Structural studies of bovine cytochrome oxidase subunit II messenger RNA and the characterization of its interaction wit...


Material Information

Structural studies of bovine cytochrome oxidase subunit II messenger RNA and the characterization of its interaction with the 28s subunit of bovine mitochondrial ribosomes
Physical Description:
xv, 145 leaves : ill. ; 29 cm.
Courtney, Bernard Clark, 1957-
Publication Date:


Subjects / Keywords:
Research   ( mesh )
Electron Transport Complex IV -- genetics   ( mesh )
RNA, Messenger -- ultrastructure   ( mesh )
RNA, Messenger -- genetics   ( mesh )
Ribosomes -- ultrastructure   ( mesh )
Mitochondria, Liver   ( mesh )
DNA, Mitochondrial   ( mesh )
Molecular Structure   ( mesh )
Binding Sites   ( mesh )
Translation, Genetic   ( mesh )
Molecular Sequence Data   ( mesh )
Base Sequence   ( mesh )
Cattle   ( mesh )
Department of Biochemistry and Molecular Biology thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Biochemistry and Molecular Biology -- UF   ( mesh )
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis (Ph.D.)--University of Florida, 1992.
Bibliography: leaves 138-144.
Statement of Responsibility:
by Bernard Clark Courtney.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 002328680
oclc - 50543575
notis - ALT2322
System ID:

This item is only available as the following downloads:

Full Text







I dedicate this research to my Lord and Savior, Jesus

Christ, to my wife, Janet Marie (Tobin) Courtney and to my

sons, Ryan Clark, Sean Michael and Quinn Stephen Courtney.


The people who should be recognized for their support,

academic, technical, and emotional would fill most of this and

several other sections. Drs. Denslow, Laipis, Lewin, and Nick

were of utmost help and encouragement in the pursuit of

realizable research as my supervisory committee. Dr. Thomas W.

O'Brien, my committee chair, receives my warmest thanks for

his endless patience and encouragement, his unusual wit, and

his boundless excitement and imagination. I am a different

and a better man for the experience and I hope he feels the

same. The members of the O'Brien lab, John Anders, Mike

Bryant, Wes Faunce, Scott Fiesler, Bob Heck, and Jiguo Liu,

have been more than mere co-workers, they have been friends in

the truest sense of the word. That friendship has been tested

many times, but the mettle of it was never more evident than,

when shoulder to shoulder, shivering and aching in the cold

room, we were visited with "the way I would do it IF I were

doing it." Of these labmates Wes and Mike must be singled out

for their meritorious counseling in the face of imminent doom

and disaster.

I would like to thank my parents, Bernard G. and Wilma A.

Courtney, for instilling in me the desire to always be a


little better and the confidence to go on regardless of the


Thanks also go to my wife's family for their care,

patience, and daughter (and for keeping the cars rolling).

The impact of the joyful, unconditional love of three

boys for their father need not be explained, only appreciated

and I do. In no manner could I have done this without the

quiet strength, support, understanding, endurance and

occasional prod of my best friend and wife, Janet.


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

LIST OF TABLES ............................................ vii

LIST OF FIGURES ..........................................viii

ABBREVIATIONS............................................... .X

ABSTRACT .................................................. xiv

BACKGROUND ................................................... 1

MATERIALS AND METHODS ...................................... 21

Preparation of Bovine Mitochondria ....................21
Preparation of Native 28S Subunits and 55S
Monosomes........................................ 22
Preparation of Derived 28S Subunits ................... 22
Preparation of Mitochondrial DNA ...................... 24
Cloning Mitochondrial Genes............................ 25
Engineering an ATPase 8/6 Transcription Vector........29
Preparation of Transcription Vectors .................. 31
Preparation of mRNA Transcripts ....................... 33
Preparation of Radiolabelled Synthetic RNA............ 33
Preparation of 5' End Labelled Nucleic Acids.......... 34
Binding Substrates to 28S Subunits .................... 35
Millipore Filter Binding Assay ........................36
Sucrose Density Gradient Centrifugation............... 36
Chemical and Enzymatic Modifications .................. 37
DNA Primers........................................... 40
Extension of Oligodeoxyribonucleotide Primers.........40
Denaturing Gels ...................................... 42
Gel Reading Algorithm.................................. 43
Scanning Densitometry................................ 44

BOVINE MITOCHONDRIAL RIBOSOMES........................ 45

Significance of 28S Ribosomal Subunit Binding
Properties ....................................... 45
Bovine Mitochondrial mRNA Binding Properties of 28S

Interaction of mRNAS of Non-Mitochondrial Origin with
Mitochondrial 28S Subunits....................... 58
Interaction of Non-message Polynucleotides with
Mitochondrial 28S Subunits ....................... 65
Competition for 28S Ribosomal Subunit Binding by
Mitochondrial and Non-mitochondrial mRNAs........69
Effects of a Ribonucleic Acid Binding Inhibitor on
28S Ribosomal Subunit Binding .................... 75
Conclusions........................................... 77

MITOCHONDRIAL COII MESSENGER RNA ...................... 84

Background and Significance of Secondary Structure
Analysis ......................................... 84
Computer Modelling of the Secondary Structure of
COIImRNA ............................... ......... 87
Enzymatic Probing of the Secondary Structure of COII
mRNA........... ................................. 90
Chemical Probing of the Secondary Structure of COII
mRNA............................ ................. 91
Refinement of the Secondary Structure of COII mRNA...103
Indications of Additional Structures on COII mRNA.... 107
Conclusions........................................... 109


Background and Significance of 28S Subunit
Interaction with COII mRNA ...................... 111
Footprints of 28S Subunits on COII mRNA .............. 114
Characterization of the 28S Subunit Interaction Site
on COII mRNA.................................... 124
Conclusions.......................................... 130

CONCLUSIONS AND FUTURE DIRECTIONS......................... 133

Summary .................... .......................... 133
Future Directions................................ .... 135

REFERENCES................................... ............. 138

BIOGRAPHICAL SKETCH...................................... 145


Table Dpa

1. Unique Properties of Mitochondrial Translation.........8

2. Oligonucleotides and Their Use ........................ 27

3. Binding Analysis of Mitochondrial mRNAs to 28S
Subunits, Native and Derived ......................... 50

4. Binding Analysis of COII mRNA to 28S Subunits.........52

5. Binding Analysis of Sucrose Density Gradients (COII)..57

6. Binding Analysis of Non-mitochondrial mRNAs........... 63

7. Binding Analysis of Sucrose Density Gradients
Poly(U) .............................................. 67

8. Binding Analysis of Non-Message Polynucleotides.......71

9. Analysis of Competition for COII mRNA Binding.........74

10. Analysis of the Scanning Densitometry of Modification
Diminutions for 28S Subunit and COII mRNA Binding...123



Figure page

1. The Three Stages of Translation ...................... 4

2. 5' Ends of Bovine Mitochondrial mRNAs............... 10

3. The RNA Binding Site on Mitochondrial Ribosomes.....12

4. Site of Chemical Modifications on Ribonucleotides...19

5. Absorbance Profiles of Ribosomes in Sucrose Density
Gradients.......................................... 23

6. Clone 38 of ATPase 8/6 for Mutagenesis.............. 26

7. Oligonucleotide-directed, Site-specific Mutagenesis
in Conjunction with Polymerase Chain Reaction......28

8. Engineering the ATPase 8/6 Clone .................... 30

9. Transcription Clones pJC213 (A8/6) and p2-6E (COII).32

10. Titration of Rnase A and DMS for the Cleavage or
Modification of COII mRNA.......................... 39

11. The Location of Primers on COII Sequence............ 41

12. Binding of Mitochondrial mRNAs to 28S Subunits,
Native and Derived ................................. 49

13. Binding of COII mRNA to 28S Subunits................ 51

14. Sucrose Density Gradients with COII mRNA............55

15. Binding Assayed by Sucrose Density Gradients of COII
mRNA and 28S Subunits............................... 56

16. C3,-endo "Pucker" and Structure of Poly(C) Single
Helix .............................................. 60

17. Binding of Non-mitochondrial mRNAs by 28S Subunits..62

18. Sucrose Density Gradients with Poly(U).............. 66


19. Non-Messages Used for Binding Assays................ 68

20. Binding of Non-message Polynucleotides
by 28S Subunits ....................................70

21. Competition for the Binding of COII on 28S Subunits.73

22. Chemical Structure of Aurin Tricarboxylic Acid......78

23. Effect of ATA on COII mRNA Binding to Ribosomes.....79

24. Secondary Structure Prediction of FOLD for COII
mRNA ............................................... 89

25. Ribonuclease Cleavages on a Extensions.............. 92

26. Ribonuclease Cleavages on P Extensions.............. 93

27. Cobra Venom Endonuclease Cleavages on 0 and a
Extensions......................................... 94

28. Nuclease Cleavages on COII Sequence................. 95

29. DMS Modifications on n Extensions................... 99

30. DMS Modifications on 3 Extensions .................. 100

31. DMS Modifications on COII Sequence................. 101

32. Model for the Secondary Structure of COII mRNA.....104

33. Potential Additional Structures on COII mRNA Model.108

34. Reduced Cleavages of RNase Ti on A Extensions......117

35. Reduced Cleavages of RNase T, on a Extensions......118

36. Reduced Cleavages of RNase T, on Extensions......119

37. Reduced Cleavages of RNase T, on Extensions......120

38. Reduced Cleavages of RNase A on / Extensions.......121

39. Densitometry of Cleavage Reductions of RNase A
and TI on / Extensions............................. 122

40. Diminished Cleavages on the COII mRNA Secondary
Structure Model ................................... 125


A 8/6.....messenger ribonucleic acid for subunits 8 and 6

of adenosine 5'-triphosphatase

Ac........ acetate

AMV(-RT)..avian myoblast virus (reverse transcriptase)

ATA.......Aurin tricarboxylic Acid

ATP.......adenosine 5'-triphosphate

ATPase .... adenosine 5'-triphosphatase

BSA.......bovine serum albumin

Ci........ Curie

COII......messenger ribonucleic acid for subunit II of

cytochrome oxidase

CPM....... Counts per minute

CPU.......Central processing unit

CTP.......cytosine 5'-triphosphate

CytB......cytochrome B

dATP......deoxyadenosine 5'-triphosphate

dCTP......deoxycytosine 5'-triphosphate

dGTP......deoxyguanosine 5'-triphosphate

DMS.......dimethyl sulfate

DNA.......deoxyribonucleic acid


ds........double stranded

dTTP......deoxythymidine 5'-triphosphate

E. coli...Escherichia coli

elF.......eukaryotic initiation factor

fMet......formylated methionine

Fmet-tRNAifMet..methionine initiator transfer ribonucleic

acid charged with formylated methionine performance liquid chromatography

GCG.......Genetics Computer Group

GDP.......guanosine 5'-diphosphate

GTP.......guanosine 5'-triphosphate

ICBR......Interdisciplinary Center for Biotechnology

IF........ initiation factor



Kd........dissociation constant




Met-tRNAiMet..methionine initiator transfer ribonucleic

acid charged with methionine




mRNA......messenger ribonucleic acid

mt........mitochondria or mitochrondrial


Ai ........microliter

Am ........micron or micrometer

AM ........micromolar

Amol...... micromole

ND........nicotinamide adenine dinucleotide (reduced)





PCR.......Polymerase Chain Reaction



poly(C)...polycytidylic acid

poly(U)...polyuridylic acid

rib(s) .... ribosome(s)

RNA.......ribonucleic acid


RNasin....ribonucleasin, a trademark product of Promega

rRNA......ribosomal ribonucleic acid



SDS....... sodium dodecylsulfate

s........ single stranded

tRNA......transfer ribonucleic acid

tRNAi..... initiator transfer ribonucleic acid


tRNAiMet..methionine initiator transfer ribonucleic acid

tRNAifMet..formylated methionine initiator transfer

ribonucleic acid


UTP.......uracil 5'-triphosphate

28S.......mitochondrial ribosomal small subunit
32p ....... radioisotope of phosphorous

39S.......mitochondrial ribosomal large subunit

55S.......mitochondrial ribosome monosome


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


Bernard Clark Courtney

December 1992

Chairman: Dr. Thomas W. O'Brien
Major Department: Biochemistry and Molecular Biology

The translation system of the mitochondrion has many

unusual characteristics when compared with other translation

systems. Among them are the unique properties of the mRNAS,

which lack the 5' cap found in eukaryotic messages and the

Shine-Dalgarno sequence of the prokaryotic system. The

mitochondrial (mt) ribosome, yet another unique element to

this system, must recognize, bind, and translate these

atypical mRNAs into proteins. The bovine system was chosen as

the model for this research because it is the source of the

majority of information on mt ribosomes, and because of the

availability of bovine liver for the preparation of mt

ribosomes and DNA to be cloned into transcription vectors for

the production of mRNAs. The mRNA for COII was chosen as a

model study because of its simple monocistronic nature and its

bovine mitochondrial origin.


The small (28S) subunit of mitochondrial ribosomes binds

the mRNAs tested with unit stoichiometry, one message per 28S

subunit and one 28S subunit per message and with a

dissociation constant of approximately 5x10-M.

The secondary structure of this message was initially

predicted using the computer algorithm and subsequently probed

with dimethyl sulfate and ribonucleases A, T1, and V, to refine

the computer generated model. The use of the same RNA

cleavers and modifiers in the presence of 28S subunits

provided "footprints" on the COII mRNA of their interaction

site(s). A region of multiple cleavage reductions (7-13

nuclease susceptible bases) was found from base 288-314 with

additional TI reductions observed elsewhere in the molecule

(guanosines 23, 24, 92, 195, 205, 209, 356, 410, 440, 453, and

457), possibly effects peripheral to the actual binding. Ten

sites of enhanced sensitivity were also seen for RNase A

throughout the molecule, but particularly between 210 and 365

in the presence of ribosomes. Some of these may indicate

areas in the RNA that have been made more available for

cleavage as a result of the interactions between the 28S

subunit and the COII mRNA.

Thus, a high affinity interaction between COII mRNA and

the small ribosomal subunit has been characterized. The COII

mRNA displays a potential 30+ nucleotide binding site

consistent with the previously described binding site on the

28S subunit.


The process by which proteins are synthesized is a three

tiered pathway that was predicted by James Watson, as simply

"DNA makes RNA makes protein." This along with "the assertion

that information passed to protein cannot return to its former

state" (i.e., proteins do not carry information for their own

biosynthesis), (T. Hunt et al., 1983), form the bases for the

Central Dogma. While this wonderfully simplifies the protein

biosynthetic process, the individual steps in the process are

rife with complexities which are inconsistent with the term

dogma. The diversity between the prokaryotic and eukaryotic

systems in the components and the methods used to complete the

protein synthetic process are differences that have and will

continue to be at the center of much investigative effort.

The first step in the process is the transcription of DNA

into RNA, ribosomal, transfer, and messenger RNAs. Ribosomal

RNAs (rRNAs) are the functional skeleton around which

ribosomal proteins assemble to form the two subunit

translation organelle, the ribosome. Three (or four for

eukaryotes) rRNAs are used in this ribonucleoprotein complex

and, in the case of prokaryotic ribosomes, these rRNA

molecules participate more directly in the translation


process. The single strand nature of these molecules and

their inherent base pairing potential allow these, as well as

any RNA molecule, to form secondary and tertiary base pairing

interactions (Gutell et al., 1985).

In rRNAs the secondary structures are extensive and

fairly well conserved, which provides for a stable long-lived

molecule upon which to build a ribosome. Transfer RNAs

(tRNAs) are small (70-130 bases) highly structured (secondary

and tertiary) RNAs that serve a dual function in the

translation process. They decode the information contained in

messenger RNAs (mRNAs) with a three base anticodon sequence

which specifies a particular amino acid and complements a

codon sequence on the mRNA. They also deliver that particular

amino acid to the ribosome for addition in polypeptide chain

elongation. These RNAs have been used extensively as models

for the development of computer algorithms for predicting

secondary structure in RNA molecules. The mRNAs are carriers

of the genetically encoded information for the synthesis of

proteins from its origin in the genome (DNA) to its site of

translation, the ribosome.

Messenger RNAs may be quite complex in their primary and

secondary structures. The primary structure can be

complicated by having its coding regions, exons, separated by

intervening non-coding regions, introns, by having a 5' cap,

or by having 5' and 3' untranslated regions. The exons must

then be spliced together and the introns removed prior to


translation of the message into protein. These RNAs are also

capable of forming secondary and tertiary structures. In

several cases, the secondary structure of mRNAs has been shown

to be important in the regulation of their translation, either

providing a binding site for a regulating element or

preventing or enhancing the ability of the ribosome to bind,

move on the mRNA, or initiate protein synthesis (Kozak, 1990;

Haile et al., 1989).

The final step, translation, can be subdivided into three

phases: initiation, elongation, and termination. While the

product is the same regardless of which system is being

observed, prokaryotic or eukaryotic, and the central

characters maintain the same roles, the cast of supporting

players can diverge considerably. In prokaryotic systems

initiation requires mRNA, IF1, IF2, IF3, fMet-tRNA1, GTP, and

both ribosomal subunits, whereas eukaryotic systems usually

operate with 5' Capped mRNA, eIFl, eIF2, eIF3, eIF4F, eIF5,

Met-tRNAiMet, GTP, ATP, and both ribosomal subunits (see

Figure 1, Darnell et al., 1990). This situation also reveals

that initiation mechanisms differ between prokaryotes and

eukaryotes. Each system assembles its components and factors

so that the end result is an initiation complex comprised of

a ribosome with a charged tRNA, at the peptidyl-tRNA site and

an mRNA in position for decoding. In all stages of

translation, the ribosome remains the central component of all

biosynthetic activities in both systems, prokaryotic and


30S or 405)
nhltaison comoie

-7-7 ------ -I



rnR_ NA

oEr ( CE,? "0
e eI,,(eP, (.,P1 eP "

Methion ne


-e-tfNA'A le9.vered
o oosoe aria
0$ ho-ea

:omo0ete ana

svntness aina

switc, trom A 'o0 se,
:ataiviea Ov
e-ongat'on f.acor 3
*n orGOParvoes-




Figure 1. The Three Stages of Translation (Darnell et al.,

1990). Translation is divided into three stages, initiation,

elongation, and termination. Each stage has its own

particular requirements for substrates and factors that differ

depending on the type of ribosome.

; |Pe o ted e ^

Figure 1 (cont.)


eukaryotic. Both systems possess ribosomes, albeit with

unique characteristics.

A third and physically unique kind of ribosome was

reported in 1967, the 55S mitochondrial ribosome (O'Brien and

Kalf, 1967; O'Brien, 1971). For years skeptics remained

unconvinced of this finding, considering the 55S "mini

ribosome" particle to be a bacterial contaminant or residual

cytoplasmic large subunit contaminant. As the data

substantiating its existence grew with time, the 55S ribosome

eventually won acceptance as part of a semi-autonomous

mitochondrial genetic system and represents a unique class of

ribosome. Subsequent investigations showed the factors

participating in the bacterial protein synthetic process are

not always interchangeable with those normally used by the 55S

ribosome (Denslow and O'Brien, 1979) indicating that the 55S

ribosomes had properties different from bacterial ribosomes.

Since that time, the sequences of many of the mammalian

mitochondrial genomes have been published (Anderson et al.,

1981, 1982; Bibb et al., 1981). These DNAs encode rRNAs,

tRNAs, and ,in the case of mammals, 13 open reading frames,

two of which are bicistronic (Chomyn et al., 1985). Some

mitochondria specific factors have been isolated (Liao and

Spremulli, 1990a), and even a few of the ribosomal protein

genes have been sequenced (Liu and O'Brien, personal

communication). Twenty-five years later, the inquiry into the

55S ribosome and the mitochondrial translation system has left


many of these unique aspects an enigma, as 80S and 70S

ribosomes continue as the preferred models in the study of

translational systems.

This notwithstanding, the mitochondrial translation

system demands elucidation by virtue of its various unique

properties (see Table 1). The mitochondrial ribosome is

characterized by a low sedimentation rate of 55S, imparted by

its relatively high protein content (85 proteins total) and a

larger mean size of ribosomal proteins by comparison with

prokaryotic ribosomes (53 proteins total)(Matthews et al.,

1982). 55S ribosomes are assembled around a 12S rRNA and a

16S rRNA (corresponding to the 16S rRNA and 23S rRNA for

prokaryotic ribosomes). Current speculation suggests that the

"extra" proteins found in 55S ribosomes may fill the voids

created by the shorter mitochondrial rRNAs (O'Brien et al.,


Mitochondrial mRNAs are remarkably different from those

of the other translation systems. Mitochondrial mRNAs, with

the exception of the message for nicotinamide adenine

dinucleotide (reduced) dehydrogenase subunit 6 (ND6), are

transcribed as a large (16+ kilobase), continuous RNA

representing the entire mitochondrial genome. ND6 and a few

of the tRNAs are transcribed in a similar manner from the

light strand. These transcripts have the unusual properties of

being intronless and having their open reading frames

punctuated by tRNAs. The current theory is that the tRNAs

Table 1

Unique Properties of Mammalian Mitochondrial Translational
Components (O'Brien, 1971, O'Brien et al., 1980, Gaines and
Attardi, 1984, and Denslow et al., 1991).

1. Sedimentation rate of 55S

2. High Protein Content

3. Low rRNA content

4. No 5' methyl cap or leader sequences

5. No Shine-Dalgarno Sequence

6. Unique (and possibly resident) factor requirements

7. Unique genetic code and tRNAs

8. Mitochondrial ribosomes are products of two genomes

9. All products of mitochondrial translation are

residents of the inner mitochondrial membrane

10. Mitochondrial ribosomes bind GTP directly

11. Mitochondrial mRNAs contain no introns

12. Mitochondrial mRNAs have neither 5' nor 3'

untranslated regions

13. Mitochondrial mRNAs are transcribed as a large 16kb

RNA that is processed into rRNAs, tRNAs, and mRNAs.


provide recognition sites for the nearly immediate processing

of the large transcript RNAs into their constituents of 11

mRNAs, 22 tRNAs, two rRNAs, and a 7S RNA of unknown function

(Ojala et al., 1981). The mRNAs have neither 5' methyl-G caps

nor S-D leader sequences (Shine and Dalgarno, 1974), nor does

the 12S rRNA possess an anti-S-D sequence. In fact,

mitochondrial mRNAs possess little or no 5' leader sequences

(see Figure 2) and no 3' untranslated sequence (Montoya et

al., 1981). Since 5' caps and leader sequences are both

absent, mitochondrial translational initiation must take place

through a novel and yet to be defined mechanism. Speculation

has suggested the presence of an internal S-D like sequence

(Saccone et al., 1985), while others speculate that the 5'

ends of the mRNAs are important (Denslow et al., 1989). In

any event the function of facilitating the localization of the

initiation codon in the decoding site, provided by the S-D

site on the 5' ends of prokaryotic mRNAs must be the function

of a separate and distinctive factor particular to the

mitochondrial translation system. Perhaps the mitochondria

possess a protein similar to Sl, which maintains the message

on the small subunit of E. coli until the S-D sequence can

function and the initiation complex can form (Calogero et al.,

1988). Alternately, the mRNAs may provide for "internal

ribosome entry sites" as those found for poliovirus mRNAs in

their 5' untranslated terminal repeat. Simply, it may be that

the binding affinity is of sufficient strength to allow for

ATPase 8






















Figure 2. 5' Ends of Bovine Mitochondrial mRNAs. Note the
extremely short, or non existent, 5' leader sequences (Montoya
et al., 1981; Hill et al., 1990). The +1 indicates the first
base of the message and start codons are underlined.


the proper positioning of the message on the small subunits,

in the vicinity of the decoding site as predicted by

experiments on prokaryotic messages with or without S-D

sequences (de Smit and Van Duin, 1990). The 55S ribosome may

recognize a particular secondary structure in the mRNA near

the 5' end, though no experimental evidence for the theory

exists. The effect of the 5' end structure, whether 5'-OH,

monophosphate or triphosphate, has been found to play no role

in the actual binding of the COII mRNA by the 28S subunit

(Liao and Spremulli, 1990a). The whole story is likely to be

more complicated, since the mitochondrion translates two

messages which are bicistronic and out of frame: ATPase 8/6

and ND 4L/4.

One possibility for the mechanism by which the

mitochondrial ribosome initiates translation is the presence

of a sequence nonspecific RNA binding domain on the

mitochondrial ribosome (Denslow et at., 1989). The roughly

thirty base long binding domain of this binding site (see

Figure 3) was disclosed by oligoribonucleotide binding studies

to determine the affinity of polynucleotides of increasing

lengths from 3-42 bases. Furthermore, an RNase protection

domain of up to eighty bases in length was provided on bound

RNA (5' end labelled). Current thinking suggests this domain

may aid in positioning the mRNA on the mitochondrial ribosome

for the start of translation and may stabilize the mRNA during

elongation. To examine the processes of initiation and



- 30b


Figure 3. The RNA Landing Pad on Mitochondrial Ribosomes.
Messenger RNA is thought to bind to a domain on the small
subunit roughly 30 bases in length. An 80 base domain of
protection from nuclease activity is provided by the subunit
(Denslow et al., 1989).




elongation, a few mitochondrial translation factors have been

isolated (Liao and Spremulli, 1990a), although the majority of

factors probably remain to be found. The peculiarly large

number of ribosomal proteins in the 55S ribosome allows for

the possibility of "resident" factors, factors that are in

constant association with the ribosome.

In addition to a unique initiation mechanism, the 55S

ribosome translates mRNA using a unique genetic code and

unique mitochondrially encoded tRNAs. This ribosome utilizes

fewer tRNAs (22) than its bacterial counterparts (61).

Interestingly, the search for a mitochondrial initiator tRNA

has been a difficult one and to this point fruitless. While

the mitochondrial genome encodes a methionyl-tRNA, no evidence

of its ability to be processed (formylated) into an initiator

tRNA has been found. Initiator tRNA obviously exists and the

possibility of its import into the mitochondria has been

considered. Keeping in mind the known properties of the

mitochondrial system, it seems likely that its elongation

mechanism may also prove unique.

Another unusual property of mitochondrial ribosomes is

their assembly from two different genomes. The proteins are

encoded by the nuclear genome, translated on 80S ribosomes,

and then imported into the mitochondria (Schieber and O'Brien,

1982) where they are assembled with the rRNAs, transcribed

from the mitochondrial genome, to form 55S ribosomes. This

mechanism of ribosome biosynthesis raises a number of


questions unique to the mitochondrial translation system: What

are the control mechanisms for the maintenance of

stoichiometric levels of ribosomal proteins to rRNAs and what

is the method of communication between the mitochondrial

machinery and that of the cytosol and the nucleus?

Another distinctive property of 55S ribosomes is that

they bind GTP directly (Denslow et al., 1991). Figure 1 shows

that GTP is involved in a number of steps in the translation

mechanism of prokaryotes and eukaryotes, but in no step does

the GTP bind to the ribosome directly. Experimentation has

shown that GTP binds to the 55S ribosome tightly (Kd = 20 Nm)

and exchanges rapidly with GDP, which has an apparently equal

affinity for the ribosome. There is, however, no evidence yet

for a functional role for GTP or GDP binding, but the binding

of GTP directly to the ribosome provides circumstantial

evidence for the presence of a resident factor. Understanding

the GTP binding function evokes still another path for further


Although the focal points of study on translation have

revolved around the eukaryotic and prokaryotic systems, the

many unique properties of the mitochondrial translation

system, outlined in Table 1, also demand attention. This is

a fundamental work inasmuch as the 55S ribosome and some

initiation and elongation factors are the only members of the

mitochondrial protein biosynthetic mechanism to have been

described. This rationale was the basis for earlier research


on the RNA molecule of the small subunit by W. Faunce, and

ongoing work by R. Heck on the RNA interaction site on the 28S

subunit, and J. Liu on the function of GTP binding. So one

might logically be directed to the "first step," initiation,

which has as an early step, the binding of mRNA by the 28S

subunit. While the ribosome binds the mRNA, the message is

being bound and these are the points of interest of the

research to be described here. The work done on the

characteristics of mRNAs/ribosome interactions in the

prokaryotic and eukaryotic systems may provide clues to the

important features to expect in the mitochondrial system. How

can this interaction between the 28S subunit and the mRNA be


Essential questions to be addressed in this research are

as follow: Is the binding of mitochondrial mRNAs in any way

specific? Does the 28S subunit bind all mRNAs similarly? Does

it bind all polynucleotides equally? What is the binding

stoichiometry? Where is(are) the ribosome binding site(s) on

mRNA? What are the characteristics of the mRNA where it is

bound by the ribosome?

Binding studies can be employed using a number of

polynucleotides to determine the extent of specificity, the

strength of the interaction, and its stoichiometry. An often

used filter binding assay provides a quantitative yet quick

and inexpensive mode for determining the strength of

interaction and specificity of a number of ligands,to include


mitochondrial mRNAs for cytochrome oxidase subunit 2 (a

monocistronic mRNA) and ATP synthase subunits 8 and 6 (a

bicistronic mRNA). The binding stoichiometry of 28S subunits

on mRNA can be analyzed by centrifugation in sucrose density

gradients, the method used in the isolation of 28S subunits


Some insights into the sequence/structural properties of

the mRNA at the site of interaction with the 28S subunit may

be revealed in the specificity studies, comparing the

properties of polynucleotides that bind to those that do not.

Additionally, the development of computer predictions of the

secondary structure of a 5' leaderless model mRNA, COII

(Figure 2), and one that has been used in previous

mitochondrial initiation studies may provide some clues to the

putative site of interaction. It would be naive to accept the

results of an RNA folding program carte blanche, especially

when the majority of folding programs is based on the

structures of tRNAs.

Therefore, the mRNA must be probed in some manner to

confirm or refine the computer prediction. The techniques

most frequently used to study the tertiary structure of

macromolecules are X-ray crystallography and electron

microscopy. X-ray crystallography is exceedingly time

consuming, expensive and has yet to be successfully applied to

RNA structures. Electron microscopy is limited in its ability

to disclose structural details in molecules like large RNAs.


The method which most efficiently discloses the secondary

structure of an RNA is the modification of the RNA followed by

primer extension termination (Moazed and Noller, 1986; Stern

et al., 1988). This primer extension technique involves

modifying the COII mRNA with either enzymes or chemical

agents. The modified mRNA is then purified and annealed to

one of several synthetic oligodeoxyribonucleotide primers

complementary to regions spaced at intervals of approximately

150 bases. The annealed primers are then extended using

reverse transcriptase (Boorstein, 1989; Knapp, 1989). The

extension is disrupted at the modified bases, providing a

means of disclosing the sites of modification. Finally, the

primer extension products are analyzed on a denaturing gel to

identify the modified bases.

The final question is answered using the same primer

extension technique, but in the presence of 28S subunits.

This technique gives a "footprint" on the mRNA by first

allowing the 28S subunit to bind the message, followed by

chemical modification or enzymatic cleavage. Then, as before,

primers are annealed and extended with reverse transcriptase.

The interruptions to extension are again disclosed by

denaturing gels and the difference in modifications/cleavage

in the presence and absence of 28S subunits is analyzed to

disclose the "footprint" (sites of blocked



In this study, mitochondrial mRNA was analyzed using a

chemical modifier, dimethyl sulfate (DMS) (Donis-Keller et

al., 1977), which reacts primarily with unpaired adenines and

cytosines, and either paired or unpaired guanines (see Figure

4)(guanines are disclosed as modified by chemically cleaving

the purine ring, a technique not used in this work). In

addition, the COII mRNA was probed using RNase A (which

cleaves 3' to pyrimidines in single-stranded RNA), RNase T,

(which cleaves 3' to unpaired guanines), and RNase V, (which

cleaves primarily double stranded RNA)(Gehrke et al., 1983).

Chemical as well as enzymatic modifiers were chosen to allow

for an additional level of analysis. First, the secondary

structure of the mRNA could be investigated by the use of

single-strand specific modifiers (RNases A and T1, and DMS) or

a double strand specific modifier (RNase V1). Secondly, the

protection from modification/cleavage in the presence of 28S

subunits provides a "footprint" of the binding site. Finally,

the disposition of the mRNA within in the binding site of the

28S subunit and its own secondary/tertiary structure can be

studied. The relatively bulky enzymatic modifiers (the

RNases) may only be effective on ribonucleotides exposed on

the surface or not in association with the small subunit. The

smaller chemical, however, should penetrate through the

secondary/tertiary structure of the 28S subunit to methylate

most pyrimidines not involved in Watson-Crick base pairing or

tertiary interactions.

\ N H. 0 O
N 4 S
sN H\ H-N3

/ N 4" 0 N


O 4*

/ N, H..''


Figure 4. Site of Chemical Modifications on Ribonucleotides.
Dots represent the sites of modification by DMS. All
modifications are blocked by Watson-Crick base-pairing except
the modification of guanine.


With these tools in hand, some of the questions

pertaining to the initiation of translation within the

mitochondrion can be approached. Does the binding of bovine

mitochodrial mRNAs require the presence of additional factors?

Is the binding strength significant for physiological

function? What is the stoichiometry of binding? What

specific attributes are necessary on messages for binding by

28S subunits? What is the nature of the mRNA at its site(s)

of interaction with the small subunit? These are the

questions to be answered in this research. The results and

the techniques used can be successfully employed in future

studies concerned with initiation of translation in the bovine

mitochondrial system.


Preparation of Bovine Mitochondria

Bovine mitochondria were isolated from fresh liver (less

than 30 minutes post mortem and provided for the most part by

the University of Florida Meat Lab) which was sliced thinly

(approximately 1.0 cm thickness) on location and submerged in

isolation medium (0.34 M sucrose, 1 mM EDTA, and 5 mM Tris-

HC1, pH 7.5 at 0-4C), as described previously (Matthews et

al., 1982). The tissue and subsequent mitochondria were

maintained in this buffer at this temperature throughout the

preparation. The sliced liver was ground coarsely in a

commercial meat grinder and then homogenized in a flow-through

homogenizer (Tekmar).

Mitochondria were resuspended to a concentration of 20 mg

protein/ml in isolation medium, treated with 100 gg/ml

digitonin, final concentration, and stirred for 15 minutes.

The digitonin-treated mitochondria (mitoplasts: mitochondria

lacking the outer membrane) were then diluted 5-fold in

isolation medium and collected by centrifugation at 11,000 x

g for 10 minutes. Mitoplasts were washed four additional

times in isolation medium and stored at -70C in freeze buffer


(40mM KC1, 15mM MgCl2, 10mM Tris-HCl, 5mM 2-mercaptoethanol,

0.05mM EDTA, 0.05mM Spermine, 0.05mM Spermidine, pH 7.5).

Preparation of Native 28S Subunits and 55S Monosomes

Mitochondrial ribosomal native 28S subunits were prepared

according to the method described by Matthews et al., 1982.

Mitochondrial ribosomes were isolated from the mitochondrial

lysate and separated in a 10-30% sucrose density gradient in

20mM MgC12, 100mM KC1, 5mM 2-mercaptoethanol, and 10mM

triethanolamine, pH 7.5. The gradients were monitored (UV) and

fractions were collected (Figure 5). The fractions containing

55S monosomes, and 28S and 39S subunits were pooled separately

and the particles were concentrated by ultracentrifugation,

45,000 rpm for 18 hours in a Beckman Ti55.2 rotor. The

subunit pellets were resuspended subunit freeze buffer (25mM

KC1, 2.5mM MgCl2, 10mM TEA and 5mM 2-mercaptoethanol, pH 7.5)

and monosomes were stored in monosome freeze buffer (25mM KC1,

5mM MgCl2, 10mM TEA and 5mM 2-mercaptoethanol, pH 7.5) and

stored at -70'C until needed.

Preparation of Derived 28S Subunits

The 55S monosomes were dissociated on 10-30% sucrose

density gradients in 5mM MgCl2, 300mM KC1, 5mM 2-

A 260 8 3 9 55S


28S 39S


Figure 5. Absorbance Profiles (260nm) of Ribosomes in Sucrose
Density Gradients. The upper gradient renders "native"
subunits and the lower gradient "derived" subunits. Note the
separation between each subunit and the monosome.


mercaptoethanol, and 10mM tris-HCl, pH 7.5. The derived 28S

subunit fractions were then pooled and the particles were

concentrated by high speed centrifugation, 45,000 rpm for 18

hours in a Beckman Ti55.2 rotor, resuspended in subunit freeze

buffer and stored at -70C until used.

Preparation of Mitochondrial DNA

Mitochondria (4 grams protein) were concentrated by

centrifugation at 11,000 x g for 10 minutes and resuspended in

STE buffer (100mM NaCi, 50mM Tris-Cl, pH 8.0, and 10mM EDTA).

These mitochondria were then lysed by the addition of SDS to

1% (w/v) final concentration. The mitochondrial lysate was

then subjected to cesium chloride (CsCl) gradient

centrifugation at 50,000 rpm in a Beckman TY65 rotor for 24

hours in the presence of ethidium bromide (EtBr). Two UV

luminescent bands are discernible at the completion of

centrifugation. The topmost and thinner band (containing

mitochondrial DNA) was removed from the gradient and the DNA

was ethanol precipitated by adding 2.5 volumes of 95% ethanol

and kept at -700C for at least one hour. The precipitate was

pelleted, washed four times with 70% EtOH, dried, and

resuspended in TE buffer (10mM Tris-Cl, pH 8.0, and 0.1mM

EDTA) to a concentration of 7-10jg/ml, with 30Ag total yield,

and stored at -20C.


Cloning Mitochondrial Genes

The mitochondrial DNA was digested by HindIII restriction

endonuclease to produce three fragments: 10.2, 4.4, and 1.7kb.

These fragments were cloned individually into a pUC 19 vector

linearized by HindIII restriction endonuclease for

amplification (as described Current Protocols in Molecular

Biology), storage, and mutagenesis. Clone 38 (Figure 6)

contained the 4.4kb fragment, which includes the gene for

ATPase 8/6. This clone was used to prepare the transcription

vector for subsequent mRNA production. In order to make an

mRNA that would most closely approximate the known mRNA

sequence, oligonucleotide-directed mutagenesis was utilized

(Kramer et al., 1984). Oligonucleotides (oligos)(Table 2)

were made by the DNA Core Facility, ICBR, to anneal to the 5'

and 3' ends of ATPase 8/6 gene in opposing orientation (Figure

7) to utilize the polymerase chain reaction. These

oligonucleotides were made up of 18 bases on their 3' ends

which complemented the 5' or 3' end of the gene exactly. Each

contained recognition sequences for a number of restriction

endonucleases (see figure 7) for cloning and, later, for

transcription purposes. The products of the PCR were

incubated with BamHI and EcoRI restriction endonucleases and

isolated by gel electrophoresis on a 1.5% low melting

temperature agarose. After EtBr treatment (0.1lg/ml) the UV

luminescent band corresponding to approximately 900 base pairs

7722/448 H-


Hindill H


- P

Figure 6. Map of Clone 38 of ATPase 8/6 for Mutagenesis.
pUC19 vector with a 4.4kb HindIII fragment from the bovine
mitochondrial circular DNA (>16kb).

Table 2

Oligonucleotides and their Uses.


TO 27

TO 28

TO 136

TO 135

TO 134

TO 133

TO 132

TO 139

TO 105


Mutagenesis of ATPase 8/6

Mutagenesis of ATPase 8/6

Primer Extension n of COII mRNA

Primer Extension a of COII mRNA

Primer Extension / of COII mRNA

Primer Extension of COII mRNA

Primer Extension 6 of COII mRNA

Binding Substrate

Binding Substrate

0 r- r-4 H

( l t .4

O ) -

u in


ZW4 .

I- jdumo..

I- m4Z-

Do- Mwzz
-C > -C

in- tz<


was removed from the gel. The transcription vector pCKSP6 (a

gift of Dr. C.W. Wu, State Univ. of New York at Stony Brook)

(lg), linearized by the same two restriction endonucleases,

was purified from its polylinker fragment by EtOH

precipitation in 2.5M NH4OAc at -70*C and two subsequent

washes with 70% EtOH. This vector was then resuspended and

added to the gel piece containing the digested PCR product.

The solution was brought to a final volume of 50A1 with ligase

buffer (40mM Tris-HCl, 10mM MgCl2, and 1mM DTT) and warmed to

65C until the agarose melted and then cooled to 37*C. T4 DNA

Ligase (10U) was then added and the mixture was incubated for

three hours at 37*C. This product was transfected into E.

coli (TG-1 and JM109), amplified, and stored as 50% glycerol

stocks at -70*C for further use.

Engineering an ATPase 8/6 Transcription Vector

The pCKSP6 vector containing the ATPase 8/6 gene required

further manipulation before it could be used in the prescribed

manner to generate an mRNA transcript. This was accomplished

by linearizing the circular DNA with BamHI restriction

endonuclease and removing the four base 5' overhanging ends

with Sl nuclease, a 5'-3' single strand specific exonuclease

(Figure 8). The clone was then re-circularized by T4 DNA

Ligase, transfected into E. coli strain TG-1, amplified, and

selected for by loss of the BamHI restriction endonuclease




GTACG A'rPse 8/6







Figure 8. Engineering the ATPase 8/6 Clone. A) Linearizing
with BamHI. B) Excision of 5' end overhangs and ligation of
the blunt ends (re-circularizing the vector). C) pJC213 with
new SphI restriction site at the 5' end of the gene.


site and the addition of an SphI restriction endonuclease

site. The clone, pJC213, was later sequenced using Thermus

aquaticus (TAQ) polymerase and dideoxynucleotide triphosphates

(as described in PCR Protocols, Innis et al., 1990) for

verification. Clones that were positive for these two traits

were then sequenced with TAQ DNA polymerase for confirmation.

pJC213 was the clone selected for further use in RNA


Preparation of Transcription Vectors

Transcription vectors pJC213 (ATPase 8/6) and p2-6E

(COII)(a gift of Dr. Philip Laipis) in bacterial strains

TG-1 and JM109, respectively, were grown to log phase in Luria

Broth (10gm NaCIl, 10gm tryptone broth, and 5gm Yeast Extract

per liter) pH 7.8, 0.5% glucose (w/v), and ampicillin

(100Ag/ml) at 370C with shaking (6-8 hours). Chloramphenicol

(20mg/liter) was then added and the incubation continued

overnight. Plasmids were harvested using the Alkaline lysis

method (described in Current Protocols in Molecular Biology,

Ausubel, et al., 1987). The dried, recovered plasmid pellet

was resuspended to img/ml in doubly distilled, deionized water

that was treated with diethylpyrocarbonate (DEPC). All water

used in this and subsequent experiments was similarly


SP6 Promoter


SP6 Promoter




Figure 9. Transcription clones pJC213 (A8/6) and p2-6E
(CoIl). A) pJC213 for ATPase 8/6. B) p2-6E for COII (a gift of
Dr. P. Laipis).


The pJC213 was then linearized with NsiI restriction

endonuclease and p2-6E was linearized with SspI restriction

endonuclease (Figure 9). Both vectors then were ethanol

precipitated, pelleted, washed with 70% EtOH, dried, and

resuspended to 0.2Ag/lLl in water and stored at 4C until


Preparation of mRNA Transcripts

ATPase 8/6, Cytochrome Oxidase II, ATPase subunit P (the

vector was the gift of Dr. A. Lewin), and poliovirus (the

vector was the gift of Dr. Flanegan) mRNAs were synthesized by

incubating 2Ag of the templates for the genes mentioned above

with 20-22.5 U SP6 RNA polymerase in 40mM Tris-HCl, 6mM MgCl2,

2mM Spermidine-(HCl)3, 10mM DTT, 100U RNAsin, 0.5mM ATP, 0.5mM

CTP, 0.5mM GTP, and 0.5mM UTP, pH 7.9, for two hours at 37C

(Krieg and Melton, 1984). The synthesis reaction was

incubated with 5 units of RNase-free DNaseI (RQ-1, Promega)

for 30 minutes at 37C. The resulting transcription products

were analyzed by 1.5% agarose gel electrophoresis for reaction

success, transcript size, and purity. Products were kept at

0OC and used within two hours.

Preparation of Radiolabelled Synthetic RNA

Labelled ATPase 8/6, COII, ATPase P subunit, and

poliovirus mRNAs were synthesized by incubating 2Ag of


appropriate template with 20-22.5 U SP6 RNA polymerase and

200/Ci [a-32P]CTP in 40mM Tris-HCl, 6mM MgCl2, 2mM Spermidine-

(HC1)3, 10mM DTT, 50U RNAsin, 0.4mM ATP, 0.4mM CTP, 0.4mM GTP,

and 0.4mM UTP, pH 7.9 for two hours at 370C. The synthesis

reaction was incubated with 5 units of RNase-free DNaseI for

30 minutes at 37C. The transcription reaction was then run

through Sephadex-50 RNase-free spin columns twice (Boehringer-

Mannheim) equilibrated with 50/5 buffer (40mM KC1, 20mM HEPES-

Cl, pH 7.3, 5mM MgOAc, and 6mM 2-mercaptoethanol) to remove

free nucleotide (label) from solution. The transcripts were

then quantified by their incorporation of radionucleotide with

specific activities usually between 5-7xl03CPM/pmoles. The

resulting transcription products were analyzed by 1.5% agarose

gel electrophoresis for success of the reaction, transcript

size, and purity. Subsequent autoradiography of the gel

confirmed that the removal of the labelled free-nucleotide had

been removed. Products were kept at 0-4*C for use within two


Preparation of 5' End Labelled Nucleic Acids

E. coli transfer RNA (tRNA)(Sigma), double stranded

(dsDNA)(Promega), single stranded DNA (ssDNA)(TO 139),

polyuridylic acid, poly(U), and polycytidylic acid, poly(C),

(Sigma) were all labelled at the 5' end. Poly(U), poly(C) and

ssDNA were labelled by incubating lOOpmoles of the appropriate


template with 100pmoles of [7-32P]ATP and 10U of T4

polynucleotide kinase in 50mM Tris-Cl, pH 7.5, 10mM MgCl2, 5mM

DTT, and 0.1mM Spermidine for 20 minutes at 370C. The other

two templates were labelled using the Exchange Reaction

(Perbal, 1988). Again 100pmoles of both template and [(-

32P]ATP were incubated with 12U of T4 polynucleotide kinase and

0.25mM ADP in 50mM Immidazole-Cl, pH 7.5, 10mM MgCl2, 5mM DTT,

and 0.1mM Spermidine for 30 minutes at 37C. Each reaction

was spun through three separate sephadex G-50 spin columns to

remove residual labelled nucleotide. Purity was tested by

autoradiography after agarose gel electrophoresis. Greater

than 95% of the free nucleotide was removed by each pass

through the sephadex so that no residual nucleotide was

detectable after the third column.

Binding Substrates to 28S Subunits

A number of substrates were bound to small subunits by

incubating 1.0, 2.0, and 5.0pmoles (0.02, 0.04, or 0.1pM,

respectively) 28S subunits with various concentrations of

radiolabelled substrates from 0.02-1.0OM (1-50 pmoles) in 504l

of 50/5 buffer for 10 minutes at 35C (Denslow et al., 1989;

Liao and Spremulli, 1990b). The reactions were stopped by the

addition of 800l of ice cold 50/5 buffer.

Similarly, 9 pmoles (0.18AM) of radiolabelled COII or

ATPase 8/6 mRNA was incubated with 4.5-90 pmoles (0.09-1.8MM)


small subunits (native and derived) in 50/5 buffer for 10

minutes at 35*C in a reaction volume of 50pl. The reactions

were stopped by the addition of 800l of ice cold 50/5 buffer.

Competition studies were carried out by mixing a

saturating amount of radiolabelled COII mRNA with varying

amounts of competitor prior to incubation with 1.0 or 5.0

pmoles (0.02 or 0.1AM) of 28S subunits in 50/5 buffer for 10

minutes at 35C in a reaction volume of 50Al. The reactions

were stopped by the addition of 8004l of ice cold 50/5 buffer.

Millipore Filter Binding Assay

After binding the experimental substrate(s), the stopped

reaction mixture was mixed gently on a vortex mixer and the

ribosomes were adsorbed to wettable 0.45Mm HAWP type millipore

filters. The filters were previously rinsed in 5 ml of 50/5

buffer to remove all glycerol. The tubes were rinsed and

vortexed with 8004l of ice cold 50/5 buffer, which was then

applied to the filters and the filters were washed twice with

8-10 ml of ice cold 50/5 buffer. Washed filters were immersed

in scintillant (ScintiVerse II, Sigma) and counted on a

Beckman LS 380 scintillation counter (Denslow et al., 1991).

Sucrose Density Gradient Centrifugation

Sucrose density gradients were employed to determine the

stoichiometry of 28S subunit/mRNA binding interactions. After


binding reactions were completed they were loaded onto 10-30%

sucrose density gradients in 50/5 buffer. These gradients

were formed in SW 27 cellulose nitrate wettablee) tubes using

a Beckman gradient former. Loading of the stopped reactions

was accomplished by layering up to 2 ml on top of the

gradients. Tubes were placed in an SW 27 rotor and

centrifuged for 18 hours at 23,000 rpm at 4*C.

The gradients were then monitored and fractionated using

a Pharmacia FPLC System, UV detector, chart recorder, and

fraction collector. To the collected fractions (2 ml) 2 ml of

ScintiVerse II (Sigma) scintillation cocktail was added and

the fractions were then counted on a Beckman LS380

scintillation counter.

Chemical Modifications and Enzymatic Cleavages

COII mRNA (9-10 pmoles, 180-200nM) was diluted in 50/5

buffer to 50Al and modified at 35C for 10 minutes by the

addition of RNase A(10-2-104'U/Ug RNA), or RNase T, (10-2-10'

4U/Ag RNA), or RNase V, (0.45-10"3U/g RNA). DMS modifications

were carried out under the same conditions but with 1-2Al of

DMS being added, gently mixed, and incubated at 35*C for 0.5

to 2 minutes. Modifications and cleavages were done in the

presence and absence of 18-20pmoles (360-400nM) of 28S

subunits. The reactions were stopped by the addition of 150Al


of ice cold stop buffer (100mM NaAc, 3.3mM EDTA, and 120 mg/ml

yeast tRNA).

The enzymatic reactions were then extracted twice with

phenol that was equilibrated with HE (10mM HEPES-CI pH 7.3 and

ImM EDTA), and once with chloroform. The mRNA was

precipitated by addition of 2.5 volumes of cold absolute

ethanol followed by incubation at -70C for 60 minutes. The

mRNA was recovered by centrifugation at 15,000 rpm in a Hermle

microfuge, washed with 70% ethanol, and dried in a Savant

Speedvac. The mRNA was resuspended in water to 0.1 pM and

stored at -70C until needed.

DMS reactions were immediately precipitated by addition

of 2.5 volumes of cold absolute ethanol and kept at -70C for

60 minutes. The mRNA was recovered by centrifugation at

15,000 rpm in a Hermle microfuge, washed with 70% ethanol, and

dried in a Savant Speedvac. The mRNA was resuspended in 200l

HE and then extracted twice with phenol that was equilibrated

with HE, and once with chloroform. The RNA was recovered by

centrifugation, washed with 70% ethanol, and dried. The RNA

was resuspended in water to 22Mg/ml (0.1MM) and stored at

-70*C until needed.

All reactions were carried out in triplicate. In

addition, each nuclease and DMS were titrated so that

approximately one chemical modification or one RNase cleavage

would occur every 150 bases to provide optimal analysis

(Figure 10).

I 2345678910 S E Q
5 'e n d !







Figure 10. Titration of RNase A and DMS Modifications of Coll
mRNA (Primer extension n) Lane 1: Control- no RNAase A; lanes
2-6: Ixl02, 7.5, 5, 2.5, and lxl0- RNase A, respectively;
lanes 7 and 9: 1il of DMS incubated for 1 and 2 minutes,
respectively; lanes 8 and 10: 2Al of DMS incubated for 1 and
2 minutes, respectively; and lanes 11-13: Sequence (A,C,G).


DNA Primers

Five DNA primers were designed to be complementary to the

COII mRNA, each 17 nucleotides long (Table 2). Each sequence

was specific for a unique area on the mRNA (i.e., recognized

for annealing to a particular mRNA sequence and no others in

that mRNA or the 12S rRNA) and designed to be separated by

150-200 nucleotide intervals (Figure 11), about the distance

that can be easily analyzed on the denaturing gels used.

Particular care was taken so that the 3' end of the

oligonucleotides would give efficient priming without

promoting spurious annealing. The guanosine and cytosine

content of these oligonucleotides was also considered, to

maintain the G+C content at about 50%, if practical. The

oligonucleotide primers were synthesized using an Applied

Biosystems Model 380B synthesizer in the University of Florida

ICBR DNA Synthesis Core Facility.

Extension of Oliqodeoxyribonucleotide Primers

Extensions were initiated by incubating 25 pmoles (500nM)

of the appropriate primers with 1.0 pmoles of the mRNA (20nM)

in 50mM KC1, 9mM MgCl2, 20mM 2-mercaptoethanol, 50mM Tris-HCl,

25AM dATP, 250pM dCTP, 250AM dGTP, 250AM dTTP, and 10-15ACi

[a-32p]dATP (>600 Ci/mmol) in a final reaction volume of

o C
o n4 U
.4 CU




5 ^

I e




I i J

c 0n



0 -



K '0


cr T
rZ 4





4' c 0
0-6 cu



50Al at 65*C for 5 minutes. The mixture was then placed into

a 50*C water bath to allow annealing to occur. The primers

were then extended by adding 10 U AMV reverse transcriptase to

the reaction mix, and incubating at 50*C for 30 minutes.

Primer extension reactions were stopped by the addition

of 150il of ice cold reverse transcription stop solution

(100mM NaAc, 6.6mM EDTA, 0.l1g/ll carrier yeast tRNA, and 90%

EtOH). The cDNA reverse transcription products were then kept

at -70*C for 60 minutes, and the cDNA was recovered by

centrifugation at 15,000 rpm in a Hermle microfuge, washed

with 70% EtOH, dried, and resuspended in 204l loading solution

(583 mg/ml bromophenol blue, 583 mg/ml xylene cyanol, 8.3mM

EDTA, and 10.8M formamide. Samples were heated for 3 minutes

at 90*C and quenched on ice before loading onto gels.

Denaturing Gels

Denaturing gels (6% acrylamide, 0.3% bisacrylamide, 8 M

urea) were formed as 3:1 wedges (to afford better linear

separation) and were run at 500C. The gels were run in TBE

buffer (10mM Tris, 10mM Boric Acid, and 0.1mM EDTA, pH 7.8) at

80 watts for approximately 3 hours. These gels were then

adhered to Whatman 3MM chromatography paper and dried on an

Ephotec gel drier under vacuum at 80C for 20 minutes. The

dried gels were then placed in autoradiographic cassettes with

Kodak XAR X-ray film and exposed at least 15 hours.


Gel Reading Algorithm

In order to determine whether a given nucleotide was

modified, its autoradiographic band intensity was compared to

the same band region in the control gel. In addition, the

intensity of the experimental band was compared to background

bands within in the same lane (this method helped offset the

problems of non-uniform background from lane to lane and from

top to bottom on the gel). Thus, if the band for a given

nucleotide was darker than the control and greater in

intensity relative to the background within the lane, that

nucleotide was scored as accessible. Occasionally, a band

would be observed even though no modifier was added. This

occurrence could be explained by a number of phenomena,

including false priming of the carrier yeast tRNA by the

primers for COII mRNA, strong secondary structural effects of

the COII mRNA, false priming of 12S rRNA (when present) by the

primers to COII mRNA, and susceptibility to RNases not

purified from the isolated 28S subunits. The priming and

extension of yeast tRNA were tested, and no evidence of

extensions was apparent. The priming of 12S rRNA was also

tested with only a few faint bands being seen in a light

background smear. The bands were usually in between the

"normal" cDNA banding pattern. In addition, all data were

first recorded against the primary structure so as not to be

prejudiced by the predicted secondary structure.


Scanning Densitometry

Scanning densitometry was done on the General Imaging

Scanner of the Protein Core Facility, ICBR, to quantify the

data from the autoradiography of denaturing gels. This proved

particular useful in determining the extent of protection

afforded by small subunit interactions.


Significance of 28S Ribosomal Subunit Binding Properties

Bovine mitochondrial ribosomes were chosen as a model

system to address a number of questions unique to the

mammalian mitochondrial translation system. These ribosomes

contain nearly twice as much protein mass as the E. coli

ribosomes, roughly half as much RNA, and still retain similar

dimensions and appearance of the latter upon electron

micrographic inspection (Hamilton and O'Brien, 1974, Lake et

al., 1976). It is, however, the ability of these ribosomes to

bind the unusual mRNAs of the mitochondrion that is of

particular interest to this work. Where the interaction

occurs on the mRNA, and what the order of assembly and

requirements for additional factors might be remain to be

answered. At the outset it must be remembered that no in

vitro protein synthetic system is available for the

mitochondrion such as is commonly used in the other systems

(e.g., reticulocyte lysates for eukaryotic systems).

The mitochondrial mRNAs have little or no 5' leader

and no 3' untranslated sequence, though they are

polyadenylated. They are not capped at their 5' end (Gaines,


et al., 1988), like most eukaryotic mRNAs, nor do they possess

a S-D like sequence. They are produced as a single transcript

(>16 kb) of the entire heavy strand of the mitochondrial

genome, which is composed of 12 intronless open reading frames

(all coding for inner mitochondrial membrane proteins) and

punctuated by 22 tRNAs which serve as processing signals. The

light strand of the mitochondrial genome encodes seven tRNAs

and one protein product, ND6, that shares all other

mitochondrial mRNA traits, including being transcribed as part

of single, full length polycistron. Additionally, the rRNAs

are transcribed in a separate event that excludes all mRNAs

and they have been quantitated at 100 fold excess over mRNAs

in HeLa cell mitochondria (Gaines and Attardi, 1984) and at a

>10 fold excess over mRNAs in adult rat liver mitochondria

(Cantatore,et al.,1984).

The small subunit is capable of binding these mRNA

molecules through a novel and yet to be defined mechanism.

Speculation has suggested the presence of an internal S-D like

sequence may be important for the binding of some of these

mRNAs but not all (Saccone et al., 1985); yet this region on

the 12S rRNA proved inaccessible to probing in the subunit

(Faunce and O'Brien, manuscript in process). Therefore, it is

less likely to be able to form base pairs with the mRNAs.

Another speculation proposes that the 5' ends of the mRNAs may

be important (Denslow et al., 1989). The 28S ribosome may

recognize a particular secondary structure in the mRNA which


may be sufficient for the binding of mRNA by the ribosome. A

third possibility for the mechanism by which the mitochondrial

ribosome binds mRNA is the presence of a sequence nonspecific

RNA binding site on the mitochondrial ribosome (Denslow et

al., 1989). The roughly thirty bases long binding domain of

this binding site was disclosed by studies binding

oligoribonucleotides of various lengths. Furthermore, an

RNase T, protection domain of up to eighty bases in length was

provided on bound RNA (5' end labelled). The use of binding

assays should provide some insight into the interaction

between mitochondrial 28S subunits and mRNAs.

Bovine Mitochondrial mRNA Binding Properties of
28S Ribosomal Subunits

In order to determine the affinity and stoichiometry of

28S subunit binding to mitochondrial mRNAs Millipore filter

binding assays and sucrose density gradients were used. The

binding conditions closely simulated those previously used

successfully (Denslow et al., 1989; Liao and Spremulli,

1990b). Initially, both native and derived 28S subunits (see

Material and Methods) were used to see whether any differences

in their ability to bind message were readily apparent. The

results are displayed as the mean of three experiments in

which the ATPase 8/6 and COII mRNAs were incubated 28S

subunits, native or derived. The reaction data were plotted

with the aid of a personal computer and Sigmaplot 4.1, and


represented as pmoles of radiolabelled message bound versus

pmoles of 28S subunits.

The apparent Kd and stoichiometry data were then

calculated by Scatchard analysis (Scatchard, 1949) using the

following equation:

r/C = -1/Kd (r) + n/Kd

where: r= pmoles 28S bound/pmoles mRNA; C= concentration of
free ligand (28S ribs); Kd= dissociation constant; and
n=number of binding sites.

Figure 12 shows the plot of the experimental data and

Table 3 shows the results of the subsequent Scatchard

analysis. Both native and derived subunits bind mitochondrial

mRNA with similar affinity and stoichiometry. Analysis

indicates that each subunit has a single binding site, and

each type of subunit binds the mRNA with the same affinity

(Kd-3-5x10-) Maximum binding of the mRNAs occurred at a

ribosome to message ratio greater than 2:1 with approximately

85-90% of the message being bound. Henceforth the 28S

subunits referred to in this work will be derived 28S subunits

due to their enhanced purity by a second sucrose density

gradient, their equivalent binding affinity to native

subunits, and the relative availability of derived subunits.


E 4 -

I Coll

2 v -
1 1 1

0 10 20 30 40 50 60 70 80 90 100
28S ribosomes pmoles

Figure 12. Binding of Mitochondrial mRNAs to 28S Subunits.
Incubations of 9 pmoles (0. 18,4M) mitochondrial mRNA, both Coll
and ATPase 8/6 mRNAs with 4.5-90 pmoles (0.09-1.8MM) 28S
subunits (native and derived) were carried out in a 504l
reaction volume at 35*C. The reactions were stopped after 10
minutes by the addition of 8004i of ice cold 50/5 buffer.
Binding was assayed by Millipore filter binding. Each data
point represents the mean of three replicate reactions.
Circles represent Coll mRNA, triangles are ATPase 8/6 mRNA,
open symbols represent native 28S subunits, and closed symbols
derived 28S subunits.

Table 3

Binding Analysis of Mitochondrial mRNAs to 28S Subunits

Mitochondrial 28S Subunit Affinity Sites on 28S
Messenger RNA Preparation Subunits
Type KdxlO"M+sd n+sd

ATPase 8/6 Native 29.7+2.4 0.88+0.10
Derived 28.0+7.0 0.96+0.13
COII Native 52.8+3.2 1.00+0.06
Derived 51.3+2.3 1.02+0.10

E 2- 2

05 10 15 20

pmoles Coll mRNA

Figure 13. Binding of Coll mRNA to 28S Subunits. A reaction
mixture of 5.0 pmoles (0.1AM) 28S subunits and 1.0-20.0 pmoles
(0.02-0.4AM)Coll mRNA was incubated for 10 minutes at 35*C and
assayed by Millipore filter binding as previously described.


Table 4

Binding Analysis of COII mRNA to 28S Subunits

Messenger RNA Affinity Sites on 28S
KdxlO^9M+sd n+sd

COII 56+10 1.3+0.07


Binding was also done holding 28S subunits constant at

concentrations of 0.04 or 0.1AM and varying the mRNA

concentration from 0.02-0.4AM. The same equations were used

for analysis of binding and the resulting plot and analysis

are displayed in Figure 13 and Table 4. These data verified

the previous binding results so the 0.1MM concentration of

ribosomes was used for most of the subsequent binding

experiments. The apparent Kd for the binding of mitochondrial

mRNAs was approximately 30-50nM and the saturation of 28S

subunits was reached at a ratio of 1.5-2 mRNAs to small

ribosomal subunits. Again a single binding site on the

ribosome was resolved.

The ability of one mRNA to bind to one small subunit did

not rule out the possibility of a single message providing

multiple binding sites for ribosomes when ribosomes were in

excess. Millipore filter assays were limited for determining

this because they adsorb the 28S subunit whether or not it

has a bound mRNA. Sucrose density gradients were decidedly

the fastest and most practical method for addressing this

question (Denslow and O'Brien personal communication). The

gradients were prepared as they were for the isolation of

derived 28S from 39S subunits except that the 50/5 binding

buffer was used instead of the 300/5 buffer, as described in

Materials and Methods. These conditions have been used to

discriminate between 28S monomers and dimerss" of the 28S

subunits (Denslow, personal communication). Here 28S subunits


(9-72 pmoles) were incubated with 9 pmoles of [32P]

radiolabelled COII mRNA, prepared as described in the

Materials and Methods section, for 10 minutes at 35*C in a

final volume of 50Al. The reaction was stopped by adding 1ml

of ice cold 50/5 buffer to the reaction mixture. Two 1ml

reactions were loaded onto the gradient for centrifugation.

These gradients were then monitored and fractionated

(Materials and Methods).

Fractions of 2 ml were collected and subsequently counted

in a Beckman LS 380 scintillation counter to determine where

the radiolabelled mRNA had sedimented. For the results of

these gradients see Figure 14. The binding data has been

plotted in Figure 15 and the analysis recorded in Table 5.

The ribosome sedimentation profiles with the radioactivity

profile representing the labelled COII message, demonstrating

that the mRNA accompanies the 28S ribosome. Control gradients

(Figure 14, A) lacking ribosomes, show the mRNA sedimenting at

an approximately 4S peak, near the top of the gradient.

Incubation in the presence of 28S subunits (Figure 14,C)

results in bound mRNA sedimenting with the 28S subunits. This

observation continues for incubations with increased amounts

of 28S subunits (Figure 14,D-F) and the amount of mRNA bound

plateaus around the 2:1 ribosome to message ratio (Figure 15).

Significantly, no radioactivity (mRNA) appears in the position

expected for 28S dimer-mRNA complexes (approximately the 35S

position in the gradient).

5 n
J !I


0.2- _;




A2S4 -

I I I ]

I -








0.1 -





2 ~

- J. ~-


S -U

Figure 14. Sucrose Density Gradients of Coll mRNA to 28S
subunit Binding. The mRNAs specific activity was 7200
CPM/pmole. A: mRNA-no ribosomes; B: Ribosomes-no mRNA; C-F:
1:1, 2:1, 4:1, and 8:1 ratio of ribosomes to mRNA
concentrations, respectively.







: 1 1 I

S0.7 -

0.6 -

v 0.5 -
S0.4 -

0.3 .

0.2 -


0.0 0- '- ^- l-
0 10 20 30 40 50 60 70 80 90 100

28S Subunits pmolea

Figure 15. Plot of Coll mRNA Binding Assayed by Sucrose
Density Gradients. 18 pmoles of Coll mRNA incubated with 0,
18, 36, 72, 144 pmoles of 28S subunits, as indicated, prior to
analysis of the binding reactions by sucrose density gradient
centrifugation (Materials and Methods).

S- I I o co o


"z; 0 I I *

z I I

SO I co 1
0 *

c o

C> I 0 l L HJ H 0

o) I N N O L

Cl) 0 '0 O O N -
to 0

Ci ) 0 w 0 0 0 ^
Do 0 ~N~
NUC n r- n r

m 1I(



Therefore, there is no evidence for two subunits binding to a

single mRNA (even at large excess of ribosomes over message),

and there appears to be only one ribosome binding site on the


On the basis of the analysis of mRNA binding, assayed by

Millipore filter binding, it can be stated that 28S subunits

bind each of these two mitochondrial messages (COII and ATPase

8/6) with essentially the same affinity (apparent Kd of about

3-5x10"M). To ensure that only one ribosome was interacting

with the message, conditions were used in subsequent

experiments (i.e., footprinting), that result in approximately

80% saturation of the mRNA (9 pmoles mRNA and 18 pmoles of 28S

subunits). The mRNA binding site on the ribosome could be

saturated at 1.5-2 messages per 28S concentration. Finally,

from the observations of mRNA binding assayed by sucrose

density gradient centrifugation it can be said that 28S

subunits bind COII mRNA with unit stoichiometry: one mRNA

molecule per 28S subunit (Table 5) and reaches maximum binding

at about the previously determined 2:1 ribosome to message


Interaction of mRNAs of Non-Mitochondrial Origin with
Mitochondrial 28S Subunits

Given the ability of 28S ribosomes to bind mitochondrial

mRNAs, it is of interest to consider whether mitochondrial

ribosomes can bind other (non-mitochondrial) kinds of mRNAs --


or is there something "special" about mt mRNAs that promotes

their initiation (and not that of non-mt mRNAs)? Four non-

mitochondrial mRNAs were chosen to examine this question.

Poly(U) and poly(C) are homopolymers of ribonucleotides

uridylic and cytidylic acid, respectively. Poly(U) has been

used for investigations of ribosome-RNA interaction studies in

all systems (Denslow et al., 1989; Kumazawa et al., 1991; Hill

et al., 1990) and here represents an essentially unstructured

(i.e., single stranded) mRNA. Poly(C) falls into a similar

category but may have some ability to form an A-type helix

even in the single stranded form, expected in the conditions

used here (Saenger, 1985). A-type helices are common to most

RNAs, whether single or double stranded, because of the "C3,-

endo pucker" of the ribose ring which brings the 03,-phosphate

of the adjacent nucleotide closer to its own 5'-phosphate.

Poly(C) helical structure differs from the common A-RNA form

in that it displays a six base per turn conformation and its

"single helices retain standard nucleotide conformation even

better than the stereochemically more demanding double

helices" (Figure 16). So, poly(C) as a single stranded

template, is expected to exhibit structural features which

differ from poly(U).

Perhaps 28S subunits will display a binding affinity

higher for unstructured single-stranded templates like poly(U)

than for mitochondrial messages. If not, then possibly the mt

mRNAs have a sequence or structural feature that may be

important for 28S subunit binding. Also, a difference in



Figure 16. C3,-endo 'Pucker' of Ribonucleotides (Stryer et al.,
1988) and Structure of Poly(C) Single Strand Helix (Saenger,


the binding properties of poly(U) and poly(C) may indicate the

effects of an alternate helical structure on binding of

templates to mitochondrial ribosomes.

Additionally, two non-mitochondrial but naturally

occurring messages, ATPase P subunit from E. coli and

poliovirus mRNAs, were chosen as well. Expression vectors for

these mRNAs were provided by Drs. A. Lewin and J. Flanegan,

respectively. These two non-mitochondrial messages would

furnish a basis for comparing their binding to that of the mt

mRNAs and for answering the question of whether non-

mitochondrial messages bind differently (or at all). All four

messages together should provide clues into the mRNA features

required to promote/permit binding.

These various mRNAs were incubated with 28S subunits in

the same fashion as COII and ATPase 8/6 messages, 1-20 pmoles

of mRNA with 5 pmoles of 28S subunits. The data were plotted

and analyzed as above (Figure 17 and Table 6). Again a single

binding site on the ribosome was exhibited, just as for the

mitochondrial mRNAs tested. Furthermore, these non-

mitochondrial mRNAs had a similar affinity for the 28S subunit

(5x0l-M) with the exception of the unusually structured

poly(C)(10-7M). Thus, the 28S subunit appears to have a

general mRNA binding site which can accommodate these and

probably other messages.

Poly(U) binding by 28S subunits was also analyzed by

sucrose density gradient centrifugation to determine whether


S2- o PolyC

2 Beta ATPase
v Poliovirus -

0 5 10 15 20
pmolea non-mit mRNA

Figure 17. Binding of Non-mitochondrial mRNAs by 28S Subunits.
1-20 pmoles (0.02-0.40iM) mRNA were incubated with 5 pmoles
(0.10MM) of 28S subunits for 10 minutes at 35*C in a 50il
final volume.

Table 6

Binding Analysis of Non-mitochondrial mRNAs.

Messenger RNA Affinity Sites on 28S
KdxlO9"M+sd n+sd

/ ATPase 34.5+8.4 1.26+0.12
Poliovirus 32.2+4.8 1.18+0.04
Poly(U) 46.4+7.5 1.03+0.20
Poly(C) 136+32 0.82+0.22


more than one subunit could bind to extended, "unstructured"

mRNA. The resulting collection of gradients are recorded in

Figure 18. Figure 18 (B and C) shows that incubation of 9

pmoles (0.18jM) poly(U) with either 18 pmoles (0.36gM) or 36

pmoles (0.72AM) of 28S subunits resulted only in poly(U)

sedimenting in the 28S peak. The lack of entities sedimenting

at values greater than 28S, where dimers of 28S would

sediment, showed that only one small subunit was bound per

poly(U). Table 7 shows that of the poly(U) in the gradient

greater than 80% was bound by the 28S subunits under

conditions where the 28S subunits concentrations were 2 and 4

fold higher than the RNA. This is similar to the binding seen

for COII mRNA under SDG binding conditions (Table 5). An

approximated Kd of 34.5nM can be calculated.

This template RNA, poly(U), and a bovine mitochondrial

mRNA are binding in similar fashion under similar conditions,

which lends some credence to the possibility of the binding

being functional (poly-Phe has been translated). Neither COII

mRNA nor poly(U) were capable of being bound by more than one

28S subunit under these conditions. It would seem that the

28S subunit may preclude other messages from binding. Perhaps

in the case of an unstructured RNA like poly(U) the molecule

is wrapped around the ribosome. This could be the case for

all RNAs or the structure of mRNAs might limit their

availability for binding. If a message was folded back on

itself, "knotted", in its tertiary form, it could potentially

shield sequences that might otherwise be bound by the

ribosome's 30 base binding domain.

Interaction of Non-message Polynucleotides with
Mitochondrial 28S Subunits

What the parameters are for the binding of a

polynucleotide by a 28S subunit cannot be defined by the

previous attempts. So several other non-message

polynucleotides were employed under the now standard

incubation conditions to determine what characteristics of a

polynucleotide might allow or prevent a ribosome from

associating with it. A double stranded DNA (3200 base

pairs)(linearized, pGEM -3zf, Promega), a single stranded DNA

(TO 139)(confirmed as ssDNA by OLIGO 4.0 Primer Analysis

Software, Rychlik, 1989), tRNA (E.coli tRNA, Sigma), as an

example of a highly structured RNA molecule normally

interacting with ribosomes (Figure 19) were used in Millipore

filter assays of 28S subunits. All polynucleotides were 5'

end labelled and all exceeded the minimum polynucleotide

length requirement of 18 bases, previously determined for

efficient binding of poly(U)(O'Brien, et al., 1990).

Small subunits were maintained at a concentration of

0.1AM while substrate concentrations were varied from 0.04 to

0.8AM. The results of the Millipore filter assays are shown

in Figure 20 and the analysis of the binding is displayed in

Table 8. Even at 20-fold excess of polynucleotide to 28S



0.2 A

0.1 -





i i 1 1 11 i 1 1 1 11


AZs4 I

Azs4 I


Figure 18. Sucrose Density Gradient with Poly(U). A: Poly(U)
alone; B: 28S subunits alone; C: 2:1 28S subunits to poly(U);
and D: 4:1 28S subunits to poly(U). Note poly(U) appears near
the top of the gradient, or accompanying the 28S subunit peak,
when present, but none in faster sedimenting complexes. 28S
subunits do not sediment in faster sedimenting complexes.

zI 0

o r- 1 o co
S0 I *

M H I I in
( 0 I I 0
0 I I
< 4 I II ( '

So 0 O

0 o 1 0

0 0 co 00

0 0 .

M 0000
00 0


4at fl Nde I
2260 2509 Nae I

9m7n 1 9' s9art
Sca 818 i fl or,- Ec 1 5
Sac 5
Kcr 21
Ava 21
Amp pGEM-3Zf(-) acBa- 26
vector Xba 32
(3199bp) Sal1 38
Acc 39
H/rc I 40
PstI 4.18
Son i 54
nHir, II 56
A -- 69



^- ___.____ *0^ A-0 Am W0 O

.o ,G oo o0oo

C .-NA
c ---- ------

Figure 19. Non-Messages Used for Binding Assays. A: pGEM-
3zf(-), Promega, linearized by HindIII before use; B: TO 139,
ssDNA template; and C: tRNA, E. coli tRNA (Sigma). 5' end
labelling described in Materials and Methods.

subunits, ribosomes were not saturated (i.e. no more than 2

pmoles of polynucleotide are bound). The small subunit's

affinity for ssDNA was about 4 fold less than that of mRNAs,

while the apparent affinity for dsDNA or tRNA was about 20- to

30-fold less than that for mRNAs (Table 8).

These results indicate that the 28S subunits have a

relatively low affinity for non-message polynucleotides. It

may be that the structure of the polynucleotides presented to

the 28S subunits is incompatible with strong binding to the

template binding site. The mRNA must offer the binding site

something that the ss and dsDNA and tRNA could not.

Competition for 28S Ribosomal Subunit Binding of COII mRNA
by Mitochondrial and Non-mitochondrial mRNAs

The binding of all tested mRNAs by the 28S subunits

raised the question as to whether they all were bound at the

same site or if they were bound at different locations on the

same subunit. If ATPase 8/6 mRNA was interacting with the 12S

rRNA (28S subunit) by a putative S-D sequence, then poly(U),

poly(C), and COII mRNA would not compete for this site because

they do not exhibit an S-D sequence.

Millipore filter assays were employed to determine if an

unlabelled RNA could effectively compete for the binding of

COII mRNA. Both mitochondrial and non-mitochondrial mRNAs

were individually mixed with a saturating concentration

(0.17gM) of radiolabelled COII mRNA (see figure 13) prior to

incubation with 5 pmoles (0.1M) 28S subunits.



0 10 20 30 40 50

pmoles non-me.sage

Figure 20. Binding of Non-message Polynucleoties by 28S
Subunits. Labelled polynucleotides, 1-40 pmoles (0. 02-0.801M) ,
were incubated ip a 50gl final volume with 5 pmoles (0.10gM)
of 28S subunits for 10 minutes at 35C.


Binding Analysis of Non-Message Polynucleotides

Polynucleic Acid Affinity Sites on 28S
Kdxl0"9M+sd n+sd

pGEM (dsDNA) 811+98 0.55+0.12
tRNA 1600+96 0.96+0.06
TO 139 (ssDNA) 223+15 0.55+0.02

The competitors were increased in relation to COII mRNA from

0.5- to 4-fold molar amounts.

The ability of each of the RNAs tested to compete for the

binding of COII mRNA with a close correlation to their

apparent dissociation constants provides confidence that the

RNAs are occupying the same site on the ribosome. The

previous use of poly(U) as a translation template adds to the

evidence that this binding is significant.

The data for these experiments were plotted as percent of

control in Figure 21. The analysis was compiled using

Enzfitter Kinetics software by Leatherbarrow (1987) and a

personal computer. The modified equation for binding

inhibition (Cantor and Schimmel, 1980) was used and is shown


Y= 1-( 1-r(I/K,) / 1+r[l+I/Ki])xlOO

where r= Kd/A, constant=0.329, (Kd for COII mRNA is 56nM, Table

4, and A= concentration of COII mRNA, 0.17AM) and Y=l-vi/vo

(vi/vo=percent inhibition).

Table 9 shows the results of the above data analysis.

ATPase 8/6 and COII mRNAs and poly(U) were essentially

equivalent competitors displaying a Ki around 50nM, which

correlates well with the apparent Kd. Poly(C) competed less

effectively for binding by the 28S subunit showing a 2 fold

weaker inhibition of COII mRNA binding then the other three.

The binding affinity for poly(C) is also 2- to 3-fold less

than COII mRNA (Table 6), possibly for unusual structural

characteristics. Again the apparent Kd and Ki correlate well.





0.2 -

0.0 I I
0 10 20 30 40

Competitor added (pmole)

Figure 21. Competition by Other Templates for the Binding of
COII mRNA on 28S Subunits. 28S subunits, 5 pmoles (0.1M) were
incubated with a mixture of 8.5 pmoles (0.17MM) of COII mRNA
(100% of Control-saturating) and increasing concentrations of
competitor from 4.3-34 pmoles (0.09-0.68MM) for 10 minutes at
350C in a 50Al final volume.


Analysis of Competition for COII mRNA Binding
by mRNA Templates

Messenger RNA K, x109M +se

COII 51.5+3.9
ATPase 8/6 46.1+0.9
Poly(U) 47.8+4.7
Poly(C) 108+7.6

Effects of Aurintricarboxylic Acid on COII mRNA Binding
to 28S Ribosomal Subunits

The 28S subunit is a complex of a 12S rRNA molecule and

33 proteins. This can be compared to the E. coli small

subunit which possesses a 16S rRNA and 21 proteins.

Speculation has frequently centered around the possibility of

the proteins taking up functional as well as "space- filling"

roles of the petite mitochondrial rRNA.

A computer aided search for possible complementary

regions between the 12S rRNA and the mitochondrial messages

has lead to speculation about the possibility of an internal

S-D-like sequence on the mitochondrial messages (Saccone et

al., 1985). Eight of the 13 mRNAs were identified as having

a sequence complementary to a region of the 12S rRNA about 700

nucleotides from the rRNA 5' end. The sequence on the 12S

rRNA is approximately 20 bases long and gives predicted

energies of interaction (AG) between -20to -30 kcal/mole.

This is 4-6 times the AG of -4.9 kcal/mole using Zuker values

for strong S-D interactions (i.e., AGGAGGU of R17 phage

protein A mRNA S-D sequence, all of which are paired). This

kind of binding would appear to be too strong to allow enough

freedom for the RNA to be translated without expending a great

amount of energy.

The 12S rRNA is not accessible to modification by

chemical agents in this region (Faunce,1991) giving evidence

that a macromolecule like mRNA is not likely to penetrate this

domain either. Furthermore, the binding of COII mRNA was

shown to have little effect on the modifications or cleavage

of 12S rRNA (Faunce, 1991). The only effect noted was to two

bases in a region approximately 500 bases 5' of the putative

S-D sequence which expressed enhanced sensitivity to RNase A

when COII message was bound. Another possibility for rRNA to

be involved in the binding interactions with the mRNA would

invoke a magnesium cross-bridging model that would be novel in

translation systems.

A final consideration would be to picture this

interaction like any other RNA binding protein-RNA

interaction. Binding of mRNA by the mitochondrial ribosomal

proteins seems the most likely to occur, since all mRNAs,

regardless of sequence and structure, were able to be bound

most likely in the same site with approximately the same

affinity by the 28S subunit. In this case a red dye,

Aurintricarboxylic Acid (ATA) (Figure 22) was used as an

inhibitor of protein-RNA interaction by the ribosome. This

dye has been used successfully as an RNase inhibitor and its

interaction on RNA binding proteins has been characterized

(Gonzalez et al., 1980). It was shown to inhibit protein

chain initiation of eukaryotic and prokaryotic ribosomes but

not the continuation of peptidyl elongation at low

concentrations (5gM)(Mathews, 1971).

To test the effects of ATA on the binding of mRNA by 28S

subunits, 5 pmoles (0.1MM) 28S subunits were preincubated for

5 minutes at 35*C with 5MM ATA. This concentration of ATA is

strongly inhibitory to RNA binding proteins (Blumenthal and

Landers, 1973 and Marcus, et al., 1970). It also is capable

of inhibiting the binding of COII mRNA from 0.17-0.68iM

(saturating levels) by the 28S subunit suggesting that the

message is being bound by proteins on the small subunit which

serve as the mRNA binding site. The data were plotted in

Figure 23. As can be readily seen, the virtually complete

ablation of the 28S subunits ability to bind COII mRNA was

observed in the presence of ATA.


The mitochondrial ribosomal small subunit was able to

bind not only mitochondrial mRNAs but non-mitochondrial mRNAs,

including homopolymers of uridine and cytosine not expected to

exist in nature. That the 28S subunit showed no preference

for homologous messages, registering dissociation constants

around 40-50nM with the exception of poly(C) (>0.1MM), implies

that all of these test templates have properties that allow

them to be bound by the mitochondrial small subunit. Since

poly(U) binds with the same affinity as the mitochondrial

messages, it appears that special secondary structures of

these messages are not a requirement for their binding by 28S

subunits. The site to which these templates are binding then

appears to be a general template binding domain, whose

0 OH



Figure 22. Chemical Structure of Aurin Tricarboxylic Acid.

I I--------, ---- I--

5 -u -ATA




0 10 20 30
pmoles Coll mRNA

Figure 23. Effect of ATA on Coll mRNA Binding to Ribosomes.
28S subunits, 5 pmoles (O.1M) were incubated with 5gM ATA
prior to the addition of 8.5-34 pmoles (0.17-0.68MM)
radiolabelled Coll mRNA. Note that 0.17gM is a saturating
concentration of Coll mRNA for 0.1 M ribosomes.


properties were inferred earlier (Denslow et al., 1989).

Perhaps the common feature of these different templates that

allows binding is the presence of (relatively) unstructured,

single stranded regions (since poly(U) binds with the same

affinity as COII and ATPase 8/6 mRNA, and competes for the

same binding site). Furthermore, since COII and ATPase 8/6

mRNAs bind with the same affinity and compete for the same

binding site, it would appear that the putative S-D

interaction of ATPase 8/6 is not contributing significantly,

if at all, to the binding of this mRNA to the ribosome.

The mitochondrion's compartmentalization, alone, is a

limiting agent which would allow only the presence of its own

mRNAs for translation by the ribosome. The fairly high

affinity measurements, the apparent single site, and

reversible binding, and the ability of several mRNA templates

to be bound point to this binding as a physiologically

meaningful interaction. The sucrose density gradients

exhibited the small subunits high affinity for COII mRNA and

poly(U) and the unit stoichiometry of binding. It is doubtful

that the mitochondrion could support much in the way of

polysomes because of the relative paucity of ribosomes and

mRNAs. Each liver cell has an average of 4-5x105

mitochondrial ribosomes (O'Brien, personal communication).

The number of messages per adult rat liver cell mitochondrion

has been determined at about 90 messages per mitochondria or

8 molecules of each message. The 12S rRNA content is 81 per

mitochondria (Cantatore et al., 1984) so that the ratio of

message to 12S rRNA (the maximum number of small subunits

possible) is about 1:1. While this does not preclude the

formation of polysomes in vivo, it does show that the

stoichiometry does not favor them. In HeLa cells each of the

11 messages are produced at a rate of roughly 1 molecule per

minute and demonstrated half-lives between 25 and 90 minutes

and the 12S rRNA is about 100 fold higher (Gelfand and

Attardi, 1981), all in a protein dense matrix where mobility

would be severely limited. Thus, it stands to reason that it

may be advantageous for the small subunit to sequester a

message with a fairly strong affinity interaction until the

balance of the translation machinery components arrive to

initiate the peptide elongation.

Furthermore, the binding appears to be predominantly to

a proteinaceous site. Previous work with RNA modifiers DMS

and l-cyclohexyl-3-(2-morpholinoethyl)-carbodiimde metho-p-

toluenesulfonate (CMCT) followed by AMV-RT extension has been

done on 12S rRNA in the presence and absence of COII mRNA.

These chemical agents are capable of modifying solution

accessible adenines and cytidines (DMS) and guanines and

uridines (CMCT) on the N1 nitrogen of the purines and the N3

nitrogen of the pyrimidines with which they react (see figure

4). These nitrogen molecules are involved in Watson-Crick

base pairing interactions and will not be modified by the

chemicals if they are paired. The experiments showed that the

12S rRNA chemical modifications were not different with or

without bound COII mRNA. The putative S-D-like sequence was

also shown to be unavailable for modification in the 28S

subunit by any method, chemical (DMS and CMCT) or enzymatic

(RNase T1, A, and Vj), and was also predicted by phylogenetic

analysis (Gutell et al., 1985) and by computer folding

analyses to be involved in base-paired helices. These data

would indicate that the 12S rRNA sequence that has regions of

complementarity in the eight mitochondrial messages is

unavailable for interaction with the mRNA.

Robert Heck has bound a bromouracil-containing (a UV

activated uridine analog), radiolabelled COII mRNA to 28S

subunits. The bound message is then UV cross-linked to

proximal elements of the binding site of the small subunit.

After degrading the mRNA and dissembling the subunit, the

crosslinked label (mRNA fragment) was found to be associated

with two ribosomal proteins with molecular weights in the 40

kd range.

The binding, then, is consistent with the speculation of

a "generic mRNA binding site" on the 28S subunit. What, then,

are the properties required for binding to this site? First,

ssRNA binds well, and DNA (ds or ss) or dsRNA bind poorly,

suggesting that the C3,-endo "Pucker" characteristic of

polyribonucleotides may be a significant feature required for

ribosome binding, as well as, an open stretch of

"unstructured" RNA. Because the affinity of poly(U) for the

binding site matches that of COII, both may be binding by the

same kind of interaction, involving predominantly

unstructured, single stranded regions of the template. This

possibility raises questions: What is the structure of COII

mRNA? Is it largely single stranded, or are only one or a few

regions single stranded/do one or more of these get bound by

the 28S subunit binding site? Where are they located?

Finally, does the binding of COII message to the small

subunits include the 5' end and therefore the initiation



Background and Significance of Secondary Structure Analysis

Some insights into the nature of the mRNA at the site of

interaction with the 28S subunit were revealed in the previous

studies dealing with affinity, specificity, stoichiometry,

reversibility, and the nature of the binding site (RNA or

protein). Access to a computer prediction of the secondary

structure of COII mRNA may provide some clues into physical

nature of the molecule and some understanding of the potential

site of interaction with the small subunit.

Computer nucleic acid folding programs predict the

secondary structure of a polynucleotide based on conventional

and nonconventional but common Watson-Crick base pairing

(i.e., the potential for guanines to pair with uridines. Most

folding programs are designed to minimize the energy of the

final folded product (maximize the number of the most

favorable base pairing interactions) and/or to meet

phylogenetic restrictions. Neither of these may provide a

complete or accurate description of the mRNA structure, but

they do give a projection on which to build a more accurate



To obtain data useful for modelling the mRNA structure,

the RNA must be physically examined in a manner that does not

disrupt its higher order structural characteristics. These

data can be used to refine the computer prediction and deliver

a comprehensive view of the physical nature of the message.

The techniques frequently used to study the structure of

macromolecules: X-ray crystallography and electron microscopy,

requires a great deal of precious reagent and/or use fixatives

and procedures that may be destabilizing to base pairing.

Furthermore, they provide only limited information on

molecules like bare RNAs.

An alternative method used to differentiate the stems and

single stranded stretches of an RNA, and the one chosen here,

is the chemical modification and enzymatic cleavage of the RNA

followed by primer extension termination (Moazed and Noller,

1986; Stern et al., 1988). This technique captures the target

RNA's bases in their higher order structure and AMV-RT driven

primer extension discloses the modified base as the extension

is disrupted (Boorstein and Craig, 1989; Knapp, 1989).

The use of chemical probing modifies both open and

potentially secluded single stranded nucleotides areas.

Dimethyl sulfate (DMS) methylates adenines (N1) and cytidines

(N3) (see Figure 4) that are not paired with another nucleotide

forming a stem or some other structure, like a pseudoknot.

Guanines are also a target for the action of this modifier but

are generally not disclosed by this method in that they do not

Full Text
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 EWNHJ00YS_02SWSQ INGEST_TIME 2012-02-20T22:03:26Z PACKAGE AA00009057_00001