskeleton; this result suggests that the
microtubules do not play a role in the association of mRNA
with the cytoskeleton (Lenk et al.. 1977; Howe and Hershey,
1984; Ornelles et al.. 1986). Disruption of vimentin


CHAPTER 7
THE INFLUENCE OF THE CYTOSKELETON ON THE REGULATION
OF c-Fos GENE EXPRESSION
Introduction
Modifications in gene expression is central to
proliferation and differentiaton of eukaryotic cells. These
processes are tightly regulated and dependent in part on the
precise expression of proto-oncogenes (for review see
Ohlsson and Pfeifer-Ohlsson, 1987? Bishop, 1983). The
proliferative response of quiescent fibroblasts to serum or
growth factor stimulation is accompanied by a rapid and
transient induction of the c-fos proto-oncogene (Cochran et
al., 1984; Greenberg et al.. 1984; Curran et al., 1985).
Within minutes of the addition of serum or growth factors to
quiescent fibroblasts the c-fos gene is transcribed and c-
fos cytoplasmic mRNA levels rapidly increase. Generally, the
expression of the c-fos gene during serum/growth factor
induction is maximal after 30 minutes and returns to pre
stimulatory levels within 1-2 hours. While the exact
function of the c-fos protein is unknown, recent evidence
suggests that this nuclear protein is part of a DNA binding
168


117
1). These results indicate that the mutation of the ATG
translation start codon of the signal peptide results in the
synthesis of a signal peptide-histone fusion mRNA
(SPH3E1ATG) that is associated with nonmembrane-bound
polysomes. In addition, the incorporation of two histidine
amino cid residues in the hydrophobic domain of the signal
peptide partially blocks the translocation of the mRNA from
nonmembrane-bound polysomes to membrane-bound polysomes.
To study the stability of the mutated signal peptide-
histone fusion mRNAs during inhibition of DNA synthesis, the
genes (SPH3E1, SPH3Elalpha# and SPH3E1ATG) were transfected
into HeLa cell monolayers by the calcium phosphate
precipitation method as described in Chapter 2. Forty-six
hours posttransfection, the cells were treated with 1 mM
hydroxyurea and samples were taken at 20 minute time
intervals. Total cellular RNA was isolated and assayed for
signal peptide-histone fusion mRNA and endogenous H3 histone
mRNA content by SI nuclease protection analysis as described
in Chapter 2. As seen in Figure 4-7 and summarized in Figure
4-10, inhibition of DNA synthesis for one hour resulted in
the destabilization of only 46% of SPH3E1 mRNA as compared
to the control cell culture (lanes 1,2 and 7,8). In sharp
contrast, SPH3E1ATG" mRNA was destabilized by 94%, which is
to the same extent as measured for endogenous H3 histone
mRNA (Figure 4-8 lanes 1,2 and 7,8? Fig. 4-10). Inhibition
of DNA synthesis for one hour with 1 mM hydroxyurea


179
cultures with culture medium containing 10% fetal calf serum
resulted in an increase in c-fos mRNA levels for 60 minutes
followed by a sharp reduction to backround levels by 120
minutes. Cytochalasin D affected c-fos gene expression with
the same kinetics as serum; however, slightly higher levels
of c-fos mRNA were measured in CD treated cells. Puromycin
treatment or CD and puromycin cotreatment caused a
continuous rise in c-fos mRNA levels for two hours, which
was the last time point examined. These results indicate
that the effects of cytochalasin D on either exponentially
growing HeLa cells or quiescent diploid fibroblasts more
closely resemble the effects of serum and growth factor
stimulation than the inhibition of protein synthesis on the
expression of c-fos mRNA levels.
Cytochalasin D Stimulates the Transcription of the c-Fos
Gene
To assess the contribution of transcription to the
increase in c-fos mRNA levels during CD treatment, we
measured the transcriptional activity of the c-fos gene by
in vitro nuclear run-on transcription analysis as described
in Chapter 2. Treatment of HeLa cells with cytochalasin D
for 15 minutes resulted in a significant increase in
transcription of the c-fos gene. The data presented in Fig.
7-4 represent a twelve fold increase in c-fos transcription
in cytochalasin D treated cells compared to untreated,


16
specific alpha-actin with respect to molecular weight
(42,000 Daltons), amino acid sequence and polymeric
structure (Tilney and Detmers, 1975; Hartwig and Stossel,
1975; Kane, 1975; Garrels and Gibson, 1976; Whalen et al.
1976; Rubenstein and Spudich, 1977). These results suggest
that beta- and gamma-actin protein in non-muscle cells may
play a role in contraction processes such as cell motility ,
chromosome movement and exocytosis (Berl et al.. 1973;
Huxley, 1973; Sanger, 1975).
Cvtoskeletal-mRNA Interactions
Early studies using stereo-electron microscopic
analysis of whole cells indicated that polysomes are
associated with the cytoskeleton structure (Wolosewick and
Porter, 1977) Closer examination of the cell interior, by
removing membranes and soluble cytoplasmic proteins with
detergent, reveals that many of the ultrastructural features
and nearly all the polysomes are retained on the
cytoskeletal framework (Lenk et al.. 1977). Furthermore, the
non-random topological distribution of the polysomes is
preserved in the cytoskeleton preparation (Fulton et al..
1980). The association of the polysomes with the
cytoskeleton does not appear to be due to trapping since
monosomes and independent ribosomal subunits are efficiently
removed by the detergent extraction (Lenk et al.. 1977;
Cervera et al.. 1981; Ornelles et al.. 1986). These results


12
The Cvtoskeleton: Microtubules. Intermediate Filaments
and Microfilaments
The cytoskeleton structure is an interconnected system
of three major classes of protein fibers, namely the
microtubules, intermediate filaments and microfilaments.
Each system represents a heterogenous population of
proteins. The microtubules, for example, are composed of 13
protofilaments of the well characterized alpha and beta
tubulin proteins (Amos and Klug, 1974). These filaments are
25 nm in diameter and appear to contain a hollow core.
Several microtubule associated proteins (MAPs) have also
been identified which appear to facilitate the assembly of
the microtubules and promote their interactions with other
cellular organelles (Sloboda et al.. 1976, Bulinski and
Borisy, 1980; Bloom et al.. 1984; Binder et al.. 1985). It
has long been recognized that in mitotic cells the
microtubules reorganize into the mitotic spindle including
the astral, kinetochore and continuous fibers which mediate
the segregation of the chromosomes (Weber et al.. 1975).
During interphase the role of the microtubules is less clear
and somewhat controversial. Studies on agents that cause the
depolymerization of the microtubules, such as colchicine and
colcemid, suggest that the microtubules play a role in cell
shape and the intracellular transport of macromolecules and
organelles (Hayden and Allen, 1984; Vale et al.. 1985).


106
LU
LU CO
CC 5
LL
1634
512
344
220
FUSION mRNA


213
Moon, R. Nicosia, R. Olsen, C. Hille, M. and Jeffery, W.
(1983) Dev. Biol. 95, 447.
Morris, T., Marashi, F., Weber, L., Hickey, E., Greenspan,
D., Bonner, J., Stein, J., and Stein, G. (1986) Proc. Natl.
Acad. Sci. U.S.A. 83, 981.
Muller, M., and Blobel, G. (1984) Proc. Natl. Acad. Sci.
U.S.A. 81, 7421.
Muller, R., Bravo, R., Burckhardt, J., and Curran, T. (1984)
Nature 312, 716.
Ng, S. (1989) J. Cell Biol. 107, 79a.
Nielsen, P., Goelz, S., and Trachsel, H. (1983) Cell Biol.
Int. Rep. 7, 245.
O'Farrell, P. (1975) J. Biol. Chem. 250, 4007.
Ohlsson, R., and Pfeifer-Ohlsson, S. (1987) Exp. Cell Res.
173. 1.
Ornelles, D., Fey, E., and Penman, S. (1986) Mol. Cell.
Biol. 6, 1650.
Osborn, M. (1983) J. Invest. Dermatol. 81, 104s.
Pandey, N. and Marzluff, W. (1987) Mol. Cell. Biol. 1_>
4557.
Pederson, T. and Robbins, E. (1970) J. Cell Biol. 4_5, 509.
Peltz, S., and Ross, J. (1987) Mol. Cell. Biol. 7, 4345.
Penman, S., Capeo, D., Fey, E., Chatterjee, P., Reiter, T.,
Ermisch, S., and Wan, K. (1983) in The modern cell biology
series: Spatial organization of eukaryotic cells. McIntosh,
J.R., Ed. (Alan R. Liss, New York) pp 385-415.
Perry, R., and Kelley, D. (1973) J. Mol. biol. 79, 681.
Plumb, M., Stein, J., and Stein, G. (1983) Nucl. Acids Res.
11. 2391.
Prescott, D. (1966) J. Cell Biol. 31, 1.
Rahmsdorf, H., Schonthal, A., Angel, P., Litfin, M., Ruther,
U., and Herrlich, P. (1987) Nucl. Acids Res. 15, 1643.
Randall, L., and Hardy, S. (1989) Science 243. 1156.


71
(32.5 cm x 44 cm) with a 60 well shark toothed comb. The gel
was run at 75 W constant, -2500 V, -40 mA for 3-5 hours
(Xylene Cyanol dye was 15 cm from the bottom of gel). The
gel was soaked in 10% acetic acid-12% methanol in the fume
hood for 1 hour, placed onto 2 sheets of 3 MM Whatman paper,
and dried under vacuum at 80C for 45 minutes. The gel was
exposed to preflashed XAR5 x-ray film at room temperature
for 12-18 hours.
Transfection of DNA into Cells
Introduction of Plasmid DNA into Bacterial Cells
An E_. coli overnight cell culture was diluted 1:100
into 100 ml LB medium and incubated at 37C until the
optical density was approximately 0.37 at 590 nm. The cells
were chilled at 0C for 10 minutes and then centrifuged at
5,000 rpm, 4C for 5 minutes in a Beckman JA20 rotor. The
pellet was resuspended in 20 ml of cold CaCl2 buffer (60 mM
CaCl2, 10 mM Pipes pH 7.0, 15% glycerol) and incubated at
0C for 30 minutes. The cells were collected by
centrifugation at 2,500 rpm for 5 minutes at 4C in a JA20
rotor, resuspended in 2.6 ml CaCl2 buffer and divided into
100-200 ul aliquots. The competent cells were either used
immediately for transformation or stored in microcentrifuge
tubes at -80C.
For transformation, 100 ul competent E^. coli cells were
mixed with 10 ul plasmid DNA (-10 ng) and incubated at 0C
for 10 minutes. The cells were then heat shocked at 37C for


3
are not polyadenylated (Adensik and Darnell, 1972). However,
the genes coding for histone proteins that are synthesized
in the absence of DNA synthesis (cell cycle independent
histones) generally contain introns and code for mRNAs that
are poly-adenylated (Engel et al.. 1982; Harvey et al..
1983; rush et al.. 1985; Wells and Kedes, 1985). Subsequent
discussions will generalize what is true only for the cell
cycle dependent histones unless otherwise indicated.
Histone Gene Regulation: Transcriptional and Post-
Transcriptional Control
The stringent requirement to package newly replicated
DNA into chromatin, the invariable 1:1 molar ratio of
histone protein to DNA and the inability of most eukaryotic
cells to store unbound histone proteins implies a temporal
and functional coupling of histone protein synthesis with
DNA replication (Prescott, 1966; Allfrey et al.. 1963).
Support for this hypothesis first came from studies on
synchronized cell cultures of Euplotes eurvstomus in which
the synthesis of histone protein was found to occur only
during the S phase of the cell cycle, the time when DNA
replication takes place (Prescott, 1966). Furthermore, the
quantity of histone protein within the cell doubled during S
phase. The temporal coupling of histone protein synthesis
with DNA synthesis has since been confirmed and well
documented by a variety of direct experimental approaches


79
removed by aspiration and occasionally material is removed
from within the sucrose pad. This material may be the source
of histone mRNA in the LER fraction, which could decrease
the true percentage of total histone message found on the
nonmembrane-bound polysomes. The HLA-B7 mRNA was located
almost "exclusively on membrane-bound polysomes. Nearly 90%
of the HLA-B7 mRNA was collectively found on the NRER and
MRER, while less than 11% was measured in the LER and
nonmembrane-bound polysomal RNA fractions.
Histone mRNA and HLA-B7 mRNA Distribution Between Csk and
Sol Fractions
Existing within the cytoplasm of eukaryotic cells is an
intricate network of protein filaments, which is referred to
as the cytoskeleton. This structure can be prepared from
HeLa cells treated with non-ionic detergent as described by
Penman and coworkers (Cervera et al.. 1981). We used this
procedure for the isolation of the cytoskeleton (Csk) and
soluble (Sol) fractions which were subsequently assayed for
mRNA content by Northern blot analysis.
Log phase, suspension grown HeLa S3 cells were
fractionated into a Csk phase and Sol phase as described in
Chapter 2. Although this fractionation procedure has been
characterized previously (Lenk and Penman, 1979), we
examined the protein heterogeneity of each fraction by two-
dimensional gel electrophoresis and immunoblotting
techniques in collaboration with Dr. Warren Schmidt. As seen


FREE
NRER
MRER
LER
FREE
NRER
MRER
LER
CONTROL CYCLO
FREE
N RER
M RER
LER
FREE
N RER
MRER
LER
i co r
>
09
*1
00
LO


15
composed of three large polypeptides with molecular weights
of 200,000, 150,000, and 68,000 Daltons (Hoffman and Lasek,
1975; Delacourte et al.. 1980). Glial filaments are
specifically found in astrocytes and other types of glial
cells and consist of a single protein, the glial fibrillary
acidic" protein (GFA), with a molecular weight of 51,000
Daltons (Schachner et al.. 1977; Goldman et al.. 1978).
Although the intermediate filament proteins from the
different cell types are heterogeneous in their chemical
composition, they assemble into similar morphological
structures consisting of filaments that are 10 nm in
diameter that are highly resistant to extraction with
detergents. It has been proposed that the intermediate
filaments contain a constant and variable region, analogous
to the structure of antibodies. The constant domain would
conserve the morphology whereas the variable domain would
confer functional specificity (Lazarides, 1980).
The microfilaments are the thinnest of the cytoskeletal
filament systems, with an average diameter of 6 nm, and are
composed primarily of actin protein. It has been estimated
that approximately 5-10% of total cellular protein is actin
which exists in an equilibrium between monomeric and
filamentous forms (Hartwig and Stossel, 1975).
Microfilaments from non-muscle cells contain two predominant
species of actin protein, namely beta- and gamma-actin. The
non-muscle actin proteins are quite similar to muscle


Figure 4-8. Effects of hydroxyurea treatment on SPH3E1ATG
mRNA.
HeLa cell monolayers were transfected with pSPH3ElATG
DNA and 46 hours post-transfection were treated with 1 mM
hydroxyurea (HU). Cells were harvested at 20 minute
intervals and total cellular RNA was prepared. The RNA (10
/g) was subjected to SI nuclease protection analysis to
quantitate the levels of SPH3E1ATG" mRNA and endogenous H3
histone mRNA. Each lane represents an individually
transfected cell culture. MW is radiolabeled Hpa II digest
of pBR322 DNA. Lanes 1 and 2, control? lanes 3 and 4, 1 mM
HU 20 minutes; lanes 5 and 6, 1 mM HU 40 minutes; lanes 7
and 8, 1 mM HU 60 minutes.


43
disrupted with 15 strokes in a tight fitting Dounce
homogenizer. The homogenate was then centrifuged at 2,000 x
g for 5 minutes. The supernatant (supernatant A) was removed
and maintained in an ice bath while the nuclear pellet
(pellet A) was resuspended in 9 ml TMN (20 mM Tris-HCl pH
7.5, 5 mM MgCl2, 25 mM NaCl, 3 mM dithiothreitol) The
nuclear suspension (pellet A) was adjusted to 1% Triton X-
100 and 1% NaDOC, vortexed briefly and centrifuged at 500 x
g for 10 minutes. The supernatant (supernatant C) from this
centrifugation step was removed and saved on ice for the
isolation of nucleus-associated rough endoplasmic reticulum
(NRER) polysomes. The supernatant from the centrifugation
step following homogenization (supernatant A) was
centrifuged at 12,000 rpm, 4C for 10 minutes in a Beckman
JA20 rotor. The resulting supernatant (supernatant B) was
removed onto ice for subsequent isolation of light
endoplasmic reticulum (LER)-associated polysomal RNA and
nonmembrane-bound polysomal RNA. The pellet (pellet B) was
resuspended in 9 ml TMN, adjusted to 1% Triton X-100 and 1%
NaDOC and centrifuged at 10,000 rpm, 4C for 10 minutes in a
JA20 rotor. The supernatant (supernatant D) was removed and
stored on ice for the isolation of mitochondria-associated
rough endoplasmic reticulum (MRER) polysomal RNA.
Exponentially growing HeLa cells (2.5 x 108 cells) were
collected by centrifugation for the isolation of total
polysomal RNA. The cell pellet was washed in PBS and


Figure 4-3. The subcellular localization of the signal
peptide-histone fusion mRNA.
HeLa cell monolayers were transfected with pSPH3El and
cultured as described in Chapter 2. Forty-six hours after
transfection, cell cultures were harvested and nonmembrane-
bound and membrane-bound polysomes were isolated. The RNAs
from these fractions were analyzed by the SI nuclease
protection assay (5 /g RNA per sample). The distribution of
mRNA species within the subcellular fractions was
quantitated by scanning laser densitometric analysis of the
autoradiograms. Lanes: 1) pBR322 Hinfl molecular weight
marker; 2) nonmembrane-bound polysoma1 RNA; and 3) membrane-
bound polysomal RNA.


20
for the polymerase, nucleocapsid and matrix proteins are all
attached to the cytoskeleton during translation (Cervera et
al. 1981). The association of free polysomes with the
cytoskeleton has also been demonstrated by electron
microscopy (Wolosewick and Porter, 1979). Free polysomes are
generally the class of polysomes involved in the synthesis
of intracellular proteins whereas membrane-bound polysomes
are generally involved in the synthesis of exported proteins
that are destined for the cell surface or secretion (for
review see Emr et al.. 1980). The association of free
polysomal mRNAs with the cytoskeleton suggests that the
polysome is anchored to a scaffold, albeit a dynamic
structure, and not freely diffusible within the cytoplasm.
The term "free polysome" is operationally defined and is a
misnomer when viewed in light of the biochemical and
microscopic studies on the cytoskeleton. The proper
reference to free polysomes is nonmembrane-bound polysomes.
The mechanism by which mRNAs assemble into membrane-
bound polysomes has been described in the signal hypothesis
(Blobel and Dobberstein, 1975). Many types of signal
peptides exist that target proteins to various subcellular
membrane compartments such as the endoplasmic reticulum,
mitochondria and chloroplasts in eukaryotes as well as the
cell membrane in prokaryotes (Date et al.. 1980; Schatz and
Butow, 1983; Muller and Blobel, 1984; Cline, 1986). The work
presented here is focused on events that occur at the


7
affinity binding sites for the newly synthesized histone
proteins. Unbound histone protein may then interact with
histone mRNA-containing polysomes resulting in a
conformational change in the mRNA that is more accessible to
ribonuclease. Growing support for this model comes from a
cell-free mRNA decay system in which the destabilization of
histone mRNA is accelerated when incubated in the presence
of histone proteins (Peltz and Ross, 1987) This result
appears to be specific for histone mRNA since the rates of
decay for gamma globin mRNA, c-myc mRNA, total poly(A)' RNA
or total poly(A)+ RNA are unaffected by the presence of
histone proteins.
Gene transfer studies demonstrated that the minimal
seguence necessary for coupling histone mRNA stability with
DNA replication lies within the 3' terminus of the mRNA
(Stauber et al.. 1986; Levine et al.. 1987; Pandey and
Marzluff, 1987). Removal of only 60 nucleotides from the 3'
end of a murine H3 histone mRNA releases the mutant mRNA
from destabilization during inhibition of DNA synthesis
(Levine et al.. 1987). Conversely, the last 30 nucleotides
of histone H3 mRNA when fused to the 3' end of human gamma
globin mRNA are sufficient to confer instability to the
chimeric mRNA when DNA synthesis is inhibited, whereas wild
type gamma globin mRNA remains stable (Pandey and Marzluff,
1987). Contained within the minimal target sequence for
destabilization is a stem-loop structure which is common to


197
perspective that cytostructure can form subcellular
compartments that are independent of membrane envelopes
(Fulton et al.. 1980; Cervera et al., 1981). Furthermore,
Singer and co-workers, using in situ hybridization
techniques, were demonstrating the localization of specific
mRNAs in discrete areas of the cytoplasm (Lawrence and
Singer, 1986).
Based on these observations and the information
available concerning histone gene expression, we proposed
that histone mRNAs and the factors that are involved in
their selective destabilization during inhibition of DNA
synthesis are co-compartmentalized in a perinuclear region.
This model predicted that the subcellular location of
histone mRNA plays a role in the coupling of histone mRNA
stability with DNA replication.
To address this question, we first determined the
normal subcellular location of histone mRNA. Subcellular
fractionation and Northern blot analysis demonstrated that
histone mRNAs are naturally located on nonmembrane-bound
polysomes that are associated with the cytoskeleton (Chapter
3). The subcellular localization of histone mRNA was then
altered by incorporating nucleotide sequences coding for a
signal peptide into an H3 histone mRNA (Chapter 4). The
modified histone gene, when transfected into HeLa cells,
expressed a signal peptide-histone fusion mRNA which was
directed to membrane-bound polysomes. The histone fusion


95
As with the protocol utilized to separate nonmembrane-
bound and membrane-bound polysomes, the cytoskeleton
preparation obtained must be operationally defined. While
this does not necessarily detract from the biological
relevance, differences in cytoskeleton preparations are
undoub£edly reflected in the biochemical composition of the
isolated complex. Additional characterization of the
cytoskeleton preparation with respect to possible variations
in regions of the complex may provide insight into
mechanisms by which specific mRNAs may be localized and
their functional properties selectively modulated.


155
Monoclonal HeLa cell lines expressing the chimeric gene
were fractionated into cytoskeleton and soluble phases as
described by Cervera et al. (1981). Subsequently, RNA from
each fraction was isolated and SPH3E1 chimeric mRNA and
endogenous H3 histone mRNA content was determined by SI
nuclease protection analysis (Chapter 2). As seen in Figure
6-1 and Table 6-1, both endogenous histone and SPH3E1 mRNAs
are predominantly associated with the cytoskeleton in
control cells (92% and 93%, respectively). Cytochalasin D
treatment [10 /g/ml] brought about a limited release of
SPH3E1 mRNA from the cytoskeleton as expected for a
membrane-bound polysomal mRNA. Only 15% of SPH3E1 mRNA
compared to 40% of endogenous histone mRNA was dissociated
from the cytoskeleton in CD treated cells. As described
above, dissociation of the ribosomes by puromycin during
cytochalasin D treatment is necessary for the efficient
release of membrane-bound polysomal mRNAs from the
cytoskeleton. Surprisingly, CD and puromycin cotreatment
also failed to release the SPH3E1 mRNA from the
cytoskeleton. Greater than 70% of the signal peptide-histone
chimeric mRNA and less than 26% of endogenous H3 histone
mRNA remained associated with the cytoskeleton after CD and
puromycin cotreatment. The selective retention of the
chimeric mRNA on the cytoskeleton during CD and puromycin
cotreatment is demonstrated by the marked increase in the
intensity of the signal obtained for the same quantity of


127
crude yeast extract (Roggenkamp et al.. 1981). In addition,
it has been reported by Weidraann et al. (1984) that
E. coli plasmid pBR322 beta-lactamase mRNA that is
synthesized and capped in vitro. when microinjected into
Xenopus oocytes, is translated into protein that is
ultimately secreted from the cell. These results suggest a
common mechanism of signal peptide recognition among
prokaryotic and eukaryotic organisms.
In an attempt to prevent the translocation of the
signal peptide-histone fusion mRNA to membrane-bound
polysomes, we initially disrupted the hydrophobic region of
the signal peptide by site directed mutagenesis. This
approach was based on previous studies demonstrating that
bacterial secretory proteins that were mutated in the
hydrophobic domain, either by insertion of charged amino
acids or by deletion, accumulate in the cytoplasm of the
bacterium (Emr et al.. 1978? Bedouelle et al.. 1980; Emr and
Bassford, Jr., 1982). The incorporation of positively
charged histidine residues in the hydrophobic domain of the
signal peptide (SPH3Elalpha) however, resulted in only a
partial block in the association of mutated signal peptide-
histone fusion mRNA with membrane-bound polysomes (Figure 4-
6). This result is not surprising in light of more recent
studies on protein export in eukaryotes, which demonstrate
that the relationship between the primary seguence of the
signal peptide and its ability to function as a secretory


34


Figure 5-1. Northern blot analysis of the cvtoskeleton and
soluble phase distribution of H3 histone. H4 histone and
HLA-B7 mRNAs in cvtochalasin D treated cells.
HeLa cells were treated with the indicated
concentration of cytochalasin D for 20 minutes and
cytoskeleton and soluble phase RNAs were isolated. Equal
quantities of RNA per sample (10 /xg/lane) were analyzed by
Northern blot analysis as described in Chapter 2. Lanes: 1)
Control/csk? 2) Control/sol? 3) CD [5 /g/ml]/csk? 4) CD [5
Mg/ml]/sol; 5) CD [10 nq/ml)/csK; 6) CD [10 Mg/ml]/sol; 7)
CD [40 4g/ml]/csk; 8) CD [40 Mg/ml]/sol.


25
replication. These results indicated that the subcellular
location plays a role in the posttranscriptional control of
histone gene regulation.
During the course of studying histone mRNA subcellular
localization it was necessary to examine histone mRNA-
cytoskeletal interactions in detail. Previously, Ornelles et
al (1986) reported that total poly-A+ mRNA, as a
heterogenous population, is released from the cytoskeleton
in HeLa cells by cytochalasin D treatment. We demonstrated
by Northern blot analysis that cytochalasin D treatment of
HeLa cells readily releases histone mRNA, which is not poly-
adenylated, from the cytoskeleton into the soluble phase.
Surprisingly, mRNA coding for the cell surface HLA-B7 class
I antigen remained associated with the cytoskeleton during
cytochalasin D treatment. Subsequent studies demonstrated
that HLA-B7 mRNA expresses at least two distinct
cytoskeletal attachment sites: a cytochalasin D sensitive
site and a puromycin sensitive site. The puromycin sensitive
site appears to be a feature of membrane-bound polysomal
mRNAs and is most likely the nascent polypeptide and/or
ribosome associated with the protein substructure of the
endoplasmic reticulum.
Cytochalasin D and puromycin cotreatment, conditions
that efficiently release HLA-B7 membrane-bound polysomal
mRNA and histone nonmembrane-bound polysomal mRNA from the
cytoskeleton into the soluble phase, did not dissociate the


217
Ullu, E., Murphy, S., and Melli, M. (1982) Cell 29, 195.
Vale, R., Schnapp, B., Reese, T., and Sheetz, M. (1985) Cell
40, 449.
van Venrooij, W., Sillekens, P., van Eekelen, A., and
Reinders, R. (1981) Exp. Cell Res. 135. 79.
Venkatesan, N., and Steele, W. (1972) Biochem. Biophys. Acta
287. 526.
Verma, I., and Sassone-Corsi, P. (1987) Cell 51, 513.
Walter, P., and Blobel, G. (1980) Proc. Natl. Acad. Sci.
U.S.A. 77, 7112.
Walter, P., and Blobel, G. (1981) J. Cell Biol. 91, 557.
Walter, P., and Blobel, G. (1982) Nature 299, 691.
Walter, P., and Blobel, G. (1983) Cell 14, 525.
Walter, P. Gilmore, R., and Blobel, G. (1984) Cell 38, 5.
Wang, Y-L. (1984) J. Cell Biol. 99, 1478.
Watson, M. (1984) Nucl. Acids Res. 12, 5145.
Weber, K. Bibring, T., and Osborn, M. (1975) Exp. Cell Res.
95, 111.
Welch, W., and Feramisco, J. (1985) Mol. Cell. Biol. 5,
1571.
Wells, D., and Kedes, L. (1985) Proc. Natl. Acad. Sci.
U.S.A. 82, 2834.
Whalen, R., Butler-Browne, G., and Gros, F. (1976) Proc.
Natl. Acad. Sci. U.S.A. 23, 2018.
Wiedman, M., Huth, A., and Rapoport, T. (1984) Nature 309,
637.
Wilson, M., and Melli, M. (1977) J. Mol. Biol. 110, 511.
Wolosewick, J., and Porter, K. (1977) Am. J. Anat. 149. 197.
Wolosewick, J., and Porter, K. (1979) J. Cell Biol. 82, 114.
Woodcock-Mitchell, J., Eichner, R., Nelson, W., and Sun, T-
T. (1982) J. Cell Biol. 95, 580.


6
HeLa cells is 45-60 minutes (Sittman et al. 1983? Plumb et
al.. 1983). Hydroxyurea, or any other compound or treatment
that inhibits DNA synthesis studied to date, results in a
rapid and selective destabilization of histone mRNA (half
life of 10-15 minutes) with little or no effect on histone
gene transcription (Sittman et al.. 1983; Plumb et al.,
1983; Baumbach et al.. 1987).
A long-standing observation concerning histone mRNA
turnover during inhibition of DNA synthesis is that the
mechanism is completely dependent on protein synthesis
(Butler and Mueller, 1973; Gallwitz, 1975; Stahl and
Gallwitz, 1977; Stimac et al.. 1983; Baumbach et al.. 1984?
Helms et al.. 1984). Pretreatment of cells with inhibitors
of protein synthesis such as high salt, cycloheximide or
puromycin followed by the addition of a DNA synthesis
inhibitor prevents the destabilization of histone mRNAs and
actually results in an increase in their steady state levels
due to the continued transcription of the histone genes.
This observation led Butler and Mueller and others to
hypothesize that the histone proteins can autoregulate their
own synthesis through a feedback loop operative at the
posttranscriptional level (Butler and Mueller, 1973; Stein
and Stein, 1984; Wu and Bonner, 1985). The model predicts
that at the natural end of S phase or when DNA synthesis is
interrupted with inhibitors a transient accumulation of
unbound histone protein may arise due to a loss in high-


CHAPTER 5
DIFFERENTIAL ASSOCIATION OF MEMBRANE-BOUND AND
NONMEMBRANE-BOUND POLYSOMES WITH CELL CYTOSTRUCTURE
Introduction
Localization of nearly all actively translated
polyribosomes on the cytoskeletal structure indicates that
polysomal mRNAs are not freely diffusible throughout the
cytoplasm (Lenk et al.. 1977; Cervera et al.. 1981; Ornelles
et al.. 1986). This observation is further supported by
subcellular fractionation studies that demonstrate the
association of both "free" (nonmembrane-bound) and membrane-
bound polysomal mRNAs with the cytoskeleton (Cervera et al,,
1981; Jeffery, 1984; Bonneau et al.. 1985).
Previous results indicate that the association of
polysomes with the cytoskeleton is mediated in part by the
mRNA, as treatments which dissociate the ribosomes from the
mRNA, such as heat shock and high salt, release the
ribosomal subunits into the soluble phase while the mRNA
remains attached to the cytoskeleton (Lenk et al.. 1977;
Cervera et al.. 1981; van Venrooij et al.. 1981; Howe and
Hershey, 1984). However, the region of the mRNA and the
cytoskeletal elements to which it is attached are not known.
130


200
mRNA, H3 and H4 histone mRNA together, or any combination of
groupings of the different classes of histone mRNAs. The
pattern for these groupings may be important in the
transport of the newly synthesized histone proteins into the
nucleus and/or the coordinate regulation of histone protein
synthesis.
Although unrelated to the problems addressed here, it
would be interesting from the standpoint of studying the
process of protein secretion to determine the subcellular
localization of the histone fusion protein. Examination of
subcellular fractions isolated from HeLa cells which are
expressing the signal peptide-histone fusion gene, by
standard biochemical techniques, such as acid extraction and
triton-urea gel electrophoresis, would provide insight into
how the fusion protein is processed and where the fusion
protein is located.
In studying the subcellular location of histone mRNA,
it was determined that histone mRNA is predominantly
associated with the cytoskeleton and that this association
is dependent on the integrity of the microfilaments
(Chapters 3 and 5). Cytochalasin D treatment, which disrupts
the actin filaments, releases histone mRNA from the
cytoskeleton into the soluble phase. This result is
consistent with the observation of Penman and co-workers
that heterogeneous poly-A+ RNA is released from the
cytoskeleton in cytochalasin D treated HeLa cells (Ornelles
et aL, 1986) .


163
SPHE1
1 2 3 4 5 6 7 8
SPH3E1ATG-
1 2 3 4 5 6 7 8
HLA-B7
hCG


7-3 Time course of c-fos mRNA accumulation during
cytochalasin D and puromycin treatments in
exponentially growing HeLa cells 181
7-4 Effect of cytochalasin D on transcription of c-fos
in HeLa cells 184
7-5 Effect of protein synthesis inhibition on
cytochalasin D induction of c-fos transcription in
HeLa cells 188
viii


56
mercaptoethanol). Cells were disrupted by homogenization (6-
12 strokes) with a Dounce homogenizer, Wheaton type A
pestle. After >90% of the cells had been lysed, nuclei were
pelleted by centrifugation at 2000 rpm for 10 minutes in an
IEC centrifuge at 4C and resuspended in nuclei storage
buffer "containing 40% glycerol (50 mM Tris-HCl pH 8.3, 5 mM
MgCl2, 0.1 mM EDTA). Nuclei were aliquoted and either snap
frozen in liquid nitrogen or used fresh in the in vitro
transcription reactions. Reactions typically contained 107
nuclei, 100 uCi alpha-32P-UTP (3000 Ci/mmole), l mM ATP,
0.25 mM GTP and CTP in a final volume of 130 ul and were
incubated for 30 minutes with intermittent shaking at 30C.
Radiolabeled RNAs were isolated by treatment of nuclei with
DNase I (100 /g/ml) in the presence of 0.6 M NaCl, 50 mM
Tris-HCl pH 7.5, 20 mM MgCl2 for 15 minutes at room
temperature. The mixture was then incubated with proteinase
K (200 jug/ml) for 3 0-60 minutes at 37C in the presence of
150 mM NaCl, 12.5 mM EDTA, 100 mM Tris-HCl pH 7.5 and 20 mM
MgCl2. Sodium acetate (pH 5.5) was added to 0.2 M and
nucleic acids extracted several times by the hot phenol
method (Clayton and Darnell, Jr., 1983; Sherrer and Darnell,
1962) To the aqueous solution of j2P-labeled RNAs, 150 /g
of yeast RNA and 2.5 volumes of 95% ethanol were added.
Precipitation was overnight at -203C. Radiolabeled
transcripts were resuspended in 10 mM Tris-HCl pH 8.0, 1 mM
EDTA and an aliquot of each sample was precipitated with 150


169
complex that may influence the transcription of specific
genes (Curran et al., 1984; Sytoyama et al., 1986; Distel et
al.. 1987; Rauscher et al., 1988).
Signal transduction models describing the induction of
c-fos gene expression due to the binding of growth factor(s)
at the plasma membrane must include a mechanism for relaying
the stimulatory signal from the cell surface, through the
cytoplasm and into the nucleus. Existing within the
cytoplasm of the cell is an extensive and intricate protein
scaffold referred to as the cytoskeleton (for review see
Nielsen et al.. 1983). In addition to its structural
properties the cytoskeleton also appears to function in cell
shape, cell motility, the intracellular transport of
macromolecules and translation of mRNA into protein.
Furthermore, the cytoskeleton may influence the regulation
of gene expression. Suspension of anchorage dependent mouse
3T6 fibroblasts induces a morphological change in cell shape
and the cytoskeleton with a concomitant reduction in DNA,
RNA and protein synthesis (Benecke et al.. 1978; Farmer et
al.. 1983). Reattachment of these cells to a solid support
restores cell shape and cytoskeleton structure as well as
DNA, RNA and protein synthesis. In addition, disruption of
the microfilaments of the cytoskeleton with cytochalasin D
in HEp-2 cells results in the enhanced expression of beta
and gamma actin protein synthesis (Tannenbaum and Godman,
1983). Removal of the drug restores the structure of the
cytoskeleton and returns actin synthesis to normal levels.


72
5 minutes, diluted with 0.9 ml prewarmed LB medium and
incubated at 37C for 1 hour. Different volumes of the
transfected bacterial cell culture (usually 10-200 ul) were
plated onto the appropriate medium dictated by the genotype
of the host and the introduced plasmid DNA.
Introduction of DNA into HeLa Cells
HeLa monolayer cell cultures were transfected with
plasmid DNA, according to Gorman et al. (1982), in a calcium
phosphate/DNA complex prepared as described by Graham and
van der Eb (1973). HeLa S3 cells growing exponentially in
suspension culture were plated into 20 ml completed EMEM
(EMEM supplemented with 5% fetal calf serum, 5% horse serum,
100 U penicillin, 100 [ig/m1 streptomycin and 1 mM glutamine)
c
at 3x10 cells per 10 cm tissue culture dish (Corning or
Falcon) and incubated overnight at 37C in 5% C02. The
monolayer cell culture was refed with 10 ml completed EMEM
and incubated at 37C in 5% C02 for 4 hours. The DNA/calcium
phosphate precipitate was prepared by adding 20 ¡q DNA in
500 ul 250 mM CaCl2 dropwise to 500 ul 2x Hebs buffer (50 mM
Hepes, 280 mM NaCl, 1.5 mM Na2HP04, pH 7.120.05) while
vortexing the solution. The DNA/CaCl2 precipitate was added
dropwise to the monolayer cell culture and incubation was
continued at 37C in 5% C02 for 4 hours. The HeLa cells were
glycerol shocked by replacing the medium with 2 ml completed
EMEM containing 15% glycerol for 1 minute. The transfected


Figure 5-3. Northern blot analysis of the cvtoskeleton and
soluble phase distribution of nonmembrane-bound and
membrane-bound polvsomal mRNAs in cvtochalasin D. puromvcin.
CD/puro and CD/cvcloheximide treated cells.
HeLa cells were cultured in the presence of the drugs
and fractionated into cytoskeleton and soluble phases as
described in Chapter 2. Equal volumes of cytoskeleton and
soluble phase RNA from each cell culture were assayed for H3
histone, H4 histone, HLA-B7 and chorionic gonadotropin mRNAs
by Northern blot analysis. Upper panel, probed for HLA-B7
and chorionic gonadotropin mRNA, and lower panel, probed for
H3 and"H4 histone mRNA. Upper and lower panels represent the
same Northern filter. Lanes: 1) Control/csk; 2) Control/sol?
3) CD/csk? 4) CD/sol; 5) Puro/csk? 6) Puro/sol? 7)
CD/puro/csk? 8) CD/puro/sol; 9) CD/cyclo/csk; 10)
CD/cyclo/sol.


94
The association of histone mRNAs with the cytoskeleton
(Figure 3-1) is consistent with studies carried out by
Penman and co-workers in which they examined the subcellular
localization of vesicular stomatitis virus (VSV) mRNAs in
HeLa cells following viral infection (Cervera et al.. 1981).
These investigators observed that in VSV-infected HeLa cells
both the viral mRNA coding for the G viral glycoprotein and
the nonmembrane-bound polysomal viral mRNAs were associated
with the cytoskeleton. While the specific nature of the
association of polysomes with the cytoskeleton remains to be
definitively established, the ability to release ribosomes
and retain mRNAs on the cytoskeleton suggests that there may
be a direct attachment of mRNAs to the cytoskeleton (Lenk et
al.. 1977; Lenk and Penman, 1979? Lemieux and Beaud, 1982).
It is therefore reasonable to speculate that the association
of mRNAs with the cytoskeleton plays a functional role in
protein synthesis and/or in the regulation of mRNA
stability. This is a particularly attractive possibility
because the association of histone mRNAs with the
cytoskeleton could permit the compartmentalization of
histone mRNA-containing polysomes in a particular region of
the cytoplasm. The potential for compartmentalization
afforded by the cytoskeleton may also be related to the
rapid and selective destabilization of histone mRNAs at the
natural end of S phase or following inhibition of DNA
synthesis.


51
were placed on areas of the sponge that were not covered by
the gel so as to direct the flow of 2Ox SSC only through the
gel. The filter was directly layered onto the gel and any
air bubbles that were trapped between the gel and filter
were carefully removed. The transfer set-up was completed by
placing two sheets of 3 MM paper followed by 2-3 boxes of
facial tissues (Kleenex brand) on top of the hybridization
filter. The buffer chamber was filled with 20x SSC and
transfer was carried out for 36 hours (note: complete
transfer is usually achieved within 14 hours). After
transfer, the filters were rinsed in 2x SSC for 10 minutes
to remove particles of agarose, air dried and baked in vacuo
at 80C for 2 hours. The filters were stored in sealed
hybridization bags at 4C.
Hybridization of Filters-Immobilized RNA (Northern Blot)
Filters (nylon or nitrocellulose) were prehybridized
for 3-4 hours in 0.5 ml hybridization buffer per cm2 filter
area at 43C for the detection of chorionic gonadotropin
alpha, HLA-B7 and c-fos mRNA and at 48C for histone mRNA.
Hybridization buffer consisted of 50% formamide, 5x SSC, lOx
Denhardt's (lOOx Denhardt's: 2% (w/v) ficoll 400, 2% (w/v)
polyvinylpyrrolidone), 1% SDS, 50 mM sodium phosphate pH
7.0, 20 /xg/ml BSA, and 250 /g/ml E^_ coli DNA. The filters
were hybridized at the appropriate temperature for 36 hours
in 0.1 ml hybridization buffer per cm2 filter containing


182
control cells. In contrast, puromycin treatment for 15
minutes did not significantly affect the transcription of
the c-fos gene. Cotreatment of HeLa cells with puromycin and
cytochalasin D for 15 minutes also stimulated c-fos gene
transcription, which indicates that inhibition of protein
synthesis during CD treatment does not affect the induction
of c-fos transcription.
Tannenbaum and co-workers have demonstrated an increase
in actin transcription during CD treatment (1983). In light
of this observation, we measured the transcription of the
beta-actin gene in our studies as a positive control.
Consistent with previous findings, CD increased actin
transcription, but only 5 fold over control cells (Fig. 7-
4). Cotreatment with CD and puromycin for 15 minutes also
increased transcription of the actin gene 5 fold compared to
controls and presumably reflects the effects of the
cytochalasin D (Fig. 7-4). The cytochalasin D stimulation of
c-fos and beta-actin transcription is unaffected by
inhibition of protein synthesis.
The elevation in c-fos and actin transcription during
cytochalasin D treatment may be part of a general phenomenon
in which the transcriptional activity of all genes becomes
increased. Alternatively, transcription of ribosomal RNA
genes by RNA polymerase I, which represents approximately
50% of the transcriptional activity measured by nuclear run-
on assays (Marzluff and Huang, 1985), may be inhibited by


CHAPTER 6
HETEROGENEOUS PATTERN FOR CYTOSKELETAL-mRNA INTERACTIONS
Introduction
In the preceding chapter we presented evidence
indicating that eukaryotic mRNAs are associated with the
cytoskeleton in a heterogeneous manner. Disruption of the
microfilaments by cytochalasin D treatment readily releases
nonmembrane-bound polysomes from the cytoskeleton into the
soluble phase, whereas membrane-bound polysomes remain
attached to the cytoskeleton. The cytoskeletal association
of membrane-bound polysomal mRNAs occurs through at least
two distinct sites: a cytochalasin D sensitive site and a
puromycin sensitive site, which is the interaction of the
nascent polypeptide/polysome with the protein substructure
of the endoplasmic reticulum (Chapter 5). To release
membrane-bound polysomes from the cytoskeleton, both
attachment sites must be disrupted as seen during
cytochalasin D and puromycin cotreatment. Previously, we
have demonstrated that a signal peptide-histone fusion mRNA
(SPH3E1) is targeted from nonmembrane-bound polysomes, the
natural site of histone protein synthesis, to membrane-bound
polysomes (Chapter 4). Relocating the histone mRNA within
153


129
These results indicate that the stability of SPH3E1
mRNA during inhibition of DNA replication is functionally
related to the change in the subcellular location of the
mRNA. The differential stability of SPH3E1 mRNA and
SPH3E1ATG* mRNA may be due to qualitative differences in the
composition of membrane-bound and nonmembrane-bound
polysomes; however, there is no previously reported evidence
to support this possibility. Alternatively, the presence of
the fusion mRNA on membrane-bound polysomes rather than
nonmembrane-bound polysomes, where histone mRNA normally
resides, may physically separate the message from the
factors that are involved in the selective destabilization
of histone mRNA during inhibition of DNA synthesis. These
results are consistent with the hypothesis that the
subcellular location of histone mRNA plays an important role
in the posttranscriptional regulation of histone gene
expression.


126
resulted in the destabilization of approximately 70% of
SPH3Elalpha mRNA, a value which is intermediate to that
observed for SPH3E1 mRNA and SPH3E1ATG mRNA (Figure 4-9
lanes 1,2 and 7,8; Fig. 4-10). The degree to which the
signal peptide-histone fusion mRNAs are degraded during
inhibition of DNA synthesis is correlated with the extent to
which the mRNA is associated with nonmembrane-bound
polysomes. These results suggest that the subcellular
localization of the histone mRNA plays a significant role in
the coupling of its stability to DNA replication.
Discussion
To our knowledge, this represents the first
demonstration that the addition of a signal peptide coding
sequence to an mRNA that is normally translated on
nonmembrane-bound polysomes results in the targeting of the
chimeric message to membrane-bound polysomes in intact
mammalian cells. The recognition of a prokaryotic signal
peptide in a eukaryotic cell is supported by both in vitro
and in vivo studies (Lingappa et al.. 1984; Talmadge et al.,
1980a; Talmadge et al., 1980b). Saccharomyces cerevisiae has
been shown to express the Eh_ coli pBR325 beta-lactamase gene
in vivo and to process the precursor protein to the
enzymatically active, mature protein (Roggenkamp et ah,
1981). The processing of the bacterial pre-protein into the
mature species has also been demonstrated in vitro using a


57
Mg yeast RNA and cold 10% TCA. TCA-precipitable counts were
determined by liquid scintillation spectrometry.
DNA excess hybridization has also been described by
Baumbach et al. (1987) In a preliminary experiment we
determined that 2 fig of H4 histone insert in pF0002 was at
least two-fold DNA excess when hybridized to in vitro
nuclear run-on transcripts. Southern blots containing
restriction endonuclease digested DNA fragments or slot
blots containing linearized plasmid DNAs were prepared.
Southern blots on nitrocellulose or slot blots on Nylon were
prehybridized in 1 M NaCl, 20 mM Tris-HCl pH 7.4, 2 mM EDTA,
0.1% SDS, 5x Denhardt's, 250 mg/ml E_¡_ coli DNA and 12.5 mM
sodium pyrophosphate at 65C for at least 6 hours.
Hybridizations were conducted at 65C for 72 hours in 1 M
NaCl, 20 mM Tris-HCl pH 7.4, 2 mM EDTA, 0.1% SDS, 2.5x
Denhardt's, 250 g/ml E^_ coli DNA with i2P-labeled
transcripts at 5xl05-lX10D TCA-precipitable counts per ml of
hybridization solution. Blots were washed at 65C for 15
minutes in fresh prehybridization solution [0.13-0.14
ml/cm2], 1 hour in 2x SSC/0.1% SDS [0.5 ml/cm2], overnight
in 2x SSC/0.1% SDS [0.5 ml/cm2] and 1 hour in 0.2x SSC/0.1%
SDS [0.5 ml/cm2] Air dried filters were autoradiographed
with XAR-5 or Cronex film and Cronex lightning Plus screens
at -70C for varying periods of time and developed with an
X-O-MAT X-ray film processor.


173
To address the possibility that the stimulation of c-fos
mRNA levels was due to the partial inhibition of protein
synthesis, we investigated the effect of puromycin at a
concentration of 0.4 mM, which inhibits protein synthesis by
more than 90 percent (Helms et al.. 1984), on c-fos gene
expression. Cells cultured in the presence of puromycin for
15 minutes exhibited 9 fold higher levels of c-fos mRNA
compared to untreated cell cultures (Fig. 7-1, lane 3; and
Figure 8-2) However, cotreatment with CD and puromycin for
only 15 minutes resulted in a 49 fold increase in c-fos mRNA
levels (Fig. 7-1, lane 4 and Fig. 7-2) The superinduction
of c-fos mRNA cellular levels by cytochalasin D and
puromycin cotreatment suggests that the stimulation by CD
alone is not due solely to its effect on protein synthesis
and may be the result of the perturbation of the
cytoskeleton.
Kinetics of c-Fos mRNA Accumulation During Cytochalasin D
Treatment
Treatment of quiescent BALB/C-3T3 mouse fibroblast
cells with growth factors or serum results in a rapid and
short lived increase in c-fos mRNA levels (Cochran et al.,
1984; Greenberg and Ziff, 1984; Muller et al.. 1984).
Generally, cellular levels of c-fos mRNA are maximal after
30 minutes of serum/growth factor treatment and return to
pre-stimulatory levels within an hour. In contrast to serum


47
Isolation of Cvtoskeleton and Soluble Phase RNA
The cytoskeleton and soluble fractions were prepared as
previously described by Penman and coworkers (Cervera et
al.. 1981). Exponentially growing HeLa cells (5 x 108 cells)
were harvested and washed in cold PBS. The cell pellet was
resuspended in 2 ml extraction buffer (10 mM Pipes pH 6.8,
100 mM KC1, 2.5 mM MgCl2, 0.3 M sucrose, 1 mM
phenylmethylsulfonyl fluoride) and Triton X-100 was
immediately added to a final concentration of 0.5%. The
sample was incubated at 0c for 3 minutes and then
centrifuged at 700 rpm in an IEC centrifuge for 3 minutes at
4C. The supernatant (SOL) was removed and stored on ice.
The pellet was resuspended in 8 ml RSB containing 1 mM
phenylmethylsulfonyl fluoride, 1% Tween 40 and 0.5% NaDOC.
The nuclei were stripped of cytoplasmic tags with 15 strokes
in a precision-bore stainless steel homogenizer with a
clearance of 0.002 inches and were pelleted at 2,000 rpm for
3 minutes at 4C in an IEC rotor. The supernatant (CSK) was
removed and saved at 0C for subseguent RNA isolation. Both
the CSK and SOL fractions were extracted in the presence of
1% SDS and 0.3 M NaCl with phenol and chloroform. Nucleic
acids were precipitated from the agueous phase with 3
volumes of 95% ethanol at -20C in the presence of 53 mM
potassium acetate. The RNA was digested with DNase I, phenol
and chloroform extracted, and ethanol precipitated as
described above for the isolation of total cellular RNA.


Table 3-1. Quantity of histone and HLA-B7 mRNAs in the
subcellular fractions.
82
H3
H4
HLA-B7
Cytoskeleton
90%
84%
97%
Soluble
10%
16%
3%
NRER
3%
3%
29%
MRER
6%
5%
60%
Free
51%
54%
5%
LER
40%
38%
6%
The distribution of histone and HLA-B7 mRNAs in the
NRER, MRER, LER and nonmembrane-bound polysomal fractions as
well as the cytoskeleton and soluble fractions was
determined by Northern blot analysis. The values have been
normalized with respect to the total yield of RNA within
each fraction.


3) SUBCELLULAR LOCATION OF HISTONE mRNAs ON CYTOSKELETON
ASSOCIATED NONMEMBRANE-BOUND POLYSOMES IN HELA S3
CELLS
Introduction 75
Results 77
Discussion 90
4) SUBCELLULAR LOCATION OF HISTONE UlRNA PLAYS A ROLE IN
THE POSTTRANSCRIPTIONAL REGULATION OF HISTONE
GENE EXPRESSION
Introduction 96
Results 97
Discussion 126
5) DIFFERENTIAL ASSOCIATION OF MEMBRANE-BOUND AND
NONMEMBRANE-BOUND POLYSOMES WITH CELL CYTOSTRUCTURE
Introduction 130
Results 131
Discussion 146
6) HETEROGENEOUS PATTERN FOR CYTOSKELETAL-ltlRNA
INTERACTIONS
Introduction 153
Results 154
Discussion 166
7} THE INFLUENCE OF THE CYTOSKELETON ON THE REGULATION
OF c-Fos GENE EXPRESSION
Introduction 168
Results 171
Discussion 190
8) SUMMARY AND FUTURE CONSIDERATIONS 196
REFERENCES 206
BIOGRAPHICAL SKETCH 219
v


110
It was possible that the stability of the histone
fusion mRNA during inhibition of DNA synthesis was due to
the targeting of the mRNA to a foreign subcellular
compartment, namely the membrane-bound polysomes.
Alternatively, the nucleotide sequences coding for the
signal "peptide may have disrupted the histone mRNA structure
in such a way that it was no longer recognized by the
factors involved in the destabilization of histone mRNA. To
test these possibilities, the signal peptide was mutated in
order to retain the fusion mRNA, without significantly
changing the mRNA structure, on nonmembrane-bound polysomes.
If the class of polysomes plays a role in the coupling of
histone mRNA stability with DNA replication then the mutated
fusion mRNAs that are retained on nonmembrane-bound
polysomes should be efficiently destabilized during
inhibition of DNA synthesis. Conversely, if the mRNA
structure has been disrupted by the nucleotide sequences
encoding the signal peptide then the mutated histone fusion
mRNAs localized on nonmembrane-bound polysomes should be
stable during inhibition DNA synthesis.
The signal peptide-histone fusion gene was mutated by
site directed mutagenesis (Zoller and Smith, 1983) as
described in Chapter 2 and the corresponding mRNAs are
schematically diagrammed in Figure 4-5. The first mutant,
pSPH3Elalpha, contains two point mutations in the region
coding for the hydrophobic domain of the signal peptide,


150
by the apparent cell type specific differences that are
observed for the components of the cytoskeleton that are
involved in the attachment of mRNA and polysomes. Adams et
al. (1983) found that polyribosomes in rat liver cells are
associated with the microfilaments and are released from the
cytoskeleton by deoxyribonuclease I (DNase I) an enzyme
that can depolymerize actin. Pretreatment of the rat liver
cells with phalloidin, a compound that stabilizes actin
filaments, abrogates the DNase I effects on the
polyribosome-cytoskeleton interactions (Adams et al.. 1983).
Consistent with this result, poly A+ RNA, translation
factors and nearly 100% of the polysomes are preferentially
associated with the cytoskeleton in HeLa cells and can be
released into the soluble phase by the disruption of the
microfilaments with cytochalasin B or cytochalasin D (Lenk
et al.. 1977; Howe and Hershey, 1984; Ornelles et al.,
1986). In contrast, Jeffery (1984) reported that the
association of actin mRNA and histone mRNA with the
cytoskeleton in ascidian eggs is not affected by
cytochalasin B treatment, which suggests that the
association is independent of the microfilaments.
Furthermore, disruption of the microfilaments by
cytochalasin B treatment does not release poly A+ histone H4
mRNA from the cytoskeleton in L6 rat myoblast cells (Bagchi
et al_i_, 1987) .


This dissertation was submitted to the Graduate Faculty of the
College of Medicine and to the Graduate School and was accepted as
partial fulfillment of the requirements for the degree of Doctor
of Philosophy.
December 1989
Dean, College of Medicine
Dean, Graduate School


78
fractions. The light endoplasmic reticulum (LER), which is
deficient in polysomes by definition, was also isolated
following this subcellular fractionation scheme. Histone and
HLA-B7 mRNA content in each of the subcellular fractions was
determined by Northern blot analysis. Briefly, RNA was
extracted from the samples, quantitated and electrophoresed
in agarose under denaturing conditions. The RNA was then
transferred to nitrocellulose filters for hybridization with
32P-labeled (nick-translated) pF0108A (H4 histone), pFF435C
(H3 histone) and pDPOOl (cDNA clone of HLA-B7 class I
antigen) DNA probes. A typical screening of subcellular
fractions with these probes is represented by the
autoradiograph in Figure 3-1, and Table 3-1 summarizes the
numerical values derived from densitometric analysis of the
Northern blot. The HLA-B7 mRNA and histone mRNA species are
each present in total cellular RNA and total polysomal RNA
which were included as positive controls. Non-membrane-bound
polysomal RNA contains approximately 55% and 51% of the
total H4 and H3 histone mRNA, respectively. Surprisingly,
the LER fraction contains approximately 38% of the H4
histone mRNA and 41% of the H3 histone mRNA, which may be a
consequence of the method by which the fraction is isolated.
The LER and non-membrane-bound polysomal sample is
centrifuged through a 2 M sucrose pad under conditions where
the LER forms a loose band on top of the pad, while the
nonmembrane-bound polysomes are pelleted. The LER band is


Figure 7-4. Effect of cvtochalasin D on transcription of el
fos in HeLa cells.
Exponentially growing HeLa cells were treated with
cytochalasin D, puromycin, or CD and puromycin for 15
minutes. Radiolabeled RNAs were prepared by in vitro nuclear
run-on transcription and hybridized to Southern blots of
electrophoretically separated restriction enzyme digested
plasmid DNAs (107 cpm/filter). Insert DNAs hybridizing to
transcripts from the corresponding genes are indicated (both
Ribo signals represent 18S ribosomal RNA transcripts).
Untreated, DMSO only; Cytochalasin D, CD [10 jug/ml]?
Puromycin, puromycin [0.04 mM] ; Puro/Cyto, cotreatment with
CD [10 g/ml] and puromycin [0.04 mM]. Autoradiograms in
upper panel for each treatment represent a 28 hour exposure.
Autoradiograms in lower panel, indicating Fos and H4 histone
transcription, represent a 4 day exposure.


48
Typical yields of RNA were 250-300 ng from CSK and 100-150
ng from SOL per 2.5 x 107 HeLa cells (50 ml exponentially
growing suspension culture).
Two-Dimensional Gel Electrophoresis and
Immunob1ottino Analysis
The cytoskeleton and soluble fractions were isolated as
described above and each sample was dialyzed in 2 mM Tris-
HC1 pH 7.5 and treated with DNase I (20 /xg/ml) and RNase A
(20 iig/ml) in the presence of 0.5 mM MgCl2 at 4C for 1
hour. The samples were dialyzed against 2 mM Tris-HCl pH 7.5
and then lyophilized. Two-dimensional gel electrophoresis
and immunoblotting analyses were performed by Dr. Warren
Schmidt (Department of Pathology, Vanderbilt University,
Nashville, TN). The protein preparations were focused in the
first dimension using equilibrium conditions as described by
O' Farrell (1975) with pH 3 to 10 or 5 to 7 ampholines. The
second dimension was electrophoresed on slab gels containing
a 3% stacking gel and 7.5% resolving gel as described by
Laemmli (1970). The proteins were then either stained with
Coomassie Brilliant Blue or electrotransferred to
nitrocellulose filters as described by Towbin et al. (1979).
Where indicated, immunotransfer analyses (Glass et al..
1981) were performed with commercially available or other
(Schmidt et al.. 1982) antibodies to actin, tubulin and
prekeratin.


73
cells were then washed with 10 ml EMEM, refed with 20 ml
completed EMEM and incubated at 37C, 5% C02.
Establishment of Polyclonal and Monoclonal Stablv
Transformed Cell Lines
Polyclonal and monoclonal cell lines were isolated by
growing HeLa cell cultures that were co-transfected with
pSV2neo and the plasmid DNA of interest in medium containing
the antibiotic G418. The plasmid pSV2-neo, constructed by
Southern and Berg (1982), carries the bacterial
aminoglycoside phosphotransferase 3' (II) gene under the
control of the SV40 early promoter and confers resistance to
the aminoglycoside antibiotic G418. The rationale behind the
co-transfection method is that HeLa cells that are competent
to take up pSV2-neo DNA are egually competent to take up the
plasmid DNA of interest, and therefore, a certain percentage
of the G418 resistant cells contain both species of DNA.
Ideally, the selectable marker should be carried by the
vector that contains the gene which is to be studied.
HeLa cells in monolayer culture were co-transfected
with the plasmid DNA to be studied and pSV2-neo DNA (molar
ratio of 20:1, respectively) by the calcium phosphate
precipitation method as described above. After 36-48 hours
post-glycerol shock, the transfected cells were resuspended
in 5 ml Puck saline A [0.04% (w/v) KC1, 0.8% (w/v) NaCl,
0.035% (w/v) NaHC03, 0.1% (w/v) glucose] containing 0.02%


Figure 4-1. Schematic diagram of endogenous H3 histone mRNA
ST519 and the beta-lactamase signal peptide-H3 histone
fusion mRNA.
The signal peptide-histone fusion mRNA is identical to
the wild type histone mRNA with the exception of the signal
peptide encoded sequences that have been inserted in frame
into the histone gene. The 5* mRNA sequences and the 3' mRNA
sequences that have been implicated in coupling histone mRNA
stability to DNA synthesis are retained in place in the
signal peptide-histone fusion mRNA (Morris et al.. 1986;
Pandey and Marzluff, 1987). Boxed areas represent histone H3
mRNA sequences, single lined areas represent beta-lactamase
signal peptide derived sequences.


Figure 4-7. Effects of hvdroxvurea treatment on SPH3E1 mRNA.
HeLa cell monolayers were transfected with pSPH3El DNA
and 46 hours post-transfection were treated with 1 mM
hydroxyurea (HU). Cells were harvested at 20 minute
intervals and total cellular RNA was prepared. The RNA (10
^g) was subjected to SI nuclease protection analysis to
quantitate the levels of wild type signal peptide-histone
fusion mRNA and endogenous H3 histone mRNA. Each lane
represents an individually transfected cell culture. MW is
radiolabeled Hpa II digest of pBR322 DNA. Lanes 1 and 2,
control? lanes 3 and 4, 1 mM HU 20 minutes? lanes 5 and 6, 1
mM HU 0 minutes? lanes 7 and 8, 1 mM HU 60 minutes.


46
resuspended in 15 ml TMN. The sample was adjusted to 1%
Triton X100 and 1% NaDOC, disrupted with several strokes in
a tight fitting Dounce homogenizer and centrifuged at 12,000
x g at 4C for 10 minutes in a JA20 rotor. The supernatant
was removed and saved at 0C.
Each of the fractions (NRER, MRER, LER, nonmembrane-
bound, and total polysomal RNA) were layered onto a 3 ml 2 M
sucrose pad (prepared in RSB containing 100 uM spermidine)
and centrifuged at 4C for 2 hours at 48,000 x g in a
Beckman Type 50.2 Ti rotor (200,000 x g) without the brake.
The LER formed a cloudy band at the top of the sucrose pad
and was carefully removed by aspiration. The remaining
fractions were pelleted under these conditions and gently
resuspended in 3 ml TMN buffer. The samples were then
extracted extensively with phenol and chloroform as
described above for the isolation of total cellular RNA. The
RNA was precipitated from solution with 3 volumes of 95%
ethanol at -2 0C in the presence of 53 mM potassium acetate.
The RNA was digested with DNase I, extracted with phenol and
chloroform, and ethanol precipitated as described above for
the preparation of total cellular RNA. Typical yields were
1.5-2.0 mg MRER and NRER RNA, 2-3 mg nonmembrane-bound RNA,
1 mg LER RNA and 4-6 mg total polysomal RNA per liter of
exponentially growing HeLa cells.


186
were incubated in an equivalent concentration of DMSO for 30
minutes. Treatment of HeLa cells with cytochalasin D for 15
minutes resulted in the expected stimulation of c-fos
transcription (Fig. 7-5). Pretreatment of HeLa cells with
puromycin did not block the CD induced stimulation of c-fos
transcription and in fact, markedly increased the
transcription of the c-fos gene (Fig. 7-5). These results
suggest that the transcription factors are present within
the cells prior to cytochalasin D treatment, which is
consistent with previous findings that the factors involved
in the increase in c-fos transcription during serum
stimulation of quiescent fibroblasts are pre-existent
(Sassone-Corsi and Verma, 1987; Greenberg et al.. 1986).
Cytochalasin D Releases c-Fos mRNA from the Cvtoskeleton
into the Soluble Phase
The marked increase in c-fos mRNA cellular levels
during cytochalasin D treatment is not completely accounted
for by the increase in c-fos transcription and may be partly
a result of enhanced mRNA stability. Previous results
indicated that cytochalasin D [>10 releases poly-A+
RNA from the cytoskeleton into the soluble phase in a dose
dependent manner (Ornelles et al.. 1986). The stabilization
of c-fos mRNA during cytochalasin D treatment may therefore
be a consequence of an alteration in mRNA subcellular
location. To begin addressing this possibility, we have


61
Site-Directed Mutagenesis
In the past, researchers had to rely on nature to
produce the proper mutations to study how the normal
biological processes function. Zoller and Smith (1983) have
designed a powerful method for selectively introducing
singlenucleotide mutations into specific sites of any
cloned, sequenced region of DNA (site-directed mutagenesis).
In addition, this procedure can be used to selectively
create deletion or insertion mutations in DNA fragments of
known sequence. The general approach of site-directed
mutagenesis is to incorporate an oligonucleotide containing
the appropriate mutation into the complementary strand of
recombinant M13 DNA. The heteroduplex is transfected into
bacteria and the resulting mutated bacteriophage are
isolated by various selection processes.
As discussed in Chapter 4, the signal peptide-histone
fusion mRNA is associated with membrane-bound polysomes and
is stable during inhibition of DNA synthesis. It was unclear
whether the uncoupling of the histone fusion mRNA stability
from DNA replication was due to the altered subcellular
location or to a change in mRNA structure, due to the
nucleotide sequence coding for the signal peptide. To
address these possibilities, we have introduced point
mutations into the signal peptide coding region with the
intention to inactivate signal peptide function. In doing
so, mRNA structure is perturbed as little as possible and


58
Radiolabelinq DNA for Northern Blot Analysis
DNA used for Northern blot analysis was uniformally
labeled by nick translation as described by Maniatis et al
(1982b) or by the random priming method of Roberts and
Wilson (1985). Radiolabeled probes prepared in this manner
are also suitable for other purposes such as Southern blot
analysis.
Random Priming Method
This protocol is very efficient for labeling small to
intermediate-sized DNA fragments (100-2000 bp in length) and
is the preferred method for work with purified DNA inserts.
Supercoiled plasmid DNA is labeled much less efficiently by
the random priming method and should be fragmented by
sonication prior to use. Routinely, specific activities of
lxlO9 cpm/jxg DNA are achieved using as little as 20-100 ng
DNA. Material for oligolabeling DNA fragments was purchased
from Bethesda Research Laboratories as a kit.
The isotope (50 uCi [a-32P] dCTP) was vacuum dried and
resuspended in 10 ul of 2.5x reaction buffer (500 mM Hepes
pH 6.6, 12.5 mM MgCl2, 25 mM beta-mercaptoethanol, 125 mM
Tris-HCl pH 8, 1 mg/ml BSA, 7.5 mg/ml synthetic DNA
oligodeoxynucleotide primers, 50 uM each dGTP, dATP, dTTP).
The DNA fragment (100 ng) in 14 ul ddH?0 was heated at 100C
for 2 minutes and immediately chilled on ice. The denatured
DNA was added to the 1.5 ml microcentrifuge tube containing


198
mRNA, in contrast to endogenous histone mRNA, was not
efficiently destabilized when DNA synthesis was interrupted
with hydroxyurea. Site directed mutagenesis analysis of the
histone fusion gene and subsequent .in vivo studies indicated
that the uncoupling of the histone fusion mRNA stability
from DA replication was due to the altered subcellular
location and was not due to the perturbation of mRNA
structure.
It has since been demonstrated by in situ hybridization
analysis that histone mRNAs are localized in "grape like
clusters" throughout the cytoplasm (Lawrence et al.. 1988).
In this regard, the initial proposal that histone mRNAs are
compartmentalized in a region close to the nucleus was
incorrect. The non-uniform distribution of histone mRNAs
within the cytoplasm, however, suggests that histone mRNAs
are compartmentalized. The stability of histone fusion mRNA
during inhibition of DNA synthesis, due to the association
of the mRNA with membrane-bound polysomes, is consistent
with the possibility that these compartments are involved in
the posttranscriptional regulation of histone gene
expression.
The proposal that histone protein synthesis is
autogenously regulated provides one possible explanation for
the role of the subcellular compartments in the
posttranscriptional regulation of histone gene expression.
As proposed by Butler and Mueller (1973) and several other


181
c-FOS


178
and growth factor stimulation, inhibition of protein
synthesis produces a larger accumulation of c-fos mRNA
levels which persists for longer periods of time, usually
several hours (Greenberg et al.. 1986; Rahmsdorf et al..
1987; Andrews et al.. 1987). To determine whether the c-fos
gene response to cytochalasin D treatment corresponds to
that observed for serum/growth factor stimulation or
inhibition of protein synthesis, we examined the kinetics of
c-fos accumulation during CD treatment. Northern blot
analysis of total cellular RNA isolated from cells treated
with 10 jig/ml cytochalasin D for various lengths of time
demonstrated that maximal induction of c-fos mRNA,
approaching a 40 fold increase over control cultures,
occured after 30 minutes and was followed by a sharp decline
to near pre-stimulatory levels within one hour (Fig. 7-3,
lanes 4-9). In contrast, inhibition of protein synthesis
with puromycin resulted in maximal accumulation of c-fos
mRNA levels after 90 minutes, nearly 130 fold higher than in
untreated cells, and the mRNA levels remained elevated for
at least two hours (Fig. 7-3, lanes 16-20). The accumulation
of c-fos mRNA in CD and puromycin co-treated cells followed
the same time course as observed for puromycin treatment and
reached higher levels than with the CD treatment and
puromycin treatment combined (Fig. 7-3, lanes 10-15).
Parallel studies were done on serum starved WI38 human
diploid fibroblast cells. Stimulation of quiescent cell


128
signal is quite variable and can withstand substantial
insertion, substitution or deletion mutations (Kaiser and
Botstein, 1986? Kaiser et al.. 1987; Randall and Hardy,
1989). Subsequently, the signal peptide ATG translation
start codon of SPH3E1 was altered by site directed
mutagenesis which completely inactivated signal peptide
function (Figure 4-6).
The relative stability of signal peptide-histone fusion
mRNA (SPH3E1) during inhibition of DNA synthesis appears to
be a result of the subcellular location of the mRNA.
Destabilization of SPH3E1ATG mRNA, which differs from
SPH3E1 mRNA by a single nucleotide, with essentially the
same kinetics and to the same extent as endogenous H3
histone mRNA indicates that SPH3E1 mRNA expresses the proper
structural confirmation for recognition by the factors that
mediate the selective destabilization of histone mRNA during
inhibition of DNA synthesis.
The difference in SPH3E1 and SPH3E1ATG mRNA stability
during inhibition of DNA synthesis may reflect a
perturbation of protein structure due to the signal peptide
at the N-terminus of SPH3E1. This possibility does not
appear likely based on the observations that histone amino
acid sequences are not required to target the mRNA for
destabilization (Pandey and Marzluff, 1987) and SPH3Elalpha
mRNA, which expresses the mutated form of the signal
peptide, is moderately destabilized during inhibition of DNA
synthesis (Figures 4-8 and 4-10).


133
Membrane-Bound Polysomal RNA is not Released from the
Cytoskeleton bv Cvtochalasin D Treatment
HLA-B7 is a class I histocompatibility antigen that is
located on the cell surface (Robb et al.. 1978). Previously,
HLA-B7 mRNA was shown to be associated with the cytoskeleton
and translated into protein on membrane-bound polysomes
(Zambetti et al., 1985). As seen in Figure 5-1, HLA-B7 mRNA
was almost exclusively associated with the cytoskeleton in
control cells (lanes 1 and 2) and remained largely
associated with the cytoskeleton in cytochalasin D treated
cells (lanes 3-8). Approximately 98% of the HLA-B7 mRNA was
associated with the cytoskeleton in untreated cell cultures
(Table 5-1). In contrast to histone mRNA and total poly-A+
RNA, only 11% of the HLA-B7 mRNA was released from the
cytoskeleton in cell cultures treated with 10 /ig/ml
cytochalasin D (Table 5-1).
The retention of HLA-B7 mRNA on the cytoskeleton during
cytochalasin D treatment may be a consequence of translating
the mRNA on membrane-bound polysomes. To examine this
possibility, the distribution of chorionic gonadotropin
alpha mRNA within the cytoskeleton and soluble fractions was
determined in cytochalasin D treated cells. Chorionic
gonadotropin is a secreted, heterodimeric protein composed
of alpha and beta subunits (Milsted et al., 1985). The alpha
subunit is expressed in exponentially growing HeLa cells
(Milsted et al.. 1985), and Northern blot analysis of


113
which results in the substitution of leucine at amino acid
position 10 (GAA-*GTA) and proline at position 12 (GGG-+GTG)
with positively charged histidine residues. The second
mutant, pSPH3ElATG', contains a single nucleotide
substitution which destroys the ATG translation initiation
codon of the signal peptide (ATG-^TTG). This mutation should
result in the initiation of translation at the natural ATG
codon of the histone coding region without the synthesis of
the signal peptide.
The distribution of the SPH3E1, SPH3Elalpha and
SPH3E1ATG signal peptide-histone fusion mRNAs in
nonmembrane-bound and membrane-bound polysome fractions from
transfected HeLa monolayer cell cultures was determined by
SI nuclease protection analysis. As seen in Figure 4-6, the
SPH3E1ATG" mRNA partitioned into the nonmembrane-bound
polysome fraction to the same extent as endogenous histone
H3 mRNA. Approximately 82% of the SPH3E1ATG" mRNA and 83%
endogenous histone mRNA was found associated with the
nonmembrane-bound polysomes (Table 4-1). In contrast, SPH3E1
mRNA was predominantly associated with membrane-bound
polysomes (Fig. 4-6). Approximately 68% of SPH3E1 mRNA was
localized in the membrane-bound polysomal RNA fraction
(Table 4-1). The SPH3Elalpha mRNA displayed a more
intermediate association with membrane-bound polysomes;
approximately 40% of the SPH3Elalpha mRNA was localized in
the membrane-bound polysome fraction and 60% in the
nonmembrane-bound polysomal fraction (Fig. 4-6 and Table 4-


35
digested with Ncol, which cuts at the ATG translation start
codon, and ligated to Ncol/dephosphorylated pSPpSTdeltaHH/E
DNA using T4 DNA ligase. The signal peptide-histone fusion
gene was subcloned into pUC8 plasmid DNA by ligating the 1.1
kbp Hpal fragment from pSPpSTdeltaHH/EpST into the Hindi
site o pUC8 to form pSPH3. The signal peptide-histone
fusion gene in pSPH3 was transcriptionally enhanced by
placing the SV40 viral enhancer element from pSVE108A
(Shiels et al.. 1985) into the EcoRI site in the polylinker
of pSPH3 to form pSPH3El (containing one enhancer element)
and pSPH3E2 (containing 2-3 enhancer elements).
DNA Isolation and Purification
Plasmid DNA Preparations
Plasmid DNAs used in the experiments described here
were harbored in Escherichia coli strain HB101 bacteria (F~,
recA 13) unless otherwise indicated. Bacteria containing the
appropriate plasmid were grown in medium containing
antibiotics (usually 25-50 jug/ml ampicillin or tetracycline)
as specified by the genotype of the plasmid or bacterium.
The following procedure is the routine method for
isolating pBR322 derived plasmid DNAs and includes an
amplification step with chloramphenicol for higher yields of
DNA (Maniatis et al.. 1982a). Plasmids that are derived from
pUC8 DNA are also prepared as follows; however, due to their
high copy number the amplification step is unnecessary. The


83
in Figure 3-2, the cytoskeleton fraction clearly contained
actin as well as the major HeLa cytokeratins (Franke et al..
1981). In addition, vimentin appeared to be enriched in the
cytoskeleton, which was confirmed by immunoblot analysis.
Extraction of the cytoskeleton required a cold, calcium
containing buffer and consequently the tubulins were
solubilized and therefore absent in the Csk fraction. The
tubulin proteins were clearly contained in the soluble
fraction as determined by immunoblot analysis. Additional
support for separation of cytoskeleton and soluble phases
comes from the observation that the bulk of the tRNA species
is in the soluble phase which is in agreement with
previously reported values (data not shown; Cervera et al.
1983). Although the HLA-B7 messenger RNA is found on
membrane-bound polysomes and the histone mRNA on
nonmembrane-bound polysomes, both species of messenger RNAs
are associated almost exclusively with the cytoskeletal
fraction. Nearly 97% of the HLA-B7 mRNA and approximately
85-90% of the H3 and H4 histone messages are retained in the
Csk (Fig. 3-1? Table 3-1).
The Effects of Metabolic Inhibitors on the Subcellular
Localization of Histone and HLA-B7 mRNA
Detectable levels of histone mRNA were observed in each
of the subcellular samples, including the Sol fraction and
the membrane-bound polysomal fractions. However, the HLA-B7


Figure 6-3. Endogenous membrane-bound polysomal mRNAs are
released from the cvtoskeleton during cotreatment with CD
and puromycin in SPH3E1 and SPH3E1ATG' expressing cell
cultures.
Cytoskeleton and soluble RNAs (10 ;g/sample) were
assayed by Northern blot analysis for HLA-B7 mRNA and
chorionic gonadotropin mRNA content as described in Chapter
2. Lane 1, control-csk; lane 2, control-sol; lane 3, CD-csk?
lane 4, CD-sol? lane 5, puro-csk? lane 6, puro-sol; lane 7,
CD/puro-csk; and lane 8, CD/puro-sol.


22
the ribosome, the signal recognition complex (SRP) binds to
the signal peptide and arrests translation (Walter and
Blobel, 1981). The SRP is an 11S ribonucleic acid-protein
complex composed of six non-identical polypeptides and one
small cytoplasmic RNA molecule (Walter and Blobel, 1980;
Walter'and Blobel, 1982; Walter and Blobel, 1983). The RNA
molecule of SRP is 300 ribonucleotides in length and shares
approximately 80% homology with a middle repetitive human
DNA family (Ullu et al. 1982; Li et al^., 1982). With the
SRP bound to the signal peptide, the complex can then
associate with the membrane of the endoplasmic reticulum
through the interaction with the SRP receptor which is also
known as the docking protein (Gilmore et al.. 1982a; Gilmore
et al.. 1982b; Meyer et al.. 1982a; Meyer et al.. 1982b).
The SRP then dissociates from the polysome with the
resultant release of the translational block. As translation
resumes the protein is co-translocated from the cytoplasm
into the endoplasmic reticulum. During translocation, a
signal peptidase which is located in the lumen of the
endoplasmic reticulum removes the signal peptide from the
protein (Maurer and Mckean, 1978; Walter et al.. 1984).
Overview of Project
The overall aim of this work is to better define the
mechanism(s) involved in coupling histone protein synthesis
with DNA replication. Specifically, we have focused our


77
DNA replication (Spaulding et al.. 1966; Robbins and Borun,
1967; Borun et al.. 1967; Baumbach et al.. 1984). To address
the relationship between the localization of histone mRNAs
and the control of cellular histone mRNA levels, we have
used cloned human histone genes to examine the distribution
of histone mRNAs on nonmembrane-bound and membrane-bound
polysomes and the association of histone mRNA-containing
polysomes with the cytoskeleton. As predicted from previous
results of in vitro translation analysis and cDNA
hybridization studies, histone mRNAs were found to reside
predominantly on nonmembrane-bound polysomes. Both the
nonmembrane-bound polysomes containing histone mRNAs and the
membrane-bound polysomes containing the cell surface class I
HLA-B7 antigen mRNAs were found associated with the
cytoskeleton.
RESULTS
Localization of Histone mRNA and HLA-B7 mRNA in Subcellular
Fractions
To determine the intracellular locations of histone
mRNA and HLA-B7 mRNA, HeLa cells were first fractionated
into nonmembrane-bound and membrane-bound polysomal
components by the procedure outlined in Fig. 2-2. The
membrane-bound polysomes are represented by the mitochondria
associated rough endoplasmic reticulum (MRER) and the
nucleus associated rough endoplasmic reticulum (NRER)


81


38
directed mutagenesis and dideoxy DNA sequencing reactions
since it can be readily isolated as a double-stranded
covalently closed circular DNA replicative form (RF) from
the infected bacteria and as a circular, single-stranded DNA
genome (SS) from the mature phage particle. In addition, M13
phage DNA has been altered genetically to facilitate cloning
and subsequent manipulations. One such derivative, M13mpl8,
is small in size, maintained in high copy number
(approximately 200 RF molecules per cell), contains a
convenient polylinker for the insertion of foreign DNA (up
to 25 kb), and expresses beta-galactosidase which is
inactivated by the insertion of DNA into the polylinker
site.
Single-stranded templates. Single-stranded M13 DNA
templates were prepared from individual "plaques" which
formed on a lawn of E_j_ coli strain JM103 bacteria
(delta(lac-proAB) supE / F'[traD36, proA*. proB+, laclq,
lacZdeltaM15]) that were infected with M13 bacteriophage or
transfected with M13 double stranded, covalently closed
circular DNA. A single plaque was removed from the agar
plate with a sterile pasteur pipette and placed in 2 ml YT
medium (0.8% Bacto-tryptone, 0.5% Bacto-yeast extract, 0.5%
NaCl) Two drops of an overnight JM103 bacterial culture
were added to the phage and incubated at 37C with agitation
for 6 hours. Longer incubation times result in deletions


218
Wu, R., and Bonner, W. (1985) Mol. Cell. Biol. 5, 2959.
Yoshizaki, K., Nakagawa, T., Kaieda, T.f Muraguchi, A.,
Yamamura, Y., and Kishimoto, T. (1982) J. Immunology 128.
1296.
Zambetti, G., Stein, J., and Stein, G. (1986) Proc. Natl.
Acad. Sci. U.S.A. 84/ 2683.
Zambetti, G., Schmidt, W., Stein, G., and Stein, J. (1985)
J. CeU. Phys. 125, 345.
Zoller, M., and Smith, M. (1983) Methods Enzymol. 100. 468.
Zumbe, A., Stahli, C., and Trachsel, H. (1982) Proc. Natl.
Acad. Sci. U.S.A. 79, 2927.


11
whereas vimentin mRNA is concentrated in perinuclear regions
in chick myoblast and fibroblast cells (Lawrence and Singer,
1986). Furthermore, actin mRNA is concentrated in the
lamellipodia, structures involved in cell locomotion
(Lawrence and Singer, 1986). The requirement for actin
protein in the lamellipodia during cell movement (Wang,
1984) suggests a functional relationship between where the
mRNA is localized and where the encoded protein is needed
(Lawrence and Singer, 1986).
Localization of mRNAs in distinct areas of the
cytoplasm may occur due to an association of the mRNA with
an underlying structure. This structure could serve as an
anchor to concentrate mRNAs in particular regions of the
cytoplasm. Consistent with this reasoning is the existence
of a complex, proteinaceous structure in the cytoplasm of
most eukaryotic cells that is referred to as the
cytoskeleton. The cytoskeleton plays a role in cell shape,
cell motility and the intracellular transport of
macromolecules and organelles (Clarke and Spudich, 1977;
Goldman et al.. 1978; and reviewed in Nielsen et al.. 1983).
In addition, there is growing support for the hypothesis
that the cytoskeleton plays a key role in regulating the
translation process (Lenk et al.. 1977; Lenk and Penman,
1979; Fulton et al.. 1980; Cervera et al.. 1981; van Verooij
et al. 1981; Nielsen et al.. 1983; Howe and Hershey, 1984).


191
Regulation of c-fos gene expression is controlled at
both the transcriptional and posttranscriptional levels. A
variety of agents such as serum, growth factors and
differentiation factors transcriptionally induce c-fos gene
expression in a rapid and transient manner (Greenberg and
Ziff, 984). In contrast, inhibition of protein synthesis
with cycloheximide or heat shock results in a slower but
more persistent accumulation of c-fos mRNA with minimal or
no effect on c-fos transcription (Andrews et al.. 1987;
Rahmsdorf et al.. 1987). The increase in c-fos mRNA levels
during inhibition of protein synthesis appears to be largely
posttranscriptionally regulated through the stabilization of
the c-fos mRNA. Previous results have demonstrated that
cytochalasin D can inhibit protein synthesis in a dose
dependent manner by releasing poly-A+ RNA from the
cytoskeleton into the soluble phase as mRNP particles
(Ornelles et al.. 1986). The concentration of cytochalasin D
used in this study was sufficient to inhibit protein
synthesis by 50 percent. However, our results suggest that
the induction of c-fos gene expression by cytochalasin D
treatment is not a result of the partial inhibition of
protein synthesis. Cytochalasin D and puromycin cotreatment
resulted in roughly an additive increase in c-fos gene
transcription, suggesting the effects of the two drugs on
transcription are due to different mechanisms. Consistent
with this reasoning, the cellular level of c-fos mRNA in


21
endoplasmic reticulum and on what is generally known for the
synthesis of proteins that are localized on the cell surface
or secreted from the cell. Most cell surface or secreted
proteins are initially synthesized with a segment of amino
acid residues at the N-terminus that is not usually found in
the mature protein. This segment is commonly referred to as
a signal peptide. Although signal peptides are not highly
conserved in sequence, they are fairly conserved in
structure and based on the compilation of 277 signal
peptides ranging from bacteria to higher eukaryotes, a
consensus structure has been deduced (Watson, 1984) In
general, signal peptides are 25-30 amino acids in length.
They contain a charged amino acid within the first five
residues, which is followed by a stretch of at least nine
hydrophobic amino acids, and terminate at a small uncharged
amino acid such as alanine, serine or glycine. It is
believed that the hydrophobic domain of the signal peptide
facilitates the transfer of the protein across the lipid
environment of specific intracellular membranes whereas the
small, uncharged amino acid is thought to direct the
protease to cleave the peptide from the mature protein
(Hortin and Boime, 1981).
Presumably, translation of mRNA encoding exported
proteins is initiated on nonmembrane-bound polysomes.
Following the synthesis of approximately 70 amino acids, the
point at which the entire signal peptide has emerged from


70
Annealing Template and Primer
The primer for synthesizing the complementary strand of
wild type and mutated M13/SPH3 recombinant DNA is a 20 bp
oligonucleotide which hybridizes to a site adjacent to the
polylinker. The annealing reaction was prepared by adding 1
ul primer (0.5 pmol), 7 ul single-stranded M13/SPH3
recombinant DNA (0.5-1.0 pmol) and 2 ul 5X sequencing buffer
(200 mM Tris-HCl pH 7.5, 100 mM MgCl2, 250 mM NaCl). The
sample was heated at 65C for 2 minutes and slowly cooled to
room temperature.
Labeling Reaction
The complementary strand was synthesized by incubating
the annealed sample with 1 ul 100 mM dithiothreitol, 2 ul
labeling nucleotide mix (1.5 uM dGTP, 1.5 uM dCTP, 1.5 uM
dTTP) 0.5 ul [a-35S] dATP [10 uCi/ul] and 2 ul Sequenase
[1.5 U/ul] at room temperature for 5 minutes. Elongation was
terminated by mixing 3.5 ul aliquots of the labeling
reaction with 2.5 ul of one of the dideoxy nucleotide
solution (ddATP, ddCTP, ddGTP, ddTTP). The dideoxy
nucleotide solutions were 80 uM dATP, 80 uM dCTP, 80 uM
dGTP, 80 uM dTTP, 50 mM NaCl and 8 uM of the appropriate
dideoxyribonucleotide. The sequencing samples were diluted
with 4 ul stop solution (95% formamide, 20 mM EDTA, 0.05%
bromophenol blue and 0.05% xylene cyanol), heated at 75C
for 2 minutes and loaded onto a 6% acrylamide-8.3 M urea gel


Figure 4-5. Schematic diagram of wild type and mutated
signal peptide-histone fusion mRNAs.
The signal peptide-histone fusion gene (SPH3E1) was
mutated by site directed mutagenesis (Zoller and Smith,
1983) as described in Chapter 2. Abbreviations: H3,
endogenous H3 histone mRNA; SPH3E1, signal peptide-histone
fusion mRNA; SPH3Elalpha, signal peptide-histone fusion mRNA
mutated in the hydrophobic region of the signal peptide (leu
at position 10 and pro at position 12 substituted with his)?
SPH3E1ATG signal peptide-histone fusion mRNA mutated in
the signal peptide translation start codon; CAP, 5' cap
structure; ATG, predicted translation start codon; atg,
internal translation start codon for histone H3 protein;
TAA, translation stop codon.


REFERENCES
Adams, A., Fey, E. G., Pike, S., Taylorson, C., White, H.,
and Rabin, B. (1983) Biochem. J. 216. 215.
Adesnik, M. and Darnell, J. (1972) J. Mol. Biol. 67., 397.
Allfrey, V., Littau, V., and Mirsky, A. (1963) Proc. Natl.
Acad. Sci. U.S.A. 49., 414.
Alterman, R., Sprecher, C., Graves, R., Marzluff, W., and
Skoultchi, A. (1985) Mol. Cell. Biol. 5, 2316.
Amos, L. and Klug, A. (1974) J. Cell Sci. 14., 523.
Andrews, G., Harding, M., Calvert, J., and Adamson, E.
(1987) Mol. Cell. Biol. 7, 117.
Bagchi, T., Larson, D., and Sells, B. (1987) Exp. Cell Res.
168. 160.
Baumbach, L., Stein, G., and Stein, J. (1987) Biochemistry
26, 6178.
Baumbach, L., Marashi, F., Plumb, M., Stein, G., and Stein,
J. (1984) Biochemistry 23., 1618.
Bedouelle, H., Bassford Jr., P., Fowler, A., Zabin, I.,
Beckwith, J., and Hofnung, M. (1980) Nature 285, 78.
Benecke, B.-J., Ben Ze'ev, A., and Penman, S. (1978) Cell
14, 931.
Bennet, G., Fellini, S., and Holtzer, H. (1978)
Differentiation 12, 71.
Berk, A., and Sharp, P. (1978) Proc. Natl. Acad. Sci. U.S.A.
75, 1274.
Berl, S., Puszkin, S., and Nicklas, W. (1973) Science 179,
441.
Binder, L., Frankfurter, A., and Rebhun, L. (1985) J. Cell
Biol. 101, 1371.
206


145
123456789 10


CHAPTER 1
INTRODUCTION
General Background
Histones are small basic proteins that are involved in
the packaging of eukaryotic DNA into chromatin. It is
estimated that more than 2 meters of DNA are contained
within the limited area of the nucleus. The condensation of
the DNA to such a great extent is due primarily to its
interactions with the histone proteins. In addition to their
structural properties histones play a functional role in
controlling gene transcription. The way in which the histone
proteins are arranged along a region of DNA appears to
affect the transcriptional properties of that DNA domain.
The complex nature of the regulation of histone protein
synthesis at the transcriptional, posttranscriptional, and
posttranslational levels makes these genes an important
model for studying eukaryotic gene expression. However, one
of the most important reasons for studying the regulation of
histone gene expression is that the synthesis of the
majority of histone proteins (cell cycle dependent histones)
occurs during DNA replication. Elucidation of the mechanisms
coupling histone protein synthesis with DNA synthesis may
1


80
60
40
20
0
125
20
T Im e(HU)
40
60


It
t:
HLA-B7
(
CSK
f

<
SOL
CSK
SOL
CSK
SOL
CSK
SOL
kO
rsj


I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as a dissertation for the
degree of Doctor of Philosophy. /
fanet L. Stein, Chair
'Professor of Immunology and
Medical Microbiology
I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as/^)diss9ir^ation for the
degree of Doctor of Philosophy.
J3cfry 5, S^in
"^Professor of Biochemistry
and Molecular Biology
I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as a dissertation for the
degree of Doctor of Philosophy.
I ^2
Jaimes B. Flanegan
professor of Immundlogy and
Medical Microbiology
I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as a dissertation for the
degree of Doctor of Philosophy.
Thomas C. Rowe
Assistant Professor of
Pharmacology and Therapeutics
I certify that I have read this study and that in my opinion
it conforms to acceptable standards of scholarly presentation and
is fully adequate, in scope and quality, as a dissertation for the
degree of Doctor of Philosophy.
Richard W. Moyer/
Professor of Immunology and
Medical Microbiology


195
expression during EGF and insulin cotreatment can be
inhibited by disruption of the cytoskeleton with
cytochalasin D. This result is consistent with the
hypothesis that the cytoskeleton may be involved in relaying
the EGF/insulin signal from the cell surface to the nucleus.
The results presented here describing the stimulation of c-
fos gene expression with cytochalasin D, although contrary
to Rebillard's results, are consistent with their
interpretation that the nucleus monitors the integrity of
the cytoskeleton and adjusts the expression of particular
genes accordingly. Rather than cytochalasin D inhibiting a
signal from the cell surface to the nucleus, the drug in our
hands may mimic the signal to stimulate c-fos gene
expression. Cell type, culture conditions and drug dosage
effects may explain the differences observed between our
results and those of Rebillard's et al. (1987) .
In summary, cytochalasin D stimulates c-fos gene
expression through increased transcription and mRNA
stability. The effect of cytochalasin D on c-fos gene
expression may be in response to the alterations in the
cytoskeleton and cell shape. Whether other elements of the
cytoskeleton such as the intermediate filaments can elicit
the same response remains to be determined. Taken together
with previous reports, the ability of cytochalasin D to
stimulate c-fos in quiescent fibroblasts with the same
kinetics and virtually to the same extent as serum supports
the role of the cytoskeleton in signal transduction.


203
the cytoskeleton. These results suggest that one mechanism
for the attachment of membrane-bound polysomes to the
cytoskeleton is through the interaction of the polysome
complex with the endoplasmic reticulum. This interaction
appears to involve the nascent polypeptide; however other
possibilities may exist.
The relocation of signal peptide-histone fusion mRNA to
membrane-bound polysomes presented the opportunity to study
the cytoskeletal interactions of membrane-bound polysomal
mRNAs in more detail. Our model predicted that histone
fusion mRNA would remain attached to the cytoskeleton during
CD treatment and would be released into the soluble phase
during CD and puromycin cotreatment. As seen in Chapter 6,
the signal peptide-histone fusion mRNA remained attached to
the cytoskeleton during cytochalasin D treatment as well as
during cytochalasin D and puromycin cotreatment. The
retention of signal peptide-histone fusion mRNA on the
cytoskeleton under these conditions demonstrates that the
histone fusion mRNA is associated with the cytoskeleton in a
manner that is different from that observed for wild type
histone mRNA, HLA-B7 membrane-bound polysomal RNA and
heterogeneous poly-A+ RNA. Although it is well documented
that mRNAs are associated with the cytoskeleton, there are
no known cytoskeleton attachment sites identified to date.
Dissection of the signal peptide-histone fusion mRNA, by
deletion and site directed mutation analysis, should prove


141
Table 5-2. Percent distribution of nonmembrane-bound and
membrane-bound polysomal RNAs in the Csk and Sol fractions
from CD. Puro, CD/puromvcin and CD/cvcloheximide treated
cells.
H3/H4
HLA-
B7
hCG
fos
Control
CSK
87%
(5)
97%
(i)
96%
(i)
80%
SOL
13%
3%
4%
20%
CD
CSK
18%
(4)
75%
(6)
72%
(6)
29%
SOL
82%
25%
28%
71%
Puro
CSK
79%
(7)
79%
(9)
85%
(5)
83%
SOL
21%
21%
15%
17%
CD/Puro
CSK
18%
(4)
25%
(3)
26%
(3)
20%
SOL
82%
75%
74%
80%
CD/Cyclo
CSK
19%
(8)
81%
(+2)
52%
(10)
40%
SOL
81%
19%
48%
60%
The
Csk
and Sol
distribution of
c-fos
mRNA during
CD,
puro, CD/puro and CD/cyclo treatment was quantitated by
normalizing the densitometric results from the autoradiogram
presented in Figure 5-2 for the yield of RNA in each
fraction. The Csk and Sol distribution of nonmembrane-bound
and membrane-bound polysomal mRNAs was quantitated directly
from the densitometric analysis. The numbers in brackets
indicate standard error.


65
for 45 minutes. The reaction was terminated by heating at
65C for 10 minutes, and unincorporated radiolabeled
nucleotides were removed by chromatography through a
Sephadex G-25 (Pharmacia) column (10 mm x 100 mm) in 50 mM
ammonium bicarbonate (pH 7.8).
The phosphorylated oligonucleotide (10-30 molar excess
over template DNA) was added to 0.5 pmol of M13 recombinant
DNA and the volume was adjusted to 10 ul with ddH20 and 1 ul
solution A (200 mM Tris-HCl pH 7.5, 100 mM MgCl2, 500 mM
NaCl, 10 mM dithiothreitol). The solution was heated at 55C
for 5 minutes and then cooled for 5 minutes at room
temperature. The primer extension reaction also contained 1
ul of 2.5 mM dCTP, 1 ul 2.5 mM dATP, 1 ul 2.5 mM dTTP, 1 ul
2.5 mM dGTP, 0.5 ul solution A and 1 unit Klenow enzyme. The
reaction proceeded at room temperature for 5 minutes and was
terminated by heating at 65C for 10 minutes. The sample was
diluted with 2 ul lOx PstI buffer (100 mM Tris-HCl pH 7.5, 1
M NaCl, 100 mM MgCl2, 1 mg/ml BSA) and 7 ul ddH20 and then
digested with 1 ul PstI restriction endonuclease [5 U/ul] at
37C for 1 hour. The sample was heated at 100C for 2
minutes, quickly cooled to 0C and diluted with 20 ul SI
loading buffer. The PstI digested samples (10 ul) were
resolved electrophoretically in 5% acrylamide (20:1;
acrylamide:bis), 7 M urea at 300 volts (-35 mA) for 1.5
hours. The gels were dried under vacuum at 80C for 1 hour
and analyzed by autoradiography. Primer extension analysis


136
Table 5-1. The percent of cytoskeleton and soluble phase
associated mRNAs in cells treated with cvtochalasin D.
H3
H4
HLA-B7
hCG
Control
Csk
93%
96%
>98%
90%
Sol
7%
4%
< 2%
10%
CD [5 Mg/ml]
Csk
24%
20%
79%
69%
Sol
76%
80%
21%
31%
CD [10 Mg/ml]
Csk
33%
31%
87%
77%
Sol
67%
69%
13%
23%
CD [40 jitg/ml]
Csk
26%
20%
64%
41%
Sol
74%
80%
36%
59%
The densitometric
results
from Figure
5.1 (hCG data not
shown) were normalized
for the
yield of RNA
. from the
cytoskeleton
and soluble fractions of each
cell sample
(note: equal quantities of cytoskeleton and soluble RNAs
were analyzed which does not take into consideration the
unequal distribution of RNA within these fractions or the
changes that occur during cytochalasin D treatment). The
values are presented as percent distribution of mRNA between
the cytoskeleton and soluble fractions.


161
MW 1 2 3 4 5 6 7 8
SPH3E1
excess probe
Endogenous H3


98
coding sequences and extends approximately 300 base pairs
beyond the TAA translation stop codon, including a region
which has been implicated in histone mRNA destabilization
when DNA synthesis is inhibited (Luscher et al.. 1985?
Pandey and Marzluff, 1987? Graves et al.. 1987). An SV40
enhancer element was incorporated into the upstream EcoRI
site in the recombinant plasmid to increase cellular levels
of the fusion transcript. The mRNA encoded by the fusion
gene consists of the first 20 nucleotides of the H3 histone
5' leader sequences followed by the untranslated leader
sequence of the beta-lactamase signal peptide, the entire
beta-lactamase signal peptide coding sequences (including
the translation start codon), the entire H3 histone coding
region which is fused in frame to the signal peptide coding
region, and the H3 histone 3' untranslated region (Figure 4-
1). The junction between the signal peptide coding region
and the H3 histone coding region has been sequenced by
Sanger's dideoxy method and the reading frame has been
conserved. All sequences required for the synthesis and
processing of the chimeric mRNA, as well as for translation
of the fusion protein, are present. The signal peptide-H3
histone fusion gene with a single SV40 enhancer element is
designated pSPH3El and an identical fusion construct
containing multiple SV40 enhancer elements in the EcoRI site
is designated pSPH3E2.


151
The association of mRNA with the cytoskeleton also
appears to change during differentiation in certain cell
types (Toh et al.. 1980). Microscopic analysis of fetal rat
lung fibroblast cells indicates that polyribosomes are
aligned in linear arrays along the actin-like stress fibers
(Toh e al. 1980). These polysomes are difficult to
visualize after disruption of the microfilaments by
cytochalasin B treatment and are presumed to be released
from the cytoskeleton into the soluble phase. Polysomes in
the more differentiated fetal lung fibroblasts are
associated with the endoplasmic reticulum and are no longer
affected by cytochalasin B treatment (Toh et al.. 1980).
These results are consistent with the possibility that the
association of polysomes with the cytoskeleton changes
during differentiation.
In conclusion, the results presented here provide
evidence for heterogeneity in the interactions of mRNA with
the cytoskeleton in HeLa cells. Membrane-bound and non-
membrane-bound polysomes are differentially associated with
the cytoskeleton. Studies on the association of mRNA with
the cytoskeleton must take into consideration these
differences when interpreting which elements are involved in
the cytoskeletal attachment of mRNA. Failure to release mRNA
from the cytoskeleton with cytochalasins may only suggest
that additional cytoskeletal attachment sites are involved.
The differential association of mRNA with the cytoskeleton


147
A+ RNA with the cytoskeleton could be disrupted by
perturbing the microfilaments with cytochalasin D in HeLa
cells (Ornelles et al.. 1986). Consistent with these
results, cytochalasin D treatment readily releases c-fos
mRNA and poly-A' histone mRNA from the cytoskeleton into the
soluble phase. Surprisingly, HLA-B7 mRNA and hCGa mRNA were
only partially released from the cytoskeleton in
cytochalasin D treated HeLa cells.
The inefficiency of cytochalasin D in releasing HLA-B7
and hCGa mRNA from the cytoskeleton appears to be a result
of the association of these mRNAs with membrane-bound
polysomes. Dissociation of the polysomes with puromycin in
the presence of cytochalasin D releases HLA-B7 and hCGa mRNA
from the cytoskeleton. In contrast, cycloheximide, an
inhibitor of protein synthesis that preserves polysomal
structure, fails to release HLA-B7 mRNA from the
cytoskeleton in cytochalasin D treated cells. These results
suggest that membrane-bound polysomal mRNAs are attached to
the cytoskeleton in at least two different ways. The first
cytoskeleton attachment site is through the mRNA/mRNP itself
and is sensitive to cytochalasin D treatment. This site is
most likely common to all cytoskeletal associated mRNAs in
HeLa cells and has been previously observed for poly A+ RNA
by Ornelles et al. (1986). The second cytoskeletal
attachment site which is characteristic of membrane-bound
polysomes appears to be mediated by the nascent polypeptide


87
fractions (Figure 3-3). This is in sharp contrast to the
HLA-B7 mRNA which appeared to be fully stable under these
conditions. Cycloheximide treatment had a stabilizing effect
on the levels of histone mRNA in each subcellular fraction
(Figure 3-3). When compared with control HeLa cells,
cycloheximide treatment resulted in a 2.1 fold (H4) and a
2.4 fold (H3) increase in the levels of histone mRNA in the
nonmembrane-bound polysomal fraction. Similar increases were
also seen for levels of histone mRNAs in the LER and
membrane-bound polysomal fractions. Inhibition of protein
synthesis with cycloheximide followed by DNA synthesis
inhibition with hydroxyurea resulted in the stabilization of
histone mRNA in each subcellular fraction, whereas the
stability of HLA-B7 mRNA remained unaffected (Figure 3-3).
Similar results were obtained when these inhibitor studies
were expanded to include the cytoskeleton and soluble
fractions (Figure 3-4). Cycloheximide treatment resulted in
a 2-3 fold increase in the representation of H3 and H4
histone mRNA over untreated cultures in both the
cytoskeleton and soluble fractions. HLA-B7 mRNA appeared to
be only moderately stabilized in the presence of this drug.
When cells were treated with hydroxyurea, histone mRNA
levels were greatly reduced in both the cytoskeleton and
soluble fractions whereas the HLA-B7 mRNA was found in
comparable levels to those present in the untreated samples.
Consistent with previous data, cotreatment of HeLa cells


29
HeLa Suspension Cell Cultures
HeLa S3 cells were grown and maintained in suspension
at 3-6xl05 cells/ml in SMEM (Joklik-modified minimum
essential medium) supplemented with 7% calf serum, 100 U/ml
penicillin, 100 /xg/ml streptomycin and 1 mM glutamine in a
warm room at 37C. Monoclonal and polyclonal HeLa S3 cell
lines, which were isolated during G418 selection of
transfected monolayer cultures, were maintained in SMEM
containing 5% fetal calf serum, 5% horse serum, 100 U/ml
penicillin, 100 /xg/ml streptomycin and 1 mM glutamine at 3-
6x10s cells/ml.
WI-38 Monolayer Cell Cultures
WI-38 cells (normal, human diploid fibroblast-like
cells; ATCC ccl 75) at passage 28, were grown as monolayers
in EMEM containing 10% fetal calf serum, 100 U/ml
penicillin, 100 /ng/ml streptomycin and 1 mM glutamine at
37C under 5% C02. Cell cultures were made quiescent by
maintaining cultures in EMEM containing 0.5% fetal calf
serum, 100 U/ml penicillin, 100 /xg/m1 streptomycin and 1 mM
glutamine at 37C under 5% C02 for 48 hours.
Metabolic Inhibitor Treatments
Inhibition of DNA synthesis was usually performed with
1-5 mM hydroxyurea for 30 to 60 minutes. The drug was always
prepared fresh in IX spinner salts (Gibco) as a 100 mM stock
solution.


131
We present evidence here to support a model for
multiple sites of attachment of eukaryotic mRNA to the
cytoskeleton. Using cytochalasin D for the disruption of
microfilaments, we have observed a differential release of
specific mRNAs from the cytoskeleton. Further
characterization of the association of mRNA with the
cytoskeleton revealed an additional attachment site that is
insensitive to cytochalasin D treatment and present only in
membrane-bound polysomal mRNAs. This second attachment site
is puromycin sensitive and appears to involve the
association of the nascent polypeptide and/or ribosome with
the remnant protein structure of the endoplasmic reticulum.
Results
Cytochalasin D Releases Polv A~ RNA from the Cytoskeleton
Disruption of actin fibers by cytochalasin D treatment
results in the release of cytoplasmic poly A+ RNA from the
cytoskeleton in a dose dependent manner (Ornelles et al..
1986). To further understand the association of eukaryotic
mRNAs with the cytoskeleton, we have studied the effects of
cytochalasin D treatment on the release of specific mRNAs,
including those that are non-polyadenylated. In addition we
have examined and compared the cytoskeletal association of
nonmembrane-bound and membrane-bound polysomal mRNAs.
Cell cycle dependent histone genes code for mRNAs that
are capped, non-polyadenylated and predominantly associated


215
Schatz, G. and Butow, R. (1983) Cell 22, 316.
Scherrer, K., and Darnell, J. (1962) Biochem. Biophys. Res.
Comm. 2/ 486.
Schliwa, M. (1982) J. Cell Biol. 92, 79.
Schmidt, W., Pardue, R., Tutt, M., Briggs, R., Brinkley, B.,
and Hnilica, L. (1982) Proc. Natl. Acad. Sci. U.S.A. T9,
3138.
Selden, L., Gershman, L., and Estes, J. (1980) Bioch.
Biophys. Res. Commun. 95, 1854.
Shiels, A., Marashi, F., Stein, J., and Stein, G. (1985)
Biochem. Biophys. Res. Comm. 127, 239.
Sierra, F., Lichtler, A., Marashi, F., Rickies, R., Van
Dyke, T., Clask, S., Wells, J., Stein, J., and Stein, G.
(1982) Proc. Natl. Acad. Sci. U.S.A. T9, 1795.
Simon, K., Perara, E., and Lingappa, V. (1987) J. Cell Biol.
104. 1165.
Sittman, D., Graves, R., and Marzluff, W. (1983) Proc. Natl.
Acad. Sci. U.S.A. 80, 1849.
Sive, H., Heintz, N., and Roeder, R. (1984) Mol. Cell. Biol.
4, 2723.
Sloboda, R., Dentler, W., and Rosenbaum, J. (1976)
Biochemistry .15, 4497.
Sood, A., Pereira, D., and Weissman, S. (1981) Proc. Natl.
Acad. Sci. U.S.A. 78, 616.
Southern, P., and Berg, P. (1982) J. Mol. Appl. Gen. 1, 327.
Spaulding, J., Kajiwara, K*, and Mueller, G. (1966) Proc.
Natl. Acad. Sci. U.S.A. 56, 1535.
Stahl, H., and Gallwitz, D. (1977) Eur. J. Biochem. 72, 385.
Stauber, C., Luscher, B., Eckner, R., Lotscher, E., and
Shumperli, D. (1986) EMBO J. 5, 3297.
Stein, G., and Borun, T. (1972) J. Cell Biol. 52, 292.
Stein, G., and Stein, J. (1984) Mol. Cell. Biochem. 64, 105*
Stein, J., Stein, G., and McGuire, P. (1977) Biochemistry
16, 2207.


Figure 4-9. Effects of hydroxyurea treatment on SPH3Elalpha
mRNA.
HeLa cell monolayers were transfected with pSPH3Elalpha
DNA and 46 hours post-transfection were treated with 1 mM
hydroxyurea (HU). Cells were harvested at 20 minute
intervals and total cellular RNA was prepared. The RNA (10
/Ltg) was subjected to SI nuclease protection analysis to
quantitate the levels of the SPH3Elalpha mRNA and endogenous
H3 histone fusion mRNA. Each lane represents an individually
transfected cell culture. MW is radiolabeled Hpa II digest
of pBR322 DNA. Lanes 1 and 2, control; lanes 3 and 4, 1 mM
HU 20 minutes; lanes 5 and 6, 1 mM Hu 40 minutes; lanes 7
and 8, 1 mM HU 60 minutes.


52
lxlO6 cpm/ml of each thermally denatured, j2P-radiolabeled
DNA probe. After hybridization, the filters were washed with
5x SSC, lx Denhardt's (0.5-1 ml/cm2 filter) for 10-15
minutes at room temperature to remove unbound probe. The
filters were then washed, in order, with the following
solutions (0.5 ml/cm2 filter area): 1) 5x SSC, lx
Denhardt's; 2) 2x SSC, 0.1% SDS; 3) lx SSC, 0.1% SDS; and 4)
O.lx SSC, 0.1% SDS at 60-65C for 30 minutes per wash.
After washing the filters were briefly air dried and then
exposed to preflashed Kodak XAR5 x-ray film at -70C with a
Cronex Lightning Plus intensifying screen (Dupont) for
varying lengths of time. Hybridization was quantitated by
scanning laser densitometric analysis of multiple exposures
of autoradiographs that were within the linear range of the
film.
SI Nuclease Protection Analysis
SI nuclease protection analysis was performed
essentially as described by Berk and Sharp (1977) and was
used in the work presented here to detect simultaneously
both endogenous H3 histone mRNA and signal peptide-histone
fusion mRNA in the same RNA sample. The SI probe was
prepared by digesting 30 ng pSPH3El, pSPH3Elalpha, or
pSPH3ElATG DNA with 3 0 units of Smal at 3 0C for 3 hours.
The 450 bp Smal fragment was isolated from a 0.8% agarose
minigel by the freeze-squeeze method. The solution


Figure 7-2. Densitometric anlavsis of steady state levels of
c-fos mRNA during cytochalasin D and puromvcin treatments.
Steady state levels of c-fos mRNA during cytochalasin D
and puromycin treatments were determined by densitometric
analysis of the Northern filters. The values presented
represent the average of the minimal fold increase in c-fos
mRNA levels during the drug treatments as compared to the
control sample. Brackets indicate standard error from three
independent experiments.


205
In summary, the work presented in this dissertation is
consistent with the hypothesis that subcellular compartments
and cell structure play an important role in the regulation
of gene expression. The influence of mRNA subcellular
location on the posttranscriptional control of histone gene
expression and possibly c-fos gene expression, indicates
that regulatory factors may be concentrated and sequestered
in specific regions or by specific structures. The
differential association of mRNA with the cytoskeleton may
be a mechanism for localization of mRNA in a particular
region of the cell, which in turn may affect the
posttranscriptional regulatory process. Lastly, the effect
of disrupting the microfilaments with cytochalasin D on c-
fos gene and beta-actin gene expression suggests that the
nucleus can transcriptionally respond in a selective manner
to the structural organization of the cytoskeleton.


THE ROLE OF SUBCELLULAR COMPARTMENTS AND STRUCTURES
IN EUKARYOTIC GENE EXPRESSION
By
GERARD PAUL ZAMBETTI
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1989


159
RNA in this sample compared with the control cytoskeleton
sample (Figure 6-1, lane 1 compared to lane 7). The
inability of CD and puromycin to release SPH3E1 chimeric
mRNA efficiently from the cytoskeleton is not unique to the
monoclonal cell line and is also observed in a polyclonal
cell population (Figure 6-2, Table 6-1). In this case, 62%
of the endogenous H3 histone mRNA and less than 5% of the
SPH3E1 mRNA is released from the cytoskeleton during
cytochalasin D treatment. In CD and puromycin co-treated
polyclonal cell cultures approximately 27% of the endogenous
histone mRNA and 62% of the SPH3E1 mRNA remain associated
with the cytoskeleton.
To investigate whether the failure to release the
signal peptide-histone mRNA from the cytoskeleton with
cytochalasin D and puromycin was related to the efficiency
of the drug treatments, we studied the association of
endogenous membrane-bound polysomal mRNAs with the
cytoskeleton. The mRNAs coding for HLA-B7, a class I cell
surface histocompatibility antigen and chorionic
gonadotropin alpha (hCGa), a secreted protein, are
translated on membrane-bound polysomes and well represented
in HeLa cells. The distribution of HLA-B7 and hCGa mRNA in
the cytoskeleton and soluble samples used in Figure 6-1 was
determined by Northern blot analysis (Figure 6-3 and Table
6-1). As expected for membrane-bound polysomal mRNAs, HLA-B7
and hCGa mRNAs were not released by cytochalasin D treatment


LIST OF TABLES
Table Page
3-1 Quantity of histone and HLA-B7 mRNAs in the
subcellular fractions 82
4-1 Quntitation of SPH3E1, SPH3Elalpha and SPH3E1ATG"
mRNA in nonmembrane-bound and membrane-bound
polysome fractions 116
5-1 The percent of cytoskeleton and soluble phase
associated mRNAs in cells treated with
cytochalasin D 13 6
5-2 Percent distribution of nonmembrane-bound and
membrane bound polysomal RNAs in the Csk and Sol
fractions from CD, Puro, CD/puromycin and
CD/cycloheximide treated cells 141
6-1 The percent of cytoskeleton associated mRNAs
isolated from cytochalasin D, puromycin and
CD/puromycin treated cells 158
ix


209
Engel, J. Sugarman, B., and Dodgson, J. (1982) Nature 297,
434.
Fanner, S., Wan, K., Ben-Ze'ev, A., and Penman, S. (1983)
Mol. Cell. Biol. 3, 182.
Fiddes, J., and Goodman, H. (1979) Nature 281. 351.
Franke, W., Schmid, E., Osborn, M., and Weber, K. (1979) J.
Cell Biol. 81, 570.
Franke, W., Schiller, D., Moll, R., Winter, S., Schmid, E.,
Engelbrecht, I., Denk, H., Krepler, R., and Platzer, B.
(1981) J. Mol. Biol. 153/ 933.
Fuchs, E., and Marchuk, D. (1983) Proc. Natl. Acad. Sci.
U.S.A. 80, 4321.
Fulton, A., Wan, K., and Penman, S. (1980) Cell 20, 849.
Gallwitz, D. (1975) Nature 257, 247.
Gallwitz, D., and Mueller, G. (1969) J. Biol. Chem. 244.
5947,
Gallwitz, D., and Briendl, M. (1972) Biochem. Biophys. Res.
Commun. 42, 1106.
Garrels, J., and Gibson, W. (1976) Cell 9, 793.
Gilmore, R., Blobel, G., and Walter, P. (1982a) J. Cell
Biol. 95, 463.
Gilmore, R., Walter, P., and Blobel, G. (1982b) J. Cell
Biol. 95, 470.
Glass, W., Briggs, R., and Hnilica, L. (1981) Science 211.
70.
Goldman, R., and Knipe, D. (1973) Cold Spring Harb. Symp.
Quant. Biol. 37., 523.
Goldman, J., Schaumberg, H., and Norton, W. (1978) J. Cell
Biol. 78, 426.
Gorman, C., Moffat, L., and Howard, B. (1982) Mol. Cell.
Biol. 2, 1044.
Graham, F., and van der Eb, A. (1973) Virology 52, 456.
Graves, R., and Marzluff, W. (1984) Mol. Cell. Biol. 4, 351.


69
briefly dried under a heat lamp and placed against pre
flashed XAR5 x-ray film without a screen for 1 hour at room
temperature. Under such low stringency each phage spot was
positive, including the negative control M13/SPH3 wild type
bacteriophage. The filter was re-washed at 35C in 6x SSC
for 5 minutes, air dried and exposed to pre-flashed XAR5
film for 1 hour. At the higher wash temperature many of the
phage spots were less intense whereas several remained
unchanged. When the filter was washed at 48C in 6x SSC for
5 minutes the negative control and nearly all the phage
spots completely disappeared. The bacteriophage (-5%) that
remained positive after the 48C wash were sequenced to
confirm the presence of the mutations.
DNA Sequencing
The site directed mutations in M13/SPH3 recombinant
DNAs were confirmed by Sanger's dideoxynucleotide chain
terminator DNA sequencing procedure using the "Sequenase
DNA Sequencing Kit" from United States Biochemical
Corporation, Cleveland, Ohio. The Sequenase enzyme is
derived from bacteriophage T7 DNA polymerase and has been
modified to efficiently use dideoxyribonucleotides, alpha-
thio deoxyribonucleotides and other nucleotide analogs that
are commonly used for sequencing. In addition, the Sequenase
enzyme expresses high processivity and low 3'-5' exonuclease
activity.


201
The release of histone mRNA from the cytoskeleton by
cytochalasin D treatment presented the opportunity to study
the role of the cytoskeleton in the posttranscriptional
regulation of histone gene expression. Histone mRNA was
released into the soluble phase with moderate amounts of
cytochalasin D [10 ug/ml], since higher levels of CD
significantly inhibit protein synthesis. The cells were then
treated with hydroxyurea to inhibit DNA replication. These
treatments resulted in the destabilization of histone mRNA
in the soluble fraction to the same extent and with the same
kinetics as the destabilization of histone mRNA in the
cytoskeleton fraction. One interpretation of this result is
that the cytoskeleton is not required for histone mRNA
destabilization during inhibition of DNA synthesis. In
addition, Ornelles et al. (1986) reported that CD treatment
of HeLa cells releases poly-A+ RNA from the cytoskeleton
into the soluble phase in monosome or mRNP form, which
suggests that histone mRNA in the soluble phase does not
need to be translated to be destabilized during inhibition
of DNA synthesis. This result is in contrast to the genetic
studies that indicate histone mRNA must be translated to
within close proximity of the translation stop codon for
destabilization to occur (Graves et al., 1987? Capasso et
al.. 1987). The release of histone mRNA into the soluble
phase as a monosome or mRNP during cytochalasin D treatment
has not yet been determined and additional studies will be


210
Graves, R., Pandey, N., Chodchoy, N., and Marzluff, W.
(1987) Cell 48, 615.
Green, L., Stein, G., and Stein, J. (1984a) Mol. Cell.
Biochem. .60, 123.
Green, L., van Antwerpen, R., Stein, J., and Stein, G.
(1984b) Science 226, 838.
Greenberg, M., and Ziff, E. (1984) Nature 311. 433.
Greenberg, M., Hermanowski, A., and Ziff, E. (1986) Mol.
Cell. Biol. 6, 1050.
Grossi de Sa, M.-F., Martins de Sa, C., Harper, F., Link-
Coux, M., Huesca, M., and Scherrer, K. (1988) J. Cell Biol.
107. 1517.
Gunning, P., Ponte, P., Okayama, H., Engel, J., Blau, H.,
and Kedes, L. (1983) Mol. Cell. Biol. 3, 787.
Hartwig, J., and Stossel, T. (1975) J. Biol. Chem. 250.
5696.
Harvey, R., Whiting, J., Coles, L., Krieg, P., and Wells, J.
(1983) Proc. Natl. Acad. Sci. U.S.A. 80/ 2819.
Hayden, J., and Allen, R. (1984) J. Cell Biol. 99, 1785.
Heintz, N., Zernik, M. and Roeder, R. (1981) Cell 24., 661.
Heintz, N., Sive, H., and Roeder, R. (1983) Mol. Cell. Biol.
3, 539.
Helms, S., Baumbach, L., Stein, G., and Stein, J. (1984)
FEBS Letters 168/ 65.
Hentschel, C., and Birnstiel, M. (1981) Cell 25, 301.
Hoffman, P., and Lasek, R. (1975) J. Cell Biol. 66, 351.
Holltrop, M., Raisz, L., and Simmons, H. (1974) J. Cell
Biol. 60, 346.
Hortin, G., and Boime, I. (1981) Cell 24, 453.
Howe, J., and Hershey, J. (1984) Cell 37, 85.
Hubbard, B., and Lazarides, E. (1979) J. Cell Biol. 80, 166.
Huberman, E., and Callahan, M. (1979) Proc. Natl. Acad. Sci.
USA 76, 1293.


LIST OF FIGURES
Figure Page
2-1 Outline of the cloning scheme for the construction of
the signal peptide-histone fusion gene 34
2-2 Outline of procedure for the isolation of subcellular
polysomal fractions 45
3-1 Northern blot analysis of H3 histone, H4 histone and
HLA-B7 mRNA in subcellular fractions 81
3-2 Two-dimensional gel electrophoresis of cytoskeleton
and soluble associated proteins 85
3-3 Northern blot analysis of H3 histone, H4 histone and
HLA-B7 mRNAs isolated from subcellular fractions of
cells treated with metabolic inhibitors 89
3-4 Northern blot analysis of H3 histone, H4 histone and
HLA-B7 mRNAs associated with the cytoskeleton and
soluble fractions from HeLa cells treated with
metabolic inhibitors 92
4-1 Schematic diagram of endogenous H3 histone mRNA ST519
and the beta-lactamase signal peptide-H3 histone
fusion mRNA 100
4-2 Expression of the signal peptide-histone fusion gene
in HeLa cells 103
4-3 The subcellular localization of the signal peptide-
histone fusion mRNA 106
4-4 The stability of the signal peptide-histone fusion
mRNA following inhibition of DNA synthesis 109
4-5 Schematic diagram of wild type and mutated signal
peptide-histone fusion mRNAs 112
4-6 Distribution of SPH3E1, SPH3Elalpha, and SPH3E1ATG*
mRNAs in nonmembrane-bound and membrane-bound
polysomal fractions 115
4-7 Effects of hydroxyurea treatment on SPH3E1 mRNA 119
vi


140


55
samples were denatured by heating at 100C for 2 minutes and
then quickly chilled at 0C to prevent reannealing prior to
loading onto the gel. Six percent polyacrylamide-8.3 M urea
(cross linked at 20:1 acrylamide:bis-acrylamide) gels with
dimensions of either 16 cm x 18 cm x 1.5 mm (1.5 mm x 6 mm
15 well comb or 1.5 mm x 4 mm 20 well comb) or 32.5 cm x 44
cm (0.5 mm x 8 mm 25 well comb) were prepared and pre-
electrophoresed for 30 minutes. The samples were loaded with
drawn-out capillary tubes or a Rainin Pipetteman (0-20 ul
capacity) using flat pipette tips. The large gels were
electrophoresed at 50 watts constant power (-1200 V, -45 A)
for 3-4 hours and the small gels were electrophoresed at 300
V (25-35 A) for 2-3 hours. The gels were placed on 2 sheets
of 3 MM Whatman paper, dried under vacuum at 80C for 1-2
hours and exposed to pre-flashed XAR5 x-ray film with a
Cronex Lightning Plus intensifying screen at -70C for
varying lengths of time. SI analysis was quantitated by
scanning laser densitometry of multiple autoradiographic
exposures that were within the linear range of the film.
In Vitro Nuclear Run-On Transcription Analysis
The in vitro nuclear run-on transcription assays were
performed by Anna Ramsey-Ewing as described by Baumbach et
al. (1987). Cells were harvested by centrifugation and the
cell pellet washed twice in cold isotonic buffer (125 mM
KC1, 30 mM Tris-HCl pH 7.9, 5 mM MgCl2, 10 mM beta-


Figure 6-2. The wild type signal peptide-histone fusion mRNA
is not released from the cytoskeleton in CD and puromvcin
co-treated polyclonal HeLa cell cultures.
Polyclonal cell cultures expressing the SPH3E1 fusion
gene were treated with cytochalasin D and puromycin for 20
minutes. Cytoskeleton and soluble RNAs were isolated and
subjected to SI nuclease protection analysis as described in
Chapter 2. Lane 1, control-csk? lane 2, control-sol; lane 3,
CD [10 Mg/ml]-csk; lane 4, CD [10 /g/ml]-sol; lane 5, CD [30
Mg/ml]-csk; lane 6, CD [30 /ig/ml ]-sol; lane 7, CD [10 /jg/ml]
and puro [0.4 mM]-csk; and lane 8, CD [10 jug/ml] and puro
[0.4 mM]-sol. MW is radiolabeled Hpa II digest of pBR322.


Figure 4-2. Expression of the signal peptide-histone fusion
gene in HeLa cells.
The expression of the signal peptide-histone fusion
gene was tested in a short term transient assay and analyzed
by an SI nuclease protection assay. HeLa cell monolayers
were transfected with 20 fiq DNA according to Gorman et al.
(1982), in a calcium phosphate/DNA complex prepared as
described by Graham and van der Eb (1973). The transfected
cells were incubated at 37C/ 5% C02 for 46 hours following
transfection. The cells were then harvested, total cellular
RNA was isolated and analyzed by SI nuclease protection
assay (200 RNA per sample) as described in Chapter 2.
Lanes 1 and 2, HeLa cells transfected with salmon sperm DNA?
lanes 3 and 4, HeLa cells transfected with pSPH3El DNA; and
lanes 5 and 6, HeLa cells transfected with pSPH3E2 DNA.
Marker lanes (M) are radiolabeled Hinfl digests of pBR322
DNA. Each lane represents RNA isolated from an independently
transfected cell culture.


7.0
T
5.2
T
4.1
T
85
PH
6.4
1
43 K~
SOL
CSK


194
of the nuclear matrix must explain the selective activation
of c-fos and actin genes and therefore does not appear
likely. In contrast, stimulation of c-fos gene expression by
cytochalasin D may be a result of the alterations in the
cytoskeleton due to the disruption of the microfilaments.
Analogous to the work presented here, it has recently been
reported that the disruption of the microtubules with
colchicine, nocodazole, or vinblastine induces the
transcription of chloramphenicol acetyl transferase genes
containing promoter elements from the human c-fos or beta-
actin gene (Ng, 1989). Interestingly, sequence analysis
indicates that actin and c-fos genes contain similar
transcription consensus sequences, referred to as serum
response elements, located in their promoter regions (Mohun
and Garret, 1987). The coordinate induction of c-fos and
actin transcription during cytochalasin D or colchicine
treatment may be mediated by the interaction of the same or
similar trans acting factor(s) with the serum response
element.
Several reports have postulated that the cytoskeleton
is involved in signal transduction (Rothstein, 1985;
Rebillard et al. 1987; Blum and Wicha, 1988). For example,
Rebillard and co-workers (1987) reported that the
stimulation of quiescent Swiss 3T3 cells with epidermal
growth factor (EGF) and insulin results in the rapid
accumulation of c-fos mRNA. The increase in c-fos gene


190
Discussion
Previous results suggest that the organization of the
cytoskeleton may influence gene expression and ultimately
cell growth and differentiation. We present evidence here
supporting the role of the cytoskeleton in regulating gene
expression. Specifically, disruption of the microfilaments
with cytochalasin D results in a rapid and dramatic
accumulation of c-fos mRNA. The induction of c-fos gene
expression during cytochalasin D treatment is accompanied by
an increase in transcription of the c-fos gene as directly
determined by in vitro nuclear run on analysis.
The effect of cytochalasin D treatment on transcription
is not limited to the c-fos gene. Consistent with previous
results actin transcription was also elevated in
cytochalasin D treated cells (Tannenbaum and Godman, 1983).
However, the stimulation of transcription of the c-fos and
actin genes was not a consequence of increased transcription
in general. RNA polymerase II transcription of several
genes, such as histone H4 (cell cycle dependent), histone
H2B (cell cycle independent), and HLA-B7 histocompatability
antigen, was unaffected in cells treated with cytochalasin
D. In addition, cytochalasin D had no effect on
transcription of the 18S ribosomal RNA genes by RNA
polymerase I or transcription of the beta-globin gene, a
transcriptionally inactive gene in HeLa cells.


49
Northern Blot Analysis
Agarose-Formaldehyde RNA Denaturing Gel Electrophoresis
RNA samples were resolved electrophoretically in 1.5%
agarose-6% (w/v) formaldehyde horizontal slab gels and
running buffer (20 mM MOPS (morpholinepropane-sulfonic acid)
pH 7.0", 5 mM sodium acetate, 1 mM EDTA, 3.7% (w/v)
formaldehyde) as described by Rave et al. (1979). Routinely
10 nq RNA per sample were dried in a Savant Speed Vac
concentrator and resuspended in 3.2 ul double distilled
water (ddH20) 5 ul formamide, 1 ul lOx MOPS (200 mM MOPS pH
7.0, 50 mM sodium acetate, 10 mM EDTA), 0.8 ul 37% (w/v)
formaldehyde and 2 ul sample buffer (0.5 ml formamide, 0.1
ml lOx MOPS, 80 ul 37% (w/v) formaldehyde and 0.32 ml 0.2%
bromophenol blue-90% glycerol). The samples were then heated
at 60C for 10 minutes, quick chilled to 0C and loaded onto
the gels. The 1.5% agarose-6% formaldehyde gels were
prepared by melting 3 g electrophoresis grade agarose in 144
ml ddH20 plus 20 ml lOx MOPS buffer in a microwave oven
(Whirlpool) at full power for 3-4 minutes. The agarose
solution was cooled to approximately 65C (warm to the
touch) while stirring and then adjusted to 6% formaldehyde
with 36 ml 37% (w/v) formaldehyde. For minigels, 35 ml
agarose-formaldehyde solution was cast in a plexiglass gel
tray (10 cm x 6.3 cm) with an 8 well comb (1 mm x 5 mm per
well). The gel was electrophoresed at 40-45 milliamps until
the bromophenol blue migrated 9 cm (approximately 1.5


104
transcribed from this gene is sufficiently stable to be
detected in a short-term transient transfection assay.
To determine the class of polysomes with which the
signal peptide-histone fusion mRNAs are associated, HeLa
cells were transfected with pSPH3El and 46 hours later
nonmembrane-bound and membrane-bound polysomes were isolated
as described in Chapter 2. RNA was isolated from both
classes of polysomes and analyzed by SI nuclease protection
assay using the Smal fragment of the signal peptide-histone
fusion gene (32P-5 1-labeled) As seen in Figure 4-3,
approximately 30% of the signal peptide-histone fusion mRNA
was found in the nonmembrane-bound polysome fraction and
approximately 70% was associated with the membrane-bound
polysomes (lanes 1 and 2 respectively). This is in contrast
to endogenous H3 histone mRNA which is represented by
approximately 90% in the nonmembrane-bound polysome and 10%
in the membrane-bound polysome fraction. The percent
distribution of the histone chimeric mRNA within the
polysomal RNA fractions was determined by densitometric
analysis of the autoradiograms and the values corrected for
the total yield of RNA in each subcellular fraction. While
the extent to which the signal peptide-histone fusion mRNA
is represented on the membrane-bound polysomes varied from
experiment to experiment, we always observed 60-90% of the
fusion mRNA in the membrane-bound polysome fraction (n=6).


142
membrane of the endoplasmic reticulum. In addition, the
nascent polypeptide is complexed with the proteins involved
in the translocation of the exported protein from the
cytoplasm into the lumen of the endoplasmic reticulum.
During the isolation of the cytoskeleton nearly all the
lipid components but not the protein substructure of the
endoplasmic reticulum membrane are removed by detergent
extraction (Lenk et al.. 1977). The cytoskeleton attachment
site associated with membrane-bound polysomes, which is
insensitive to cytochalasin D, may be the nascent
polypeptide complexed with the remnant protein structure of
the endoplasmic reticulum.
To test the possibility that the nascent polypeptide
serves as an additional anchor for the attachment of mRNA to
the cytoskeleton, we co-treated HeLa cells with cytochalasin
D and puromycin. Puromycin inhibits protein synthesis by
interrupting polypeptide chain formation which results in
the release of the nascent polypeptide and ribosoraal
subunits from the mRNA. The distribution of membrane-bound
and nonmembrane-bound polysomal RNA within the cytoskeleton
and soluble fractions was then determined by Northern blot
analysis.
As seen in Figure 5-3, treatment of HeLa cells with
cytochalasin D and puromycin released membrane-bound
polysomal raRNAs from the cytoskeleton (lanes 7 and 8).
Approximately 25% of HLA-B7 mRNA and 20% of hCGa mRNA


Table 4-1. Quantitation of SPH3E1, SPH3Elalpha and
SPH3E1ATG" inRNA in nonmembrane-bound and membrane-bound
polysome fractions.
116
Nonmembrane-bound Membrane-bound
SPH3E1 mRNA
32%
68%
SPH3Elalpha mRNA
60%
40%
SPH3E1ATG' mRNA
82%
18%
Endogenous H3 mRNA
83%
17%
The distribution of SPH3E1, SPH3Elalpha, SPH3E1ATG-,
and endogenous H3 histone mRNAs in nonmembrane-bound and
membrane-bound polysome fractions was quantitated by
densitometric analysis of the SI nuclease protection assay
presented in Figure 4-6. The densitometric values obtained
represent the quantity of mRNA in 5 g RNA. The values were
then adjusted to reflect the total yield of RNA recovered in
each fraction.


97
membrane-bound polysomes. The ability of the cell to
selectively recognize and destabilize histone mRNA in a
foreign subcellular compartment could then be examined.
Based on the data presented here, we propose that the
subcellular location of histone mRNA, with respect to the
class of polysomes in which they are associated with, plays
a significant role in the coupling of histone mRNA stability
with DNA replication.
Results
The construction of the signal peptide-histone fusion
gene is described in Chapter 2 and outlined in Figure 2-1.
The upstream flanking region of the chimeric gene contains
210 base pairs of the 5' regulatory sequences of the cell
cycle-dependent human H3 histone gene pST519 including the
TATAA and CCAAT consensus sequences. This region is followed
by the H3 mRNA cap site and sequences encoding the initial
20 nucleotides of the non-translated H3 histone leader. The
fusion gene is, therefore, under transcriptional control of
the H3 histone gene promoter, and a segment of the H3 leader
which has been implicated in the coupling of histone mRNA
levels to DNA synthesis is present (Morris et al.. 1986).
The H3 histone leader segment is fused to the untranslated
leader sequences of the beta-lactamase signal peptide
including the ATG translation start codon. The H3 histone
structural gene is fused in frame to the signal peptide


212
Lin, D. Tobin, K., Grumet, M., and Lin, S. (1980) J. Cell
Biol. 84., 455.
Lingappa, V., Chaidez, J., Yost, C., and Hedgpeth, J. (1984)
Proc. Natl. Acad. Sci. U.S.A. 81, 456.
Luscher, B., Stauber, C., Schindler, R., and Schumperli, D.
(1985) Proc. Natl. Acad. Sci. U.S.A. 82, 4389.
Maniatis T., Fritsch, E., and Sambrook, J. (1982a) in
Molecular Cloning. (Cold Spring Harbor, New York) pp 86-90.
Maniatis T., Fritsch, E., and Sambrook, J. (1982b) in
Molecular Cloning. (Cold Spring Harbor, New York) pp 109-
112.
Marashi, F., Helms, S., Shiels, A., Silverstein, S.,
Greenspan, D., Stein, G., and Stein, J. (1986) Can. J.
Bioch. Cell Biol. 64, 277.
Marzluff, W., and Huang, R. (1985) in Transcription and
Translation. Hames, B. and Higgins, S. Eds. (IRL Press,
Oxford) pp 89-129.
Maurer, R., and Mckean, D. (1978) J. Biol. Chem. 253. 6315.
Meyer, D. Krause, E., and Dobberstein, B. (1982a) Nature
297. 647.
Meyer, R. Louvard, D., and Dobberstein, B. (1982b) J. Cell
Biol. 92, 579.
Milcarek, C., and Penman, S. (1974) J. Mol. Biol. 89, 327.
Milsted, A., Silver, B., Cox, R., and Nilson, J. (1985)
Endocrinology 117. 2033.
Miranda, A., Godman, G., Deitch, A., and Tannenbaum, S.
(1974a) J. Cell Biol. 61, 481.
Miranda, A., Godman, G., and Tannenbaum, S. (1974b) J. Cell
Biol. 62, 406.
Mitchell, R., Zokas, L., Schreiber, R., and Verma, I. (1985)
Cell 40, 209.
Mohun, T., and Garret, N. (1987) Develop. 101. 393.
Moll, R., Franke, W., Schiller, D., Geiger, B., and Krepler,
R. (1982) Cell 21, 11.


5
in conjunction with the decrease in proliferative activity
(Bird et al.. 1985; Challoner et al.. 1989; Stein et al..
1989) .
The expression of histone genes during the cell cycle
is regulated at both the transcriptional and
posttrnscriptional levels. Analysis by hybrid selection of
RNA radiolabeled in synchronized HeLa cells demonstrates a
burst of 3- to 5- fold in transcription of the histone genes
within the first two hours of S phase, followed by a
reduction to a basal level that remains constant until the
following S phase (Plumb et al.. 1983a). Similar results
were obtained when histone gene transcription during the
cell cycle was measured by nuclear run-on analysis (Sive et
al.. 1984; Baumbach et al.. 1987). Steady state levels of
histone mRNA, however, were elevated 15-20 fold during S
phase and closely paralleled the rates of DNA synthesis
(Plumb et al.. 1983a; Heintz et al.. 1983; Baumbach et al..
1984) Maximal accumulation of histone mRNA occurred at 5
hours into S phase and returned to a basal level as DNA
synthesis ended. The timing and extent of the increase in
histone transcription can not account for the marked
increase in histone mRNA steady state levels and must be a
consequence of a posttranscriptional control mechanism.
Posttranscriptional regulation of histone gene expression is
further supported by DNA synthesis inhibitor studies. The
estimated half life of histone mRNA in exponentially growing


Figure 4-6. Distribution of SPH3E1. SPH3Elalpha. and
SPH3E1ATG mRNAs in nonmembrane-bound and membrane-bound
polvsomal fractions.
HeLa cell monolayers were transfected with the signal
peptide-histone fusion genes by the calcium phosphate
precipitation method and 46 hours posttransfection
nonmembrane-bound and membrane-bound polysomes were
isolated. RNA (5 jug) from each subcellular fraction was
assayed for signal peptide-histone fusion mRNA and
endogenous H3 histone mRNA content by SI nuclease protection
analysis as described in Chapter 2. Abbreviations: SPH3E1,
wild type signal peptide-histone fusion mRNA; alpha,
hydrophobic mutant of SPH3E1; ATG, signal peptide
translation initiation codon mutant of SPH3E1. Lane 1,
nonmembrane-bound polysomal RNA; Lane 2, membrane-bound
polysomal RNA.


Figure 4-4. The stability of the signal peptide-histone
fusion mRNA following inhibition of DNA synthesis.
HeLa cell monolayers were transfected with pSPH3El DNA
and cultured as described in Chapter 2. Forty-six hours
after transfection, half of the transfected HeLa cell
cultures were treated with 1 mM hydroxyurea in completed
medium for 1 hour at 37C, 5% C02. Cell cultures were
harvested and nonmembrane-bound and membrane-bound polysomes
were isolated. Total cellular RNA from both the control and
hydroxyurea-treated samples was also isolated. The RNA was
analyzed by the SI nuclease protection assay as described in
Chapter 2, using 5 nq RNA per sample. The distribution of
mRNA species within the subcellular fractions was
quantitated by densitometric analysis of the autoradiograms.
Lanes: 1) control (untreated cells) nonmembrane-bound
polysomes; 2) control, membrane-bound polysomes; 3)
hydroxyurea, nonmembrane-bound polysomes; 4) hydroxyurea,
membrane-bound polysomes; 5) and 6) control, total cellular
RNA; and 7) and 8) hydroxyurea, total cellular RNA. Lanes 5-
8 represent RNA isolated from independently transfected cell
cultures.


86
mRNA was located almost exclusively in the cytoskeleton and
membrane-bound polysome fractions. The widespread
subcellular localization of histone mRNA may be due to non
specific trapping of nonmembrane-bound polysomes in each of
the fractions during the isolation procedure. Alternatively,
this observation may reflect the presence of mRNAs encoding
variant histone proteins that could be targeted to the
different cytoplasmic compartments. Inhibition of DNA
synthesis with hydroxyurea results in a rapid
destabilization of cell cycle dependent histone mRNAs,
whereas cell cycle independent histone mRNAs remain stable
in the presence of this drug (Baumbach et al.. 1984; Heintz
et al.. 1983; DeLisle et al.. 1983; Graves and Marzluff,
1984; Plumb et al.. 1983). Conversely, inhibition of protein
synthesis with cycloheximide results in a 2-3 fold increase
in cytoplasmic levels of histone mRNAs (Butler and Mueller,
1973; Baumbach et al.. 1984). Differential sensitivity of
histone mRNAs in the subcellular fractions to hydroxyurea or
cycloheximide treatment would favor the latter suggestion
that cell cycle dependent and independent histone gene
transcripts may selectively reside in specific cytoplasmic
compartments.
Treatment of exponentially growing HeLa cells with
hydroxyurea, as described in Chapter 2, resulted in a
complete and selective destabilization of histone mRNA in
the LER, nonmembrane-bound and membrane-bound polysomal


CHAPTER 8
SUMMARY AND FUTURE CONSIDERATIONS
Tre initial emphasis of the work presented in this
dissertation was focused on studying the posttranscriptional
regulation of histone gene expression. When this project was
undertaken little was known concerning the coupling of
histone mRNA stability to DNA replication. It was clear that
the destabilization of histone mRNA during inhibition of DNA
synthesis was selective and required ongoing protein
synthesis (Butler and Mueller, 1973; Gallwitz, 1975; Stahl
and Gallwitz, 1977; Stimac et al.. 1983; Baumbach et al..
1984; Helms et al.. 1984). In addition, it was known that
the kinetics of histone gene expression were extremely rapid
which is evident by the appearance of translatable histone
mRNAs on polysomes within minutes following transcription
(Borun et al.. 1967) and the equally rapid transfer of newly
synthesized histone protein into the nucleus (Spaulding et
al.. 1966; Robbins and Borun, 1967). At that time, several
research groups were providing evidence that suggested the
cell is organized in a highly ordered manner. Penman and co
workers, while studying the structural and functional
properties of the cytoskeleton, were developing the
196


100
CAP
I
ATG
TAA
5'
[
3'
pST 5 19
CAP
I
ATG
ir/L-
a;g
TAA
i
pSPH3E1


101
To test for expression of the signal peptide-histone
fusion gene in human cells, we transfected pSPH3El and
pSPH3E2 DNA into HeLa cell monolayers by the calcium
phosphate precipitation method as described in Chapter 2
(Gorman et al., 1982; Graham and van der Eb, 1973). Forty-
six hours post-transfection cells were harvested and total
cellular RNA was isolated. The RNA was subjected to SI
nuclease protection analysis by a modification of Berk and
Sharp (1978). The probe used in the SI nuclease assays was
the Smal fragment from pSPH3El which was radiolabeled at its
5' termini as described in Chapter 2. The probe is
complimentary to endogenous HeLa histone H3 mRNA from the 5*
end-labeled Smal site within the protein coding region to
the signal peptide-histone fusion junction and therefore
protects a 130 nucleotide region of the H3 mRNA from SI
nuclease digestion. As seen in Figure 4-2, when total
cellular RNA from two independent transfections of HeLa
cells with pSPH3El or pSPH3E2 DNA was analyzed by SI
nuclease digestion, a 130 nucleotide fragment corresponding
to endogenous H3 histone mRNA was protected as well as an
approximately 280 nucleotide fragment corresponding to the
signal peptide-histone fusion mRNA species (lanes 3-6). In
HeLa cells transfected with salmon sperm DNA, only the 130
nucleotide fragment was detected (lanes 1 and 2). These data
demonstrate that the signal peptide-histone fusion gene is
capable of expression in HeLa cells and that the mRNA