Citation
Expression of the TPI gene of Saccharomyces Cerevisiae is controlled by a single complex upstream activating sequence containing binding sites for three Trans-acting factors: REB1, RAP1, and GCR1

Material Information

Title:
Expression of the TPI gene of Saccharomyces Cerevisiae is controlled by a single complex upstream activating sequence containing binding sites for three Trans-acting factors: REB1, RAP1, and GCR1
Creator:
Scott, Edward William, 1964-
Publication Date:
Language:
English
Physical Description:
x, 136 leaves : ill. ; 29 cm.

Subjects

Subjects / Keywords:
DNA ( jstor )
Enzymes ( jstor )
Gels ( jstor )
Genetic mutation ( jstor )
In vitro fertilization ( jstor )
Oligonucleotides ( jstor )
Plasmids ( jstor )
Saccharomyces cerevisiae ( jstor )
Sequencing ( jstor )
Yeasts ( jstor )
Saccharomyces cerevisiae -- genetics ( mesh )
Triose-Phosphate Isomerase -- genetics ( mesh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1992.
Bibliography:
Includes bibliographical references (leaves 124-135).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Edward William Scott V.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
27132943 ( OCLC )
ocm27132943
0027182529 ( ALEPH )

Downloads

This item has the following downloads:

ELIHRZ9P2_9IMSWZ.xml

AA00012908_00001.pdf

expressionoftpig00scot_0089.txt

AA00012908_00001_0087.txt

expressionoftpig00scot_0021.txt

AA00012908_00001_0013.txt

expressionoftpig00scot_0118.txt

expressionoftpig00scot_0003.txt

expressionoftpig00scot_0018.txt

expressionoftpig00scot_0040.txt

expressionoftpig00scot_0123.txt

AA00012908_00001_0142.txt

expressionoftpig00scot_0145.txt

AA00012908_00001_0099.txt

AA00012908_00001_0069.txt

expressionoftpig00scot_0091.txt

AA00012908_00001_0125.txt

AA00012908_00001_0015.txt

expressionoftpig00scot_0085.txt

AA00012908_00001_0017.txt

expressionoftpig00scot_0113.txt

AA00012908_00001_0082.txt

expressionoftpig00scot_0004.txt

AA00012908_00001_0126.txt

expressionoftpig00scot_0124.txt

expressionoftpig00scot_0048.txt

AA00012908_00001_0005.txt

expressionoftpig00scot_0071.txt

AA00012908_00001_0074.txt

AA00012908_00001_0068.txt

expressionoftpig00scot_0047.txt

AA00012908_00001_0032.txt

AA00012908_00001_0091.txt

AA00012908_00001_0046.txt

AA00012908_00001_0141.txt

AA00012908_00001_0044.txt

AA00012908_00001_0064.txt

expressionoftpig00scot_0066.txt

expressionoftpig00scot_0084.txt

AA00012908_00001_0039.txt

expressionoftpig00scot_0143.txt

AA00012908_00001_0006.txt

AA00012908_00001_0149.txt

expressionoftpig00scot_0043.txt

AA00012908_00001_0132.txt

expressionoftpig00scot_0141.txt

AA00012908_00001_0066.txt

AA00012908_00001_0092.txt

expressionoftpig00scot_0087.txt

expressionoftpig00scot_0128.txt

AA00012908_00001_0019.txt

expressionoftpig00scot_0127.txt

AA00012908_00001_0051.txt

AA00012908_00001_0002.txt

AA00012908_00001_0009.txt

AA00012908_00001_0029.txt

AA00012908_00001_0112.txt

AA00012908_00001_0120.txt

expressionoftpig00scot_0064.txt

AA00012908_00001_0133.txt

AA00012908_00001_0136.txt

expressionoftpig00scot_0007.txt

expressionoftpig00scot_0136.txt

expressionoftpig00scot_0101.txt

AA00012908_00001_0018.txt

AA00012908_00001_0031.txt

expressionoftpig00scot_0024.txt

AA00012908_00001_0010.txt

expressionoftpig00scot_0119.txt

AA00012908_00001_0144.txt

expressionoftpig00scot_0030.txt

AA00012908_00001_0081.txt

AA00012908_00001_0145.txt

expressionoftpig00scot_0100.txt

AA00012908_00001_0003.txt

expressionoftpig00scot_0065.txt

expressionoftpig00scot_0022.txt

AA00012908_00001_0067.txt

expressionoftpig00scot_0068.txt

AA00012908_00001_0076.txt

expressionoftpig00scot_0115.txt

expressionoftpig00scot_0037.txt

expressionoftpig00scot_0073.txt

expressionoftpig00scot_0081.txt

AA00012908_00001_0115.txt

AA00012908_00001_0096.txt

expressionoftpig00scot_0095.txt

AA00012908_00001_0121.txt

expressionoftpig00scot_0038.txt

AA00012908_00001_0119.txt

expressionoftpig00scot_0108.txt

AA00012908_00001_0049.txt

AA00012908_00001_0147.txt

AA00012908_00001_0117.txt

AA00012908_00001_0055.txt

expressionoftpig00scot_0072.txt

expressionoftpig00scot_0067.txt

AA00012908_00001_0059.txt

expressionoftpig00scot_0019.txt

expressionoftpig00scot_0012.txt

AA00012908_00001_0065.txt

expressionoftpig00scot_0023.txt

expressionoftpig00scot_0130.txt

AA00012908_00001_0100.txt

AA00012908_00001_0094.txt

AA00012908_00001_0022.txt

expressionoftpig00scot_0086.txt

expressionoftpig00scot_0075.txt

expressionoftpig00scot_0025.txt

AA00012908_00001_0140.txt

expressionoftpig00scot_0076.txt

expressionoftpig00scot_0140.txt

AA00012908_00001_0084.txt

expressionoftpig00scot_0131.txt

AA00012908_00001_0052.txt

expressionoftpig00scot_0125.txt

expressionoftpig00scot_0049.txt

expressionoftpig00scot_0009.txt

AA00012908_00001_0040.txt

expressionoftpig00scot_0138.txt

expressionoftpig00scot_0120.txt

AA00012908_00001_0128.txt

expressionoftpig00scot_0135.txt

AA00012908_00001_0077.txt

expressionoftpig00scot_0013.txt

expressionoftpig00scot_0031.txt

expressionoftpig00scot_0008.txt

AA00012908_00001_0078.txt

expressionoftpig00scot_0134.txt

expressionoftpig00scot_0083.txt

AA00012908_00001_0057.txt

expressionoftpig00scot_0137.txt

expressionoftpig00scot_0006.txt

AA00012908_00001_0148.txt

expressionoftpig00scot_0069.txt

AA00012908_00001_0061.txt

expressionoftpig00scot_0090.txt

AA00012908_00001_0143.txt

AA00012908_00001_0072.txt

AA00012908_00001_pdf.txt

AA00012908_00001_0079.txt

expressionoftpig00scot_0002.txt

expressionoftpig00scot_0126.txt

expressionoftpig00scot_0098.txt

expressionoftpig00scot_0080.txt

expressionoftpig00scot_0104.txt

expressionoftpig00scot_0088.txt

expressionoftpig00scot_0027.txt

expressionoftpig00scot_0139.txt

expressionoftpig00scot_0110.txt

expressionoftpig00scot_0010.txt

AA00012908_00001_0106.txt

AA00012908_00001_0098.txt

expressionoftpig00scot_0032.txt

expressionoftpig00scot_0077.txt

expressionoftpig00scot_0044.txt

AA00012908_00001_0108.txt

expressionoftpig00scot_0112.txt

expressionoftpig00scot_0017.txt

AA00012908_00001_0025.txt

expressionoftpig00scot_0033.txt

AA00012908_00001_0037.txt

ELIHRZ9P2_9IMSWZ_xml.txt

expressionoftpig00scot_0109.txt

AA00012908_00001_0146.txt

AA00012908_00001_0118.txt

AA00012908_00001_0123.txt

AA00012908_00001_0024.txt

expressionoftpig00scot_0053.txt

AA00012908_00001_0054.txt

AA00012908_00001_0038.txt

AA00012908_00001_0134.txt

expressionoftpig00scot_0063.txt

expressionoftpig00scot_0096.txt

expressionoftpig00scot_0148.txt

expressionoftpig00scot_0028.txt

expressionoftpig00scot_0079.txt

expressionoftpig00scot_0034.txt

AA00012908_00001_0086.txt

AA00012908_00001_0045.txt

expressionoftpig00scot_0014.txt

AA00012908_00001_0056.txt

expressionoftpig00scot_0051.txt

expressionoftpig00scot_0058.txt

AA00012908_00001_0124.txt

AA00012908_00001_0028.txt

AA00012908_00001_0090.txt

AA00012908_00001_0041.txt

AA00012908_00001_0026.txt

expressionoftpig00scot_0036.txt

expressionoftpig00scot_0099.txt

AA00012908_00001_0114.txt

AA00012908_00001_0063.txt

expressionoftpig00scot_0055.txt

expressionoftpig00scot_0132.txt

expressionoftpig00scot_0092.txt

expressionoftpig00scot_0078.txt

expressionoftpig00scot_0144.txt

expressionoftpig00scot_0041.txt

expressionoftpig00scot_0059.txt

AA00012908_00001_0058.txt

expressionoftpig00scot_0001.txt

AA00012908_00001_0097.txt

AA00012908_00001_0080.txt

AA00012908_00001_0107.txt

expressionoftpig00scot_0107.txt

AA00012908_00001_0053.txt

expressionoftpig00scot_0114.txt

AA00012908_00001_0033.txt

AA00012908_00001_0127.txt

AA00012908_00001_0102.txt

AA00012908_00001_0042.txt

AA00012908_00001_0027.txt

AA00012908_00001_0093.txt

AA00012908_00001_0073.txt

expressionoftpig00scot_0045.txt

expressionoftpig00scot_0035.txt

expressionoftpig00scot_0117.txt

expressionoftpig00scot_0050.txt

AA00012908_00001_0034.txt

AA00012908_00001_0111.txt

expressionoftpig00scot_0133.txt

expressionoftpig00scot_0056.txt

expressionoftpig00scot_0000.txt

expressionoftpig00scot_0147.txt

expressionoftpig00scot_0111.txt

expressionoftpig00scot_0062.txt

AA00012908_00001_0137.txt

AA00012908_00001_0048.txt

expressionoftpig00scot_0121.txt

AA00012908_00001_0023.txt

expressionoftpig00scot_0020.txt

expressionoftpig00scot_0146.txt

expressionoftpig00scot_0015.txt

AA00012908_00001_0020.txt

AA00012908_00001_0139.txt

expressionoftpig00scot_0060.txt

AA00012908_00001_0011.txt

expressionoftpig00scot_0097.txt

AA00012908_00001_0071.txt

expressionoftpig00scot_0116.txt

AA00012908_00001_0095.txt

expressionoftpig00scot_0026.txt

AA00012908_00001_0085.txt

AA00012908_00001_0116.txt

expressionoftpig00scot_0093.txt

expressionoftpig00scot_0094.txt

AA00012908_00001_0105.txt

expressionoftpig00scot_0103.txt

expressionoftpig00scot_0142.txt

AA00012908_00001_0083.txt

AA00012908_00001_0131.txt

expressionoftpig00scot_0057.txt

AA00012908_00001_0016.txt

expressionoftpig00scot_0129.txt

expressionoftpig00scot_0106.txt

AA00012908_00001_0007.txt

AA00012908_00001_0129.txt

AA00012908_00001_0050.txt

expressionoftpig00scot_0042.txt

AA00012908_00001_0110.txt

AA00012908_00001_0001.txt

AA00012908_00001_0089.txt

AA00012908_00001_0070.txt

AA00012908_00001_0008.txt

AA00012908_00001_0047.txt

expressionoftpig00scot_0052.txt

expressionoftpig00scot_0061.txt

AA00012908_00001_0122.txt

AA00012908_00001_0138.txt

AA00012908_00001_0060.txt

AA00012908_00001_0109.txt

AA00012908_00001_0030.txt

expressionoftpig00scot_0105.txt

expressionoftpig00scot_0054.txt

AA00012908_00001_0062.txt

expressionoftpig00scot_0046.txt

AA00012908_00001_0075.txt

AA00012908_00001_0130.txt

expressionoftpig00scot_0029.txt

expressionoftpig00scot_0082.txt

AA00012908_00001_0014.txt

expressionoftpig00scot_0102.txt

AA00012908_00001_0043.txt

expressionoftpig00scot_0039.txt

AA00012908_00001_0012.txt

expressionoftpig00scot_0005.txt

AA00012908_00001_0101.txt

expressionoftpig00scot_0070.txt

AA00012908_00001_0135.txt

AA00012908_00001_0104.txt

AA00012908_00001_0004.txt

AA00012908_00001_0036.txt

expressionoftpig00scot_0016.txt

expressionoftpig00scot_0011.txt

AA00012908_00001_0035.txt

AA00012908_00001_0113.txt

AA00012908_00001_0088.txt

expressionoftpig00scot_0074.txt

expressionoftpig00scot_0122.txt

AA00012908_00001_0021.txt

AA00012908_00001_0103.txt


Full Text










EXPRESSION OF THE TPIGENE OF SACCHAROMYCES CEREVISIAE
IS CONTROLLED BY A SINGLE COMPLEX UPSTREAM ACTIVATING
SEQUENCE CONTAINING BINDING SITES FOR THREE TRANS-ACTING
FACTORS: REB1, RAP1, AND GCR1







BY

EDWARD WILLIAM SCOTT V


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


1992








DEDICATION


To my wife Rochelle for her love and support in all things I do.

Father, first and foremost this belongs to you. Your rearing has

produced my thirst for knowledge about all I touch. Your example

demonstrates that education is not a stagnant entity cast off after

college, rather, it is a lifelong and enlightening endeavor.

Why, all delights are vain; but that most vain,
Which with pain purchased doth inherit pain:
As, painful to pore upon a book
To seek the light of truth; while truth the while
Doth falsely blind the eyesight of his look:
Light seeking light doth light of light beguile:
So, ere you find where light in darkness lies,
Your light grows dark by the losing of your eyes.

Love's Labour's Lost
William Shakespeare









ACKNOWLEDGMENTS


I would like to extend my deepest appreciation to my mentor

Dr. Henry V. Baker for providing superb direction to my graduate

education and personal development as a scientist. My deepest

regards to the members of my dissertation committee: Dr. Lonnie O.

Ingram, Dr. Alfred S. Lewin, Dr. Richard W. Moyer, and Dr. Thomas C.

Rowe for their invaluable insights. I thank the Department of

Immunology and Medical Microbiology for providing a climate

conducive to intellectual growth and exchange.









TABLE OF CONTENTS


page

ACKNOWLEDGMENTS ...................................... .............. .......................... iii

LIST OF TABLES .......................................................................................... vi

LIST OF FIGURES ............................................................................................... vii

ABSTRACT ..................................................................................................... ix

INTRODUCTION ................................................................................................. 1

MATERIALS AND METHODS ....................................................................... 17

S trains ..................................................................................................... 17
Media and Growth Conditions ....................................... ........ 17
Nucleic Acid Manipulations .......................................................... 19
Generation of Double Strand DNA Oligonucleotides ............. 19
DNA Sequencing ............................................................................ 20
Primer Extension ........................................................ ................. 22
Plasmid Construction ................................................ ................ 22
Constructs for sequencing ............................................... 22
5' deletion scheme ............................................... ........... 23
Internal deletion scheme .................................... ......... 26
Constructs to assess UAS activity ............................... 29
Mutant UAS oligonucleotides driving expression of
the TPI::lacZfusion ...................................... ......... 29
Site-Directed Mutagenesis ........................................... .......... 30
Transformation ................................................................................. 32
Screen for Unit Copy Integrants of the TPI::IacZ Fusions
at U R A 3 .............................................................................................. 32
P-Galactosidae Assays ................................................ ............ 35
In vitro DNasel Protection Assays ............................................ 36
DNA Band Shift Assays ............................................................. 38
In vivo Methylation Protection Analysis ................................. 38









R ES U LTS ........................................................................................................ 46

The Mature 5' Ends of Steady-State TPI Transcripts are
Unaffected by a gcrl Mutation ................................... ........ 46
Identification of the 5' Boundary of the TPI Controlling
R eg io n ................................................................................................... 4 8
An Upstream Activating Sequence Activity for TPI
Resides from Position -377 to -327 in the 5'
Noncoding Region ................................................................... 59
Internal Deletions Indicate Single UAS Element
Responsible for TPI Transcription ........................................ 63
Mutational Analysis of UASTPI ....................................................... 68
In vitro DNase I Protection Assays Reveal Binding of
the REB1 Site and the RAP1 Site ............................................ 75
DNA Band Shift Assays Demonstrate REB1 Binding to
TPI5' Noncoding Region ........................................... .......... 81
In vivo Methylation Protection Assays .................................... 82
Site-Directed Mutagenesis of Transcription Factor
Binding Sites in the UAS of TPI .......................................... 90

D ISC USS IO N ................................................................................................. 99

R EFER EN C ES ........................................................................................................ 124

BIOGRAPHICAL SKETCH ................................................................................. 136









LIST OF TABLES


a....

Table 1. Strains .......... ......................................................................... 18

Table 2. Oligonucleotides ............................................................................. 21









LIST OF FIGURES


Figure 1.


Figure

Figure


Figure 4.



Figure 5.

Figure 6.




Figure 7.

Figure 8.



Figure 9.


Figure 10.


Figure 11.


Scheme to generate 5' deletions in the TPI
5' noncoding region ................................... ........... 25

Scheme to create internal deletions .......................... 28

Scheme for integration of TPI::IacZ gene
fusion constructs at URA3 ................................... 34

Primer extension analysis of the 5' ends of
the TPI transcript in wild-type and
gcrl-deletion mutant strains ........................... 50

Sequence of TPI 5' Noncoding Region .......................... 52

Effect of 5' deletions upon expression of a
TPI::lacZ fusion integrated in unit copy
at the URA3 locus in wild-type and
gcrl-deletion mutant strains ............................ 57

Identification of UASTPI ................................................... ...... 62

Effect of internal deletions on p-galactosidase
activity expressed from a TPI::lacZ
gene fusion ................................... ........... ....... .... 67

Summary composite of the 5' noncoding
region of TPI .......................................... ................. 69

Mutational analysis of UASTPI utilizing
mutant oligonucleotides .......................................... 72

In vitro DNase I protection assays demonstrating
protection of the RAP1-binding site ............... 78

vii








Figure 12. In vitro DNase I protection assays demonstrating
protection of the REB1-binding site ................ 80

Figure 13. DNA band shift assays demonstrating REB1 binding
to the 5' noncoding region of TPI ........................ 84

Figure 14. Genomic footprinting of the bottom strand
of the TPI 5' noncoding region ............................. 87

Figure 15. Genomic footprinting of the top strand
of the TPI 5' noncoding region ............................. 92

Figure 16. Effect of site-directed mutations upon
3-galactosidase activity expressed
from a TPI::lacZfusion ............................................ 94

Figure 17. Composite summary of the TPI controlling
region ............................................................................... 109

Figure 18. Model of protein interactions at UASTPI ......................... 122


viii














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

EXPRESSION OF THE TPI GENE OF SACCHAROMYCES CEREVISIAE IS
CONTROLLED BY A SINGLE COMPLEX UPSTREAM ACTIVATING SEQUENCE
CONTAINING BINDING SITES FOR THREE TRANS-ACTING FACTORS:
REB1, RAP1, AND GCR1

By

Edward William Scott V

May 1992

Chairman: Henry V. Baker, Ph. D.
Major Department: Immunology and Medical Microbiology

In Saccharomyces cerevisiae the enzymes of glycolysis

constitute 30-60 percent of the soluble protein. GCR1 gene function

is required for high-level glycolytic gene expression. This study

involved a biochemical and genetic characterization of TPI, a gene

affected by gcrl lesions. Primer extension experiments on TPI

transcripts isolated from wild-type and gcrl mutant strains mapped

the initiation or "I" site of transcription to a pair of adenines 29 and

30 bases upstream of the start of translation. To delineate the TPI








controlling region, a TPI::lacZ gene fusion was employed. Nuclease

Bal31 deletion analysis identified a single upstream activating

sequence, UASTPI, responsible for high-level TPI expression. DNase I

protection and in vivo dimethyl sulfate methylation protection

assays indicated the binding of three trans-acting factors to four

sites within UASTPI. GCR1 binds two sites: 1) a sequence element

which functionally requires the CTTCC pentamer from position -375

to -370; 2) an element requiring the CATCC pentamer from position

-335 to -330. REB1 binds to an almost perfect consensus binding

site from position -397 to -387. RAP1 binds from -358 to -346.

REB1 and RAP1 binding at UASTPI are independent of GCR1 binding in

vivo. Site-directed mutation of the REB1-binding site reduced the

expression of the TPI::IacZ gene fusion by two-fold. Mutation of the

RAP1 or both GCR1-binding sites abolished expression. Thus, TPI

absolutely requires RAP1 and GCR1 binding for expression and

requires REB1 for full expression. This work suggests a mechanism

of high-level glycolytic gene expression mediated primarily through

the actions of RAP1 and GCR1.













INTRODUCTION


A fundamental process in biology is the regulation of gene

expression. For it is this process which allows a single fertilized

human egg to develop into an organism of 1014 cells in just nine

months. Furthermore, the study of cancer has shown that when the

process of gene regulation is perturbed a neoplastic transformation

may result. The yeast Saccharomyces cerevisiae offers an ideal

model system to study the regulation of gene expression. Over the

years a sophisticated genetic system has developed that allows one

to manipulate a gene in vitro and then reintroduce the mutated gene

back into the genome to assess the effect of the manipulation in

vivo. In addition, transcription factors in yeast often have homologs

in higher eukaryotes. For example, the JUN oncoprotein binds the

same recognition sequence and has extensive amino acid homologies

with the yeast transcriptional activator GCN4. JUN is even able to

functionally complement a gcn4 mutation in yeast (Struhl, 1988).

It has long been known that upon neoplastic transformation in

1








2

certain types of cancer there is an increase in aerobic glycolysis

(Warburg, 1930). Saccharomyces cerevisiae utilizes aerobic

glycolysis to a much greater extent than respiration (Lagunas,

1986). The enzymatic pathway of glycolysis in yeast is well

established. The enzymes of glycolysis, while few in number,

compose between 30-60% of the total soluble protein (Fraenkel,

1982). This observation suggests that the genes encoding these

enzymes are among the most highly expressed in yeast. Indeed,

mRNA encoding glycolytic enzymes has been demonstrated to be a

major fraction of total yeast mRNA (Holland et al., 1977; Holland and

Holland, 1978). The regulation of the genes encoding the glycolytic

enzymes is currently receiving much study, but no overall consensus

regulatory mechanisms have yet been identified, rather some

similarities in regulatory elements and factors have been noted.

These similarities will be addressed subsequently.

Mutations affecting the flux of metabolites through the

glycolytic pathway tend to map to single loci and affect single

enzymes (Fraenkel, 1982). However, Clifton et al. (1978) isolated a

mutant that has severely reduced levels of most glycolytic enzymes.

Yet genetic analysis showed that this strain contains a mutation








3

that segregates as a single gene. Due to its pleiotrophic nature the

gene was named GCR1 for glycolysis regulation. Strains harboring a

gcrl mutation express the genes encoding the enzymes of glycolysis

at approximately 5-10% of wild-type levels (Clifton et al., 1978;

Clifton and Fraenkel, 1981; Baker, 1986). gcrl mutants exhibit poor

growth on glucose while retaining adequate growth on non-

fermentable carbon sources (Clifton and Fraenkel, 1981). The

reduction in enzyme levels has been shown to be mirrored by a

corresponding reduction in steady-state mRNA levels for several

affected enzymes (Holland et al., 1987; Santangelo and Tornow,

1990; Scott et al., 1990). GCR1 has been cloned (Kawasaki and

Fraenkel, 1982) and sequenced (Baker, 1986; Holland et al., 1987).

DNA sequence analysis indicates that GCR1 encodes a polypeptide of

844 amino acids with a molecular weight of 94,414 Daltons.

Much attention and effort has been devoted to the study of gene

regulation in S. cerevisiae and other systems. The plieotrophic

nature of mutations in GCR1 suggests that the gene is involved in

the coordinate regulation of expression of the genes encoding

glycolytic enzymes. However, the residual level of expression of

aldolase, triose-phosphate isomerase, glyceraldehyde-3-phosphate











dehydrogenase, phosphoglycerate kinase, and pyruvate kinase in a

gcrl mutant is somewhat inducible by glucose, relative to

expression under gluconeogenic conditions (Baker, 1986). Insights

into the mechanism by which the GCR1 gene product exerts its

effect will prove valuable in the exploration of glycolytic gene

regulation.

This study will attempt to elucidate the cis and trans-acting

elements involved in the expression of TPI, the sole gene encoding

triose-phosphate isomerase activity in Saccharomyces cerevisiae.

Triose-phosphate isomerase catalizes the reversible isomerization

of dihydroxyacetone phosphate to glyceraldehyde 3-phosphate during

glycolysis. TPI has been shown to be dependent upon GCR1 for full

expression; in a gcrl mutant strain TPI gene expression is reduced

17-fold (Clifton and Fraenkel, 1981). By mapping the elements of

the TPI controlling region, it should prove possible to map the DNA

sequence through which the GCR1 gene product acts and to determine

if GCR1 acts alone or in conjunction with other sites or factors. In

order to facilitate the identification of the site of GCR1 action in

the TPI controlling region, expression of a TPI::IacZ fusion will be

analyzed in isogeneic wild-type and gcrl mutant strains. The








5

identification of other trans acting factors required for the full

expression of TPI is also a primary concern. This introduction will

serve as a literature review of the studies to date concerning

glycolytic gene expression in Saccharomyces cerevisiae.

In addition to the GCR1 gene product, this study has shown that

expression of TPI requires the binding of a trans-acting protein

known as repressor activator protein 1 (RAP1) (Shore et al., 1987)/

general regulatory factor 1 (GRF1) (Buchman et al., 1988)/

translation upstream factor (TUF) (Huet et al., 1985), (Scott et al.,

1990). RAP1/GRF1/TUF will be referred to hereafter as RAP1.

RAP1 was first purified as a binding activity which bound the

HMR(E) silencer locus. RAP1 binding is required for repression of

transcription by the HMR(E) silencer (Shore and Nasmyth, 1987);

however, RAP1 binding is also required for the activation of

transcription for many genes (Huet et al., 1985; Capieaux et al.,

1989). Shore and Nasmyth (1987) demonstrated that two different

binding sites for RAP1 derived from two different UAS's are able to

restore HMR silencer function in vivo when substituted for the

native RAP1-silencer-binding site. When the native RAP1- silencer-

binding site was destroyed the silencer no longer functioned and a 1











transcript was produced, restoration of the native site or

substitution of the UAS derived RAP1-sites restored silencer

function and al transcript was no longer produced. These findings

indicate additional factors must play a role in RAP1 dependent

activation or repression.

RAP1 involvement in gene expression is widespread. RAP1

binds and is required for the function of the HMR and HML loci of the

mating type locus (Kimmerly et al., 1988; Hofmann et al., 1989),

genes encoding ribosomal proteins and other proteins of the basic

translational machinery (Huet et al., 1985), the H+-ATPase gene

(Capieaux et al., 1989), and the HIS4 gene (Devlin et al., 1991).

In addition to TPI, RAP1 binding has also been shown to be

necessary for the expression of several other genes encoding

glycolytic enzymes. Capieaux et al. (1989) were the first to note

near consensus RAP1-binding sites in the regulatory regions of GPD,

PGK, PYK, EN01, PDC, and ADH1. Subsequently, mutational analysis

has shown those sites to be essential for the expression of PGK

encoding phosphoglycerate kinase (Ogden et al., 1986; Chambers et

al., 1989), EN01 encoding an isozyme of enolase (Buchman et al.,

1988), PYK encoding pyruvate kinase (Buchman et al., 1988), ADH1








7

encoding alcohol dehydrogenase isozyme one (Buchman et al., 1988;

Santangelo and Tornow, 1990), and PDC1 encoding pyruvate

decarboxylase (Kellerman and Hollenberg, 1988). RAP1 binding has

also been reported to be required for the expression of EN02 (Brindle

et al., 1990). Recent studies with both temperature sensitive rap1

mutants and rap1s mutants, selected for defective silencing of the

HMR locus, have been able to separate the suppression and activation

functions of the RAP1 protein (Kurtz and Shore, 1991; Sussel and

Shore, 1991). However, both sets of mutations map to the carboxyl

terminus of the RAP1 protein and no clear domains have been

defined.

RAP1 is a phosphoprotein (Tsang et al., 1990). RAP1 binding

in vitro to the PYK controlling region has been shown to be reduced

by phosphatase treatment (Tsang et al., 1990). However, in addition

to phosphorylation, RAP1 binding in the in vitro binding assay was

also dependent upon binding site context (Tsang et al., 1990). The

significance of RAP1 phosphorylation state in vivo is unknown.

The controlling regions of PGK, EN01, EN02, and PYK have all

been mapped. In the case of PGK the upstream activating sequence

has been shown to be comprised of three main elements (Stanway et










al., 1989). The central portion of the UAS contains the activator

core (AC) to which RAP1 binding has been demonstrated (Chambers

et al., 1989). Upstream of the activator core is a region designated

Yfp, first recognized as a site of strong DNasel protection (Chambers

et al., 1988). The Yfp region binds ARS binding factor 1 (ABF1),

another trans-acting binding protein similar in properties to RAP1

(Chambers et al., 1990; Buchman and Kornberg, 1990). The third

region identified in the UAS of PGK consists of three repeats of a

pentamer sequence motif, CTTCC (Ogden et al., 1986). In

experiments with fragments from the PGK promoter which carry

both the RAP1 binding site and the CTTCC sequence motif, there is

DNasel protection of the RAP1-binding site and evidence of

heightened DNasel sensitivity at the CTTCC pentamer motif

(Chambers et al., 1988). However, no clear footprint was observed.

Chambers et al. (1988) suggested that the heightened DNase I

sensitivity at the CTTCC sequence element was due to interactions

with RAP1. Of the three main elements of the PGK UAS, no single

element is able to activate expression alone (Stanway et al., 1989).

The activator core with the Yfp is able to drive about one-third the

wild-type levels of expression. The activator core with the three











CTTCC pentamers has wild type levels of expression even without

the strong ABF1-binding site in Yfp (Ogden et al., 1986). Deletion of

the CTTCC pentamer from -432 to -428 reduces expression by 50%,

deletion of two of the repeats (-432 to -428 and -449 to -445)

reduces expression 75%, and removal of all three CTTCC motifs

reduces expression to less than 10% of wild-type (Chambers et al.,

1988). As seen with other glycolytic genes, full expression of PGK

depends on GCR1.

Both EN01 and EN02 have two upstream activating sequence

elements (Cohen et al., 1987). EN01 also contains an upstream

repressor sequence (URS) responsible for repression of EN01 in

nonfermentable carbon sources. The UAS1 of both EN01 (Buchman et

al., 1988; Brindle et al., 1990) and EN02 bind RAP1 (Brindle et al.,

1990). In the case of UAS1 of EN01 a CTTCC pentamer has been

identified. The pentamers presence has been shown to be required

for full UAS1 activity, but its absence does not affect the ability of

RAP1 to bind in vitro (Buchman et al., 1988). UAS2 of EN01 binds a

factor designated EBF1, Enolase Binding Factor 1. UAS2 of EN02

binds ABF1 and overlaps UAS1. In EN02 RAP1 binding to UAS1 and

ABF1 binding at the overlapping UAS2 may be competitive (Brindle et











al., 1990). The competitive nature of the interaction of RAP1 and

ABF1 at the UAS of EN02 is in contrast to the interactions that

occur between these factors at the UAS of PGK, where binding is not

thought to be competitive.

Expression of both enolase genes depends on GCR1 (Holland et

al., 1987), and, in the case of EN02, a small deletion in the area of

overlap of the two UAS's is able to relieve the requirement of GCR1

for full expression (Holland et al., 1990). However, when the UAS

elements of EN02 was moved immediately upstream of the TATA

boxes of EN01 or EN02 they were able to confer expression but

expression remained high in a gcrl mutant strain. Thus, Holland et

al. (1990) suggest there may be an effect due to the positioning of

the UAS elements close to the TATA box which can alleviate the

requirement for GCR1.

However, when UAS1 and UAS2 of EN02 (positions -491 to -

443 with respect to the start of the EN02 structural gene) were

cloned upstream of a TPI::IacZ gene fusion, replacing the native UAS

element, no expression was seen (J. Anderson and H. Baker, personal

communication). Furthermore, the fragment was unable to bind

RAP1 in vitro. These results directly contradict those reported by










Holland et al. (1990).

PYK has two UAS elements: UAS1 (-653 to -634), UAS2 (-811

to -714). PYK is also controlled by an URS (-468 to -344) to repress

expression during growth on nonfermentable carbon sources

(Nishizawa et al., 1989). Full expression of PYK requires GCR1

(Clifton and Fraenkel, 1981). The UAS1 binds RAP1 and has a CTTCC

pentamer that enhances UAS activity (Buchman et al., 1988).

Mutation of the CTTCC pentamer to CAACC reduced UAS activity by

16-fold. Deletion of the RAP1-binding site of UAS1 prevents

expression of PYK (McNeil et al., 1990). The pentamer is not

required for binding of RAP1 to UAS1 in vitro (Buchman et al., 1988).

The overall organization of UAS1 of PYK is very similar to the

UAS of PGK (Buchman et al., 1988; Chambers et al., 1990). Both PYK

and PGK have ABF1-binding sites adjacent to their RAP1-binding

sites and CTTCC sequence elements. The role of ABF1 binding in

expression of PYK has yet to be determined. In addition, another

factor known as REB1 (Morrow et al., 1989)/ QBP (Brandl and Struhl,

1990)/ Y (Fedor et al., 1988), hereafter called REB1, binds in the

UAS2 area of PYK (Chasman et al., 1990).

REB1 has been shown to affect nucleosome positioning in UAS











elements (Fedor et al., 1988; Chasman et al., 1990; Brandl and

Struhl, 1990), and, aids in gene activation in conjunction with

additional transcription factors (Chasman et al., 1990; Wang et al.,

1990). The REB1 gene has been cloned (Ju et al., 1990) and the

protein purified (Morrow et al., 1990). REB1 is a phosphorylated

polypeptide with an apparent molecular mass of 125,000 Daltons

(Morrow et al., 1990). REB1 shares homology with the DNA binding

domain of the oncogene myb, and chemical characteristics such as

hydrophilicity, abundant glutamines, and numerous hydroxyl-

containing amino acid residues in common with RAP1, ABF1, GCN4,

and HSF1 (Morrow et al., 1990).

A newly espoused function that may play a role in the

activation of PYK transcription is an "adaptor" or "modulator"

protein (Berger et al., 1990; Pugh and Tjian, 1990; Kelleher et al.,

1990; Liu and Green, 1990). An adaptor is postulated to act as a

bridge via protein-protein interactions between a positive trans

acting activator protein bound at UAS elements some distance

upstream of the basic promoter elements, and proteins such as TFIID

which bind the TATA box (Davison et al., 1983; Parker and Topol,

1984; Sawadogo and Roeder, 1985; Nakajima et al., 1988). The








13

interaction of the trans-activator and the adaptor is essential to aid

in the recruitment of TFIID or other basic transcription factors to

the TATA box. RAP1 dependent expression of PYK has been shown to

require functional GAL11/SPT13 (Nishizawa et al., 1990).

Nishizawa et al. suggest that GAL11/SPT13 was acting as an adaptor

protein in PYK expression. GAL11/SPT13 was first identified as a

gene required for full expression of some genes regulated by GAL4

(Nogi and Fukasawa, 1980), or as a suppressor for auxotrophic

mutants induced by Ty insertion (Fassler and Winston, 1988).

However, Nishizawa et al. note that Fassler and Winston (1989) did

not find decreased transcription of PYK1 in gal11/spt13 cells.

Nishizawa et al. suggest that the disparity in the observations may

be due to allele specific differences of the gal11/spt13 mutations.

The matter is by no means well resolved. Nishizawa et al. (1990)

showed that the requirement for GAL11 was alleviated by moving

the RAP1 binding site closer to the TATA element. Thus, the

requirement for the putative adaptor was alleviated by moving the

binding site for the second trans-activator closer to the TATA

element, supposedly allowing a direct interaction with the basic

transcription machinery such as TFIID.











ADH1 also has a UAS that binds RAP1 (Buchman et al., 1988).

Transcription of ADH1 is reduced in a gcrl strain (Santangelo and

Tornow, 1990). Santangelo and Tornow have reported that the RAP1

site of the ADH1 UAS was able to confer responsiveness to GCR1

when it replaced the UAS of LAC4 (Santangelo and Tornow, 1990).

Therefore, Santangelo and Tornow suggest that GCR1 acts through

the RAP1 binding site. Contrasting this observation is the PGK

controlling region where the activator core region to which RAP1

binds is insufficient to activate transcription (Stanway et al.,

1989). In addition, both EN01 and PYK UAS's require the CTTCC

pentamer as well as the RAP1-binding site for full UAS activity

(Buchman et al., 1988). This contradiction over RAP1 binding alone

being sufficient for activation of transcription has yet to be

resolved.

The exact mechanism by which RAP1 is able to activate

transcription of TPI, PGK, EN01, EN02, PYK, PDC1, and ADH1 has yet

to be determined. Expression of each of these genes also depends on

GCR1. The nature of the interactions of RAP1, ABF1, REB1 and GCR1

are also unknown. Some indications of the role of RAP1 binding may

be provided by studies that link RAP1 with DNA structure, both in








15

DNA loop formation (Hofmann et al., 1989) and interactions with the

nuclear scaffolding (Cardenas et al., 1990). The RAP1 protein also

binds yeast telomeric repeat sequences (Conrad et al., 1990), and

telomeres are shortened in a conditionally lethal (ts) rap1 mutant at

nonpermissive temperatures (Conrad et al., 1990). RAP1, ABF1, and

REB1 are abundant proteins (Buchman et al., 1988; Morrow et al.,

1989) and thus they may be more involved in a common mechanism

of transcription rather than specific regulation of any given gene.

The GCR1 gene product appears to be a specific regulatory

protein (or one of many) for the genes encoding glycolytic enzymes.

As with many specific regulatory proteins in yeast, such as GAL4

(Bram and Kornberg, 1982), GCR1 is expressed at low levels is S.

cerevisiae (Baker, 1986). If GCR1(or any other protein) provides for

the specific activation of the genes of glycolysis how does it exert

its effect in the context of RAP1, REB1, and ABF1 binding at UAS

elements?

Uemura and Fraenkel (1991) (Uemura and Fraenkel, 1991) have

recently isolated GCR2 which like GCR1 has a pleiotropic effect upon

the expression of most of the enzymes of glycolysis. The pattern of

affected enzymes in a gcr2 or a gcrl strain was quite similar with a











90-95% reduction seen in the expression of most glycolytic

enzymes. EN01 transcript levels were compared in GCR2 and gcr2

backgrounds. The defect in enzymatic activity seen in a gcr2 mutant

is mirrored by a reduction in EN01 mRNA. Interestingly, the gcr2

mutant exhibits only a partial growth defect on glucose where a

gcrl mutant has a severe growth defect on glucose. GCR2 has been

cloned by complementation but its sequence has not been reported.

How GCR2 will fit into the overall regulation of the expression of

the genes encoding glycolytic enzymes and how it relates to GCR1

remains to be seen.













MATERIALS AND METHODS


Strains

The strains of Saccharomyces cerevisiae and Escherichia coli

used in this study are shown in Table 1.



Media and Growth Conditions

Yeast cultures were grown in YP medium (Sherman et al.,

1983) supplemented with 2% glucose or 2% glycerol and 2% lactate.

Selection was carried out in YNB supplemented with the appropriate

carbon source and 0.0025 % histidine, 0.0025% leucine, 0 0025%

tryptophan,and 0.1% case amino acids. All yeast cultures were grown

at 300C. E. coli strains were grown in LB broth or minimal medium

63 supplemented with thiamine hydrochloride (1~g/ml), and amino

acids (25pg/ml) (Miller, 1972). Ampicillin was added to 100 pg/ml

for selection. All E. coli cultures were grown at 370C.











Table 1. STRAINS
Strain
E.coli
KK2186


MC1061


TG1


S. cerevisiae
S150-2B


HBY4


JF1052



DFY642


(aenntvna


Sourop Reference


supE, sbcB15, hsdR4,
rpsL, thi, A(lac-proAB)
F' [traD36, proAB+, laclq,
lacZAml 5]

hsdR, mcrB, araD139,
A(araABC-leu)7679
AlacX74, galU, galK,
rpsL, thi

supE, hsdA5,thi,
A(lac-proAB)
F' [traD36, proAB+,
laclq, lacZAM 15]


MATa, leu2-3,112,
his3A, trp 1-289, ura3-52

MA Ta, gcrlA::HIS3,
leu2-3, 112, his3A,
trp 1-289, ura3-52

MA Ta, leu2, ura3-52,
his4-917, lys2-1288,
spt13-20 (LEU2)

MA Ta, leu2-3,112,
ura3-52


(Zagursky and
Berman, 1984)


(Meissner et al.,
1987)




(Gibson, 1984)


(D. Shore)


(Scott et al.,
1990)


(J. Fassler)



(D. Fraenkel)


Pnntv,


G











Nucleic Acid Manipulations

Techniques used throughout this study are derived from the

standard reference manuals (Sambrook et al., 1989; Sherman et al.,

1983; Current Protocols in Molecular Biology, 1989) except for

deviations noted. S. cerevisiae DNA and RNA was prepared by the

methods of Sherman et al. (1983) and Struhl and Davis (1981),

respectively.



Generation of Double-Strand DNA Oliaonucleotides

Most double-stranded oligonucleotides were generated by a

modification of the method of Oliphant et al. (1986). Single-

stranded oligonucleotides were synthesized (University of Florida

Interdisciplinary Center for Biotechnology Research) with the

desired sequence flanked by restriction sites, typically a Hindill

site on the 5' end and a Sphl site on the 3' end. One restriction site

was made into an 8-10 bp palindrome by the addition of G and C

residues. The oligonucleotides were self-annealed via the

reinforced palindrome and the 3' ends extended with Klenow

fragment of DNA polymerase I in the presence of dNTPs. The double-

stranded extension products were then gel purified via











polyacrylamide gel electrophoresis (PAGE), digested with the

appropriated restriction enzymes, and cloned into the polylinker

region of pUC18 or other suitable plasmid vectors. This method was

used to generate the double-stranded form of oligonucleotides HB16,

HB19, HB21, HB22, HB24, HB25 listed in Table 2. Additional double-

stranded oligonucleotides were generated for cloning by annealing

complementary oligonucleotide pairs, also shown in Table 2.

Followed by digestion with the appropriate restriction enzymes, and

cloning.



DNA Sequencing

DNA sequencing was carried out by the dideoxy chain

elongation termination method of Sanger (1977) as modified by U.S.

Biochemicals to utilize the Sequenase enzyme. Sequencing primers

were synthesized by the University of Florida Interdisciplinary

Center for Biotechnology Research and are listed in Table 2.
















Table 2. OLIGONUCLEOTIDES
Name Sequence
Forward 5'-GTAAAACGACGGCCACT-3'

HB01 5-ATGTGTGGAATTGTGAGCGG3'

HB03 5-CAAGCTTGTCGACAAGCTTG-3'

HB05 -CCACCGACAAAGAAAGTTCTAGCC-3'

HB06 5'-ACATGCATGCATGT-3'

HB07 CAAGCTGTTCTAAAT CAGCTTOCTCTATTGATGTTACATGGACGCATGC-3'
HB08 S5GCATGCTC GTGTAATCAATAGAGGAACT TCTTA CAA TG

HB09 S-GGCATGCCAACATGTATGGGTTCCAAGCTTG3'

HB 11 GGCATGCCAACCTGATGGGTTCCAAGCTTG3'

HB14 SGGAA TA CCTCTATTGATGACACCTTTTCT TCCAGGCAT
HB15 SGCATGAACTGATGCCAAAA TGTCCAGGTGTAACA

HB16 SGAAGC T ACTTTTC TTCCTCTAGATGTACACCTAA
GTTGCATGCC-3'
H B 17 S3CTG CTGATGAGAAAAGGGTGCCA GTAACATCAAC
ACAAGCTTCC-3'

HB19 STAAGCTrA TCTATGATGAACCT A TCAGGTGCATCCAGTT

HB21 5GGGGCTGCAACAAAA TGTCC TGTAACATAATG G

HB22 'S~GGCTCATCCAGGTGTAACATCAATTA

HB23 5'-CGCAAACCGCCTCTCC-3'

HB24 S-TAA CTTC ATGATGTCAC GATATCTGCAGTGGCATCCAGG

H B25 SGCTAACTACAACTCTATTGATGTACACCTGGAC CTTTCTGGC CTGTTGCATGC

HB28 S-GTACACCTGGAGATATCTGCAGTGGCATCCAG3'

HB29 5'-GACTTTTCAGCAACCTCTATT-3'

HB30 '-CCTTTTCTGGCAACAGTTTrAATC-3'

HB31 5-GGAAGATTGAACGTTCTAAG-3'

HB32 S-GGAACCCATACATGTTGGTGGAAG3'

HB33 S-GTAACAGGGAAGGAAAGGCAGC3'

HB42 5-CTGTGAGGACC-3'

HB53 S-GATGTTACACCAGATTACCCGTTCTCTGGCATCCAG-'

HB54 5-GACCTTAATACATTCAG-3'

HB60 5-GCATTAGCATGCGTAACAAACCACC-3











Primer extension

Total RNA was isolated from yeast strains S150-2B and HBY4

and used to determine the 5' end of the TPI transcripts. Total RNA,

101ig from wild-type and 251g from gcrl-deletion mutant, was

annealed with a radiolabeled primer HB05 which hybridized near the

5' end of the TPI structural gene. The primer was extended with 10

units of reverse transcriptase in the presence of dNTPs at 3mM. The

extension products were analyzed by denaturing PAGE. A sequencing

ladder generated from the 5' noncoding region of TPI with primer

HB05 was used as a molecular weight standard.



Plasmid Construction

Constructs for sequencing Single-stranded templates of the

5' noncoding region of TPI were generated for sequencing in the

following manner. M13mp18 and M13mp19 RF II double-stranded

DNA was prepared. Portions of the 5' noncoding region of TPI were

subcloned from plasmid pHB51 into M13mp18 or M13mp19 RF II DNA.

Single-stranded templates were prepared by standard techniques

(Sambrook et al., 1989), utilizing E. coli strain KK2186. Templates

were sequenced as above, utilizing the forward primer.








23

5' deletion scheme The scheme to introduce the Ba/31-induced

deletions in the 5' noncoding region of TPI is shown in Figure 1.

Plasmid pHB110 contains 3.5 kb of additional yeast DNA 5' to the

start of the TPI structural gene fused in frame to lacZ. The fusion

construct was linearized with Tth1111 at a unique site 853 bp

upstream of the structural gene. The linearized material was then

treated for various times with exonuclease Ba131. Following Ba131

treatment, Klenow fragment was used to fill in the ends. Hindlll-

Sali-Hindlll linkers (CAAGCTTGTCGACAAGCTTG, HB03) were added.

The material was then digested with Sail which cuts within the

linker and within the polycloning site of pHB110. This digest

removes all yeast derived DNA 5' to the deletion endpoint. The

desired fragments, containing the TPI::IacZ fusion and the remaining

5' noncoding sequence, were gel purified, ligated, and used to

transform E. coli MC1061. Plasmid DNA was prepared from

individual transformants, and the precise deletion endpoints were

determined by double-stranded DNA sequencing using primer HB01.

Once the desired constructs were identified, they were

subcloned on Hindlll fragments into the Hindlll site of Ylp56. The

orientation of the fusion with respect to the URA3 selectable
























Figure 1. Scheme to generate 5' deletions in the TPI 5' noncoding
region. The TPI::lacZ fusion plasmid pHB110 contains a unique
T th 111 site at position -853 with respect to the start of
translation. The plasmid was linearized with Tth111 and treated
with exonuclease Bal31 for various times. HindllI-Sall-Hindlll
linkers (HB03, table 2) were added. The material was then digested
with Sa/I which cuts once in the polycloning region of pHB110 and in
the linker. After gel purification of the vector band, the material
was recircularized via ligation and used to transform E.coli.
Plasmid DNA was prepared from the transformants and precise
deletion endpoints were determined by DNA sequence analysis.
Restriction sites are as follows: E,EcoRI. S, Sail. H, Hindlll. T,
Tth 111.

















I Tth1111 digest
Bal 31 digest
I I ~


T

SBal31


if


PI g~ga1









26

marker was determined by restriction endonuclease analysis with

BamHI.

Internal deletion scheme Plasmid pES35 contains the TPI::IacZ

fusion with 853 bp of 5' DNA on a Hindlll fragment subcloned into

Ylp56 at Hindlll. The Sphl restriction site at position -220 with

respect to the start of the structural gene was a convenient origin

for exonuclease Bal31 digestion. Therefore, the Sphl site at -220

was rendered unique by a Kpnl dropout of pES35, which removed a

second Sphl site in the polylinker region of pES35, generating

plasmid pES90. The scheme to introduce Ba/31-induced deletions

into pES90 is shown in Figure 2. Plasmid pES90 was linearized with

Sphl and treated for various times with exonuclease BAL31. The

Klenow fragment of DNA polymerase I was used to fill in the ends of

the remaining material. Sphl linkers (HB06) were added. The

material was then digested with Sphl and Hindlll to liberate DNA

distal to the Sphl site at -220. The Hindlll-Sphl fragment that

corresponds to the Ba131 treated DNA of yeast origin distal to the

Sphl site was gel purified and subcloned into pES90 which had been

digested with Hindlll and Sphl. The resulting material was

transformed into E. coli MC1061. Plasmid DNA was prepared, and


























Figure 2. Scheme to create internal deletions. The TPI::lacZ fusion construct pES90 contains a unique
Sphl restriction site at position -220. Plasmid pES90 was linearized with Sphl and digested for
various times with exonuclease Ba131. Sphl linkers were ligated on and the material was digested
with Hindlll and Sphl. The material remaining 5' to position -220 was isolated by gel purification and
subcloned into pES90 digested with Hindlll and Sphl. Deletion endpoints were determined by
sequencing with primer HB05.













Sph I
-220


1 AS? I I UAS I


TATA


ITPI::IacZ


1) Open at Sph I

2) Delete with Bal31

3) Add Sph I Linkers


Isolate Hind III/ Sphl Fragment
Subclone to Sphl at -220 (wt.)


Sph I
-220


tfTATA


ITPI::acZ I


Hind III
-853


Hind III
-853


ASn


I V--V


I -~c~--








29

precise deletion endpoints were determined by double-stranded DNA

sequencing using primer HB05.

Constructs to assess UAS activity Plasmid plCZ312 is an

integrative yeast shuttle vector that contains a CYC1::IacZ fusion

(Guarente and Mason, 1983). The native CYC1 UAS elements were

removed by Hindlll-Sphl digest and replaced with various fragments

containing portions of the TPI 5' noncoding region. All constructs

were confirmed by sequencing, using primer HB23 which anneals in

the vector sequence of plCZ312 immediately upstream of the

polycloning site. Plasmid pES90 was digested with Hindlll and Sphl

to remove sequences distal to position -220 of the TPI::IacZ fusion.

Such a digest removes sequence required for expression of the

fusion. Fragments containing portions of the TPI 5' noncoding region

distal to -220 were subsequently subcloned and tested for UAS

activity. Constructs were confirmed by sequencing with primer

HB01.

Mutant UAS oligonucleotides driving expression of the TPI::IacZ

fusion Double-stranded oligonucleotides capable of UASTPI

function, or mutant derivatives thereof, were generated via mutually

primed synthesis. These double-stranded oligonucleotides contained











a Hindlll site at their 5' end, and a Sphl site at their 3' end. The

oligonucleotides were isolated from pUC18 by Hindlll-Sphl digest

and subcloned into the HindllI-Sphl sites of pES90. Thus, the

oligonucleotides replaced the native UASTPI element in pES90. The

replacement of UASTPI in the resulting plasmids was confirmed by

DNA sequence analysis utilizing primer HB01.



Site-Directed Mutagenesis

The 1.4 kb Smal fragment of plasmid pES90, containing the 5'

noncoding region of TPI, was cloned into the Smal site of M13mp18.

The ligation mixture was used to transform E. coli TG-1 and lysates

were prepared from 10 individual plaques. RF II DNA was prepared

from each isolate and the orientation of the insert determined by

Hindlll restriction analysis. One phage, MES-TPImp18B, contains the

strand of 5' noncoding region that corresponds to the antisense

strand of the TPI structural gene, and was used as the target for

site-directed mutagenesis. Oligonucleotides with the desired

mutations were synthesized such that the area of mismatch was

flanked by regions of complementarity of 10 to 12 nucleotides.

These oligonucleotides, HB28-33, are listed in Table 2.








31

Materials utilized to introduce the mutations were supplied by

a kit available from Amersham. The kit is based on the method of

Eckstein and co-workers (Taylor et al., 1985). The method provides

strand-specific selection based on the inability of the restriction

enzyme Ncil to cleave DNA containing thionucleotides. The mutant

oligonucleotide was annealed to the target, extended with the large

fragment of DNA polymerase I in the presence of the thionucleotide

(dCTPaS).

The resulting material was ligated with T4 DNA ligase to form

complete and unnicked double-stranded plasmid. Excess single-

standed target template was removed via passing through a filter

which binds single, but not double-stranded DNA. Then the parental

strand was specifically nicked with Ncil, which did not cut the

thionucleotide containing mutant daughter strand. The nicked

material was then partially digested with exonuclease III. The

intact daughter strand was then used as a template with DNA

polymerase I and circularized with T4 DNA ligase.

The new double-stranded construct contained the desired

mutation on both strands. The material was used to transform E.

coli strain TG-1, and lysates were prepared from 10 individual











plaques. Introduction of the desired mutation was confirmed by

sequencing with primer HB42. RF II DNA was prepared from desired

mutant phage and the Hindlll-Sphl fragment was subcloned into

pES90 to form the mutant constructs. All pES90 derived mutant

constructs were also confirmed by sequencing with primer HB42

prior to integration and assay.



Transformation

The method of Enea et al. (1975) was used to transform E. coli.

Saccharomyces cerevisiae was transformed by the method of Ito et

al. (1983), selecting for uracil prototropy. Integration of the

TPI::IacZ fusion constructs at the URA3 locus was achieved by

linearizing the plasmids with Stul which cuts at a unique site

within URA3. The scheme for integration is shown in Figure 3.



Screen for Unit Copy Integrants of TPI::IacZ fusions at URA3

Integrative transformation of yeast cells can lead to tandem

insertional events (Orr-Weaver et al., 1983), Figure 3. To screen for

unit copy integrants, the genomic structure of the URA3 locus was
























Figure 3. Scheme for integration of TPI::lacZ gene fusion
constructs at URA3. A) depicts a plasmid containing a TPI::IacZ
fusion construct and a URA3 selectable marker juxtaposed above the
ura3-52 locus of the yeast genome. The plasmid is linearized within
URA3 to direct integration via homologous recombination. Sad sites
are indicated. B) represents unit copy integrant at ura3-52.
Digestion with Sad gives rise to three fragments A, B, and C. C)
represents a tandem integration event. Upon digestion with Sad,
five fragments are generated: A, 2B, C, and a fourth fragment, D,
which is diagnostic of a tandem insertion event.

















Sac I


Sac I Sac I Sac I Sac I


URA3 TPI' 'lacZ 'TPI ura3
--------------------------------------.. ----------- ..... ....... ....... -c --.............

A. B. C.



Sac I Sac I Sac I Sac I Sac I Sac I


URA3 TPI' 'lacZ TPI URA3 TPI' 'lacZ TPI ura3

A. B. D. B. C.
A. B.D. B. C.


Sac I








35

determined by Southern blot hybridization analysis. Genomic DNA

was isolated from individual transformants. The DNA was then

digested with Sad, run on a 0.8% agarose 1X TBE gel, and capillary

blotted to a Gene Screen nylon membrane according to the

manufactures instructions (DuPont). The filter was hybridized with

a Ylp56 probe generated by random-primer extension. Hybridizations

were carried out at 420C for 16 hours in 50% formamide, 0.2% BSA,

0.2% polyvinyl-pyrrolidone (M.W. 40,000), 0.2% ficoll (M.W. 400,000),

50mM Tris pH 7.5, 0.1% Na-pyrophosphate, 1% SDS. Hybridized

membranes were washed according to manufactures instructions.

Visualization was via autoradiography.



5-Galactosidase Assays

Strains to be assayed were grown from a single colony to an

A6oo between 0.5 and 1.5 at 300C. P-galactosidase assays were

performed essentially by the method of Miller (1972). The units

reported correspond to AA420/minute/A6oo of the initial culture.











In vitro DNasel Protection Assays

The method used in the in vitro DNasel protection assays was a

modification of the method of Singh et al. (1986). The 228 bp Hindlll-

Sphl fragment of pES40-23 or the 169 bp Hindll-Sphl fragment of

pES34 were end-labeled by filling in the Hindlll site with the

Klenow fragment in the presence of 32P-dATP. A protein extract

was prepared from yeast S150-2B by lysis with a French pressure

cell at 20,000 psi. Rabbit reticulocyte lysates (RRL) containing

RAP1 protein were generated via in vitro transcription and

translation, and provided to me by C. Lopez.

Multiple aliquots of 2p1 of the end-labeled fragments (20,000

cpm/ul) were incubated with either 5g1 of yeast extract or 5!1 RRL

containing RAP1 in 181il of 1X binding buffer [12mM HEPES pH7.5,

60mM KCI, 5mM MgCI2, 4mM Tris, 0.6mM DTT, 10% glycerol, 0.26

ug/ul poly(dl-dC), and 0.3 pg/Lpl BSA]. Incubation was for 20 minutes

at room temperature. 0.5 units or 1.0 units of DNase I was then

added to quadruplicate aliquots. The reactions were incubated at

room temperature for 2 minutes, then stopped by the addition of

10l of Stop solution (0.25M EDTA, 25% glycerol) on ice. The entire











reaction was then loaded onto a 5% polyacrylamide, 0.5X TBE gel

running at 5 volts/cm. The gel had been pre-run for at least 1 hour.

The samples were electrophoresed at 7.5 volts/cm until the

bromphenol blue tracking dye, loaded along with the samples, was 2

cm from the bottom of the gel. The wet gel was wrapped in plastic

film and exposed to X-ray film overnight at 40C.

The developed film was used as a guide to excise the shifted

and unshifted portions of the radiolabeled fragment from the gel.

Identical positions from the quadruplicate samples were pooled for

extraction from the gel slice. Fragments were extracted from the

gel fragments by incubation at 370C overnight in 3 volumes of 0.5M

NH4Ac, 1mM EDTA. The extracted fragments were precipitated by

the addition of 2.5 volumes ethanol and centrifugation. The pelleted

fragments were washed with 70% ethanol and dried. The dried

pellets were Cherenkov counted and resuspended to 4,000 cpm/pl in

formamide sequencing dye.

A control DNase I ladder was generated from the unprotected

fragments by digestion for 2 minutes with 0.02 units of DNase I.

The control ladder was pelleted, counted, and resuspended in the

same manner as the sample fragments. A sequencing ladder was











generated from M13mp18 with the forward sequencing primer to

serve as a molecular weight standard.

The shifted sample fragments, bracketed by the unshifted and

control DNasel ladder, were electrophoresed next to the sequencing

standard on a 0.4mm-7% polyacrylamide, 8M urea, 0.5X TBE gel at 50

volts/cm. After electrophoresis, the gel was blotted to Whatman 3M

paper and dried. The dried gel was exposed to X-ray film for

visualization.



DNA Band Shift Assays

DNA band shift assays were performed as previously described

(Scott et al., 1990). E. coli extracts were prepared by lysis with a

French pressure cell at 20,000 psi.



In vivo Methylation Protection Analysis

Yeast strains were cultured in 2 liters of YP medium (Sherman

et al., 1983) supplemented with either 2% glucose or 2% lactate and

2% glycerol. Cultures were harvested at an A600oo of 1.0 by

centrifugation. Cells were washed twice in 137mM NaCI,2.7mM KCI,

4.3mM NaPO4, 1.4mM KPO4 pH7.4 (PBS), and concentrated to 1x 108








39

cells per ml in PBS (35ml final volume). 5ml aliquots were placed in

50ml disposable, sealable plastic tubes on ice. Dimethyl sulfate

(DMS) was added to 0.5% final concentration to various aliquots (5

gl/aliquot), and incubated from 1-6 minutes at room temperature.

Reactions were quenched by the addition of 45ml of ice-cold PBS.

Treated cells were harvested by centrifugation (1,000x g for 10

min.) and washed twice in 35ml ice-cold PBS. DMS waste was

allowed to decay in a fume hood for one week prior to disposal.

Methylated genomic DNA was prepared essentially by the

method of Sherman et al. (1983), except all incubations were carried

out at 370C or less to prevent unwanted cleavage events. The DNA

preparation was treated with RNase A at a final concentration of

50pg/ml for one hour. Two phenol extractions, one

phenol/chloroform extraction, and one chloroform extraction were

performed. 1/10 volume of 5M NH4Ac was added, followed by

precipitation with 2 volumes of ethanol. DNA was pelleted by a low

speed spin (500x g, 5 min.), washed twice with 70% ethanol, and

dried in vacuo. The dried pellets were resuspended in 250 p1 10mM

Tris pH 8.0, 1.0 mM EDTA (TE); and the A260/280 was determined.








40

Control DNA was prepared simultaneously from an untreated aliquot

of the original cells.

501ig of methylated DNA from each time point and 2001ig of

untreated DNA was cleaved to completion with Avail. 1/10 volume

5M NH4Ac was added, followed by 2.5 volumes ethanol. Samples

were centrifuged 10,000x g for 30 minutes. The pellet was washed

twice with 70% ethanol and dried in vacuo. Methylated DNA was

resuspended in 2001ll 1M piperidine and incubated at 950C for 30

minutes. 40g1l aliquots were removed for alkaline agarose gel

electrophoresis. 16il of 3M NaAc pH 6.0 was added to remaining

sample, followed by precipitation with 2.5 volumes ethanol.

Samples were centrifuged 10,000x g for 30 minutes, washed twice

with 70% ethanol, and dried in vacuo. 501i of TE pH 8.0 was used to

resuspend samples, and the concentration of the was determined

spectrophotometrically.

The aliquot of the piperidine cleavage reaction was ethanol

precipitated and resuspended in formamide sequencing dye. Samples

from the various timepoints were denatured at 950C and loaded onto

a 1.5% Agarose, 50mM NaOH, 1.0mM EDTA gel next to molecular











weight markers. The gel was electrophoresed in a 50mM NaOH,

1.0mM EDTA running buffer at 30 volts and 175 milliamps overnight.

The gel was neutralized in 1M Tris pH 7.5 for one hour, stained with

0.5ug/ml ethidium bromide, and visualized with UV light. Samples

generated a smear with a definite peak size of fragment. Time

points with a peak fragment size of approximately 500-800 bp were

used for subsequent steps.

Control sequences were generated from the 200p.g of Avail

digested untreated DNA. The DNA was divided into four aliquots and

ethanol precipitated, washed twice with 70% ethanol, and dried.

Control G-reactions were as follows. One aliquot was resuspended

in 200L1 DMS buffer (50mM Na-cacodylate, 1.0mM EDTA pH 8.0). 11l

of DMS was added for 1.5 minutes at room temp. 50pl ice-cold DMS

stop solution was added (1.5M NaAc pH 7.0, 1M 2-mercaptoethanol),

and 750pl ice-cold ethanol was added. The reaction mixture was

then placed on dry ice.

The control A+G-reactions were as follows. The DNA aliquot

was resuspended in 11 l ddH2O. 25pl of concentrated formic acid

was added at room temperature and incubated 5 minutes. 200pl ice-











cold hydrazine stop solution was added (0.3M NaAc pH 7.5, 0.1mM

EDTA). 750g1l of ethanol was added, and the mixture placed on dry

ice.

The control C-reactions were as follows. The DNA aliquot was

resuspended in 5il ddH20, and 15pl of 5M NaCI added. 301l hydrazine

was added and incubated 16 minutes. The reaction was stopped by

the addition of 200pl ice-cold hydrazine stop solution. 750 l of

ethanol was added, and the mixture was placed on dry ice.

The control T+C-reactions were as follows. The DNA aliquot

was resuspended in 20pl ddH20. 30 l of hydrazine was added and

incubated for 16 minutes. 200L1 ice-cold hydrazine stop solution

was added. 7501l of ethanol was added, and the mixture placed on

dry ice.

All reactions were processed as follows. Samples were

centrifuged at 10,000x g for 15 minutes, and pellets washed with

70% ethanol. (Hydrazine waste was neutralized by placing in a

saturated ferric chloride/water solution.) Pellets were resuspended

in 225gl ddH20, 25p1l of 3M NaAc pH 6.0 and 75011 of ethanol was

added. Following centrifugation at 10,000x g for 30 minutes,











pellets were washed twice with 70% ethanol, and dried. Pellets

were dissolved in 1001l of 1M piperidine and incubated at 950C for

30 minutes. 11 i of 3M NaAc pH 6.0, and 230p1 ethanol was added.

Samples were centrifuged 30 minutes, washed twice with 70%

ethanol and dried in vacuo overnight. Pellets were dissolved in

1001pl TE pH 8.0 and the concentration of the material was

determined spectrophotometrically.

51.g of all sample and control DNA's were lyophilized,

resuspended in 4]l formamide sequencing dye, and denatured at 950C.

A genomic sequencing ladder followed by the in vivo methylated

samples bracketed with G-reaction control ladders were

electrophoresed in a 60cm, 7% (40:1.3) polyacrylamide, 8M urea gel

in 0.5X TBE running buffer at 3500 volts and 35 milliamps. After

running, the gel is lifted using Whatman 541-sfc paper, and

electroblotted to a Hybond N+ membrane (Amersham). The DNA was

UV cross-linked to the membrane with a FisherBiotech 312 nm

variable intensity transilluminator at full power for 5 minutes. The

membrane was prehybridized in 20ml of hybridization buffer [1.0%

bovine serum albumen, 7.0% SDS, 1.0mM EDTA, (0.5M Na) HPO4 pH 7.2]

at 630C in a roller incubator.











The TPI specific probe was prepared as follows. Single-

stranded M13 phage containing the sense (mES2-2) or anti-sense

(mESTPImp18B) strand of TPI, spanning the 5' non-coding region and

the beginning of the structural gene, was prepared by standard

techniques (Current Protocols in Molecular Biology). 6111 of phage

(0.25p~g/gl) was incubated with 5.l of primer(0.5pM/pl), and 2.5.l of

10X klenow buffer (0.5M Tris pH 8.0, 2M NaCI), at 500C for 30

minutes. The sense probe was prepared by annealing primer HB54 to

mESTPImpl8B. The antisense probe was made by annealing primer

HB60 to mES2-2. The following reagents were then added in order:

511 of 0.1M DTT, 5gl of 50mM MgCI2, 2pl of 3.0mM dNTP-dATP mix,

10 units of DNA polymerase I-large subunit, and 1011 of 32P- dATP

at 3,000 Ci/mM. The antisense reaction also included 2 1l of Avail to

cleave the probe to the appropriate length. The reaction was

incubated at 370C for 45 minutes. 120.l of formamide sequencing

dye was then added. The probe was denatured at 950C for 10

minutes and run into a 6% (40:1.3) polyacrylamide, 8M urea gel. The

wet gel was exposed to Polaroid type 57 film for 15 minutes. The

film was developed and used as a guide to excise the probe from the








45

gel. The gel slice was crushed and 8ml of hybridization buffer was

added.

The pre-hybridization solution was removed from the

membrane and the gel/hybridization solution mixture added to the

roller tube. Hybridization was performed at 630C overnight.

Following hybridization, the membrane was washed in 500ml

aliquots of (40mM Na) HPO4, 1.0mM EDTA, 1% SDS at 600C until wash

was no longer radioactive when examined with a geiger counter. The

damp membrane was wrapped in plastic wrap and exposed to film for

visualization.













RESULTS


The Mature 5' Ends of Steady-State TPI Transcripts are Unaffected

by a gcrl Mutation

Previous studies suggested that gcrl lesions bring about a

reduction in the levels of mRNAs specifying glycolytic enzymes

(Clifton and Fraenkel, 1981; Holland et al., 1987). An initial

objective was to characterize the TPI transcript in both wild-type

and gcrl mutant strains of Saccharomyces cerevisiae. RNA gel

transfer hybridization experiments in both strains demonstrated a

reduced steady-state level of the TPI transcript in a gcrl mutant

strain (Scott et al., 1990). The reduction in steady-state levels of

the transcripts suggests that the GCR1 gene product may play a role

in the transcriptional regulation of the genes encoding glycolytic

enzymes.

To further characterize the TPI transcript, it was of interest

to determined if the mature 5' end(s) of the transcripts were

affected by deletion of GCR1. Differences in the mature 5' end of the

46








47

TPI transcript isolated from wild-type and gcrl mutant strains may

reflect altered transcriptional start sites. Such a result may

indicate the existence of two promoter elements, one of which is

GCR1-dependent and one of which is GCR1-independent. The TPI

transcript remaining in the gcrl mutant could originate from a GCR1-

independent promoter and initiate at a unique start site. A GCR1

independent TPI transcript may be the product of a different

transcriptional mechanism than the abundant transcript in wild-type

strains. Transcripts of a single gene derived from independent

mechanisms are known to exist for HIS4 which has both inducible

and basal transcripts initiating from independent sites (Arndt et al.,

1987; Pellman et al., 1990).

Mapping the mature 5' end of the TPI transcript was

accomplished using primer extension analysis. 10glg of wild-type

(S150-2B) or 25ig of gcrl mutant (HBY4) total RNA was annealed

with a radiolabeled oligonucleotide primer (HB05) corresponding to

the start of the TPI structural gene. Extension reactions were

carried out with reverse transcriptase in the presence of

deoxyribonucleotide triphosphates. The first base incorporated

corresponded to position +2 with respect to the adenine of the








48

initiation codon. Extension products were resolved on a denaturing

polyacrylamide gel next to a DNA sequence of the TPI 5'

nontranslated region which was generated with the same primer

used in the extension reactions. Figure 4 shows the results of the

primer extension experiment. In both the wild-type and gcrl mutant

strain the predominant mature 5' ends of the TPI transcript were

identical. The ends corresponded to a pair of adenines at positions

-29 to -30 with respect to the initiation codon. Therefore, it

appears likely that TPI transcription is controlled by a single

promoter element, and that the residual expression of TPI observed

in gcrl mutant strains is the result of transcription originating at

the native start site.



Identification of the 5' Boundary of the TPI Controlling Region

In order to map the controlling elements) of TPI a TPI::IacZ

gene fusion was utilized. This gene fusion produced a protein that

was a hybrid between triose-phosphate isomerase and 3-

galactosidase, which retained P-galactosidase activity. Use of the

fusion in trans to TPI allowed normal expression of TPI which is

required for cellular growth while allowing the manipulation of the
























Figure 4. Primer extension analysis of the 5' ends of the TPI
transcript in wild-type and gcrl-deletion mutant strains. (A) Lane
1, extension products generated with reverse transcriptase from
10pg of total RNA isolated from wild-type strain S150-2B
hybridized with a radiolabeled primer corresponding to the start of
the TPI structural gene. Lane 2, extension products from similar
reaction carried out with 25p1g total RNA from gcrl mutant strain
HBY4. Lanes 3 through 6, DNA sequencing ladder generated with the
aforementioned primer from the TPI 5' nontranslated region. Bold
double arrow indicates the major extension products. P, denotes the
position of the unextended primer. (B) DNA sequence of the TPI 5'
nontranslated region. Double arrows indicate the bases that
correspond to the predominant 5' ends of the TPI transcript.



























C
T
T
G
C
T
T
A
A
A
T
C
T
A
T
-rn-A
----A
C
T
A
C
A
A
A
A
A
A
C
A
C
A
T
A
C
A
T
A
A
A
C
T
A
A
A
A








51

TPI::lacZ fusion. The effect of the manipulations on the expression

of TPI::IacZ could be measured indirectly by monitoring (-

galactosidase activity.

The plasmid pHB110, a derivative of pUC18, harbors the

TPI::IacZ gene fusion with 3.5 kilobase pairs (kbp) of DNA 5' to the

start of TPI. 3.5 kbp. of DNA 5' to TPI was initially included in order

to ensure that all cis-acting regulatory elements necessary for TPI

expression were included in the construct. Sequencing of the 5'

non-coding region identified a unique Tth1111 site at position -853,

Figure 5. Subcloning experiments showed that all sequences

sufficient for high-level expression of the fusion resided within the

Tth1111 site. Therefore, the site could be used as an origin for a

series of nested deletions created through the action of exonuclease

Bal31. Figure 1 shows the scheme used to create the Bal31

deletions from plasmid pHB110. These deletions were used to map

the 5' boundary of the region sufficient for high-level expression of

the TPI::IacZ gene fusion.

The fusion construct was linearized with Tth 1111 and treated

for various times with exonuclease Bal31. Hindlll/SaIl/Hindlll

linkers (HB03, table 2) were then ligated to the material. This









FIGURE 5. Sequence of TPI 5' Noncoding Region 52

S-1191
acgtcatcgatgaatataatgaattaaacagtggtgttcgtatatgtgaagatatgagatatga

tccacatggtaaacagaaagatgcattttggccgagaggacttaataatactggtggtgtttac

gaaaataatgaagataatatttgtgaagggaagcctggaaaatgttatctgcaatatcgggtta

aggatgagccaagaataagggaacaagattttggtaatttccaaaaaatcaatagcatgcagg

acgttatgaagaagagatctacgtatggtcatttcttcttcagattccctcatggagaaagtgc
S-853
ggcagatgtatataacaaagtcgccagtttccaagagactttattcaggcacttccatgatagg

caagagagaagacccagagatgttgttgtcctagttacacatggtatttattccagagtattcc

tgatgaatggtttagatggacatacgaagagtttgaatcgtttaccaatgttcctaacgggagc
1 -658 (Previously Known Sequence)
gtaatggtgatggaactggacgaatccatcaatagatacgtcctgaggaccgtgctacccaaa

tggactgattgtgagggagacctaactacatagtgtttaagattacggatatttaacttactta

gaataatgccatttttttgagttataataatcctacgttagtgtgagcgggatttaaactgtgag

gaccttaatacattcagacacttctgacggtatcaccctacttattcccttcgagattatatct

aggaacccatcaggttggtg g aagattacccgttctaagacttttcagcttcctctattgatgt

tacacctggacaccccttttctggcatccagtttttaatcttcagtggcatgtgagattctccg

aaattaattaaagcaatcacacaattctctcggataccacctcggttgaaactgacaggtggtt
S-220
tgttacacatactaatgcaaaggagcctatatacctttggctcggctgctgtaacagggaatat

aaagggcagcataatttaggagtttagtgaacttgcaacatttactattttcccttcttacgta

aatatttttctttttaattctaaatcaatctttttcaattttttgtttgtattcttttcttgcttaa
S+1
atctataactacaaaaaacacatacataaactaaaaATG








53

linker was designed such that upon subsequent digestion with Sail

and religation of the vector each deletion endpoint would be marked

with a Hindlll site. Precise deletion endpoints were determined for

65 individual constructs by using the dideoxy sequencing method of

Sanger (1977) as modified by U.S. Biochemicals. Once the 5' deletion

series was obtained it was necessary to place the constructs back

into yeast to determine the effects of the deletions upon expression

of the fusion. Of the 65 plasmids sequenced 13 were chosen because

they provided a well spaced, nested set of 5' deletions for study.

The constructs were subcloned into the yeast integrative

plasmid 56 (Yip56) which contains a URA3 selectable marker for

yeast. Yip56 also contains an origin of replication and an ampicillin

resistance determinant for propagation and selection in E. coli. An

integrative yeast shuttle vector was chosen to avoid effects on

fusion expression due to plasmid copy number discrepancies brought

on by the high segregation rate of yeast plasmids (Botstein and

Davis, 1982). The isogeneic uracil auxotrophs S150-2B and HBY4

were transformed to uracil prototropy with the deletion constructs.

Integration of the fusions was directed to the URA3 locus by

transforming with plasmid DNA linearized with Stul which cuts









54

within URA3. This procedure served to direct the site of integration

to the URA3 locus via homologous recombination (Orr-Weaver et al.,

1983). Transformation competent yeast cells can take up multiple

copies of plasmid DNA, raising the possibility that tandem

integration events can occur (see Figure 3). Therefore, screening for

unit copy integrants was required in order to assure the most

accurate expression data possible for the 5' deletion series.

Genomic DNA was isolated from individual transformants and

digested with Sad, then subjected to Southern blot analysis probing

for URA3. Sacl cut outside of the URA3 locus but within the plasmid.

The URA3 probe hybridized to a single band in experiments with DNA

isolated from the parental strain, S150-2B. A unit copy integrant

was distinguishable by the presence of two fragments which

hybridized to the probe, labeled A and C in Figure 3. These are the

two junction fragments between the yeast chromosome and the

integrated plasmid. Upon a tandem integration a characteristic

band, totally plasmid derived, was also observed, labeled fragment D

in Figure 3. Presence of the totally plasmid derived fragment D was

used as a diagnostic for multiple integration events. Utilizing this











screen, unit copy integrants were obtained for all constructs in

S150-2B (GCR1) and all but two constructs in HBY4 (gcrl).

The stability of the fusion constructs integrated at URA3 was

determined by growing a strain carrying an integrated construct in

non-selective media (YPD) for >10 generations. The cells were then

plated on non-selective media and 100 individual colonies were

screened for their ability to grow on selective media (YNB with Glu,

His Leu, Trp). Each colony tested was URA+. Furthermore, each

colony still expressed p-galactosidase activity, which indicated the

presence of the integrated fusion construct. Integration into the

yeast genome was a stable method for carrying the fusion with a

segregation rate of less than one percent.

Once strains were isolated that carried the 5' deletion series

integrated in unit copy, they were assayed for 3-galactosidase

activity in order to determine the effect of the deletions on the

expression of the TPI::IacZ gene fusion. p-galactosidase assays were

carried out in duplicate on three separate occasions. Figure 6

depicts the assay results. Plasmid 92-9 contains the entire 3.5 kb

of 5' non-coding region. When integrated in the S150-2B background,

276 Miller units of activity were expressed. 92-9 integrated in the


























Figure 6. Effect of 5' deletions upon expression of a TPI::IacZ fusion integrated in unit copy at the
URA3 locus in wild-type and gcrl-deletion mutant strains. Constructs 92-9, 92-9R, and 35-2, 35-2R
contain the same respective deletion endpoints but are integrated in opposite orientation at the URA3
locus. Deletion endpoints are indicated with respect to the start of translation. 1-galactosidase
activities were determined by the method of Miller (1970) in duplicate from at least three independent
cultures. Strains were grown to an optical density A600oo of approximately 1.0 in YP medium
supplemented with either 2% glycerol and 2% lactate (YPGL), or 2% glucose (YPD).















-220 +1
Sph I ATG
1 I


g- Galactosidase Activity
S150-2B (GCR1) HBY4 (gcrl)


YPGL


500

500
-TPI I""MacZ lw*W'TPI M


ITPIr B i'i-acZ :ii i I


I- TPI' -I TplacZ T:P I l


-490

-420


-392


'-3!
92-9
'-3
92-9R

35-2

35-2R

36-2

37-2

34-1

70-1

83-1

73-2

82-1

77-2

78-1

76-1

38-4

39-2

No Fusion


-337

-330


HTPr imm acZ TPi


PIr B a acz f1rTPI -a

P' I:i:iacZ ri:H7pl


1TPI' iii:llacZ B ai1'TPI 8


HTP'r M-:-dijacz S.TPI J


BTpr IMiacZ I.MTPI


ETPIr iacziETrPI I
-299
fTP irTiiiiiIlacZ i TPI i


-278

-192


fTWPI H i WllacZ : TPI


-179
ITPI'liacZ TP
+63
I:T71I lacZ Pill


YPD


YPGL


Mean SD Mean SD Mean SD


276 11

254 4

210 11

276 18

241 36

266 9

251 36

107 8

27 7

19 8

16 9

14 3

11 10

36 11

17 11

14 7

9 3


222 6

215 21

236 25

251 8

238 33

255 43

241 24

69 4

21 3

9 6

6 3

6 3

8 4

10 7

5 4

4 3

4 3


19 3

17 1

21 4

ND

22 s

20 3

20 2

29 3

18 3

4 4

3 3

13 4

10 8

ND

12 1

8 8

3 3


Fusion #


-377

-348


"fn


I---~~.. --.~--~











gcrl mutant HBY4 background only expressed 19 units of activity.

Thus the gcrl mutation resulted in a 12 fold reduction in the level of

expression of the TPI::IacZ fusion. Clifton and Fraenkel (1981)

previously reported a 17 fold reduction in the level of triose-

phosphate isomerase in a gcrl mutant background. Thus, expression

of the TPI::IacZ, like TPI itself, is dependent on GCR1 for full

expression. Plasmids 92-9 and 92-9R, 35-2 and 35-2R contain the

same respective 5' deletion endpoint, but the TPI::lacZ fusion was

integrated in opposite orientation with respect to URA3. No effect

on expression of the fusion due to orientation with respect to URA3

was observed. Strains harboring deletions up to position -392 still

express high levels of 13-galactosidase activity. However, deletion

of an additional 15 bp to position -377 reduces expression

approximately two-fold to 107 units of activity. Deletion to -348

or beyond abolished expression of the fusion. Based on these

results, the 5' boundary of the region sufficient for high-level

expression of TPI::IacZ must reside within 392 to 377 base pairs

from the start of the TPI structural gene, when the fusion is

integrated at the URA3 locus.











An Upstream Activating Sequence Activity for TPI Resides from

Position -377 to -327 in the 5' Non-Coding Region

The 5' boundary of the TPI controlling region mapped by the

initial deletions was probably the boundary of an upstream

activating sequence (UAS). UAS's of yeast are similar to enhancer

elements of higher eukaryotes. UAS's are the sites of interaction

between positive trans-acting proteins and specific DNA target

sequences, and serve to activate transcription. Others in the

laboratory who utilized fragments isolated from the initial deletion

series in DNA band shift assays, demonstrated a specific protein

nucleic acid interaction involving the region between -377 and -327

(Scott et al., 1990). Further band shift experiments determined that

the protein responsible for the shift was RAP1. As noted in the

introduction, it has been recently reported that RAP1 binding is

important for the UAS activity of other genes encoding glycolytic

enzymes such as PGK1 (Ogden et al., 1986; Chambers et al., 1989),

EN01 (Buchman et al., 1988; Brindle et al., 1990), EN02 (Brindle et

al., 1990), PYK (Buchman et al., 1988), PDC1 (Kellerman and

Hollenberg, 1988), and ADHi (Buchman et al., 1988; Santangelo and

Tornow, 1990). I wanted to determine if the region downstream of











position -392 of TPI encoded an UAS, and if RAP1 binding was

required for expression of TPI.

To address this question, fragments of the TPI controlling

region were used to replace the native UAS elements of a CYC1::IacZ

gene fusion. A Hindlll-Sphl fragment (-392 to -220) isolated from

the last 5' deletion construct to drive high-level expression of

TPI::IacZ (34-1) was able to restore expression to a CYC1::lacZ gene

fusion which had its native UAS elements removed (Figure 7). Cells

containing a CYC1::IacZ gene fusion which had a 66 base pair

oligomer (HB14 and HB15, Table2)(TP/ sequence from -377 to -327)

substituted for the native UAS elements of CYC1 were able to

express 99 units of P-galactosidase activity. This level of

expression was comparable to the 94 units of activity expressed

using the longer TPI fragment, 34-1 H-S (-392 to -220) or the 101

units expressed by cells containing the 81 base pair oligonucleotide

(HB16 and HB17, Table 2)(-392 to -327). Therefore, UAS activity is

bounded by positions -377 to -327 which contains the RAP1-binding

site from -358 to -346. The sequence from -377 to -327 will be

termed UASTPI.

The same 51 base pairs was able to express 129 units of P-

galactosidase activity from the TPI::IacZ fusion which was deleted


























Figure 7. Identification of UASTPI. Various fragments containing portions of the 5' nontranslated
region of TPI were cloned before either a CYC1::IacZ construct deleted of the native UAS elements or a
TPI::IacZ construct deleted of the native UAS elements. p-galactosidase assays were performed from
cultures grown in YPD as described in figure 5 Fragment 34-1 H-S contains the sequence form -392
to -220 of TPI. 81mer contains from position -392 to -327. 66mer contains from position -377 to -
327. 58mer contains from -377 to -335. 69RBSM is identical to the 66mer except that the sequence
from positions -349 to -339 (CACCCCTTTTC) was replaced with the sequence AACCCATCAGG.










B Galactosidase Activity


CYC1:: lacZ

UASless CYC1:: lacZ

34-1 H-S:: CYC1:: lacZ

81 mer:: CYC 1:: lacZ

66mer:: CYC1:: lacZ




35-2

78-1

58mer:: TPI:: lacZ

66mer:: TPI:: lacZ

69 RBSM:: TPI:: lacZ


" UAS TATA CYC1::lacZ

( ) --TATA CYC1::lacZ

(34-1 H-S) -- TATA CYC ::lacZ

(81 mer) TATA CYC1::lacZ

(66mer) -- TATA CYC1 ::lacZ




TATA TPI ::lacZ


STATA TPI ::lacZ

58mer ( )--TATA --- TPI ::lacZ
)-TATA TPI ::IacZ

66mer ( )--TATA TPI ::lacZ

69 RBSMJ ( -TATA--- TPI::lacZ


Mean
70

8

94

101

99




236

8

20

129

18


Construct


Arrangement









63

of all sequences 5' to the Sphl site at -220. However, a 58 base pair

oligonucleotide (HB07 and HB08, table 2) with TPI noncoding

sequence from -377 to -335 was unable to restore expression the

TPI::IacZ fusion, Figure 7. An oligonucleotide (69RBSM or HB19,

table 2) identical to the native 51 b.p. region from -377 to -327

except for mutations in the RAP1-binding site was prepared. The

mutant oligonucleotide was placed before the same deleted TPI::IacZ

fusion as the wild-type oligonucleotide to assess the role of the

RAP1-binding site in UASTPI function. The mutated oligonucleotide

was unable to drive expression of the fusion demonstrating that the

RAP1-binding site from -358 to -346 is essential for UAS activity.

Others in the laboratory performed DNA band shift assays with these

oligonucleotides which indicated that RAP1 binds with much reduced

affinity if at all to the mutant oligonucleotide in vitro (Scott et al.,

1990).



Internal Deletions Indicate Single UAS Element Responsible for TPI

Transcription

The 5' deletion series facilitated the mapping of an upstream

activating sequence, UASTPI, that was sufficient to drive expression









64

of TPI. However, controlling regions in Saccharomyces cerevisiae

are often composed of multiple UAS elements, any of which are

independently capable of driving expression (Guarente et al., 1984;

Cohen et al., 1986; Cohen et al., 1987; Johnston, 1987; Nishizawa et

al., 1989). One-tailed deletion series are able to map only the last

element remaining which is sufficient for expression. In order to

determine the number of UAS elements present in the controlling

region of TPI, a second set of deletions was generated which began

internal to the known UASTPI at position -220 and extending towards

and through UASTPI located from -377 to -327. Figure 2 cartoons the

scheme utilized to generate the internal deletion series.

Deletions originating internal to and extending through UASTPI

of the TPI::lacZ fusion in plasmid pES90 ("internal" deletions) were

created originating from a unique Sphl site at position -220.

Linearized plasmid pES90 was digested with exonuclease Bal 31,

followed by the addition of Sphl linkers (HB06, table 2). To assure

an intact TATA element, required for expression, the deletion

products distal to the Sphl site were isolated after Hindlll digest by

gel purification then subcloned into plasmid pES90 which had been

digested with Hindlll and Sphl. The final "internal" deletion series











had a Sphl site at position -220 that served to fix the 3' deletion

endpoint at position -220. All "internal" deletion constructs had

position -853 as their common 5' end. Precise 3' deletion end-points

and junctions were determine by sequencing using primer HB05,

Table 2, and are shown in Figure 8. The "internal" deletion series

was integrated into both wild-type (S150-2B) and gcrl mutant

(HBY4) strains of yeast. Unit copy integrants were confirmed by

Southern blot analysis (Southern, 1975) as previously detailed. P-

galactosidase assays by the method of Miller (1972) were carried

out in duplicate on three independent occasions.

Figure 8 depicts the "internal" deletion constructs and the

results of the p-galactosidase assays from lysates of strains with

unit copy integrants of the various plasmids. 392 base pairs of 5'

non-coding region was sufficient for high-level expression of the

TPI::IacZ fusion. The initial "internal" deletion construct contained

5' sequence from -853 to -300 and from -220 through the structural

gene. This construct retained the known UASTPI sequence, and

produced 138 units of p-galactosidase activity when integrated into

the wild-type strain, S150-2B. 52 units of P-galactosidase activity

























Figure 8. Effect of internal deletions on P-galactosidase activity expressed from a TPI::IacZ gene
fusion. The cartoon depicts the extent of the internal deletions with precise deletion endpoints
indicated. p-galactosidase assays were performed in duplicate by the method of Miller (1972) on three
individual occasions. Plasmids with deletions in the 5' noncoding region of TPI were integrated in unit
copy at the URA3 locus in both wild-type (S150-2B) and gcrl mutant (HBY4) strains of yeast.
















Sph I
-220
i


IIIAII IUACs I


IUAS?


Il)-AS I


r~"~


-392


-377


-348

^ -300 _


-


TPI::IacZ I


\TPI::lacZ I


\TPI::IacZ 1


TPI::lacZ 1





TPI::lacZ I


TPI::lacZ l


ITPI::lacZ I

No Fusion


-347


D- Galactosidase Activity
8150.28 (GCR1) HBY4 (gcrl)
Mean SD Mean SD

21011 21 4


251 36 20 2


1078 29 3


27 7 183


13819 52 19


52 3 20 4


15 2 7 2


14 2


12 5


853


'"""' '


_._-


S-367
853











were expressed in the gcrl mutant background, only a 2.5 fold

reduction in expression. The lessening of the severity of the

reduction caused by the gcrl lesion may be due to a position effect

such as was seen for EN02 gene (Holland et al., 1990). The 138 units

of activity in the wild-type background corresponds with the

expression observed when only UASTPI was driving expression of the

fusion (Figure 7). Deletion of an additional 36 base pairs from the 5'

noncoding region of the TPI::IacZ gene fusion, removal of sequence

from -336 to -220, reduced expression of the fusion approximately

three-fold to 52 units of 1-galactosidase activity. The strain

harboring the construct that removed sequence from -347 to -220

yielded background levels of expression. Constructs which are

deleted of the known UAS element but retain all sequences to

position -853, are unable to drive expression. Therefore, no

additional UAS elements lie distal to UASTPI.



Mutational Analysis of UASTPi

UASTPI has been mapped to a 51 base pair region from position

-377 to -327 with respect to the start of the structural gene.

Figure 9 shows the sequence of the 5' non-coding region of TPI
















37-2 [266] 34-1 [251] 70-1 [107]
-430 (-420) -392) (-377)
1 | f **000
TATATCTAGGAACCCATCAGGTTGGTGGAAGATTACCCGTTCTAAGACTTTTCAGCTTCCTCTAT
(-377)



83-1 [27] 73-2 [19] 77-2 [16]
(-348) l(-337) 1(-314)

TGATGTTACACCTGGACACCCCTTTCTIGCATCCAGTTTTTAATCTTCAGGGCATGTGAGATTC
(-327)





-234

TCCGAAATTAATTAAAGCAATCACACAATTCTCTCGGATACCACCTCGGTTGAAACTGACAGGTGA










Figure 9. Summary composite of the 5' noncoding region of TPI.
Deletion endpoints of the 5' deletion series are denoted with the
construct number in bold print, positions are in parentheses. p-
galactosidase activities expressed by wild-type strains harboring
the 5' deletions are indicated in brackets. UASTPI is underlined. The
RAP1-binding site is underscored by dashes with mismatches with
the consensus RAP1-binding site underscored by X's. The CTTCC and
CATCC pentamer motifs are overdotted.








70

flanking the UAS element. The RAP1-binding site from -358 to -346

has been shown to be required for UAS activity in wild-type yeast

(Figure 7). Recently, other laboratories have demonstrated the

requirement of a CTTCC pentamer motif for the full expression of

PGK (Stanway et al., 1989), PYK, and EN01 (Buchman et al., 1988).

Examination of UASTPI revealed a CTTCC pentamer located from

position -375 to -370, and a closely related CATCC pentamer from

-335 to -330 (Figure 9).

Mutant variations of UASTPI were generated and cloned at

position -220 before a TPI::IacZ fusion. The constructs were

integrated in unit copy into wild-type (S150-2B) and gcrl mutant

(HBY4) strains of yeast. UASTPI, -377 to -327 (HB14 and HB15, Table

2), expressed 115 units of P-galactosidase activity in the wild-type

background and 18 units of activity in the gcrl mutant background,

Figure 10. It was interesting to note that expression of the UASTPI

construct was reduced 10-fold in the gcrl mutant background. Thus,

GCR1 must act through UASTPI sequence, or through a sequence

downstream of position -220. The role of the RAP1-binding site in

UASTPI activity was addressed by an oligonucleotide, HB24 (Table 2),
























Figure 10. Mutational analysis of UASTPI utilizing mutant
oligonucleotides. Double-stranded oligonucleotides with portion of
the 5' noncoding region of TPI were cloned before a TPI::IacZ fusion
at position -220. Positions within the 5' noncoding region are
indicated. The large box denotes the RAP1-binding site from
position -358 to -346. The 5' small box denotes the CTTCC pentamer
motif located from -375 to -370. The 3' small box denotes the
CATCC pentamer motif from position -335 to -330. Filled boxes
represent mutated motifs as detailed in the text. Unit copy
integrants of the various constructs in wild-type (S150-2B) or gcrl
mutant (HBY4) yeast were assayed for (3-galactosidase activity.
Assays were performed in duplicate on three sperate occasions.

















-2


-2

-392 i- -27
-392 m-- L."~1 ^"
-2

-377 m~ 1
- S-327

-2





-2327
-377
2-



-37 -327
L -2-*


I TPI::lacZ I


TPI::lacZ


TPI::lacZ


TPI::lacZ


TPI::lacZ


TPI::lacZ


TPI::lacZ


20


20


20
1


-220


-220
-220


I TPI::lacZ I


TPI::IacZ


72

3-Galactosidase Activity
S150-2B (GCRI) HBY4 (gcrl)
Mean SD Mean SD


161 8 9 1


115 7 18 5


116 9 12 5


14 3 10 3


45 4 21 7


82 4 21 7


36 7 20 5


27 11


9 1


28 4


8 2


!0











identical to the native TPI sequence, but for replacing the RAP1-

binding site with an EcoRI and Pstl restriction site. When the HB24

derived mutant UASTPI was placed before the TPI::IacZ fusion, p-

galactosidase activity was not expressed in either strain

background. Thus, UASTPI activity has an absolute requirement for

the RAP1-binding site.

In order to determine if a RAP1-binding site was sufficient for

UAS activity a consensus RAP1 site (AACCCATACATG),

oligonucleotide HB09, was cloned before the fusion. The consensus

RAP1 site had been shown to bind RAP1 protein (Scott et al., 1990).

As a control a mutant RAP1-binding site (AACCCATCAGG)

(oligonucleotide HB11), unable to bind RAP1, was also cloned before

the TPI::IacZ fusion. The consensus RAP1-binding site was not

sufficient to drive expression of the TPI::IacZ fusion, Figure 10, and

resulted in only 27 units of 3-galactosidase activity.

To address the role of the CTTCC and CATCC pentamer motifs

in UASTPI function, double-stranded oligonucleotides were generated

that changed either the CTTCC to CAACC (HB21), the CATCC to

CAACC (HB22), or both (HB24) while retaining a functional RAP1-

binding site. Similar mutations had been shown to reduce expression











driven by the UAS elements of the PYK and EN01 (Buchman et al.,

1988). These oligonucleotides were cloned at position -220 before

the TPI::IacZ fusion, integrated in unit copy, and the strains were

assayed for p-galactosidase activity.

Mutating the CTTCC from -375 to -370 to CAACC reduced

expression of the TPI::IacZ fusion construct in wild-type yeast to 45

units of activity, a 2.5-fold reduction. 21 units of p-galactosidase

activity were expressed in the gcrl mutant background. Changing

the CATCC from -335 to -330 to CAACC reduced activity by one

third to 82 units in the wild-type background and 21 units of

activity in the gcrl mutant strain. Mutating both pentamers reduced

activity to 36 units in the wild-type strain. The double mutant

containing construct expressed 20 units of p-galactosidase activity

in the gcrl mutant strain.

Both pentamers were required for full UAS activity with

mutations in the CTTCC having a larger effect than mutations in the

CATCC. However, it should be noted that the oligonucleotides

employed in these experiments contained only the sequence from

-377 to -327 known to be sufficient for UASTPI activity (Figure 7).











Expression driven by UASTPI was 115 units of 13-galactosidase

activity while the full length (-853) fusion construct expressed 161

units of activity.



In vitro DNase I Protection Assays Reveal Binding of the REB1 Site

and the RAP1 Site

Upstream activating sequence elements are sites where trans-

acting factors bind to cis-acting elements to mediate expression of

the cognate gene. UASTPI was initially used in a series of DNA

bandshift assays by others in the laboratory with crude protein

extracts derived from yeast. Using UASTPI and the RAP1-binding site

mutant derivative, HB19 or 69RBSM from the oligonucleotide cloning

experiments, they were able to demonstrate RAP1 binding to UASTPI

in a RAP1-binding site dependent manner (Scott et al., 1990). Since

UASTPI was specifically bound by protein, DNase I protection assays

as modified by Singh et al. (1986) were performed on the region to

determine the precise areas of interaction.

In order to detect the region protected by the RAP1

protein, the 169 base pair Hindlll-Sphl fragment from the 5' deletion

pES34-1 was used as a target in the DNasel protection assay. In








76

addition to the S150-2B protein extract, a rabbit reticulocyte lysate

(RRL) containing RAP1 generated via in vitro transcription and

translation was used in the binding reactions. The RRL containing

RAP1 was generously provided by M. Cecilia Lopez. Figure 11 shows

the results of the DNase I protection assay. A region of heightened

DNase I sensitivity was observed bounded by positions -288 to -285.

Protection from DNase I cleavage was seen from position -365 to

-343. Similar results were obtained with both the yeast protein

extract and the in vitro generated RAP1 in RRL. No regions of DNase

I protection were seen that affected the CTTCC or CATCC pentamer

motifs when the yeast protein extracts were used. However, Michael

A. Huie has been able to detect an area of DNase I protection

centered about the CTTCC when he used a purified MBP-GCR1 fusion

protein in the binding reactions (Huie et al., 1992).

The initial target used in the DNase I protection assay was the

234 base pair Hindll-Sphl fragment from the 5' deletion construct

pES40-23. Protein extracts were prepared from the yeast S150-2B.

The results of the DNase I protection analysis on this fragment are

shown in Figure 12. Surprisingly, an area of protection was

observed bounded by positions -397 to -386. Inspection of the
























Figure 11. In vitro DNase I protection assays demonstrating
protection of the RAP1-binding site. The last four lanes are a DNA
sequencing ladder generated with M13mp18 and the forward primer.
This ladder serves as a molecular weight marker. The (-) reaction
indicates DNase I cleavage was performed on a naked radiolabeled
Hindlll-Sphl fragment from plasmid pES34 and serves as the control
ladder. The second and third lanes, labeled 0.5 and 1.0 respectively,
are the DNase I cleavage reactions generated by 0.5 and 1.0 units of
DNase I performed in the presence of in vitro generated RAP1. The
fourth and fifth lanes are the DNasel cleavage reactions of 0.5 and
1.0 units of DNase I performed in the presence of yeast protein
extract. Regions of DNase I protection and hypersensitivity are
demarcated and the sequence indicated.











DNasel
0(-) d ,- 0,( ) A C G
(-) (-) A C G T


-286

-290 I-


; ** ... -i ... ..... .= ... ,


i "
vc



'f '
U:;


ml


"Aw ": 4,w .q w
IP l ,
- O '4


-343 H
O
O
O
0




S-
O


Sc
I-
0




F-



-366 -
10


C,

*
























Figure 12. In vitro DNase I protection assays demonstrating
protection of the REB1-binding site. The first four lanes are a DNA
sequencing ladder generated with M13mp18 and the forward primer.
This ladder serves as a molecular weight marker. The next two
lanes are the products of DNase I digestion of a radiolabeled Hindlll-
Sphl fragment from plasmid pES40-23. (+) extract indicates the
DNase I digestion was carried out in the presence of yeast protein
extract. The (-) extract reaction was performed on naked DNA and
serves as the control Dnase I cleavage ladder. The region of DNase I
protection is demarcated and the sequence indicated.










A C G T (+)(-)


atiiu H-397


,1 C.>, Oo. o

S'.I-- -388
MIc




di n





*- **
'i" ss," *11*'1
S^-( II
*;- ^-
^-^
**1~ *








81

protected sequence revealed that it was a one base pair mismatch

from a consensus REB1-binding site proposed by Chasman et al.

(1990). REB1 has been implicated in nucleosome phasing in

transcriptional control elements (Brandl and Struhl, 1990).

However, it should be noted that the REB1-binding site can be

deleted in the 5' deletion series with no effect on expression (pES34-

1, figure 6), when the fusion construct is integrated at the URA3

locus.



DNA Band Shift Assays Demonstrate REB1 Binding to TPI 5' Non-

Coding Region

The DNase I protection assays demonstrated an area of

protection over a region containing a one base pair mismatch from a

consensus REB1-binding site. DNA band shift assays were performed

to determine if REB1 binds the TPI 5' non-coding region. The target

fragment was an end-labeled 139 bp Avall-Fokl fragment (positions

-487 to -348). This fragment contains the sequence protected in the

DNase I assays, as well as the near consensus REB1-binding site.

Others in the laboratory have demonstrated that a fragment

containing sequence from -392 to -348 is incapable of interacting








82

with wild-type yeast extracts is DNA band shift assays ( H. Baker,

personal communication). Extracts were prepared from the wild-

type yeast (S150-2B), or E. coli expressing REB1 from a REB1 insert

in the expression vector pET11A, and E. coli with pET11A and no

insert (kindly provided by B. Morrow and J. Warner). Results of the

DNA band shift assays are shown in Figure 13.

The wild-type yeast extract and the E. coli extract containing

REB1 gave rise to positive band shift assays when incubated with

the fragment from -487 to -348. The E. coli extract without the

REB1 insert failed to band shift. These results indicate that REB1 is

capable of binding a region of the 5' non-coding region of TPI that

contains a near consensus REB1-binding site.



In vivo Methylation Protection Assays

The DNase I protection and DNA band shift assays were able to

provide information about the ability of the trans-acting factors

RAP1, REB1, and GCR1 to bind to the TPI controlling region under in

vitro conditions. The ability to demonstrate an area of protection

indicated that the factors have the ability to interact with specific

sites, but, did not prove that the interactions occur in vivo. To
























Figure 13. DNA band shift assays demonstrating REB1 binding to the
5' noncoding region of TPI. An Avall-Fokl restriction fragment was
isolated from pES90. Radiolabeled DNA fragment was incubated in
binding buffer with E.coli extracts without or with REB1, and wild-
type yeast extract (S150-2B). The first lane serves as a control for
the migration of the fragment alone. Nucleoprotein complexes were
resolved from free DNA by nondenaturing polyacrylamide gel
electrophoresis and were revealed by autoradiography. f, free
unbound probe.





















an


, FRAG.ALONE
* E.coli/pllA
E. coli/pl11A-REB1
S150-2B








85

determine the DNA sequences bound by protein in the cell, dimethyl

sulfate (DMS) methylation protection assays genomicc footprinting)

(Ephrussi et al., 1985) were carried out on various strains of S.

cerevisiae. The strains utilized in the genomic footprints were

S150-2B (wild-type), DFY 642 (wild-type), HBY4 (gcr-1), and JF

1052 (spt13), Table 1. The gcr-1 strain was chosen because the

mutation is known to affect expression of TPI (Clifton and Fraenkel,

1981). SPT13/GAL11 gene function is required for the full

expression of PYK1 (Nishizawa et al., 1990).

Yeast were treated DMS, DNA harvested, prepared and blotted

to a nylon membrane as described in Materials and Methods. Top and

bottom strand radiolabeled probes, as defined in Figure 14, were

prepared to hybridize to the TPI controlling region. Figure 14 shows

the methylation protection pattern of the top strand of the TPI

controlling region from all four strains. The actual footprinting

controls and results are on the left of the figure. On the right a

cartoon represents the sequence of the UASTPI region with the

protected bases denoted. The first four lanes in Figure 14 are

genomic sequencing reactions that provided molecular weight

markers. The genomic "G" sequencing reaction was also the naked
























Figure 14. Genomic footprinting of the bottom strand of the TPI 5'
noncoding region. The initial four lanes are genomic sequencing
reactions of the TPI 5' noncoding region. The G sequencing ladder
serves as the control ladder for the genomic footprinting reactions.
Lanes 1-4 are genomic footprinting reactions carried out in wild-
type (S150-2B) yeast. The reactions in lanes 1 and 2 were generated
in wild-type yeast grown in YP media supplemented with 2% glucose
(YPD) and represent 3 and 4 minutes of DMS treatment, respectively.
Lanes 3 and 4 were grown in YP media supplemented with 2%
glycerol and 2% lactate and represent 4 and 5 minutes of DMS
treatment, respectively. Lanes 5 and 6 are the products of 4 and 5
minutes of DMS treatment in a gcrl (HBY4) mutant strain of yeast.
Lanes 7 and 8 are similar treatment of a second wild-type strain
(DFY642). Lanes 9 and 10 are similar treatment of a spt13 (JF1052)
mutant strain. Guanine residues protected within the GCR1-binding
site are denoted by an (*). Residues protected within the RAP1-
binding site are denoted by (0). Residues protected within the REB1-
binding site are denoted by (A). The right portion of the figure
depicts the double stranded sequence of the TPI 5' noncoding region.
The bottom strand is the righthand most strand as depicted. The
sequence motifs known to play a role in TPI expression are stippled.
Protected guanine residues are denoted as above.











T G wild-type gcrl w-t 2
+ + I I r-1 I--
C CAG 1 2 3 4G 5 6G 78


sptl3

G 910G


I


=8


--4


mn 87

r E-





E-'



U
00




00
-" 0O



E-



U


00 -
00





E-e





E-



U0


43


n 00
o-0 U

nI -
tIn








88

DNA control for the footprinting reactions. In the wild-type strains

grown in YP medium supplemented with glucose, four major areas of

protection were observed. The most distal region of protection was

four guanine residues. Three of the four bases fall within the near

consensus binding site for REB1, and all four bases are within the

region protected from DNase I digestion in vitro. The next area of

protection was the guanine doublet that base pairs with the cytosine

doublet in the CTTCC pentamer motifs. The CTTCC pentamer was

shown to be protected by GCR1 in vitro, (Huie et al., 1992). The next

region of methylation protection seen in the wild-type background

was five guanine residues within the RAP1-binding site. The final

area of protection seen on the top strand of the wild-type strain

was the guanine doublet that corresponds to the cytosine doublet of

the CATCC pentamer. In vitro DNase I protection by GCR1 of the

CATCC pentamer was suggested but not well resolved in the

footprinting gel (Huie et al., 1992). The pattern of protection seen

in both wild-type strains, S150-2B and DFY 642, are identical. Thus,

the genomic footprinting analysis was able to detect protein

interactions at all known binding sites within the UASTPI region in

wild-type yeast grown in glucose.








89

The effect of carbon source on factor binding-site occupancy

state at UASTPI was also tested. Lanes 3 and 4 of Figure 14 show

that there was no effect of the carbon sources tested on the pattern

of protection observed.

Lanes 5 and 6 of Figure 14 are the footprinting reactions

generated by DMS treatment of a gcrl mutant strain of yeast grown

in YP medium supplemented with 2% glycerol and 2% lactic acid. The

guanine doublets in the pentamer motifs were not protected. Both

sites were susceptible to methylation and cleavage to the same

degree as the control reactions. However, the areas of protection

corresponding to the REB1 and RAP1-binding sites were still

present. The GCR1 dependent binding of the pentamer motifs was

not required for the binding of the REB1 and RAP1-binding sites in

the TPI controlling region.

Mutation of the sptl3/gall1 locus had no effect upon the

protection of UASTPI in vivo. Lanes 11,12 of Figure 14 are the

footprints generated in the spt13 backgrounds. The areas of

protection from methylation are identical to those seen in the wild-

type strain. The REB1, RAP1, and GCR1-binding sites are all still

bound.








90

Figure 15 depicts the genomic footprint of the TPI controlling

region probed for the bottom strand, as defined in the figure, in both

wild-type and gcrl mutant strains of yeast. Only one area of

protection was seen in both strains. The guanine at position -392

was protected from methylation. Position -392 corresponds to the

last base of the near consensus REB1-binding site.



Site-Directed Mutagenesis of Transcription Factor Binding Sites in

the UAS of TPI

Once the sites bound by transcription factors had been

identified, it was important to mutate those sites in order to assess

their role in TPI gene expression. Site-directed mutagenesis was

utilized to introduce mutations in single and pairwise combinations

of factor-binding sites before the TPI::IacZ gene fusion construct.

Plasmid pES90, used as a target for mutagenesis, contained 853

base pairs of material 5' to the start of the structural gene. This

construct was used to insure that the mutations were in a context

as close as possible to the native loci.

Figure 16 depicts the mutations made in the TPI controlling

region and the results of the p-galactosidase assays performed. The




Full Text
Figure 11. In vitro DNase I protection assays demonstrating
protection of the RAP1-binding site. The last four lanes are a DNA
sequencing ladder generated with M13mp18 and the forward primer.
This ladder serves as a molecular weight marker. The (-) reaction
indicates DNase I cleavage was performed on a naked radiolabeled
Hin\\\-Sph\ fragment from plasmid pES34 and serves as the control
ladder. The second and third lanes, labeled 0.5 and 1.0 respectively,
are the DNase I cleavage reactions generated by 0.5 and 1.0 units of
DNase I performed in the presence of in vitro generated RAP1. The
fourth and fifth lanes are the DNasel cleavage reactions of 0.5 and
1.0 units of DNase I performed in the presence of yeast protein
extract. Regions of DNase I protection and hypersensitivity are
demarcated and the sequence indicated.


3
that segregates as a single gene. Due to its pleiotrophic nature the
gene was named GCR1 for glycolysis regulation. Strains harboring a
gcr1 mutation express the genes encoding the enzymes of glycolysis
at approximately 5-10% of wild-type levels (Clifton et al., 1978;
Clifton and Fraenkel, 1981; Baker, 1986). gcr1 mutants exhibit poor
growth on glucose while retaining adequate growth on non-
fermentable carbon sources (Clifton and Fraenkel, 1981). The
reduction in enzyme levels has been shown to be mirrored by a
corresponding reduction in steady-state mRNA levels for several
affected enzymes (Holland et al., 1987; Santangelo and Tornow,
1990; Scott et al., 1990). GCR1 has been cloned (Kawasaki and
Fraenkel, 1982) and sequenced (Baker, 1986; Holland et al., 1987).
DNA sequence analysis indicates that GCR1 encodes a polypeptide of
844 amino acids with a molecular weight of 94,414 Daltons.
Much attention and effort has been devoted to the study of gene
regulation in S. cerevisiae and other systems. The plieotrophic
nature of mutations in GCR1 suggests that the gene is involved in
the coordinate regulation of expression of the genes encoding
glycolytic enzymes. However, the residual level of expression of
aldolase, triose-phosphate isomerase, glyceraldehyde-3-phosphate


Nogi, Y. and Fukasawa, T. (1980). A novel mutation that affects
utilization of galactose in Saccharomyces cerevisiae. Genet. 2,
115-120.
132
Ogden, J.E., Stanway, C., Kim, S., Mellor, J., Kingsman, A.J., and
Kingsman, S.M. (1986). Efficient expression of the Saccharomyces
cerevisiae PGK gene depends on an upstream activating sequence but
does not require TATA sequences. Mol. Cell. Biol. 6, 4335-4343.
Oliphant, A.R., Nussbaum, A.L., and Struhl, K. (1986). Cloning
random-sequence oligodeoxynucleotides. Gene 44, 177-183.
Orr-Weaver, T.L., Szostak, J.W., and Rothstein, R.J. (1983). Genetic
applications of yeast transformation with linear and gapped
plasmids. In Methods in enzymology: volume 101. R. Wu, L. Grossman,
and K. Moldave, eds. (New York: Academic Press), pp. 228-244.
Parker, C.S. and Topol, J. (1984). A Drosophila RNA polymerasell
transcription factor contains a promoter-region specific
DNA-binding activity. Cell 36, 357-369.
Pellman, D., McLaughlin, M.E., and Fink, G.R. (1990). TATA-dependent
and TATA-independent transcription at the HIS4 gene of yeast.
Nature 348, 82-85.
Pugh, B.F. and Tjian, R. (1990). Mechanism of transcriptional
activation by Sp1: evidence for coactivators. Cell 61, 1187-1197.
Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). Molecular
Cloning- A laboratory Manual (Cold Spring Harbor: Cold Spring Harbor
Laboratory Press).
Sanger, F., Nicklen, S., and Coulson, A.R. (1977). DNA sequencing with
chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74,
5463-5467.


89
The effect of carbon source on factor binding-site occupancy
state at UASjpi was also tested. Lanes 3 and 4 of Figure 14 show
that there was no effect of the carbon sources tested on the pattern
of protection observed.
Lanes 5 and 6 of Figure 14 are the footprinting reactions
generated by DMS treatment of a gcr1 mutant strain of yeast grown
in YP medium supplemented with 2% glycerol and 2% lactic acid. The
guanine doublets in the pentamer motifs were not protected. Both
sites were susceptible to methylation and cleavage to the same
degree as the control reactions. However, the areas of protection
corresponding to the REB1 and RAP1-binding sites were still
present. The GCR1 dependent binding of the pentamer motifs was
not required for the binding of the REB1 and RAP1-binding sites in
the TPI controlling region.
Mutation of the spt13/gal11 locus had no effect upon the
protection of UAStpi in vivo. Lanes 11,12 of Figure 14 are the
footprints generated in the spt13 backgrounds. The areas of
protection from methylation are identical to those seen in the wild-
type strain. The REB1, RAP1, and GCRI-binding sites are all still
bound.


59
An Upstream Activating Sequence Activity for TPI Resides from
Position -377 to -327 in the 5' Non-Coding Region
The 5' boundary of the TPI controlling region mapped by the
initial deletions was probably the boundary of an upstream
activating sequence (UAS). UAS's of yeast are similar to enhancer
elements of higher eukaryotes. UAS's are the sites of interaction
between positive trans-acting proteins and specific DNA target
sequences, and serve to activate transcription. Others in the
laboratory who utilized fragments isolated from the initial deletion
series in DNA band shift assays, demonstrated a specific protein
nucleic acid interaction involving the region between -377 and -327
(Scott et al., 1990). Further band shift experiments determined that
the protein responsible for the shift was RAP1. As noted in the
introduction, it has been recently reported that RAP1 binding is
important for the UAS activity of other genes encoding glycolytic
enzymes such as PGK1 (Ogden et al., 1986; Chambers et al., 1989),
EN01 (Buchman et al., 1988; Brindle et al., 1990), EN02 (Brindle et
al., 1990), PYK (Buchman et al., 1988), PDC1 (Kellerman and
Hollenberg, 1988), and ADH1 (Buchman et al., 1988; Santangelo and
Tornow, 1990). I wanted to determine if the region downstream of


115
However, when the REB1-binding site was mutated in the context of
853 base pairs of TPI 5' noncoding sequence, a 2-fold reduction in
expression was seen by the Miller assay (Figure 16) or a 5-fold
reduction in specific activity (M.C. Lopez personal communication).
Intimating that REB1 may counter the action of inhibitory sequences
which reside 5' to the REB1 -binding site. REB1 has been implicated
in nucleosome positioning (Bram and Kornberg, 1982; Brandi and
Struhl, 1990; Fedor et al1988; Chasman et al1990). Chasman et
al. (1990) also demonstrated that UASgal functions independently of
the REB1/GRF2-binding site in test constructs. However, a REB1-
binding site is required for full expression of the GAL1 and GAL10
genes (Chasman et al., 1990). If the role of REB1 is to exclude
nucleosomes, thereby opening chromatin structure for transcription,
then the discrepancy between the results of the 5' deletion series
and site directed mutants may be explained. REB1 binding was not
needed for UAStpi function in the 5' deletion series when the
deletion endpoint was located directly downstream of the actively
transcribed URA3 locus. This may be due to a nucleosome free
region created by active transcription and, therefore, REB1 was not
needed. The site-directed mutant constructs, however, had


5
identification of other trans acting factors required for the full
expression of TPI is also a primary concern. This introduction will
serve as a literature review of the studies to date concerning
glycolytic gene expression in Saccharomyces cerevisiae.
In addition to the GCR1 gene product, this study has shown that
expression of TPI requires the binding of a trans-acting protein
known as repressor activator protein 1 (RAP1) (Shore et al., 1987)/
general regulatory factor 1 (GRF1) (Buchman et al., 1988)/
translation upstream factor (TUF) (Huet et al., 1985), (Scott et al.,
1990). RAP1/GRF1/TUF will be referred to hereafter as RAP1.
RAP1 was first purified as a binding activity which bound the
HMR(E) silencer locus. RAP1 binding is required for repression of
transcription by the HMR(E) silencer (Shore and Nasmyth, 1987);
however, RAP1 binding is also required for the activation of
transcription for many genes (Huet et al., 1985; Capieaux et al.,
1989). Shore and Nasmyth (1987) demonstrated that two different
binding sites for RAP1 derived from two different UAS's are able to
restore HMR silencer function in vivo when substituted for the
native RAP1-silencer-binding site. When the native RAP1- silencer
binding site was destroyed the silencer no longer functioned and a 1


7
encoding alcohol dehydrogenase isozyme one (Buchman et al., 1988;
Santangelo and Tornow, 1990), and PDC1 encoding pyruvate
decarboxylase (Kellerman and Hollenberg, 1988). RAP1 binding has
also been reported to be required for the expression of EN02 (Brindle
et al., 1990). Recent studies with both temperature sensitive rap^
mutants and rap is mutants, selected for defective silencing of the
HMR locus, have been able to separate the suppression and activation
functions of the RAP1 protein (Kurtz and Shore, 1991; Sussel and
Shore, 1991). However, both sets of mutations map to the carboxyl
terminus of the RAP1 protein and no clear domains have been
defined.
RAP1 is a phosphoprotein (Tsang et al., 1990). RAP1 binding
in vitro to the PYK controlling region has been shown to be reduced
by phosphatase treatment (Tsang et al., 1990). However, in addition
to phosphorylation, RAP1 binding in the in vitro binding assay was
also dependent upon binding site context (Tsang et al., 1990). The
significance of RAP1 phosphorylation state in vivo is unknown.
The controlling regions of PGK, EN01, EN02, and PYK have all
been mapped. In the case of PGK the upstream activating sequence
has been shown to be comprised of three main elements (Stanway et


72
-853
-220
n I In
-392
n [~~lr
220
327 |
-377 I In'
-327
-377
-220
-220
TPI::lacZ
I TPI::lacZ
I TPI::lacZ
TPI::lacZ
B-Galactosidase Activity
S150-2B (GCR1)
Mean SD
HBY4 (gcr1)
Mean SD
161
8
9 1
115
7
18 5
116
9
12 5
14
3
10 3
45
4
21 7
82
4
21 7
36
7
20 5
-220
I TPI::lacZ
TPI::lacZ
27 11
9 1
28 4
8 2


116
approximately 400 bases of TPI non-coding region between UAStpi
and the URA3 locus. Nucleosomes could form in this region, and
without REB1 binding to prevent formation, UAStpi could be packaged
in a less active configuration.
Double mutations were introduced between the REB1-binding
site and either of the two GCR1-binding sites. If REB1 and GCR1
interacted in the expression of TPI, then mutation of the REB1-
binding site and either GCR1-binding site would have the same
effect. If REB1 and GCR1 do not interact then the results may vary
depending upon which GCR1 -binding site was mutated. Mutation of
the REB1-binding site and the CTTCC pentamer resulted in a 4-fold
reduction in expression. Mutating both the REB1 -binding site and the
CATCC pentamer resulted in an 8-fold reduction in expression. The
results demonstrated that the effects seen in the double mutants
retained much of their individual characteristics, varying depending
upon which GCR1-binding site was mutated. Thereby indicating that
REB1 and GCR1 do not interact at the protein level in the expression
of TPI.
The site-directed mutagenesis indicates an absolute
requirement for RAP1 and GCR1 for the expression of TPI. The REB1-


RESULTS
46
The Mature 5' Ends of Steady-State TPI Transcripts are
Unaffected by a gcr1 Mutation 46
Identification of the 5' Boundary of the TPI Controlling
Region 48
An Upstream Activating Sequence Activity for TPI
Resides from Position -377 to -327 in the 5'
Noncoding Region 59
Internal Deletions Indicate Single UAS Element
Responsible for TPI Transcription 63
Mutational Analysis of UASjpi 68
In vitro DNase I Protection Assays Reveal Binding of
the REB1 Site and the RAP1 Site 75
DNA Band Shift Assays Demonstrate REB1 Binding to
TPI 5' Noncoding Region 81
In vivo Methylation Protection Assays 82
Site-Directed Mutagenesis of Transcription Factor
Binding Sites in the UAS of TPI 90
DISCUSSION 99
REFERENCES 124
BIOGRAPHICAL SKETCH 136
v


64
of TPI. However, controlling regions in Saccharomyces cerevisiae
are often composed of multiple UAS elements, any of which are
independently capable of driving expression (Guarente et al., 1984;
Cohen et al., 1986; Cohen et al., 1987; Johnston, 1987; Nishizawa et
al., 1989). One-tailed deletion series are able to map only the last
element remaining which is sufficient for expression. In order to
determine the number of UAS elements present in the controlling
region of TPI, a second set of deletions was generated which began
internal to the known UASjpi at position -220 and extending towards
and through UAStpi located from -377 to -327. Figure 2 cartoons the
scheme utilized to generate the internal deletion series.
Deletions originating internal to and extending through UASjpi
of the TPI::lacZ fusion in plasmid pES90 ("internal" deletions) were
created originating from a unique Sph\ site at position -220.
Linearized plasmid pES90 was digested with exonuclease Bal 31,
followed by the addition of SphI linkers (HB06, table 2). To assure
an intact TATA element, required for expression, the deletion
products distal to the Sph\ site were isolated after H/ndlll digest by
gel purification then subcloned into plasmid pES90 which had been
digested with H/ndlll and Sph\. The final "internal" deletion series


58
gcr1 mutant HBY4 background only expressed 19 units of activity.
Thus the gcr1 mutation resulted in a 12 fold reduction in the level of
expression of the TPI::lacZ fusion. Clifton and Fraenkel (1981)
previously reported a 17 fold reduction in the level of triose-
phosphate isomerase in a gcr1 mutant background. Thus, expression
of the TPIr.lacZ, like TPI itself, is dependent on GCR1 for full
expression. Plasmids 92-9 and 92-9R, 35-2 and 35-2R contain the
same respective 5' deletion endpoint, but the TPIr.lacZ fusion was
integrated in opposite orientation with respect to URA3. No effect
on expression of the fusion due to orientation with respect to URA3
was observed. Strains harboring deletions up to position -392 still
express high levels of p-galactosidase activity. However, deletion
of an additional 15 bp to position -377 reduces expression
approximately two-fold to 107 units of activity. Deletion to -348
or beyond abolished expression of the fusion. Based on these
results, the 5' boundary of the region sufficient for high-level
expression of TPIr.lacZ must reside within 392 to 377 base pairs
from the start of the TPI structural gene, when the fusion is
integrated at the URA3 locus.


22
Primer extension
Total RNA was isolated from yeast strains S150-2B and HBY4
and used to determine the 5' end of the TPI transcripts. Total RNA,
10pg from wild-type and 25pg from gcr\ -deletion mutant, was
annealed with a radiolabeled primer HB05 which hybridized near the
5' end of the TPI structural gene. The primer was extended with 10
units of reverse transcriptase in the presence of dNTPs at 3mM. The
extension products were analyzed by denaturing PAGE. A sequencing
ladder generated from the 5' noncoding region of TPI with primer
HB05 was used as a molecular weight standard.
Plasmid Construction
Constructs for sequencing Single-stranded templates of the
5' noncoding region of TPI were generated for sequencing in the
following manner. M13mp18 and M13mp19 RF II double-stranded
DNA was prepared. Portions of the 5' noncoding region of TPI were
subcloned from plasmid pHB51 into M13mp18 or M13mp19 RF II DNA.
Single-stranded templates were prepared by standard techniques
(Sambrook et al., 1989), utilizing E. coli strain KK2186. Templates
were sequenced as above, utilizing the forward primer.


81
protected sequence revealed that it was a one base pair mismatch
from a consensus REBI-binding site proposed by Chasman et al.
(1990). REB1 has been implicated in nucleosome phasing in
transcriptional control elements (Brandi and Struhl, 1990).
However, it should be noted that the REB1 -binding site can be
deleted in the 5' deletion series with no effect on expression (pES34-
1, figure 6), when the fusion construct is integrated at the URA3
locus.
DNA Band Shift Assays Demonstrate REB1 Binding to TPI 5' Non-
Codina Region
The DNase I protection assays demonstrated an area of
protection over a region containing a one base pair mismatch from a
consensus REB1-binding site. DNA band shift assays were performed
to determine if REB1 binds the TPI 5' non-coding region. The target
fragment was an end-labeled 139 bp Ava\\-Fok\ fragment (positions
-487 to -348). This fragment contains the sequence protected in the
DNase I assays, as well as the near consensus REB1-binding site.
Others in the laboratory have demonstrated that a fragment
containing sequence from -392 to -348 is incapable of interacting


36
In vitro DNasel PrQtectiQn A??qy The method used in the in vitro DNasel protection assays was a
modification of the method of Singh et al. (1986). The 228 bp Hind\\\-
Sph\ fragment of pES40-23 or the 169 bp Hind\\\-Sph\ fragment of
pES34 were end-labeled by filling in the Hind\\\ site with the
Klenow fragment in the presence of 32p-dATP. A protein extract
was prepared from yeast S150-2B by lysis with a French pressure
cell at 20,000 psi. Rabbit reticulocyte lysates (RRL) containing
RAP1 protein were generated via in vitro transcription and
translation, and provided to me by C. Lopez.
Multiple aliquots of 2pl of the end-labeled fragments (20,000
cpm/ul) were incubated with either 5pl of yeast extract or 5pl RRL
containing RAP1 in 18pil of 1X binding buffer [12mM HEPES pH7.5,
60mM KCI, 5mM MgCl2, 4mM Tris, 0.6mM DTT, 10% glycerol, 0.26
ug/ul poly(dl-dC), and 0.3 |ig/|il BSA]. Incubation was for 20 minutes
at room temperature. 0.5 units or 1.0 units of DNase I was then
added to quadruplicate aliquots. The reactions were incubated at
room temperature for 2 minutes, then stopped by the addition of
10pl of Stop solution (0.25M EDTA, 25% glycerol) on ice. The entire


131
Mejean, C., Pons, F., Benyamin, Y., and Roustan, C. (1989). Antigenic
probes locate binding sites for the glycolytic enzymes
glyceraldehyde-3-phosphate dehydrogenase, aldolase, and
phosphofructokinase on actin momomer in microfilaments. Biochem.
J. 264, 671-677.
Miles, L.A., Dahlberg, C.M., Plescia, J., Felez, J., Kanefusa, K., and
Plow, E.F. (1991). Role of cell-surface lysines in plasminogen
binding to cells: identification of a-enolase as a candidate
plasminogen receptor. Biochem. 30, 1682-1691.
Miller, J.H. (1972). In Experiments in molecular genetics. Cold
Spring Harbor: Cold Spring Harbor Laboratory),
Morrow, B.E., Johnson, S.P., and Warner, J.R. (1989). Proteins that
bind the yeast rDNA enhancer. J. Biol. Chem. 264, 9061-9068.
Morrow, B.E., Ju, Q., and Warner, J.R. (1990). Purification and
characterization of the yeast rDNA binding protein REB1. J. Biol.
Chem. 265, 20778-20783.
Nakajima, N., Horikoshi, M., and Roeder, R.G. (1988). Factors involved
in specific transcription by mammalian RNA polymerase II:
purification, genetic specificity, and TATA box-promoter
interactions of TFIID. Mol. Cell. Biol. 8, 4028-4040.
Nishizawa, M., Araki, R., and Teranishi, Y. (1989). Identification of an
upstream activating sequence and an upstream repressible sequence
of the pyruvate kinase gene of the yeast Saccharomyces cerevisiae.
Mol. Cell. Biol. 9, 442-451.
Nishizawa, M., Suzuki, Y., Nogi, Y., Matsumoto, K., and Fukasawa, T.
(1990). Yeast Gall 1 protein mediates the transcriptional activation
signal of two different transacting factors, Gal4 and general
regulatory factor l/repressor/activator site binding protein
I/translation upstream factor. Proc. Natl. Acad. Sci. USA 87,
5373-5377.


34
URA3 r TPI' IEL
ura3
C.
c.
Sac I Sac I Sac I
UR A3 TPI' 'lacZ 'TPI
Sac I Sac I
t t
URA3 TPI'
lacZ 'TPI
Sac I
ura3
A.
B.
D.
B.
C.


54
within URA3. This procedure served to direct the site of integration
to the URA3 locus via homologous recombination (Orr-Weaver et al.,
1983). Transformation competent yeast cells can take up multiple
copies of plasmid DNA, raising the possibility that tandem
integration events can occur (see Figure 3). Therefore, screening for
unit copy integrants was required in order to assure the most
accurate expression data possible for the 5' deletion series.
Genomic DNA was isolated from individual transformants and
digested with Sacl, then subjected to Southern blot analysis probing
for URA3. Sacl cut outside of the URA3 locus but within the plasmid.
The URA3 probe hybridized to a single band in experiments with DNA
isolated from the parental strain, S150-2B. A unit copy integrant
was distinguishable by the presence of two fragments which
hybridized to the probe, labeled A and C in Figure 3. These are the
two junction fragments between the yeast chromosome and the
integrated plasmid. Upon a tandem integration a characteristic
band, totally plasmid derived, was also observed, labeled fragment D
in Figure 3. Presence of the totally plasmid derived fragment D was
used as a diagnostic for multiple integration events. Utilizing this


29
precise deletion endpoints were determined by double-stranded DNA
sequencing using primer HB05.
Constructs to assess UAS activity Plasmid plCZ312 is an
integrative yeast shuttle vector that contains a CYC1::lacZ fusion
(Guarente and Mason, 1983). The native CVC1 UAS elements were
removed by Hind 111-Sph\ digest and replaced with various fragments
containing portions of the TPI 5' noncoding region. All constructs
were confirmed by sequencing, using primer HB23 which anneals in
the vector sequence of plCZ312 immediately upstream of the
polycloning site. Plasmid pES90 was digested with H/ndlll and SphI
to remove sequences distal to position -220 of the TPI::lacZ fusion.
Such a digest removes sequence required for expression of the
fusion. Fragments containing portions of the TPI 5' noncoding region
distal to -220 were subsequently subcloned and tested for UAS
activity. Constructs were confirmed by sequencing with primer
HB01.
Mutant UAS oligonucleotides driving expression of the TPI::lacZ
fusion Double-stranded oligonucleotides capable of UASjpi
function, or mutant derivatives thereof, were generated via mutually
primed synthesis. These double-stranded oligonucleotides contained


LIST OF TABLES
£age
Table 1. Strains 18
Table 2. Oligonucleotides 21


UNIVERSITY OF FLORIDA


122
TPI
Figure 18
Model of protein interactions at UASypi.


Figure 6. Effect of 5' deletions upon expression of a TPI::lacZ fusion integrated in unit copy at the
URA3 locus in wild-type and gcrf-deletion mutant strains. Constructs 92-9, 92-9R, and 35-2, 35-2R
contain the same respective deletion endpoints but are integrated in opposite orientation at the URA3
locus. Deletion endpoints are indicated with respect to the start of translation, p-galactosidase
activities were determined by the method of Miller (1970) in duplicate from at least three independent
cultures. Strains were grown to an optical density A6oo of approximately 1.0 in YP medium
supplemented with either 2% glycerol and 2% lactate (YPGL), or 2% glucose (YPD).


82
with wild-type yeast extracts is DNA band shift assays ( H. Baker,
personal communication). Extracts were prepared from the wild-
type yeast (S150-2B), or E. coli expressing REB1 from a REB1 insert
in the expression vector pET11A, and E. coli with pET11A and no
insert (kindly provided by B. Morrow and J. Warner). Results of the
DNA band shift assays are shown in Figure 13.
The wild-type yeast extract and the E. coli extract containing
REB1 gave rise to positive band shift assays when incubated with
the fragment from -487 to -348. The E. coll extract without the
REB1 insert failed to band shift. These results indicate that REB1 is
capable of binding a region of the 5' non-coding region of TPI that
contains a near consensus REB1-binding site.
In vivo Methvlation Protection Assays
The DNase I protection and DNA band shift assays were able to
provide information about the ability of the trans-acting factors
RAP1, REB1, and GCR1 to bind to the TPI controlling region under in
vitro conditions. The ability to demonstrate an area of protection
indicated that the factors have the ability to interact with specific
sites, but, did not prove that the interactions occur in vivo. To


9
CTTCC pentamers has wild type levels of expression even without
the strong ABF1-binding site in YfP (Ogden et al., 1986). Deletion of
the CTTCC pentamer from -432 to -428 reduces expression by 50%,
deletion of two of the repeats (-432 to -428 and -449 to -445)
reduces expression 75%, and removal of all three CTTCC motifs
reduces expression to less than 10% of wild-type (Chambers et al.,
1988). As seen with other glycolytic genes, full expression of PGK
depends on GCR1.
Both EN01 and EN02 have two upstream activating sequence
elements (Cohen et al., 1987). EN01 also contains an upstream
repressor sequence (URS) responsible for repression of EN01 in
nonfermentable carbon sources. The UAS1 of both EN01 (Buchman et
al., 1988; Brindle et al., 1990) and EN02 bind RAP1 (Brindle et al.,
1990). In the case of UAS1 of EN01 a CTTCC pentamer has been
identified. The pentamers presence has been shown to be required
for full UAS1 activity, but its absence does not affect the ability of
RAP1 to bind in vitro (Buchman et al., 1988). UAS2 of EN01 binds a
factor designated EBF1, Enolase Binding Factor 1. UAS2 of EN02
binds ABF1 and overlaps UAS1. In EN02 RAP1 binding to UAS1 and
ABF1 binding at the overlapping UAS2 may be competitive (Brindle et


41
weight markers. The gel was electrophoresed in a 50mM NaOH,
1.0mM EDTA running buffer at 30 volts and 175 milliamps overnight.
The gel was neutralized in 1M Tris pH 7.5 for one hour, stained with
0.5ug/ml ethidium bromide, and visualized with UV light. Samples
generated a smear with a definite peak size of fragment. Time
points with a peak fragment size of approximately 500-800 bp were
used for subsequent steps.
Control sequences were generated from the 200pg of Avail
digested untreated DNA. The DNA was divided into four aliquots and
ethanol precipitated, washed twice with 70% ethanol, and dried.
Control G-reactions were as follows. One aliquot was resuspended
in 200|il DMS buffer (50mM Na-cacodylate, 1.0mM EDTA pH 8.0). 1 pil
of DMS was added for 1.5 minutes at room temp. 50pl ice-cold DMS
stop solution was added (1.5M NaAc pH 7.0, 1M 2-mercaptoethanol),
and 750pl ice-cold ethanol was added. The reaction mixture was
then placed on dry ice.
The control A+G-reactions were as follows. The DNA aliquot
was resuspended in 11 pi ddH2. 25pl of concentrated formic acid
was added at room temperature and incubated 5 minutes. 200pl ice-


DEDICATION
To my wife Rochelle for her love and support in all things I do.
Father, first and foremost this belongs to you. Your rearing has
produced my thirst for knowledge about all I touch. Your example
demonstrates that education is not a stagnant entity cast off after
college, rather, it is a lifelong and enlightening endeavor.
Why, all delights are vain; but that most vain,
Which with pain purchased doth inherit pain:
As, painful to pore upon a book
To seek the light of truth; while truth the while
Doth falsely blind the eyesight of his look:
Light seeking light doth light of light beguile:
So, ere you find where light in darkness lies,
Your light grows dark by the losing of your eyes.
Love's Labour's Lost
William Shakespeare


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
EXPRESSION OF THE TP/GENE OF SACCHAROMYCES CEREVISIAE\S
CONTROLLED BY A SINGLE COMPLEX UPSTREAM ACTIVATING SEQUENCE
CONTAINING BINDING SITES FOR THREE TRANS-ACTING FACTORS:
REB1, RAP1, AND GCR1
By
Edward William Scott V
May 1992
Chairman: Henry V. Baker, Ph. D.
Major Department: Immunology and Medical Microbiology
In Saccharomyces cerevisiae the enzymes of glycolysis
constitute 30-60 percent of the soluble protein. GCR1 gene function
is required for high-level glycolytic gene expression. This study
involved a biochemical and genetic characterization of TPI, a gene
affected by gcr1 lesions. Primer extension experiments on TPI
transcripts isolated from wild-type and gcr1 mutant strains mapped
the initiation or "I" site of transcription to a pair of adenines 29 and
30 bases upstream of the start of translation. To delineate the TPI


Nucleic Acid Manipulations
Techniques used throughout this study are derived from the
standard reference manuals (Sambrook et al1989; Sherman et al.,
1983; Current Protocols in Molecular Biology, 1989) except for
deviations noted. S. cerevisiae DNA and RNA was prepared by the
methods of Sherman et al. (1983) and Struhl and Davis (1981),
respectively.
Generation of Double-Strand DNA Oligonucleotides
Most double-stranded oligonucleotides were generated by a
modification of the method of Oliphant et al. (1986). Single-
stranded oligonucleotides were synthesized (University of Florida
Interdisciplinary Center for Biotechnology Research) with the
desired sequence flanked by restriction sites, typically a H/ndlll
site on the 5' end and a SphI site on the 3' end. One restriction site
was made into an 8-10 bp palindrome by the addition of G and C
residues. The oligonucleotides were self-annealed via the
reinforced palindrome and the 3' ends extended with Klenow
fragment of DNA polymerase I in the presence of dNTPs. The double-
stranded extension products were then gel purified via


102
detected in the gcr1 mutant strain for each of the transcripts
tested. Transcription of the HIS4 gene shows a similar pattern of
transcript levels when the gcn4 gene is mutated (Arndt et al., 1987).
The HIS4 gene has been shown to have two promoter elements, one
dependent upon GCN4 and responsible for high-level expression, and
one dependent upon BAS1/BAS2 and responsible for basal level
transcription. Each HIS4 promoter element gives rise to transcripts
that have different sites of initiation.
Primer extension experiments (Figure 4) revealed that unlike
these other systems, the TPI transcript has one major mature 5' end.
The predominant start site is unaffected by mutations in GCR1. In
both wild-type and gcr1 mutant strains the mature 5' end of the TPI
transcript mapped to a pair of adenines 29 and 30 base pairs
upstream of the start of translation, mapping an initiation (i) site of
a single promoter element for TPI. The single promoter requires
GCR1 for full expression of TPI, but is still capable of driving the
low level expression seen in the gcr1 mutant strain.
Mutational analysis of the 5' noncoding region of TPI was
carried out utilizing a TPI::lacZ gene fusion in trans to TPI. Use of
the fusion allowed normal TPI activity required for cellular growth.


110
The fact that such interactions are possible, particularly when
adding high-levels of a purified protein such as the RRL- RAP1 and
MBP-GCR1 footprints, was not proof that the interactions were
occurring with the cell. DMS methylation protection assays
(genomic footprinting) were carried out on several strains of
Saccharomyces cerevisiae to analyze the in vivo protein-DNA
interactions within the TPI controlling region. The composite
summary, Figure 17, indicates the position of the protected bases in
the wild-type strains, S150-2B and DFY 642. Portions of all sites
protected in vitro from DNase I digestion are protected from DMS
methylation in vivo. The REB1 and RAP1 -binding sites are bound,
presumably by REB1 and RAP1 respectively. The CTTCC and CATCC
pentamers are also bound.
The genomic footprint of the TPI controlling region in the
isogeneic gcr1 deletion strain, HBY4, confirms that GCR1 is
responsible for the protection seen at the pentamer motifs. In the
gcr1 strain, the REB1 and RAP1-binding sites are still occupied, but,
neither the CTTCC nor the CATCC pentamer is protected, see Figure
14. The GCR1 dependent protection of the pentamer motifs, coupled
with the in vitro DNase I protection of those sites by MBP-GCR1, and


123
ablates expression, Figure16. RAP1 is an abundant protein (Shore
and Nasmyth, 1987) that is known to play a role in both silencing and
activating transcription (Shore and Nasmyth, 1987). RAP1 is also
associated with the nuclear matrix (Cardenas et al., 1990) and
causes bending of DNA (Vignais and Sentenac, 1989). Therefore, it
seems likely that RAP1 also provides a critical DNA structure or
configuration required for expression. The specificity of the
individual transcriptional elements is probably provided by factors
such as GCR1. GCR1 binds UAStpi at the CTTCC and CATCC pentamer
motifs. The binding affinity of GCR1 to naked DNA appears to be
several orders of magnitude less than the binding of RAP1 to naked
DNA (Baker, 1991). However, the GCR1-binding sites are almost
fully occupied in vivo, Figure 14. It is likely that the action of RAP1
binding facilitates the binding of GCR1. Given that the binding of
RAP1 and GCR1 appear to be on opposite sides of the helix this
facilitation of binding may not involve direct protein-protein
interactions. Rather, GCR1 may preferentially bind its recognition
site in a bent confirmation. In this model GCR1 is responsible for
the activation signals that eventually lead to the recruitment of RNA
polymerase and the initiation of transcription.


126
Chambers, A., Stanway, C., Kingsman, A.J., and Kingsman, S.M. (1988).
The UAS of the yeast PGK gene is composed of multiple functional
elements. Nucleic Acids Res. 16, 8245-8260.
Chambers, A., Stanway, C., Tsang, J.S.H., Henry, Y., Kingsman, A.J., and
Kingsman, S.M. (1990). ARS binding factor 1 binds adjacent to RAP1
at the UASs of the yeast glycolytic genes PGK and PYK1. Nucleic
Acids Res. 18, 5393-5399.
Chambers, A., Tsang, J.S.H., Stanway, C., Kingsman, A.J., and
Kingsman, S.M. (1989). Transcriptional control of the Saccharomyces
cerevisiae PGK gene by RAP1. Mol. Cell. Biol. 9, 5516-5524.
Chasman, D.I., Le, N.F., Buchman, A.R., LaPointe, J.W., Lorch, Y., and
Kornberg, R.D. (1990). A yeast protein that influences the chromatin
structure of UASq and functions as a powerful auxiliary gene
activator. Genes Dev. 4, 503-514.
Clifton, D. and Fraenkel, D.G. (1981). The gcr (glycolysis regulation)
mutation of Saccharomyces cerevisiae. J. Biol. Chem. 256,
13074-13078.
Clifton, D., Weinstock, S.B., and Fraenkel, D.G. (1978). Glycolysis
mutants in Saccharomyces cerevisiae. Genet. 88, 1-11.
Cohen, R., Holland, J.P., Yokoi, T., and Holland, M.J. (1986).
Identification of a regulatory region that mediates
glucose-dependent induction of the Saccharomyces cerevisiae
enolase gene EN02. Mol. Cell. Biol. 6, 2287-2297.
Cohen, R., Yokoi, T., Holland, J.P., Pepper, A.E., and Holland, M.J.
(1987). Transcription of the constitutively expressed yeast enolase
gene EN01 is mediated by positive and negative c/s-acting
regulatory sequences. Mol. Cell. Biol. 7, 2753-2761.


8
al., 1989). The central portion of the UAS contains the activator
core (AC) to which RAP1 binding has been demonstrated (Chambers
et al., 1989). Upstream of the activator core is a region designated
YfP, first recognized as a site of strong DNasel protection (Chambers
et al., 1988). The Yfp region binds ARS binding factor 1 (ABF1),
another trans-acting binding protein similar in properties to RAP1
(Chambers et al., 1990; Buchman and Kornberg, 1990). The third
region identified in the UAS of PGK consists of three repeats of a
pentamer sequence motif, CTTCC (Ogden et al., 1986). In
experiments with fragments from the PGK promoter which carry
both the RAP1 binding site and the CTTCC sequence motif, there is
DNasel protection of the RAP1 -binding site and evidence of
heightened DNasel sensitivity at the CTTCC pentamer motif
(Chambers et al., 1988). However, no clear footprint was observed.
Chambers et al. (1988) suggested that the heightened DNase I
sensitivity at the CTTCC sequence element was due to interactions
with RAP1. Of the three main elements of the PGK UAS, no single
element is able to activate expression alone (Stanway et al., 1989).
The activator core with the YfP is able to drive about one-third the
wild-type levels of expression. The activator core with the three


21
Table 2,
OLIGONUCLEOTIDES
Name
Sequence
Forward
5'-GT AAAACGACGGCCACT-3'
HB01
5-AT GT GT GGAATTGT GAGCGG-3'
HB03
5'-C AAGCTT GT CGACAAGCTTG-3'
HB05
5-CCACCGACAAAGAAAGTT CT AGCC-3'
HB06
5'-AC AT GCAT GCAT GT -3'
HB07
HB08
ETC AAGCTTGTT CTAAGACTTTTC AGCTT CCTCTATT GATGTTAC ACTTGGACGC ATGCC-3'
5-GGCATGCGT CCAAGTGTAACATCAATAGAGGAAGCTGAAAAGTCTTAGAACAAGCTTG-3'
HB09
5-GGCAT GCCAAC AT GT ATGGGTT CCAAGCTTG-3'
HB11
5-GGCAT GCCAACCTGAT GGGTTCCAAGCTT G-3'
HB14
HB15
5-GGAAGCTTAGCTTCCTCTATT GAT GTTACACCT GGACACCCCTTTTCT GGCATCCAGTTGCATGCC-3'
5-GQCATGCAACTQGATGCCAGAAAAGGGGTGTCCA3GTGTAACATCAATAGAGGAAGCTAAGCTTCC-3'
HB16
5-GGAAGCTTGTTCTAAGACTTTTCAGCTT CCTCTATT GATGTTACACCT GGACACCCCTTTT CT GGCAT CCA
GTTGCATGCC-3'
HB17
5-GQCATQCAACTQGATQCCAGAAAAGQQGTGTCCAGGTGTAACATCAATAGAGGAAQCTGAAAAGTCTTAGA
ACAAGCTTCC-3'
HB19
5-GCTAAGCTTAGCTT CCT CT ATT GAT GTT ACACCTGGAAACCCATCAGGTGGCATCCAGTT GCATGCAAC-3'
HB21
Ei-GGGGCATGCAACTGGATGCCAGAAAAGGQGTGTCCAGGTGTAACAT CAATAGAGGTTGCTAAGCTTAGG3'
HB22
5-GGQQCATGCAACTGGTTGCCAGAAAAGGQGTGTCCAQGTGTAACATCAATAGAGGAAGCTAAGCTTAGC-3'
HB23
5'-CGCAAACCGCCT CTCC-3'
HB24
5-GCTAAGCTT AGCTT CCT CT ATT GAT GTT ACACCTGGAGATATCTGCAGTGGCAT CCAGTTGCATGCAAC-3'
HB25
5-GCTAAGCTTAGCAACCTCTATTGATGTTACACCTGGACACCCCTTTTCTGGCAACCTGTTGCATGCAAC-3'
HB28
5-GTTACACCTGGAGAT AT CT GCAGT GGCATCCAG-3'
HB29
5-GACTTTT CAGC AACCT CT ATT-3'
HB30
5'-CCTTTTCT GGCAAACAGTTTTT AATC-3
HB31
5-GGAAGATT GAACGTTCT AAG-3'
HB32
5-GGAACCCATAC AT GTT GGT GGAAG-3'
HB33
5-GT AACAGGGAAGCGAAAGGGCAGC-3'
HB42
5-CTGTGAGGACC-3'
HB53
5-GATGTTAG ACCAGATTACCCGTT CT CT GGCAT CCAG-3'
HB54
5-GACCTT AAT ACATT C AG-3'
HB60
5-GCATTAGCATGCGTAACAAACCACC-3'


controlling region, a TPI::lacZ gene fusion was employed. Nuclease
Bal31 deletion analysis identified a single upstream activating
sequence, UASjpi, responsible for high-level TPI expression. DNase I
protection and in vivo dimethyl sulfate methylation protection
assays indicated the binding of three trans-acting factors to four
sites within UAStpi. GCR1 binds two sites: 1) a sequence element
which functionally requires the CTTCC pentamer from position -375
to -370; 2) an element requiring the CATCC pentamer from position
-335 to -330. REB1 binds to an almost perfect consensus binding
site from position -397 to -387. RAP1 binds from -358 to -346.
REB1 and RAP1 binding at UASjpi are independent of GCR1 binding in
vivo. Site-directed mutation of the REB1-binding site reduced the
expression of the TPI::lacZ gene fusion by two-fold. Mutation of the
RAP1 or both GCR1-binding sites abolished expression. Thus, TPI
absolutely requires RAP1 and GCR1 binding for expression and
requires REB1 for full expression. This work suggests a mechanism
of high-level glycolytic gene expression mediated primarily through
the actions of RAP1 and GCR1.
x


134
Struhl, K. (1988). The JUN oncoprotein, a vertebrate transcription
factor, activates transcription in yeast. Nature 332, 649-650.
Struhl, K. and Davis, R.W. (1981). Transcription of the his3 gene
region in Saccharomyces cerevisiae. J. Mol. Biol. 152, 535-552.
Sussel, L. and Shore, D. (1991). Separation of transcriptional
activation and silencing functions of the F?AP7-encoded
repressor/activator protein 1: Isolation of viable mutants affecting
both silencing and telomere length. Proc. Natl. Acad. Sci. USA 88,
7749-7753.
Taylor, I.C.A., Workman, J.L., Schuetz, T.J., and Kingston, R.E. (1991).
Facilitated binding of GAL4 and heat shock factor to nucleosomal
templates: differential function of DNA-binding domains. Genes Dev.
5, 1285-1298.
Taylor, J.W., Ott, J., and Eckstein, F. (1985). The rapid generation of
oligonucleotide-directed mutations at a high frequency using
phosphorothioate-modified DNA. Nucleic Acids Res. 13, 8765-8785.
Tsang, J.S.H., Henry, Y.A.L., Chambers, A., Kingsman, A.J., and
Kingsman, S.M. (1990). Phosphorylation influences the binding of the
yeast RAP1 protein to the upstream activating sequence of the PGK
gene. Nucleic Acids Res. 18, 7331-7337.
Uemura, H. and Fraenkel, D.G. (1991). gcr2, a new mutation affecting
glycolytic gene expression in Saccharomyces cerevisiae. Mol. Cell.
Biol. 10, 6389-6396.
Vignais, M.L. and Sentenac, A. (1989). Asymmetric DNA bending
induced by the yeast multifuntional factor TUF. J. Biol. Chem. 264,
8463-8466.
Walsh, J.L., Keith, T.J., and Knull, H.R. (1989). Glycolytic enzyme
interactions with tubulin and microtubules. Biochim. Biophys. Acta
999, 64-70.


Figure 10. Mutational analysis of UAStpi utilizing mutant
oligonucleotides. Double-stranded oligonucleotides with portion of
the 5' noncoding region of TPI were cloned before a TPIr.lacZ fusion
at position -220. Positions within the 5' noncoding region are
indicated. The large box denotes the RAP1-binding site from
position -358 to -346. The 5' small box denotes the CTTCC pentamer
motif located from -375 to -370. The 3' small box denotes the
CATCC pentamer motif from position -335 to -330. Filled boxes
represent mutated motifs as detailed in the text. Unit copy
integrants of the various constructs in wild-type (S150-2B) or gcr1
mutant (FIBY4) yeast were assayed for p-galactosidase activity.
Assays were performed in duplicate on three sperate occasions.


135
Wang, H., Nicholson, P.R., and Stillman, D.J. (1990). Identification of
a Saccharomyces cerevisiae DNA-binding protein involved in
transcriptional regulation. Mol. Cell. Biol. 10, 1743-1753.
Warburg, O. (1930). The metabolism of tumors (London: Constable).
Yu, J., Donoviel, M.S., and Young, E.T. (1989). Adjacent upstream
activating sequence elements synergistically regulate transcription
of ADH2 in Saccharomyces cerevisiae. Mol. Cell. Biol. 9, 34-42.
Zagursky, R.J. and Berman, M.L. (1984). Cloning vectors that yield
high levels of single-stranded DNA for rapid DNA sequencing. Gene
27, 183.


ACKNOWLEDGMENTS
I would like to extend my deepest appreciation to my mentor
Dr. Henry V. Baker for providing superb direction to my graduate
education and personal development as a scientist. My deepest
regards to the members of my dissertation committee: Dr. Lonnie O.
Ingram, Dr. Alfred S. Lewin, Dr. Richard W. Moyer, and Dr. Thomas C.
Rowe for their invaluable insights. I thank the Department of
Immunology and Medical Microbiology for providing a climate
conducive to intellectual growth and exchange.


Fusion # P- Galactosidase Activity
- -3500
-220
Sph I
l
4-1
ATG
Ltpi tsa
S150-2B (GCR1)
YPGL YPD
Mean SD Mean SO
HBY4 (gcr1)
YPGL
Mean SD
92-9
HacZ [ m-TPt ESI
276
11
222 6
19 3
- -3500
254
215 21
17 1
92-9 R
y| ip r fe:!
llacZ E ESlTPI EiJ
4
-853
210
236 25
21 4
35-2
htpi' ta
lacZ 1 BSaTPI ESI
11
-853
276
251 8
ND
35-2R
ESI TP r M
HacZ E BSJ'TPI ES
18
36-2
-490
241
238 33
22 5
ta tp! m
llacZ [ fSHTPl m
36
37-2
-420
266
255 43
20 3
KTPr ES3
1 acZ t &1TPI
9
34-1
-392
251
241 24
20 2
l&TPI' M
: llacZ EiilHTPI M
36
70-1
-377
107
69 4
29 3
EiiTPi m
llacZ t m'TPl Eil
8
-348
27
21 3
18 3
83-1
TPI' ESI
HacZ t ElTPI ESI
7
73-2
-337
19
htpi' m
HacZ 1 BsJTPI m
8
9 6
4 4
-330
16
82-1
^tpi !:'H
HacZ 1 fcil'TPI Eil
9
6 3
3 3
-299
14
13 4
77-2
htpi' ea
HacZ t S'TPI M
3
6 3
78-1
-278
11
8 4
10 8
I3TPI' Esa
HacZ 1 ffl 1P ESI
10
-192
36
10 7
ND
76-1
ESI TP r m
HacZ | E3TPI E43
11
-179
17
12 1
38-4
BiTPr m
HacZ ¡ E2ITPI m
11
5 4
39-2
+63
14
8 8
B'TPI 1
HacZ | HiHTPl Hi!
7
4 3
No Fusion
9
3
4 3
3 3


Figure 8. Effect of internal deletions on p-galactosidase activity expressed from a TPI::lacZ gene
fusion. The cartoon depicts the extent of the internal deletions with precise deletion endpoints
indicated. (3-galactosidase assays were performed in duplicate by the method of Miller (1972) on three
individual occasions. Plasmids with deletions in the 5' noncoding region of TPI were integrated in unit
copy at the UR A3 locus in both wild-type (S150-2B) and gcr1 mutant (HBY4) strains of yeast.


105
ADH1 (Buchman et al., 1988; Santangelo and Tornow, 1990).
Although, the existence of a RAP1-binding site and UAS element in
the region of the EN02 5' noncoding region suggested by the Holland
group appears doubtful. J. Anderson demonstrated that the region
from -491 to -443 of the 5' noncoding region of EN02 was unable to
bind RAP1 or act as a UAS before a TPI::lacZ fusion. Chambers et al.
(1989) suggested a putative RAP1-binding site within the 5'
noncoding region of TP I from position -420 to -410. The importance
of this site was brought into question by the 5' deletion studies
when construct 34-1 with an endpoint of -392 was still able to
drive high-level expression, see Figure 6. Subsequent DNA band shift
experiments, performed by others in the laboratory, demonstrated
that -420 to -410 was not a RAP1-binding site (Scott et al., 1990).
A CTTCC sequence motif is also found in UASypi. CTTCC pentamer
motifs have been demonstrated to play a role in the UAS activity of
PGK (Ogden et al., 1986; Stanway et al., 1989), EN01 (Buchman et al.,
1988), and PYK (Buchman et al., 1988; McNeil et al., 1990). UASjpi
contains a CTTCC from -375 to -370, and a closely related CATCC
sequence from -335 to -340.


Figure 14. Genomic footprinting of the bottom strand of the TPI 5'
noncoding region. The initial four lanes are genomic sequencing
reactions of the TPI 5' noncoding region. The G sequencing ladder
serves as the control ladder for the genomic footprinting reactions.
Lanes 1-4 are genomic footprinting reactions carried out in wild-
type (S150-2B) yeast. The reactions in lanes 1 and 2 were generated
in wild-type yeast grown in YP media supplemented with 2% glucose
(YPD) and represent 3 and 4 minutes of DMS treatment, respectively.
Lanes 3 and 4 were grown in YP media supplemented with 2%
glycerol and 2% lactate and represent 4 and 5 minutes of DMS
treatment, respectively. Lanes 5 and 6 are the products of 4 and 5
minutes of DMS treatment in a gcr1 (HBY4) mutant strain of yeast.
Lanes 7 and 8 are similar treatment of a second wild-type strain
(DFY642). Lanes 9 and 10 are similar treatment of a spt13 (JF1052)
mutant strain. Guanine residues protected within the GCR1 -binding
site are denoted by an (*). Residues protected within the RAP1-
binding site are denoted by (O). Residues protected within the REB1-
binding site are denoted by (A). The right portion of the figure
depicts the double stranded sequence of the TPI 5' noncoding region.
The bottom strand is the righthand most strand as depicted. The
sequence motifs known to play a role in TPI expression are stippled.
Protected guanine residues are denoted as above.


the genetic evidence for the importance of CTTCC and CATCC
pentamers in TPI gene expression, proves that GCR1 acts through the
CTTCC and CATCC pentamer motifs to drive expression of TPI.
GCR1 binding is not required for the binding of either REB1 or
RAP1, as is clearly demonstrated by the continued occupation of
those sites in the gcr1 mutant strain. The presence of CTTCC
pentamer motifs that have been genetically identified to play a role
in the expression of PGK (Stanway et al., 1989), PYK, and EN01
(Buchman et al., 1988) suggests that GCR1 acting through those
pentamers as well. Indeed GCR1 binding has been demonstrated to
the CTTCC elements in the 5' noncoding regions of PGK, EN01, PYK,
and ADH1(y\uie et al., 1992). This suggests that GCR1 acts through a
CTTCC or closely related pentamer motif, such as CATCC, in all
genes affected by mutations in GCR1.
The genomic footprints of the TPI controlling region generated
in a spt13 mutant strain was identical to wild-type. Therefore
neither the spt13 mutation does not affect the binding of REB1,
RAP1, or GCR1. Furthermore, there is no evidence that the SPT13
gene product binds the TPI controlling region to a sufficient degree


109
-430
37-2 [100%]
'(-420)
34-1 [95%]
|(-392)
t3
70-1 [27%]
(-377)
x
f AT AT CTAGGACCCAT CAGGTT GGTGGAAG ATT ACCCGTT CT AAG ACTTTT CGCTTCCT CTAT
AAAA A A
REB1-DNasel
Footprint
(-377)
GCR1-DNasel Footprint
83-1 [8%]
(-348)
73-2 [4%]
(-337)
TGATGTTACACCTGGACACCCCT
77-2 [2%]
(-314)
AA*
A Ai
r CTGGCAT CCAGT
AA
f AAT CTT CAGT GGCAT GT GAG ATT C
RAP1-DNasel Footprint
(-327)
GCR1-DNasel Footprint
TCCG AAATT A ATT AAAGC AAT C ACACAATT CT CTCGGATACC ACCTCGGTT G AAACT GACAGGT G
-234
I
RAP1-DNasel Hypersensitivity
Figure 17. Composite summary of the TPI controlling region. Bold
numbers denote 5' deletion series used to map the TPI UAS element.
Position of deletion endpoints are denoted in parentheses. The
effect of each deletion on expression of TPI::lacZ is indicated in
brackets. The sequence underlined from -377 to -327 was able to
confer UAS activity to a CYC1::lacZ fusion. Core binding sites are
denoted as follows: REB1 (), GCR1 (), and RAP1 (_). Mismatches
from consensus binding sites are denoted with an X. Stippling
demarcates areas identified by in vitro DNasel footprinting studies,
as indicated on the figure. denotes G residues protected in cells
treated with DMS on the DNA strand opposite the strand depicted in
the figure. A denotes G residues protected on the DNA strand
depicted.


39
cells per ml in PBS (35ml final volume). 5ml aliquots were placed in
50ml disposable, sealable plastic tubes on ice. Dimethyl sulfate
(DMS) was added to 0.5% final concentration to various aliquots (5
pl/aliquot), and incubated from 1-6 minutes at room temperature.
Reactions were quenched by the addition of 45ml of ice-cold PBS.
Treated cells were harvested by centrifugation (1,000x g for 10
min.) and washed twice in 35ml ice-cold PBS. DMS waste was
allowed to decay in a fume hood for one week prior to disposal.
Methylated genomic DNA was prepared essentially by the
method of Sherman et al. (1983), except all incubations were carried
out at 370C or less to prevent unwanted cleavage events. The DNA
preparation was treated with RNase A at a final concentration of
50pg/ml for one hour. Two phenol extractions, one
phenol/chloroform extraction, and one chloroform extraction were
performed. 1/10 volume of 5M NH4AC was added, followed by
precipitation with 2 volumes of ethanol. DNA was pelleted by a low
speed spin (500x g, 5 min.), washed twice with 70% ethanol, and
dried in vacuo. The dried pellets were resuspended in 250 pi 10mM
Tris pH 8.0, 1.0 mM EDTA (TE); and the A260/280 was determined.


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.
Henry V. Baker, Chair
Assistant 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 a dissertation for the degree
of Doctor of Philosophy.
Lonnie O. Ingram
Professor of Microbiology and Cell
Science
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.
ml
Alfred Lewin
Associate 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 a dissertation for the degree
of Doctor of Philosophy.
-
Richard W. Moyer
Professor and Chaifrmjan of Immunology
and Medical Microbiology


107
Constructs that mutated the pentamer motifs demonstrated
that they are required for full UAS activity. Mutating the CTTCC
reduced expression 2.5-fold, and mutating the CATCC reduced
expression 33%. Therefore, UAStpi demonstrates an absolute
requirement for the RAP1-binding site from -358 to -346 for
activity, and requires both the CTTCC and the CATCC pentamers for
full activity. The experiments mutating UAStpi in isolation would
indicate that the CTTCC pentamer plays a greater role in expression
than the CATCC pentamer. However, while experiments with
isolated elements can indicate an motif plays a role in expression,
caution should be observed in ordering the overall importance of
individual motifs.
In addition to indicating that certain sequence motifs play a
role in UAStpi activity, the constructs driven by UAStpi were also
shown to respond to a gcr1 mutation with a 10 fold reduction in
expression. Therefore, GCR1 acts either through UAStpi or through
sequences proximal to position -220. The pentamer motifs were a
good candidate for the site of GCR1 action, as they were a common
motif important in the expression of many of the genes affected by
lesions in gcr1 (Stanway et al., 1989; Buchman et al., 1988). The


45
gel. The gel slice was crushed and 8ml of hybridization buffer was
added.
The pre-hybridization solution was removed from the
membrane and the gel/hybridization solution mixture added to the
roller tube. Hybridization was performed at 630C overnight.
Following hybridization, the membrane was washed in 500ml
aliquots of (40mM Na) HPO4, 1.0mM EDTA, 1% SDS at 60OC until wash
was no longer radioactive when examined with a geiger counter. The
damp membrane was wrapped in plastic wrap and exposed to film for
visualization.


Figure 4. Primer extension analysis of the 5' ends of the TPI
transcript in wild-type and grcri-deletion mutant strains. (A) Lane
1, extension products generated with reverse transcriptase from
10pg of total RNA isolated from wild-type strain S150-2B
hybridized with a radiolabeled primer corresponding to the start of
the TPI structural gene. Lane 2, extension products from similar
reaction carried out with 25pg total RNA from gcr1 mutant strain
HBY4. Lanes 3 through 6, DNA sequencing ladder generated with the
aforementioned primer from the TPI 5' nontranslated region. Bold
double arrow indicates the major extension products. P, denotes the
position of the unextended primer. (B) DNA sequence of the TPI 5'
nontranslated region. Double arrows indicate the bases that
correspond to the predominant 5' ends of the TPI transcript.


55
screen, unit copy integrants were obtained for all constructs in
S150-2B (GCR1) and all but two constructs in HBY4 (gcr1).
The stability of the fusion constructs integrated at URA3 was
determined by growing a strain carrying an integrated construct in
non-selective media (YPD) for >10 generations. The cells were then
plated on non-selective media and 100 individual colonies were
screened for their ability to grow on selective media (YNB with Glu,
His Leu, Trp). Each colony tested was URA+. Furthermore, each
colony still expressed p-galactosidase activity, which indicated the
presence of the integrated fusion construct. Integration into the
yeast genome was a stable method for carrying the fusion with a
segregation rate of less than one percent.
Once strains were isolated that carried the 5' deletion series
integrated in unit copy, they were assayed for p-galactosidase
activity in order to determine the effect of the deletions on the
expression of the TPI::lacZ gene fusion, p-galactosidase assays were
carried out in duplicate on three separate occasions. Figure 6
depicts the assay results. Plasmid 92-9 contains the entire 3.5 kb
of 5' non-coding region. When integrated in the S150-2B background,
276 Miller units of activity were expressed. 92-9 integrated in the


90
Figure 15 depicts the genomic footprint of the TPI controlling
region probed for the bottom strand, as defined in the figure, in both
wild-type and gcr1 mutant strains of yeast. Only one area of
protection was seen in both strains. The guanine at position -392
was protected from methylation. Position -392 corresponds to the
last base of the near consensus REB1-binding site.
Site-Directed Mutagenesis of Transcription Factor Binding Sites in
the UAS of TPI
Once the sites bound by transcription factors had been
identified, it was important to mutate those sites in order to assess
their role in TPI gene expression. Site-directed mutagenesis was
utilized to introduce mutations in single and pairwise combinations
of factor-binding sites before the TPI::lacZ gene fusion construct.
Plasmid pES90, used as a target for mutagenesis, contained 853
base pairs of material 5' to the start of the structural gene. This
construct was used to insure that the mutations were in a context
as close as possible to the native loci.
Figure 16 depicts the mutations made in the TPI controlling
region and the results of the p-galactosidase assays performed. The


, FRAG. ALONE
E. coli/ p11A
E. coli/ p11A-REB1
j S150-2B
oo


elements (Fedor et al., 1988; Chasman et al., 1990; Brandi and
Struhl, 1990), and, aids in gene activation in conjunction with
additional transcription factors (Chasman et al., 1990; Wang et al.,
1990). The REB^ gene has been cloned (Ju et al., 1990) and the
protein purified (Morrow et al., 1990). REB1 is a phosphorylated
polypeptide with an apparent molecular mass of 125,000 Daltons
(Morrow et al., 1990). REB1 shares homology with the DNA binding
domain of the oncogene myb, and chemical characteristics such as
hydrophilicity, abundant glutamines, and numerous hydroxyl-
containing amino acid residues in common with RAP1, ABF1, GCN4,
and HSF1 (Morrow et al., 1990).
A newly espoused function that may play a role in the
activation of PYK transcription is an "adaptor" or "modulator"
protein (Berger et al., 1990; Pugh and Tjian, 1990; Kelleher et al.,
1990; Liu and Green, 1990). An adaptor is postulated to act as a
bridge via protein-protein interactions between a positive trans
acting activator protein bound at UAS elements some distance
upstream of the basic promoter elements, and proteins such as TFIID
which bind the TATA box (Davison et al., 1983; Parker and Topol,
1984; Sawadogo and Roeder, 1985; Nakajima et al., 1988). The


130
Kellerman, E. and Hollenberg, C.P. (1988). The glucose- and
ethanol-dependent regulation of PDC1 from Saccharomyces
cerevisiae are controlled by two distinct promoter regions. Curr
Genet 14, 337-344.
Kimmerly, W.J., Buchman, A.R., Kornberg, R.D., and Rie, J. (1988).
Roles of two DNA-binding factors in replication, segregation and
transcriptional repression mediated by a yeast silencer. EMBO J. 7,
2241-2253.
Kurtz, S. and Shore, D. (1991). RAP1 protein activates and silences
transcription of mating-type genes in yeast. Genes Dev. 5, 616-628.
Lagunas, R. (1986). Misconceptions about energy metabolism of
Saccharomyces cerevisiae. Yeast 2, 221-228.
Liu, F. and Green, M.R. (1990). A specific member of the ATF
transcription factor family can mediate transcription activation by
the adenovirus E1a protein. Cell 61, 1217-1224.
Lutstorf, U. and Megnet, R. (1968). Multiple forms of alcohol
dehydrogense in Saccharomyces cerevisiae. Arch. Biochem. Biophys.
126, 933-944.
McNeil, J.B., Dykshoorn, P., Huy, J.N., and Small, S. (1990). The
DNA-binding protein RAP1 is required for efficient transcriptional
activation of the yeast PYK glycolytic gene. Curr Genet 18, 405-412.
Meissner, P.S., Sisk, W.P., and Berman, M.L. (1987). Bacteriophage
lambda cloning system for the construction of directional cDNA
libraries. Proc. Natl. Acad. Sci. USA 84, 4171.
Meisterernst, M., Horikoshi, M., and Roeder, R.G. (1990). Recombinant
yeast TFIID, a general transcription factor, mediates activation by
the gene-specific factor USF in a chromatin assembly assay. Proc.
Natl. Acad. Sci. USA 87, 9153-9157.


74
driven by the UAS elements of the PYK and EN01 (Buchman et al.,
1988). These oligonucleotides were cloned at position -220 before
the TPIr.lacZ fusion, integrated in unit copy, and the strains were
assayed for p-galactosidase activity.
Mutating the CTTCC from -375 to -370 to CAACC reduced
expression of the TPIr.lacZ fusion construct in wild-type yeast to 45
units of activity, a 2.5-fold reduction. 21 units of (3-galactosidase
activity were expressed in the gcr1 mutant background. Changing
the CATCC from -335 to -330 to CAACC reduced activity by one
third to 82 units in the wild-type background and 21 units of
activity in the gcr1 mutant strain. Mutating both pentamers reduced
activity to 36 units in the wild-type strain. The double mutant
containing construct expressed 20 units of (3-galactosidase activity
in the gcr1 mutant strain.
Both pentamers were required for full UAS activity with
mutations in the CTTCC having a larger effect than mutations in the
CATCC. However, it should be noted that the oligonucleotides
employed in these experiments contained only the sequence from
-377 to -327 known to be sufficient for UAStpi activity (Figure 7).


42
cold hydrazine stop solution was added (0.3M NaAc pH 7.5, 0.1 mM
EDTA). 750|il of ethanol was added, and the mixture placed on dry
ice.
The control C-reactions were as follows. The DNA aliquot was
resuspended in 5pl ddH2, and 15pl of 5M NaCI added. 30pl hydrazine
was added and incubated 16 minutes. The reaction was stopped by
the addition of 200pl ice-cold hydrazine stop solution. 750(il of
ethanol was added, and the mixture was placed on dry ice.
The control T+C-reactions were as follows. The DNA aliquot
was resuspended in 20pl dd^O. 30|il of hydrazine was added and
incubated for 16 minutes. 200pl ice-cold hydrazine stop solution
was added. 750pl of ethanol was added, and the mixture placed on
dry ice.
All reactions were processed as follows. Samples were
centrifuged at 10,000x g for 15 minutes, and pellets washed with
70% ethanol. (Hydrazine waste was neutralized by placing in a
saturated ferric chloride/water solution.) Pellets were resuspended
in 225pl ddHgO, 25|il of 3M NaAc pH 6.0 and 750pl of ethanol was
added. Following centrifugation at 10,000x g for 30 minutes,


30
a Hind\\\ site at their 5' end, and a SphI site at their 3' end. The
oligonucleotides were isolated from pUC18 by Hind\\\-Sph\ digest
and subcloned into the Hind\\\-Sph\ sites of pES90. Thus, the
oligonucleotides replaced the native UASjpi element in pES90. The
replacement of UASjpi in the resulting plasmids was confirmed by
DNA sequence analysis utilizing primer HB01.
Site-Directed Mutagenesis
The 1.4 kb Sma\ fragment of plasmid pES90, containing the 5'
noncoding region of TPI, was cloned into the Smal site of M13mp18.
The ligation mixture was used to transform E. coli TG-1 and lysates
were prepared from 10 individual plaques. RF II DNA was prepared
from each isolate and the orientation of the insert determined by
Hind\\\ restriction analysis. One phage, MES-TPImp18B, contains the
strand of 5' noncoding region that corresponds to the antisense
strand of the TPI structural gene, and was used as the target for
site-directed mutagenesis. Oligonucleotides with the desired
mutations were synthesized such that the area of mismatch was
flanked by regions of complementarity of 10 to 12 nucleotides.
These oligonucleotides, HB28-33, are listed in Table 2.


118
1990), to activate transcription. These results suggest that it may
be the role of RAP1 and REB1 to facilitate the binding of GCR1 by
providing the proper DNA structure. The results of the site-directed
mutagenesis and the 5' deletion series indicate that protein-protein
interactions are unlikely to occur between REB1 and GCR1. REB1 is,
therefore, most likely involved only in chromatin structure and not
transcriptional activation as suggested by Chasman et al. (1990) or
transcriptional repression as suggested by Wang et al. (1990).
Regardless of whether or not REB1 provides an actual activation
signal, the REB1-binding site is required for the full expression of
TPI. Therefore, UASypi should be expanded to positions -401 to
-327, thereby including the REB1-binding site.
The individual roles of RAP1 and GCR1 remain less well
defined. Both proteins are absolutely required for high-level
transcriptional activation of TPI. Binding of RAP1 alone is
insufficient for high-level activation. One possibility is that RAP1
and GCR1 are coactivators of transcription. In this model,
interactions between the two proteins are required for activation.
Alternatively, the role of RAP1 could be the facilitation of GCR1
binding to UAStpi. GCR1 is the specific transcriptional activator in


Sph I
-220
rsn \
\TPI::lacZ 1
-392
1 TPI::lacZ 1
ruxsi \
-377
1 TPI::lacZ 1

-348
1 TPI::lacZ 1
-853
-853
-853
STl lUAS I -300
IUAS? I -367
\TPI::lacZ
\TPI::lacZ I
No Fusion
B- Galactosidase Activity
S150.2B [GCR1)
Mean SD
HBY4 (gcr1)
Mean SD
210 11
21 4
251 36
20 2
107 s
29 3
27 7
18 3
138 19
52 19
52 3
20 4
15 2
7 2
14 2
6 2
12 5
3 1
CD
^1


68
were expressed in the gcr1 mutant background, only a 2.5 fold
reduction in expression. The lessening of the severity of the
reduction caused by the gcr1 lesion may be due to a position effect
such as was seen for EN02 gene (Holland et al., 1990). The 138 units
of activity in the wild-type background corresponds with the
expression observed when only UAStpi was driving expression of the
fusion (Figure 7). Deletion of an additional 36 base pairs from the 5'
noncoding region of the TPIr.lacZ gene fusion, removal of sequence
from -336 to -220, reduced expression of the fusion approximately
three-fold to 52 units of (3-galactosidase activity. The strain
harboring the construct that removed sequence from -347 to -220
yielded background levels of expression. Constructs which are
deleted of the known UAS element but retain all sequences to
position -853, are unable to drive expression. Therefore, no
additional UAS elements lie distal to UAStpi-
Mutational Analysis of UAStpi
UAStpi has been mapped to a 51 base pair region from position
-377 to -327 with respect to the start of the structural gene.
Figure 9 shows the sequence of the 5' non-coding region of TPI


47
TPI transcript isolated from wild-type and gcr1 mutant strains may
reflect altered transcriptional start sites. Such a result may
indicate the existence of two promoter elements, one of which is
GCR1-dependent and one of which is GCR1-independent. The TPI
transcript remaining in the gcr1 mutant could originate from a GCR1-
independent promoter and initiate at a unique start site. A GCR1
independent TPI transcript may be the product of a different
transcriptional mechanism than the abundant transcript in wild-type
strains. Transcripts of a single gene derived from independent
mechanisms are known to exist for H/S4 which has both inducible
and basal transcripts initiating from independent sites (Arndt et al.,
1987; Pellman et al., 1990).
Mapping the mature 5' end of the TPI transcript was
accomplished using primer extension analysis. 10pg of wild-type
(S150-2B) or 25pg of gcr1 mutant (HBY4) total RNA was annealed
with a radiolabeled oligonucleotide primer (HB05) corresponding to
the start of the TPI structural gene. Extension reactions were
carried out with reverse transcriptase in the presence of
deoxyribonucleotide triphosphates. The first base incorporated
corresponded to position +2 with respect to the adenine of the


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
Associate Professor of Pharmacology
and Therapeutics
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.
May 1992
Dean, College of Medicine
Dean, Graduate School


51
TPIr.lacZ fusion. The effect of the manipulations on the expression
of TPIr.lacZ could be measured indirectly by monitoring p-
galactosidase activity.
The plasmid pHB110, a derivative of pUC18, harbors the
TPIr.lacZ gene fusion with 3.5 kilobase pairs (kbp) of DNA 5' to the
start of TPI. 3.5 kbp. of DNA 5' to TPI was initially included in order
to ensure that all c/s-acting regulatory elements necessary for TPI
expression were included in the construct. Sequencing of the 5'
non-coding region identified a unique 77/71111 site at position -853,
Figure 5. Subcloning experiments showed that all sequences
sufficient for high-level expression of the fusion resided within the
77/71111 site. Therefore, the site could be used as an origin for a
series of nested deletions created through the action of exonuclease
Bal31. Figure 1 shows the scheme used to create the Bal31
deletions from plasmid pHB110. These deletions were used to map
the 5' boundary of the region sufficient for high-level expression of
the TPIr.lacZ gene fusion.
The fusion construct was linearized with 77/71111 and treated
for various times with exonuclease Bal31. H/ndlll/Sa/l/H/ndlll
linkers (HB03, table 2) were then ligated to the material. This


133
Santangelo, G.M. and Tornow, J. (1990). Efficient transcription of the
glycolytic gene ADH1 and three translational component genes
requires the GCR1 product, which can act through TUF/GRF/RAP
binding sites. Mol. Cell. Biol. 10, 859-862.
Sawadogo, M. and Roeder, R.G. (1985). Interaction of a gene specific
transcription factor with the adenovirus major late promoter
upstream of the TATA region. Cell 43, 165-175.
Scott, E.W., Allison, H.E., and Baker, H.V. (1990). Characterization of
TPI gene expression in isogeneic wild-type and grcri-deletion mutant
strains of Saccharomyces cerevisiae. Nucleic Acids Res. 18,
7099-7107.
Sherman, F., Fink, G.R., and Hicks, J.B. (1983). Methods in Yeast
Genetics Laboratory Manual (Cold Spring Harbor: Cold Spring Harbor
Laboratory Press).
Shore, D. and Nasmyth, K. (1987). Purification and cloning of a DNA
binding protein from yeast that binds to both silencer and activator
elements. Cell 51, 721-732.
Shore, D., Stillman, D.J., Brand, A.H., and Nasmyth, K. (1987).
Identification of silencer binding proteins from yeast: possible roles
in SIR control and DNA replication. EMBO J. 6, 461-467.
Singh, H., Sen, R., Baltimore, D., and Sharp, P.A. (1986). A nuclear
factor that binds to a conserved sequence motif in transcriptional
control elements of immunoglobulin genes. Nature 319, 154-158.
Southern, E.M. (1975). Detection of specific sequences among DNA
fragments seperated by gel electrophoresis. J. Mol. Biol. 98,
503-517.
Stanway, C.A., Chambers, A., Kingsman, A.J., and Kingsman, S.M.
(1989). Characterization of the transcriptional potency of
sub-elements of the UAS of the yeast PGK gene in a PGK
mini-promoter. Nucleic Acids Res. 17, 9205-9218.


Construct Arrangement
CYC1:: lacZ
£>
UAS TATA
UAS|ess CYC1:: lacZ
( ) TATA
34-1 H-S:: CYC1:: lacZ
(34-1 H-S) TATA
81mer:: CYC1:: lacZ
(81 mer) £ TATA
66mer:: CYC1:: lacZ
£>
(66mer) TATA
35-2
S
2 TATA
78-1
TATA
58mer:: TPI:: lacZ
$
58mer \ )-tata
66mer:: TPI:: lacZ
$ $
( j-TATA
69 RBSM:: TPI:: lacZ
$ $
39 RBgM( )_TATA
3 Galactosidase Activity
Mean
CYC1 ::lacZ 70
CYC1 ::lacZ 8
CYC1 ::lacZ 94
CYC1 ::lacZ 101
CYC1 ::lacZ 99
TPI ::lacZ 236
TPI ::lacZ 8
TPI ::lacZ 20
TPI ::lacZ 129
TPI ::lacZ 18
CD
N>
SD
9
6
3
5
7
25
4
4
11
4


EXPRESSION OF THE TP/GENE OF SACCHAROMYCES CEREVISIAE
IS CONTROLLED BY A SINGLE COMPLEX UPSTREAM ACTIVATING
SEQUENCE CONTAINING BINDING SITES FOR THREE TPAA/S-ACTING
FACTORS: REB1, RAP1, AND GCR1
BY
EDWARD WILLIAM SCOTT V
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
1992

DEDICATION
To my wife Rochelle for her love and support in all things I do.
Father, first and foremost this belongs to you. Your rearing has
produced my thirst for knowledge about all I touch. Your example
demonstrates that education is not a stagnant entity cast off after
college, rather, it is a lifelong and enlightening endeavor.
Why, all delights are vain; but that most vain,
Which with pain purchased doth inherit pain:
As, painful to pore upon a book
To seek the light of truth; while truth the while
Doth falsely blind the eyesight of his look:
Light seeking light doth light of light beguile:
So, ere you find where light in darkness lies,
Your light grows dark by the losing of your eyes.
Love's Labour's Lost
William Shakespeare

ACKNOWLEDGMENTS
I would like to extend my deepest appreciation to my mentor
Dr. Henry V. Baker for providing superb direction to my graduate
education and personal development as a scientist. My deepest
regards to the members of my dissertation committee: Dr. Lonnie O.
Ingram, Dr. Alfred S. Lewin, Dr. Richard W. Moyer, and Dr. Thomas C.
Rowe for their invaluable insights. I thank the Department of
Immunology and Medical Microbiology for providing a climate
conducive to intellectual growth and exchange.

TABLE OF CONTENTS
page
ACKNOWLEDGMENTS iii
LIST OF TABLES vi
LIST OF FIGURES vii
ABSTRACT ix
INTRODUCTION 1
MATERIALS AND METHODS 17
Strains 17
Media and Growth Conditions 17
Nucleic Acid Manipulations 19
Generation of Double Strand DNA Oligonucleotides 19
DNA Sequencing 20
Primer Extension 22
Plasmid Construction 22
Constructs for sequencing 22
5' deletion scheme 23
Internal deletion scheme 26
Constructs to assess UAS activity 29
Mutant UAS oligonucleotides driving expression of
the TPi.JacZfusion 29
Site-Directed Mutagenesis 30
Transformation 32
Screen for Unit Copy Integrants of the TPI::lacZ Fusions
at URA3 32
p-Galactosidae Assays 35
In vitro DNasel Protection Assays 36
DNA Band Shift Assays 38
In vivo Methylation Protection Analysis 38

RESULTS
46
The Mature 5' Ends of Steady-State TPI Transcripts are
Unaffected by a gcr1 Mutation 46
Identification of the 5' Boundary of the TPI Controlling
Region 48
An Upstream Activating Sequence Activity for TPI
Resides from Position -377 to -327 in the 5'
Noncoding Region 59
Internal Deletions Indicate Single UAS Element
Responsible for TPI Transcription 63
Mutational Analysis of UASjpi 68
In vitro DNase I Protection Assays Reveal Binding of
the REB1 Site and the RAP1 Site 75
DNA Band Shift Assays Demonstrate REB1 Binding to
TPI 5' Noncoding Region 81
In vivo Methylation Protection Assays 82
Site-Directed Mutagenesis of Transcription Factor
Binding Sites in the UAS of TPI 90
DISCUSSION 99
REFERENCES 124
BIOGRAPHICAL SKETCH 136
v

LIST OF TABLES
£age
Table 1. Strains 18
Table 2. Oligonucleotides 21

LIST OF FIGURES
page
Figure 1. Scheme to generate 5' deletions in the TPI
5' noncoding region 25
Figure 2. Scheme to create internal deletions 28
Figure 3. Scheme for integration of TPI::lacZ gene
fusion constructs at URA3 34
Figure 4. Primer extension analysis of the 5' ends of
the TPI transcript in wild-type and
gcr1-deletion mutant strains 50
Figure 5. Sequence of TPI 5' Noncoding Region 52
Figure 6. Effect of 5' deletions upon expression of a
TPI::lacZ fusion integrated in unit copy
at the URA3 locus in wild-type and
gfcri-deletion mutant strains 57
Figure 7. Identification of UAStpi 62
Figure 8. Effect of internal deletions on (3-galactosidase
activity expressed from a TPI::lacZ
gene fusion 67
Figure 9. Summary composite of the 5' noncoding
region of TPI 69
Figure 10. Mutational analysis of UASjpi utilizing
mutant oligonucleotides 72
Figure 11. In vitro DNase I protection assays demonstrating
protection of the RAP1-binding site 78
v i i

Figure 12. In vitro DNase I protection assays demonstrating
protection of the REB1 -binding site 80
Figure 13. DNA band shift assays demonstrating REB1 binding
to the 5' noncoding region of TPI 84
Figure 14. Genomic footprinting of the bottom strand
of the TPI 5' noncoding region 87
Figure 15. Genomic footprinting of the top strand
of the TPI 5' noncoding region 92
Figure 16. Effect of site-directed mutations upon
(3-galactosidase activity expressed
from a TPI::lacZ fusion 94
Figure 17. Composite summary of the TPI controlling
region 109
Figure 18. Model of protein interactions at UAStpi 122
vi i i

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
EXPRESSION OF THE TP/GENE OF SACCHAROMYCES CEREVISIAE\S
CONTROLLED BY A SINGLE COMPLEX UPSTREAM ACTIVATING SEQUENCE
CONTAINING BINDING SITES FOR THREE TRANS-ACTING FACTORS:
REB1, RAP1, AND GCR1
By
Edward William Scott V
May 1992
Chairman: Henry V. Baker, Ph. D.
Major Department: Immunology and Medical Microbiology
In Saccharomyces cerevisiae the enzymes of glycolysis
constitute 30-60 percent of the soluble protein. GCR1 gene function
is required for high-level glycolytic gene expression. This study
involved a biochemical and genetic characterization of TPI, a gene
affected by gcr1 lesions. Primer extension experiments on TPI
transcripts isolated from wild-type and gcr1 mutant strains mapped
the initiation or "I" site of transcription to a pair of adenines 29 and
30 bases upstream of the start of translation. To delineate the TPI

controlling region, a TPI::lacZ gene fusion was employed. Nuclease
Bal31 deletion analysis identified a single upstream activating
sequence, UASjpi, responsible for high-level TPI expression. DNase I
protection and in vivo dimethyl sulfate methylation protection
assays indicated the binding of three trans-acting factors to four
sites within UASypi. GCR1 binds two sites: 1) a sequence element
which functionally requires the CTTCC pentamer from position -375
to -370; 2) an element requiring the CATCC pentamer from position
-335 to -330. REB1 binds to an almost perfect consensus binding
site from position -397 to -387. RAP1 binds from -358 to -346.
REB1 and RAP1 binding at UASjpi are independent of GCR1 binding in
vivo. Site-directed mutation of the REB1-binding site reduced the
expression of the TPI::lacZ gene fusion by two-fold. Mutation of the
RAP1 or both GCR1-binding sites abolished expression. Thus, TPI
absolutely requires RAP1 and GCR1 binding for expression and
requires REB1 for full expression. This work suggests a mechanism
of high-level glycolytic gene expression mediated primarily through
the actions of RAP1 and GCR1.
x

INTRODUCTION
A fundamental process in biology is the regulation of gene
expression. For it is this process which allows a single fertilized
human egg to develop into an organism of 10^ cells in just nine
months. Furthermore, the study of cancer has shown that when the
process of gene regulation is perturbed a neoplastic transformation
may result. The yeast Saccharomyces cerevisiae offers an ideal
model system to study the regulation of gene expression. Over the
years a sophisticated genetic system has developed that allows one
to manipulate a gene in vitro and then reintroduce the mutated gene
back into the genome to assess the effect of the manipulation in
vivo. In addition, transcription factors in yeast often have homologs
in higher eukaryotes. For example, the JUN oncoprotein binds the
same recognition sequence and has extensive amino acid homologies
with the yeast transcriptional activator GCN4. JUN is even able to
functionally complement a gcn4 mutation in yeast (Struhl, 1988).
It has long been known that upon neoplastic transformation in
1

2
certain types of cancer there is an increase in aerobic glycolysis
(Warburg, 1930). Saccharomyces cerevisiae utilizes aerobic
glycolysis to a much greater extent than respiration (Lagunas,
1986). The enzymatic pathway of glycolysis in yeast is well
established. The enzymes of glycolysis, while few in number,
compose between 30-60% of the total soluble protein (Fraenkel,
1982). This observation suggests that the genes encoding these
enzymes are among the most highly expressed in yeast. Indeed,
mRNA encoding glycolytic enzymes has been demonstrated to be a
major fraction of total yeast mRNA (Holland et al., 1977; Holland and
Holland, 1978). The regulation of the genes encoding the glycolytic
enzymes is currently receiving much study, but no overall consensus
regulatory mechanisms have yet been identified, rather some
similarities in regulatory elements and factors have been noted.
These similarities will be addressed subsequently.
Mutations affecting the flux of metabolites through the
glycolytic pathway tend to map to single loci and affect single
enzymes (Fraenkel, 1982). However, Clifton et al. (1978) isolated a
mutant that has severely reduced levels of most glycolytic enzymes.
Yet genetic analysis showed that this strain contains a mutation

3
that segregates as a single gene. Due to its pleiotrophic nature the
gene was named GCR1 for glycolysis regulation. Strains harboring a
gcr1 mutation express the genes encoding the enzymes of glycolysis
at approximately 5-10% of wild-type levels (Clifton et al., 1978;
Clifton and Fraenkel, 1981; Baker, 1986). gcr1 mutants exhibit poor
growth on glucose while retaining adequate growth on non-
fermentable carbon sources (Clifton and Fraenkel, 1981). The
reduction in enzyme levels has been shown to be mirrored by a
corresponding reduction in steady-state mRNA levels for several
affected enzymes (Holland et al., 1987; Santangelo and Tornow,
1990; Scott et al., 1990). GCR1 has been cloned (Kawasaki and
Fraenkel, 1982) and sequenced (Baker, 1986; Holland et al., 1987).
DNA sequence analysis indicates that GCR1 encodes a polypeptide of
844 amino acids with a molecular weight of 94,414 Daltons.
Much attention and effort has been devoted to the study of gene
regulation in S. cerevisiae and other systems. The plieotrophic
nature of mutations in GCR1 suggests that the gene is involved in
the coordinate regulation of expression of the genes encoding
glycolytic enzymes. However, the residual level of expression of
aldolase, triose-phosphate isomerase, glyceraldehyde-3-phosphate

4
dehydrogenase, phosphoglycerate kinase, and pyruvate kinase in a
gcr1 mutant is somewhat inducible by glucose, relative to
expression under gluconeogenic conditions (Baker, 1986). Insights
into the mechanism by which the GCR1 gene product exerts its
effect will prove valuable in the exploration of glycolytic gene
regulation.
This study will attempt to elucidate the cis and trans-acting
elements involved in the expression of TPI, the sole gene encoding
triose-phosphate isomerase activity in Saccharomyces cerevisiae.
Triose-phosphate isomerase catalizes the reversible isomerization
of dihydroxyacetone phosphate to glyceraldehyde 3-phosphate during
glycolysis. TPI has been shown to be dependent upon GCR1 for full
expression; in a gcr1 mutant strain TPI gene expression is reduced
17-fold (Clifton and Fraenkel, 1981). By mapping the elements of
the TPI controlling region, it should prove possible to map the DNA
sequence through which the GCR1 gene product acts and to determine
if GCR1 acts alone or in conjunction with other sites or factors. In
order to facilitate the identification of the site of GCR1 action in
the TPI controlling region, expression of a TPI::lacZ fusion will be
analyzed in isogeneic wild-type and gcrI mutant strains. The

5
identification of other trans acting factors required for the full
expression of TPI is also a primary concern. This introduction will
serve as a literature review of the studies to date concerning
glycolytic gene expression in Saccharomyces cerevisiae.
In addition to the GCR1 gene product, this study has shown that
expression of TPI requires the binding of a trans-acting protein
known as repressor activator protein 1 (RAP1) (Shore et al., 1987)/
general regulatory factor 1 (GRF1) (Buchman et al., 1988)/
translation upstream factor (TUF) (Huet et al., 1985), (Scott et al.,
1990). RAP1/GRF1/TUF will be referred to hereafter as RAP1.
RAP1 was first purified as a binding activity which bound the
HMR(E) silencer locus. RAP1 binding is required for repression of
transcription by the HMR(E) silencer (Shore and Nasmyth, 1987);
however, RAP1 binding is also required for the activation of
transcription for many genes (Huet et al., 1985; Capieaux et al.,
1989). Shore and Nasmyth (1987) demonstrated that two different
binding sites for RAP1 derived from two different UAS's are able to
restore HMR silencer function in vivo when substituted for the
native RAP1-silencer-binding site. When the native RAP1- silencer¬
binding site was destroyed the silencer no longer functioned and a 1

6
transcript was produced, restoration of the native site or
substitution of the UAS derived RAP1-sites restored silencer
function and a1 transcript was no longer produced. These findings
indicate additional factors must play a role in RAP1 dependent
activation or repression.
RAP1 involvement in gene expression is widespread. RAP1
binds and is required for the function of the HMR and HML loci of the
mating type locus (Kimmerly et al., 1988; Hofmann et al., 1989),
genes encoding ribosomal proteins and other proteins of the basic
translational machinery (Huet et al., 1985), the H+-ATPase gene
(Capieaux et al., 1989), and the H/S4 gene (Devlin et al., 1991).
In addition to TPI, RAP1 binding has also been shown to be
necessary for the expression of several other genes encoding
glycolytic enzymes. Capieaux et al. (1989) were the first to note
near consensus RAP1-binding sites in the regulatory regions of GPD,
PGK, PYK, EN01, PDC, and ADH1. Subsequently, mutational analysis
has shown those sites to be essential for the expression of PGK
encoding phosphoglycerate kinase (Ogden et al., 1986; Chambers et
al., 1989), EN01 encoding an isozyme of enolase (Buchman et al.,
1988), PYK encoding pyruvate kinase (Buchman et al., 1988), ADH1

7
encoding alcohol dehydrogenase isozyme one (Buchman et al., 1988;
Santangelo and Tornow, 1990), and PDC1 encoding pyruvate
decarboxylase (Kellerman and Hollenberg, 1988). RAP1 binding has
also been reported to be required for the expression of EN02 (Brindle
et al., 1990). Recent studies with both temperature sensitive rap*\
mutants and rapis mutants, selected for defective silencing of the
HMR locus, have been able to separate the suppression and activation
functions of the RAP1 protein (Kurtz and Shore, 1991; Sussel and
Shore, 1991). However, both sets of mutations map to the carboxyl
terminus of the RAP1 protein and no clear domains have been
defined.
RAP1 is a phosphoprotein (Tsang et al., 1990). RAP1 binding
in vitro to the PYK controlling region has been shown to be reduced
by phosphatase treatment (Tsang et al., 1990). However, in addition
to phosphorylation, RAP1 binding in the in vitro binding assay was
also dependent upon binding site context (Tsang et al., 1990). The
significance of RAP1 phosphorylation state in vivo is unknown.
The controlling regions of PGK, EN01, EN02, and PYK have all
been mapped. In the case of PGK the upstream activating sequence
has been shown to be comprised of three main elements (Stanway et

8
al., 1989). The central portion of the UAS contains the activator
core (AC) to which RAP1 binding has been demonstrated (Chambers
et al., 1989). Upstream of the activator core is a region designated
YfP, first recognized as a site of strong DNasel protection (Chambers
et al., 1988). The Yfp region binds ARS binding factor 1 (ABF1),
another trans-acting binding protein similar in properties to RAP1
(Chambers et al., 1990; Buchman and Kornberg, 1990). The third
region identified in the UAS of PGK consists of three repeats of a
pentamer sequence motif, CTTCC (Ogden et al., 1986). In
experiments with fragments from the PGK promoter which carry
both the RAP1 binding site and the CTTCC sequence motif, there is
DNasel protection of the RAP1 -binding site and evidence of
heightened DNasel sensitivity at the CTTCC pentamer motif
(Chambers et al., 1988). However, no clear footprint was observed.
Chambers et al. (1988) suggested that the heightened DNase I
sensitivity at the CTTCC sequence element was due to interactions
with RAP1. Of the three main elements of the PGK UAS, no single
element is able to activate expression alone (Stanway et al., 1989).
The activator core with the YfP is able to drive about one-third the
wild-type levels of expression. The activator core with the three

9
CTTCC pentamers has wild type levels of expression even without
the strong ABF1-binding site in YfP (Ogden et al., 1986). Deletion of
the CTTCC pentamer from -432 to -428 reduces expression by 50%,
deletion of two of the repeats (-432 to -428 and -449 to -445)
reduces expression 75%, and removal of all three CTTCC motifs
reduces expression to less than 10% of wild-type (Chambers et al.,
1988). As seen with other glycolytic genes, full expression of PGK
depends on GCR1.
Both EN01 and EN02 have two upstream activating sequence
elements (Cohen et al., 1987). EN01 also contains an upstream
repressor sequence (URS) responsible for repression of EN01 in
nonfermentable carbon sources. The UAS1 of both EN01 (Buchman et
al., 1988; Brindle et al., 1990) and EN02 bind RAP1 (Brindle et al.,
1990). In the case of UAS1 of EN01 a CTTCC pentamer has been
identified. The pentamers presence has been shown to be required
for full UAS1 activity, but its absence does not affect the ability of
RAP1 to bind in vitro (Buchman et al., 1988). UAS2 of EN01 binds a
factor designated EBF1, Fnolase Binding Factor 1. UAS2 of EN02
binds ABF1 and overlaps UAS1. In EN02 , RAP1 binding to UAS1 and
ABF1 binding at the overlapping UAS2 may be competitive (Brindle et

al., 1990). The competitive nature of the interaction of RAP1 and
ABF1 at the UAS of EN02 is in contrast to the interactions that
occur between these factors at the UAS of PGK, where binding is not
thought to be competitive.
Expression of both enolase genes depends on GCR1 (Holland et
al., 1987), and, in the case of EN02, a small deletion in the area of
overlap of the two UAS's is able to relieve the requirement of GCR1
for full expression (Holland et al., 1990). However, when the UAS
elements of EN02 was moved immediately upstream of the TATA
boxes of EN01 or EN02 they were able to confer expression but
expression remained high in a gcr1 mutant strain. Thus, Holland et
al. (1990) suggest there may be an effect due to the positioning of
the UAS elements close to the TATA box which can alleviate the
requirement for GCR1.
However, when UAS1 and UAS2 of EN02 (positions -491 to -
443 with respect to the start of the EN02 structural gene) were
cloned upstream of a TPI::lacZ gene fusion, replacing the native UAS
element, no expression was seen (J. Anderson and H. Baker, personal
communication). Furthermore, the fragment was unable to bind
RAP1 in vitro. These results directly contradict those reported by

Holland et al. (1990).
PYK has two UAS elements: UAS1 (-653 to -634), UAS2 (-811
to -714). PYK is also controlled by an URS (-468 to -344) to repress
expression during growth on nonfermentable carbon sources
(Nishizawa et al., 1989). Full expression of PYK requires GCR1
(Clifton and Fraenkel, 1981). The UAS1 binds RAP1 and has a CTTCC
pentamer that enhances UAS activity (Buchman et al., 1988).
Mutation of the CTTCC pentamer to CAACC reduced UAS activity by
16-fold. Deletion of the RAP1-binding site of UAS1 prevents
expression of PYK (McNeil et al., 1990). The pentamer is not
required for binding of RAP1 to UAS1 in vitro (Buchman et al., 1988).
The overall organization of UAS1 of PYK is very similar to the
UAS of PGK (Buchman et al., 1988; Chambers et al., 1990). Both PYK
and PGK have ABF1-binding sites adjacent to their RAP1-binding
sites and CTTCC sequence elements. The role of ABF1 binding in
expression of PYK has yet to be determined. In addition, another
factor known as REB1 (Morrow et al., 1989)/ QBP (Brandi and Struhl,
1990)/ Y (Fedor et al., 1988), hereafter called REB1, binds in the
UAS2 area of PYK (Chasman et al., 1990).
REB1 has been shown to affect nucleosome positioning in UAS

elements (Fedor et al., 1988; Chasman et al., 1990; Brandi and
Struhl, 1990), and, aids in gene activation in conjunction with
additional transcription factors (Chasman et al., 1990; Wang et al.,
1990). The REB^ gene has been cloned (Ju et al., 1990) and the
protein purified (Morrow et al., 1990). REB1 is a phosphorylated
polypeptide with an apparent molecular mass of 125,000 Daltons
(Morrow et al., 1990). REB1 shares homology with the DNA binding
domain of the oncogene myb, and chemical characteristics such as
hydrophilicity, abundant glutamines, and numerous hydroxyl-
containing amino acid residues in common with RAP1, ABF1, GCN4,
and HSF1 (Morrow et al., 1990).
A newly espoused function that may play a role in the
activation of PYK transcription is an "adaptor" or "modulator"
protein (Berger et al., 1990; Pugh and Tjian, 1990; Kelleher et al.,
1990; Liu and Green, 1990). An adaptor is postulated to act as a
bridge via protein-protein interactions between a positive trans
acting activator protein bound at UAS elements some distance
upstream of the basic promoter elements, and proteins such as TFIID
which bind the TATA box (Davison et al., 1983; Parker and Topol,
1984; Sawadogo and Roeder, 1985; Nakajima et al., 1988). The

13
interaction of the trans-activator and the adaptor is essential to aid
in the recruitment of TFIID or other basic transcription factors to
the TATA box. RAP1 dependent expression of PYK has been shown to
require functional GAL11/SPT13 (Nishizawa et al., 1990).
Nishizawa et al. suggest that GAL11/SPT13 was acting as an adaptor
protein in PYK expression. GAL11/SPT13 was first identified as a
gene required for full expression of some genes regulated by GAL4
(Nogi and Fukasawa, 1980), or as a suppressor for auxotrophic
mutants induced by Ty insertion (Fassler and Winston, 1988).
However, Nishizawa et al. note that Fassler and Winston (1989) did
not find decreased transcription of PYK1 in gall 1/spt13 cells.
Nishizawa et al. suggest that the disparity in the observations may
be due to allele specific differences of the gall 1 /spt13 mutations.
The matter is by no means well resolved. Nishizawa et al. (1990)
showed that the requirement for GAL11 was alleviated by moving
the RAP1 binding site closer to the TATA element. Thus, the
requirement for the putative adaptor was alleviated by moving the
binding site for the second trans-activator closer to the TATA
element, supposedly allowing a direct interaction with the basic
transcription machinery such as TFIID.

ADH1 also has a UAS that binds RAP1 (Buchman et al., 1988).
Transcription of ADH1 is reduced in a gcr1 strain (Santangelo and
Tornow, 1990). Santangelo and Tornow have reported that the RAP1
site of the ADH1 UAS was able to confer responsiveness to GCR1
when it replaced the UAS of LAC4 (Santangelo and Tornow, 1990).
Therefore, Santangelo and Tornow suggest that GCR1 acts through
the RAP1 binding site. Contrasting this observation is the PGK
controlling region where the activator core region to which RAP1
binds is insufficient to activate transcription (Stanway et al.,
1989). In addition, both EN01 and PYK UAS's require the CTTCC
pentamer as well as the RAP1-binding site for full UAS activity
(Buchman et al., 1988). This contradiction over RAP1 binding alone
being sufficient for activation of transcription has yet to be
resolved.
The exact mechanism by which RAP1 is able to activate
transcription of TPI, PGK, EN01, EN02, PYK, PDC1, and ADH1 has yet
to be determined. Expression of each of these genes also depends on
GCR1. The nature of the interactions of RAP1, ABF1, REB1 and GCR1
are also unknown. Some indications of the role of RAP1 binding may
be provided by studies that link RAP1 with DNA structure, both in

DNA loop formation (Hofmann et al., 1989) and interactions with the
nuclear scaffolding (Cardenas et al., 1990). The RAP1 protein also
binds yeast telomeric repeat sequences (Conrad et al., 1990), and
telomeres are shortened in a conditionally lethal (ts) rap1 mutant at
nonpermissive temperatures (Conrad et al., 1990). RAP1, ABF1, and
REB1 are abundant proteins (Buchman et al., 1988; Morrow et al.,
1989) and thus they may be more involved in a common mechanism
of transcription rather than specific regulation of any given gene.
The GCR1 gene product appears to be a specific regulatory
protein (or one of many) for the genes encoding glycolytic enzymes.
As with many specific regulatory proteins in yeast, such as GAL4
(Bram and Kornberg, 1982), GCR1 is expressed at low levels is S.
cerevisiae (Baker, 1986). If GCR1(or any other protein) provides for
the specific activation of the genes of glycolysis how does it exert
its effect in the context of RAP1, REB1, and ABF1 binding at UAS
elements?
Uemura and Fraenkel (1991) (Uemura and Fraenkel, 1991) have
recently isolated GCR2 which like GCR1 has a pleiotropic effect upon
the expression of most of the enzymes of glycolysis. The pattern of
affected enzymes in a gcr2 or a gcr1 strain was quite similar with a

90-95% reduction seen in the expression of most glycolytic
enzymes. EN01 transcript levels were compared in GCR2 and gcr2
backgrounds. The defect in enzymatic activity seen in a gcr2 mutant
is mirrored by a reduction in EN01 mRNA. Interestingly, the gcr2
mutant exhibits only a partial growth defect on glucose where a
gcr1 mutant has a severe growth defect on glucose. GCR2 has been
cloned by complementation but its sequence has not been reported.
How GCR2 will fit into the overall regulation of the expression of
the genes encoding glycolytic enzymes and how it relates to GCR1
remains to be seen.

MATERIALS AND METHODS
Strains
The strains of Saccharomyces cerevisiae and Escherichia coli
used in this study are shown in Table 1.
Media and Growth Conditions
Yeast cultures were grown in YP medium (Sherman et al.,
1983) supplemented with 2% glucose or 2% glycerol and 2% lactate.
Selection was carried out in YNB supplemented with the appropriate
carbon source and 0.0025 % histidine, 0.0025% leucine, 0 0025%
tryptophan,and 0.1% case amino acids. All yeast cultures were grown
at 30OC. E. coli strains were grown in LB broth or minimal medium
63 supplemented with thiamine hydrochloride (1pg/ml), and amino
acids (25pg/ml) (Miller, 1972). Ampicillin was added to 100 pg/ml
for selection. All E. coli cultures were grown at 370C.

Table 1. STRAINS
Strain
Genotvpe
Source. Reference
E.coli
KK2186
supE, sbcB15, hsdR4,
rpsL,thi, A(lac-proAB)
F [traD36, proAB+, laclq,
lacZAml 5]
(Zagursky and
Berman, 1984)
MC1061
hsdR, mcrB, araD139,
A(araABC-leu)7679
AlacX74, galU, galK,
rpsL, thi
(Meissner et al.,
1987)
TG1
supE, hsdA5,thi,
A(lac-proAB)
F [traD36, proAB+,
laclq, /acZAM15]
(Gibson, 1984)
S. cerevisiae
S150-2B
MAT a, leu 2-3,112,
his3A,trp1 -289, ura3-52
(D. Shore)
HBY4
MAT a, gcr1A::HIS3,
Ieu2-3,112, his3A,
trp 1-289, ura3-52
(Scott et al.,
1990)
JF1052
MAT a, Ieu2, ura3-52,
his4-917, Iys2-1288,
spt 13-20 {LEU2)
(J. Fassler)
DFY642
MATa, leu2-3,112,
ura3-52
(D. Fraenkel)

Nucleic Acid Manipulations
Techniques used throughout this study are derived from the
standard reference manuals (Sambrook et al., 1989; Sherman et al.,
1983; Current Protocols in Molecular Biology, 1989) except for
deviations noted. S. cerevisiae DNA and RNA was prepared by the
methods of Sherman et al. (1983) and Struhl and Davis (1981),
respectively.
Generation of Double-Strand DNA Oligonucleotides
Most double-stranded oligonucleotides were generated by a
modification of the method of Oliphant et al. (1986). Single-
stranded oligonucleotides were synthesized (University of Florida
Interdisciplinary Center for Biotechnology Research) with the
desired sequence flanked by restriction sites, typically a H/ndlII
site on the 5' end and a SphI site on the 3' end. One restriction site
was made into an 8-10 bp palindrome by the addition of G and C
residues. The oligonucleotides were self-annealed via the
reinforced palindrome and the 3' ends extended with Klenow
fragment of DNA polymerase I in the presence of dNTPs. The double-
stranded extension products were then gel purified via

20
polyacrylamide gel electrophoresis (PAGE), digested with the
appropriated restriction enzymes, and cloned into the polylinker
region of pUC18 or other suitable plasmid vectors. This method was
used to generate the double-stranded form of oligonucleotides HB16,
HB19, HB21, HB22, HB24, HB25 listed in Table 2. Additional double-
stranded oligonucleotides were generated for cloning by annealing
complementary oligonucleotide pairs, also shown in Table 2.
Followed by digestion with the appropriate restriction enzymes, and
cloning.
DNA Sequencing
DNA sequencing was carried out by the dideoxy chain
elongation termination method of Sanger (1977) as modified by U.S.
Biochemicals to utilize the Sequenase enzyme. Sequencing primers
were synthesized by the University of Florida Interdisciplinary
Center for Biotechnology Research and are listed in Table 2.

21
Table 2,
OLIGONUCLEOTIDES
Name
Sequence
Forward
5'-GT AAAACGACGGCC ACT-3'
HB01
5-AT GT GT GGAATTGT GAGCGG-3'
HB03
5'-C AAGCTT GT CGACAAGCTTG-3'
HB05
5-CCACCGACAAAGAAAGTT CT AGCC-3'
HBOG
5'-AC AT GCAT GCAT GT -3'
HB07
HB08
5-CAAGCTTGTT CTAAGACTTTTC AGCTT CCTCTATT GATGTTAC ACTTGGACGC ATGCC-3'
5-GGCATGCGT CCAAGTGTAACATCAATAGAGGAAGCTGAAAAGTCTTAGAACAAGCTTG-3'
HB09
5-GGCAT GCCAAC AT GT ATGGGTT CCAAGCTTG-3'
HB11
5-GGCAT GCCAACCTGAT GGGTTCCAAGCTT G-3'
HB14
HB15
5-GGAAGCTTAGCTTCCTCTATT GAT GTTACACCT GGACACCCCTTTTCT GGCATCCAGTTGCATGCC-3'
5-GQCATGCAACTQGATGCCAGAAAAGGGGTGTCCA3GTGTAACATCAATAGAGGAAGCTAAGCTTCC-3'
HB16
5-GGAAGCTTGTTCTAAGACTTTTCAGCTT CCTCTATT GATGTTACACCT GGACACCCCTTTT CT GGCAT CCA
GTTGCATGCC-3'
HB17
5-GQCATQCAACTQGATQCCAGAAAAGQQGTGTCCAGGTGTAACATCAATAGAGGAAQCTGAAAAGTCTTAGA
ACAAGCTTCC-3'
HB19
5-GCTAAGCTTAGCTT CCT CT ATT GAT GTT ACACCTGGAAACCCATCAGGTGGCATCCAGTT GCATGCAAC-3'
HB21
5-GGGGCATGCAACTGGATGCCAGAAAAGGQGTGTCCAGGTGTAACAT CAATAGAGGTTGCTAAGCTTAGG3'
HB22
5-GGQQCATGCAACTGGTTGCCAGAAAAGGQGTGTCCAQGTGTAACATCAATAGAGGAAGCTAAGCTTAGC-3'
HB23
5'-CGCAAACCGCCT CTCC-3'
HB24
5-GCTAAGCTT AGCTT CCT CT ATT GAT GTT ACACCTGGAGATATCTGCAGTGGCAT CCAGTTGCATGCAAC-3'
HB25
5-GCTAAGCTTAGCAACCTCTATTGATGTTACACCTGGACACCCCTTTTCTGGCAACCTGTTGCATGCAAC-3'
HB28
5-GTTACACCTGGAGAT AT CT GCAGT GGCATCCAG-3'
HB29
5'-GACTTTT CAGC AACCT CT ATT-3'
HB30
5'-CCTTTTCT GGCAAACAGTTTTT AATC-3’
HB31
5-GGAAGATT GAACGTTCT AAG-3'
HB32
5-GGAACCCATAC AT GTT GGT GGAAG-3'
HB33
5-GT AACAGGGAAGCGAAAGGGCAGC-3'
HB42
5-CTGTGAGGACC-3'
HB53
5-GATGTTAC ACCAGATTACCCGTT CT CT GGCAT CCAG-3'
HB54
5-GACCTT AAT ACATT C AG-3'
HB60
5-GCATTAGCATGCGTAACAAACCACC-3'

22
Primer extension
Total RNA was isolated from yeast strains S150-2B and HBY4
and used to determine the 5' end of the TPI transcripts. Total RNA,
10pg from wild-type and 25pg from gcrl -deletion mutant, was
annealed with a radiolabeled primer HB05 which hybridized near the
5' end of the TPI structural gene. The primer was extended with 10
units of reverse transcriptase in the presence of dNTPs at 3mM. The
extension products were analyzed by denaturing PAGE. A sequencing
ladder generated from the 5' noncoding region of TPI with primer
HB05 was used as a molecular weight standard.
Plasmid Construction
Constructs for sequencing Single-stranded templates of the
5' noncoding region of TPI were generated for sequencing in the
following manner. M13mp18 and M13mp19 RF II double-stranded
DNA was prepared. Portions of the 5' noncoding region of TPI were
subcloned from plasmid pHB51 into M13mp18 or M13mp19 RF II DNA.
Single-stranded templates were prepared by standard techniques
(Sambrook et al., 1989), utilizing E. coli strain KK2186. Templates
were sequenced as above, utilizing the forward primer.

23
5' deletion scheme The scheme to introduce the Ba/31-induced
deletions in the 5' noncoding region of TPI is shown in Figure 1.
Plasmid pHB110 contains 3.5 kb of additional yeast DNA 5' to the
start of the TPI structural gene fused in frame to lacZ. The fusion
construct was linearized with Tth1111 at a unique site 853 bp
upstream of the structural gene. The linearized material was then
treated for various times with exonuclease Bal31. Following Ba/31
treatment, Klenow fragment was used to fill in the ends. HindIII-
Sal\-Hlnd\\\ linkers (CAAGCTTGTCGACAAGCTTG, HB03) were added.
The material was then digested with Sal I which cuts within the
linker and within the polycloning site of pHB110. This digest
removes all yeast derived DNA 5' to the deletion endpoint. The
desired fragments, containing the TPT.:lacZ fusion and the remaining
5' noncoding sequence, were gel purified, ligated, and used to
transform E. coli MC1061. Plasmid DNA was prepared from
individual transformants, and the precise deletion endpoints were
determined by double-stranded DNA sequencing using primer HB01.
Once the desired constructs were identified, they were
subcloned on Hind\\\ fragments into the Hind\\\ site of Ylp56. The
orientation of the fusion with respect to the URA3 selectable

Figure 1. Scheme to generate 5' deletions in the TPI 5' noncoding
region. The TPIr.lacZ fusion plasmid pHB110 contains a unique
Jth111\ site at position -853 with respect to the start of
translation. The plasmid was linearized with Jth111\ and treated
with exonuclease Bal31 for various times. H/'nd 11 l-Sa/l-H /nd 111
linkers (HB03, table 2) were added. The material was then digested
with Sail which cuts once in the polycloning region of pHB110 and in
the linker. After gel purification of the vector band, the material
was recircularized via ligation and used to transform E.coli.
Plasmid DNA was prepared from the transformants and precise
deletion endpoints were determined by DNA sequence analysis.
Restriction sites are as follows: E,EcoRI. S, Sail. H, H/'ndlll. T,

25
Tth111l digest
Bal 31 digest
♦
T
T

26
marker was determined by restriction endonuclease analysis with
BamHl
Internal deletion scheme Plasmid pES35 contains the TPI::lacZ
fusion with 853 bp of 5' DNA on a Hind\\\ fragment subcloned into
Ylp56 at Hind\\\. The SphI restriction site at position -220 with
respect to the start of the structural gene was a convenient origin
for exonuclease Bal31 digestion. Therefore, the Sph I site at -220
was rendered unique by a Kpn\ dropout of pES35, which removed a
second Sph I site in the polylinker region of pES35, generating
plasmid pES90. The scheme to introduce Sa/31-induced deletions
into pES90 is shown in Figure 2. Plasmid pES90 was linearized with
Sph\ and treated for various times with exonuclease BAL31. The
Klenow fragment of DNA polymerase I was used to fill in the ends of
the remaining material. Sph\ linkers (HB06) were added. The
material was then digested with Sph I and Hind\\\ to liberate DNA
distal to the Sph I site at -220. The H/ndlll-Sp/7l fragment that
corresponds to the Ba/31 treated DNA of yeast origin distal to the
Sph\ site was gel purified and subcloned into pES90 which had been
digested with Hind\\\ and Sph\. The resulting material was
transformed into E. coli MC1061. Plasmid DNA was prepared, and

Figure 2. Scheme to create internal deletions. The TPI::lacZ fusion construct pES90 contains a unique
Sph\ restriction site at position -220. Plasmid pES90 was linearized with Sphl and digested for
various times with exonuclease Ba/31. Sphl linkers were ligated on and the material was digested
with Hinó\\\ and Sph\. The material remaining 5' to position -220 was isolated by gel purification and
subcloned into pES90 digested with H/ndlll and Sph\. Deletion endpoints were determined by
sequencing with primer HB05.

Hind III
-853
|lUAS? I I LIAS I
Hind III
-853
V
Isolate Hirió III / Sph\ Fragment
Subclone to Sph\ at -220 (wt.)
Sph I
-220
TATA
TPI::lacZ
Sph I
-220
TATA
1 TPI::lacZ
1) Open at Sph I
2) Delete with Bal 31
3) Add Sph I Linkers
rv>
CD

29
precise deletion endpoints were determined by double-stranded DNA
sequencing using primer HB05.
Constructs to assess UAS activity Plasmid plCZ312 is an
integrative yeast shuttle vector that contains a CYC1::lacZ fusion
(Guarente and Mason, 1983). The native CVC1 UAS elements were
removed by Hind 111-Sph\ digest and replaced with various fragments
containing portions of the TPI 5' noncoding region. All constructs
were confirmed by sequencing, using primer HB23 which anneals in
the vector sequence of plCZ312 immediately upstream of the
polycloning site. Plasmid pES90 was digested with H/ndlll and SphI
to remove sequences distal to position -220 of the TPI::lacZ fusion.
Such a digest removes sequence required for expression of the
fusion. Fragments containing portions of the TPI 5' noncoding region
distal to -220 were subsequently subcloned and tested for UAS
activity. Constructs were confirmed by sequencing with primer
HB01.
Mutant UAS oligonucleotides driving expression of the TPI::lacZ
fusion Double-stranded oligonucleotides capable of UASjpi
function, or mutant derivatives thereof, were generated via mutually
primed synthesis. These double-stranded oligonucleotides contained

30
a Hind\\\ site at their 5' end, and a SphI site at their 3' end. The
oligonucleotides were isolated from pUC18 by H/ndlll-Spfrl digest
and subcloned into the Hind\\\-Sph\ sites of pES90. Thus, the
oligonucleotides replaced the native UASjpi element in pES90. The
replacement of UASjpi in the resulting plasmids was confirmed by
DNA sequence analysis utilizing primer HB01.
Site-Directed Mutagenesis
The 1.4 kb Sma\ fragment of plasmid pES90, containing the 5'
noncoding region of TPI, was cloned into the Smal site of M13mp18.
The ligation mixture was used to transform E. coli TG-1 and lysates
were prepared from 10 individual plaques. RF II DNA was prepared
from each isolate and the orientation of the insert determined by
Hind\\\ restriction analysis. One phage, MES-TPImp18B, contains the
strand of 5' noncoding region that corresponds to the antisense
strand of the TPI structural gene, and was used as the target for
site-directed mutagenesis. Oligonucleotides with the desired
mutations were synthesized such that the area of mismatch was
flanked by regions of complementarity of 10 to 12 nucleotides.
These oligonucleotides, HB28-33, are listed in Table 2.

31
Materials utilized to introduce the mutations were supplied by
a kit available from Amersham. The kit is based on the method of
Eckstein and co-workers (Taylor et al., 1985). The method provides
strand-specific selection based on the inability of the restriction
enzyme Nci\ to cleave DNA containing thionucleotides. The mutant
oligonucleotide was annealed to the target, extended with the large
fragment of DNA polymerase I in the presence of the thionucleotide
(dCTPaS).
The resulting material was ligated with T4 DNA ligase to form
complete and unnicked double-stranded plasmid. Excess single-
standed target template was removed via passing through a filter
which binds single, but not double-stranded DNA. Then the parental
strand was specifically nicked with Nci\, which did not cut the
thionucleotide containing mutant daughter strand. The nicked
material was then partially digested with exonuclease III. The
intact daughter strand was then used as a template with DNA
polymerase I and circularized with T4 DNA ligase.
The new double-stranded construct contained the desired
mutation on both strands. The material was used to transform E.
coli strain TG-1, and lysates were prepared from 10 individual

32
plaques. Introduction of the desired mutation was confirmed by
sequencing with primer HB42. RF II DNA was prepared from desired
mutant phage and the Hiná\\\-Sph\ fragment was subcloned into
pES90 to form the mutant constructs. All pES90 derived mutant
constructs were also confirmed by sequencing with primer HB42
prior to integration and assay.
T ransformation
The method of Enea et al. (1975) was used to transform E. coli.
Saccharomyces cerevisiae was transformed by the method of Ito et
al. (1983), selecting for uracil prototropy. Integration of the
TPI::lacZ fusion constructs at the UR A3 locus was achieved by
linearizing the plasmids with StuI which cuts at a unique site
within URA3. The scheme for integration is shown in Figure 3.
Screen for Unit Copy Integrants of TPI::lacZ fusions at URA3
Integrative transformation of yeast cells can lead to tandem
insertional events (Orr-Weaver et al., 1983), Figure 3. To screen for
unit copy integrants, the genomic structure of the URA3 locus was

Figure 3. Scheme for integration of TPI::lacZ gene fusion
constructs at URA3. A) depicts a plasmid containing a TPIr.lacZ
fusion construct and a URA3 selectable marker juxtaposed above the
ura3-52 locus of the yeast genome. The plasmid is linearized within
URA3 to direct integration via homologous recombination. Sacl sites
are indicated. B) represents unit copy integrant at ura3-52.
Digestion with Sacl gives rise to three fragments A, B, and C. C)
represents a tandem integration event. Upon digestion with Sacl,
five fragments are generated: A, 2B, C, and a fourth fragment, D,
which is diagnostic of a tandem insertion event.

34
URA3 r TP!' 'Iasi TP/ „ ual
c.
c.
Sac I Sac I Sac I
UR A3 TP!' 'lacZ TP!
Sac I Sac I
t t
URA3 TP!'
lacZ TPI
Sac I
ura3
A.
B.
D.
B.
C.

35
determined by Southern blot hybridization analysis. Genomic DNA
was isolated from individual transformants. The DNA was then
digested with Sad, run on a 0.8% agarose 1X TBE gel, and capillary
blotted to a Gene Screen nylon membrane according to the
manufactures instructions (DuPont). The filter was hybridized with
a Ylp56 probe generated by random-primer extension. Hybridizations
were carried out at 420C for 16 hours in 50% formamide, 0.2% BSA,
0.2% polyvinyl-pyrrolidone (M.W. 40,000), 0.2% ficoll (M.W. 400,000),
50mM Tris pH 7.5, 0.1% Na-pyrophosphate, 1% SDS. Hybridized
membranes were washed according to manufactures instructions.
Visualization was via autoradiography.
p-Galactosidase Assays
Strains to be assayed were grown from a single colony to an
A6oo between 0.5 and 1.5 at 30OC. p-galactosidase assays were
performed essentially by the method of Miller (1972). The units
reported correspond to AA42o/minute/A6oo of the initial culture.

36
In vitrQ DNasel PrQtectiQn A??qy The method used in the in vitro DNasel protection assays was a
modification of the method of Singh et al. (1986). The 228 bp Hind\\\-
Sph\ fragment of pES40-23 or the 169 bp Hind\\\-Sph\ fragment of
pES34 were end-labeled by filling in the Hind\\\ site with the
Klenow fragment in the presence of 32p-dATP. A protein extract
was prepared from yeast S150-2B by lysis with a French pressure
cell at 20,000 psi. Rabbit reticulocyte lysates (RRL) containing
RAP1 protein were generated via in vitro transcription and
translation, and provided to me by C. Lopez.
Multiple aliquots of 2pl of the end-labeled fragments (20,000
cpm/ul) were incubated with either 5pl of yeast extract or 5pl RRL
containing RAP1 in 18pil of 1X binding buffer [12mM HEPES pH7.5,
60mM KCI, 5mM MgCl2, 4mM Tris, 0.6mM DTT, 10% glycerol, 0.26
ug/ul poly(dl-dC), and 0.3 pg/pl BSA]. Incubation was for 20 minutes
at room temperature. 0.5 units or 1.0 units of DNase I was then
added to quadruplicate aliquots. The reactions were incubated at
room temperature for 2 minutes, then stopped by the addition of
10(il of Stop solution (0.25M EDTA, 25% glycerol) on ice. The entire

37
reaction was then loaded onto a 5% polyacrylamide, 0.5X TBE gel
running at 5 volts/cm. The gel had been pre-run for at least 1 hour.
The samples were electrophoresed at 7.5 volts/cm until the
bromphenol blue tracking dye, loaded along with the samples, was 2
cm from the bottom of the gel. The wet gel was wrapped in plastic
film and exposed to X-ray film overnight at 40C.
The developed film was used as a guide to excise the shifted
and unshifted portions of the radiolabeled fragment from the gel.
Identical positions from the quadruplicate samples were pooled for
extraction from the gel slice. Fragments were extracted from the
gel fragments by incubation at 370C overnight in 3 volumes of 0.5M
NH4Ac, 1mM EDTA. The extracted fragments were precipitated by
the addition of 2.5 volumes ethanol and centrifugation. The pelleted
fragments were washed with 70% ethanol and dried. The dried
pellets were Cherenkov counted and resuspended to 4,000 cpm/pl in
formamide sequencing dye.
A control DNase I ladder was generated from the unprotected
fragments by digestion for 2 minutes with 0.02 units of DNase I.
The control ladder was pelleted, counted, and resuspended in the
same manner as the sample fragments. A sequencing ladder was

38
generated from M13mp18 with the forward sequencing primer to
serve as a molecular weight standard.
The shifted sample fragments, bracketed by the unshifted and
control DNasel ladder, were electrophoresed next to the sequencing
standard on a 0.4mm-7% polyacrylamide, 8M urea, 0.5X TBE gel at 50
volts/cm. After electrophoresis, the gel was blotted to Whatman 3M
paper and dried. The dried gel was exposed to X-ray film for
visualization.
DNA Band Shift Assays
DNA band shift assays were performed as previously described
(Scott et al., 1990). E. coli extracts were prepared by lysis with a
French pressure cell at 20,000 psi.
In vivo Methvlation Protection Analysis
Yeast strains were cultured in 2 liters of YP medium (Sherman
et al., 1983) supplemented with either 2% glucose or 2% lactate and
2% glycerol. Cultures were harvested at an A6oo of 1-0 by
centrifugation. Cells were washed twice in 137mM NaCI,2.7mM KCI,
4.3mM NaPCU, 1.4mM KPO4 pH7.4 (PBS), and concentrated to 1x 108

39
cells per ml in PBS (35ml final volume). 5ml aliquots were placed in
50ml disposable, sealable plastic tubes on ice. Dimethyl sulfate
(DMS) was added to 0.5% final concentration to various aliquots (5
pl/aliquot), and incubated from 1-6 minutes at room temperature.
Reactions were quenched by the addition of 45ml of ice-cold PBS.
Treated cells were harvested by centrifugation (1,000x g for 10
min.) and washed twice in 35ml ice-cold PBS. DMS waste was
allowed to decay in a fume hood for one week prior to disposal.
Methylated genomic DNA was prepared essentially by the
method of Sherman et al. (1983), except all incubations were carried
out at 370C or less to prevent unwanted cleavage events. The DNA
preparation was treated with RNase A at a final concentration of
50pg/ml for one hour. Two phenol extractions, one
phenol/chloroform extraction, and one chloroform extraction were
performed. 1/10 volume of 5M NH4AC was added, followed by
precipitation with 2 volumes of ethanol. DNA was pelleted by a low
speed spin (500x g, 5 min.), washed twice with 70% ethanol, and
dried in vacuo. The dried pellets were resuspended in 250 pi 10mM
Tris pH 8.0, 1.0 mM EDTA (TE); and the A260/280 was determined.

40
Control DNA was prepared simultaneously from an untreated aliquot
of the original cells.
50|ig of methylated DNA from each time point and 200pg of
untreated DNA was cleaved to completion with Avail. 1/10 volume
5M NH4AC was added, followed by 2.5 volumes ethanol. Samples
were centrifuged 10,000x g for 30 minutes. The pellet was washed
twice with 70% ethanol and dried in vacuo. Methylated DNA was
resuspended in 200pl 1M piperidine and incubated at 95°C for 30
minutes. 40pl aliquots were removed for alkaline agarose gel
electrophoresis. 16|il of 3M NaAc pH 6.0 was added to remaining
sample, followed by precipitation with 2.5 volumes ethanol.
Samples were centrifuged 10,000x g for 30 minutes, washed twice
with 70% ethanol, and dried in vacuo. 50pil of TE pH 8.0 was used to
resuspend samples, and the concentration of the was determined
spectrophotometrically.
The aliquot of the piperidine cleavage reaction was ethanol
precipitated and resuspended in formamide sequencing dye. Samples
from the various timepoints were denatured at 95°C and loaded onto
a 1.5% Agarose, 50mM NaOH, 1 .OmM EDTA gel next to molecular

41
weight markers. The gel was electrophoresed in a 50mM NaOH,
1.0mM EDTA running buffer at 30 volts and 175 milliamps overnight.
The gel was neutralized in 1M Tris pH 7.5 for one hour, stained with
0.5ug/ml ethidium bromide, and visualized with UV light. Samples
generated a smear with a definite peak size of fragment. Time
points with a peak fragment size of approximately 500-800 bp were
used for subsequent steps.
Control sequences were generated from the 200pg of Avail
digested untreated DNA. The DNA was divided into four aliquots and
ethanol precipitated, washed twice with 70% ethanol, and dried.
Control G-reactions were as follows. One aliquot was resuspended
in 200|il DMS buffer (50mM Na-cacodylate, 1.0mM EDTA pH 8.0). 1 pil
of DMS was added for 1.5 minutes at room temp. 50pl ice-cold DMS
stop solution was added (1.5M NaAc pH 7.0, 1M 2-mercaptoethanol),
and 750pl ice-cold ethanol was added. The reaction mixture was
then placed on dry ice.
The control A+G-reactions were as follows. The DNA aliquot
was resuspended in 11 pi ddH2Ü. 25pl of concentrated formic acid
was added at room temperature and incubated 5 minutes. 200pl ice-

42
cold hydrazine stop solution was added (0.3M NaAc pH 7.5, 0.1 mM
EDTA). 750|il of ethanol was added, and the mixture placed on dry
ice.
The control C-reactions were as follows. The DNA aliquot was
resuspended in 5pl ddH2Ü, and 15pl of 5M NaCI added. 30pl hydrazine
was added and incubated 16 minutes. The reaction was stopped by
the addition of 200pl ice-cold hydrazine stop solution. 750pl of
ethanol was added, and the mixture was placed on dry ice.
The control T+C-reactions were as follows. The DNA aliquot
was resuspended in 20pl dd^O. 30|il of hydrazine was added and
incubated for 16 minutes. 200pl ice-cold hydrazine stop solution
was added. 750pl of ethanol was added, and the mixture placed on
dry ice.
All reactions were processed as follows. Samples were
centrifuged at 10,000x g for 15 minutes, and pellets washed with
70% ethanol. (Hydrazine waste was neutralized by placing in a
saturated ferric chloride/water solution.) Pellets were resuspended
in 225pl ddHgO, 25|il of 3M NaAc pH 6.0 and 750pl of ethanol was
added. Following centrifugation at 10,000x g for 30 minutes,

43
pellets were washed twice with 70% ethanol, and dried. Pellets
were dissolved in 10Opil of 1M piperidine and incubated at 95°C for
30 minutes. 11 pi of 3M NaAc pH 6.0, and 230pl ethanol was added.
Samples were centrifuged 30 minutes, washed twice with 70%
ethanol and dried in vacuo overnight. Pellets were dissolved in
1 OOpil TE pH 8.0 and the concentration of the material was
determined spectrophotometrically.
5pg of all sample and control DNA's were lyophilized,
resuspended in 4pl formamide sequencing dye, and denatured at 950C.
A genomic sequencing ladder followed by the in vivo methylated
samples bracketed with G-reaction control ladders were
electrophoresed in a 60cm, 7% (40:1.3) polyacrylamide, 8M urea gel
in 0.5X TBE running buffer at 3500 volts and 35 milliamps. After
running, the gel is lifted using Whatman 541-sfc paper, and
electroblotted to a Hybond N+ membrane (Amersham). The DNA was
UV cross-linked to the membrane with a FisherBiotech 312 nm
variable intensity transilluminator at full power for 5 minutes. The
membrane was prehybridized in 20ml of hybridization buffer [1.0%
bovine serum albumen, 7.0% SDS, 1.0mM EDTA, (0.5M Na) HPO4 pH 7.2]
at 630C in a roller incubator.

44
The TPI specific probe was prepared as follows. Single-
stranded M13 phage containing the sense (mES2-2) or anti-sense
(mESTPImp18B) strand of TPI, spanning the 5' non-coding region and
the beginning of the structural gene, was prepared by standard
techniques (Current Protocols in Molecular Biology). 6pl of phage
(0.25pg/|il) was incubated with 5pl of primer(0.5pM/pl), and 2.5|il of
10X klenow buffer (0.5M Tris pH 8.0, 2M NaCI), at 50OC for 30
minutes. The sense probe was prepared by annealing primer HB54 to
mESTPImp18B. The antisense probe was made by annealing primer
HB60 to mES2-2. The following reagents were then added in order:
5pl of 0.1 M DTT, 5pl of 50mM MgCl2, 2pl of 3.0mM dNTP-dATP mix,
10 units of DNA polymerase Marge subunit, and 10pl of 32p. dATP
at 3,000 Ci/mM. The antisense reaction also included 2pl of Avail to
cleave the probe to the appropriate length. The reaction was
incubated at 370C for 45 minutes. 120(il of formamide sequencing
dye was then added. The probe was denatured at 950C for 10
minutes and run into a 6% (40:1.3) polyacrylamide, 8M urea gel. The
wet gel was exposed to Polaroid type 57 film for 15 minutes. The
film was developed and used as a guide to excise the probe from the

45
gel. The gel slice was crushed and 8ml of hybridization buffer was
added.
The pre-hybridization solution was removed from the
membrane and the gel/hybridization solution mixture added to the
roller tube. Hybridization was performed at 630C overnight.
Following hybridization, the membrane was washed in 500ml
aliquots of (40mM Na) HPO4, 1.0mM EDTA, 1% SDS at 60OC until wash
was no longer radioactive when examined with a geiger counter. The
damp membrane was wrapped in plastic wrap and exposed to film for
visualization.

RESULTS
The Mature 5' Ends of Steadv-State TPI Transcripts are Unaffected
by a gen Mutation
Previous studies suggested that gcr1 lesions bring about a
reduction in the levels of mRNAs specifying glycolytic enzymes
(Clifton and Fraenkel, 1981; Holland et al., 1987). An initial
objective was to characterize the TPI transcript in both wild-type
and gcr1 mutant strains of Saccharomyces cerevisiae. RNA gel
transfer hybridization experiments in both strains demonstrated a
reduced steady-state level of the TPI transcript in a gcr1 mutant
strain (Scott et al., 1990). The reduction in steady-state levels of
the transcripts suggests that the GCR1 gene product may play a role
in the transcriptional regulation of the genes encoding glycolytic
enzymes.
To further characterize the TPI transcript, it was of interest
to determined if the mature 5' end(s) of the transcripts were
affected by deletion of GCR1. Differences in the mature 5' end of the
46

47
TPI transcript isolated from wild-type and gcr1 mutant strains may
reflect altered transcriptional start sites. Such a result may
indicate the existence of two promoter elements, one of which is
GCR1-dependent and one of which is GCR1-independent. The TPI
transcript remaining in the gcr1 mutant could originate from a GCR1-
independent promoter and initiate at a unique start site. A GCR1
independent TPI transcript may be the product of a different
transcriptional mechanism than the abundant transcript in wild-type
strains. Transcripts of a single gene derived from independent
mechanisms are known to exist for H/S4 which has both inducible
and basal transcripts initiating from independent sites (Arndt et al.,
1987; Pellman et al., 1990).
Mapping the mature 5' end of the TPI transcript was
accomplished using primer extension analysis. 10pg of wild-type
(S150-2B) or 25pg of gcr1 mutant (HBY4) total RNA was annealed
with a radiolabeled oligonucleotide primer (HB05) corresponding to
the start of the TPI structural gene. Extension reactions were
carried out with reverse transcriptase in the presence of
deoxyribonucleotide triphosphates. The first base incorporated
corresponded to position +2 with respect to the adenine of the

48
initiation codon. Extension products were resolved on a denaturing
polyacrylamide gel next to a DNA sequence of the TPI 5'
nontranslated region which was generated with the same primer
used in the extension reactions. Figure 4 shows the results of the
primer extension experiment. In both the wild-type and gcr1 mutant
strain the predominant mature 5' ends of the TPI transcript were
identical. The ends corresponded to a pair of adenines at positions
-29 to -30 with respect to the initiation codon. Therefore, it
appears likely that TPI transcription is controlled by a single
promoter element, and that the residual expression of TPI observed
in gcr1 mutant strains is the result of transcription originating at
the native start site.
Identification of the 5' Boundary of the TPI Controlling Region
In order to map the controlling element(s) of TPI a TPI::lacZ
gene fusion was utilized. This gene fusion produced a protein that
was a hybrid between triose-phosphate isomerase and (3-
galactosidase, which retained (3-galactosidase activity. Use of the
fusion in trans to TPI allowed normal expression of TPI which is
required for cellular growth while allowing the manipulation of the

Figure 4. Primer extension analysis of the 5' ends of the TPI
transcript in wild-type and grcri-deletion mutant strains. (A) Lane
1, extension products generated with reverse transcriptase from
10pg of total RNA isolated from wild-type strain S150-2B
hybridized with a radiolabeled primer corresponding to the start of
the TPI structural gene. Lane 2, extension products from similar
reaction carried out with 25pg total RNA from gcr1 mutant strain
HBY4. Lanes 3 through 6, DNA sequencing ladder generated with the
aforementioned primer from the TPI 5' nontranslated region. Bold
double arrow indicates the major extension products. P, denotes the
position of the unextended primer. (B) DNA sequence of the TPI 5'
nontranslated region. Double arrows indicate the bases that
correspond to the predominant 5' ends of the TPI transcript.

| — T3 -h| Qh>[»»hh»
>
H
I
i in
°
n:
I I,,, 111
II I III
GO
O > H > O > O 5
H
•>>0>H0>>H>H0H>]
I
8
co
o
I
$;
cn
o

51
TPIr.lacZ fusion. The effect of the manipulations on the expression
of TPIr.lacZ could be measured indirectly by monitoring p-
galactosidase activity.
The plasmid pHB110, a derivative of pUC18, harbors the
TPIr.lacZ gene fusion with 3.5 kilobase pairs (kbp) of DNA 5' to the
start of TPI. 3.5 kbp. of DNA 5' to TPI was initially included in order
to ensure that all c/'s-acting regulatory elements necessary for TPI
expression were included in the construct. Sequencing of the 5'
non-coding region identified a unique 77/71111 site at position -853,
Figure 5. Subcloning experiments showed that all sequences
sufficient for high-level expression of the fusion resided within the
77/71111 site. Therefore, the site could be used as an origin for a
series of nested deletions created through the action of exonuclease
Bal31. Figure 1 shows the scheme used to create the Bal31
deletions from plasmid pHB110. These deletions were used to map
the 5' boundary of the region sufficient for high-level expression of
the TPIr.lacZ gene fusion.
The fusion construct was linearized with 77/71111 and treated
for various times with exonuclease Bal31. H/ndlll/Sa/l/H/ndlll
linkers (HB03, table 2) were then ligated to the material. This

FIGURE 5. Sequence of TPI 5’ Noncodina Region
52
I -1191
acgtcatcgatgaatataatgaattaaacagtggtgttcgtatatgtgaagatatgagatatga
tccacatggtaaacagaaagatgcattttggccgagaggacttaataatactggtggtgtttac
gaaaataatgaagataatatttgtgaagggaagcctggaaaatgttatctgcaatatcgggtta
aggatgagccaagaataagggaacaagattttggtaatttccaaaaaatcaatagcatgcagg
acgttatgaagaagagatctacgtatggtcatttcttcttcagattccctcatggagaaagtgc
I -853
aacaaatatatataacaaaatcaccaatttccaaaaaactttattcaaacacttccataataaa
caagagagaagacccagagatgttgttgtcctagttacacatggtatttattccagagtattcc
tgatgaatggtttagatggacatacgaagagtttgaatcgtttaccaatgttcctaacgggagc
I -658 (Previously Known Sequence)
gtaatggtgatggaactggacgaatccatcaatagatacgtcctgaggaccgtgctacccaaa
tggactgattgtgagggagacctaactacatagtgtttaagattacggatatttaacttactta
gaataatgccatttttttgagttataataatcctacgttagtgtgagcgggatttaaactgtgag
gaccttaatacattcagacacttctgacggtatcaccctacttattcccttcgagattatatct
aggaacccatcaggttggtggaagattacccgttctaagacttttcagcttcctctattgatgt
tacacctggacaccccttttctggcatccagtttttaatcttcagtggcatgtgagattctccg
aaattaattaaagcaatcacacaattctctcggataccacctcggttgaaactgacaggtggtt
I -220
tgttaccicatcictaatgcaaaggagcctatatacctttggctcggctgctgtaacagggaatat
aaagggcagcataatttaggagtttagtgaacttgcaacatttactattttcccttcttacgta
aatatttttctttttaattctaaatcaatctttttcaattttttgtttgtattcttttcttgcttaa
I +1
atctataactacaaaaaacacatacataaactaaaaATG

53
linker was designed such that upon subsequent digestion with Sal I
and religation of the vector each deletion endpoint would be marked
with a H/'ndlll site. Precise deletion endpoints were determined for
65 individual constructs by using the dideoxy sequencing method of
Sanger (1977) as modified by U.S. Biochemicals. Once the 5' deletion
series was obtained it was necessary to place the constructs back
into yeast to determine the effects of the deletions upon expression
of the fusion. Of the 65 plasmids sequenced 13 were chosen because
they provided a well spaced, nested set of 5' deletions for study.
The constructs were subcloned into the yeast integrative
plasmid 56 (Yip56) which contains a URA3 selectable marker for
yeast. Yip56 also contains an origin of replication and an ampicillin
resistance determinant for propagation and selection in E. coli. An
integrative yeast shuttle vector was chosen to avoid effects on
fusion expression due to plasmid copy number discrepancies brought
on by the high segregation rate of yeast plasmids (Botstein and
Davis, 1982). The isogeneic uracil auxotrophs S150-2B and HBY4
were transformed to uracil prototropy with the deletion constructs.
Integration of the fusions was directed to the URA3 locus by
transforming with plasmid DNA linearized with Stu\ which cuts

54
within URA3. This procedure served to direct the site of integration
to the URA3 locus via homologous recombination (Orr-Weaver et al.,
1983). Transformation competent yeast cells can take up multiple
copies of plasmid DNA, raising the possibility that tandem
integration events can occur (see Figure 3). Therefore, screening for
unit copy integrants was required in order to assure the most
accurate expression data possible for the 5' deletion series.
Genomic DNA was isolated from individual transformants and
digested with Sacl, then subjected to Southern blot analysis probing
for URA3. Sacl cut outside of the URA3 locus but within the plasmid.
The URA3 probe hybridized to a single band in experiments with DNA
isolated from the parental strain, S150-2B. A unit copy integrant
was distinguishable by the presence of two fragments which
hybridized to the probe, labeled A and C in Figure 3. These are the
two junction fragments between the yeast chromosome and the
integrated plasmid. Upon a tandem integration a characteristic
band, totally plasmid derived, was also observed, labeled fragment D
in Figure 3. Presence of the totally plasmid derived fragment D was
used as a diagnostic for multiple integration events. Utilizing this

55
screen, unit copy integrants were obtained for all constructs in
S150-2B (GCR1) and all but two constructs in HBY4 (gcr1).
The stability of the fusion constructs integrated at URA3 was
determined by growing a strain carrying an integrated construct in
non-selective media (YPD) for >10 generations. The cells were then
plated on non-selective media and 100 individual colonies were
screened for their ability to grow on selective media (YNB with Glu,
His Leu, Trp). Each colony tested was URA+. Furthermore, each
colony still expressed p-galactosidase activity, which indicated the
presence of the integrated fusion construct. Integration into the
yeast genome was a stable method for carrying the fusion with a
segregation rate of less than one percent.
Once strains were isolated that carried the 5' deletion series
integrated in unit copy, they were assayed for p-galactosidase
activity in order to determine the effect of the deletions on the
expression of the TPI::lacZ gene fusion, p-galactosidase assays were
carried out in duplicate on three separate occasions. Figure 6
depicts the assay results. Plasmid 92-9 contains the entire 3.5 kb
of 5' non-coding region. When integrated in the S150-2B background,
276 Miller units of activity were expressed. 92-9 integrated in the

Figure 6. Effect of 5' deletions upon expression of a TPI::lacZ fusion integrated in unit copy at the
URA3 locus in wild-type and gcr1-deletion mutant strains. Constructs 92-9, 92-9R, and 35-2, 35-2R
contain the same respective deletion endpoints but are integrated in opposite orientation at the URA3
locus. Deletion endpoints are indicated with respect to the start of translation, p-galactosidase
activities were determined by the method of Miller (1970) in duplicate from at least three independent
cultures. Strains were grown to an optical density A6oo of approximately 1.0 in YP medium
supplemented with either 2% glycerol and 2% lactate (YPGL), or 2% glucose (YPD).

Fusion # P- Galactosidase Activity
- -3500
-220
Sph I
l
♦ 1
ATG
Ltpi' ESI
S150-2B (GCR1)
YPGL YPD
Mean SD Mean SO
HBY4 (gcr1)
YPGL
Mean SD
92-9
llacZ E m-TPt ESI
276
11
222 6
19 3
- -3500
254
215 21
17 1
92-9 R ‘
yj ip r fcxj
llacZ E ESI TP 1 ESI
4
-853
210
236 25
21 4
35-2
htpi' \¡a
lacZ IítBMTPI ESI
11
-853
276
251 8
ND
35-2R
eíitpi m
llacZ E BSJ'TPI ESI
18
36-2
-490
241
238 33
22 5
til TP1 m
llacZ E ESll TPl EiS
36
37-2
-420
266
255 43
20 3
WTPr M
HacZ t ^'TPI 1ft
9
34-1
-392
251
241 24
20 2
HTPI' EE3
HacZ ESiMTPI M
36
70-1
-377
107
69 4
29 3
EsiTPi m
llacZ E KSTPI ESJ
8
-348
27
21 3
18 3
83-1
EÍTPI SSI
HacZ t ESlTPI ESI
7
73-2
-337
19
^tpi m
HacZ E ESHTPI ESI
8
9 6
4 4
-330
16
82-1
^TPI !:'H
llacZ E ES3TPI ESI
9
6 3
3 3
-299
14
13 4
77-2
htpi' tsa
HacZ t ®'TPI EÍS
3
6 3
78-1
-278
11
8 4
10 8
STPI' ESSE
HacZ E ESI 1P EK1
10
-192
36
10 7
ND
76-1
Eii tp i'm
Hacz t eaTPi Esa
11
-179
17
12 1
38-4
BiTPi' M
HacZ ¡ Eil'TPI m
11
5 4
39-2
+63
14
8 8
H'TPI 1
llacZ | eshtpi esi
7
4 3
No Fusion
9
3
4 3
3 3

58
gcr1 mutant HBY4 background only expressed 19 units of activity.
Thus the gcr1 mutation resulted in a 12 fold reduction in the level of
expression of the TPI::lacZ fusion. Clifton and Fraenkel (1981)
previously reported a 17 fold reduction in the level of triose-
phosphate isomerase in a gcr1 mutant background. Thus, expression
of the TPI::lacZ, like TPI itself, is dependent on GCR1 for full
expression. Plasmids 92-9 and 92-9R, 35-2 and 35-2R contain the
same respective 5' deletion endpoint, but the TPI::lacZ fusion was
integrated in opposite orientation with respect to URA3. No effect
on expression of the fusion due to orientation with respect to URA3
was observed. Strains harboring deletions up to position -392 still
express high levels of p-galactosidase activity. However, deletion
of an additional 15 bp to position -377 reduces expression
approximately two-fold to 107 units of activity. Deletion to -348
or beyond abolished expression of the fusion. Based on these
results, the 5' boundary of the region sufficient for high-level
expression of TPI::lacZ must reside within 392 to 377 base pairs
from the start of the TPI structural gene, when the fusion is
integrated at the URA3 locus.

59
An Upstream Activating Sequence Activity for TPI Resides from
Position -377 to -327 in the 5' Non-Coding Region
The 5' boundary of the TPI controlling region mapped by the
initial deletions was probably the boundary of an upstream
activating sequence (UAS). UAS's of yeast are similar to enhancer
elements of higher eukaryotes. UAS's are the sites of interaction
between positive trans-acting proteins and specific DNA target
sequences, and serve to activate transcription. Others in the
laboratory who utilized fragments isolated from the initial deletion
series in DNA band shift assays, demonstrated a specific protein
nucleic acid interaction involving the region between -377 and -327
(Scott et al., 1990). Further band shift experiments determined that
the protein responsible for the shift was RAP1. As noted in the
introduction, it has been recently reported that RAP1 binding is
important for the UAS activity of other genes encoding glycolytic
enzymes such as PGK1 (Ogden et al., 1986; Chambers et al., 1989),
EN01 (Buchman et al., 1988; Brindle et al., 1990), EN02 (Brindle et
al., 1990), PYK (Buchman et al., 1988), PDC1 (Kellerman and
Hollenberg, 1988), and ADH1 (Buchman et al., 1988; Santangelo and
Tornow, 1990). I wanted to determine if the region downstream of

60
position -392 of TPI encoded an UAS, and if RAP1 binding was
required for expression of TPI.
To address this question, fragments of the TPI controlling
region were used to replace the native UAS elements of a CYC1 ::lacZ
gene fusion. A Hind\\\-Sph\ fragment (-392 to -220) isolated from
the last 5' deletion construct to drive high-level expression of
TPI::lacZ (34-1) was able to restore expression to a CYC1 ::lacZ gene
fusion which had its native UAS elements removed (Figure 7). Cells
containing a CYC1::lacZ gene fusion which had a 66 base pair
oligomer (HB14 and HB15, Table2){TPI sequence from -377 to -327)
substituted for the native UAS elements of CYC1 were able to
express 99 units of (3-galactosidase activity. This level of
expression was comparable to the 94 units of activity expressed
using the longer TPI fragment, 34-1 H-S (-392 to -220) or the 101
units expressed by cells containing the 81 base pair oligonucleotide
(HB16 and HB17, Table 2)(-392 to -327). Therefore, UAS activity is
bounded by positions -377 to -327 which contains the RAP1-binding
site from -358 to -346. The sequence from -377 to -327 will be
termed UAStpi-
The same 51 base pairs was able to express 129 units of p-
galactosidase activity from the TPI::lacZ fusion which was deleted

Figure 7. Identification of UAStpi- Various fragments containing portions of the 5' nontranslated
region of TPI were cloned before either a CYC1 ::lacZ construct deleted of the native UAS elements or a
TPI::lacZ construct deleted of the native UAS elements, p-galactosidase assays were performed from
cultures grown in YPD as described in figure 5 . Fragment 34-1 H-S contains the sequence form -392
to -220 of TPI. 81mer contains from position -392 to -327. 66mer contains from position -377 to -
327. 58mer contains from -377 to -335. 69RBSM is identical to the 66mer except that the sequence
from positions -349 to -339 (CACCCCTTTTC) was replaced with the sequence AACCCATCAGG.

Construct Arrangement
CYC1:: lacZ
—HAS TATA
UAS|ess CYC1:: lacZ
**
( ) TATA
34-1 H-S:: CYC1:: lacZ
&
(34-1 H-S) TATA
81mer:: CYC1:: lacZ
£>
(81 mer) £ TATA
66mer:: CYC1:: lacZ
£>
(66mer) ¿ TATA
35-2
¿S’
2 TATA
78-1
A TATA
58mer:: TPI:: lacZ
$
58mer \ )-tata —
66mer:: TPI:: lacZ
<$â–  <{P
( j-TATA
69 RBSM:: TPI:: lacZ
$ $
SS Rbsm| )_tata —
3 - Galactosidase Activity
Mean
CYC1 ::lacZ 70
CYC1 ::lacZ 8
CYC1 ::lacZ 94
CYC1 ::lacZ 101
CYC1 ::lacZ 99
TPI ::lacZ 236
TPI ::lacZ 8
â–  TPI ::lacZ 20
â–  TPI ::lacZ 129
TPI ::lacZ 18
CD
N>
SD
9
6
3
5
7
25
4
4
11
4

63
of all sequences 5' to the SphI site at -220. However, a 58 base pair
oligonucleotide (HB07 and HB08, table 2) with TPI noncoding
sequence from -377 to -335 was unable to restore expression the
TPI::lacZ fusion, Figure 7. An oligonucleotide (69RBSM or HB19,
table 2) identical to the native 51 b.p. region from -377 to -327
except for mutations in the RAP1-binding site was prepared. The
mutant oligonucleotide was placed before the same deleted TPI::lacZ
fusion as the wild-type oligonucleotide to assess the role of the
RAPI-binding site in UAStpi function. The mutated oligonucleotide
was unable to drive expression of the fusion demonstrating that the
RAP1-binding site from -358 to -346 is essential for UAS activity.
Others in the laboratory performed DNA band shift assays with these
oligonucleotides which indicated that RAP1 binds with much reduced
affinity if at all to the mutant oligonucleotide in vitro (Scott et al.,
1990).
Internal Deletions Indicate Single UAS Element Responsible for TPI
Transcription
The 5’ deletion series facilitated the mapping of an upstream
activating sequence, UAStpi, that was sufficient to drive expression

64
of TPI. However, controlling regions in Saccharomyces cerevisiae
are often composed of multiple UAS elements, any of which are
independently capable of driving expression (Guarente et al., 1984;
Cohen et al., 1986; Cohen et al., 1987; Johnston, 1987; Nishizawa et
al., 1989). One-tailed deletion series are able to map only the last
element remaining which is sufficient for expression. In order to
determine the number of UAS elements present in the controlling
region of TPI, a second set of deletions was generated which began
internal to the known UASjpi at position -220 and extending towards
and through UAStpi located from -377 to -327. Figure 2 cartoons the
scheme utilized to generate the internal deletion series.
Deletions originating internal to and extending through UAStpi
of the TPI::lacZ fusion in plasmid pES90 ("internal" deletions) were
created originating from a unique Sph\ site at position -220.
Linearized plasmid pES90 was digested with exonuclease Bal 31,
followed by the addition of SphI linkers (HB06, table 2). To assure
an intact TATA element, required for expression, the deletion
products distal to the Sph I site were isolated after H/ndlll digest by
gel purification then subcloned into plasmid pES90 which had been
digested with Hind\\\ and Sph\. The final "internal" deletion series

65
had a SphI site at position -220 that served to fix the 3' deletion
endpoint at position -220. All "internal" deletion constructs had
position -853 as their common 5' end. Precise 3' deletion end-points
and junctions were determine by sequencing using primer HB05,
Table 2, and are shown in Figure 8. The "internal" deletion series
was integrated into both wild-type (S150-2B) and gcr1 mutant
(HBY4) strains of yeast. Unit copy integrants were confirmed by
Southern blot analysis (Southern, 1975) as previously detailed, p-
galactosidase assays by the method of Miller (1972) were carried
out in duplicate on three independent occasions.
Figure 8 depicts the "internal" deletion constructs and the
results of the p-galactosidase assays from lysates of strains with
unit copy integrants of the various plasmids. 392 base pairs of 5'
non-coding region was sufficient for high-level expression of the
TPI::lacZ fusion. The initial "internal" deletion construct contained
5' sequence from -853 to -300 and from -220 through the structural
gene. This construct retained the known UAStpi sequence, and
produced 138 units of p-galactosidase activity when integrated into
the wild-type strain, S150-2B. 52 units of p-galactosidase activity

Figure 8. Effect of internal deletions on p-galactosidase activity expressed from a TPI::lacZ gene
fusion. The cartoon depicts the extent of the internal deletions with precise deletion endpoints
indicated. (3-galactosidase assays were performed in duplicate by the method of Miller (1972) on three
individual occasions. Plasmids with deletions in the 5' noncoding region of TPI were integrated in unit
copy at the UR A3 locus in both wild-type (S150-2B) and gcr1 mutant (HBY4) strains of yeast.

Sph I
-220
ruÁsn \
\TPI::lacZ 1
-392
1 TPI::lacZ 1
ruÁsn \
-377
1 TPI::lacZ 1
Í
-348
1 TPI::lacZ 1
-853
-853
-853
rOÁSTl I UAS I -300
IUAS? I -367
\TPI::lacZ
\TPI::lacZ I
No Fusion
B- Galactosidase Activity
S150.2B [GCR1)
Mean SD
HBY4 (gcr1)
Mean SD
210 11
21 4
251 36
20 2
107 s
29 3
27 7
18 3
138 19
52 19
52 3
20 4
15 2
7 2
14 2
6 2
12 5
3 1
CD
-si

68
were expressed in the gcr1 mutant background, only a 2.5 fold
reduction in expression. The lessening of the severity of the
reduction caused by the gcr1 lesion may be due to a position effect
such as was seen for EN02 gene (Holland et al., 1990). The 138 units
of activity in the wild-type background corresponds with the
expression observed when only UAStpi was driving expression of the
fusion (Figure 7). Deletion of an additional 36 base pairs from the 5'
noncoding region of the TPI::lacZ gene fusion, removal of sequence
from -336 to -220, reduced expression of the fusion approximately
three-fold to 52 units of (3-galactosidase activity. The strain
harboring the construct that removed sequence from -347 to -220
yielded background levels of expression. Constructs which are
deleted of the known UAS element but retain all sequences to
position -853, are unable to drive expression. Therefore, no
additional UAS elements lie distal to UAStpi-
Mutational Analysis of UAStpi
UAStpi has been mapped to a 51 base pair region from position
-377 to -327 with respect to the start of the structural gene.
Figure 9 shows the sequence of the 5' non-coding region of TPI

69
-430
37-2 [266]
'(-420)
34-1 [251]
'(-392)
70-1 [107]
(-377)
T AT AT CTAGG AACCCAT CAGGTTGGT GG AAG ATT ACCCGTT CT AAG ACTTTT C AGCTT CCT CT AT
(-377)
83-1 [27]
(-348)
73-2 [19]
‘(-337)
77-2 [16]
(-314)
TGATGTTACACCTGGACACCCCTTTTCTGGCATCCAGTTTTTAATCTTCAGTGGCATGTGAGATTC
x
(-327)
T CCG AAATT AATT AAAGC AAT CACAC AATT CT CTCGG ATACC ACCT CGGTT G A AACT G AC AGGTG'
-234
Á
Figure 9. Summary composite of the 5' noncoding region of TPI.
Deletion endpoints of the 5' deletion series are denoted with the
construct number in bold print, positions are in parentheses. (3-
galactosidase activities expressed by wild-type strains harboring
the 5' deletions are indicated in brackets. UAStpi is underlined. The
RAP1-binding site is underscored by dashes with mismatches with
the consensus RAP1-binding site underscored by X's. The CTTCC and
CATCC pentamer motifs are overdotted.

70
flanking the UAS element. The RAP1-binding site from -358 to -346
has been shown to be required for UAS activity in wild-type yeast
(Figure 7). Recently, other laboratories have demonstrated the
requirement of a CTTCC pentamer motif for the full expression of
PGK (Stanway et al., 1989), PYK, and EN01 (Buchman et al., 1988).
Examination of UAStpi revealed a CTTCC pentamer located from
position -375 to -370, and a closely related CATCC pentamer from
-335 to -330 (Figure 9).
Mutant variations of UAStpi were generated and cloned at
position -220 before a TPI::lacZ fusion. The constructs were
integrated in unit copy into wild-type (S150-2B) and gcr1 mutant
(HBY4) strains of yeast. UAStpi, -377 to -327 (HB14 and HB15, Table
2), expressed 115 units of (3-galactosidase activity in the wild-type
background and 18 units of activity in the gcr1 mutant background,
Figure 10. It was interesting to note that expression of the UAStpi
construct was reduced 10-fold in the gcr1 mutant background. Thus,
GCR1 must act through UAStpi sequence, or through a sequence
downstream of position -220. The role of the RAP1-binding site in
UAStpi activity was addressed by an oligonucleotide, HB24 (Table 2),

Figure 10. Mutational analysis of UAStpi utilizing mutant
oligonucleotides. Double-stranded oligonucleotides with portion of
the 5' noncoding region of TPI were cloned before a TPI::lacZ fusion
at position -220. Positions within the 5' noncoding region are
indicated. The large box denotes the RAP1-binding site from
position -358 to -346. The 5' small box denotes the CTTCC pentamer
motif located from -375 to -370. The 3' small box denotes the
CATCC pentamer motif from position -335 to -330. Filled boxes
represent mutated motifs as detailed in the text. Unit copy
integrants of the various constructs in wild-type (S150-2B) or gcr1
mutant (FIBY4) yeast were assayed for p-galactosidase activity.
Assays were performed in duplicate on three sperate occasions.

72
-853
-220
n I In
-392
n [~~lrí
220
327 |
-377 I In'
-327
-377
-220
-220
TPI::lacZ
I TPI::lacZ
I TPI::lacZ
TPI::lacZ
B-Galactosidase Activity
S150-2B (GCR1)
Mean SD
HBY4 (gcr1)
Mean SD
161
8
9 1
115
7
18 5
116
9
12 5
14
3
10 3
45
4
21 7
82
4
21 7
36
7
20 5
-220
I TPI::lacZ
TPI::lacZ
27 11
9 1
28 4
8 2

73
identical to the native TPI sequence, but for replacing the RAP1-
binding site with an EcoRI and Pst\ restriction site. When the HB24
derived mutant UAStpi was placed before the TPI::lacZ fusion, (3-
galactosidase activity was not expressed in either strain
background. Thus, UAStpi activity has an absolute requirement for
the RAP1-binding site.
In order to determine if a RAP1-binding site was sufficient for
UAS activity a consensus RAP1 site (AACCCATACATG),
oligonucleotide HB09, was cloned before the fusion. The consensus
RAP1 site had been shown to bind RAP1 protein (Scott et al., 1990).
As a control a mutant RAP1-binding site (AACCCATCAGG)
(oligonucleotide HB11), unable to bind RAP1, was also cloned before
the TPI::lacZ fusion. The consensus RAP1-binding site was not
sufficient to drive expression of the TPI::lacZ fusion, Figure 10, and
resulted in only 27 units of (3-galactosidase activity.
To address the role of the CTTCC and CATCC pentamer motifs
in UAStpi function, double-stranded oligonucleotides were generated
that changed either the CTTCC to CAACC (HB21), the CATCC to
CAACC (HB22), or both (HB24) while retaining a functional RAP1-
binding site. Similar mutations had been shown to reduce expression

74
driven by the UAS elements of the PYK and EN01 (Buchman et al.,
1988). These oligonucleotides were cloned at position -220 before
the TPIr.lacZ fusion, integrated in unit copy, and the strains were
assayed for p-galactosidase activity.
Mutating the CTTCC from -375 to -370 to CAACC reduced
expression of the TPIr.lacZ fusion construct in wild-type yeast to 45
units of activity, a 2.5-fold reduction. 21 units of (3-galactosidase
activity were expressed in the gcr1 mutant background. Changing
the CATCC from -335 to -330 to CAACC reduced activity by one
third to 82 units in the wild-type background and 21 units of
activity in the gcr1 mutant strain. Mutating both pentamers reduced
activity to 36 units in the wild-type strain. The double mutant
containing construct expressed 20 units of (3-galactosidase activity
in the gcr1 mutant strain.
Both pentamers were required for full UAS activity with
mutations in the CTTCC having a larger effect than mutations in the
CATCC. However, it should be noted that the oligonucleotides
employed in these experiments contained only the sequence from
-377 to -327 known to be sufficient for UAStpi activity (Figure 7).

75
Expression driven by UAStpi was 115 units of p-galactosidase
activity while the full length (-853) fusion construct expressed 161
units of activity.
In vitro DNase I Protection Assays Reveal Binding of the REB1 Site
and the RAP1 Site
Upstream activating sequence elements are sites where trans¬
acting factors bind to c/s-acting elements to mediate expression of
the cognate gene. UAStpi was initially used in a series of DNA
bandshift assays by others in the laboratory with crude protein
extracts derived from yeast. Using UAStpi and the RAP1-binding site
mutant derivative, HB19 or 69RBSM from the oligonucleotide cloning
experiments, they were able to demonstrate RAP1 binding to UAStpi
in a RAP1-binding site dependent manner (Scott et al., 1990). Since
UAStpi was specifically bound by protein, DNase I protection assays
as modified by Singh et al. (1986) were performed on the region to
determine the precise areas of interaction.
In order to detect the region protected by the RAP1
protein, the 169 base pair Hinó\\\-Sph\ fragment from the 5' deletion
pES34-1 was used as a target in the DNasel protection assay. In

76
addition to the S150-2B protein extract, a rabbit reticulocyte lysate
(RRL) containing RAP1 generated via in vitro transcription and
translation was used in the binding reactions. The RRL containing
RAP1 was generously provided by M. Cecilia Lopez. Figure 11 shows
the results of the DNase I protection assay. A region of heightened
DNase I sensitivity was observed bounded by positions -288 to -285.
Protection from DNase I cleavage was seen from position -365 to
-343. Similar results were obtained with both the yeast protein
extract and the in vitro generated RAP1 in RRL. No regions of DNase
I protection were seen that affected the CTTCC or CATCC pentamer
motifs when the yeast protein extracts were used. However, Michael
A. Huie has been able to detect an area of DNase I protection
centered about the CTTCC when he used a purified MBP-GCR1 fusion
protein in the binding reactions (Huie et al., 1992).
The initial target used in the DNase I protection assay was the
234 base pair Hind\\\-Sph\ fragment from the 5' deletion construct
pES40-23. Protein extracts were prepared from the yeast S150-2B.
The results of the DNase I protection analysis on this fragment are
shown in Figure 12. Surprisingly, an area of protection was
observed bounded by positions -397 to -386. Inspection of the

Figure 11. In vitro DNase I protection assays demonstrating
protection of the RAP1-binding site. The last four lanes are a DNA
sequencing ladder generated with M13mp18 and the forward primer.
This ladder serves as a molecular weight marker. The (-) reaction
indicates DNase I cleavage was performed on a naked radiolabeled
Hinó\\\-Sph\ fragment from plasmid pES34 and serves as the control
ladder. The second and third lanes, labeled 0.5 and 1.0 respectively,
are the DNase I cleavage reactions generated by 0.5 and 1.0 units of
DNase I performed in the presence of in vitro generated RAP1. The
fourth and fifth lanes are the DNasel cleavage reactions of 0.5 and
1.0 units of DNase I performed in the presence of yeast protein
extract. Regions of DNase I protection and hypersensitivity are
demarcated and the sequence indicated.

-366
CO
CO
N>
CO
O
ro
CO
o>
5' TGATGTTAnAnnmnAr'Ano^^-r
' • • I N I 11 I I II* I III I
» » #»t r • * im* o
^ Ml !• INIIfa Ml t IINNNMHM h
^1
oo
DNasel

Figure 12. In vitro DNase I protection assays demonstrating
protection of the REB1-binding site. The first four lanes are a DNA
sequencing ladder generated with M13mp18 and the forward primer.
This ladder serves as a molecular weight marker. The next two
lanes are the products of DNase I digestion of a radiolabeled Hind\\\-
Sph\ fragment from plasmid pES40-23. (+) extract indicates the
DNase I digestion was carried out in the presence of yeast protein
extract. The (-) extract reaction was performed on naked DNA and
serves as the control Dnase I cleavage ladder. The region of DNase I
protection is demarcated and the sequence indicated.

I
I
'ti n « ti n \
uu in iim
ui «un tuiu
VIVI lUUUUUftll
• W IN
it • tt :t«
limitan >
ui uimm -
mini «■■««
i m
1
II 1
. ' \
*
5’ TACCCGTTCT 3'
CO
CO
00
CD
00
â– 'J
CD
O

81
protected sequence revealed that it was a one base pair mismatch
from a consensus REBI-binding site proposed by Chasman et al.
(1990). REB1 has been implicated in nucleosome phasing in
transcriptional control elements (Brandi and Struhl, 1990).
However, it should be noted that the REB1 -binding site can be
deleted in the 5' deletion series with no effect on expression (pES34-
1, figure 6), when the fusion construct is integrated at the URA3
locus.
DNA Band Shift Assays Demonstrate REB1 Binding to TPI 5' Non-
Codina Region
The DNase I protection assays demonstrated an area of
protection over a region containing a one base pair mismatch from a
consensus REB1-binding site. DNA band shift assays were performed
to determine if REB1 binds the TPI 5' non-coding region. The target
fragment was an end-labeled 139 bp Ava\\-Fok\ fragment (positions
-487 to -348). This fragment contains the sequence protected in the
DNase I assays, as well as the near consensus REB1-binding site.
Others in the laboratory have demonstrated that a fragment
containing sequence from -392 to -348 is incapable of interacting

82
with wild-type yeast extracts is DNA band shift assays ( H. Baker,
personal communication). Extracts were prepared from the wild-
type yeast (S150-2B), or E. coli expressing REB1 from a REB1 insert
in the expression vector pET11A, and E. coli with pET11A and no
insert (kindly provided by B. Morrow and J. Warner). Results of the
DNA band shift assays are shown in Figure 13.
The wild-type yeast extract and the E. coli extract containing
REB1 gave rise to positive band shift assays when incubated with
the fragment from -487 to -348. The E. coll extract without the
REB1 insert failed to band shift. These results indicate that REB1 is
capable of binding a region of the 5' non-coding region of TPI that
contains a near consensus REB1-binding site.
In vivo Methvlation Protection Assays
The DNase I protection and DNA band shift assays were able to
provide information about the ability of the trans-acting factors
RAP1, REB1, and GCR1 to bind to the TPI controlling region under in
vitro conditions. The ability to demonstrate an area of protection
indicated that the factors have the ability to interact with specific
sites, but, did not prove that the interactions occur in vivo. To

Figure 13. DNA band shift assays demonstrating REB1 binding to the
5' noncoding region of TPI. An Ava\\-Fok\ restriction fragment was
isolated from pES90. Radiolabeled DNA fragment was incubated in
binding buffer with E.coli extracts without or with REB1, and wild-
type yeast extract (S150-2B). The first lane serves as a control for
the migration of the fragment alone. Nucleoprotein complexes were
resolved from free DNA by nondenaturing polyacrylamide gel
electrophoresis and were revealed by autoradiography, f, free
unbound probe.

•# FRAG. ALONE
• E. coli/ p11A
E. coli/ p11A-REB1
j S150-2B
oo

85
determine the DNA sequences bound by protein in the cell, dimethyl
sulfate (DMS) methylation protection assays (genomic footprinting)
(Ephrussi et al., 1985) were carried out on various strains of S.
cerevisiae. The strains utilized in the genomic footprints were
S150-2B (wild-type), DFY 642 (wild-type), HBY4 (gcr-1), and JF
1052 (spt13), Table 1. The gcr-1 strain was chosen because the
mutation is known to affect expression of TPI (Clifton and Fraenkel,
1981). SPT13/GAL11 gene function is required for the full
expression of PYK1 (Nishizawa et al., 1990).
Yeast were treated DMS, DNA harvested, prepared and blotted
to a nylon membrane as described in Materials and Methods. Top and
bottom strand radiolabeled probes, as defined in Figure 14, were
prepared to hybridize to the TPI controlling region. Figure 14 shows
the methylation protection pattern of the top strand of the TPI
controlling region from all four strains. The actual footprinting
controls and results are on the left of the figure. On the right a
cartoon represents the sequence of the UAStpi region with the
protected bases denoted. The first four lanes in Figure 14 are
genomic sequencing reactions that provided molecular weight
markers. The genomic "G" sequencing reaction was also the naked

Figure 14. Genomic footprinting of the bottom strand of the TPI 5'
noncoding region. The initial four lanes are genomic sequencing
reactions of the TPI 5' noncoding region. The G sequencing ladder
serves as the control ladder for the genomic footprinting reactions.
Lanes 1-4 are genomic footprinting reactions carried out in wild-
type (S150-2B) yeast. The reactions in lanes 1 and 2 were generated
in wild-type yeast grown in YP media supplemented with 2% glucose
(YPD) and represent 3 and 4 minutes of DMS treatment, respectively.
Lanes 3 and 4 were grown in YP media supplemented with 2%
glycerol and 2% lactate and represent 4 and 5 minutes of DMS
treatment, respectively. Lanes 5 and 6 are the products of 4 and 5
minutes of DMS treatment in a gcr1 (HBY4) mutant strain of yeast.
Lanes 7 and 8 are similar treatment of a second wild-type strain
(DFY642). Lanes 9 and 10 are similar treatment of a spt13 (JF1052)
mutant strain. Guanine residues protected within the GCR1 -binding
site are denoted by an (*). Residues protected within the RAP1-
binding site are denoted by (O). Residues protected within the REB1-
binding site are denoted by (A). The right portion of the figure
depicts the double stranded sequence of the TPI 5' noncoding region.
The bottom strand is the righthand most strand as depicted. The
sequence motifs known to play a role in TPI expression are stippled.
Protected guanine residues are denoted as above.

87
T G wild-type gcr1 w-t 2 spt13
+ + I 1 I 1 I 1 I 1
CCAG 1234G 5 6G 78 G9 10G
CM.
ro
I
o-
!
O 0
mm
< P
Qg
< p
mm
mm
mm
00
o o
< m
1
s
Eh .
O U
Eh <
< m
0 O
Eh '
Eh .
Eh <
O 0
Eh C
m
o.
Eh
¡i
0:0
0 0
C Eh
0 0
Eh <£
Eh «3
Eh <2
Eh <2
0 0
C Eh
0 0
<2 Eh
Eh <
0 0
Eh
S:
0
m
£h
P ,
< P
0 0
< m
<2 p
0 0
-0 0
i i
LD ro
*
*
O
O
O
O
O

88
DNA control for the footprinting reactions. In the wild-type strains
grown in YP medium supplemented with glucose, four major areas of
protection were observed. The most distal region of protection was
four guanine residues. Three of the four bases fall within the near
consensus binding site for REB1, and all four bases are within the
region protected from DNase I digestion in vitro. The next area of
protection was the guanine doublet that base pairs with the cytosine
doublet in the CTTCC pentamer motifs. The CTTCC pentamer was
shown to be protected by GCR1 in vitro, (Huie et al., 1992). The next
region of methylation protection seen in the wild-type background
was five guanine residues within the RAP1-binding site. The final
area of protection seen on the top strand of the wild-type strain
was the guanine doublet that corresponds to the cytosine doublet of
the CATCC pentamer. In vitro DNase I protection by GCR1 of the
CATCC pentamer was suggested but not well resolved in the
footprinting gel (Huie et al., 1992). The pattern of protection seen
in both wild-type strains, S150-2B and DFY 642, are identical. Thus,
the genomic footprinting analysis was able to detect protein
interactions at all known binding sites within the UAStpi region in
wild-type yeast grown in glucose.

89
The effect of carbon source on factor binding-site occupancy
state at UASjpi was also tested. Lanes 3 and 4 of Figure 14 show
that there was no effect of the carbon sources tested on the pattern
of protection observed.
Lanes 5 and 6 of Figure 14 are the footprinting reactions
generated by DMS treatment of a gcr1 mutant strain of yeast grown
in YP medium supplemented with 2% glycerol and 2% lactic acid. The
guanine doublets in the pentamer motifs were not protected. Both
sites were susceptible to methylation and cleavage to the same
degree as the control reactions. However, the areas of protection
corresponding to the REB1 and RAP1-binding sites were still
present. The GCR1 dependent binding of the pentamer motifs was
not required for the binding of the REB1 and RAP1-binding sites in
the TPI controlling region.
Mutation of the spt13/gal11 locus had no effect upon the
protection of UAStpi in vivo. Lanes 11,12 of Figure 14 are the
footprints generated in the spt13 backgrounds. The areas of
protection from methylation are identical to those seen in the wild-
type strain. The REB1, RAP1, and GCRI-binding sites are all still
bound.

90
Figure 15 depicts the genomic footprint of the TPI controlling
region probed for the bottom strand, as defined in the figure, in both
wild-type and gcr1 mutant strains of yeast. Only one area of
protection was seen in both strains. The guanine at position -392
was protected from methylation. Position -392 corresponds to the
last base of the near consensus REB1-binding site.
Site-Directed Mutagenesis of Transcription Factor Binding Sites in
the UAS of TPI
Once the sites bound by transcription factors had been
identified, it was important to mutate those sites in order to assess
their role in TPI gene expression. Site-directed mutagenesis was
utilized to introduce mutations in single and pairwise combinations
of factor-binding sites before the TPI::lacZ gene fusion construct.
Plasmid pES90, used as a target for mutagenesis, contained 853
base pairs of material 5' to the start of the structural gene. This
construct was used to insure that the mutations were in a context
as close as possible to the native loci.
Figure 16 depicts the mutations made in the TPI controlling
region and the results of the p-galactosidase assays performed. The

Figure 15. Genomic footprinting of the top strand of the TPI 5'
noncoding region. The initial lanes are genomic sequencing reactions
as indicated. Guanine sequencing reactions (G) serve as a control
ladder for the genomic footprinting reactions. Wild-type (S150-2B)
or gcr1 mutant (HBY4) yeast were treated with DMS for the time in
minutes indicated above the lanes. The sole area of protection due
to REB1 binding is denoted by (A). The right portion of the figure
depicts the double stranded sequence of the TPI 5' noncoding region.
The top strand is the lefthand most strand as depicted. The
sequence motifs known to play a role in TPI expression are stippled.
Protected guanines are denoted as above.

-404 -324
92
ro in
^ i i
Ü CJ
o—►o u
^ i i
i - -
in ro

Figure 16. Effect of site-directed mutations upon (3-galactosidase activity expressed from a TPI::lacZ
fusion. The sequence of the TPI 5' noncoding region targeted by site-directed mutagenesis is depicted
at the top of the figure. The known binding motifs are cartooned in the next line. Mutations
introducing a known binding motif are denoted by open boxes. Mutations that destroy the activity of a
known motif are stippled, see text for precise details. The results of duplicate (3-galactosidase assays
performed on three separate occasions are listed. Assays were performed on unit copy integrants of
the various constructs in wild-type (S150-2B) and gcr1 mutant (HBY4) strains of yeast. SD indicates
standard deviation.

GGAACCCATCAGGTTGGTGGAAGATTACCCGTTCTAAGACTTTTCAGCTTCCTCTATTGATGTTACACCTGGACACCCCTTTTCTGGCATCCAG
REB1
1 RAP1 | | REbT
REB1
REB1
â– [321
REB1
REB1
-I GCR1 h
HGCRt h
-I GCR1 h
H GCR1 h
H GCR1 h
RAP1 | IGCR1 y-
RAP1 | |GCR1 |—
RAPI | 1 GCR1 y-
RAPl | 1GCR1 y-
Igcri I—
REB1
RAPI
RAP1
RAP1
RAP1
] mi—
■Igcri I—
HD-
«fmm IgcrTV-
REB1 | Igcri I—
1 RAP1 [ 1 REB1 |-
B-Galactosidase Activity
S150-2B
(GCR1)
Mean SD
HBY4
(9cr1)
Mean SD
132 6
9 4
132 12
5 2
56 20
5 3
70 2
6 3
16 1
6 4
21 1
4 2
32 3
6 3
11 3
6 3
41 4
5 2
17 2
4 2
55 3
5 3
10 1
5 2
13 1 4 2
CD
4^

95
initial mutation utilizes oligonucleotide HB32 to introduce an
additional RAP1-binding site into the 5' noncoding region of TPI at
position -420. No effect was seen upon expression in either wild-
type or gcr1 mutant strains. Thus, changes can be made in the region
without affecting expression.
Mutation of the REB1-binding site from 5'-AGATTACCCG-3' to
5'-AGATTGAACG-3' (HB31) resulted in a 2.5-fold reduction in
expression from 132 units of activity to 56 units. This reduction
was somewhat surprising, as the REB1-binding site can be deleted in
the 5’ deletion series without affecting expression (Figure 6).
UAStpi activity was also seen to be independent of the REB1-binding
site in the UAS mapping experiments (Figure 7). Flowever, in the
context of native yeast sequence, mutations in the REB1-binding site
reduce expression.
Changing the CTTCC pentamer to CAACC (HB29) resulted in a 2-
fold reduction in expression to 70 units of p-galactosidase activity.
Mutating the CATCC to CAAAC (HB30) resulted in a 4-fold reduction
to 32 units of activity. Therefore, both elements were required for
full expression. But, unlike the experiments where an isolated
UAStpi was mutated, the mutation of the CATCC element had a

96
greater effect on expression than mutation of the CTTCC element.
However, the mutation introduced into the CATCC element was more
severe in the site-directed mutagenesis than in the oligonucleotide
system, CAAAC versus CAACC.
As in the previous experiments, however, mutation of the RAP1
site to 5'-GATATCTGCAG-3' (HB28) or replacing the RAP1 site with
the REB1 site 5'-AGATTACCCGTTCT-3' (HB53) resulted in abolition
of expression. The TATA box of the TPI promoter was mapped to
positions -175 to -169 by mutations which changed the putative
site from 5'-TATAAG-3' to 5'-GCGAAG-3' (HB33). The mutation
abolished expression of the fusion confirming the location of the
TATA element.
Next a series of double mutants were constructed and assayed,
Figure 15. The double mutants were made from combinations of
single mutations that did not completely destroy expression on their
own. Mutating both pentamer motifs reduced expression to 11 units
of p-galactosidase activity. This double mutation would prevent the
binding of GCR1 to UAStpi, but the in vivo footprinting data would
suggest that REB1 and RAP1 were still bound.

97
Double mutants of the REBI-binding site and either the CTTCC
or CATCC pentamers resulted in reductions that appeared to be
essentially combinations of the individual mutations. Mutating the
REB1 binding site and the CTTCC pentamer resulted in a 4-fold
reduction in expression to 41 units of (3-galactosidase activity. The
4-fold reduction was most likely an additive effect of two
individual mutations, each of which caused a 2-fold reduction.
Mutation of the REB1-binding site and the CATCC pentamer reduced
expression 8-fold to a near background level of 17 units. Since the
mutations in the individual pentamer motifs, GCR1-binding sites,
retained there individual characteristics in combination with REBI-
binding site mutations, REB1 and GCR1 probably do not directly
interact at the protein level.
Mutations that replaced the RAP1-binding site with the REBI-
binding site, and either allowed the native REBI-binding site to
remain or mutating it, abolished expression. Overall, the mutants
affecting the REBI-binding site would suggest that, although the
binding of REB1 was required, REB1 cannot substitute for RAP1 and
probably does not directly interact with GCR1.

98
The final double mutant served to determine the effect of
moving the RAP1-binding site from its position bracketed by the two
GCR1-binding sites to a position upstream of both. The initial single
mutant that replaced a RAP1-like site with a RAP1 -binding site
showed no effects on expression. Next, the second mutation was
carried out to destroy the RAP1-binding site at position -358 to
-346. This double mutant moved the RAP1-binding site upstream by
70 base pairs, and resulted in a 2.4-fold reduction in expression to
55 units of activity. Activity was not completely destroyed,
however. Positioning of the RAP1-binding site was critical for full
expression, but was not absolutely essential. Whether, the residual
expression remaining in the double mutant was due to RAP1
interaction with other sites, such as the CTTCC pentamer, remains
to be determined.
Others in the laboratory have measured the specific activity of
(3-galactosidase expressed per milligram of total protein in these
strains (M. C. Lopez personal communication). The pattern of (3-
galactosidase expression determined by specific activity correlates
very well with the assay results detailed above.

DISCUSSION
In summary, this study has determined the initiation or "I" site
of TPI transcription to be a pair of adenines positioned 29 to 30
bases upstream of the start of translation. The TATA box utilized
for high-level TPI transcription was mapped to position -175 to
-169. Mutational analysis indicated that a single complex upstream
activating sequence, UAStpi, is responsible for expression of TPI.
Three trans-acting factors: REB1, RAP1, and GCR1 have been
identified that bind UAStpi and are required for full expression of
TPI.
The aim of this project was to investigate the mechanism of
high level glycolytic gene expression in Saccharomyces cerevislae.
The GCR1 gene product is known to be required for high-level
glycolytic gene expression (Clifton et al., 1978; Clifton and
Fraenkel, 1981; Baker, 1986). gcr1 lesions result in a 90-95%
reduction in the specific activity of most of the glycolytic enzymes.
In order to identify the cis-acting element through which GCR1
99

100
exerts its effect, and other cis and trans-acting factors which may
be involved in high level glycolytic gene expression, the study
focused on the expression of TPI, a gene whose expression is
severely reduced by mutations in GCR1 (Clifton and Fraenkel, 1981).
TPI encodes, and is the sole source of triose-phosphate isomerase.
TPI expression normally accounts for approximately 2% of the total
soluble protein in S. cerevisiae (The Enzymes, Vol. VI), however,
expression is reduced 17-fold in a gcr1 mutant (Clifton and
Fraenkel, 1981). Understanding the cis and frans-acting elements
required for the high-level expression of TPI should aid in
understanding the common mechanisms required for the high-level
expression of most genes encoding glycolytic enzymes.
In order to understand the mechanism of high-level glycolytic
gene expression it was desirable to choose a gene that is unaffected
by the regulation of metabolic flux through glycolysis. For example,
the ADH2 gene, encoding an isozyme of alcohol dehydrogenase, is
repressed by glucose (Lutstorf and Megnet, 1968; Yu et al., 1989).
The ADH2 gene product favors the production of acetaldehyde from
ethanol, a reaction equilibrium preferred under gluconeogenic
conditions.

101
Many glycolytic enzymes have also been implicated in
interactions other than reactions required for glycolysis. One
isozyme of enolase has been identified as a possible plasminogen
receptor on the cell surface of human cells (Miles et al., 1991).
Pyruvate kinase, aldolase, glyceraldehyde-3-phosphate
dehydrogenase, phosphofructokinase, and lactate dehydrogenase have
all been found in association with actin and tubulin in mammals
(Walsh et al., 1989; Mejean et al., 1989). Phosphoglycerate kinase
has even been implicated in human lagging strand DNA replication
(Jindal and Vishwanatha, 1990). As far as is known, however, triose-
phosphate isomerase, encoded by TPI, is solely involved in
glycolysis.
GCR1 is thought to exert its effect on the expression of TPI,
and other affected genes, at the level of transcription (Clifton and
Fraenkel, 1981; Holland et al., 1987). Early experiments involving in
vitro transcription of isolated RNA (Clifton and Fraenkel, 1981).
Later, Northern blot analysis on TPI (Scott et al., 1990), PGI, PGK,
PYK ( Baker, unpublished result), and GAPDH (Holland et al., 1987)
have all shown reduced steady-state levels of transcript in a gcr1
strain when compared to wild-type. Specific message was still

102
detected in the gcr1 mutant strain for each of the transcripts
tested. Transcription of the HIS4 gene shows a similar pattern of
transcript levels when the gcn4 gene is mutated (Arndt et al., 1987).
The HIS4 gene has been shown to have two promoter elements, one
dependent upon GCN4 and responsible for high-level expression, and
one dependent upon BAS1/BAS2 and responsible for basal level
transcription. Each HIS4 promoter element gives rise to transcripts
that have different sites of initiation.
Primer extension experiments (Figure 4) revealed that unlike
these other systems, the TPI transcript has one major mature 5' end.
The predominant start site is unaffected by mutations in GCR1. In
both wild-type and gcr1 mutant strains the mature 5' end of the TPI
transcript mapped to a pair of adenines 29 and 30 base pairs
upstream of the start of translation, mapping an initiation (i) site of
a single promoter element for TPI. The single promoter requires
GCR1 for full expression of TPI, but is still capable of driving the
low level expression seen in the gcr1 mutant strain.
Mutational analysis of the 5' noncoding region of TPI was
carried out utilizing a TPI::lacZ gene fusion in trans to TPI. Use of
the fusion allowed normal TPI activity required for cellular growth.

103
Expression of the fusion could indirectly be followed by p-
galactosidase activity assay. The Miller assay (Miller, 1972) was
chosen to follow p-galactosidase activity because of the simplicity
of the assay.
5' deletion analysis suggested that all sequences required for
the high level expression of a TPI::lacZ fusion resided with 392
bases of the start of the structural gene. Deletion clone 34-1
(-392) still retained high-level expression, whereas, 70-1 (-377)
expresses only 29-43% relative activity (see Figure 6). These
results suggested that the 5' limit of an upstream activating
sequence (UAS) was bounded by positions -392 and -377 with
respect to the start of translation.
Mapping of a UAS element from TPI was accomplished using a
CYC1::lacZ test construct. The UAS1 and UAS2 elements of CYC1
(Guarente et al., 1984) were replaced with portions of the TPI 5' non¬
coding region in the CYC1 ::lacZ construct. A 66 base pair
oligonucleotide, containing 51 bases of TPI 5' non-coding region
from position -377 to -327, was able to restore expression to the
CYC1 ::lacZ fusion, see Figure 7. The same oligonucleotide was able
to restore expression to a TPI::lacZ fusion deleted to position -220.

104
Therefore, an upstream activating sequence activity for TPI (UAStpi)
resides from position -377 to -327.
Multiple UAS elements driving expression of yeast genes are
not uncommon (Yu et al., 1989; Cohen et al., 1987; Guarente et al.,
1984). The 5' deletion series was only able to map the most
proximal UAS capable of high-level expression of TPI. Therefore,
the 5' deletion series did not rule out additional UAS elements distal
to UAStpi- A deletion series initiating from position -220 was
generated that extended through and 5' to UAStpi. Deletion of UAStpi
abolished expression of the TPI::lacZ fusion, even though several
hundred bases of 5' noncoding sequence remained distal to the
deletion endpoints, see Figure 8. Based on these results, UAStpi is
the sole upstream activating sequence driving high-level expression
of TPI.
The 51 bases from -377 to -327 that confer UAStpi activity
contain an almost perfect consensus RAP1/GRF1/TUF-binding site
from -358 to -346 (Scott et al., 1990). During the course of this
study RAP1 -binding sites were shown to be required for the
expression of PGK (Chambers et al., 1989), EN01 and EN02 (Buchman
et al., 1988; Brindle et al., 1990), PYK (Buchman et al., 1988), and

105
ADH1 (Buchman et al., 1988; Santangelo and Tornow, 1990).
Although, the existence of a RAP1 -binding site and UAS element in
the region of the EN02 5' noncoding region suggested by the Holland
group appears doubtful. J. Anderson demonstrated that the region
from -491 to -443 of the 5' noncoding region of EN02 was unable to
bind RAP1 or act as a UAS before a TPI::lacZ fusion. Chambers et al.
(1989) suggested a putative RAP1-binding site within the 5'
noncoding region of TP I from position -420 to -410. The importance
of this site was brought into question by the 5' deletion studies
when construct 34-1 with an endpoint of -392 was still able to
drive high-level expression, see Figure 6. Subsequent DNA band shift
experiments, performed by others in the laboratory, demonstrated
that -420 to -410 was not a RAP1-binding site (Scott et al., 1990).
A CTTCC sequence motif is also found in UASypi. CTTCC pentamer
motifs have been demonstrated to play a role in the UAS activity of
PGK (Ogden et al., 1986; Stanway et al., 1989), EN01 (Buchman et al.,
1988), and PYK (Buchman et al., 1988; McNeil et al., 1990). UASjpi
contains a CTTCC from -375 to -370, and a closely related CATCC
sequence from -335 to -340.

106
The role of the RAPI-binding site and the pentamer motifs in
the function of UASjpi was tested. Oligonucleotides that encode
UAStpi, or mutant variations thereof, were cloned before a TPI::lacZ
fusion which had been deleted of the native UAS sequence element(s)
and assayed for expression. The 51 base pairs constituting UAStpi
cloned before TPI::lacZ were able to express 120 units of (3-
galactosidase activity on average, Figures 7 and 10. A mutant,
69RBSM of Figure 7, that replaced most of the RAP1-binding site
from -358 to -346 with the sequence from -420 to -410 was unable
to drive expression. Similarly, when the RAP1-binding site was
mutated by the introduction of an EcoRI and a Pst\ restriction site,
UAStpi activity was abolished, Figure 10. A RAP1-binding site alone
was insufficient to restore expression to the TPI::lacZ fusion
deleted of the native UAS element(s). RAP1 binding is required but
not sufficient for high-level expression of the TPI::lacZ fusion.
Similar results have been reported for PGK (Stanway et al., 1989),
and the EN01 and PYK (Buchman et al., 1988). The assertion of
Santangelo and Tornow (1990), that a RAP1-binding site is
sufficient for UAS activity and is the site through which GCR1
exerts its effect, appears untenable.

107
Constructs that mutated the pentamer motifs demonstrated
that they are required for full UAS activity. Mutating the CTTCC
reduced expression 2.5-fold, and mutating the CATCC reduced
expression 33%. Therefore, UAStpi demonstrates an absolute
requirement for the RAP1-binding site from -358 to -346 for
activity, and requires both the CTTCC and the CATCC pentamers for
full activity. The experiments mutating UAStpi in isolation would
indicate that the CTTCC pentamer plays a greater role in expression
than the CATCC pentamer. However, while experiments with
isolated elements can indicate an motif plays a role in expression,
caution should be observed in ordering the overall importance of
individual motifs.
In addition to indicating that certain sequence motifs play a
role in UAStpi activity, the constructs driven by UAStpi were also
shown to respond to a gcr1 mutation with a 10 fold reduction in
expression. Therefore, GCR1 acts either through UAStpi or through
sequences proximal to position -220. The pentamer motifs were a
good candidate for the site of GCR1 action, as they were a common
motif important in the expression of many of the genes affected by
lesions in gcr1 (Stanway et al., 1989; Buchman et al., 1988). The

108
oligonucleotides generated in this study were subsequently used to
show GCR1 binding to UAStpi in a CTTCC dependent manner (Baker,
1991).
DNase I protection assays on the controlling region of TPI were
able to demonstrate binding to the near consensus REB1-binding
site, Figure 12, and the RAP1-binding site, Figurel 1, with crude
yeast protein extracts. The protection seen over the RAP1-binding
site with crude extracts was confirmed to be due to RAP1 by DNase I
protection assays utilizing rabbit reticulocyte lysates (RRL)
containing RAP1 generated via in vitro transcription and translation.
DNA band shift assays demonstrated that REB1 is able to bind to a
fragment that contains the region of DNasel protection, Figure 13.
Figure 17 is a composite summary of the results of this study.
The areas of DNase I protection are denoted by stippled boxes.
Recently, Huie et al. (1992) have demonstrated DNase I protection by
a MBP-GCR1 fusion protein centered about the CTTCC and CATCC
pentamer motifs in UAStpi (Huie et al., 1992).
The DNase I footprinting assays demonstrated that protein
interactions are possible at the REB1 and RAP1-binding sites, and
the pentamer motifs under experimental conditions defined in vitro.

109
-430
37-2 [100%]
'(-420)
34-1 [95%]
|(-392)
t3
®.
70-1 [27%]
(-377)
x â–¡ â–¡â–¡â–¡â–¡â–¡â–¡â–¡*
f AT AT CTAGGÁACCCAT CAGGTT GGTGGAAG ATT ACCCGTT CT AAG ACTTTT CÁGCTTCCT CT AT
AAAA A A A
REB1-DNasel
Footprint
(-377)
GCR1-DNasel Footprint
83-1 [8%]
(-348)
73-2 [4%]
‘(-337)
T GAT GTT ACACCT GG AC ACCCCT
77-2 [2%]
â– (-314)
AA*
A Ai
TCTGGCAT CCAGT
AA
f AAT CTT CAGT GGCAT GT GAG ATT C
RAP1-DNasel Footprint
(-327)
GCR1-DNasel Footprint
TCCG AAATT A ATT AAAGC AAT C ACACAATT CT CTCGGATACC ACCTCGGTT G AAACT GACAGGT G
-234
I
RAP1-DNasel Hypersensitivity
Figure 17. Composite summary of the TPI controlling region. Bold
numbers denote 5' deletion series used to map the TPI UAS element.
Position of deletion endpoints are denoted in parentheses. The
effect of each deletion on expression of TPI::lacZ is indicated in
brackets. The sequence underlined from -377 to -327 was able to
confer UAS activity to a CYC1::lacZ fusion. Core binding sites are
denoted as follows: REB1 (□), GCR1 (•), and RAP1 (_). Mismatches
from consensus binding sites are denoted with an X. Stippling
demarcates areas identified by in vitro DNasel footprinting studies,
as indicated on the figure. â–² denotes G residues protected in cells
treated with DMS on the DNA strand opposite the strand depicted in
the figure. A denotes G residues protected on the DNA strand
depicted.

110
The fact that such interactions are possible, particularly when
adding high-levels of a purified protein such as the RRL- RAP1 and
MBP-GCR1 footprints, was not proof that the interactions were
occurring with the cell. DMS methylation protection assays
(genomic footprinting) were carried out on several strains of
Saccharomyces cerevisiae to analyze the in vivo protein-DNA
interactions within the TPI controlling region. The composite
summary, Figure 17, indicates the position of the protected bases in
the wild-type strains, S150-2B and DFY 642. Portions of all sites
protected in vitro from DNase I digestion are protected from DMS
methylation in vivo. The REB1 and RAP1 -binding sites are bound,
presumably by REB1 and RAP1 respectively. The CTTCC and CATCC
pentamers are also bound.
The genomic footprint of the TPI controlling region in the
isogeneic gcr1 deletion strain, HBY4, confirms that GCR1 is
responsible for the protection seen at the pentamer motifs. In the
gcr1 strain, the REB1 and RAP1-binding sites are still occupied, but,
neither the CTTCC nor the CATCC pentamer is protected, see Figure
14. The GCR1 dependent protection of the pentamer motifs, coupled
with the in vitro DNase I protection of those sites by MBP-GCR1, and

the genetic evidence for the importance of CTTCC and CATCC
pentamers in TPI gene expression, proves that GCR1 acts through the
CTTCC and CATCC pentamer motifs to drive expression of TPI.
GCR1 binding is not required for the binding of either REB1 or
RAP1, as is clearly demonstrated by the continued occupation of
those sites in the gcr1 mutant strain. The presence of CTTCC
pentamer motifs that have been genetically identified to play a role
in the expression of PGK (Stanway et al., 1989), PYK, and EN01
(Buchman et al., 1988) suggests that GCR1 acting through those
pentamers as well. Indeed GCR1 binding has been demonstrated to
the CTTCC elements in the 5' noncoding regions of PGK, EN01, PYK,
and ADH1(y\uie et al., 1992). This suggests that GCR1 acts through a
CTTCC or closely related pentamer motif, such as CATCC, in all
genes affected by mutations in GCR1.
The genomic footprints of the TPI controlling region generated
in a spt13 mutant strain was identical to wild-type. Therefore
neither the spt13 mutation does not affect the binding of REB1,
RAP1, or GCR1. Furthermore, there is no evidence that the SPT13
gene product binds the TPI controlling region to a sufficient degree

112
to provide DMS methylation protection. It is also possible that
SPT13 binds to sites that are adenine and thymine rich.
The implications on the roles of REB1 and RAP1 binding on the
expression of TPI are twofold. First, neither REB1 nor RAP1 binding
alone is sufficient for the high-level expression of TPI as was
demonstrated by the genomic footprinting analysis of the gcr1
mutant strain. Therefore, the role of GCR1 in high-level expression
is not to facilitate the binding of REB1 or RAP1. Secondly, neither
REB1 nor RAP1 binding alone is sufficient for high-level expression
of TPI. The GCR1 dependent protection of the CTTCC and CATCC
pentamer motifs makes the hypothesis of Santangelo and Tornow
(1990), that GCR1 acts through a RAP1-binding site, appear
untenable.
The 5' and internal deletion series have mapped a single
upstream activating sequence for TPI. UAStpi activity was
demonstrated to lie from position -377 to -327. A RAP1-binding
site from -358 to -346 was shown to be absolutely required for
UAStpi activity. Two pentamer sequence motifs were also
demonstrated to be required for full UAS activity. Furthermore, in
vitro and in vivo footprinting analysis have demonstrated RAP1 and

11 3
GCR1 binding to UAStpi. REB1 was also bound to a near consensus
binding site from -401 to -391, a site which could be deleted in the
5' deletion series without affecting expression. Each of the
deletional and mutational analyses used to generate these results
changed the sequence context of the elements being assayed. The 5'
deletions removed yeast sequences distal to the deletion endpoint.
The internal deletions altered spacing with respect to the putative
TATA box. The mutational analysis using oligonucleotides carrying
UAStpi, did both.
Site-directed mutagenesis was used to mutate sequences
involved in the expression of TPI. The TPI::lacZ fusion construct
which was mutated had 853 bases of yeast sequence 5' to the TPI
structural gene. As with all constructs, the site-directed mutants
were integrated into the yeast genome at the URA3 locus. Therefore,
chromatin structures which normally form 5' to TPI at least had the
possibility of forming. The results of the site-directed mutational
analysis provided the most accurate assessment of the importance
of sequence elements in the expression of TPI. Only the actual
introduction of the mutations into the TPI locus itself would provide
more accurate information.

114
Mutation of the RAP1-binding site reduced expression of the
TPIr.lacZ fusion to background levels. Mutation of both GCR1-binding
sites also reduced expression to background levels. The individual
mutation of CATCC to CAAAC resulted in a 4-fold reduction, and
mutation of CTTCC to CAACC resulted in a 2-fold reduction in
expression. The degree of reduction in TPIr.lacZ expression due to
the individual mutation of the pentamer motifs, as established by
site-directed mutagenesis, was the reverse established when
mutating UAStpi alone. Therefore, the mutation of the pentamers in
the integrated fusion constructs was confirmed by PCR
amplification with subsequent sequence analysis. Both integrated
plasmids carried the appropriate mutation. The CATCC pentamer
was mutated to CAACC in the isolated UAStpi, compared to the
CAAAC mutation introduced by site-directed mutagenesis. The
mutational analysis was able to establish the importance of the
pentamer motifs in TPI expression, but was not able to rank their
individual contributions to expression.
The necessity for performing the site-directed mutagenesis
was confirmed by mutations which affected the REB1-binding site.
The REBI-binding site was dispensable in the 5' deletion series.

115
However, when the REB1-binding site was mutated in the context of
853 base pairs of TPI 5' noncoding sequence, a 2-fold reduction in
expression was seen by the Miller assay (Figure 16) or a 5-fold
reduction in specific activity (M.C. Lopez personal communication).
Intimating that REB1 may counter the action of inhibitory sequences
which reside 5' to the REB1 -binding site. REB1 has been implicated
in nucleosome positioning (Bram and Kornberg, 1982; Brandi and
Struhl, 1990; Fedor et al1988; Chasman et al1990). Chasman et
al. (1990) also demonstrated that UASgal functions independently of
the REB1/GRF2-binding site in test constructs. However, a REB1-
binding site is required for full expression of the GAL1 and GAL10
genes (Chasman et al., 1990). If the role of REB1 is to exclude
nucleosomes, thereby opening chromatin structure for transcription,
then the discrepancy between the results of the 5' deletion series
and site directed mutants may be explained. REB1 binding was not
needed for UAStpi function in the 5' deletion series when the
deletion endpoint was located directly downstream of the actively
transcribed URA3 locus. This may be due to a nucleosome free
region created by active transcription and, therefore, REB1 was not
needed. The site-directed mutant constructs, however, had

116
approximately 400 bases of TPI non-coding region between UAStpi
and the URA3 locus. Nucleosomes could form in this region, and
without REB1 binding to prevent formation, UASjpi could be packaged
in a less active configuration.
Double mutations were introduced between the REB1-binding
site and either of the two GCR1-binding sites. If REB1 and GCR1
interacted in the expression of TPI, then mutation of the REB1-
binding site and either GCR1-binding site would have the same
effect. If REB1 and GCR1 do not interact then the results may vary
depending upon which GCR1 -binding site was mutated. Mutation of
the REB1-binding site and the CTTCC pentamer resulted in a 4-fold
reduction in expression. Mutating both the REB1 -binding site and the
CATCC pentamer resulted in an 8-fold reduction in expression. The
results demonstrated that the effects seen in the double mutants
retained much of their individual characteristics, varying depending
upon which GCR1-binding site was mutated. Thereby indicating that
REB1 and GCR1 do not interact at the protein level in the expression
of TPI.
The site-directed mutagenesis indicates an absolute
requirement for RAP1 and GCR1 for the expression of TPI. The REB1-

117
binding site is required in the context of native yeast sequence. The
genomic footprinting results demonstrated that GCR1 binding was
not required for RAP1 or REB1 binding. Thus, neither RAP1 or REB1
binding is sufficient for expression, and the role of GCR1 is not to
facilitate RAP1 or REB1 binding. Recent studies have emphasized
the importance of chromatin structure in transcriptional activation,
reviewed by Felenfeld (1992). Apparently there are two classes of
transcriptional activator. Cloned yeast TFIID is capable of
preventing nucleosome formation when added to DNA only prior to
nucleosome formation (Meisterernst et al., 1990). The second class,
GAL4 for example, is able to bind DNA in chromatin in vitro,
suggesting that GAL4 may invade nucleosomes (Taylor et al., 1991;
Felsenfeld, 1992). The binding of GAL4 to the GAL4-binding sites in
chromatin is at a 10 to 100-fold lower affinity than naked DNA. The
in vivo significance of this binding ability is unknown (Taylor et al.,
1991). GAL4 is known to require the binding of REB1 for the
activation of GAL1 and GAL10 (Chasman et al., 1990), suggesting
that GAL4 may not be able to invade chromatin in vivo. GCN4 is also
known to require auxiliary factors, such as ABF1, RAP1 and REB1
(Buchman and Kornberg, 1990; Arndt et al., 1987; Chasman et al.,

118
1990), to activate transcription. These results suggest that it may
be the role of RAP1 and REB1 to facilitate the binding of GCR1 by
providing the proper DNA structure. The results of the site-directed
mutagenesis and the 5' deletion series indicate that protein-protein
interactions are unlikely to occur between REB1 and GCR1. REB1 is,
therefore, most likely involved only in chromatin structure and not
transcriptional activation as suggested by Chasman et al. (1990) or
transcriptional repression as suggested by Wang et al. (1990).
Regardless of whether or not REB1 provides an actual activation
signal, the REB1-binding site is required for the full expression of
TPI. Therefore, UAStpi should be expanded to positions -401 to
-327, thereby including the REB1-binding site.
The individual roles of RAP1 and GCR1 remain less well
defined. Both proteins are absolutely required for high-level
transcriptional activation of TPI. Binding of RAP1 alone is
insufficient for high-level activation. One possibility is that RAP1
and GCR1 are coactivators of transcription. In this model,
interactions between the two proteins are required for activation.
Alternatively, the role of RAP1 could be the facilitation of GCR1
binding to UAStpi. GCR1 is the specific transcriptional activator in

1 1 9
this model. Protein-protein interactions in this model would merely
serve to stabilize the binding of GCR1.
Direct interactions between RAP1 and GCR1 may not occur in
UAStpi as the bases protected from DMS methylation tend to be on
opposite sides of the DNA helix, suggesting RAP1 and GCR1 are also
on opposite sides of the helix. Therefore RAP1, like REB1, may be
needed to provide proper DNA structure for GCR1 binding. RAP1 is
known to be part of the nuclear scaffold (Cardenas et al., 1990), and
to be able to bend DNA (Vignais and Sentenac, 1989). The relative
binding affinity of GCR1 for linear UAStpi appears to be much less
than the affinity of RAP1 for UAStpi (Baker, 1991). Perhaps the
preferred binding substrate for GCR1 is bent DNA. RAP1 would be
required to provide the substrate for GCR1 binding and the
subsequent activation of TPI. Given the abundance of RAP1 and RAP1-
binding sites in yeast, a role for RAP1 in providing proper DNA
structure seems plausible. RAP1 would provide the proper
structural context for the binding of transcription factors that are
then able to perform there role, whether that be activation or
repression of transcription. Protein-protein interactions between
RAP1 and other factors may still play a role in this model, but, the

120
overall characteristics of the control element would be due to the
other factors, not RAP1.
This study suggests a common mechanism for the high-level
expression of genes encoding glycolytic enzymes. RAP1 and GCR1
have been shown to be required for the high-level expression of all
genes encoding glycolytic enzymes that have been investigated to
date (Scott et al., 1990; Stanway et al., 1989; Chambers et al., 1989;
Buchman et al., 1988; Brindle et al., 1990; Holland et al., 1987;
McNeil et al., 1990; Santangelo and Tornow, 1990). Therefore, it
seems likely that activation by RAP1 and GCR1 is the common theme
used for high-level glycolytic gene expression. Differences in
expression between alleles is likely due to the binding of additional
factors such as ABF1 (Chambers et al., 1990; Brindle et al., 1990),
REB1 (Chasman et al., 1990), and EBF2 (Brindle et al., 1990). One
test of the commonality of RAP1 and GCR1 binding will be to carry
out genomic footprinting experiments on several additional
glycolytic gene promoters to determine if RAP1 and GCR1 are bound
in each case.
The data presented in this study suggests a possible model for
the activation of high-level TPI expression. The model is depicted in

121
Figure 18. REB1 binding is suggested to be required for expression
of the native TPI locus by the approximately 2-fold reduction in
expression seen when the REB1-binding site was mutated by site-
directed mutagenesis. REB1 most likely plays a role in chromatin
structure as previously suggested (Brandi and Struhl, 1990; Fedor et
al., 1988; Chasman et al., 1990) since its binding site can be deleted
without affecting expression when the deletion construct was
integrated immediately downstream of an actively transcribed gene,
Figure 6. Direct protein-protein interactions between REB1 and
GCR1 are unlikely considering the results of the double mutations of
the REBI-binding site and either of the GCRI-binding sites. The
double mutants retained much of the character of the individual
mutations. If protein-protein interactions were required between
REB1 and GCR1 both double mutants would have exhibited the same
effect.
Both RAP1 and GCR1 have been shown to be absolutely required
for the high level expression of TPI. Mutation of the RAP1-binding
site or a double mutation of both GCR1 -binding sites completely

122
TPI
Figure 18
Model of protein interactions at UASypi.

123
ablates expression, Figure16. RAP1 is an abundant protein (Shore
and Nasmyth, 1987) that is known to play a role in both silencing and
activating transcription (Shore and Nasmyth, 1987). RAP1 is also
associated with the nuclear matrix (Cardenas et al., 1990) and
causes bending of DNA (Vignais and Sentenac, 1989). Therefore, it
seems likely that RAP1 also provides a critical DNA structure or
configuration required for expression. The specificity of the
individual transcriptional elements is probably provided by factors
such as GCR1. GCR1 binds UAStpi at the CTTCC and CATCC pentamer
motifs. The binding affinity of GCR1 to naked DNA appears to be
several orders of magnitude less than the binding of RAP1 to naked
DNA (Baker, 1991). However, the GCR1-binding sites are almost
fully occupied in vivo, Figure 14. It is likely that the action of RAP1
binding facilitates the binding of GCR1. Given that the binding of
RAP1 and GCR1 appear to be on opposite sides of the helix this
facilitation of binding may not involve direct protein-protein
interactions. Rather, GCR1 may preferentially bind its recognition
site in a bent confirmation. In this model GCR1 is responsible for
the activation signals that eventually lead to the recruitment of RNA
polymerase and the initiation of transcription.

REFERENCES
(1972). The Enzymes, Vol. VI (New York: Academic Press).
(1989). Current Protocols in Molecular Biology (Ney York: Greene
Publishing Associates and Wiley-lnterscience).
Arndt, K.T., Styles, C., and Fink, G.R. (1987). Multiple global
regulators control HIS4 transcription in yeast. Science 237,
874-880.
Baker, H.V. (1986). Glycolytic gene expression in Saccharomyces
cerevisiae: nucleotide sequence of GCR1, null mutants, and evidence
for expression. Mol. Cell. Biol. 6, 3774-3784.
Baker, FI.V. (1991). GCR1 of Saccharomyces cerevisiae encodes a DNA
binding protein whose binding is abolished by mutations in the
CTTCC sequence motif. Proc. Natl. Acad. Sci. USA 88, 9443-9447.
Berger, S.L., Cress, W.D., Cress, A., Triezenberg, S.J., and Guarente, L.
(1990). Selective inhibition of activated but not basal transcription
by the acidic activation domain of VP16: evidence for
transcriptional adaptors. Cell 61, 1199-1208.
Botstein, D. and Davis, R.W. (1982). Principles and practice of
recombinant DNA research with yeast. In The molecular biology of
the yeast Saccharomyces: Metabolism and gene expression. J.N.
Strathern, E.W. Jones, and J.R. Broach, eds. (Cold Spring Flarbor: Cold
Spring Flarbor Laboratory), pp. 607-638.
124

125
Bram, R. and Kornberg, R.D. (1982). Specific protein binding to far
upstream activating sequences in polymerase II promoters. Proc.
Natl. Acad. Sci. USA 43, 47.
Brandi, C.J. and Struhl, K. (1990). A nucleosome-positioning sequence
is required for GCN4 to activate transcription in the absence of a
TATA element. Mol. Cell. Biol. 10, 4256-4265.
Brindle, P.K., Holland, J.P., Willett, C.E., Innin, M.A., and Holland, M.J.
(1990). Multiple factors bind the upstream activation sites of the
yeast enolase genes EN01 and EN02: ABFI protein, like repressor
activator protein RAP1, binds c/s-acting sequences which modulate
repression or activation of transcription. Mol. Cell. Biol. 10,
4872-4885.
Buchman, A.R., Kimmerly, W.J., Riñe, J., and Kornberg, R.D. (1988).
Two DNA-binding factors recognize specific sequences at silencers,
upstream activating sequences, autonomously replicating sequences,
and telomeres in Saccharomyces cerevisiae. Mol. Cell. Biol. 8,
210-225.
Buchman, A.R. and Kornberg, R.D. (1990). A yeast ARS-binding protein
activates transcription synergistically in combination with other
weak activating factors. Mol. Cell. Biol. 10, 887-897.
Buchman, A.R., Lúe, N.F., and Kornberg, R.D. (1988). Connections
between transcriptional activators, silencers, and telomers as
revealed by functional analysis of a yeast DNA-binding protein. Mol.
Cell. Biol. 8, 5086-5099.
Capieaux, E., Vignais, M.L., Sentenac, A., and Goffeau, A. (1989). The
yeast H+-ATPase gene is controlled by the promoter binding factor
TUF. J. Biol. Chem. 264, 7437-7446.
Cardenas, M.E., LaRoche, T.L., and Gasser, S.M. (1990). The
composition and morphology of yeast nuclear scaffolds. J. Cell Sci.
96, 439-450.

126
Chambers, A., Stanway, C., Kingsman, A.J., and Kingsman, S.M. (1988).
The UAS of the yeast PGK gene is composed of multiple functional
elements. Nucleic Acids Res. 16, 8245-8260.
Chambers, A., Stanway, C., Tsang, J.S.H., Henry, Y., Kingsman, A.J., and
Kingsman, S.M. (1990). ARS binding factor 1 binds adjacent to RAP1
at the UASs of the yeast glycolytic genes PGK and PYK1. Nucleic
Acids Res. 18, 5393-5399.
Chambers, A., Tsang, J.S.H., Stanway, C., Kingsman, A.J., and
Kingsman, S.M. (1989). Transcriptional control of the Saccharomyces
cerevisiae PGK gene by RAP1. Mol. Cell. Biol. 9, 5516-5524.
Chasman, D.I., Lúe, N.F., Buchman, A.R., LaPointe, J.W., Lorch, Y., and
Kornberg, R.D. (1990). A yeast protein that influences the chromatin
structure of UASq and functions as a powerful auxiliary gene
activator. Genes Dev. 4, 503-514.
Clifton, D. and Fraenkel, D.G. (1981). The gcr (glycolysis regulation)
mutation of Saccharomyces cerevisiae. J. Biol. Chem. 256,
13074-13078.
Clifton, D., Weinstock, S.B., and Fraenkel, D.G. (1978). Glycolysis
mutants in Saccharomyces cerevisiae. Genet. 88, 1-11.
Cohen, R., Holland, J.P., Yokoi, T., and Holland, M.J. (1986).
Identification of a regulatory region that mediates
glucose-dependent induction of the Saccharomyces cerevisiae
enolase gene EN02. Mol. Cell. Biol. 6, 2287-2297.
Cohen, R., Yokoi, T., Holland, J.P., Pepper, A.E., and Holland, M.J.
(1987). Transcription of the constitutively expressed yeast enolase
gene EN01 is mediated by positive and negative c/s-acting
regulatory sequences. Mol. Cell. Biol. 7, 2753-2761.

127
Conrad, M.N., Wright, J.H., Wolf, A.J., and Zakian, V.A. (1990). RAP1
protein interacts with yeast telomeres in vivo: overproduction
alters telomere structure and decreases chromosome stability. Cell
63, 739-750.
Davison, B.L., Egly, J.M., Mulvihill, E.R., and Chambón, P. (1983).
Formation of stable preinitiation complexes between eukaryotic
class B transcription factors and promoter sequences. Nature 301,
680-686.
Devlin, C., Tice-Baldwin, K., Shore, D., and Arndt, K.T. (1991). Rap1 is
required for BAS1/BAS2- and GCN4-dependent transcription of the
yeast HIS4 gene. Mol. Cell. Biol. 11, 3642-3651.
Enea, U., Vovis, G.F., and Zinder, N.D. (1975). Genetic studies with
heteroduplex DNA of bacteriophage f1. Asymmetric segregation, base
correction and implications for the mechanism of genetic
recombination. J. Mol. Biol. 96, 495-509.
Ephrussi, A., Church, G.M., Tonegawa, S., and Gilbert, W. (1985). B
lineage-specific interactions of an immunoglobulin enhancer with
cellular factors in vivo. Science 227, 134-140.
Fassler, J.S. and Winston, F. (1988). Isolation and analysis of a novel
class of suppressor of Ty insertion mutations in Saccharomyces
cerevisiae. Genet. 118, 203-212.
Fassler, J.S. and Winston, F. (1989). The Saccharomyces cerevisiae
SPT13/GAL11 gene has both positive and negative regulatory roles
in transcription. Mol. Cell. Biol. 9, 5602-5609.
Fedor, M.J., Lúe, N.F., and Kornberg, R.D. (1988). Statistical
positioning of nucleosomes by specific protein-binding to an
upstream activating sequence in yeast. J. Mol. Biol. 204, 109-127.
Felsenfeld, G. (1992). Chromatin as an essential part of the
transcriptional mechanism. Nature 335, 219-224.

128
Fraenkel, D.G. (1982). Carbohydrate metabolism. In The molecular
biology of the yeast Saccharomyces: Metabolism and gene
expression. J.N. Strathern, E.W. Jones, and J.R. Broach, eds. (Cold
Spring Harbor: Cold Spring Harbor Laboratory), pp. 1-38.
Gibson, T.J. (1984). Studies on the Epstein-Barr virus genome. Ph.d
thesis, Cambridge University, England, thesis
Guarente, L., Lalonde, B., Gifford, P., and Alani, E. (1984). Distinctly
regulated tandem upstream activation sites mediate catabolite
repression of the CYC1 gene of S. cerevisiae. Cell 36, 503-511.
Guarente, L. and Mason, T. (1983). Heme Regulates Transcription of
the CYC1 gene of S. cerevisiae via an upstream activation site. Cell
32, 1279-1286.
Hofmann, J.F., Laroche, T., Brand, A.H., and Gasser, S.M. (1989). RAP-1
factor is necessary for DNA loop formation in vitro at the silent
mating type locus HML. Cell 57, 725-737.
Holland, J.P., Brindle, P.K., and Holland, M.J. (1990). Sequences within
an upstream activation site in the yeast enolase gene EN02 modulate
repression of EN02 expression in strains carrying a null mutation in
the positive regulatory gene GCR1. Mol. Cell. Biol. 10, 4863-4871.
Holland, M.J., Hager, G.L., and Rutter, W.J. (1977). Characterization of
Purified Poly(adenylic acid)-containing messenger ribonucleic acid
from Saccharomyces cerevisiae. Biochem. 16, 8-16.
Holland, M.J. and Holland, J.P. (1978). Isolation and identification of
yeast messenger ribonucleic acids coding for enolase,
glyceraldehyde-3-phosphate dehydrogenase, and phosphoglycerate
kinase. Biochem. 4900, 4907.
Holland, M.J., Yokoi, T., Holland, J.P., Myambo, K., and Innis, M.A.
(1987). The GCR1 gene encodes a positive transcriptional regulator
of the enolase and glyceraldehyde-3-phosphate dehydrogenase gene
families in Saccharomyces cerevisiae. Mol. Cell. Biol. 7, 813-820.

129
Huet, J., Cottrelle, P., Cool, M., Vignais, M.L., Thiele, D., Marck, C.,
Buhler, J.M., Sentenac, A., and Fromageot, P. (1985). A general
upstream binding factor for genes of the yeast translational
apparatus. EMBO J. 4, 3539-3547.
Huie, M.A., Scott, E.W., Lopez, M.C., Hornstra, I.K., Yang, T.P., and
Baker, H.V. (1992). Characterization of the DNA binding activity of
GCR1: In vivo evidence for two GCR1-binding sites in the upstream
activating sequence of TPI of Saccharomyces cerevisiae. Submitted
Ito, H., Fikuda, Y., Murata, K., and Kimura, A. (1983). Transformation
of intact yeast cells treated with alkali cations. J. Bacteriol. 153,
163-168.
Jindal, H.K. and Vishwanatha, J.K. (1990). Functional identity of a
primer recognition protein as phosphglycerate kinase. J. Biol. Chem.
265, 6540-6543.
Johnston, M. (1987). A model fungal gene regulatory mechanism: the
GAL genes of Saccharomyces cerevisiae. Microbiol. Rev. 51, 458-476.
Ju, Q., Morrow, B.E., and Warner, J.R. (1990). REB1, a yeast
DNA-binding protein with many targets, is essential for cell growth
and bears some resemblance to the oncogene myb. Mol. Cell. Biol. 10,
5226-5234.
Kawasaki, G. and Fraenkel, D.G. (1982). Cloning of yeast glycolysis
genes by complemetation. Biochem. Biophys. Res. Commun. 108,
1107-1112.
Kelleher, R.J., Flanagan, P.M., and Kornberg, R.D. (1990). A novel
mediator between activator proteins and the RNA polymerase II
transcription apparatus. Cell 61, 1209-1215.

130
Kellerman, E. and Hollenberg, C.P. (1988). The glucose- and
ethanol-dependent regulation of PDC1 from Saccharomyces
cerevisiae are controlled by two distinct promoter regions. Curr
Genet 14, 337-344.
Kimmerly, W.J., Buchman, A.R., Kornberg, R.D., and Riñe, J. (1988).
Roles of two DNA-binding factors in replication, segregation and
transcriptional repression mediated by a yeast silencer. EMBO J. 7,
2241-2253.
Kurtz, S. and Shore, D. (1991). RAP1 protein activates and silences
transcription of mating-type genes in yeast. Genes Dev. 5, 616-628.
Lagunas, R. (1986). Misconceptions about energy metabolism of
Saccharomyces cerevisiae. Yeast 2, 221-228.
Liu, F. and Green, M.R. (1990). A specific member of the ATF
transcription factor family can mediate transcription activation by
the adenovirus E1a protein. Cell 61, 1217-1224.
Lutstorf, U. and Megnet, R. (1968). Multiple forms of alcohol
dehydrogense in Saccharomyces cerevisiae. Arch. Biochem. Biophys.
126, 933-944.
McNeil, J.B., Dykshoorn, P., Huy, J.N., and Small, S. (1990). The
DNA-binding protein RAP1 is required for efficient transcriptional
activation of the yeast PYK glycolytic gene. Curr Genet 18, 405-412.
Meissner, P.S., Sisk, W.P., and Berman, M.L. (1987). Bacteriophage
lambda cloning system for the construction of directional cDNA
libraries. Proc. Natl. Acad. Sci. USA 84, 4171.
Meisterernst, M., Horikoshi, M., and Roeder, R.G. (1990). Recombinant
yeast TFIID, a general transcription factor, mediates activation by
the gene-specific factor USF in a chromatin assembly assay. Proc.
Natl. Acad. Sci. USA 87, 9153-9157.

131
Mejean, C., Pons, F., Benyamin, Y., and Roustan, C. (1989). Antigenic
probes locate binding sites for the glycolytic enzymes
glyceraldehyde-3-phosphate dehydrogenase, aldolase, and
phosphofructokinase on actin momomer in microfilaments. Biochem.
J. 264, 671-677.
Miles, L.A., Dahlberg, C.M., Plescia, J., Felez, J., Kanefusa, K., and
Plow, E.F. (1991). Role of cell-surface lysines in plasminogen
binding to cells: identification of a-enolase as a candidate
plasminogen receptor. Biochem. 30, 1682-1691.
Miller, J.H. (1972). . In Experiments in molecular genetics. Cold
Spring Harbor: Cold Spring Harbor Laboratory),
Morrow, B.E., Johnson, S.P., and Warner, J.R. (1989). Proteins that
bind the yeast rDNA enhancer. J. Biol. Chem. 264, 9061-9068.
Morrow, B.E., Ju, Q., and Warner, J.R. (1990). Purification and
characterization of the yeast rDNA binding protein REB1. J. Biol.
Chem. 265, 20778-20783.
Nakajima, N., Horikoshi, M., and Roeder, R.G. (1988). Factors involved
in specific transcription by mammalian RNA polymerase II:
purification, genetic specificity, and TATA box-promoter
interactions of TFIID. Mol. Cell. Biol. 8, 4028-4040.
Nishizawa, M., Araki, R., and Teranishi, Y. (1989). Identification of an
upstream activating sequence and an upstream repressive sequence
of the pyruvate kinase gene of the yeast Saccharomyces cerevisiae.
Mol. Cell. Biol. 9, 442-451.
Nishizawa, M., Suzuki, Y., Nogi, Y., Matsumoto, K., and Fukasawa, T.
(1990). Yeast Gall 1 protein mediates the transcriptional activation
signal of two different transacting factors, Gal4 and general
regulatory factor l/repressor/activator site binding protein
I/translation upstream factor. Proc. Natl. Acad. Sci. USA 87,
5373-5377.

Nogi, Y. and Fukasawa, T. (1980). A novel mutation that affects
utilization of galactose in Saccharomyces cerevisiae. Genet. 2,
115-120.
132
Ogden, J.E., Stanway, C., Kim, S., Mellor, J., Kingsman, A.J., and
Kingsman, S.M. (1986). Efficient expression of the Saccharomyces
cerevisiae PGK gene depends on an upstream activating sequence but
does not require TATA sequences. Mol. Cell. Biol. 6, 4335-4343.
Oliphant, A.R., Nussbaum, A.L., and Struhl, K. (1986). Cloning
random-sequence oligodeoxynucleotides. Gene 44, 177-183.
Orr-Weaver, T.L., Szostak, J.W., and Rothstein, R.J. (1983). Genetic
applications of yeast transformation with linear and gapped
plasmids. In Methods in enzymology: volume 101. R. Wu, L. Grossman,
and K. Moldave, eds. (New York: Academic Press), pp. 228-244.
Parker, C.S. and Topol, J. (1984). A Drosophila RNA polymerasell
transcription factor contains a promoter-region specific
DNA-binding activity. Cell 36, 357-369.
Pellman, D., McLaughlin, M.E., and Fink, G.R. (1990). TATA-dependent
and TATA-independent transcription at the HIS4 gene of yeast.
Nature 348, 82-85.
Pugh, B.F. and Tjian, R. (1990). Mechanism of transcriptional
activation by Sp1: evidence for coactivators. Cell 61, 1187-1197.
Sambrook, J., Fritsch, E.F., and Maniatis, T. (1989). Molecular
Cloning- A laboratory Manual (Cold Spring Harbor: Cold Spring Harbor
Laboratory Press).
Sanger, F., Nicklen, S., and Coulson, A.R. (1977). DNA sequencing with
chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74,
5463-5467.

133
Santangelo, G.M. and Tornow, J. (1990). Efficient transcription of the
glycolytic gene ADH1 and three translational component genes
requires the GCR1 product, which can act through TUF/GRF/RAP
binding sites. Mol. Cell. Biol. 10, 859-862.
Sawadogo, M. and Roeder, R.G. (1985). Interaction of a gene specific
transcription factor with the adenovirus major late promoter
upstream of the TATA region. Cell 43, 165-175.
Scott, E.W., Allison, H.E., and Baker, H.V. (1990). Characterization of
TPI gene expression in isogeneic wild-type and grcri-deletion mutant
strains of Saccharomyces cerevisiae. Nucleic Acids Res. 18,
7099-7107.
Sherman, F., Fink, G.R., and Hicks, J.B. (1983). Methods in Yeast
Genetics Laboratory Manual (Cold Spring Harbor: Cold Spring Harbor
Laboratory Press).
Shore, D. and Nasmyth, K. (1987). Purification and cloning of a DNA
binding protein from yeast that binds to both silencer and activator
elements. Cell 51, 721-732.
Shore, D., Stillman, D.J., Brand, A.H., and Nasmyth, K. (1987).
Identification of silencer binding proteins from yeast: possible roles
in SIR control and DNA replication. EMBO J. 6, 461-467.
Singh, H., Sen, R., Baltimore, D., and Sharp, P.A. (1986). A nuclear
factor that binds to a conserved sequence motif in transcriptional
control elements of immunoglobulin genes. Nature 319, 154-158.
Southern, E.M. (1975). Detection of specific sequences among DNA
fragments seperated by gel electrophoresis. J. Mol. Biol. 98,
503-517.
Stanway, C.A., Chambers, A., Kingsman, A.J., and Kingsman, S.M.
(1989). Characterization of the transcriptional potency of
sub-elements of the UAS of the yeast PGK gene in a PGK
mini-promoter. Nucleic Acids Res. 17, 9205-9218.

134
Struhl, K. (1988). The JUN oncoprotein, a vertebrate transcription
factor, activates transcription in yeast. Nature 332, 649-650.
Struhl, K. and Davis, R.W. (1981). Transcription of the his3 gene
region in Saccharomyces cerevisiae. J. Mol. Biol. 152, 535-552.
Sussel, L. and Shore, D. (1991). Separation of transcriptional
activation and silencing functions of the F?AP7-encoded
repressor/activator protein 1: Isolation of viable mutants affecting
both silencing and telomere length. Proc. Natl. Acad. Sci. USA 88,
7749-7753.
Taylor, I.C.A., Workman, J.L., Schuetz, T.J., and Kingston, R.E. (1991).
Facilitated binding of GAL4 and heat shock factor to nucleosomal
templates: differential function of DNA-binding domains. Genes Dev.
5, 1285-1298.
Taylor, J.W., Ott, J., and Eckstein, F. (1985). The rapid generation of
oligonucleotide-directed mutations at a high frequency using
phosphorothioate-modified DNA. Nucleic Acids Res. 13, 8765-8785.
Tsang, J.S.H., Henry, Y.A.L., Chambers, A., Kingsman, A.J., and
Kingsman, S.M. (1990). Phosphorylation influences the binding of the
yeast RAP1 protein to the upstream activating sequence of the PGK
gene. Nucleic Acids Res. 18, 7331-7337.
Uemura, H. and Fraenkel, D.G. (1991). gcr2, a new mutation affecting
glycolytic gene expression in Saccharomyces cerevisiae. Mol. Cell.
Biol. 10, 6389-6396.
Vignais, M.L. and Sentenac, A. (1989). Asymmetric DNA bending
induced by the yeast multifuntional factor TUF. J. Biol. Chem. 264,
8463-8466.
Walsh, J.L., Keith, T.J., and Knull, H.R. (1989). Glycolytic enzyme
interactions with tubulin and microtubules. Biochim. Biophys. Acta
999, 64-70.

135
Wang, H., Nicholson, P.R., and Stillman, D.J. (1990). Identification of
a Saccharomyces cerevisiae DNA-binding protein involved in
transcriptional regulation. Mol. Cell. Biol. 10, 1743-1753.
Warburg, O. (1930). The metabolism of tumors (London: Constable).
Yu, J., Donoviel, M.S., and Young, E.T. (1989). Adjacent upstream
activating sequence elements synergistically regulate transcription
of ADH2 in Saccharomyces cerevisiae. Mol. Cell. Biol. 9, 34-42.
Zagursky, R.J. and Berman, M.L. (1984). Cloning vectors that yield
high levels of single-stranded DNA for rapid DNA sequencing. Gene
27, 183.

136
BIOGRAPHICAL SKETCH
Edward William Scott V was born in 1964 and is not yet dead.
Collegiate education was provided by the University of Chicago,
culminating in a Bachelor of Arts with Honors from the College of
Biological Sciences in 1985. Edward feels himself greatly indebted to
that institution (at least until the year 2005). Graduate education
commenced in 1987 at the University of Florida and has proven to be a
fruitful endeavor. Edward was wedded to the former Miss Rochelle
McFarland in 1991.

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.
Henry V. Baker, Chair
Assistant 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 a dissertation for the degree
of Doctor of Philosophy.
Lonnie O. Ingram
Professor of Microbiology and Cell
Science
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.
ml
Alfred Lewin
Associate 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 a dissertation for the degree
of Doctor of Philosophy.
..... ~—
Richard W. Moyer
Professor and Chaifrmjan 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 a dissertation for the degree
of Doctor of Philosophy.
Thomas C. Rowe
Associate Professor of Pharmacology
and Therapeutics
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.
May 1992
Dean, College of Medicine
Dean, Graduate School

UNIVERSITY OF FLORIDA



69
-430
37-2 [266]
'(-420)
34-1 [251]
'(-392)
70-1 [107]
(-377)
T AT AT CTAGG AACCCAT CAGGTTGGT GG AAG ATT ACCCGTT CT AAG ACTTTT C AGCTT CCT CT AT
(-377)
83-1 [27]
(-348)
73-2 [19]
(-337)
77-2 [16]
(-314)
TGATGTTACACCTGGACACCCCTTTTCTGGCATCCAGTTTTTAATCTTCAGTGGCATGTGAGATTC
x
(-327)
T CCG AAATT AATT AAAGC AAT CACAC AATT CT CTCGG ATACC ACCT CGGTT G A AACT G AC AGGTG'
-234
A
Figure 9. Summary composite of the 5' noncoding region of TPI.
Deletion endpoints of the 5' deletion series are denoted with the
construct number in bold print, positions are in parentheses. (3-
galactosidase activities expressed by wild-type strains harboring
the 5' deletions are indicated in brackets. UAStpi is underlined. The
RAP1-binding site is underscored by dashes with mismatches with
the consensus RAP1-binding site underscored by X's. The CTTCC and
CATCC pentamer motifs are overdotted.


96
greater effect on expression than mutation of the CTTCC element.
However, the mutation introduced into the CATCC element was more
severe in the site-directed mutagenesis than in the oligonucleotide
system, CAAAC versus CAACC.
As in the previous experiments, however, mutation of the RAP1
site to 5'-GATATCTGCAG-3' (HB28) or replacing the RAP1 site with
the REB1 site 5'-AGATTACCCGTTCT-3' (HB53) resulted in abolition
of expression. The TATA box of the TPI promoter was mapped to
positions -175 to -169 by mutations which changed the putative
site from 5'-TATAAG-3' to 5'-GCGAAG-3' (HB33). The mutation
abolished expression of the fusion confirming the location of the
TATA element.
Next a series of double mutants were constructed and assayed,
Figure 15. The double mutants were made from combinations of
single mutations that did not completely destroy expression on their
own. Mutating both pentamer motifs reduced expression to 11 units
of p-galactosidase activity. This double mutation would prevent the
binding of GCR1 to UAStpi, but the in vivo footprinting data would
suggest that REB1 and RAP1 were still bound.


88
DNA control for the footprinting reactions. In the wild-type strains
grown in YP medium supplemented with glucose, four major areas of
protection were observed. The most distal region of protection was
four guanine residues. Three of the four bases fall within the near
consensus binding site for REB1, and all four bases are within the
region protected from DNase I digestion in vitro. The next area of
protection was the guanine doublet that base pairs with the cytosine
doublet in the CTTCC pentamer motifs. The CTTCC pentamer was
shown to be protected by GCR1 in vitro, (Huie et al., 1992). The next
region of methylation protection seen in the wild-type background
was five guanine residues within the RAP1-binding site. The final
area of protection seen on the top strand of the wild-type strain
was the guanine doublet that corresponds to the cytosine doublet of
the CATCC pentamer. In vitro DNase I protection by GCR1 of the
CATCC pentamer was suggested but not well resolved in the
footprinting gel (Huie et al., 1992). The pattern of protection seen
in both wild-type strains, S150-2B and DFY 642, are identical. Thus,
the genomic footprinting analysis was able to detect protein
interactions at all known binding sites within the UAStpi region in
wild-type yeast grown in glucose.


98
The final double mutant served to determine the effect of
moving the RAP1-binding site from its position bracketed by the two
GCR1-binding sites to a position upstream of both. The initial single
mutant that replaced a RAP1-like site with a RAP1 -binding site
showed no effects on expression. Next, the second mutation was
carried out to destroy the RAP1-binding site at position -358 to
-346. This double mutant moved the RAP1-binding site upstream by
70 base pairs, and resulted in a 2.4-fold reduction in expression to
55 units of activity. Activity was not completely destroyed,
however. Positioning of the RAP1-binding site was critical for full
expression, but was not absolutely essential. Whether, the residual
expression remaining in the double mutant was due to RAP1
interaction with other sites, such as the CTTCC pentamer, remains
to be determined.
Others in the laboratory have measured the specific activity of
(3-galactosidase expressed per milligram of total protein in these
strains (M. C. Lopez personal communication). The pattern of (3-
galactosidase expression determined by specific activity correlates
very well with the assay results detailed above.


DNA loop formation (Hofmann et al., 1989) and interactions with the
nuclear scaffolding (Cardenas et al., 1990). The RAP1 protein also
binds yeast telomeric repeat sequences (Conrad et al., 1990), and
telomeres are shortened in a conditionally lethal (ts) rap1 mutant at
nonpermissive temperatures (Conrad et al., 1990). RAP1, ABF1, and
REB1 are abundant proteins (Buchman et al., 1988; Morrow et al.,
1989) and thus they may be more involved in a common mechanism
of transcription rather than specific regulation of any given gene.
The GCR1 gene product appears to be a specific regulatory
protein (or one of many) for the genes encoding glycolytic enzymes.
As with many specific regulatory proteins in yeast, such as GAL4
(Bram and Kornberg, 1982), GCR1 is expressed at low levels is S.
cerevisiae (Baker, 1986). If GCR1(or any other protein) provides for
the specific activation of the genes of glycolysis how does it exert
its effect in the context of RAP1, REB1, and ABF1 binding at UAS
elements?
Uemura and Fraenkel (1991) (Uemura and Fraenkel, 1991) have
recently isolated GCR2 which like GCR1 has a pleiotropic effect upon
the expression of most of the enzymes of glycolysis. The pattern of
affected enzymes in a gcr2 or a gcr1 strain was quite similar with a


Figure 2. Scheme to create internal deletions. The TPI::lacZ fusion construct pES90 contains a unique
Sph\ restriction site at position -220. Plasmid pES90 was linearized with Sphl and digested for
various times with exonuclease Ba/31. Sphl linkers were ligated on and the material was digested
with Hind\\\ and Sph\. The material remaining 5' to position -220 was isolated by gel purification and
subcloned into pES90 digested with H/ndlll and Sph\. Deletion endpoints were determined by
sequencing with primer HB05.


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 ELIHRZ9P2_9IMSWZ INGEST_TIME 2013-01-23T14:41:23Z PACKAGE AA00012908_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


136
BIOGRAPHICAL SKETCH
Edward William Scott V was born in 1964 and is not yet dead.
Collegiate education was provided by the University of Chicago,
culminating in a Bachelor of Arts with Honors from the College of
Biological Sciences in 1985. Edward feels himself greatly indebted to
that institution (at least until the year 2005). Graduate education
commenced in 1987 at the University of Florida and has proven to be a
fruitful endeavor. Edward was wedded to the former Miss Rochelle
McFarland in 1991.


108
oligonucleotides generated in this study were subsequently used to
show GCR1 binding to UAStpi in a CTTCC dependent manner (Baker,
1991).
DNase I protection assays on the controlling region of TPI were
able to demonstrate binding to the near consensus REB1-binding
site, Figure 12, and the RAP1-binding site, Figurel 1, with crude
yeast protein extracts. The protection seen over the RAP1-binding
site with crude extracts was confirmed to be due to RAP1 by DNase I
protection assays utilizing rabbit reticulocyte lysates (RRL)
containing RAP1 generated via in vitro transcription and translation.
DNA band shift assays demonstrated that REB1 is able to bind to a
fragment that contains the region of DNasel protection, Figure 13.
Figure 17 is a composite summary of the results of this study.
The areas of DNase I protection are denoted by stippled boxes.
Recently, Huie et al. (1992) have demonstrated DNase I protection by
a MBP-GCR1 fusion protein centered about the CTTCC and CATCC
pentamer motifs in UAStpi (Huie et al., 1992).
The DNase I footprinting assays demonstrated that protein
interactions are possible at the REB1 and RAP1-binding sites, and
the pentamer motifs under experimental conditions defined in vitro.


11 3
GCR1 binding to UAStpi. REB1 was also bound to a near consensus
binding site from -401 to -391, a site which could be deleted in the
5' deletion series without affecting expression. Each of the
deletional and mutational analyses used to generate these results
changed the sequence context of the elements being assayed. The 5'
deletions removed yeast sequences distal to the deletion endpoint.
The internal deletions altered spacing with respect to the putative
TATA box. The mutational analysis using oligonucleotides carrying
UAStpi, did both.
Site-directed mutagenesis was used to mutate sequences
involved in the expression of TPI. The TPI::lacZ fusion construct
which was mutated had 853 bases of yeast sequence 5' to the TPI
structural gene. As with all constructs, the site-directed mutants
were integrated into the yeast genome at the URA3 locus. Therefore,
chromatin structures which normally form 5' to TPI at least had the
possibility of forming. The results of the site-directed mutational
analysis provided the most accurate assessment of the importance
of sequence elements in the expression of TPI. Only the actual
introduction of the mutations into the TPI locus itself would provide
more accurate information.


ADH1 also has a UAS that binds RAP1 (Buchman et al., 1988).
Transcription of ADH1 is reduced in a gcr1 strain (Santangelo and
Tornow, 1990). Santangelo and Tornow have reported that the RAP1
site of the ADH1 UAS was able to confer responsiveness to GCR1
when it replaced the UAS of LAC4 (Santangelo and Tornow, 1990).
Therefore, Santangelo and Tornow suggest that GCR1 acts through
the RAP1 binding site. Contrasting this observation is the PGK
controlling region where the activator core region to which RAP1
binds is insufficient to activate transcription (Stanway et al.,
1989). In addition, both EN01 and PYK UAS's require the CTTCC
pentamer as well as the RAP1-binding site for full UAS activity
(Buchman et al., 1988). This contradiction over RAP1 binding alone
being sufficient for activation of transcription has yet to be
resolved.
The exact mechanism by which RAP1 is able to activate
transcription of TPI, PGK, EN01, EN02, PYK, PDC1, and ADH1 has yet
to be determined. Expression of each of these genes also depends on
GCR1. The nature of the interactions of RAP1, ABF1, REB1 and GCR1
are also unknown. Some indications of the role of RAP1 binding may
be provided by studies that link RAP1 with DNA structure, both in


44
The TPI specific probe was prepared as follows. Single-
stranded M13 phage containing the sense (mES2-2) or anti-sense
(mESTPImp18B) strand of TPI, spanning the 5' non-coding region and
the beginning of the structural gene, was prepared by standard
techniques (Current Protocols in Molecular Biology). 6pl of phage
(0.25pg/|il) was incubated with 5pl of primer(0.5pM/pl), and 2.5pil of
10X klenow buffer (0.5M Tris pH 8.0, 2M NaCI), at 50OC for 30
minutes. The sense probe was prepared by annealing primer HB54 to
mESTPImp18B. The antisense probe was made by annealing primer
HB60 to mES2-2. The following reagents were then added in order:
5pl of 0.1 M DTT, 5pl of 50mM MgCl2, 2pl of 3.0mM dNTP-dATP mix,
10 units of DNA polymerase Marge subunit, and 10pl of 32p. dATP
at 3,000 Ci/mM. The antisense reaction also included 2pl of Avail to
cleave the probe to the appropriate length. The reaction was
incubated at 370C for 45 minutes. 120(il of formamide sequencing
dye was then added. The probe was denatured at 950C for 10
minutes and run into a 6% (40:1.3) polyacrylamide, 8M urea gel. The
wet gel was exposed to Polaroid type 57 film for 15 minutes. The
film was developed and used as a guide to excise the probe from the


Hind III
-853
|lUAS? I I LIAS I
Hind III
-853
V
Isolate Hiri III / Sph\ Fragment
Subclone to Sph\ at -220 (wt.)
Sph I
-220
TATA
TPI::lacZ
Sph I
-220
TATA
1 TPI::lacZ
1) Open at Sph I
2) Delete with Bal 31
3) Add Sph I Linkers
rv>
CD


REFERENCES
(1972). The Enzymes, Vol. VI (New York: Academic Press).
(1989). Current Protocols in Molecular Biology (Ney York: Greene
Publishing Associates and Wiley-lnterscience).
Arndt, K.T., Styles, C., and Fink, G.R. (1987). Multiple global
regulators control HIS4 transcription in yeast. Science 237,
874-880.
Baker, H.V. (1986). Glycolytic gene expression in Saccharomyces
cerevisiae: nucleotide sequence of GCR1, null mutants, and evidence
for expression. Mol. Cell. Biol. 6, 3774-3784.
Baker, FI.V. (1991). GCR1 of Saccharomyces cerevisiae encodes a DNA
binding protein whose binding is abolished by mutations in the
CTTCC sequence motif. Proc. Natl. Acad. Sci. USA 88, 9443-9447.
Berger, S.L., Cress, W.D., Cress, A., Triezenberg, S.J., and Guarente, L.
(1990). Selective inhibition of activated but not basal transcription
by the acidic activation domain of VP16: evidence for
transcriptional adaptors. Cell 61, 1199-1208.
Botstein, D. and Davis, R.W. (1982). Principles and practice of
recombinant DNA research with yeast. In The molecular biology of
the yeast Saccharomyces: Metabolism and gene expression. J.N.
Strathern, E.W. Jones, and J.R. Broach, eds. (Cold Spring Flarbor: Cold
Spring Flarbor Laboratory), pp. 607-638.
124


76
addition to the S150-2B protein extract, a rabbit reticulocyte lysate
(RRL) containing RAP1 generated via in vitro transcription and
translation was used in the binding reactions. The RRL containing
RAP1 was generously provided by M. Cecilia Lopez. Figure 11 shows
the results of the DNase I protection assay. A region of heightened
DNase I sensitivity was observed bounded by positions -288 to -285.
Protection from DNase I cleavage was seen from position -365 to
-343. Similar results were obtained with both the yeast protein
extract and the in vitro generated RAP1 in RRL. No regions of DNase
I protection were seen that affected the CTTCC or CATCC pentamer
motifs when the yeast protein extracts were used. However, Michael
A. Huie has been able to detect an area of DNase I protection
centered about the CTTCC when he used a purified MBP-GCR1 fusion
protein in the binding reactions (Huie et al., 1992).
The initial target used in the DNase I protection assay was the
234 base pair Hind\\\-Sph\ fragment from the 5' deletion construct
pES40-23. Protein extracts were prepared from the yeast S150-2B.
The results of the DNase I protection analysis on this fragment are
shown in Figure 12. Surprisingly, an area of protection was
observed bounded by positions -397 to -386. Inspection of the


35
determined by Southern blot hybridization analysis. Genomic DNA
was isolated from individual transformants. The DNA was then
digested with Sad, run on a 0.8% agarose 1X TBE gel, and capillary
blotted to a Gene Screen nylon membrane according to the
manufactures instructions (DuPont). The filter was hybridized with
a Ylp56 probe generated by random-primer extension. Hybridizations
were carried out at 420C for 16 hours in 50% formamide, 0.2% BSA,
0.2% polyvinyl-pyrrolidone (M.W. 40,000), 0.2% ficoll (M.W. 400,000),
50mM Tris pH 7.5, 0.1% Na-pyrophosphate, 1% SDS. Hybridized
membranes were washed according to manufactures instructions.
Visualization was via autoradiography.
p-Galactosidase Assays
Strains to be assayed were grown from a single colony to an
A6oo between 0.5 and 1.5 at 30OC. p-galactosidase assays were
performed essentially by the method of Miller (1972). The units
reported correspond to AA42o/minute/A6oo of the initial culture.


RESULTS
The Mature 5' Ends of Steadv-State TPI Transcripts are Unaffected
by a gen Mutation
Previous studies suggested that gcr1 lesions bring about a
reduction in the levels of mRNAs specifying glycolytic enzymes
(Clifton and Fraenkel, 1981; Holland et al., 1987). An initial
objective was to characterize the TPI transcript in both wild-type
and gcr1 mutant strains of Saccharomyces cerevisiae. RNA gel
transfer hybridization experiments in both strains demonstrated a
reduced steady-state level of the TPI transcript in a gcr1 mutant
strain (Scott et al., 1990). The reduction in steady-state levels of
the transcripts suggests that the GCR1 gene product may play a role
in the transcriptional regulation of the genes encoding glycolytic
enzymes.
To further characterize the TPI transcript, it was of interest
to determined if the mature 5' end(s) of the transcripts were
affected by deletion of GCR1. Differences in the mature 5' end of the
46


114
Mutation of the RAP1-binding site reduced expression of the
TPIr.lacZ fusion to background levels. Mutation of both GCR1-binding
sites also reduced expression to background levels. The individual
mutation of CATCC to CAAAC resulted in a 4-fold reduction, and
mutation of CTTCC to CAACC resulted in a 2-fold reduction in
expression. The degree of reduction in TPIr.lacZ expression due to
the individual mutation of the pentamer motifs, as established by
site-directed mutagenesis, was the reverse established when
mutating UAStpi alone. Therefore, the mutation of the pentamers in
the integrated fusion constructs was confirmed by PCR
amplification with subsequent sequence analysis. Both integrated
plasmids carried the appropriate mutation. The CATCC pentamer
was mutated to CAACC in the isolated UAStpi, compared to the
CAAAC mutation introduced by site-directed mutagenesis. The
mutational analysis was able to establish the importance of the
pentamer motifs in TPI expression, but was not able to rank their
individual contributions to expression.
The necessity for performing the site-directed mutagenesis
was confirmed by mutations which affected the REB1-binding site.
The REBI-binding site was dispensable in the 5' deletion series.


Table 1. STRAINS
Strain
Genotvpe
Source. Reference
E.coli
KK2186
supE, sbcB15, hsdR4,
rpsL,thi, A(lac-proAB)
F [traD36, proAB+, laclq,
lacZAml 5]
(Zagursky and
Berman, 1984)
MC1061
hsdR, mcrB, araD139,
A(araABC-leu)7679
AlacX74, galU, galK,
rpsL, thi
(Meissner et al.,
1987)
TG1
supE, hsdA5,thi,
A(lac-proAB)
F [traD36, proAB+,
laclq, /acZAM15]
(Gibson, 1984)
S. cerevisiae
S150-2B
MAT a, leu 2-3,112,
his3A,trp1 -289, ura3-52
(D. Shore)
HBY4
MAT a, gcr1A::HIS3,
Ieu2-3,112, his3A,
trp 1-289, ura3-52
(Scott et al.,
1990)
JF1052
MAT a, Ieu2, ura3-52,
his4-917, Iys2-1288,
spt 13-20 {LEU2)
(J. Fassler)
DFY642
MATa, leu2-3,112,
ura3-52
(D. Fraenkel)


I
I
"i n i 11 q \
m in i i m
iu nn u tuiu
i i i \ tvvvui
1 WWW
in uimuim"
nim nnxin
^^nn
o
t
t
W
it
5' TACCCGTTCT 3'
CO
co
CO
co
CD
'J
CO
o


31
Materials utilized to introduce the mutations were supplied by
a kit available from Amersham. The kit is based on the method of
Eckstein and co-workers (Taylor et al., 1985). The method provides
strand-specific selection based on the inability of the restriction
enzyme Nci\ to cleave DNA containing thionucleotides. The mutant
oligonucleotide was annealed to the target, extended with the large
fragment of DNA polymerase I in the presence of the thionucleotide
(dCTPaS).
The resulting material was ligated with T4 DNA ligase to form
complete and unnicked double-stranded plasmid. Excess single-
standed target template was removed via passing through a filter
which binds single, but not double-stranded DNA. Then the parental
strand was specifically nicked with Nci\, which did not cut the
thionucleotide containing mutant daughter strand. The nicked
material was then partially digested with exonuclease III. The
intact daughter strand was then used as a template with DNA
polymerase I and circularized with T4 DNA ligase.
The new double-stranded construct contained the desired
mutation on both strands. The material was used to transform E.
coli strain TG-1, and lysates were prepared from 10 individual


90-95% reduction seen in the expression of most glycolytic
enzymes. EN01 transcript levels were compared in GCR2 and gcr2
backgrounds. The defect in enzymatic activity seen in a gcr2 mutant
is mirrored by a reduction in EN01 mRNA. Interestingly, the gcr2
mutant exhibits only a partial growth defect on glucose where a
gcr1 mutant has a severe growth defect on glucose. GCR2 has been
cloned by complementation but its sequence has not been reported.
How GCR2 will fit into the overall regulation of the expression of
the genes encoding glycolytic enzymes and how it relates to GCR1
remains to be seen.


104
Therefore, an upstream activating sequence activity for TPI (UAStpi)
resides from position -377 to -327.
Multiple UAS elements driving expression of yeast genes are
not uncommon (Yu et al., 1989; Cohen et al., 1987; Guarente et al.,
1984). The 5' deletion series was only able to map the most
proximal UAS capable of high-level expression of TPI. Therefore,
the 5' deletion series did not rule out additional UAS elements distal
to UAStpi- A deletion series initiating from position -220 was
generated that extended through and 5' to UAStpi. Deletion of UAStpi
abolished expression of the TPI::lacZ fusion, even though several
hundred bases of 5' noncoding sequence remained distal to the
deletion endpoints, see Figure 8. Based on these results, UAStpi is
the sole upstream activating sequence driving high-level expression
of TPI.
The 51 bases from -377 to -327 that confer UAStpi activity
contain an almost perfect consensus RAP1/GRF1/TUF-binding site
from -358 to -346 (Scott et al., 1990). During the course of this
study RAP1 -binding sites were shown to be required for the
expression of PGK (Chambers et al., 1989), EN01 and EN02 (Buchman
et al., 1988; Brindle et al., 1990), PYK (Buchman et al., 1988), and


53
linker was designed such that upon subsequent digestion with Sal I
and religation of the vector each deletion endpoint would be marked
with a Hin6\\\ site. Precise deletion endpoints were determined for
65 individual constructs by using the dideoxy sequencing method of
Sanger (1977) as modified by U.S. Biochemicals. Once the 5' deletion
series was obtained it was necessary to place the constructs back
into yeast to determine the effects of the deletions upon expression
of the fusion. Of the 65 plasmids sequenced 13 were chosen because
they provided a well spaced, nested set of 5' deletions for study.
The constructs were subcloned into the yeast integrative
plasmid 56 (Yip56) which contains a URA3 selectable marker for
yeast. Yip56 also contains an origin of replication and an ampicillin
resistance determinant for propagation and selection in E. coli. An
integrative yeast shuttle vector was chosen to avoid effects on
fusion expression due to plasmid copy number discrepancies brought
on by the high segregation rate of yeast plasmids (Botstein and
Davis, 1982). The isogeneic uracil auxotrophs S150-2B and HBY4
were transformed to uracil prototropy with the deletion constructs.
Integration of the fusions was directed to the URA3 locus by
transforming with plasmid DNA linearized with Stu\ which cuts


48
initiation codon. Extension products were resolved on a denaturing
polyacrylamide gel next to a DNA sequence of the TPI 5'
nontranslated region which was generated with the same primer
used in the extension reactions. Figure 4 shows the results of the
primer extension experiment. In both the wild-type and gcr1 mutant
strain the predominant mature 5' ends of the TPI transcript were
identical. The ends corresponded to a pair of adenines at positions
-29 to -30 with respect to the initiation codon. Therefore, it
appears likely that TPI transcription is controlled by a single
promoter element, and that the residual expression of TPI observed
in gcr1 mutant strains is the result of transcription originating at
the native start site.
Identification of the 5' Boundary of the TPI Controlling Region
In order to map the controlling element(s) of TPI a TPI::lacZ
gene fusion was utilized. This gene fusion produced a protein that
was a hybrid between triose-phosphate isomerase and (3-
galactosidase, which retained (3-galactosidase activity. Use of the
fusion in trans to TPI allowed normal expression of TPI which is
required for cellular growth while allowing the manipulation of the


T G wild-type gcr1 w-t 2 spt13
CCAG1234G5 6G 78 G9 10G
CM
ro
l
o
I
Q 0
up:
< P
Qg:
< P:
00
0 0
P:M
a o
0 0
< P
B
0 u
P C
< P
0 U
gg
P <
0 0
Eh c
0
0
Eh
PI .
00
0 0
0 0
P p 3
p <2
p <
0 o
C P
0 0
< P
p c
u o
p
Si
0
0
Sl
P:
P ,
< m
0 0
< P
< P
0 0
0 0
l l
LD m
87
*
*
o
o
o
o
o
*
*


70
flanking the UAS element. The RAP1-binding site from -358 to -346
has been shown to be required for UAS activity in wild-type yeast
(Figure 7). Recently, other laboratories have demonstrated the
requirement of a CTTCC pentamer motif for the full expression of
PGK (Stanway et al., 1989), PYK, and EN01 (Buchman et al., 1988).
Examination of UAStpi revealed a CTTCC pentamer located from
position -375 to -370, and a closely related CATCC pentamer from
-335 to -330 (Figure 9).
Mutant variations of UAStpi were generated and cloned at
position -220 before a TPI::lacZ fusion. The constructs were
integrated in unit copy into wild-type (S150-2B) and gcr1 mutant
(HBY4) strains of yeast. UAStpi, -377 to -327 (HB14 and HB15, Table
2), expressed 115 units of (3-galactosidase activity in the wild-type
background and 18 units of activity in the gcr1 mutant background,
Figure 10. It was interesting to note that expression of the UAStpi
construct was reduced 10-fold in the gcr1 mutant background. Thus,
GCR1 must act through UAStpi sequence, or through a sequence
downstream of position -220. The role of the RAP1-binding site in
UAStpi activity was addressed by an oligonucleotide, HB24 (Table 2),


97
Double mutants of the REBI-binding site and either the CTTCC
or CATCC pentamers resulted in reductions that appeared to be
essentially combinations of the individual mutations. Mutating the
REB1 binding site and the CTTCC pentamer resulted in a 4-fold
reduction in expression to 41 units of (3-galactosidase activity. The
4-fold reduction was most likely an additive effect of two
individual mutations, each of which caused a 2-fold reduction.
Mutation of the REB1-binding site and the CATCC pentamer reduced
expression 8-fold to a near background level of 17 units. Since the
mutations in the individual pentamer motifs, GCR1-binding sites,
retained there individual characteristics in combination with REBI-
binding site mutations, REB1 and GCR1 probably do not directly
interact at the protein level.
Mutations that replaced the RAP1-binding site with the REBI-
binding site, and either allowed the native REBI-binding site to
remain or mutating it, abolished expression. Overall, the mutants
affecting the REBI-binding site would suggest that, although the
binding of REB1 was required, REB1 cannot substitute for RAP1 and
probably does not directly interact with GCR1.


43
pellets were washed twice with 70% ethanol, and dried. Pellets
were dissolved in 10Opil of 1M piperidine and incubated at 95C for
30 minutes. 11 pi of 3M NaAc pH 6.0, and 230pl ethanol was added.
Samples were centrifuged 30 minutes, washed twice with 70%
ethanol and dried in vacuo overnight. Pellets were dissolved in
10Opil TE pH 8.0 and the concentration of the material was
determined spectrophotometrically.
5pg of all sample and control DNA's were lyophilized,
resuspended in 4pl formamide sequencing dye, and denatured at 950C.
A genomic sequencing ladder followed by the in vivo methylated
samples bracketed with G-reaction control ladders were
electrophoresed in a 60cm, 7% (40:1.3) polyacrylamide, 8M urea gel
in 0.5X TBE running buffer at 3500 volts and 35 milliamps. After
running, the gel is lifted using Whatman 541-sfc paper, and
electroblotted to a Hybond N+ membrane (Amersham). The DNA was
UV cross-linked to the membrane with a FisherBiotech 312 nm
variable intensity transilluminator at full power for 5 minutes. The
membrane was prehybridized in 20ml of hybridization buffer [1.0%
bovine serum albumen, 7.0% SDS, 1.0mM EDTA, (0.5M Na) HPO4 pH 7.2]
at 630C in a roller incubator.


23
5' deletion scheme The scheme to introduce the Ba/31-induced
deletions in the 5' noncoding region of TPI is shown in Figure 1.
Plasmid pHB110 contains 3.5 kb of additional yeast DNA 5' to the
start of the TPI structural gene fused in frame to lacZ. The fusion
construct was linearized with Tth1111 at a unique site 853 bp
upstream of the structural gene. The linearized material was then
treated for various times with exonuclease Bal31. Following Ba/31
treatment, Klenow fragment was used to fill in the ends. HindIII-
Sal\-Hlnd\\\ linkers (CAAGCTTGTCGACAAGCTTG, HB03) were added.
The material was then digested with Sal I which cuts within the
linker and within the polycloning site of pHB110. This digest
removes all yeast derived DNA 5' to the deletion endpoint. The
desired fragments, containing the TPT.:lacZ fusion and the remaining
5' noncoding sequence, were gel purified, ligated, and used to
transform E. coli MC1061. Plasmid DNA was prepared from
individual transformants, and the precise deletion endpoints were
determined by double-stranded DNA sequencing using primer HB01.
Once the desired constructs were identified, they were
subcloned on Hind\\\ fragments into the Hind\\\ site of Ylp56. The
orientation of the fusion with respect to the URA3 selectable


117
binding site is required in the context of native yeast sequence. The
genomic footprinting results demonstrated that GCR1 binding was
not required for RAP1 or REB1 binding. Thus, neither RAP1 or REB1
binding is sufficient for expression, and the role of GCR1 is not to
facilitate RAP1 or REB1 binding. Recent studies have emphasized
the importance of chromatin structure in transcriptional activation,
reviewed by Felenfeld (1992). Apparently there are two classes of
transcriptional activator. Cloned yeast TFIID is capable of
preventing nucleosome formation when added to DNA only prior to
nucleosome formation (Meisterernst et al., 1990). The second class,
GAL4 for example, is able to bind DNA in chromatin in vitro,
suggesting that GAL4 may invade nucleosomes (Taylor et al., 1991;
Felsenfeld, 1992). The binding of GAL4 to the GAL4-binding sites in
chromatin is at a 10 to 100-fold lower affinity than naked DNA. The
in vivo significance of this binding ability is unknown (Taylor et al.,
1991). GAL4 is known to require the binding of REB1 for the
activation of GAL1 and GAL10 (Chasman et al., 1990), suggesting
that GAL4 may not be able to invade chromatin in vivo. GCN4 is also
known to require auxiliary factors, such as ABF1, RAP1 and REB1
(Buchman and Kornberg, 1990; Arndt et al., 1987; Chasman et al.,


-404 -324
92
ro in
^ i i
U
oo o
^ i i
i -
in ro


32
plaques. Introduction of the desired mutation was confirmed by
sequencing with primer HB42. RF II DNA was prepared from desired
mutant phage and the Hin\\\-Sph\ fragment was subcloned into
pES90 to form the mutant constructs. All pES90 derived mutant
constructs were also confirmed by sequencing with primer HB42
prior to integration and assay.
T ransformation
The method of Enea et al. (1975) was used to transform E. coli.
Saccharomyces cerevisiae was transformed by the method of Ito et
al. (1983), selecting for uracil prototropy. Integration of the
TPI::lacZ fusion constructs at the UR A3 locus was achieved by
linearizing the plasmids with StuI which cuts at a unique site
within URA3. The scheme for integration is shown in Figure 3.
Screen for Unit Coov Integrants of TPI::lacZ fusions at URA3
Integrative transformation of yeast cells can lead to tandem
insertional events (Orr-Weaver et al., 1983), Figure 3. To screen for
unit copy integrants, the genomic structure of the URA3 locus was


MATERIALS AND METHODS
Strains
The strains of Saccharomyces cerevisiae and Escherichia coli
used in this study are shown in Table 1.
Media and Growth Conditions
Yeast cultures were grown in YP medium (Sherman et al.,
1983) supplemented with 2% glucose or 2% glycerol and 2% lactate.
Selection was carried out in YNB supplemented with the appropriate
carbon source and 0.0025 % histidine, 0.0025% leucine, 0 0025%
tryptophan,and 0.1% case amino acids. All yeast cultures were grown
at 30OC. E. coli strains were grown in LB broth or minimal medium
63 supplemented with thiamine hydrochloride (1pg/ml), and amino
acids (25pg/ml) (Miller, 1972). Ampicillin was added to 100 pg/ml
for selection. All E. coli cultures were grown at 370C.


Figure 13. DNA band shift assays demonstrating REB1 binding to the
5' noncoding region of TPI. An Ava\\-Fok\ restriction fragment was
isolated from pES90. Radiolabeled DNA fragment was incubated in
binding buffer with E.coli extracts without or with REB1, and wild-
type yeast extract (S150-2B). The first lane serves as a control for
the migration of the fragment alone. Nucleoprotein complexes were
resolved from free DNA by nondenaturing polyacrylamide gel
electrophoresis and were revealed by autoradiography, f, free
unbound probe.


63
of all sequences 5' to the SphI site at -220. However, a 58 base pair
oligonucleotide (HB07 and HB08, table 2) with TPI noncoding
sequence from -377 to -335 was unable to restore expression the
TPI::lacZ fusion, Figure 7. An oligonucleotide (69RBSM or HB19,
table 2) identical to the native 51 b.p. region from -377 to -327
except for mutations in the RAP1-binding site was prepared. The
mutant oligonucleotide was placed before the same deleted TPI::lacZ
fusion as the wild-type oligonucleotide to assess the role of the
RAPI-binding site in UAStpi function. The mutated oligonucleotide
was unable to drive expression of the fusion demonstrating that the
RAP1-binding site from -358 to -346 is essential for UAS activity.
Others in the laboratory performed DNA band shift assays with these
oligonucleotides which indicated that RAP1 binds with much reduced
affinity if at all to the mutant oligonucleotide in vitro (Scott et al.,
1990).
Internal Deletions Indicate Single UAS Element Responsible for TPI
Transcription
The 5 deletion series facilitated the mapping of an upstream
activating sequence, UAStpi, that was sufficient to drive expression


Figure 1. Scheme to generate 5' deletions in the TPI 5' noncoding
region. The TPI::lacZ fusion plasmid pHB110 contains a unique
Jth111\ site at position -853 with respect to the start of
translation. The plasmid was linearized with Jth111\ and treated
with exonuclease Bal31 for various times. H/'nd 11 l-Sa/l-H /nd 111
linkers (HB03, table 2) were added. The material was then digested
with Sail which cuts once in the polycloning region of pHB110 and in
the linker. After gel purification of the vector band, the material
was recircularized via ligation and used to transform E.coli.
Plasmid DNA was prepared from the transformants and precise
deletion endpoints were determined by DNA sequence analysis.
Restriction sites are as follows: E,EcoRI. S, Sail. H, H/'ndlll. T,


101
Many glycolytic enzymes have also been implicated in
interactions other than reactions required for glycolysis. One
isozyme of enolase has been identified as a possible plasminogen
receptor on the cell surface of human cells (Miles et al., 1991).
Pyruvate kinase, aldolase, glyceraldehyde-3-phosphate
dehydrogenase, phosphofructokinase, and lactate dehydrogenase have
all been found in association with actin and tubulin in mammals
(Walsh et al., 1989; Mejean et al., 1989). Phosphoglycerate kinase
has even been implicated in human lagging strand DNA replication
(Jindal and Vishwanatha, 1990). As far as is known, however, triose-
phosphate isomerase, encoded by TPI, is solely involved in
glycolysis.
GCR1 is thought to exert its effect on the expression of TPI,
and other affected genes, at the level of transcription (Clifton and
Fraenkel, 1981; Holland et al., 1987). Early experiments involving in
vitro transcription of isolated RNA (Clifton and Fraenkel, 1981).
Later, Northern blot analysis on TPI (Scott et al., 1990), PGI, PGK,
PYK ( Baker, unpublished result), and GAPDH (Holland et al., 1987)
have all shown reduced steady-state levels of transcript in a gcr1
strain when compared to wild-type. Specific message was still


127
Conrad, M.N., Wright, J.H., Wolf, A.J., and Zakian, V.A. (1990). RAP1
protein interacts with yeast telomeres in vivo: overproduction
alters telomere structure and decreases chromosome stability. Cell
63, 739-750.
Davison, B.L., Egly, J.M., Mulvihill, E.R., and Chambn, P. (1983).
Formation of stable preinitiation complexes between eukaryotic
class B transcription factors and promoter sequences. Nature 301,
680-686.
Devlin, C., Tice-Baldwin, K., Shore, D., and Arndt, K.T. (1991). Rap1 is
required for BAS1/BAS2- and GCN4-dependent transcription of the
yeast HIS4 gene. Mol. Cell. Biol. 11, 3642-3651.
Enea, U., Vovis, G.F., and Zinder, N.D. (1975). Genetic studies with
heteroduplex DNA of bacteriophage f1. Asymmetric segregation, base
correction and implications for the mechanism of genetic
recombination. J. Mol. Biol. 96, 495-509.
Ephrussi, A., Church, G.M., Tonegawa, S., and Gilbert, W. (1985). B
lineage-specific interactions of an immunoglobulin enhancer with
cellular factors in vivo. Science 227, 134-140.
Fassler, J.S. and Winston, F. (1988). Isolation and analysis of a novel
class of suppressor of Ty insertion mutations in Saccharomyces
cerevisiae. Genet. 118, 203-212.
Fassler, J.S. and Winston, F. (1989). The Saccharomyces cerevisiae
SPT13/GAL11 gene has both positive and negative regulatory roles
in transcription. Mol. Cell. Biol. 9, 5602-5609.
Fedor, M.J., Le, N.F., and Kornberg, R.D. (1988). Statistical
positioning of nucleosomes by specific protein-binding to an
upstream activating sequence in yeast. J. Mol. Biol. 204, 109-127.
Felsenfeld, G. (1992). Chromatin as an essential part of the
transcriptional mechanism. Nature 335, 219-224.


38
generated from M13mp18 with the forward sequencing primer to
serve as a molecular weight standard.
The shifted sample fragments, bracketed by the unshifted and
control DNasel ladder, were electrophoresed next to the sequencing
standard on a 0.4mm-7% polyacrylamide, 8M urea, 0.5X TBE gel at 50
volts/cm. After electrophoresis, the gel was blotted to Whatman 3M
paper and dried. The dried gel was exposed to X-ray film for
visualization.
DNA Band Shift Assays
DNA band shift assays were performed as previously described
(Scott et al., 1990). E. coli extracts were prepared by lysis with a
French pressure cell at 20,000 psi.
In vivo Methvlation Protection Analysis
Yeast strains were cultured in 2 liters of YP medium (Sherman
et al., 1983) supplemented with either 2% glucose or 2% lactate and
2% glycerol. Cultures were harvested at an A6oo of 1-0 by
centrifugation. Cells were washed twice in 137mM NaCI,2.7mM KCI,
4.3mM NaPC>4, 1.4mM KPO4 pH7.4 (PBS), and concentrated to 1x 108


13
interaction of the trans-activator and the adaptor is essential to aid
in the recruitment of TFIID or other basic transcription factors to
the TATA box. RAP1 dependent expression of PYK has been shown to
require functional GAL11/SPT13 (Nishizawa et al., 1990).
Nishizawa et al. suggest that GAL11/SPT13 was acting as an adaptor
protein in PYK expression. GAL11/SPT13 was first identified as a
gene required for full expression of some genes regulated by GAL4
(Nogi and Fukasawa, 1980), or as a suppressor for auxotrophic
mutants induced by Ty insertion (Fassler and Winston, 1988).
However, Nishizawa et al. note that Fassler and Winston (1989) did
not find decreased transcription of PYK1 in gall 1/spt13 cells.
Nishizawa et al. suggest that the disparity in the observations may
be due to allele specific differences of the gall 1/spt13 mutations.
The matter is by no means well resolved. Nishizawa et al. (1990)
showed that the requirement for GAL11 was alleviated by moving
the RAP1 binding site closer to the TATA element. Thus, the
requirement for the putative adaptor was alleviated by moving the
binding site for the second trans-activator closer to the TATA
element, supposedly allowing a direct interaction with the basic
transcription machinery such as TFIID.


al., 1990). The competitive nature of the interaction of RAP1 and
ABF1 at the UAS of EN02 is in contrast to the interactions that
occur between these factors at the UAS of PGK, where binding is not
thought to be competitive.
Expression of both enolase genes depends on GCR1 (Holland et
al., 1987), and, in the case of EN02, a small deletion in the area of
overlap of the two UAS's is able to relieve the requirement of GCR1
for full expression (Holland et al., 1990). However, when the UAS
elements of EN02 was moved immediately upstream of the TATA
boxes of EN01 or EN02 they were able to confer expression but
expression remained high in a gcr1 mutant strain. Thus, Holland et
al. (1990) suggest there may be an effect due to the positioning of
the UAS elements close to the TATA box which can alleviate the
requirement for GCR1.
However, when UAS1 and UAS2 of EN02 (positions -491 to -
443 with respect to the start of the EN02 structural gene) were
cloned upstream of a TPI::lacZ gene fusion, replacing the native UAS
element, no expression was seen (J. Anderson and H. Baker, personal
communication). Furthermore, the fragment was unable to bind
RAP1 in vitro. These results directly contradict those reported by


129
Huet, J., Cottrelle, P., Cool, M., Vignais, M.L., Thiele, D., Marck, C.,
Buhler, J.M., Sentenac, A., and Fromageot, P. (1985). A general
upstream binding factor for genes of the yeast translational
apparatus. EMBO J. 4, 3539-3547.
Huie, M.A., Scott, E.W., Lopez, M.C., Hornstra, I.K., Yang, T.P., and
Baker, H.V. (1992). Characterization of the DNA binding activity of
GCR1: In vivo evidence for two GCR1-binding sites in the upstream
activating sequence of TPI of Saccharomyces cerevisiae. Submitted
Ito, H., Fikuda, Y., Murata, K., and Kimura, A. (1983). Transformation
of intact yeast cells treated with alkali cations. J. Bacteriol. 153,
163-168.
Jindal, H.K. and Vishwanatha, J.K. (1990). Functional identity of a
primer recognition protein as phosphglycerate kinase. J. Biol. Chem.
265, 6540-6543.
Johnston, M. (1987). A model fungal gene regulatory mechanism: the
GAL genes of Saccharomyces cerevisiae. Microbiol. Rev. 51, 458-476.
Ju, Q., Morrow, B.E., and Warner, J.R. (1990). REB1, a yeast
DNA-binding protein with many targets, is essential for cell growth
and bears some resemblance to the oncogene myb. Mol. Cell. Biol. 10,
5226-5234.
Kawasaki, G. and Fraenkel, D.G. (1982). Cloning of yeast glycolysis
genes by complemetation. Biochem. Biophys. Res. Commun. 108,
1107-1112.
Kelleher, R.J., Flanagan, P.M., and Kornberg, R.D. (1990). A novel
mediator between activator proteins and the RNA polymerase II
transcription apparatus. Cell 61, 1209-1215.


INTRODUCTION
A fundamental process in biology is the regulation of gene
expression. For it is this process which allows a single fertilized
human egg to develop into an organism of 10^ cells in just nine
months. Furthermore, the study of cancer has shown that when the
process of gene regulation is perturbed a neoplastic transformation
may result. The yeast Saccharomyces cerevisiae offers an ideal
model system to study the regulation of gene expression. Over the
years a sophisticated genetic system has developed that allows one
to manipulate a gene in vitro and then reintroduce the mutated gene
back into the genome to assess the effect of the manipulation in
vivo. In addition, transcription factors in yeast often have homologs
in higher eukaryotes. For example, the JUN oncoprotein binds the
same recognition sequence and has extensive amino acid homologies
with the yeast transcriptional activator GCN4. JUN is even able to
functionally complement a gcn4 mutation in yeast (Struhl, 1988).
It has long been known that upon neoplastic transformation in
1


Figure 7. Identification of UAStpi- Various fragments containing portions of the 5' nontranslated
region of TPI were cloned before either a CYC1 ::lacZ construct deleted of the native UAS elements or a
TPI::lacZ construct deleted of the native UAS elements, p-galactosidase assays were performed from
cultures grown in YPD as described in figure 5 Fragment 34-1 H-S contains the sequence form -392
to -220 of TPI. 81mer contains from position -392 to -327. 66mer contains from position -377 to -
327. 58mer contains from -377 to -335. 69RBSM is identical to the 66mer except that the sequence
from positions -349 to -339 (CACCCCTTTTC) was replaced with the sequence AACCCATCAGG.


85
determine the DNA sequences bound by protein in the cell, dimethyl
sulfate (DMS) methylation protection assays (genomic footprinting)
(Ephrussi et al., 1985) were carried out on various strains of S.
cerevisiae. The strains utilized in the genomic footprints were
S150-2B (wild-type), DFY 642 (wild-type), HBY4 (gcr-1), and JF
1052 (spt13), Table 1. The gcr-1 strain was chosen because the
mutation is known to affect expression of TPI (Clifton and Fraenkel,
1981). SPT13/GAL11 gene function is required for the full
expression of PYK1 (Nishizawa et al., 1990).
Yeast were treated DMS, DNA harvested, prepared and blotted
to a nylon membrane as described in Materials and Methods. Top and
bottom strand radiolabeled probes, as defined in Figure 14, were
prepared to hybridize to the TPI controlling region. Figure 14 shows
the methylation protection pattern of the top strand of the TPI
controlling region from all four strains. The actual footprinting
controls and results are on the left of the figure. On the right a
cartoon represents the sequence of the UAStpi region with the
protected bases denoted. The first four lanes in Figure 14 are
genomic sequencing reactions that provided molecular weight
markers. The genomic "G" sequencing reaction was also the naked


75
Expression driven by UAStpi was 115 units of p-galactosidase
activity while the full length (-853) fusion construct expressed 161
units of activity.
In vitro DNase I Protection Assays Reveal Binding of the REB1 Site
and the RAP1 Site
Upstream activating sequence elements are sites where trans
acting factors bind to c/s-acting elements to mediate expression of
the cognate gene. UAStpi was initially used in a series of DNA
bandshift assays by others in the laboratory with crude protein
extracts derived from yeast. Using UAStpi and the RAP1-binding site
mutant derivative, HB19 or 69RBSM from the oligonucleotide cloning
experiments, they were able to demonstrate RAP1 binding to UAStpi
in a RAP1-binding site dependent manner (Scott et al., 1990). Since
UAStpi was specifically bound by protein, DNase I protection assays
as modified by Singh et al. (1986) were performed on the region to
determine the precise areas of interaction.
In order to detect the region protected by the RAP1
protein, the 169 base pair Hin\\\-Sph\ fragment from the 5' deletion
pES34-1 was used as a target in the DNasel protection assay. In


106
The role of the RAPI-binding site and the pentamer motifs in
the function of UASjpi was tested. Oligonucleotides that encode
UAStpi, or mutant variations thereof, were cloned before a TPI::lacZ
fusion which had been deleted of the native UAS sequence element(s)
and assayed for expression. The 51 base pairs constituting UAStpi
cloned before TPI::lacZ were able to express 120 units of (3-
galactosidase activity on average, Figures 7 and 10. A mutant,
69RBSM of Figure 7, that replaced most of the RAP1 -binding site
from -358 to -346 with the sequence from -420 to -410 was unable
to drive expression. Similarly, when the RAP1-binding site was
mutated by the introduction of an EcoRI and a Pst\ restriction site,
UAStpi activity was abolished, Figure 10. A RAP1-binding site alone
was insufficient to restore expression to the TPI;;/acZ fusion
deleted of the native UAS element(s). RAP1 binding is required but
not sufficient for high-level expression of the TPI::lacZ fusion.
Similar results have been reported for PGK (Stanway et al., 1989),
and the EN01 and PYK (Buchman et al., 1988). The assertion of
Santangelo and Tornow (1990), that a RAP1-binding site is
sufficient for UAS activity and is the site through which GCR1
exerts its effect, appears untenable.


95
initial mutation utilizes oligonucleotide HB32 to introduce an
additional RAP1-binding site into the 5' noncoding region of TPI at
position -420. No effect was seen upon expression in either wild-
type or gcr1 mutant strains. Thus, changes can be made in the region
without affecting expression.
Mutation of the REB1-binding site from 5'-AGATTACCCG-3' to
5'-AGATTGAACG-3' (HB31) resulted in a 2.5-fold reduction in
expression from 132 units of activity to 56 units. This reduction
was somewhat surprising, as the REB1-binding site can be deleted in
the 5 deletion series without affecting expression (Figure 6).
UAStpi activity was also seen to be independent of the REB1-binding
site in the UAS mapping experiments (Figure 7). Flowever, in the
context of native yeast sequence, mutations in the REB1-binding site
reduce expression.
Changing the CTTCC pentamer to CAACC (HB29) resulted in a 2-
fold reduction in expression to 70 units of p-galactosidase activity.
Mutating the CATCC to CAAAC (HB30) resulted in a 4-fold reduction
to 32 units of activity. Therefore, both elements were required for
full expression. But, unlike the experiments where an isolated
UAStpi was mutated, the mutation of the CATCC element had a


73
identical to the native TPI sequence, but for replacing the RAP1-
binding site with an EcoRI and Pst\ restriction site. When the HB24
derived mutant UAStpi was placed before the TPI::lacZ fusion, (3-
galactosidase activity was not expressed in either strain
background. Thus, UAStpi activity has an absolute requirement for
the RAP1-binding site.
In order to determine if a RAP1-binding site was sufficient for
UAS activity a consensus RAP1 site (AACCCATACATG),
oligonucleotide HB09, was cloned before the fusion. The consensus
RAP1 site had been shown to bind RAP1 protein (Scott et al., 1990).
As a control a mutant RAP1-binding site (AACCCATCAGG)
(oligonucleotide HB11), unable to bind RAP1, was also cloned before
the TPI::lacZ fusion. The consensus RAP1-binding site was not
sufficient to drive expression of the TPI::lacZ fusion, Figure 10, and
resulted in only 27 units of (3-galactosidase activity.
To address the role of the CTTCC and CATCC pentamer motifs
in UAStpi function, double-stranded oligonucleotides were generated
that changed either the CTTCC to CAACC (HB21), the CATCC to
CAACC (HB22), or both (HB24) while retaining a functional RAP1-
binding site. Similar mutations had been shown to reduce expression


121
Figure 18. REB1 binding is suggested to be required for expression
of the native TPI locus by the approximately 2-fold reduction in
expression seen when the REB1-binding site was mutated by site-
directed mutagenesis. REB1 most likely plays a role in chromatin
structure as previously suggested (Brandi and Struhl, 1990; Fedor et
al., 1988; Chasman et al., 1990) since its binding site can be deleted
without affecting expression when the deletion construct was
integrated immediately downstream of an actively transcribed gene,
Figure 6. Direct protein-protein interactions between REB1 and
GCR1 are unlikely considering the results of the double mutations of
the REBI-binding site and either of the GCR1-binding sites. The
double mutants retained much of the character of the individual
mutations. If protein-protein interactions were required between
REB1 and GCR1 both double mutants would have exhibited the same
effect.
Both RAP1 and GCR1 have been shown to be absolutely required
for the high level expression of TPI. Mutation of the RAP1-binding
site or a double mutation of both GCR1 -binding sites completely


6
transcript was produced, restoration of the native site or
substitution of the UAS derived RAP1-sites restored silencer
function and a1 transcript was no longer produced. These findings
indicate additional factors must play a role in RAP1 dependent
activation or repression.
RAP1 involvement in gene expression is widespread. RAP1
binds and is required for the function of the HMR and HML loci of the
mating type locus (Kimmerly et al., 1988; Hofmann et al., 1989),
genes encoding ribosomal proteins and other proteins of the basic
translational machinery (Huet et al., 1985), the H+-ATPase gene
(Capieaux et al., 1989), and the H/S4 gene (Devlin et al., 1991).
In addition to TPI, RAP1 binding has also been shown to be
necessary for the expression of several other genes encoding
glycolytic enzymes. Capieaux et al. (1989) were the first to note
near consensus RAP1-binding sites in the regulatory regions of GPD,
PGK, PYK, EN01, PDC, and ADH1. Subsequently, mutational analysis
has shown those sites to be essential for the expression of PGK
encoding phosphoglycerate kinase (Ogden et al., 1986; Chambers et
al., 1989), EN01 encoding an isozyme of enolase (Buchman et al.,
1988), PYK encoding pyruvate kinase (Buchman et al., 1988), ADH1


LIST OF FIGURES
page
Figure 1. Scheme to generate 5' deletions in the TPI
5' noncoding region 25
Figure 2. Scheme to create internal deletions 28
Figure 3. Scheme for integration of TPI::lacZ gene
fusion constructs at URA3 34
Figure 4. Primer extension analysis of the 5' ends of
the TPI transcript in wild-type and
gcr1-deletion mutant strains 50
Figure 5. Sequence of TPI 5' Noncoding Region 52
Figure 6. Effect of 5' deletions upon expression of a
TPI::lacZ fusion integrated in unit copy
at the URA3 locus in wild-type and
gcr1-deletion mutant strains 57
Figure 7. Identification of UAStpi 62
Figure 8. Effect of internal deletions on (3-galactosidase
activity expressed from a TPI::lacZ
gene fusion 67
Figure 9. Summary composite of the 5' noncoding
region of TPI 69
Figure 10. Mutational analysis of UASjpi utilizing
mutant oligonucleotides 72
Figure 11. In vitro DNase I protection assays demonstrating
protection of the RAP1-binding site 78
v i i


11 9
this model. Protein-protein interactions in this model would merely
serve to stabilize the binding of GCR1.
Direct interactions between RAP1 and GCR1 may not occur in
UAStpi as the bases protected from DMS methylation tend to be on
opposite sides of the DNA helix, suggesting RAP1 and GCR1 are also
on opposite sides of the helix. Therefore RAP1, like REB1, may be
needed to provide proper DNA structure for GCR1 binding. RAP1 is
known to be part of the nuclear scaffold (Cardenas et al., 1990), and
to be able to bend DNA (Vignais and Sentenac, 1989). The relative
binding affinity of GCR1 for linear UAStpi appears to be much less
than the affinity of RAP1 for UAStpi (Baker, 1991). Perhaps the
preferred binding substrate for GCR1 is bent DNA. RAP1 would be
required to provide the substrate for GCR1 binding and the
subsequent activation of TPI. Given the abundance of RAP1 and RAP1-
binding sites in yeast, a role for RAP1 in providing proper DNA
structure seems plausible. RAP1 would provide the proper
structural context for the binding of transcription factors that are
then able to perform there role, whether that be activation or
repression of transcription. Protein-protein interactions between
RAP1 and other factors may still play a role in this model, but, the


40
Control DNA was prepared simultaneously from an untreated aliquot
of the original cells.
50|ig of methylated DNA from each time point and 200pg of
untreated DNA was cleaved to completion with Avail. 1/10 volume
5M NH4AC was added, followed by 2.5 volumes ethanol. Samples
were centrifuged 10,000x g for 30 minutes. The pellet was washed
twice with 70% ethanol and dried in vacuo. Methylated DNA was
resuspended in 200pl 1M piperidine and incubated at 95C for 30
minutes. 40|il aliquots were removed for alkaline agarose gel
electrophoresis. 16|il of 3M NaAc pH 6.0 was added to remaining
sample, followed by precipitation with 2.5 volumes ethanol.
Samples were centrifuged 10,000x g for 30 minutes, washed twice
with 70% ethanol, and dried in vacuo. 50pil of TE pH 8.0 was used to
resuspend samples, and the concentration of the was determined
spectrophotometrically.
The aliquot of the piperidine cleavage reaction was ethanol
precipitated and resuspended in formamide sequencing dye. Samples
from the various timepoints were denatured at 95C and loaded onto
a 1.5% Agarose, 50mM NaOH, 1 .OmM EDTA gel next to molecular


100
exerts its effect, and other cis and trans-acting factors which may
be involved in high level glycolytic gene expression, the study
focused on the expression of TPI, a gene whose expression is
severely reduced by mutations in GCR1 (Clifton and Fraenkel, 1981).
TPI encodes, and is the sole source of triose-phosphate isomerase.
TPI expression normally accounts for approximately 2% of the total
soluble protein in S. cerevisiae (The Enzymes, Vol. VI), however,
expression is reduced 17-fold in a gcr1 mutant (Clifton and
Fraenkel, 1981). Understanding the cis and frans-acting elements
required for the high-level expression of TPI should aid in
understanding the common mechanisms required for the high-level
expression of most genes encoding glycolytic enzymes.
In order to understand the mechanism of high-level glycolytic
gene expression it was desirable to choose a gene that is unaffected
by the regulation of metabolic flux through glycolysis. For example,
the ADH2 gene, encoding an isozyme of alcohol dehydrogenase, is
repressed by glucose (Lutstorf and Megnet, 1968; Yu et al., 1989).
The ADH2 gene product favors the production of acetaldehyde from
ethanol, a reaction equilibrium preferred under gluconeogenic
conditions.


EXPRESSION OF THE TP/GENE OF SACCHAROMYCES CEREVISIAE
IS CONTROLLED BY A SINGLE COMPLEX UPSTREAM ACTIVATING
SEQUENCE CONTAINING BINDING SITES FOR THREE TPAA/S-ACTING
FACTORS: REB1, RAP1, AND GCR1
BY
EDWARD WILLIAM SCOTT V
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
1992


Figure 12. In vitro DNase I protection assays demonstrating
protection of the REB1-binding site. The first four lanes are a DNA
sequencing ladder generated with M13mp18 and the forward primer.
This ladder serves as a molecular weight marker. The next two
lanes are the products of DNase I digestion of a radiolabeled Hind\\\-
Sph\ fragment from plasmid pES40-23. (+) extract indicates the
DNase I digestion was carried out in the presence of yeast protein
extract. The (-) extract reaction was performed on naked DNA and
serves as the control Dnase I cleavage ladder. The region of DNase I
protection is demarcated and the sequence indicated.


60
position -392 of TPI encoded an UAS, and if RAP1 binding was
required for expression of TPI.
To address this question, fragments of the TPI controlling
region were used to replace the native UAS elements of a CYC1 ::lacZ
gene fusion. A Hind\\\-Sph\ fragment (-392 to -220) isolated from
the last 5' deletion construct to drive high-level expression of
TPI::lacZ (34-1) was able to restore expression to a CYC1 ::lacZ gene
fusion which had its native UAS elements removed (Figure 7). Cells
containing a CYC1::lacZ gene fusion which had a 66 base pair
oligomer (HB14 and HB15, Table2){TPI sequence from -377 to -327)
substituted for the native UAS elements of CYC1 were able to
express 99 units of (3-galactosidase activity. This level of
expression was comparable to the 94 units of activity expressed
using the longer TPI fragment, 34-1 H-S (-392 to -220) or the 101
units expressed by cells containing the 81 base pair oligonucleotide
(HB16 and HB17, Table 2)(-392 to -327). Therefore, UAS activity is
bounded by positions -377 to -327 which contains the RAP1-binding
site from -358 to -346. The sequence from -377 to -327 will be
termed UAStpi-
The same 51 base pairs was able to express 129 units of p-
galactosidase activity from the TPI::lacZ fusion which was deleted


Figure 12. In vitro DNase I protection assays demonstrating
protection of the REB1 -binding site 80
Figure 13. DNA band shift assays demonstrating REB1 binding
to the 5' noncoding region of TPI 84
Figure 14. Genomic footprinting of the bottom strand
of the TPI 5' noncoding region 87
Figure 15. Genomic footprinting of the top strand
of the TPI 5' noncoding region 92
Figure 16. Effect of site-directed mutations upon
(3-galactosidase activity expressed
from a TPI::lacZfusion 94
Figure 17. Composite summary of the TPI controlling
region 109
Figure 18. Model of protein interactions at UAStpi 122
vi i i


37
reaction was then loaded onto a 5% polyacrylamide, 0.5X TBE gel
running at 5 volts/cm. The gel had been pre-run for at least 1 hour.
The samples were electrophoresed at 7.5 volts/cm until the
bromphenol blue tracking dye, loaded along with the samples, was 2
cm from the bottom of the gel. The wet gel was wrapped in plastic
film and exposed to X-ray film overnight at 40C.
The developed film was used as a guide to excise the shifted
and unshifted portions of the radiolabeled fragment from the gel.
Identical positions from the quadruplicate samples were pooled for
extraction from the gel slice. Fragments were extracted from the
gel fragments by incubation at 370C overnight in 3 volumes of 0.5M
NH4Ac, 1mM EDTA. The extracted fragments were precipitated by
the addition of 2.5 volumes ethanol and centrifugation. The pelleted
fragments were washed with 70% ethanol and dried. The dried
pellets were Cherenkov counted and resuspended to 4,000 cpm/pl in
formamide sequencing dye.
A control DNase I ladder was generated from the unprotected
fragments by digestion for 2 minutes with 0.02 units of DNase I.
The control ladder was pelleted, counted, and resuspended in the
same manner as the sample fragments. A sequencing ladder was


112
to provide DMS methylation protection. It is also possible that
SPT13 binds to sites that are adenine and thymine rich.
The implications on the roles of REB1 and RAP1 binding on the
expression of TPI are twofold. First, neither REB1 nor RAP1 binding
alone is sufficient for the high-level expression of TPI as was
demonstrated by the genomic footprinting analysis of the gcr1
mutant strain. Therefore, the role of GCR1 in high-level expression
is not to facilitate the binding of REB1 or RAP1. Secondly, neither
REB1 nor RAP1 binding alone is sufficient for high-level expression
of TPI. The GCR1 dependent protection of the CTTCC and CATCC
pentamer motifs makes the hypothesis of Santangelo and Tornow
(1990), that GCR1 acts through a RAP1-binding site, appear
untenable.
The 5' and internal deletion series have mapped a single
upstream activating sequence for TPI. UAStpi activity was
demonstrated to lie from position -377 to -327. A RAP1-binding
site from -358 to -346 was shown to be absolutely required for
UAStpi activity. Two pentamer sequence motifs were also
demonstrated to be required for full UAS activity. Furthermore, in
vitro and in vivo footprinting analysis have demonstrated RAP1 and


128
Fraenkel, D.G. (1982). Carbohydrate metabolism. In The molecular
biology of the yeast Saccharomyces: Metabolism and gene
expression. J.N. Strathern, E.W. Jones, and J.R. Broach, eds. (Cold
Spring Harbor: Cold Spring Harbor Laboratory), pp. 1-38.
Gibson, T.J. (1984). Studies on the Epstein-Barr virus genome. Ph.d
thesis, Cambridge University, England, thesis
Guarente, L., Lalonde, B., Gifford, P., and Alani, E. (1984). Distinctly
regulated tandem upstream activation sites mediate catabolite
repression of the CYC1 gene of S. cerevisiae. Cell 36, 503-511.
Guarente, L. and Mason, T. (1983). Heme Regulates Transcription of
the CYC1 gene of S. cerevisiae via an upstream activation site. Cell
32, 1279-1286.
Hofmann, J.F., Laroche, T., Brand, A.H., and Gasser, S.M. (1989). RAP-1
factor is necessary for DNA loop formation in vitro at the silent
mating type locus HML. Cell 57, 725-737.
Holland, J.P., Brindle, P.K., and Holland, M.J. (1990). Sequences within
an upstream activation site in the yeast enolase gene EN02 modulate
repression of EN02 expression in strains carrying a null mutation in
the positive regulatory gene GCR1. Mol. Cell. Biol. 10, 4863-4871.
Holland, M.J., Hager, G.L., and Rutter, W.J. (1977). Characterization of
Purified Poly(adenylic acid)-containing messenger ribonucleic acid
from Saccharomyces cerevisiae. Biochem. 16, 8-16.
Holland, M.J. and Holland, J.P. (1978). Isolation and identification of
yeast messenger ribonucleic acids coding for enolase,
glyceraldehyde-3-phosphate dehydrogenase, and phosphoglycerate
kinase. Biochem. 4900, 4907.
Holland, M.J., Yokoi, T., Holland, J.P., Myambo, K., and Innis, M.A.
(1987). The GCR1 gene encodes a positive transcriptional regulator
of the enolase and glyceraldehyde-3-phosphate dehydrogenase gene
families in Saccharomyces cerevisiae. Mol. Cell. Biol. 7, 813-820.


50


DISCUSSION
In summary, this study has determined the initiation or "I" site
of TPI transcription to be a pair of adenines positioned 29 to 30
bases upstream of the start of translation. The TATA box utilized
for high-level TPI transcription was mapped to position -175 to
-169. Mutational analysis indicated that a single complex upstream
activating sequence, UAStpi, is responsible for expression of TPI.
Three trans-acting factors: REB1, RAP1, and GCR1 have been
identified that bind UAStpi and are required for full expression of
TPI.
The aim of this project was to investigate the mechanism of
high level glycolytic gene expression in Saccharomyces cerevisiae.
The GCR1 gene product is known to be required for high-level
glycolytic gene expression (Clifton et al., 1978; Clifton and
Fraenkel, 1981; Baker, 1986). gcr1 lesions result in a 90-95%
reduction in the specific activity of most of the glycolytic enzymes.
In order to identify the cis-acting element through which GCR1
99


20
polyacrylamide gel electrophoresis (PAGE), digested with the
appropriated restriction enzymes, and cloned into the polylinker
region of pUC18 or other suitable plasmid vectors. This method was
used to generate the double-stranded form of oligonucleotides HB16,
HB19, HB21, HB22, HB24, HB25 listed in Table 2. Additional double-
stranded oligonucleotides were generated for cloning by annealing
complementary oligonucleotide pairs, also shown in Table 2.
Followed by digestion with the appropriate restriction enzymes, and
cloning.
DNA Sequencing
DNA sequencing was carried out by the dideoxy chain
elongation termination method of Sanger (1977) as modified by U.S.
Biochemicals to utilize the Sequenase enzyme. Sequencing primers
were synthesized by the University of Florida Interdisciplinary
Center for Biotechnology Research and are listed in Table 2.


FIGURE 5. Sequence of TPI 5 Noncodina Region
52
I -1191
acgtcatcgatgaatataatgaattaaacagtggtgttcgtatatgtgaagatatgagatatga
tccacatggtaaacagaaagatgcattttggccgagaggacttaataatactggtggtgtttac
gaaaataatgaagataatatttgtgaagggaagcctggaaaatgttatctgcaatatcgggtta
aggatgagccaagaataagggaacaagattttggtaatttccaaaaaatcaatagcatgcagg
acgttatgaagaagagatctacgtatggtcatttcttcttcagattccctcatggagaaagtgc
I -853
aacaaatatatataacaaaatcaccaatttccaaaaaactttattcaaacacttccataataaa
caagagagaagacccagagatgttgttgtcctagttacacatggtatttattccagagtattcc
tgatgaatggtttagatggacatacgaagagtttgaatcgtttaccaatgttcctaacgggagc
I -658 (Previously Known Sequence)
gtaatggtgatggaactggacgaatccatcaatagatacgtcctgaggaccgtgctacccaaa
tggactgattgtgagggagacctaactacatagtgtttaagattacggatatttaacttactta
gaataatgccatttttttgagttataataatcctacgttagtgtgagcgggatttaaactgtgag
gaccttaatacattcagacacttctgacggtatcaccctacttattcccttcgagattatatct
aggaacccatcaggttggtggaagattacccgttctaagacttttcagcttcctctattgatgt
tacacctggacaccccttttctggcatccagtttttaatcttcagtggcatgtgagattctccg
aaattaattaaagcaatcacacaattctctcggataccacctcggttgaaactgacaggtggtt
I -220
tgttaccicatcictaatgcaaaggagcctatatacctttggctcggctgctgtaacagggaatat
aaagggcagcataatttaggagtttagtgaacttgcaacatttactattttcccttcttacgta
aatatttttctttttaattctaaatcaatctttttcaattttttgtttgtattcttttcttgcttaa
I +1
atctataactacaaaaaacacatacataaactaaaaATG


65
had a SphI site at position -220 that served to fix the 3' deletion
endpoint at position -220. All "internal" deletion constructs had
position -853 as their common 5' end. Precise 3' deletion end-points
and junctions were determine by sequencing using primer HB05,
Table 2, and are shown in Figure 8. The "internal" deletion series
was integrated into both wild-type (S150-2B) and gcr1 mutant
(HBY4) strains of yeast. Unit copy integrants were confirmed by
Southern blot analysis (Southern, 1975) as previously detailed, p-
galactosidase assays by the method of Miller (1972) were carried
out in duplicate on three independent occasions.
Figure 8 depicts the "internal" deletion constructs and the
results of the p-galactosidase assays from lysates of strains with
unit copy integrants of the various plasmids. 392 base pairs of 5'
non-coding region was sufficient for high-level expression of the
TPI::lacZ fusion. The initial "internal" deletion construct contained
5' sequence from -853 to -300 and from -220 through the structural
gene. This construct retained the known UAStpi sequence, and
produced 138 units of p-galactosidase activity when integrated into
the wild-type strain, S150-2B. 52 units of p-galactosidase activity


120
overall characteristics of the control element would be due to the
other factors, not RAP1.
This study suggests a common mechanism for the high-level
expression of genes encoding glycolytic enzymes. RAP1 and GCR1
have been shown to be required for the high-level expression of all
genes encoding glycolytic enzymes that have been investigated to
date (Scott et al., 1990; Stanway et al., 1989; Chambers et al., 1989;
Buchman et al., 1988; Brindle et al., 1990; Holland et al., 1987;
McNeil et al., 1990; Santangelo and Tornow, 1990). Therefore, it
seems likely that activation by RAP1 and GCR1 is the common theme
used for high-level glycolytic gene expression. Differences in
expression between alleles is likely due to the binding of additional
factors such as ABF1 (Chambers et al., 1990; Brindle et al., 1990),
REB1 (Chasman et al., 1990), and EBF2 (Brindle et al., 1990). One
test of the commonality of RAP1 and GCR1 binding will be to carry
out genomic footprinting experiments on several additional
glycolytic gene promoters to determine if RAP1 and GCR1 are bound
in each case.
The data presented in this study suggests a possible model for
the activation of high-level TPI expression. The model is depicted in


4
dehydrogenase, phosphoglycerate kinase, and pyruvate kinase in a
gcr1 mutant is somewhat inducible by glucose, relative to
expression under gluconeogenic conditions (Baker, 1986). Insights
into the mechanism by which the GCR1 gene product exerts its
effect will prove valuable in the exploration of glycolytic gene
regulation.
This study will attempt to elucidate the cis and trans-acting
elements involved in the expression of TPI, the sole gene encoding
triose-phosphate isomerase activity in Saccharomyces cerevisiae.
Triose-phosphate isomerase catalizes the reversible isomerization
of dihydroxyacetone phosphate to glyceraldehyde 3-phosphate during
glycolysis. TPI has been shown to be dependent upon GCR1 for full
expression; in a gcr1 mutant strain TPI gene expression is reduced
17-fold (Clifton and Fraenkel, 1981). By mapping the elements of
the TPI controlling region, it should prove possible to map the DNA
sequence through which the GCR1 gene product acts and to determine
if GCR1 acts alone or in conjunction with other sites or factors. In
order to facilitate the identification of the site of GCR1 action in
the TPI controlling region, expression of a TPI::lacZ fusion will be
analyzed in isogeneic wild-type and gcrI mutant strains. The


Figure 3. Scheme for integration of TPI::lacZ gene fusion
constructs at URA3. A) depicts a plasmid containing a TPIr.lacZ
fusion construct and a URA3 selectable marker juxtaposed above the
ura3-52 locus of the yeast genome. The plasmid is linearized within
URA3 to direct integration via homologous recombination. Sacl sites
are indicated. B) represents unit copy integrant at ura3-52.
Digestion with Sacl gives rise to three fragments A, B, and C. C)
represents a tandem integration event. Upon digestion with Sacl,
five fragments are generated: A, 2B, C, and a fourth fragment, D,
which is diagnostic of a tandem insertion event.


2
certain types of cancer there is an increase in aerobic glycolysis
(Warburg, 1930). Saccharomyces cerevisiae utilizes aerobic
glycolysis to a much greater extent than respiration (Lagunas,
1986). The enzymatic pathway of glycolysis in yeast is well
established. The enzymes of glycolysis, while few in number,
compose between 30-60% of the total soluble protein (Fraenkel,
1982). This observation suggests that the genes encoding these
enzymes are among the most highly expressed in yeast. Indeed,
mRNA encoding glycolytic enzymes has been demonstrated to be a
major fraction of total yeast mRNA (Holland et al., 1977; Holland and
Holland, 1978). The regulation of the genes encoding the glycolytic
enzymes is currently receiving much study, but no overall consensus
regulatory mechanisms have yet been identified, rather some
similarities in regulatory elements and factors have been noted.
These similarities will be addressed subsequently.
Mutations affecting the flux of metabolites through the
glycolytic pathway tend to map to single loci and affect single
enzymes (Fraenkel, 1982). However, Clifton et al. (1978) isolated a
mutant that has severely reduced levels of most glycolytic enzymes.
Yet genetic analysis showed that this strain contains a mutation


Figure 15. Genomic footprinting of the top strand of the TPI 5'
noncoding region. The initial lanes are genomic sequencing reactions
as indicated. Guanine sequencing reactions (G) serve as a control
ladder for the genomic footprinting reactions. Wild-type (S150-2B)
or gcrl mutant (HBY4) yeast were treated with DMS for the time in
minutes indicated above the lanes. The sole area of protection due
to REB1 binding is denoted by (A). The right portion of the figure
depicts the double stranded sequence of the TPI 5' noncoding region.
The top strand is the lefthand most strand as depicted. The
sequence motifs known to play a role in TPI expression are stippled.
Protected guanines are denoted as above.


125
Bram, R. and Kornberg, R.D. (1982). Specific protein binding to far
upstream activating sequences in polymerase II promoters. Proc.
Natl. Acad. Sci. USA 43, 47.
Brandi, C.J. and Struhl, K. (1990). A nucleosome-positioning sequence
is required for GCN4 to activate transcription in the absence of a
TATA element. Mol. Cell. Biol. 10, 4256-4265.
Brindle, P.K., Holland, J.P., Willett, C.E., Innin, M.A., and Holland, M.J.
(1990). Multiple factors bind the upstream activation sites of the
yeast enolase genes EN01 and EN02: ABFI protein, like repressor
activator protein RAP1, binds c/s-acting sequences which modulate
repression or activation of transcription. Mol. Cell. Biol. 10,
4872-4885.
Buchman, A.R., Kimmerly, W.J., Rie, J., and Kornberg, R.D. (1988).
Two DNA-binding factors recognize specific sequences at silencers,
upstream activating sequences, autonomously replicating sequences,
and telomeres in Saccharomyces cerevisiae. Mol. Cell. Biol. 8,
210-225.
Buchman, A.R. and Kornberg, R.D. (1990). A yeast ARS-binding protein
activates transcription synergistically in combination with other
weak activating factors. Mol. Cell. Biol. 10, 887-897.
Buchman, A.R., Le, N.F., and Kornberg, R.D. (1988). Connections
between transcriptional activators, silencers, and telomers as
revealed by functional analysis of a yeast DNA-binding protein. Mol.
Cell. Biol. 8, 5086-5099.
Capieaux, E., Vignais, M.L., Sentenac, A., and Goffeau, A. (1989). The
yeast H+-ATPase gene is controlled by the promoter binding factor
TUF. J. Biol. Chem. 264, 7437-7446.
Cardenas, M.E., LaRoche, T.L., and Gasser, S.M. (1990). The
composition and morphology of yeast nuclear scaffolds. J. Cell Sci.
96, 439-450.


GGAACCCATCAGGTTGGTGGAAGATTACCCGTTCTAAGACTTTTCAGCTTCCTCTATTGATGTTACACCTGGACACCCCTTTTCTGGCATCCAG
REB1
1 RAP1 | | REbT
REB1
REB1
[321
REB1
REB1
-I GCRl h
HGCRt h
-I GCR1 h
-| GCRl h
H GCR1 h
RAP1 | 1 GCRl y-
RAP1 | ¡GCR1 |
RAPI | 1 GCRl y-
RAPl | 1 GCRl y-
iM-
REBl
RAPI
RAP1
RAP1
RAP1
] mi
GCR1 |
mm-
mmm IgcrTV-
REB1 | 1 GCRl |
1 RAP1 [ 1 REB1 h
B-Galactosidase Activity
S150-2B
(GCR1)
Mean SD
HBY4
(.gen)
Mean SD
132 6
9 4
132 12
5 2
56 20
5 3
70 2
6 3
16 1
6 4
21 1
4 2
32 3
6 3
11 3
6 3
41 4
5 2
17 2
4 2
55 3
5 3
10 1
5 2
13 1 4 2
CD
4^


TABLE OF CONTENTS
page
ACKNOWLEDGMENTS iii
LIST OF TABLES vi
LIST OF FIGURES vii
ABSTRACT ix
INTRODUCTION 1
MATERIALS AND METHODS 17
Strains 17
Media and Growth Conditions 17
Nucleic Acid Manipulations 19
Generation of Double Strand DNA Oligonucleotides 19
DNA Sequencing 20
Primer Extension 22
Plasmid Construction 22
Constructs for sequencing 22
5' deletion scheme 23
Internal deletion scheme 26
Constructs to assess UAS activity 29
Mutant UAS oligonucleotides driving expression of
the TPI::lacZfusion 29
Site-Directed Mutagenesis 30
Transformation 32
Screen for Unit Copy Integrants of the TPI::lacZ Fusions
at URA3 32
p-Galactosidae Assays 35
In vitro DNasel Protection Assays 36
DNA Band Shift Assays 38
In vivo Methylation Protection Analysis 38


26
marker was determined by restriction endonuclease analysis with
BamHI.
Internal deletion scheme Plasmid pES35 contains the TPI::lacZ
fusion with 853 bp of 5' DNA on a Hind\\\ fragment subcloned into
Ylp56 at H/'ndlll. The SphI restriction site at position -220 with
respect to the start of the structural gene was a convenient origin
for exonuclease Bal31 digestion. Therefore, the Sph\ site at -220
was rendered unique by a Kpn\ dropout of pES35, which removed a
second Sph I site in the polylinker region of pES35, generating
plasmid pES90. The scheme to introduce Ba/31-induced deletions
into pES90 is shown in Figure 2. Plasmid pES90 was linearized with
Sph\ and treated for various times with exonuclease BAISl. The
Klenow fragment of DNA polymerase I was used to fill in the ends of
the remaining material. Sph\ linkers (HB06) were added. The
material was then digested with Sph I and Hind\\\ to liberate DNA
distal to the Sph I site at -220. The H/ndlll-Sp/7l fragment that
corresponds to the Ba/31 treated DNA of yeast origin distal to the
Sph\ site was gel purified and subcloned into pES90 which had been
digested with Hind\\\ and Sph\. The resulting material was
transformed into E. coli MC1061. Plasmid DNA was prepared, and


25
Tth111l digest
Bal 31 digest

T


103
Expression of the fusion could indirectly be followed by p-
galactosidase activity assay. The Miller assay (Miller, 1972) was
chosen to follow (3-galactosidase activity because of the simplicity
of the assay.
5' deletion analysis suggested that all sequences required for
the high level expression of a TPI::lacZ fusion resided with 392
bases of the start of the structural gene. Deletion clone 34-1
(-392) still retained high-level expression, whereas, 70-1 (-377)
expresses only 29-43% relative activity (see Figure 6). These
results suggested that the 5' limit of an upstream activating
sequence (UAS) was bounded by positions -392 and -377 with
respect to the start of translation.
Mapping of a UAS element from TPI was accomplished using a
CYC1::lacZ test construct. The UAS1 and UAS2 elements of CYC1
(Guarente et al., 1984) were replaced with portions of the TPI 5' non
coding region in the CYC1 ::lacZ construct. A 66 base pair
oligonucleotide, containing 51 bases of TPI 5' non-coding region
from position -377 to -327, was able to restore expression to the
CYC1 ::lacZ fusion, see Figure 7. The same oligonucleotide was able
to restore expression to a TPI::lacZ fusion deleted to position -220.


-366
CO
CO
5' tgatgttacacctggacacccct
JO ro
CD 00
O CT)
TTAA 3'
Â¥
m
m
m
m*
Ml *
* *# ifI fjf
i ihimi i in m l
* $ i ill
i
* f
i Cl 0
ft 0 1 ? Iff Mt
Milt lit I III I o
o

M I H
MMif000 * i I0H 1
vi
oo
DNasel


Holland et al. (1990).
PYK has two UAS elements: UAS1 (-653 to -634), UAS2 (-811
to -714). PYK is also controlled by an URS (-468 to -344) to repress
expression during growth on nonfermentable carbon sources
(Nishizawa et al., 1989). Full expression of PYK requires GCR1
(Clifton and Fraenkel, 1981). The UAS1 binds RAP1 and has a CTTCC
pentamer that enhances UAS activity (Buchman et al., 1988).
Mutation of the CTTCC pentamer to CAACC reduced UAS activity by
16-fold. Deletion of the RAP1-binding site of UAS1 prevents
expression of PYK (McNeil et al., 1990). The pentamer is not
required for binding of RAP1 to UAS1 in vitro (Buchman et al., 1988).
The overall organization of UAS1 of PYK is very similar to the
UAS of PGK (Buchman et al., 1988; Chambers et al., 1990). Both PYK
and PGK have ABF1-binding sites adjacent to their RAP1-binding
sites and CTTCC sequence elements. The role of ABF1 binding in
expression of PYK has yet to be determined. In addition, another
factor known as REB1 (Morrow et al., 1989)/ QBP (Brandi and Struhl,
1990)/ Y (Fedor et al., 1988), hereafter called REB1, binds in the
UAS2 area of PYK (Chasman et al., 1990).
REB1 has been shown to affect nucleosome positioning in UAS


Figure 16. Effect of site-directed mutations upon (3-galactosidase activity expressed from a TPI::lacZ
fusion. The sequence of the TPI 5' noncoding region targeted by site-directed mutagenesis is depicted
at the top of the figure. The known binding motifs are cartooned in the next line. Mutations
introducing a known binding motif are denoted by open boxes. Mutations that destroy the activity of a
known motif are stippled, see text for precise details. The results of duplicate (3-galactosidase assays
performed on three separate occasions are listed. Assays were performed on unit copy integrants of
the various constructs in wild-type (S150-2B) and gcr1 mutant (HBY4) strains of yeast. SD indicates
standard deviation.