Structural studies of the RNA of the small subunit of bovine mitochondrial ribosomes

MISSING IMAGE

Material Information

Title:
Structural studies of the RNA of the small subunit of bovine mitochondrial ribosomes
Physical Description:
Book
Creator:
Faunce, Wesley H., 1964-
Publication Date:

Record Information

Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 25961284
oclc - 25605556
System ID:
AA00022826:00001


This item is only available as the following downloads:


Full Text














STRUCTURAL STUDIES OF THE RNA OF THE SMALL SUBUNIT OF
BOVINE MITOCHONDRIAL RIBOSOMES
















By

WESLEY H. FAUNCE III


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

UNIVERSITY OF FLORIDA


1991















DEDICATION



I dedicate this research to my Fiancee, Leticia Conte

and my parents, Wesley and Ruth Faunce.















ACKNOWLEDGEMENTS


There are numerous people who deserve acknowledgement

on this page for their academic, technical, and emotional

support. Foremost is the supervisory committee, Dr. Boyce,

Dr. West, Dr. Hauswirth, and Dr. Nick, without whom this

research would not have been possible. I would like to

single out Dr. Nick for special thanks for his

consultations on some of the methods employed. The

supervisory chair, my mentor, Dr. Thomas W. O'Brien,

deserves extra special accolades for his insight, wisdom,

and creativity which were a tremendous boon to this

research project. I would like to thank him for the

opportunity to make magic in his laboratory (without having

to mine my own bauxite, like they did in olden times).

I would also like to thank the members of the O'Brien

lab, Scott Fiesler, Bob Heck, Jigulo Liu, Mietek

Piatysczek, and Pat Gillevet. Members of the lab deserving

special mention are Mike Bryant, John Anders, and Bernie

Courtney for their numerous contributions, both academic

and automotive. I would also like to acknowledge the two

newest members of our lab, Jeff Smith and John Sarzier, who

were friends of mine before joining the lab and,


iii









surprisingly, have remained friends even after working with

me. Also deserving of special mention is Dr. Nancy Denslow

for her expert guidance and endless patience with my

questions.

In addition to academic and technical support, I have

received a great deal of encouragement from my closest

friends outside of the lab, Steve Donahue, Kelly Haselden,

Steve Atkins, Sam Jalet, Ruth Faunce and Wes Faunce.

Finally, I would like to thank my best friend and fiancee,

Leticia Conte, who has been invaluable in the emotional

support category as well as the prodding category (when

needed).















TABLE OF CONTENTS

page

ACKNOWLEDGEMENTS ........................................ iii

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

ABBREVIATIONS ............................................. x

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

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

MATERIALS AND METHODS .................................... 20

Preparation of Bovine Mitochondria .................. 20
Preparation of Native 28S Subunits and 55S
Monosomes ...................................... 20
Preparation of Derived 28S Subunits ................. 21
Preparation of IF3 .... .............................. 21
Preparation of Short Synthetic RNA................ 21
Preparation of ATPase 8/6 and Cytochrome Oxidase
Subunit II mRNA................................ .22
Binding of RNA Transcripts to 28S Derived
Subunits........................................ 22
Binding of IF3 to 28S Derived Subunits ..............23
Millipore Filter Binding Assay ...................... 23
Chemical and Enzymatic Modifications ................23
DNA Primers ..... . ..................................... .... ....... 25
Extension of Oligodeoxyribonucleotide Primers....... 25
Denaturing Gels ..................................... 28
Gel Reading Algorithm............................... 28

SOLVENT ACCESSIBLE NUCLEOTIDES IN THE 28S SUBUNIT........ 30

Significance of Chemical Probing .................... 30
Identification of Solvent Accessible Regions of 12S
RNA in the Small Subunit....................... 32
Environmental Dependence of Small Subunit Solvent
Accessibility .................................. 38
Comparison of Chemically Accessible Regions of
Mitochondrial 12S rRNA and E. coli 16S rRNA .... 42
Refinement of the 12S rRNA in subunit Secondary
Structure ................... . .. ................... ....50









DEVELOPMENT OF A THREE DIMENSIONAL UNDERSTANDING OF THE
SMALL SUBUNIT........................................ 55

Significance of Three Dimensional Modeling .......... 55
Identification of Surface Accessible Regions of 12S
RNA in the Small Subunit ....................... 56
Development of a Three Dimensional Model of the Small
Subunit from Electron Microscopy................ 62
Localization of Surface Accessible 12S RNA Secondary
Structures Within the Small Subunit ............ 69

THE EFFECT OF RIBOSOMAL SUBSTRATES ON NUCLEOTIDE
ACCESSIBILITY IN THE SMALL SUBUNIT................... 75

Significance of Ribosome-Substrate Interaction...... 75
Effect of Transcripts on 12S RNA Reactivity......... 77
Effect of Initiation Factor 3 on 12S RNA Reactivity.89

CONCLUSIONS AND FURTHER DIRECTIONS ....................... 97

Summary ............................................. 97
Future Experiments .................................... 99


APPENDIX I .............................................. 102

Chemically and Enzymatically Accessible Nucleotides
of 12S rRNA..................................... 102

APPENDIX II ............................................. ll

Stem by Stem Comparison of Bovine Mitoribosomal 12S
rRNA and E. coli 16S rRNA ..................... ll

REFERENCES .............................................. 116

BIOGRAPHICAL SKETCH ..................................... 123















LIST OF FIGURES


page
1. The Three Stages of Translation......................... 2

2. 5' Ends of Mitochondrial mRNAs.......................... 8

3. The RNA Landing Pad..................................... 9

4. Functional Roles of E. coli 16S rRNA .................. 13

5. Comparison of E. coli 16S rRNA and bovine mitochondrial
12S rRNA ............................................ 15

6. Site of Chemical Modifications on Ribonucleotides .... 18

7. Titration of RNAse V,.................................. 26

8. The Location of Primers on the Proposed 12S RNA
Secondary Structure ........................... ...... 27

9. Chemical Modifications 3' to Primer G .................34

10. Chemical Modifications 3' to Primer C ................35

11. Location of Chemical Modifications on the 12S Secondary
Structure ........................................... 36

12. A26 Profiles of Ribosomes in Sucrose Density
Gradients ........................................... 39

13. Chemical Modifications 3' to Primer C in the Native
Subunit ............................................. 40

14. Summary of Chemical Modifications on E. coli 16S
rRNA ................................................ 43

15. Relative Protein and RNA Masses of 28S and 30S
Subunit ............................................. 44

16. Models for the Arrangement of Protein and RNA Mass
Within the Ribosome................................. 46

17. Densitometric Analysis of Chemical Modifications 3' to
Primer C ............................................ 52


vii









18. Refinement of the Secondary Structure of Stem 18A....53

19. RNAse A/Ti Cleavages of Stem 18 ...................... 57

20. RNAse A/Ti Cleavages of Stem 15 ...................... 59

21. RNAse A/Ti and V1 Cleavages 3' to Primer G........... 60

22. Location of Enzymatic Cleavages on the 12S Secondary
Structure ........................................... 61

23. Electron micrograph of 28S Subunits .................. 64

24. Three Dimensional Model of the Small Subunit......... 65

25. Topographic Layers of the Small Subunit.............. 67

26. Three Dimensional Computer Model of the Small
Subunit ............................................. 68

27. Identification of Analogous Secondary Structures
Between Bovine Mitoribosomal 12S rRNA and E. ccli 16S
rRNA ................................................ 70

28. Comparison of Two Independent Models of the 30S
Subunit ............................................. 71

29. Location of Surface Accessible Secondary Structures
Within the Three Dimensional Model ..................73

30. Sequence of BH ....................................... 79

31. Cytochrome Oxidase Subunit II mRNA Secondary
Structure ........................................... 80

32. Accessibility of the Shine-Dalgarno-Like Sequence in
the 28S Subunit..................................... 82

33. Effect of COII on 12S RNA Accessibility.............. 84

34. Location of the COII Effect on the 12S RNA Secondary
Structure ........................................... 85

35. Sequence of Pan...................................... 87

36. Accessibility of 12S RNA to the Pan Probe............ 88

37. Effect of IF3 on 12S RNA Accessibility............... 91

38. Location of the IF3 Effect on the 12S RNA Secondary
Structure....... ..................................... 92


viii









39. Effect of IF3 on E. coli 16S rRNA Accessibility...... 94

40. Effect of IF3 on the Stem 15 Area of 12S rRNA........ 95















ABBREVIATIONS


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

of adenosine 5'-triphosphatase

Ac........ acetate

AMV .......avian myoblast virus

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

ATPase .... adenosine 5'-triphosphatase

BSA.......bovine serum albumin

Ci........Curie

CMCT......l-cyclohexyl-3-(2-morpholinoethyl)-carbodiimide

metho-p-toluene sulfonate

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

cytochrome oxidase

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

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

E. coli...Escherichia coli









elF.......eukaryotic initiation factor

exp.......experiment

fMet......formylated methionine

fMet-tRNAifMet..methionine initiator transfer ribonucleic

acid charged with formylated methionine

FPLC......fast performance liquid chromatography

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

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

ICBR...... Interdisciplinary Center for Biotechnology

IF........bacterial initiation factor

in mito...refers to manipulations done on 12S RNA while it

is still inside the mitochondrion

in subunit..refers to manipulations done on 12S RNA while

it is still inside the small subunit

Kd........kilodalton

M ......... molar

Met.......methionine

Met-tRNAiMet..methionine initiator transfer ribonucleic

acid charged with methionine

min .......minute

mg........milligram

mM........millimolar

mmol......millimole

mRNA......messenger ribonucleic acid

mt........mitochondria









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

dehydrogenase

nM........nanomolar

pM........picomolar

poly(U)...polyuracil

RNA.......ribonucleic acid

RNAse.....ribonuclease

RNAsin .... ribonucleasin

r-protein..ribosomal protein

rRNA......ribosomal ribonucleic acid

S.........Svedberg

SDS.......sodium dodecylsulfate

S5 ........ small subunit mitoribosomal protein #5

tRNA......transfer ribonucleic acid

tRNAi..... initiator transfer ribonucleic acid

tRNAiMet..methionine initiator transfer ribonucleic acid

tRNAifMet..formylated methionine initiator transfer

ribonucleic acid

U ......... unit

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

uCi.......microCurie

uM ........micromolar

2D........two dimensional

3D........three dimensional

28Sd SU...derived mitochondrial ribosomal small subunit

28Sn SU...native mitochondrial ribosomal small subunit


xii









32P.......radioisotope of phosphorous

35S.......radioisotope of sulphur

55S M.....mitochondrial ribosome


xiii















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

STRUCTURAL STUDIES OF THE RNA OF THE SMALL SUBUNIT OF
BOVINE MITOCHONDRIAL RIBOSOMES

By

WESLEY H. FAUNCE III

AUGUST 1991


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


The mitochondrial ribosome has a number of unique

properties as compared with other types of ribosomes. The

relatively high protein content (nearly twice that of

prokaryotic ribosomes) and low RNA content (approximately

half that of prokaryotic ribosomes) of mitochondrial

ribosomes raises a number of questions. This research is

concerned with how the mitochondrial ribosome functions

with such a low RNA content and what the conformation of

the RNA may be during different functional states of the

ribosome. Bovine mitoribosomes were chosen as the model

system for this research, since the majority of information

on mitoribosomes is based on the bovine system and also,

because of the availability of bovine liver for the

preparation of mitoribosomes.


xiv









To understand the structure of the 12S mitoribosomal

RNA and to investigate changes in the structure under

various experimental conditions, the 12S RNA was probed in

the 28S subunit with dimethyl sulfate, l-cyclohexyl-3-(2-

morpholinoethyl)-carbodiimide metho-p-toluene sulfonate,

and ribonucleases A, T1, and V1. The modification data

allowed refinement of the existing predicted 12S rRNA

structure by eliminating stem 18A and flaring the ends of

stems 4, 9a, and 15. This study also permitted comparison

with the more thoroughly studied E. coli 16S rRNA. In

addition, by analyzing the enzymatic data alone, the

surface accessible regions of 12S rRNA can be identified.

The more readily surface exposed structures include stems

1/2, 18, and 26b. Also, the accessibility of 12S rRNA in

subunit was studied under various experimental conditions

including the small subunit bound to IF3 and assorted RNA

transcripts. It was shown that the presence of COII mRNA

enhanced 12S RNA accessibility to RNAse A at nucleotides

U232 and U234. Also, in the presence of IF3, RNAse A

activity was enhanced at nucleotide U871.















BACKGROUND


Protein biosynthesis, an exceedingly complex, multi-

step pathway which begins with transcription of DNA into

RNA, leads through translation of mRNA into proteins, and

ends with the final processing of the protein product and

translocation of this product to its proper site within, or

outside of, the cell. Each step in the biosynthesis of

proteins requires a number of macromolecules and factors.

In addition, each step also has multiple points (and

various types) of regulation which add to the complexity of

the process. In spite of the complexity of protein

biosynthesis, translation remains the crux of protein

manufacture.

Translation itself can be divided into three discrete

stages: initiation, elongation, and termination. Each of

the three stages of translation differ in factor

requirements between prokaryotic and eukaryotic systems.

For example, initiation in prokaryotic systems requires

mRNA, IF1, IF2, IF3, fMet-tRNAifMet, GTP, and both

ribosomal subunits, where as eukaryotic systems require 5'

Capped mRNA, eIFl, eIF2, eIF3, eIF4F, eIF5, Met-tRNAiMet,

GTP, ATP, and both ribosomal subunits (see Figure 1,






















(la) INITIATION


30S Initiation complex


PROKARYOTE ,30


30S /
small
GTP *ubunit
.wu




GTP
4 + (( + 40S small +
L ,ubunit


+ mRNA


ATP + mR


EUKARYOTE


40S Initiation corn

I;,,

NA

ADP


J 70S or OOS
initiation complex


I 'AUG y i
+^< Large subunit ^ A i ^
(SOS or 6US)


Las ge l"
paex ,




9I3..


".--------------------


( (b ELONGATION



I I u o A U G I













PROK EUK
Tu a EF1
Ts EF 1


A_ U G UUU/ CUG UAG7


^ \ ayftlas
synthesis
and tranlocatlon

A to P switch catalyz
by G (PflOK) or EF, IE


-GTP GP+P
GTP GDP + P)


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 and
these requirements differ depending on the type of
ribosome.


INTIfATION COMPONENTS PROKARYOTIC EUKARYOTIC
Small
C^ ribomaKTl I WAUG mRN2 30S small l~~ G NA 408 small
ubunht *ubunit s ubunt
(1 M or 40S) naation A911i*i6, Iiito
0 0 0 facio 50S large factor. oSOS W
Large I&bLIt aubunit
dbosomal
subunit MWat*
US f r'M NtRNA GTP Met-tRNA GTP
or06 SO ATP
f AT"P


Translocation
completed
and tRNA-
positioned


ad
JKI






















(c) TERMINATION


..-.' .. ". Termination
/ 1 ycodon
A*U B^UUU U UG UAGI UUU CUG .AG

li t ] r Translocation
I-j | GTP GOP ;E @.GTP
P PROK: 3TF5
EUK: 1TF


UUU CUSG UAGI
\GTP
Rtidyk i
NAA


Polypaptide


"9


Figure 1 (cont.).


'T


ON
I
I









4

Darnell et al., 1990). Further study of Figure 1 shows that

the mechanism of initiation also differs somewhat between

prokaryotes and eukaryotes. Each system choreographs its

components and factors so that the end result is an

assembled ribosome with a charged tRNA, at the peptidyl-

tRNA site. In all stages of translation, the ribosome

remains the focus of all mechanistic activities, with both

systems having their own particular ribosome (80S for

eukaryotes and 70S for prokaryotes).

In 1967, a third, unique kind of ribosome was

discovered, the 55S mitochondrial ribosome (O'Brien and

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

unconvinced, declaring the 55S "mini ribosome" a bacterial

contaminant or a cytoplasmic large subunit contaminant.

Over the years, a substantial body of data emerged,

supporting the existence of a distinct mitochondrial

translation system and mitoribosomes gradually became

accepted as a unique class of ribosomes. It was shown that

some bacterial protein synthesis factors were not

interchangeable with the 55S ribosome (Denslow arid O'Brien,

1979) indicating that the mitoribosomes had properties

different from bacterial ribosomes. Since that time, the

rRNA sequences have been published, some mitochondrial

specific factors isolated, and even a few of the r-protein

genes have been sequenced (Guttel et al., 1985, Liao and

Spremulli, 1990a). But, even twenty-four years after its









5

discovery, the mitoribosome and the mitochondrial

translation system remain somewhat obscure and poorly

understood, since the emphasis of study on translational

systems continues to be on the 80S and 70S ribosomes.

Although outside the mainstream of ribosomology, the

mitochondrial translation system begs further study due to

its many unique properties (see Table I).

The mitoribosome's low sedimentation rate of 55S is

due to its relatively high protein content (85 proteins

total) as compared with prokaryotic ribosomes (53 proteins

total). In addition, the mean size distribution of

mitoribosomal proteins is larger than the prokaryotic r-

proteins (Matthews et al., 1982). Another contributing

factor to the slow sedimentation rate is the relatively low

rRNA content of mitoribosomes. Mitoribosomes possess a 12S

rRNA and a 16S rRNA where as prokaryotic ribosomes possess

a 16S rRNA and a 23S rRNA. Current speculation suggests

that the "extra" mitoribosomal proteins may be filling in

the structural voids created by the foreshortened

mitochondrial rRNAs (O'Brien et al., 1980).

Another area in which mitochondrial translation is

unique is in its initiation. Mitochondrial mRNAs have no 5'

methyl caps or anti-Shine-Dalgarno leader sequences, nor

does the 12S rRNA possess a Shine-Dalgarno sequence. In

fact, mitochondrial mRNAs possess little or no 5' leader















Table I



Unique Properties of Mitochondrial Translation (O'Brien,
1971, O'Brien et al., 1980, 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 in-house) factor requirements

7. Unique genetic code and tRNAs

8. Mitoribosomes are products of two different genomes

9. mitoribosomes bind GTP directly









7
sequence (see Figure 2). Since 5' caps and leader sequences

are both absent, mitochondrial translation initiation must

take place through a novel, and as yet undiscovered,

mechanism. Speculation has suggested the presence of an

internal Shine-Dalgarno like sequence (Saccone et al.,

1985), while other speculation proposes that the 5' ends of

the mRNAs are important (Denslow et al., 1989) The

mitoribosome may recognize a particular secondary structure

in the mRNA near the 5' end or perhaps the 5' phosphate

itself may be sufficient for recognition of mRNA by the

mitoribosome. Unfortunately, there is no experimental

evidence for either theory, and the whole story is likely

to be more complicated, since the mitochondrion translates

some messages which are bicistronic and are out of frame,

such as ATPase 8/6. One clue as to the mechanism of

mitochondrial translation initiation is the presence of a

sequence non-specific RNA "landing pad" on the mitoribosome

(Denslow et at., 1989). The binding domain of this landing

pad is roughly thirty bases long (see Figure 3) as shown by

binding studies of oligonucleotides of various lengths. In

addition, this mitoribosomal.domain affords RNAse

protection for bound RNA up to eighty bases in length.

Current thinking suggests this domain may aid in

positioning the mRNA on the mitoribosome prior to the start

of translation and may help keep the mRNA aligned during

elongation. In order to better understand initiation and





























COI CAUG UUCAUUAACCGCUG

COII UAUG GCAUAUCCCAUACA

COIII UAAUG ACAACCAAACUCAU

NDI AAAUG UUCAUAAUUAACAU

NDII AUA AACCCAAUUAUCUU

NDIII AUA AAUUUAAUACUAGC

CYTB ACUAAUG ACUAACAUUCGAAA





















Figure 2. 5' Ends of Bovine Mitochondrial mRNAs. Note the
extremely short, or non existent, 5' leader sequences (Hill
et al., 1990).


























BINDING
301


. __RNASE [RDTECTILIN
80b


Figure 3. The RNA Landing Pad. Messenger RNA is thought to
bind to a domain on the small subunit roughly 30 bases in
length. An additional 50 bases of mRNA are protected from
nuclease activity by the subunit (Denslow et al., 1989).









10
elongation, a few mitoribosomal factors are in the process

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

majority of factors probably remain undiscovered. Due to

the relatively large number of mitoribosomal proteins, the

possibility of resident factors also exists; that is,

factors that are in constant association with the ribosome.

In addition to a unique initiation mechanism, the

mitoribosome translates mRNA using a unique genetic code

and unique tRNAs. In fact, the mitoribosome utilizes fewer

tRNAs (22) than its bacterial counterparts (62).

Interestingly, the search for a mitochondrial initiator

tRNA has been a difficult one and some have found evidence

suggesting that initiator tRNA (and perhaps other tRNAs)

are actually imported into the mitochondria (Clayton,

personal communication). 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 mitoribosomes is that they

are the product of two different genomes. The proteins are

encoded by the nuclear genome, translated on cytosolic

ribosomes, and then imported into the mitochondria

(Schieber and O'Brien, 1982). The rRNAs, in the other hand,

are transcribed from the mitochondrial genome and then

assembled with the imported proteins to form mitoribosomes.

This mechanism of ribosome biosynthesis raises a number of

questions unique to the mitochondrial translation system.









11

For example, how are the r-proteins imported? Also, how are

stoichiometric levels of r-proteins relative to rRNAs

maintained and how does the mitochondrial machinery

communicate with that of the cytosol and the nucleus?

Another distinctive property of mitoribosomes 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

mitoribosome tightly (Kd = 20 nM) and exchanges rapidly

with GDP which has an apparently equal affinity for the

mitoribosome. Unfortunately, there is no evidence as to

what the GTP or GDP is doing, but the binding of GTP

directly to the ribosome provides circumstantial evidence

as to the presence of an in-house factor (perhaps an IF-2

like factor). Again, the mitoribosome provokes distinctive

questions such as what is the function of GTP?

Although the primary focus of ribosomologists has been

on the eukaryotic and prokaryotic systems, the unique

properties of the mitochondrial translation system,

outlined in Table I, demand that this system be studied as

well; but, how best to study this system? In order to

obtain the greatest amount of data on these properties in

the most efficient manner, it is reasonable that the rRNA

of the small subunit be studied first. This reasoning is









12

based on the fact that much of initiation takes place with

only the small subunit, the RNA landing pad is found on the

small subunit, and it is the small subunit which binds GTP.

Furthermore, its seems prudent to study the one RNA

molecule of the small subunit rather than all of its

thirty-three proteins. Also, the functional aspects of the

small subunit RNA have been investigated in other systems,

providing a rich comparison base.

Figure 4 displays some of the emerging functional

roles assigned to various regions of E. coli 16S rRNA

(Dahlberg,1989). The 16S rRNA has been shown to be involved

in both the initiation and the termination of protein

biosynthesis. During initiation 16S rRNA binds the 5'

untranslated region of mRNAs through its Shine-Dalgarno

sequence (Jacob et al., 1987, Hui and Deboer, 1987). When

an UGA termination sequence is encountered, the rRNA again

interacts with the mRNA (Murgola et al., 1988). 16S rRNA

has also been cross-linked to the wobble base of tRNAs

suggesting its involvement in tRNA positioning (Steiner et

al., 1988). In addition, 16S rRNA probably plays a role in

subunit association, since various regions have been cross-

linked to the large subunit (Brimacombe, et al., 1988).

Furthermore, it is suggested that 16S rRNA is involved in

binding various factors, such as IF3 (Muralikrishna and

Wickstrom, 1989).

























-~ A .*


'4-


UN I T
:A. TjC I7I


'4 '4

A S
'4 *co'oC %c.au.
O c
G '4OC.4CG .'4'4GQi &
~'4.'4'4CCG'4
*q~ -.0
--A '4
'4" '4~. ~


U4 -

"-<:- -. C" *':4"- ; ^ ^

,(C 4'' ,,c~o~.c. 1c Z .- : .-'5 4j*c:; Ujc;





WOBBLE BASE: '' ,- *., *~~cc- Q ,
1OF tRNA ^ '^T--,^ J ^c^
'' U,



ADA .. t


"~A'4 C00I AT
.cc .4400 I 0= FCC\--\
C ROSSI NK --- '4flACA4C'404
WO B E B S '. C-"., A \T 0~\-|
F^ ^ tRN ** i" D.4c0.tc
W ^ ~ ~~ ~ .-'L*. ______




'4*'4Aw -*>-lrb\L G *" Q EIBU



05" '4' :


'4-. C

.4CC 000* i CC '.o4
< 7cMc~ C.C*:., ."<*; 0= ['coD .N


'4


C SITE


Figure 4. Functional Roles of E. coli 16S rRNA (Dahlberg et
al., 1989). Several regions of E. coli 16S rRNA have been
shown to be involved in distinct functions of bacterial
translation as indicated above.


ci:









14

Before studying the mitoribosomal small subunit rRNA,

it is pertinent to ask three questions regarding the

investigation strategy. First, does the mitoribosomal 12S

rRNA perform the same functions as the corresponding 16S

rRNA in E. coli? Second, is the 12S rRNA involved in any

unique properties of the mitoribosome? And third, what is

the most effective method to address the first two

questions?

A comparison of the secondary structures of E. coli

16S rRNA and bovine mitochondrial 12S rRNA (Figure 5)

answers, in part, question one. Figure 5 shows that the 12S

rRNA may naively be interpreted as a truncated version of

the 16S rRNA with many regions of secondary structure

missing, including the Shine-Dalgarno sequence. Therefore,

the answer to the first question is 12S rRNA can not

possibly perform all of the functions of 16S rRNA.

The second question, of whether or not 12S rRNA is

involved in unique properties of the mitoribosome, can be

addressed by asking if 12S rRNA binds mRNAs or if the 12S

rRNA undergoes the same conformational changes that 16S

rRNA undergoes during factor.binding. In order to answer

the first two questions the third problem, of the method of

investigation, must first be solved.

What is the most efficient method for studying the

properties and functions of 12S rRNA? It is possible that























... .... ...... ..

----\? \ .. ; -.. .... ........ ..



-. I "= -.........,, ,, ,'"
i'< ," ..... - *< '*1 : ,
,, -! n i- '=' .' .... ^,.yi





,I ,


















Figure 5. Comparison of E. coli 16S rRNA and bovine
mitochondrial 12S rRNA. Note the analogous structures
between the two types of rRNAs. Also, note that 16S rRNA
has some "extra" RNA, shown in bold.









16

each function of the 12S rRNA would have some effect on its

secondary and tertiary structure. Since techniques used to

study the tertiary structure of macromolecules, such as X-

ray crystallography, are exceedingly time consuming,

expensive and complex, and the limited amounts of

mitoribosomal particles available precludes these methods,

it reasons that secondary structure studies of the 12S

rRNA, under various experimental conditions (each selected

to mimic a different functional state), would provide the

greatest amount of data and understanding in the most

efficient manner.

Secondary structures have been proposed for numerous

16S like rRNA species, including E. coli 16S rRNA and

bovine mitoribosomal 12S rRNA (Guttel et al., 1985). The

proposed structures were modeled on E. coli 16S rRNA, the

only 16S-like rRNA that has undergone extensive

experimental scrutiny. Almost all of the remaining

structures are based solely on computer modeling. The

computer modeling takes advantage of phylogenetic sequence

comparisons and notes regions of secondary structure that

are conserved through compensatory base changes. The

modeling, however, is incapable of taking into account

factors other than the RNA sequence.

To investigate the secondary structure of 12S rRNA,

the RNA was probed, in subunit, using the primer extension

technique (Moazed and Noller, 1986, Stern et al., 1988)









17
under differing experimental conditions. The primer

extension model involves modifying the 12S rRNA in the 28S

subunit with either enzymes or chemical reagents. The

modified rRNA is then purified and annealed to one of

several synthetic oligodeoxyribonucleotide primers spaced

at intervals of approximately 150 bases. Annealed primers

are then extended using reverse transcriptase, which pauses

at the modified bases. Finally, the primer extension

products are analyzed on a denaturing gel to identify the

modified residues.

In the present study, mitoribosomal 12S rRNA was

analyzed, in subunit, using the chemical modifiers DMS

(which reacts primarily with unpaired adenines and

cytosines, and either paired or unpaired guanines) and CMCT

(which reacts primarily with single-stranded uracils and

guanines) (see Figure 6). In addition, the 12S rRNA was

probed in intact subunits using RNAse A (which cleaves 3'

to pyrimidines in single-stranded RNA), RNAse T, (which

cleaves 3V to guanine residues in unpaired RNA), and RNAse

V, (which cleaves primarily double stranded RNA). These

modifiers were chosen so that the data could be analyzed in

a number of ways. First, the secondary structure could be

investigated by the use of single-stranded specific

modifiers (RNAses A and T1, DMS, and CMCT) or a double

strand specific modifier (RNAse Vj). Secondly, the



















A U

H

., >-- c..-c

= "C-
H t '
-C.
4P~


^.N O-H
- \ r/


Figure 6. Site of Chemical Modifications on
Ribonucleotides. Dots represent the sites of modification
by DMS and X's represent the sites of modification by CMCT.
All modifications are blocked by Watson-Crick base-pairing
except the DMS modification of guanine.









19

disposition of the RNA within in the subunit can be

studied. Through the use of relatively bulky enzymatic

modifiers (the RNAses) only those residues exposed on the

surface of the small subunit would be modified.

With an effective method to study the secondary

structure and accessibility of the 12S rRNA in subunit, the

rRNA can be investigated under a variety of experimental

conditions. In this study, 12S rRNA was examined under a

number of experimental conditions, including derived and

native subunits, with transcripts bound to the small

subunit and with IF3 bound to the small subunit. These

conditions enable a number of questions about the 12S

rRNA's role in translation to be addressed. Are factors

possibly co-isolated with the native mitoribosomal

subunits? Does mRNA binding to the mitoribosomal "landing

pad" have a conformational effect on rRNA and if so, is it

specific to each type of mRNA? And finally, does small

subunit binding of IF3 alter the conformation or

accessibility of the rRNA to chemical modification? This

research represents an effort to shed light on these

questions and others.













MATERIALS AND METHODS


Preparation of Bovine Mitochondria


Bovine mitochondria were isolated from fresh liver

(less than 30 minutes post mortem) which was sliced and

kept at 0-4C for the duration of the isolation. The sliced

liver was ground and then homogenized by a flow through

Tekmar homogenizer in isolation medium (0.34 M sucrose, 1

mM EDTA, and 5 mM Tris-HCl, pH 7.5, as described ,previously

(Matthews et al., 1982). Mitochondria were resuspended at a

concentration of 20 mg protein/ml in isolation buffer and

exposed to digitonin at a final concentration of 100 ug/ml

for 15 minutes with constant stirring. The mitochondria

were then diluted 5-fold and collected by centrifugation at

11,000 x g for 10 minutes and washed four additional times

in isolation buffer.



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

mitochondrial lysate and separated in a 10-30% sucrose









21

density gradient in 20 mM MgCL2, 100 mM KC1, 5 mM 2-

mercaptoethanol, and 10 mM triethanolamine, pH 7.5. 55S and

28S fractions were pooled separately, concentrated by high

speed centrifugation, and stored at -70C until needed.



Preparation of Derived 28S Subunits



Concentrated 55S monosomes were dissociated on 10-30%

sucrose density gradients in 5 mM MgCl2, 300 mM KC1, 5 inmM

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

28S subunits were then pooled and concentrated by high

speed centrifugation and stored at -70C until used.



Preparation of IF3



E. coli ribosomes were prepared as described

previously (Denslow and O'Brien, 1978) and were washed in a

high salt buffer containing 1 M NH4ClI, 10 mM MgCl2, 5 mM 2-

mercaptoethanol, and 20 mM triethanolamine, pH 7.5. IF3 was

the generous gift of Claudio Gualerzi..



Preparation of Short Synthetic RNA



A synthetic RNA 30-mer with a 5' AUG sequence was

prepared by Robert Heck following previous methodology

(Lowary et al., 1986). DNA template (127TP6, 200 nM) was









22

incubated with 10 U of T7 RNA polymerase in 40 mM Tris-HCl,

8mM MgCl2, 1mM spermidine, 5mM DTT, 0.01% Triton X-100, 50

ug/ml BSA, 1.9 mM ATP, 1.9 mM CTP, 1.9 mM GTP, and 1.9 mM

UTP, pH 7.5 for two hours at 37C. The RNA product was then

ethanol precipitated, washed and purified on a FPLC Mono Q

column.



Preparation of ATPase 8/6 and Cytochrome Oxidase II mRNA



ATPase 8/6 and Cytochrome Oxidase II mRNAs were

prepared by Bernie Courtney by incubating 2 ug of

appropriate template with 22.5 U SP6 RNA polymerase in 40

mM Tris-HCl, 6 mM MgCl2, 2mM Spermidine, 10 mM DTT, 100 U

RNAsin, 2.5 mM ATP, 2.5 mM CTP, 2.5 mM GTP, and 2.5 mM UTP,

pH 7.9 for two hours at 37oC. The RNA product was ethanol

precipitated, washed four times in 70% ethanol, and

resuspended in 75 mM KCl, 10 mM MgC12, 20 mM Tris-HCl, pH

7.9.



Binding of RNA Transcripts to 28S Derived Subunits



RNA was bound to derived small subunits (at five-fold

excess of RNA to ribosomes) by incubating 360 nM 28S

subunits with 1.8 uM appropriate RNA transcript in 75 mM

KCl, 10 mM MgCl2, 20 mM Tris-HCl, 6 mM 2-mercaptoethanol,

0.01% Triton X-100, pH 7.8 for 10 minutes on ice (Denslow









23

et al., 1989, Liao and Spremulli, 1990). Ribosomes were

saturated by the RNA transcripts as revealed by millipore

assay.



Binding of IF3 to 28S Derived Subunits



IF3 was bound to derived small subunits by incubating

360 nM 28S subunits with 1.8 uM IF3 in 50 mM KCl, 5 mM

MgCl2, 20 mM Tris-HCl, 6 mM 2-mercaptoethanol, 0.01% Triton

X-100, pH 7.8 for 10 minutes on ice. A five-fold excess of

IF3 to ribosomes was used to insure saturation of the

ribosomes (Denslow et al., 1988).



Millipore Filter Binding Assay



After binding the experimental substrate, the

ribosomes were adsorbed to millipore filters and washed

with an additional 8 ml of the particular binding solution.

Washed filters were immersed in scintillant and counted

(Denslow et al., 1991).



Chemical and Enzymatic Modifications



Small subunits (360 nM) were dissolved in substrate

solution (substrate solution is: 20 mM potassium HEPES or

20 mM potassium Borate, 6 mM 2-mercaptoethanol, 0.01%









24

Triton X-100, plus the KC1 and MgCl2 dictated by the given

experiment) and warmed to 35C for 30 minutes prior to

modification. The samples were cooled to 0C and modified

by addition of RNAse A, RNAse T1, RNAse V1, DMS, or

CMCT. The samples were treated for 30 minutes at 0C.

Enzymatic reactions were stopped by bringing the

solution's final concentration to 75 mM NaAc, 2.5 mM EDTA,

88 mg/ml yeast tRNA, and 0.5% SDS. The reaction contents

were then extracted twice with phenol and once with

chloroform. The RNA was precipitated by addition of 2.5

volumes of cold 95% ethanol at -70C for 30 minutes. The

RNA was recovered by centrifugation, washed with 70%

ethanol, and dried. The RNA was resuspended in water to

0.72 nM and stored at -70C until needed.

Chemical reactions were stopped by bringing the final

concentration of the solution to 88 mg/ml yeast tRNA. The

reaction contents were precipitated by addition of 2.5

volumes of cold 95% ethanol at -70C for 30 minutes. The

RNA was recovered by centrifugation, washed with 70%

ethanol, and dried. The RNA was resuspended in 0.3 M NaAc,

2.5 mM EDTA, and 0.5% SDS. The reaction contents were then

extracted twice with phenol and once with chloroform. The

RNA was recovered by centrifugation, washed with 70%

ethanol, and dried. The RNA was resuspended in water to

0.72 nM and stored at -70C until needed. All chemical

reactions were carried out in duplicate on at least two









25

different mitoribosomal preparations. In addition, each

modification reagent was titrated such that approximately

one chemical modification or one RNAse cleavage would occur

every 150 bases (see Figure 7).



DNA Primers



Seven DNA primers were designed to be complementary to

the 12S rRNA such that each was 17 nucleotides long and had

a G+C content of at least 50%. The primers each had one or

more G's or C's at the 3' end of the oligonucleotides for

efficient priming. In addition, the primers were designed

to be complementary to regions of the 12S rRNA that were

spaced at roughly 150 base intervals (see Figure 8). The

oligonucleotide primers were synthesized using an Applied

Biosystems Model 380B synthesizer in the University of

Florida ICBR DNA Synthesis Core Facility. It should be

noted that no information is available 3' of base 910 due

to the proximity of primer A.



Extension of Oligodeoxyribonucleotide Primers



Appropriate primers (5 uM) were annealed to modified

rRNA (35 pM) in 50 mM KCl, 9 mM MgCl2, 20 mM 2-

mercaptoethanol, 50 mM Tris-HCl, 25 uM dATP, 250 uM dCTP,













6E:^-Q

6 c


Figure 7. Titration of RNAse V,. RNAse V1, along with all
other reagents used was titrated such that approximately
one cleavage event would take place every 150 bases (the
length covered by one primer). Note that at high
concentrations of RNAse V1 a complete primer extension is
not possible and that at low concentrations of RNAse V1
there is no discernable difference with the control lane.








































,h, -A O A A AAAA
\,A AA AA

A A
A A-AOJ




A-C4









A-C C Q,4A
a- I
0"1
A A'
-c







ACC





.c ', 'U
A A
,' c, C,.'






C AA











A A cA U -
A A













A-A


Uu A A"u A
A C,














AAC











C AAA
CAA U-




















A-U
A '



G TAA4
fC'w |c-CA



A c^




C A.


C4. 0 C-G




N A A -...A

A 0


A A
AAA


C-C
I-c


Au \ u






A -
A--



A "
A
U AU A

CU N
\CC, C
AA -


: -. ' I
AA00 AC

' '


A C -
A 0VA AA




: --- uA















;-0
G^ II
AAC
A',



-C^ A" A
AA
C A









A- C
:C A









v-U
A A U A




C -C


Figure 8. The Location of Primers on the Proposed 12S RNA

Secondary Structure.


A I
u CA


A

CA
A,'


"'CACG\



U
A' I'* 0\
C A

C AA









28

250 uM dGTP, 250 uM dTTP, 20 uCi a[35S] dATP (1300

Ci/mmol), and 80 U RNAsin. The primers were then extended

using 10 U AMV reverse transcriptase, at 49C for 30

minutes.

Primer extension reactions were stopped by bringing

the reaction solutions to a final concentration of 20 mM

EDTA, 0.08 mg/ml carrier yeast tRNA, and 300 mM NaAc. The

reaction contents were then precipitated by incubation in

2.5 volumes of 95% ethanol at -70C for 30 minutes. The RNA

was then recovered by centrifugation, dried, and

resuspended to 72 pM in loading solution (583 mg/ml

bromophenol blue, 583 mg/ml xylene cyanol, 8.3 mM EDTA, and

10.8 M formamide. Samples were heated to 90oC for 2 minutes

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 50C. The gels were run in TBE

buffer at approximately 60 watts.



Gel Reading Algorithm



In order to determine whether a given nucleotide was

modified, its autoradiographic band intensity was compared









29

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 will be disclosed even though no

modifier was added. This occurrence can be explained by a

number phenomena, including false priming of the carrier

yeast tRNA, secondary structural effects of the 12S rRNA,

false priming of 12S rRNA by other 12S rRNA molecules, and

susceptibility to RNAses during the isolation procedure. In

addition, all data was first recorded against the primary

structure so as not to be prejudiced by the predicted

secondary structure.













SOLVENT ACCESSIBLE NUCLEOTIDES IN THE 28S SUBUNIT


Significance of Chemical Probing



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). Due to these compositional

differences, bovine mitochondrial ribosomes are actually

slightly larger than their bacterial counterparts based on

both particle mass and actual physical dimensions. In this

sense, however, it appears that the mass of the ribosomal

subunits has been conserved, relative to E. coli ribosomes,

by replacement of several RNA structural domains by the

extra proteins contained in these ribosomes. As a result,

the protein-RNA mass ratio in these protein rich ribosomes

is 3.3 times that of E. coli ribosomes (Matthews et al.,

1982). These compositional differences raise questions

related to the structural organization of the proteins and

the diminutive rRNA in these mitoribosomes.









31

Chemical probing of the 12S RNA was carried out in

order to gain insight into the effect of these extra

proteins and to obtain clues to the higher-order structure

of the small subunit. Secondary structures which are

largely inaccessible to chemical modification are likely to

be bound by ribosomal proteins or possibly involved in

tertiary RNA interactions. Secondary structures which are

readily accessible to chemical reagents are unlikely to be

tightly bound by protein. By distinguishing between these

two groups it is possible to gain an understanding as to

which RNA structures may be intimately involved with the

protein mass of the ribosome.

The small subunit RNA was also probed in the small

subunit to investigate differences between mitoribosomes

and eubacterial ribosomes. This study provides a comparison

between the chemical accessibilities of bovine 12S rRNA in

the 28S subunit and similar information for the E. coli 16S

rRNA in the 30S subunit (Moazed and Noller, 1986). Since

the mitochondrial 12S RNA is missing some analogous

structural domains found in E. coli 16S RNA, but has nearly

twice the amount of protein, proteins may be substituting

for these missing RNA regions in the three-dimensional

structure. By comparing the chemical accessibilities of

analogous stems, it could be determined which secondary

structures had a decrease in chemical accessibility

relative to E. coli RNA. These regions with relatively low









32

chemical reactivity might be localized near extra protein-

RNA contact points in the mitochondrial ribosome.,

In addition, probing the 12S RNA provides an

experimental test of the predicted structure. The predicted

structure of bovine mitoribosomal 12S RNA (Guttel et al.,

1985) is based largely on phylogenetic sequence comparisons

and notes regions of secondary structure that are conserved

through compensatory base changes. The modeling, however,

is incapable of taking into account factors other than the

RNA primary sequence, such as the effect of ribosomal

proteins on RNA secondary structure. Since the protein and

RNA composition of the mitoribosome are so different, it is

expected that the predicted structure may need some

refinement by experimental probing. This type of study

should provide the necessary data to test the predicted

secondary structure.





Identification of Solvent Accessible Regions of 12S RNA in
the Small Subunit





In order to locate the accessible regions of the 12S

RNA in the small subunit, 12S RNA was chemically modified

in small subunits with both DMS and CMCT. After a

nucleotide is modified by chemical agents, it can not be









33

transcribed by AMV reverse transcriptase. When titrated

properly, one out of every 150 bases of RNA (the

approximate length covered in primer extension reactions by

each primer) should be chemically modified. During the

primer extension reaction, the AMV reverse transcriptase

pauses at the site of chemical modifications and this

effect is detected as a band on the corresponding

autoradiograph.

Representative results of 12S rRNA modification and

subsequent primer extension are shown in Figures 9 and 10.

Figure 9 displays the 5' region of the 12S RNA and the

chemical modifications in that region. This figure shows

evidence of chemical modifications at nucleotides C15 and

A16, both near the 5' end of the molecule. In addition,

chemical modifications of U34 can also be seen. Figure 10,

covering the area from nucleotide U457 to C549, shows a

number of DMS and CMCT modifications in this area and a

particularly strong DMS modification at A514. In addition,

a number of CMCT modifications are evident in the stem 18A

area from U457 to U478. Figure 11, representing at least

two replicate reactions covering the entire molecule,

presents a summary of all nucleotides, accessible by the

single-strand specific reagents DMS and CMCT. Figure 11

shows several regions of the secondary structure that are

highly reactive with chemical probes. These structures

















0 &. C 0 C

5 iErz-,












... n i'"' .,..



..- -t
.O, -. ., .... . -

























Figure 9. Chemical Modifications 3' to Primer G. Sites of
chemical modification, as revealed by primer extension, are
indicated at the right of the figure. Seq refers to
sequencing lanes, Mod refers to modifications, 0 refers to
no chemicals added, D refers to DMS modifications, and C
refers to CMCT modifications.
-*- --*~~tr

a-A













chmialmoifcaio,^^ as reeae bypie xtnin.r
iniaeda h righ of th figre Seq reest










rFegrs to CMhemiaModifications. 3 oPie .Stso















ODC


C& c. C oc



-a- g


J


on -


-i~UAGI



A>f cO


S' co





-Asc t-
",-_. I.5-t c






. M55
NA- ,A.,sc




NA50


Figure 10. Chemical Modifications 3' to Primer C. Sites of
chemical modification, as revealed by primer extension, are
indicated at the right of the figure. Seq refers to
sequencing lanes, Mod refers to modifications, 0 refers to
no chemicals added, D refers to DMS modifications, and C
refers to CMCT modifications.


























S AAA U AAA /
A
AU "

eA
AA A. A
AU". 15 A

15o "a
/ A* UY A LA/, W
aeCuIJA* 4-C
14 a-
A-u
LA-A CL


.A

A



0 -0
GCc
L*A
C C


A U
A-A

A 11 U--





AO AU AAU-A
:"-,







K%
OA Ao *0 :A U. A
10% 4-




9\ A-
A





% A 0 .UA
*U U



AAOO L A o A 4b-




9- A 5A
9a~~- t~%eAu


A A
Ou A A U.





=O, A, r I CU
O U ,A 5



A W 7-A-U







A'-Q-
A--
A A S

9 A-
** L' , C



te t ^--
\ IA, A 14.-J--




A-U
A A
LA


0,.
u A,
41-c A*
O'Y _" Al
4-c a
.o-c .8

O*J C. Un
*AA

A7 A A
0.- ** A.i
A O A
7 4 AS 0 A
A


20


AL



AU-, 26a
c-4 Co
u u
C-U,

r,-a
25 o-o to'
A
CAA

4 A--U


* C
\ U
c-4









24
u+ u













22
C,- C.


* LA U*

A .*

A A


A A

C A

9 28




C-4 A
C-A C
30 _, 3
L A 4
C C- A

C A A
C-A i
--L-A C..
29 C- A

C-A A



G-A U
6-4 4 i~ u~


.30 _*UiL 233
UAC A
Y 0, A
C A U
A-u
4-.
231 U-
G-C
0-A
-A C
1 A
C A
U-A
32 -

A-
0 u


4
/A

++ A


*
,A C
C


C k AU AA
A aL


U A



A7 0
A7 C A CG
AA
-7 A
A7 AA.


Figure 11. Location of Chemical Modifications on the 12S

Secondary Structure. Dots indicate the sites of

modification by DMS and CMCT plotted on the experimentally

refined secondary structure. The algorithm for refining the

structure is outlined in the text. No data is available 3'

to nucleotide 910, due to the proximity of primer A.


6.UO









37

include stem 8, the stem 9A area, and stems 18, 23, and

26A. The relatively high reactivity of these regions

suggests that they are not obscured by proteins. In

addition, it seems likely that the most accessible RNA

structures would be the ones that would serve functional

roles. For example, the E. coli 16S RNA analogue to

mitoribosomal stem 18 is highly accessible to chemical

attack and has been shown to be involved in subunit

association (Brimacombe et al., 1988). Since the analogous

region in both ribosomes is highly conserved in sequence,

structure and chemical accessibility, stem 18 (in

mitoribosomes) is probably also involved in subunit

association.

Along with highly accessible structures, Figure 11

shows a number of areas that are quite resistant to

chemical probing. These regions include stems 11, stems 27

through 32, the single stranded region between stems 20 and

29, and the bulge loop between stems 24 and 25. Apparently

these structures are tightly bound by protein or are

involved in tertiary interactions of the RNA. Since these

regions are so inaccessible to modification by the

relatively small chemical reagents, these regions may not

be directly involved in functional roles of translation

proper; however, they may be playing important roles in the

assembly of the ribosome and the maintenance of its three-

dimensional structure.















Environmental Dependence of Small Subunit Solvent
Accessibility



After having determined the chemical accessibility of

derived subunits in solution, questions were raised about

what effect different environments of the small subunit

might have on the chemical reactivity of the 12S RNA. Would

the 12S RNA be as accessible in the native subunit?

Native subunits were also probed with chemical

reagents in an attempt to discern differences between

native and derived small subunits. During the isolation of

mitoribosomes, the ribosomes are purified by centrifugation

into two sucrose density gradients which are monitored at

260 nm. The difference between the native and derived

subunits is that in the first gradient monosomes and native

subunits are isolated (see Figure 12). The monosomes are

then concentrated and resuspended in a higher monovalent to

divalent cation solution, which dissociates the subunits.

The derived subunits are then isolated from the second

sucrose density gradient.

Nucleotide A514 shows consistently enhanced reactivity

with DMS in the native subunit (see Figure 13) as compared

to the derived subunit. Since native subunits show

consistently enhanced reactivity at nucleotide A514 with

























































Figure 12. A260 Profiles of Ribosomes in Sucrose Density
Gradients. Native subunits and monosomes are isolated from
the first gradient. Derived subunits are isolated from the
second gradient.
























( 'N\ 0 0 0 D


Figure 13. Chemical Modifications 3' to Primer C in the
Native Subunit. Sites of chemical modification, as revealed
by primer extension, are indicated at the right of the
figure. This figure shows the enhanced reactivity at A514
in both the monosome and native subunits.


- k5sk'









41

DMS as compared to derived subunits, this result can be

interpreted in at least two different ways.

The native subunits were probed in the hope of finding

additional evidence that native subunits are isolated with

a bound initiation factor 3-like molecule. Previous

evidence to support this theory included the presence of

additional proteins on two dimensional gels of native small

subunits as compared to their derived counterparts. In

addition, native subunits can not be reassociated as

derived subunits can. If present, an IF3-like factor may

serve to make A514 more accessible. This region, which is

thought to be located in the cleft of the small subunit

(see Development of a Three Dimensional Model) may be

involved in important functional roles such as mRNA binding

or tRNA decoding.

The second explanation is that the native

configuration of the small subunit may actually be a more

realistic representation of the chemical reactivity than

that form the derived subunits. Since the native subunits

are subjected to the dissociation salt wash, this may

explain the difference rather than any special conformation

imparted by potential initiation factors.









42

Comparison of Chemically Accessible Regions of
Mitochondrial 12S rRNA and E. coli 16S rRNA





A general comparison between mitoribosomal 12S RNA and

E. coli 16S rRNA can be made using the chemical

modification data in this study, using DMS and CMCT, and

the observations of Moazed and Noller (1986). For this

comparison, guanosine nucleotides were disregarded since

the available E. coli data were based on a mixture of DMS,

CMCT, and Kethoxal modifications. A summary of the data of

Moazed and Noller (1986) is shown in Figure 14. Guanosines

modified by CMCT alone were not distinguished from those

modified by kethoxal. In making the comparisons between the

two ribosomal species, only moderate to strong

modifications were considered. Analogous secondary

structural regions between the 12S RNA and the 16S RNA were

identified by visual inspection of the secondary structure

proposed for these two rRNAs (Guttel et al., 1986). The

stem numbering scheme used in this text for E. coli 16S RNA

is according to Brimacombe (1988).

Figure 15 is a schematic representation of both 28S

and 30S subunits. This figure shows the relatively high

abundance of proteins for the 28S subunit (33 proteins with

an average mass of 24.9 Kd for 28S subunits versus 21

proteins with an average mass of 19 Kd for 30S subunits)













43




7*o,

66000 ^006 0-8 e 11 i
', 'o *o oo r 1
00- 0600002-000
... ... Oeoeo "oo K ooo* "' oo*'.oi -o''.:- .
Iii.oo oooo,. o ooo oo Me w i. .
S00 |00 A0 -0 I 0 l "
I000 .. .. . .o, ,I I-
S*G**o 00 00 o 96 "
-... oo o 000 0
0-00
0-0 0 0000 0 0 :- Iwo
0 -.0-000- -0 .
0-o - 0- oo -0 0 0
0'0 0-064 0 O OS

1- 0 0- 0
:0, o' 7 .



00 -. 0 000 00 900000. 0000 0 00I
@0. .- 0 e o0 0- : *0-0
. ..,;.. "' o .- - .o-
0600 0.00. o ,*@ 3 .. ,0, 9. 0 o
000000000 000 00 0 0 e' o 01 0
oooo O a000- 0 z\00 0 0 o 0 0 0 0" 0 0l
m. c .. o .o o -+ -o
.:o 00000 0 0.; 0 ,0
0o," 00. 00 B -oO O
00000000.00000 0 0.0 : *
0 0 at. 'O 0 -
00..
0 0 0 0 0 0 0 0 0 0 %.00
00 0 00 0000 '0 j0 -.07
o o 0 0 -s 0 9oO". o ..e. .
.,. o S +. oO-- :'oo,_. *.0. o
o t a 00 0 0.0 0 a 00 0
0 *00 0 00000090 -. 0tt 00 .. .,0 0 o- a
900 0-
0 't .... 0. 00 00 0 0 ao
o_ 0 +0o0000 00 0 00 0-,0 -
0 / 0 o o o o-, o


o o .. 0 ,0,0 0 o. o ,0 0


% .,go g Oo % o, +: o -.o+


.0 S 0 .0
oDo
*-0 0 000 0 o 0 -0 -oB r
00 o -00 0 0 0oooo-*









a^ a* -F ""
*0o 0 000 .0 0,o-o 0 I ,, 0
0. o 0;' -000t .. 00.. 0. ..... 0 .."
.."o
u +oo o-o o 0o. o0 oooooooeoll o 0 0 ioi

000, 0
000 0 0 ,s0 0 0




" .-Do' -
o o" @-o- .05 o g o



0000 00-0 0. 0. 00 .^0
0 o00 0 0 00 0. 00.. 0 fill



'00 0 00 t 000,0
0 0c o ,, *O o 3 0o to0 0
p0tG.0 0 0 jO 000// 0 lm0 000


.oo "+t-' o --', 0-




o0 0-0 o ;-0
0 00 000 -- 000
0. o00 oo o


1 0 , -.


o .o o s -s: ,.
00 o 0 -00
0 0 0000 0 0 *.0
0 o o-o i
Oooo ,+o *0 0-0 .




S .000 Ste.: 0of 0- 0 0
al l 00,i 00 00 0 0

009 600. o .0 09 0 000 0 0000
oo ....9000 .90.000 0
t 0 0 .o






0-0




0000r 14. sumar ofteCeia oiictoso .g-
0 .'-@ 0 0- H 0 x 0 -0-0 ,. e












16 0RN Mo0d n 0-.r 1986.tOe irlsrersn














represent nucleotides that are moderately accessible, and
triangles represent nucleotides that are readily
accessible.
i00 0
^o0+', o % 0o-
0'++oo-0


000o. ,o% 0o'-' .o-

0,0., 0 ....,
o~ *.o '
02: o o +
ll5.00,i 40 0 0 ,000 O0O0.i










Figure 14..Summary of the Chemical Modifications on E. cl
16+S +RNA (Moaz+ed and Noller, 1986). Open circles represent
nucleotides that are not modified, small dots represent
nucleotides that are marginally accessible, large dots
represent nucleotides that are moderately accessible, and
triangles represent nucleotides that are readily
accessible.



















0000 000000
0000 000000
0000 00D000
000 000000
000 000000
000 000


Figure 15. Relative Protein and RNA Masses of 28S and 30S
Subunits. E. coli ribosomes (left) contain one 16S rRNA
molecule and 21 proteins of an average molecular weight of
19 Kd. Bovine mitoribosomes have a single 12S rRNA molecule
and 33 proteins weighing, on average, 24.9 Kd each. The
areas of the circles are in proportion to the average
molecular weights of the proteins.









45

and its consequently low amount of rRNA (Matthews et al.,

1982). From these data it can be proposed that the RNA and

proteins are arranged in one of two ways. The presence of

extra proteins may provide additional protein-RNA contact

points and thus effectively swamp the RNA (see Figure 16).

In this case, the 12S RNA would be relatively inaccessible

to chemical agents as compared to the 16S RNA. The other

possibility is that the protein-RNA contact points are

essentially the same in both kinds of ribosomes and the

extra proteins would be added to the "outside" in

additional protein-protein interactions. In this second

case, it is expected that the reactivity of the RNA would

be somewhat similar in both types of ribosomes.

In general, the E. coli 16S rRNA in subunit is less

accessible than the mitochondrial 12S RNA in subunit. While

the E. coli ribosome contains appreciably more RNA than the

mitochondrial ribosome, a higher percentage of 16S RNA

comprises stem structures (59% of 16S RNA is base paired

compared to 45% for 12S RNA). From a consideration of the

secondary structure of the RNA it is reasonable that a

higher percentage of the 12S.RNA will be accessible to

single strand chemical modifying agents. However,

considering that the mitoribosome has nearly twice the

protein mass of the E. coli ribosome, it is surprising that

so much of the mitoribosomal RNA is chemically reactive.

Given that the mitoribosome has a relatively high protein






















































Figure 16. Models for the Arrangement of Protein and RNA
Mass Within the Ribosome. The additional proteins of the
mitoribosome may be arranged on the outside of an existing
core structure (left), thereby maintaining the majority of
the RNA-protein contact points or the additional proteins
may be in direct contact with the RNA, effectively swamping
it.









47

content and its RNA maintains a high chemical reactivity,

the most likely conclusion is that the mitoribosomal

proteins must be arranged in a three-dimensional structure

such that the majority of their primary interactions are of

a protein-protein nature rather than of the protein-RNA

type, and thus should not obscure chemical access to the

rRNA.

It has been speculated that the extra proteins might

be substituting for the "missing" RNA of mitoribosomes. To

investigate this possibility, areas of secondary structure

in the 16S RNA which have no analogue in the 12S RNA were

identified. These missing structures are located on the 12S

RNA in the following areas: a missing stem on the loop

between stems 5 and 6, three missing stems at the loop end

of stem 7, a missing stem between stems 8 and 6, a missing

stem between stems 6 and 9, two missing stems between stems

4 and 11, a missing stem on the loop between stems 13 and

14, two missing stem on the loop end of stem 18A, a missing

stem on the loop between stems 24 and 25, two missing stems

on the loop end of stem 26A, and a missing stem on the loop

end of stem 26B. The location of these structures is

indicated in Figure 5. If these regions were replaced by

protein, it might be expected that the neighboring RNA

regions would be less chemically reactive than their E.

coli counterparts. Of the above candidates, only one, the

region between stems 24 and 25 fits this criterion. This









48

region of the subunit is quite likely proteinaceous whereas

its E. coli counterpart is probably RNA rich. The remaining

regions mentioned above are all more accessible in the 12S

RNA than the 16S RNA. Two possibilities exist for these

areas. First, the "loss" of the dispensable RNA structures

leaves these regions exposed to chemical modification and

the extra proteins are arranged elsewhere in the three-

dimensional structure. The second possibility is that extra

proteins may be "filling space" in these areas without

being intimately associated with these RNA structures. This

study can't distinguish between these possibilities.

However, the second possibility, that the "extra" proteins

reside in the regions of missing RNA structures, seems more

likely based on the similar appearance of the ribosomes

under electron micrographic examination.

In comparing individual stems between the two

ribosomal species, a number of differences are noted

between the accessibility of E. coli ribosomal and

mitoribosomal RNA; however, only a few regions are sharply

contrasted (see Figures 11 and 14). For example, stem 18A,

stem 19 and stem 23 are much-more susceptible to chemical

modification in the mitoribosome than the corresponding

stems in the E. coli ribosome. However, the single stranded

region between stems 20 and 29 is more accessible in E.

coli. The differential chemical reactivities suggest that

although the secondary structure of the RNA is conserved,









49

the environment in which the two different species of RNA

lie differs considerably. The differences between the two

types of subunits with respect to stems 18A and 19 are not

surprising because the sequences of these structures is not

highly conserved. The differences in chemical

susceptibility between stem 23 and the single-stranded

region between stems 20 and 29 is, however, surprising

because of the high nucleotide conservation in these areas

(17 out of 28 conserved bases and 12 out of 17 bases,

respectively). Although these regions are conserved and may

be functionally important, the protein structure around

them may not be. Therefore a major local structural

difference in these proteins is the most likely cause of

the accessibility differences. The "additional" proteins

found in mitoribosomes may act in a number of ways to

explain the differential reactivities of the nucleotides.

The extra proteins may be located on the outer surface of a

particular region, thus masking the RNA. This situation may

be occurring in stem/loop 28. In addition, the extra

proteins may be located towards the interior of the subunit

and act to force the 12S RNA into more accessible

conformations, perhaps as in the case of stem 18A.

Although the two species of RNA show differences in

chemical reactivity in some regions, it can not be

concluded that these regions don't perform analogous

functions. However, if two regions are conserved in both









50

primary and secondary structure, as well as in chemical

reactivity, it seems likely that these portions of the RNA

share analogous functions. That is the case with stem 18.

Stem 18 and its E. coli analogue, stem 24, are

conserved in terms of both chemical susceptibility and RNA

sequence. This structure is proposed to have important

functional roles, being involved in the decoding site and

in subunit association in E. coli (Dahlberg, 1989).

Considering the conservation of structure and chemical

reactivity, it is likely that mitoribosomal stem 18 has the

same subunit association function as does its E. coli

counterpart as well as being localized to the tRNA decoding

site like its E. coli analogue.





Refinement of the 12S rRNA in subunit Secondary Structure





Experimental deviations from the predicted structure

of Guttel were divided into two categories: predicted stems

which were accessible to the single-strand specific

modifiers on both sides of a predicted stem, and predicted

stems which were accessible to single-strand specific

chemical modifiers on only one side of the stem. Of the two

categories, only the first (stems accessed by single-strand

specific reagents on both sides) was used to refine the









51

secondary structure. Stem 18A was the only major region

which fell into this category.

Figure 11 shows the results of probing with CMCT in

the stem 18A area in subunit. Inspection of Figure 10

reveals uracils 457, 460, 461, 474, 475, 477, and 478 are

each modified. In addition, a densitometric profile of this

region is shown in Figure 17. Since stem 18A was accessible

to the single-strand specific reagent CMCT at each uracil

in the stem (see Figure 10), this suggests that this stem

is not present in the subunit or it is under going

substantial breathing. Both potential structures are shown

in Figure 18. Unfortunately, this study can not distinguish

between the two possibilities. Comparing this secondary

structure to the analogous structure in E. coli reveals

that the 12S version differs dramatically in primary

structure. Stem 18A is extremely A:U rich, whereas its

counterpart in E. coli is G:C rich. In addition, the 12S

secondary structure is a greatly foreshortened version of

the 16S. analogue, suggesting whatever purpose is served by

this RNA structure in E. coli has become largely

unimportant or is being served by another part of the

mitoribosome. Although stem 18A may not be present in the

small subunit, it is possible that this stem may form and

be important during the assembly of the ribosome. Since

little is known regarding the assembly of mitochondrial




























5x-cr i-aA







c/ \n


















Figure 17. Densitometric Analysis of Chemical Modifications
3' to Primer C. The presence of a peak in the lower profile
greater in height than that in the upper profile indicates
a modification by CMCT in the stem 18A region.





















A--U
U-A
400 C-G U A
\ uAG-C C C
U


CACAG AUA
CA A


18o
A AU

AA AC A
AAuG-c UUAU U A A A C
C-G-
G-C
G-C


A-U
U-A

C -C
AU AG- C
UA


G A L .J
A U U A ... '
SAUG-C
C-G-
G-C
G-C


- 450


CA
A U
IJIJ A
A

A A
AC

18a


Figure 18. Refinement of the Secondary Structure of Stem
18A. The refined structure, which is supported by the
experimental data, is on the left. The structure on the
right is the original predicted structure (Gutell et al.,
1985).


U
U
A
C











ribosomes and no reasonable test yet exists for the

mitochondrial system this possibility cannot be dismissed.

A third possibility for the ready access of stem 18 to CMCT

modification is that the isolation procedure itself may

disrupt stem 18A.

Other than stem 18A, only minor refinements were

required of the predicted structure. These additional

refinements include flaring the ends of stems 4 and 9A, and

removing the three base pair stem 15 due to the single

strand specific modifications in these regions (see Figure

11). Since these adjustments are minor and restricted to

the ends of stems (in the case of stem 15, both ends), they

probably represent stem breathing in these regions.

Predicted stems which were susceptible to single-

strand specific reagents on only one side include stems 7,

26A, and 26B (see Figure 11). The reactivity of these stems

with single-strand specific modifiers may also be the

result of breathing. Most double-stranded nucleic acids are

dynamic and have a tendency to breath. During a breath,

stems may be reactive with single-strand specific reagents.

In addition, base pairs at the ends of stems (such as those

mentioned above) are especially prone to breathing since

they are stabilized by base stacking energy from 'only one

side.















DEVELOPMENT OF A THREE DIMENSIONAL UNDERSTANDING OF THE
SMALL SUBUNIT





Significance of Three Dimensional Modeling





In order to gain insight into the spatial structure of

the small subunit, a three-dimensional computer model of

the bovine mitoribosomal small subunit was constructed. The

three dimensional model of the small subunit was generated

taking advantage of phylogenetic sequence comparisons and

secondary structure comparisons between mitochondrial and

E. coli rRNA (Guttel et al., 1985), and electron

micrographs of the small subunit of mitoribosomes (O'Brien,

in preparation). By visual inspection of the secondary

structures of E. coli 16S rRNA and bovine mitoribosomal 12S

RNA, 12S RNA analogues for most of the 16S RNA secondary

structure domains were found, while some secondary

structures in the 16S RNA had no apparent analogue in the

smaller 12S RNA. By examination of electron micrographs of

various perspectives (provided by T. O'Brien), a three

dimensional clay model of the small subunit was









56

constructed. Topographic elevations of the clay model were

determined and entered into the computer. Finally, by using

the models of both Brimacombe (1988) and Noller (1988) for

the three dimensional dispositions of 16S rRNA stems,

extrapolations were made to place the analogous 12S rRNA

stems into the computer model of the small mitoribosomal

subunit. This model is not forwarded as the absolute

structure of the small subunit, but rather as a best

current estimate based on available information, since it

seems to be a reasonable assumption that conserved

structures should occupy similar spatial domains relative

to each other, within the very similar space envelope of

each ribosome subunit.





Identification of Surface Accessible Regions of 12S RNA in
the Small Subunit





Information on the disposition of 12S RNA in the

subunit can be gained by examining the enzyme cleavage

sites. Chemical modifiers are small molecules which may

penetrate the 28S subunit to modify internal residues. The

RNAses however, are relatively large molecules and their

action should be confined to RNA residues exposed on the

surface of the subunit. Figure 19 displays those






















U c-,


6('4t -


~1DO
N


- 4w~m~
,m-w,.


*p ~(A~C
I. -'~A~











~ -KA7-~


Figure 19. RNAse A/T1 Cleavages on Stem 18. Sites of
enzymatic cleavage, as revealed by primer extension, are
indicated to the right of the figure. Seq refers to
sequencing lanes, Mod refers to modifications, 0 refers to
no enzymes added, A refers to RNAse A/TI.


mf -
^^^^^f OWNtnhb "eug^









58

nucleotides in stem 18 that are exposed to RNAse attack,

notably C410, C411, C412, U423, and U427. Note the apparent

gradient of RNAse activity in the C410 to C412 region,

perhaps due to the fact that C410 is closer to a predicted

stem than is C412. Figure 20 is representative of the stem

15 area and shows that U338, U352, U357, and U363 are all

accessible to RNAse attack, suggesting these areas are

single-stranded. In addition, the stem 1/ stem 2 pseudoknot

also appears to be accessible to enzymatic cleavage. Figure

21 shows that C15 and A16 are accessible to RNAse A (it

should be noted that adenines are not normally susceptible

to RNAse A attack and A16 may actually be a naturally

occurring analog of adenine which is susceptible to RNAse A

attack), whereas A16, G17, and C18 all appear to be

accessible to RNAse V1. Since A16 is accessible by both

RNAse A and V1, A16 may be "breathing", that is in

transition between single and double-stranded states. It

should also be noted that the precise recognition pattern

for RNAse V1 is not double stranded RNA, but rather helical

RNA (which usually infers a double-stranded state),

therefore, A16 may not be double-stranded, but simply

retaining the helical conformation of its double-stranded

neighbor, G17. Figure 22 shows that many of these surface

nucleotides are scattered throughout the molecule. There

also exist clusters of RNAse-accessible bases on specific



















K) CA C-


', 0 \


-:




,~

4


-U35{




- U3&3


Figure 20. RNAse A/T, Cleavages on Stem 15. Sites of
enzymatic cleavage, as revealed by primer extension, are
indicated to the right of the figure. Seq refers to
sequencing lanes, Mod refers to modifications, 0 refers to
no enzymes added, A refers to RNAse A/Ti.


\Y$~% ~bV


- 033s?


















5i-(^ A o0T.












OPPWI)
\



























..Sites of enzymatic cleavage, as revealed by primer








extension, are indicated to the right of the figure. Seq
refers to sequencing lanes, Mod refers to modifications, 0

refers to no enzymes added, A refers to RNAse A/T1, V
refers to RNAse V1.





























P
IA AAUA A
,A A

A
AA? 15

US A
A

15a %
/ AA A A -'
"HOUUA* / "-4


20

C U U.


A -
C'*


U- JA MeiU U I
AA

C A

U..A 26.
4u

25 SV
C-4 A
0-0

--A-U
A- A A

_U
-A

*u \ u*
a u .;
QUU-A
c'"";--
C-U
A A-



Cc uC- --M


A -A 22
A-U
U A.4
:- A


^
a 22



*-. a .- a.e"^ , UiJ A
'A
A 26
U


A A
u u




9a U, A
A A UA U

9. :UA O"A A
A

0-4-A
-C A 5

ACa/ ,A,
c aC








A~g

A~A
U A 8
AU A

0%


U, A go- A-U r

A A7
I* -4<


#-IA
A A
A


U @
C A C
A A_ 0
A U
A A
.-0 A

29 c-4 A
". A
U-A


A-U
A A U A

A0 0, 1 "


U C A
C A U
A-U
O-C
U1 ]-A
G-C
C-4
-A 4
U A
C A
V-*
32
SVw
0 A


A', 1, AA AC A

a C A 0
AU
U, u

U A"


Figure 22. Location of Enzymatic Cleavages on the 12S RNA

Secondary Structure. Crosses indicate sites of RNAse V,

cleavage and triangles indicate sites of cleavage by either

RNAse A or Ti. No data is available 3' to nucleotide 910,

due to the proximity of primer A.









62

secondary structures. Such accessible structures include

stems 18 and 15. In general, the 12S rRNA appears to be

essentially a buried molecule, with limited but specific

surface nucleotides. The relative inaccessibility may be

accounted for by the relatively high protein content of the

mitoribosomal subunit, containing nearly twice as much

protein as the E. coli 30S subunit.

Areas of interest which seemed accessible to enzymatic

cleavage include stems 15 and 18, which probably make up

the outer rim of the small subunit shelf (see Figures 19,

20, and 28). It should be noted that the counterpart to

stem 18 in E. coli has been implicated in a number of

functions,including subunit association. Other regions

which are surface accessible (as determined by RNAse

cleavage) include stems 5 and 30.





Development of a Three Dimensional Model of the Small
Subunit from Electron Microscopy





In order to develop a better understanding of

ribosomal functioning and structure, it was felt that two

dimensional modeling was insufficient and therefore the

existing data on mitoribosomal small subunits was pooled to









63

evolve a three-dimensional model of the mitoribosomal small

subunit.

To develop a three dimensional model of the small

subunit, electron micrographs of various views of the small

subunit were used to construct a clay model of the subunit

(see Figure 23). By comparing the evolving clay model, in

profile, against electron micrographs, reshaping the model,

and repeating this process hundreds of times, the hard clay

model shown in Figure 24 was created. After development of

the clay model, it became necessary to transform it to a

computer image so that it could be readily manipulated and

correlated with other data. To create the computer image,

first a duplicate soft clay model was constructed and then

sectioned. Each of these sections were traced on cellulose

acetate sheets and then the acetate overlays were mounted

on the computer screen. Following the mounting, the

coordinates for each tracing were entered into a Gateway

2000 386 computer with the use of a Logitech mouse and

AUTOCAD 10.0, a computer-aided design software package.

Figure 25 shows the tracings used to develop the model.

After each tracing was entered, the layers were assigned

elevations according to their positions in the clay model

(see Figure 26). The longest dimension of the small subunit

was estimated to be approximately 290 angstroms.



























































Figure 23. Electron micrograph of 285 Subunits. Note the
various views of the small subunits presented (O'Brien, in
preparation).



























































Figure 24a. Three Dimensional Model of the Small Subunit
(domal view). Some of the surface accessible stems have
been added to the model in pencil (see Figures 29).


























































Figure 24b. Three Dimensional Model of the Small Subunit
interfaciall view). Some of the surface accessible stems
have been added to the model in pencil (see Figures 29).



























































Figure 25. Topographic Layers of the Small Subunit. These
layers represent the individual sections of the sliced soft
clay model. The shelf is towards the upper right of this
figure.



























































Figure 26. Three Dimensional Computer Model of the Small
Subunit. This view of the small subunit is from the
interfacial side, tilted 20 from the X-Y plane, towards
the viewer.









69

Localization of Surface Accessible 12S RNA Secondary
Structures Within the Small Subunit





To make the newly constructed computer model more

useful, secondary structures which are surface accessible

were entered into the model. In the absence of crosslinking

and neutron scattering data, analogy to the E. coli 30S

subunit data was relied on. The analogue to each of the 12S

structures in 16S rRNA was located by visual inspection and

a few examples of this are shown in Figure 27. A complete

listing of analogous secondary structures is shown in

Appendix II. In order to more accurately place the surface

accessible stems in three dimensions by analogy with the

30S subunit, two existing models of the 30S subunit were

utilized. Inspection of the two models shown in Figure 28

shows that stems placed in three dimensions by correlating

neutron scattering data with either RNA crosslinking or RNA

footprinting result in essentially similar models. If a

given stem was shown to crosslink to (or be footprinted by)

two or more proteins, that stem was tilted between the

proteins to reflect this fact. In addition, Noller's model

took in to consideration the distance between stem governed

by the single-stranded linkers. The Brimacombe model did

not take this factor into consideration. After compiling a

satisfactory hybrid model for 30S subunits from the data of























26A
, ,18 *4 -



: i8A 24




... .. 1": . r ......


30 RNA





12S RNA


16S RNA


Figure 27. Identification of Analogous Secondary Structures
Between Bovine Mitoribosomal 12S rRNA and E. coli 16S rRNA.
A complete comparison between individual secondary
structures is listed in Appendix II.


'' '. '"




, ;: = .






















































Figure 28. Comparison of Two Independent Models of the 30S
Subunit. The Noller model (Stern, et al., 1988), left, is
based on RNA footprinting. The Brimacombe model (Brimacombe
et al., 1988) is based largely on RNA crosslinking studies.
Neutron scattering was used to determine the relative
positions of the proteins in both cases (Engelman and
Moore, 1972).









72

Noller and Brimacombe (Stern et al., 1988, Brimacombe et

al., 1988), analogous surface accessible stems of the

mitoribosomes were located in the newly constructed

computer model of the 28S subunit. For example, the E. coli

analog to stem 18 is found in the shelf on the interfacial

side in both E. coli models therefore stem 18 was placed in

the shelf on the interfacial side of the mitoribosomal

model (see Figure 29). In order to proportion the stems

correctly, the mitoribosomes were estimated to be

approximately the same size and shape as their bacterial

counterparts (based on electron micrographic data) and the

stems were assumed to be roughly 20 Angstroms wide. Since

there is no supporting data for the tilt of stems within

the mitoribosomal subunit other than analogy with the E.

coli system, no tilt or rotation was given to the stems in

the mitoribosomal model. In addition, the loop ends of

stems are represented by open circles. The resulting model,

with surface accessible stems in place, is shown in Figure

29.

This figure does not represent the absolute

configuration of these stems in the small subunit, but

rather the best estimate given the available data. For

example, it can be stated with some confidence that stem 18

is surface accessible due to its reactivity with RNAse, its

loop end is toward the tip of the shelf and its body is

located within the shelf, on the interfacial side nearer





























1'
U


Figure 29. Location of Surface Accessible Secondary
Structures Within the Three Dimensional Model. This view is
from the domal side of the small subunit, with the subunit
tilted 20 away from the X-Y plane towards the viewer. Each
layer is approximately 12 angstroms thick. Open circles
represent the loop ends of stems.









74

the medial aspect of the subunit. However, exact

coordinates can not be given, nor can items such as the

degree of rotation of the stem be accurately estimated.















THE EFFECT OF RIBOSOMAL SUBSTRATES ON NUCLEOTIDE
ACCESSIBILITY IN THE SMALL SUBUNIT





Significance of Ribosome-Substrate Interaction





Having gained a better understanding of the structure

of the mitoribosomal small subunit, it became desirable to

determine how the structure was altered, if at all, during

ribosome-substrate interactions and to see what conclusions

could be drawn regarding ribosome function from changes in

the chemical and enzymatic reactivity of the 12S RNA. Two

binding interactions were studied, the binding of the small

subunit with RNA transcripts and the binding of initiation

factor 3 by the small subunit.

The binding of mRNA to the small subunit is thought to

occur in a region roughly thirty bases in length (see

Figure 3). This estimate is based on studies of binding

affinity versus oligonucleotide length (Denslow et al.,

1989). Affinity of RNA for the small subunit increases with

increasing length up to a length of thirty bases. After

this point, increased length of the RNA substrate does not









76

raise binding strength. In addition, studies involving the

use of RNAses show that up to 80 bases are protected by the

small subunit. From these data, it can be surmised that the

RNA binding site on the ribosome is approximately 30 bases

in length and that an additional 50 bases of the substrate

are protected from RNAse activity, presumably by 'steric

interference from the small subunit.

These data raise a number of questions. Does the

binding of RNA to the small subunit induce conformational

changes in the ribosome or block accessible residues from

modification? If so, are these conformational changes

transcript specific, or are they general changes, induced

by every transcript? Also, what is the binding site

composed of RNA, protein, or both?

Little is known about the interaction of factors with

the small subunit. No initiation factor 3-like protein has

yet been isolated from the mitochondria, but E. coli

initiation factor 3 binds mitoribosomal small subunits with

nearly the same affinity as it does with E. coli small

subunits (Denslow et al., 1988). In addition, E. coli IF3

inhibits mitoribosomal subunit association. Since E. coli

IF3 appears to function in the same manner in both

translation systems, it can be used to investigate the

nature of the IF3 binding site on the small subunit. This

study could provide clues to answer a number of questions.

Is the binding site mostly proteinaceous or is it composed









77

primarily of RNA? Does IF3 induce a detectable

conformational change in the small subunit? Finally, how

does the conformational change induced on the mitoribosomal

subunit (if any) compare the conformational change induced

on the E. coli subunit?









Effect of Transcripts on 12S RNA Reactivity





In order to study the effect of transcript binding on

the structure of the mitoribosomal small subunit, two

methods were employed. First, the small subunit was

preloaded with ribosomal substrates (as determined by

Millipore filter assay, see Methods) and then probed with

chemical and enzymatic agents to determine whether or not

any structural changes induced in the small subunit could

be detected through its RNA. Also, ultraviolet crosslinking

of a photoreactive transcript to the small subunit was used

to gauge whether or not any portion of the RNA binding site

on the small subunit was composed of rRNA (in collaboration

with Robert Heck).

The transcripts used to study RNA-subunit interaction

were the cytochrome oxidase subunit II mRNA and BH, a short









78

oligonucleotide (see Figure 30). BH was chosen to evaluate

the earlier binding versus length studies, to see if

binding exclusively to the "RNA landing pad" had any

conformational effect on the small subunit. BH was designed

to be free of any innate secondary structure, and contained

an AUG codon near the 5' end to simulate more closely a

short mRNA. COII mRNA was selected for a number of reasons

(in collaboration with Bernard Courtney). The use of COII

mRNA served the purpose of using a "natural" mitochondrial

RNA (see Figure 31). Would the conformational effect (if

any) be dependent on the mRNA species? Would the .extra

length of the natural mRNA have additional conformational

effects on the small subunit? Can the conformational

effects specific to COII mRNA (also, if any) be competed

for with poly(U)? These were some of the questions this

portion of the project attempted to address.

BH had no apparent effect on the accessibility of 12S

RNA to enzymatic or chemical agents. This suggests several

things. The 12S rRNA is probably not directly involved in

binding mRNA. If 12S rRNA were involved in binding mRNA,

then it would be expected that the portion of rRNA involved

would be exposed to chemical or enzymatic agents. Also, in

the presence of BH, it is expected that these proposed

sites would then be blocked from reacting with the chemical

or enzymatic probes. Since BH has no effect, it seems





































5' GAU GCA CAC CCA ACC ACC ACA ACA CCA CCC























Figure 30. Sequence of BH (courtesy of Robert Heck). This
oligonucleotide was designed to be an unstructured probe of
the RNA landing pad. Note the 5' AUG start codon.








































S2 40


















Figure 31. Cytochrome Oxidase Subunit II mRNA Secondary
Structure (courtesy of Bernie Courtney). This is the
proposed structure of bovine COII mRNA based on computer
analysis using the Squiggles program of the Genetics
Computer Group.









81

likely that 12S rRNA is not directly involved in mRNA

binding to any significant extent. It is possible that 12S

RNA is interacting through magnesium crossbridging to the

mRNA through its phosphate backbone with no detectable

conformational changes. In this binding mode, it would not

be expected to block access to modifiable reporter groups

(likely to be in contact with underlying protein) on the

rRNA. While this possibility can not be discounted by the

methods used, it does seem unlikely, and such a

translational mechanism would be the first of its kind

discovered.

There also has been proposed the presence of a Shine-

Dalgarno-like interaction on the mitoribosomal subunit,

based on computer sequence analyses of the 12S RNA (Saccone

et al., 1985). This area is purported to be located in an

atypical area, the stem 22-stem 24 area. However, Figure 32

shows this area to be inaccessible to chemical

modification. Since this area is inaccessible to chemical

modification, it seems unlikely that this region is

involved in binding mRNA transcripts. However, it is

possible that this area may become accessible and bind mRNA

in the presence of some as yet to be discovered factor, an

effect not seen under the conditions of this study.

Despite the probable absence of a Shine-Dalgarno-like

site on the 28S subunit, COII mRNA binding does have a





















DE:f^


6%. (<


S0


U


fta*























Figure 32. Inaccessibility of the Shine-Dalgarno-like
Sequence in the 28S Subunit. The proposed Shine-Dalgarno-
like region is apparently inaccessible to all of the probes
used in this study. Seq refers to sequencing lanes, Mod
refers to modifications, 0 refers to no enzymes added, A
refers to RNAse A/T,, V refers to RNAse V1, D refers to
DMS, and C refers to CMCT.









83
conformational effect on the small subunit. Figure 33 shows

that COII enhances RNAse A activity at nucleotides U232 and

U234 (see Figure 34). Since it has been shown that COII is

competitively inhibited by poly(U) under these conditions,

in the absence of factors (Denslow and O'Brien, in

preparation) it can be assumed that at least some portion

of COII mRNA is binding at the RNA landing pad. Therefore,

apparently some other portion of COII is having a message-

specific effect. Small subunits preloaded with p6ly(U) of

an average length of 300 bases was unable to block the COII

mRNA induced enhancement of RNAse A activity at residues

U232 and U234 (data not shown). Since it was shown earlier

that COII and poly(U) exhibit competitive binding

inhibition of each other, the result is somewhat confusing.

The most likely explanation is that over the course of the

one hour reaction time, the poly(U) and COII mRNA can

exchange, and during the COII bound time interval, the COII

specific effects are captured by RNAse A. This supposition

is consistent with all of the available data including the

relatively short reaction time of the previous experiment

(Denslow, et al., 1989) (10 minutes) and the relatively

long reaction time of this experiment (one hour). It should

be noted that functional binding of COII mRNA to the small

subunit may require the presence of COII mRNA specific

factors not available at the time of this study. Precedent


















inoo
o o /C
.- C Z..3'\





\ ,I
% %


Figure 33. Effect of COII on 12S RNA Accessibility. Note
the COII induced RNAse cleavages indicated at the right of
this figure. Seq refers to sequencing lanes, Mod refers to
modifications, 0 refers to no enzymes added, A refers to
RNAse A/TI.


!SE:6<

*MPC



t 40-


(cL:3 -










A





C


cc


C
G 250
G
U
C


G-U


-A


G-C


C-G
VC U-A
U


/ AU


10


/


U
J-A


U-A
G-C
G-C
G-C


Figure 34. Location of the COII Effect on the 12S RNA
Secondary Structure. Triangles indicate the location of the
COII induced cleavages.


C
C
G
A
.C