The chloramphenicol binding site of mammalian mitochondrial ribosomes

MISSING IMAGE

Material Information

Title:
The chloramphenicol binding site of mammalian mitochondrial ribosomes
Uncontrolled:
Mitochondrial ribosomes
Physical Description:
xvi, 137 leaves : ill. ; 29 cm.
Language:
English
Creator:
Harville, Terry O
Publication Date:

Subjects

Subjects / Keywords:
Anti-Bacterial Agents   ( mesh )
Binding, Competitive   ( mesh )
Binding Sites   ( mesh )
Chloramphenicol   ( mesh )
Ribosomes   ( mesh )
Biochemistry and Molecular Biology thesis Ph.D   ( mesh )
Dissertations, Academic -- Biochemistry and Molecular Biology -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida.
Bibliography:
Bibliography: leaves 130-136.
Statement of Responsibility:
by Terry O. Harville.
General Note:
Photocopy of typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000344454
oclc - 09364222
notis - ABY1247
sobekcm - AA00006112_00001
System ID:
AA00006112:00001

Full Text

















THE CHLORAMPHENICOL BINDING SITE OF
MAMMALIAN MITOCHONDRIAL RIBOSOMES











BY

TERRY O. HARVILLE


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



UNIVERSITY OF FLORIDA


1982




















ACKNOWLEDGEMENTS


I would like to gratefully acknowledge Robert

Sheehan, Robert Cheatham, and Jane Edwards, for their

technical support, especially in the area of ribosome

preparation. In addition, I would like to thank Jane

Edwards for her technical expertise in performing

two-dimensional PAGE. I would like to acknowledge Mary

Conde for providing information about the irreversible

inactivation of mitochondrial ribosomes by iodoamphenicol

and for the preparation of the bacterial ribosomes.

I acknowledge the good times and criticisms provided

by Gretchen Schieber and Sal Pietromonaco, during my

graduate career.

Nancy Denslow receives a special acknowledgement for

donating the tRNA substrates and also as an influential

scientist in shaping my understanding of mitochondrial

ribosomes.

Tom O'Brien receives a most grateful acknowledgement.

He is truly the most observant student of the empirical

sciences and human nature that I know. His guidance has

been, above all, a remarkable experience for me.










I would like to thank my wife, Jean, and my parents

for their patience.

Finally, I would like to thank Donna VanPuymbrouck,

who provided the secretarial skills to bring about this

manuscript.


















TABLE OF CONTENTS


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

LIST OF TABLES.................... ......................viii

LIST OF FIGURES ..................... .................. ix

ABBREVIATIONS.........................................xii

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

CHAPTER I INTRODUCTION................................. 1

Properties of Mammalian Mitochondrial Ribosomes......1
Size of the Ribosome ....... .....................
RNA Content......................... ........... 4
Protein Content.......................... .......4
The Peptidyl Transferase Center....................... 5
Substrates: Amino Acyl-tRNA and
Peptidyl-tRNA Binding Sites.................... 6
Catalytic Activity.............................. 7
The Peptidyl Transferase Reaction................ 8
The Chloramphenicol Binding Site.....................9
Properties of Chloramphenicol................... 9
Features of the Chloramphenicol Binding Site....11
Chloramphenicol Resistance......................12
Affinity Labelling of Ribosomes.....................13
Traditional Affinity Labelling..................13
Ribosomal Affinity Labelling....................14
The Basic Approach to Affinity Labelling the Peptidyl
Transferase Center and the Chloramphenicol Binding
Site............................................... 16
Reversible Binding............................. 16
Irreversible Inactivation.......................17
Labelling Parallels Irreversible Inactivation...18
Labelling Saturates Mole per Mole............... 19
Competitive Inhibition of the Affinity
Labelling Reaction............................ 20
Problems Associated with Affinity Labelling.....20
Summary ............................................. 23











CHAPTER II MATERIALS AND METHODS.......................25

Ribosome Preparation...............................25
Mitochondrial Ribosomes......................... 25
E. coli Ribosomes...............................27
Thin Layer Chromatography...........................28
Synthesis of the CAP Analogs.......................29
Iodoamphenicol (IAP)
(D-threo-2-iodoacetamido-l-p-nitrophenyl-
1,3-propanediol)..............................29
p-Amino-chloramphenicol (ACAP)
(D-threo-l-p-aminophenyl-2-dichloroacetamido-
1,3-propanediol)..............................30
p-Azido-chloramphenicol (PAP)
(D-threo-l-p-azidophenyl-2-dichloroacetamido-
1,3-propanediol).............................31
p-Azidobenzoamido-chloramphenicol (FAP)
(D-threo-l-p-azidobenzoyl-p-amidophenyl-2-
dichloroacetamido-l,3-propanediol) ............32
Chloramphenicol-p-azidobenzoate (PAF)
(D-threo-2-p-azidobenzoamido-l-p-nitropi vl-
1,3-propanediol).............................. 33
p-Hydroxylamino-chloramphenicol (HACAP)
(D-threo-2-dichloroacetamido-l-p-hydroxyl-
aminophenyl-1,3-propanediol)..................34
p-Nitroso-chloramphenicol (NOCAP)
(D-threo-2-dichloroacetamido-l-p-nitrosophenyl-
1,3-propandiol) ............................34
The Peptidyl Transferase Assay.....................35
Apparent Dissociation Constant (K'd)
Calculations.......................................36
Structure-Activity Relationship of CAP Analogs......37
Labelling Reactions............ ....................38
Sucrose Density Gradient Separation of Labelled
Ribosomal Subunits................................39
Polyacrylamide Gel Electrophoresis (PAGE)............40
One-Dimensional SDS PAGE.........................40
Two-Dimensional PAGE........................... 42
Fluorography .............................. .......... 44
Densitometry........................................45
Cutting and Digestion of Protein Bands or Spots
from Polyacrylamide Gels.......................... 45
Liquid Scintillation Counting (LSC).................46

CHAPTER III RESULTS.. ................................47

Functional Studies of the Chloramphenicol Binding
Site Through the Use of Chloramphenicol Analogs...48
Chloramphenicol Analogs Modified in the Para-
Nitro Moiety..................................51
Chloramphenicol (CAP)....................... 51
p-Amino-chloramphenicol (ACAP).............. 54
p-Hydroxylamino-chloramphenicol (HACAP)......56










p-Nitroso-chloramphenicol(NOCAP)............ 56
p-Azido-chloramphenicol (PAP)................58
p-Azidobenzoamido-chloramphenicol (FAP).....58
Chloramphenicol Analogs Modified in the
Dichloroacetate Chain........................ 60
Iodoamphenicol (IAP)......................60
Chloramphenicol-p-azidobenzoate (PAF).......60
Structure-Activity Relationships of Chlor-
amphenicol Analogs...........................64
Affinity Labelling for Structural (and Functional?)
Proteins of the Peptidyl Transferase Center
and the Chloramphenicol Binding Site of
Mammalian Mitochondrial Ribosomes..................68
Iodoamphenicol as the Affinity Probe............69
Effect of iodoamphenicol on the peptidyl
transferase activity of mitochondrial
ribosomes................ ................ 69
Reversible effect.......................69
Irreversible effect ..................... 69
1-D PAGE analyses of [ H]iodoamphenicol
labelling of ribosomal large subunit
proteins.................. ........... 72
1-D PAGE analyses of [ Cliodoacetamide
labelling of ribosomal large subunit
proteins.............................. ...77
Effect of chloramphenicol on the labelling
of ribosomal large subunit proteins by
[ H]iodoamphenicol........................81
Effect of chloramphenicol on the labelling
of ribosomal large subunit proteins by
[ C]iodoacetamide....................... 83
2-D PAGE analyses of [ H]iodoamphenicol
labelling of ribosomal large subunit
proteins ....... ......... ....85
2-D PAGE analyses of I H]iodoamphenicol
labelling of ribosomal proteins in
the monosome..............................89
Effects of N-acetyl-AA-tRNA and AA-tRNA on
the [ H]iodoamphenicol labelling
of ribosomal large subunit proteins.......94
p-Nitroso-chloramphenicol (NOCAP) as the Affinity
Probe......................................... 98
Effect of p-nitroso-chloramphenicol on the
peptidyl transferase activity of
mitochondrial ribosoes....................99
1-D PAGE analysis of [ C]NOCAP labelling
of ribosomal large subunit proteins in the
absence and presence of chloramphenicol.... 101

CHAPTER IV DISCUSSION.... ...........................107

APPENDIX I BUFFERS ................................. 123









APPENDIX II RIBOSOMAL QUANTITATION BY UV SPECTROSCOPY,
MOLECULAR WEIGHTS OF CHLORAMPHENICOL AND
THE CHLORAMPHENICOL ANALOGS............125

APPENDIX III PHOTOLYTIC STUDIES OF FAP............... 126

BIBLIOGRAPHY ............................................130

BIOGRAPHICAL SKETCH.................................... 137

















LIST OF TABLES


I. Properties of Ribosomes.......................... 3

II. Apparent Dissociation Constants for Chloramphenicol
and the Chloramphenicol Analogs..................53

III. Peptidyl Transferase Activity of Mitochondrial
Ribosomes Treated with lodoamphenicol,
Chloramphenicol or lodoacetamide..................73

IV. Peptidyl Transferase Activity of Mitochondrial
Ribosomes after a Prolonged Treatment with
Chloramphenicol or p-Nitroso-Chloramphenicol.....100


viii













LIST OF FIGURES


1. Comparison of the structures of chloramphenicol
and the photo-activatable chloramphenicol
analogs.......................... ............. 49

2. Comparison of the structures of chloramphenicol
and the chloramphenicol reduction analogs.....50

3. Comparison of the peptidyl transferase activities
of mitochondrial ribosomes and bacterial
ribosomes in the presence of chloramphenicol
(CAP).................... ............... ..... 52

4. Comparison of the peptidyl transferase activities
of mitochondrial ribosomes and bacterial
ribosomes in the presence of the chloramphenicol
reduction analogs........... ...................55

5. Comparison of the peptidyl transferase activities
of mitochondrial ribosomes and bacterial
ribosomes in the presence of p-azido-
chloramphenicol (PAP)........................57

6. Comparison of the peptidyl transferase activities
of mitochondrial ribosomes and bacterial
ribosomes in the presence of p-azidobenzo-
amido-chloramphenicol (FAP) ................... 59

7. Comparison of the structures of chloramphenicol,
iodoamphenicol, and idoacetamide...............61

8. Comparison of the peptidyl transferase activities
of mitochondrial ribosomes in the presence of
chloramphenicol (CAP) or iodoamphenicol
(IAP) .........................................62

9. Comparison of the peptidyl transferase activities
of mitochondrial ribosomes and bacterial
ribosomes in the presence of chloramphenicol-
p-azidobenzoate (PAF) ........................63

10. Structure-Activity relationships of
chloramphenicol analogs with para-nitro
substitutions................................66

11. The peptidyl transferase activity of
mitochondrial ribosomes after prolonged
incubations with iodoamphenicol (IAP)..........71









12. One-dimensional PAGE analyses of the
I H]iodoamphenicol ([ H]IAP) labelling
of mitochondrial ribosomal large subunit
proteins .......... .......... ........ ..... ... 75

13. Comparison of label3incorporated into individual
protein bands of [ H]iodoamphenicol
([ H]IAP) labelled mitochondrial
ribosomal large subunit proteins separated
by 1-D PAGE....................................76

14. On-edimensional PAGE analyses of the
[ C]iodoacetamide ([ C]IAM) labelling
of mitochondrial ribosomal large subunit
proteins ..................................... 79

15. Comparison of labell4ncorporated into individual
protein bands of [ C]iodoacetamide
([ C]IAM) labelled mitochondrial
ribosomal large subunit proteins separated
by 1-D PAGE....................................80

16. Comparison of label incorporated into individual
protein bands separated by 1-D PAGE, from
mitoShondrial ribosomal large subunits labelled
by [ H]iodoamphenicol ([ H]IAP) in the
absence and presence of chloramphenicol
(CAP).........................................82

17. Comparison of the tracings of densitometric scans
from the fluorograms of 1-D PAGE separated
mitochondrial ri osomal large subunit proteins
called with [ C]iodoacetamide
([ C]IAM), where chloramphenicol (CAP) was
included or not included in the labelling
reactions......................................84

18. Two-dimensional PAGE analysis of the
[ H]iodoamphenicol ([ H]IAP) labelling
of mitochondrial ribosomal large subunit
proteins, where chloramphenicol (CAP) was absent
or present in the labelling reaction..........86

19. Comparison of label incorporated into individual
proteins separated by 2-D PAGE, from
mitoShondrial ribosomal Iarge subunits labelled
by [ H]iodoamphenicol ([ H]IAP) in the
absence and presence of chloramphenicol
(CAP)............... ............................. 87









20. Comparison of the ratio of [3H]iodo-
amphenicol ([ H]IAP) incorporated into
mitochondrial ribosomal large subunit
proteins in the absence and presence of
cilorampehnicol (CAP) to the level of
[ H]IAP incorporated into the
proteins...................................... 88

21. Two-dimensional PAGE analysis of the
[ H]iodoamphenicol ([ H]IAP) labelling of
mitochondrial ribosomal large subunit proteins
labelled in the monosome in the absence and
presence of chloramphenicol (CAP)..............91

22. Twg-dimensional PAGE analysis of the
[ H]iodoamphenicol ([ H]IAP) labelling
of mitochondrial ribosomal small subunit
proteins labelled in the monosome in the absence
and presence of chlorampenicol (CAP)..........93

23. One-dimensional PAGE analyses of the effect of
N-acetyl-AA-tRNA and JA-tRNA on the
[ H]iodoamnhenicol ([ H]IAP) labelling
of mitochondrial large subunit proteins....... 96

24. One-dimensional PAGE analysis of4the effects of
chloramphenicol (CH) on the [ C]nitroso-
chloramphenicol ([ C]NOCAP) labelling of
mitochondrial ribosomal large subunit
proteins......................................103
14
25. Comparison of the raio of [ C]nitroso-
chloramphenicol ([ C]NOCAP) incorporated
into mitochondrial ribosomal large subunit
proteins in the absence and presence of
chloramphenicol (CAP)............................105

26. Model of the topological map of proteins from the
interfacial region of the mitochondrial
ribosomal large subunit......................114

27. Detailed model of the chloramphenicol binding
site.............................. ............ 121

28. Spectra of the photolytic decay of
p-azidobenzoamido-chloramphenicol
(FAP)..........................................127

29. Determination of the half-time of the photolytic
decay of p-azidobenzoamido-chloramphenicol
(FAP)........................................ 128














ABBREVIATIONS





Chloramphenicol and Chloramphenicol Analogs


Chloramphenicol

p-Anino-chloramphenicol

p-Azidobenzoamido-chloranphenicol

p-Hydroxylamino-chloramphenicol

Iodoamphenicol

Chloraiphenicol-p-azidobenzoate

p-Azido-chloramphenicol




Other Abbreviations Used


AA-tRNA

% C



CPM

f-Met-tRNA

H-bond

IAM

LSC

K' d

2ME

mRNA


Amino acyl transfer RNA

Bisacrylamide concentration divided by
the acrylamide plus the bisacrylamide
concentration x 100

Counts per minute

Formyl-methionyl transfer RNA

Hydrogen bond

Iodoacetamide

Liquid scintillation counting

Apparent disociation constant

2-Mercaptoethanol

Messenger RNA
xii


CAP

ACAP

FAP

HACAP

IAP

PAF

PAP










N-Acetyl-AA-tRNA

NMR

l-D

PAGE

PTase

rRNA

S

SDS

% T


2-D

TLC

tRNA


N-Acetyl-amino acyl transfer RNA

Nuclear magnetic resonance

One-dimensional

Polyacrylamide gel electrophoresis

Peptidyl transferase

Ribosomal RNA

Svedberg unit

Sodium dodecyl sulfate

Total acrylamide plus the bisacrylamide
concentration x 100


Two-dimensional

Thin layer chromatography

Transfer RNA


xiii

















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


THE CHLORAMPHENICOL BINDING SITE OF
MAMMALIAN MITOCHONDRIAL RIBOSOMES

By


Terry O. Harville


December, 1982



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


Chloramphenicol (CAP) is an antibiotic of therapeutic

importance which inhibits protein synthesis in mammalian

mitochondria as well as in its intended target, bacteria.

CAP binds with high affinity to a site on the large subunits

of ribosomes from bacteria and mitochondria. These ribo-

somes resemble each other in a primary functional manner, in

that each is involved in a complete translational system.

Yet, these ribosomes are very different in their physical-

chemical properties. In spite of differences in the

properties of these ribosomes, the binding of CAP to the

Peptidyl Transferase (PTase) Center and the conservation of


xiv









protein synthesis suggests that the fundamental functional

groups required for protein synthesis and CAP binding have

been maintained on both kinds of ribosomes.

CAP is a relatively small antibiotic with a well

defined bacterial site of action. Previous studies with

bacterial systems have implied that CAP could be divided

into three distinct regions that are required for activity.

In the present study, CAP was chemically modified in two of

these regions, the dichloroacetate chain and the para-

nitrophenyl ring, and the resultant analogs were tested for

the inhibition of the PTase activities of mitochondrial and

bacterial ribosomes. In this manner, the CAP binding site

from these ribosomes was examined to determine if any of the

analogs could discriminate between the ribosomes and to

identify features of the CAP molecule which were important

for binding to this site. The resultant data provided

evidence that the para-nitro moiety was hydrogen bonded to

some component of the ribosome at the CAP binding site and

that a limited volume or bulk could be accommodated in the

position of the dichloroacetate chain. Some of the CAP

analogs contained electrophilic groups, so that these CAP

analogs could act as "affinity probes" for proteins of the

CAP binding site. Protein L2 of the mitochondrial ribosome

was identified, through the use of iodoamphenicol as a CAP

affinity probe, as being within the domain of the PTase

Center and CAP binding site of these ribosomes.









Using the functional and structural data obtained,

models are presented, detailing the CAP binding site of

mammalian mitochondrial ribosomes.


xvi
















CHAPTER I
INTRODUCTION



Properties of Mammalian
Mitochondrial Ribosomes



Protein synthesis occurs on a macromolecular complex of

protein and ribosomal RNA (rRNA) determined to be an

important entity of the cell more than 20 years ago. The

ribosomee" became the word coined to represent this

ribonucleoprotein complex (1). Ribosomes were found not

only in bacteria and within the cytoplasm of the eukaryotic

cell, but, in addition, during the mid-1960's were found by

O'Brien and Kalf to inhabit the subcellular organelle, the

mitochondria, of mammals (2, 3). These mammalian

mitochondrial ribosomes are a component of a complete

translation system within mitochondria, where they use

mitochondrial mRNA's and tRNA's to synthesize about a dozen

different proteins most of which belong to the electron

transport system and ultimately reside in the inner

mitochondrial membrane (4, 5). The protein synthesizing

capability of the mitochondrial ribosome is required for the

eukaryotic cell to maintain the integrity of its energy

transducing system.








The mammalian mitochondrial ribosomes resemble pro-

karyotic ribosomes and eukaryotic cytoplasmic ribosomes in

their basic organization, in that each consist of a large

number of proteins and rRNA partitioned into two subunits

which house the individual functions necessary to bring

about protein synthesis. The major functional site on the

ribosome is the Peptidyl Transferase (PTase) Center which is

located on the larger of the two subunits of each of the

ribosome types The PTase Center is also the prime site

for antibiotic action and details of its function and the

binding of antibiotics will be discussed in later sections

(6).

Size of the Ribosome

The mammalian mitochondrial ribosome was found to have

a sedimentation coefficient of 55S which is much smaller

than the values observed for prokaryotic ribosomes and

eukaryotic cytoplasmic ribosomes with sedimentation

coefficients of about 70S and 80S, respectively (Table I)

(7, 8). Initially it was thought that mitochondrial

ribosomes were smaller than their counterparts, based on the

sedimentation coefficients observed (9). Even some

investigators suggested that the 55S ribosome was a

degradation artifact of a larger ribosome. Determination of

a molecular weight of 2.8 million for the mammalian

mitochondrial ribosome, which is slightly larger than the

molecular weight of 2.7 million observed for bacterial

ribosomes, helped to end the "small or degraded ribosome"




















TABLE I




Properties of Ribosomes


Ribosome Type


Mammalian
Mitochondrial

E. coli


Mammalian
Cytoplasmic


Buoyant
MW* Density


55S

70S



80S


2.8

2.7



4.5


1.44

1.64



1.58


rRNA
MW*


0.35
9.54

0.56
1.10

0.7
1.5


Proteins


80-90

54



70-80


* (MW x 10-6 Daltons)








notion connotated by a sedimentation coefficient of 55S (7,

10). The rather low sedimentation coefficient, 55S, is the

result of a low buoyant density of about 1.44 g/ml for the

mitochondrial ribosome as compared to about 1.64 g/ml for

bacterial ribosomes (Table I) (7, 8, 11).

RNA Content

The low buoyant density and consequently the low

sedimentation coefficient reflect not a smaller size of the

mitochondrial ribosome when compared with the bacterial

ribosome, but they reflect an alteration in the RNA and

protein content of the mitochondrial ribosome. The

mammalian mitochondrial ribosome has two RNA species which

sediment at 16S and 12S, and have molecular weight of 0.54

million and 0.35 million respectively (7, 8). These

ribosomal RNA's are about one-half the size of the

comparable rRNA's from E. coli ribosomes which have

molecular weights of 1.1 million and 0.56 million,

respectively, and sediment at 23S and 16S, respectively

(12).

Protein Content

The low buoyant density, the low sedimentation co-

efficient, and the smaller rRNA's indicate that

mitochondrial ribosomes must have the bulk of their total

mass contributed to by protein when compared to the

bacterial ribosome. This is, in fact, the case. The

mammalian mitochondrial ribosome has about 90 proteins which

total to 1.81 million daltons and which comprise about









two-third's of the ribosomal mass (13). In comparison, the

bacterial ribosome has 54 proteins of 0.98 million daltons

and this amounts to slightly more than one-third of the

bacterial ribosomal mass (14). The excess protein content

of mammalian mitochondrial ribosomes, when compared with

bacterial ribosomes, leads to interesting questions as to

what functional and structural roles these proteins may have

in addition to the primary protein synthetic activity of the

ribosome.

Some of these proteins may have replaced the rRNA in

the mitochondrial ribosome, yet others must be performing

the same functional tasks as -. analogous proteins from the

bacterial ribosome. The mammalian mitochondrial ribosome

with 52 proteins in its large subunit contains almost as

many proteins as the entire bacterial ribosome, which has 54

proteins. Several of these proteins are certainly involved

in the peptidyl transferase activity and some of these must

be intimately associated with the chloramphenicol (CAP)

binding site. The present study examines these functionally

important proteins and resolves questions about the nature

of the chloramphenicol binding site and the Peptidyl

Transferase Center.

The Peptidyl Transferase Center

The function of the ribosome is to synthesize proteins

by catalyzing the formation of peptide bonds between amino

acids at the primary functional site on the large ribosomal

subunit in the region termed the Peptidyl Transferase








Center. The catalytic activity at this site is called

peptidyl transferase.

Why has such a compleK organelle evolved to house this

function and to bring about protein biosynthesis? A

proteolytic-type enzyme could synthesize peptide bonds, but

the essential features for the fidelity of protein synthesis

would be lost. The ribosome provides an essential platform

for the substrates and factors of protein synthesis to come

together in an orderly fashion for the faithful translation

of the genetic material, which apparently would not occur in

a less complex system.

Substrates: Amino Acyl-tRNA and Peptidyl tRNA Binding Sites

The Peptidyl Transferase Center is characterized by two

adjacent binding sites for the substrates of peptide bond

formation. Operationally these have been defined as the A

site, for amino acyl-tRNA binding, and P site, for

peptidyl-tRNA binding (6, 15). The P site accommodates the

growing peptide chain attached to tRNA or the initial

f-Met-tRNA. The A site is where the incoming amino

acyl-tRNA binds prior to peptide bond formation. After

transfer of the peptide chain, from the P site, to the amino

moiety of the amino acyl-tRNA at the A site and removal of

the uncharged tRNA from the P site, the new peptidyl-tRNA

"translocates" to the vacated P site, opening the A site for

a new amino acyl-tRNA and a new round of peptide bond

synthesis (16). This process repeats until a protein has

been synthesized.








Catalytic Activity

Molecularly, protein synthesis is not well understood.

What enzymatic functions are required for peptide bond

formation and protein biosynthesis? In recent years, two

models have been proposed to explain the catalysis of

peptide bond formation, i.e. peptidyl transferase. In the

first model, peptide bond synthesis occurs by essentially

the reversal of proteolysis, as if an acid protease such as

chymotrypsin was performing the back reaction (17, 18, 19).

This model invokes an active serine and histidine at the

PTase Center (17). The major consequence of this model is

the "acyl-ribosome" intermediate which would be generatedd by

the serine attack on peptidyl-tRNA (17). No such

intermediate has been generated by classical enzymological

techniques, however (17). Another model, the "template"

model, simply uses a ribosome as a template to bind the

substrates and bring them into close proximity and the

correct orientation for the chemical attack of the amino

moiety of amino acyl-tRNA on the peptidyl-tRNA ester (17).

A further revision of this model includes catalytic groups,

such as histidine, to facilitate the reaction (17). This

model is more favorable in light of the lack of evidence to

support the first model and from the fact that the

peptidyl-tRNA ester can adequately substitute for the acyl

intermediate required by the ribosome (17, 18, 19).








The Peptidyl Transferase Reaction

Several years ago, an assay was developed to study the

mechanism of peptide bond formation (20, 21, 22). The PTase

assay became available when the action of puromycin, acting

like an amino acyl-tRNA acceptor and terminating protein

synthesis, was understood (23, 24). Puromycin binds to the

A site, resembling amino acyl-tRNA, and will covalently

attack the peptidyl-tRNA at the P site, through the action

of peptidyl transferase, forming a peptidyl-puromycin adduct

which can be isolated and quantified (22). The extent of

adduct formation provides information about the activity of

the PTase Center. This in vitro a jay is convenient

in that only the ribosomal large subunit, a peptidyl-tRNA

donor, and puromycin are required for the reaction to occur

(22). The need for extraneous factors has been circumvented

by the reaction conditions (15, 72). Many of the

antibiotics that inhibit protein synthesis do so by binding

to the PTase Center, subsequently preventing the peptidyl

transferase reaction (6, 25). The binding of these

antibiotics is routinely measured by the PTase assay where a

decrease in peptidyl transferase activity is detected as a

decrease in the puromycin-adduct formed. Classically, this

has been an excellent technique for examining the role of

chloramphenicol in the inhibition of protein biosynthesis

(20, 26).










The Chloramphenicol Binding Site

The antibiotic chloramphenicol (CAP) was first

discovered and reported during the late 1940's (27). After

its structure was fully elucidated, it became the first

antibiotic to have a commercially available synthetic route

of production, rather than biosynthetic (28).

The antibiotic property of chloramphenicol was found to

be its ability to inhibit protein synthesis within bacteria

and further studies provided evidence that the target of

action was the ribosome (29, 30, 31).

CAP was lat r shown to bind with high affinity to a

site on the large ribosomal subunit of bacteria and to

inhibit the PTase reaction (32, 33). Using similar

techniques, the same conclusions were made in regard to the

binding of CAP and its mode of action on mitochondrial

ribosomes (34).

Properties of Chloramphenicol

Chloramphenicol is a rather small antibiotic with

molecular weight of 323. The systematic name for CAP is

D-(-)-threo-2-dichloroacetamido-l-para-nitrophenyl-

1,3-propanediol. Two interesting functional groups can be

noted from inspection of the name, the para-nitro moiety and

the dichloroacetate chain. These are unusual features for a

biologically synthesized compound and bring about a

uniqueness for CAP (28, 35). The molecule CAP can be








divided into three structural parts that are required for

its biological activity: 1) para-nitrophenyl ring,

2) dichloroacetate chain, and 3) 2-aminopropanediol chain

(29, 35, 36). Each of these positions has been modified,

individually or collectively, to produce analogs which have

been tested for inhibition of bacterial growth. In this

manner, structure-activity relationships for these three

positions were ascertained (36). It was proposed that the

electronegativity of the nitro moiety and the size and

electronegativity of the dichloro moiety played crucial

roles in providing for the ability of chloramphenicol to

inhibit protein synthesis in bacteria (29, 36). The

consensus of the results from various modifications suggest

that the CAP binding site can accommodate some minor

variation in the nitrophenyl ring and the dichloroacetate

chain, but that the 2-aainopropanediol chain must remain

stereochemically intact (29, 36, 37). It is of interest that

none of the CAP analogs tested attained the inhibitory level

of the parent compound, CAP.

Using the structure-activity relationships proposed

according to the analyses of the inhibition of bacterial

growth by CAP analogs, a derivative of chloramphenicol was

synthesized in the expectations that it would be a better

antibiotic than CAP (38). When tested though, this analog

was shown to be much inferior to CAP (38, 39). As of the

present, no derivative of CAP has been found to inhibit

protein synthesis in bacteria better than CAP itself. In








the Results and Discussion chapters, I will report some

structural modifications of chloramphenicol and the

consequences of the changes with respect to the inhibition

of the PTase activities of bacterial and mitochondrial

ribosomes and I will suggest new structure-activity

relationships for CAP and its binding to the ribosome.

Features of the Chloramphenicol Binding Site

Several techniques have been applied to elucidate

structural and functional components of the CAP binding site

of bacterial ribosomes. Among the more useful techniques

have been: 1) the salt-stripping of ribosomal proteins,

while following the loss in the ability of -'P to bind to

the ribosomes, 2) the reconstitution of the CAP binding

activity, by adding individual ribosomal proteins back to

protein-deficient ribosomal particles, and 3) the affinity

labelling of ribosomal proteins by CAP affinity analogs (40,

41, 42, 43). The ribosomal proteins identified by these

procedures have been assigned to be structural and/or

functional proteins of the CAP binding site and PTase

Center. The bacterial ribosomal proteins L2, L16, and L27

(with molecular weights of about 30,000, 18,000, and 12,000,

respectively) are the main proteins selected by these

techniques for the PTase Center and CAP binding site from

E. coli.

There are features of the CAP binding site that are not

readily ascribed to individual proteins. These mainly

involve the interactions that allow CAP to bind to the








ribosome. One way of examining for the essential

interactions has already been mentioned briefly (see

previous section) and has been used primarily by medicinal

chemists in search of superior CAP analogs. This procedure

entails the selective modification of portions of the CAP

molecule and then assaying for the antibiotic activity of

the analog in comparison to CAP. The major conclusions of

these studies were that the: 1) D-threo conformation is

absolutely required, 2) the electronegativity and molar

volume are the important features of the dichloroacetate

chain, and 3) the electronegativity of the p-nitro moiety is

required (29, 35, 36).

NMR spectroscopy has been recently applied to examine

the binding of CAP to bacterial ribosomes (44). This

technique does not allow one to identify actual components

of the CAP binding site, but general features of the binding

interactions have been ascertained. From the broadening of

peaks in the NMR spectra, it was inferred that the hydroxyls

of the aminopropanediol chain of CAP form hydrogen bonds to

components of the ribosome in the anchoring of CAP within

its binding site on the bacterial ribosome (44).

I will report in the Results and Discussion chapters

the important components and features of the chloramphenicol

binding site of mitochondrial ribosomes.

Chloramphenicol Resistance

Some bacteria and eukaryotic cells develop the ability

to exhibit a resistance to the effects of CAP. The








resistance usually results from the induction of an enzyme

that can inactivate CAP or by an alteration of the ribosome

so that CAP does not interact in an inhibitory manner with

its ribosomal binding site (35, 45, 46). In eukaryotes, CAP

resistance generally results from a mutant mitochondrial

ribosome and the mutation can be traced to the mitochondrial

DNA, where the lesion is localized to the mitochondrial

large ribosomal RNA gene (47, 48). Several CAP resistant

mitochondrial DNA's have been sequenced and single base

mutations have been localized about 300 nucleotides from the

3' end of the large rRNA in resistant cell lines (49, 50,

51). The base substitutions occur in single-stranded

regions near the base of a helix in the proposed secondary

structure of the rRNA (50, 51). The mechanism of CAP

resistance appears to be an alteration in the binding of

ribosomal proteins to the affected region of rRNA, thereby

decreasing the ability of CAP to bind to the ribosome (51).

It is important to note that CAP resistant ribosomes, which

show a decrease in the ability to bind CAP, generally

exhibit a decrease in their protein synthetic capacity as

well (46, 47, 48).

Affinity Labelling of Ribosomes

Traditional Affinity Labelling

Historically, affinity labelling studies were done to

determine which amino acid residues may be involved at the

active site of the enzyme in question (52, 53, 54, 55). An

alkylating or acylating compound was expected to react with








the enhanced nucleophilic moiety at the active site,

providing a means of localizing important amino acid

residues. The labelling compound could be made active-site-

directed by having the same structural groups as the

enzyme's substrate (53, 55). Many of the amino acid side

groups are potential nucleophiles towards the electrophilic

affinity labelling species and the enhanced nucleophilicity

of the active site provides a means of increased reaction

for the affinity probe (52, 53, 54, 55). As a control for

the specific labelling reaction, the enzyme could be

incubated with a tight binding substrate prior to the

addition of the labelling reagent. In this situation the

lack of label on the previously identified amino acid

supports the fact that the substrate binding to the active

site can effectively block the labelling reaction, and

therefore, that the identified amino acid is within the

active site (53, 55).

Ribosomal Affinity Labelling

For macromolecular complexes the affinity labelling

approach is intended to provide information other than

which amino acids are in active sites. Rather, the

identification of individual proteins and regions of RNA

forming the binding sites is the goal of affinity labelling

studies involving ribosomes (56, 57, 58). To this end,

various ribosomal substrates have been converted to

electrophilic species by de novo synthesis or the addition

of an electrophilic moiety to the parent molecule. These








compounds can thus serve as reactive affinity probes towards

their respective binding sites on the ribosome. In this

way, structural and functional roles can be assigned to

proteins that react specifically with particular affinity

probes.

Among the first affinity probes used on bacterial

ribosomes were two chloramphenicol analogs, iodoamphenicol

and bromoamphenicol (59, 60). Iodoamphenicol (IAP) and

bromoamphenicol (BAP) were shown to bring about an

irreversible inactivation of the bacterial ribosomal

protein synthetic capacity that CAP could not (59, 60, 61).

Incubation of the ribosome in the presence of [14 C]IAP

showed covalent incorporation into both subunits upon

sucrose gradient analysis (59, 61). Polyacrylamide gel

electrophoresis allowed identification of protein L16 from

the bacterial ribosome large subunit as the chloramphenicol

binding protein (61). Reversible and irreversible

inhibition were also shown for BAP (60). Using [ C]BAP,

though, L2 and L27 were identified as the primary peptidyl

transferase proteins in bacterial ribosomes by another group

(60). These early experiments were not carefully

controlled, however, and competitive inhibitors were not

used to demonstrate blockage of the "specific" labelling

reaction. For the IAP experiments a non-specific reagent of

similar chemical reactivity, iodoacetamide (IAM), was used

in one series of experiments to show a different overall

labelling pattern than that obtained from IAP labelling








(61). This control was intended to demonstrate that the

enhanced labelling on L16 by [ C]IAP was not simply due

to a protein containing a powerful nucleophile.

Substrate analogs, usually substituted tRNA's, have

also been used as affinity probes of the Peptidyl

Transferase Center (56, 57, 58). Although more difficult to

prepare and purify, these tRNA affinity probes have a much

higher affinity for the Peptidyl Transferase Center and

thereby, potentially, a much more specific reaction. The

tRNA affinity probes have labelled many proteins including

L2, L27, and L16, (56, 57, 58). These proteins are probably

atimately associated with the Peptidyl Transferase Center

and most likely are very close to the actual CAP binding

site of bacterial ribosomes, as revealed by their labelling

by CAP affinity analogs.

The Basic Approach to Affinity Labelling the PTase Center
and the CAP Binding Site

In this section, I have compiled the general criteria

that should be used in the affinity labelling study.

Previous studies using affinity labelling to identify the

proteins involved in the CAP binding site have been somewhat

ambiguous because of the lack of complete experimental

evidence. I have sought to overcome the weaknesses of

previous studies by following the guidelines listed below.

Reversible Binding

To be useful in an affinity labelling study, the

affinity probe must be able to interact with its binding








site in a manner similar to its parent compound. That is,

any derivatization must not abolish the active site

direction of the parent molecule. For the peptidyl

transferase activity, reversible inhibition can be measured

for the antibiotic and its affinity probe derivative by the

peptidyl transferase assay. In this way, a functional

interaction with the Peptidyl Transferase Center can be

ascertained for the affinity probe, providing a rationale

for continued testing of its utility as an affinity probe.

For electrophilic affinity probes, the time of exposure with

the ribosomes is minimized to limit the amount of covalent

adduct formation and, therefore, irreversible inactivation

that would be measured as reversible binding in the peptidyl

transferase assay. For similar reasons, photo-affinity

probes would be assayed in the dark. Reversible binding to

the Peptidyl Transferase Center is one of the most useful

criteria for a functional interaction with the ribosome.

Irreversible Inactivation

It would be expected that if a binding site were

occluded, then a function associated with the site would

diminish. This could occur with saturating amounts of a

reversible inhibitor under which conditions the site is

effectively always occupied or by the formation of a

covalent linkage within the site so that the site is

permanently occupied by the covalently-attached probe. The

electrophilic affinity probe can occlude its binding site if

a properly oriented nucleophilic group in the binding site








can react with the probe, whereas a photo-affinity probe

requires only excitation by light to bring about covalent

attachment within the site (55). Total inactivation of

peptidyl transferase could be expected if the affinity probe

and its binding site meet the above criteria. In actuality,

for a correctly oriented nucleophile to be present on the

ribosome is an unlikely event, because of lack of the

"classical active sites" containing nucleophilic amino

acids. And since photo-incorporation rarely proceeds to

saturation total inactivation of the PTase activity seems

unlikely (17, 58).

Labelling Parallels Irreversible Inactivation

When the conditions for interaction with the binding

site by the affinity probe have been fulfilled and the probe

is available with a radioisotope incorporated into its

molecular structure, a series of experiments can be

undertaken to study the covalent attachment of the affinity

probe to ribosomal components. The "hot" affinity probe is

incubated with the ribosomes for various periods of time,

using the concentrations previously determined to produce a

response in the ribosomal peptidyl transferase activity.

The extracted ribosomal proteins, separated by

unidimensional and bidimensional gel electrophoresis and

subjected to fluorography, are then examined for the extent

of covalently-attached label. Ideally, specific labelling

(active site directed labelling) is expected to occur with

increasing time or probe concentration in a manner that








parallels the time-concentration dependence of the

irreversible inactivation of the peptidyl transferase

activity. Under these conditions the extent of label

appearing in a specific ribosomal protein should reflect the

extent of irreversible inactivation of the ribosome (55).

That is, the loss of activity should correlate with the

amount of affinity probe on the protein being labelled.



Labelling Saturates Mole Per Mole

The ribosomal peptidyl transferase activity involves

one region of the large subribosomal particle. The

labelling via an affinity probe for a protein in a

functional binding site is expected to saturate at a

stoichiometric relationship of one affinity probe molecule

on a protein, if only one reactive nucleophile is present on

the protein. Ideally, the labelling kinetics are expected

to correlate with the kinetics of irreversible inactivation

of the Peptidyl Transferase Center (55). Saturation of

labelling in this manner implies a unique point of

attachment in the site for the affinity probe and a specific

relationship for the probe and the site. For proteins not

associated with the binding site, a non-saturating labelling

pattern will be observed, which resembles a non-affinity

directed labelling pattern as observed with a general

labelling reagent (55).








Competitive Inhibition of the Affinity Labelling Reaction

Definitive assignment of a ribosomal protein to the

Peptidyl Transferase Center occurs when the labelling of the

protein by an affinity probe can be blocked by a competitive

binder to the same site. The best examples would involve

utilizing the parent molecule to block a derivatized

compound synthesized from the parent molecule, such as

chloramphenicol, to competitively block the CAP affinity

analogs.

Once interaction with the ribosome has been shown for

an affinity probe and labelling on a protein via the probe

saturates in a manner that can be blocKeid by a competitive

inhibitor, then assured assignment of the protein to the

ribosomal Peptidyl Transferase Center can be made. Having

met these criteria successfully for chloramphenicol affinity

analogs, one will be able to identify proteins at the CAP

binding site.

Problems Associated with Affinity Labelling

Affinity probes can be classified into two categories,

photo-affinity probes and electophilic affinity probes, each

with special problems. Photo-activatable affinity labelling

reagents in theory are the best active site labellers to be

found, but in practice have not worked that well with the

ribosome (58). The affinity labelling reaction is time,

concentration and light fluence dependent for the

photo-affinity probe. It may also be nucleophile dependent,








if an electrophilic species is the light-generated product.

The chemistry of photo-addition is largely unknown and a

variety of reaction and rearrangement products may be

produced. Reactions with water and buffer may also occur as

predominant side reactions, competing with active site

labelling. The rapid exchange of CAP or its photo-affinity

analogs with the CAP binding site may be fast enough to

allow a "photo-activated" intermediate in solution to

diffuse into the CAP binding site and then covalently attach

to protein, but the "photo-activated" probe may diffuse away

from the site before covalency occurs. The reaction of

photo-activated probes with the Pep-Aldyl Transferase Center,

though, should be enhanced by the active site direction of

the photo-affinity analog, but the probe may act as a

competitive blocker of itself if non-activated probe is

bound at the site.

Electrophilic affinity labelling reagents have another

problem, in that all nucleophiles on ribosomal proteins will

react with the affinity probe according to their

accessibility on the ribosome. Also, water or buffers may

react with the affinity probe effectively decomposing it,

before sufficient labelling can occur. Furthermore, thiol

agents such as 2-mercaptoethanol (2ME) certainly will react

exceptionally well toward most electrophilic reagents, so

the ribosomes must be free of 2ME prior to affinity probe

incubation. Very powerful nucleophiles on ribosomal

proteins at locations other than the Peptidyl Transferase








Center and the CAP binding site may also react with the

affinity probe, giving a false positive affinity labelling

reaction. However, as a control for this reaction, the

competitive blockage by the parent compound should not

affect the labelling on these proteins, making the issue

less confusing. An equally important consideration is that

a correctly oriented nucleophile is required in the binding

site of chloramphenicol for each particular electrophilic

affinity probe used. If this in fact were the case, then

complete, irreversible inactivation of the peptidyl

transferase activity could be achieved at the saturation of

labelling in the CAP binding site. Such correctly oriented

nucleophiles would be fortuitous and the amount of

irreversible inhibition will depend upon the proximity of

the labelled amino acid side group to the CAP binding site.

Chloramphenicol exchanges rapidly at its binding site

on E. coli ribosomes (44). Since mammalian

mitochondrial ribosomes and bacterial ribosomes have a

similar dissociation constant (Kd) for CAP, it is

expected that CAP will also exchange rapidly at the

mitochondrial ribosomal binding site (34). The affinity

probe should be exposed to nucleophilic amino acid side

groups as it moves in or out of the CAP binding site so that

"affinity" directed labelling is enhanced on proteins with

available nucleophiles in or near the CAP binding site. The

disadvantage of not having an extremely low Kd is that a

higher concentration of affinity probe is required to allow









significant labelling of the CAP binding site, and this may

lead to more non-specific labelling of nucleophiles on the

ribosome surface. Also a concentration of CAP at 4 mM or

higher is required to completely obscure the CAP binding

site as a competitor of the affinity labelling probe (44).

Another problem is that the modification of chlor-

amphenicol to produce an affinity probe may also bring about

the loss of binding to the CAP site. This is easily

discovered in the PTase assay testing procedure and provides

valuable information about the positions of the CAP molecule

pertinent for binding and inhibition of protein synthesis.

Summary

Mammalian mitochondrial ribosomes are similar to

bacterial ribosomes in their general features, but differ

markedly in their specific properties. One common feature is

the binding of chloramphenicol and thereby the inhibition of

the peptidyl transferase activities of the ribosomes. The

CAP binding site can be studied by various techniques to

identify structural and functional components of the site.

The most useful techniques involve modification of

chloramphenicol and examining the CAP analogs as inhibitors

of the PTase activity or as affinity probes for the CAP

binding site. These procedures have allowed the

identification of proteins at the bacterial CAP binding

site. In addition, features of the CAP molecule that are

required for binding, and the complementary ribosomal









features required for the binding of CAP have been

identified.

In the present study, I will report the ribosomal

proteins from mammalian mitochondria involved with the CAP

binding site and propose a topological model that maps the

position of these proteins. From the analyses of the

binding of CAP and the CAP analogs to the ribosomes, I have

developed a model describing the features of chloramphenicol

that are required for CAP to bind to the chloramphenicol

binding site.
















CHAPTER II
MATERIALS AND METHODS



A list of the buffers and their compositions appears in

Appendix I. Appendix II contains the standards for

quantifying ribosomes by UV spectroscopy and the molecular

weights of the chloramphenicol analogs.

Ribosome Preparation

Mitochondrial Ribosomes

The mitochondrial ribosomes were prepared essentially

as has been previously reported (2, 3, 8, 13, 34). All

preparations began with livers of freshly killed cows

transported to the laboratory in ice. Four to eight kg of

fresh liver were ground in a meat grinder and diluted with

four volumes of Buffer A or B. Subsequently, the mixture

was passed through a coarse-mesh plastic screen and

homogenized by a high-frequency cavatation device (a Tekmar

Super Dispax, Model SD-45K).

Unbroken cells and nuclei were removed from the

mitochondria by pumping the homogenate at 880 ml/min through

a Vernitron CFR-2 continuous flow rotor spinning at 10,000

rpm in a Vernitron LCA-2 centrifuge or by centrifuging the

homogenate in individual buckets of a JA-10 rotor in a








Beckman J-21 centrifuge at 3,000 rpm for 10 minutes. The

mitochondria were harvested from the combined supernatants

by continuous-flow centrifugation in a Beckman JCF-Z rotor

at 18,000 rpm with a flow rate of 440 ml/min or by pelleting

the mitochondria in the buckets of a Beckman JA-10 rotor at

10,000 rpm for 10 minutes. The mitochondria were washed

three times by resuspension into Buffer A and centrifugation

at 10,000 rpm for ten minutes in a JA-10 rotor.

To isolate the ribosomes, the mitochondria were

resuspended to a concentration of 20 mg protein/ml in Buffer

C and lysed by the addition of Triton-Xl00 to a final

concentration of 2%. The resulting lysate was clarified by

centrifugation in a Beckman JA-10 rotor at 10,000 rpm for 45

minutes. Fibrous DEAE (DE52, Whatman) was added to the

lysate, at the amount of 1 g DEAE per 1 g of mitochondria.

After stirring for 30 minutes, the fluid was filtered from

the DEAE. The ribosomes are eluted from the DEAE with

Buffer D. The ribosomal solution was concentrated by

ultrafiltration in a DC-10 Amicon Hollow Fiber Filter

ultrafiltration device to a volume of one liter. The

ribosomes are pelleted by centrifugation for 20 hours at

35,000 rpm in a Beckman type 35 rotor through a 45% sucrose

cushion of Buffer E. The ribosomal pellets were resuspended

into 10 ml Buffer E containing puromycin and incubated for

five minutes at 370C to remove any nascent polypeptide chains

from the ribosomes. The ribosomal solution was clarified

by centrifugation in a Beckman type 65 rotor for ten minutes








at 20,000 rpm. The supernatant was loaded onto a zonal

gradient (Beckman Ti 14 rotor) of 10% to 30% (linear)

sucrose in Buffer F for the 55S monosome preparation or of

Buffer G to dissociate the 55S into 39S and 28S subunits.

The ribosomes were centrifuged for 12 hours at 30,000 rpm in

the former case and for 17 hours at 30,000 rpm for the

latter case. The gradient was fractionated into 20 ml

fractions collected by an LKB 2112 fraction collector while

being continuously monitored by a Uvicord UVA ultraviolet

monitor. The fractions containing ribosomes were pooled and

centrifuged for 16 hours at 50,000 rpm in a Beckman Ti 60

rotor. The pellets were resuspended into Buffer H and

aliquoted for storage at -700C.

E. coli Ribosomes

E. coli ribosomes were prepared by Mary Conde as

reported (34). E. coli K-12, strain Hfr DIO RNase- or

strain 1200F- end A 1100 rns A Sn- were grown to

early exponential phase in nutrient broth at 370 in a

rotory shaking incubator. The cells were harvested by low

speed centrifugation (JA-10 Beckman rotor, 6,000 rpm for ten

minutes). The cells were resuspended in 10 mis of cold

Buffer I and sonicated three times of 20 seconds each. The

cellular debris was removed by centrifugation (Beckman type

65 rotor, 15,000 rpm for 10 minutes) after which the

supernatant was centrifuged in a Beckman type 65 rotor for

two hours at 50,000 rpm. The 50S ribosomal large subunits

were prepared by placing 20 to 30 A260 of 70S ribosomes








into 2 ml of Buffer J, loading the solution on the top of a

10% to 30% linear sucrose gradient of Buffer J in a Beckman

SW 27 rotor and centrifuging at 27,000 rpm for nine hours.

The gradients were fractionated under continuous monitoring

by a Gilford 2400 spectrophotometer. The fractions

containing subunits were pooled and centrifuged in a Beckman

type 65 rotor for two hours at 50,000 rpm. The pellets were

resuspended into Buffer K and stored frozen at -700C.

Thin Layer Chromatography (TLC)

The synthetic reactions of CAP analog production

required careful monitoring. I used silica gel TLC plates

(Merck 60 F 254) with CHC13:CT30H (95:5, 9:1 or 3:1)

and ethyl acetate as the eluants for the monitoring steps

and to help gauge the purity of the products. Each analog

was developed on several chromatograms in the different

solvent systems until I was convinced that the product was

clean from reactants or unwanted products. In most cases a

10 to 100 fold excess of the analog was spotted on the

plates to reveal less than 10% contamination. In

conjunction with other analyses such as melting points, UV

spectroscopy, IR spectroscopy, NMR spectroscopy, and Mass

spectroscopy all of the analogs appeared to be greater than

99% pure, by having less than 1% contaminants.








Synthesis of the CAP Analogs

lodoamphenicol (IAP)
(D-threo-2-iodoacetamido-l-p-nitrophenyl-l,3-propanediol)

lodoamphenicol was synthesized essentially according to

published procedures (59,62). Iodoacetic acid (Aldrich) was

recrystallized and stored in the dark under desiccant at

-200C. The iodoacetic acid (1.0 g) was dissolved into 25

ml dry dioxane and N-hydroxysuccinimide (Sigma) (0.7 g) was

added. Dicyclohexylcarbodiimide (Aldrich) (1.3 g) in 10 ml

dry dioxane was added to the mixture. The reaction was

allowed to proceed at room temperature for one hour in the

dark after which the dicyclohexylurea was removed by

centrifugation. The supernatant wu dried by rotary

evaporation and recrystallized from isopropanol.

Chloramphenicol base (Sigma) (0.5 g) was dissolved into 20

ml cold ethyl acetate (0C). Iodoacetylsuccimide ester

(0.7 g) was added slowly over several minutes and the

mixture was stirred at 0C for two hours. An additional

one hour at room temperature incubation followed after which

50 ml ethyl acetate and 50 ml 0.1 N H2SO4 were added

to provide a two-phase extraction system. The organic phase

was obtained and worked up to provide iodoamphenicol. TLC

analysis showed a single resolvable spot which migrated to a

distinct position slightly slower than where CAP does in the

solvent systems employed. Alkaline reaction conditions and

aluminum oxide as the chromatography medium were avoided

since these tend to cause the decomposition of IAP into a








morpholino adduct by the internal hydroxyl attack at the

iodo-containing carbon.

[3H]Iodoamphenical was synthesized in a micro-scale

reaction, scaled down from the one listed above.

[3H]Iodoacetic acid (100 mCi/mmole, Amersham) was the

starting material. Instead of a two-phase separation, the

crude [3H]IAP was applied to a silica gel (Merck) column

(2.5 cm x 10 cm) and eluted with CHC13:CH3OH (9:1).

One-half ml fractions were collected and the [3H]IAP

containing fractions were pooled and concentrated. The

[3H]IAP residue was dissolved in 95% ETOH and stored in

the dark at -200C. TLC analysis was periodically performed

prior to affinity labelling reactions to check for

decomposition of the [3H]IAP.

p-Amino-chloramphenicol (ACAP)
(D-threo-l-p-aminophenyl-2-dichloroacetamido-l,3-propanediol)


ACAP was synthesized from CAP (Sigma) by sodium

dithionite reduction (63). CAP (5 g) was placed in 100 ml

0.1 M NaHCO3, pH 7.0, and sodium dithionite was added

slowly. The reduced CAP was extracted by ethyl acetate (3 x

100 ml) or was acidified and precipitated as the ACAP HCL.

The ACAP HCL was sometimes neutralized and extracted into

ethylacetate for recrystallization. ACAP was somewhat

yellow in color and remained near the origin on silica gel

TLC analyses. ACAP produced in this manner was used in the

PAP and FAP syntheses.








p-Azido-chloramphenicol (PAP)
(D-threo-l-p-azidophenyl-2-dichloroacetamido-l,3-propanediol)

PAP was synthesized from chloramphenicol (Sigma) as the

starting material. CAP was reduced by hydrogen gas over

activated pallidium-charcoal or by sodium dithionite

reduction (63, 64, 65). The reduced CAP was diazotized by

NaNO2 (0.25 g) in 50 ml 20% H2SO4 at 0C for 30

minutes after which sodium azide (0.5 g) was added followed

by a 30 minute incubation at 0C. A very pure PAP was

recovered from the aqueous phase by three ethyl acetate

extractions (50 ml each). The extracts were dried to a

yellow oil which was recrystallized from isopropanol to give

a very clean product. All reactions were carried out in th

dark to prevent photolysis of the PAP. I performed a

stability test by allowing PAP in methanol to remain in a

sealed glass vesicle on the countertop. It took more than

three days of normal room light exposure to provide a

significant decomposition of PAP. The UV spectrum of PAP

was very useful in determining its purity since it is very

distinct from CAP, and the photolytic reaction could be

followed spectrophotometrically (see Appendix III), further

distinguishing PAP from CAP.








p-Azidobenzoamido-chloramphenicol (FAP)
(D-threo-l-p-azidobenzoyl-p-amidophenyl-2-dichloroacetamido-
1,3-propanediol)

For the synthesis of FAP and chloramphenicol-p-

azidobenzoate (PAF), the activated ester of p-azidobenzoic

acid was required. All reactions were performed in the dark

to prevent photo-decomposition of the materials.

p-Azidobenzoic acid was synthesized from p-aminobenzoic

acid (Calbiochem) similar to the manner that PAP was

synthesized. p-Aminobenzoic acid (0.5 mole) was dissolved

in 300 mls of cold water with 100 ml H2SO4.

NaNO2 (0.6 mole) in 150 ml cold water was added portion-

wise over 30 minutes with continuous stirring. After an

additional one hour of stirring at 0C, sodium azide (1.0

mole) in 100 ml cold water was added portion-wise over 30

minutes and then allowed to react an additional one-half

hour. The tan, mud-like product was filtered and

recrystallized from ethanol. TLC analysis gave one

resolvable spot and the UV spectrum was consistent with

p-azidobenzoic acid (66).

The active ester of p-azidobenzoic acid was synthesized

in a like manner to the synthesis of the active ester of

iodoacetic acid. p-Azidobenzoic acid (24.5 g) and

N-hydroxysuccinimide were dissolved into 250 ml of dry

dioxane and chilled on ice to 0C. Dicyclohexyl-

carbodiimide (31 g) in 50 mls of dry dioxane at 0C was

added to the mixture and the solution was allowed to warm

to room temperature and incubate overnight. The








dicyclohexylurea was removed by filtration and the orange

filtrate was evaporated to cream colored crystals which were

recrystallized from dioxane.

FAP was synthesized from ACAP (10.0 mmole) and

p-azidobenzoylsuccinimide ester (10.0 mmole) in 50 ml of

ethyl acetate. The reaction was performed at 0C for two

hours and subsequently overnight at room temperature. The

product had precipitated as an orange clay-like material

from the ethyl acetate. This was washed several times with

water and then dissolved in DMSO. TLC and spectral analyses

showed that the FAP was free from contaminating materials.

(Photolytic experiments with FAP are given in Appendix III.)

Chloramphenicol-p-azidobenzoate (PAF)
(D-threo-2-p-azidobenzoamido-l-p-nitrophenyl-1,3-
propanediol)
PAF was synthesized from CAP base (Sigma) (10.0 mmole)

and from p-azidobenzoylsuccinimide ester (10.0 mmole)

according to the scheme given for FAP. The reactants were

dissolved in cold ethyl acetate (50 mis) and incubated for

two hours at 0C and then overnight at room temperature.

Likewise an orange clay-like precipitate occurred which was

washed with water and then dissolved in DMSO. TLC and

spectral analysis showed PAF to be free from contaminants.








p-Hydroxylamino-chloramphenicol (HACAP)
(D-threo-2-dichloracetamido-l-p-hydroxylaminophenyl-
1,3-propanediol)

HACAP was synthesized from CAP essentially following

the published procedure (67). CAP (10 g) and NH4Cl

(3 g) were vigorously stirred in 120 ml of 25% ethanol.

Zinc dust (7.8 g) was added slowly over 10 minutes after

which the solution was allowed to stir an additional 15

minutes. Thirty ml of water was added and the suspension

was filtered. The filtrate was saturated with NaCl and

extracted three times with 150 ml ethyl acetate. The

extracts were dried and the residue was dissolved in 95%

ethanol. UV and NMR spectral analyses exhibited results

that were identical to the published spectra of HACAP (67).

HACAP was stored at -200C until used.

p-Nitroso-chloramphenicol (NOCAP)
(D-threo-2-dichloroacetamido-l-p-nitrosophenyl-l,3-
propanediol)

NOCAP was also synthesized essentially following the

published procedure (67). CAP (10 g) was reduced to HACAP

as described above except the procedure was ended prior to

ethyl acetate extraction. The aqueous phase containing the

HACAP was chilled on ice to 0C and poured slowly into

100 ml of FeC13- 6H20 (20 g) at 0C. The solution

immediately turned green. An additional 30 minutes of

stirring at 0C was allowed. The solution was extracted

three times with 75 ml of diethyl ether. The ethereal phase

was evaporated to a blue-green residue. The residue was

chromatographed on a silica gel (Merck) column (2.5 cm x 30








cm) with CH3C1:CH3OH (95:5) as the eluent. A blue

band trailed by a yellow band were resolved and the blue

band was collected and recrystallized from diethyl ether to

give pale blue crystals. UV spectroscopy gave a spectrum

identical to authentic NOCAP (67). [ C]NOCAP was

synthesized from [14C]chloramphenicol (14.9 mCi/mmole)

(Amersham) in a reaction scaled down from the scheme given

above. The [ 14C]NOCAP migrated coincident with the

unlabelled material with no radioactivity elsewhere on the

TLC plate (silica gel, CH3Cl:CH3OH, 95:5).

The Peptidyl Transferase Assay

The PTase assay employed has been published in part

previously (34). 0.5 to 1.0 A260 of ribosomes were

required for each data point to be collected in the assay.

Each concentration of antibiotic was assayed in duplicates

or triplicates which were averaged for the final value.

Many of the antibiotics were assayed on several different

occasions with no change in the expected inhibition value.

The ribosomes were placed in 90 pl of Buffer K to which

50 pl of 95% ETOH or 50 pl of antibiotic solution in 95%

ETOH was added. The reaction was initiated by the addition

of 10 pV of N-acetyl-[3H]leucyl-tRNA (10,000 cpm) to

the ribosomal mixture (total volume 150 pl). (The

N-acetyl-[ H]leucyl-tRNA was received as a gift from

Nancy Denslow.) The reaction was allowed to proceed for ten

minutes at room temperature and terminated by the addition

of 10 pl of 10 N KOH. A further incubation for three








minutes at 400C followed. The solutions were neutralized

by the addition of 1 ml of 1 M sodium phosphate pH 7.0.

Each sample was extracted with 1.5 ml ethyl acetate and the

organic phase was separated from the aqueous phase by low

speed centrifugation. One ml of the upper phase, which

contained the N-acetyl-[3H]leucyl-puromycin adduct, was

collected and added to 10 ml ACS (Amersham) liquid

scintillant and then the samples were quantified by liquid

scintillation counting. Sample reactions without ribosomes

were used to register the background and reactions with

ribosomes but without antibiotic present were used to give

the untreated control values. The results are presented as

a percentage of the untreated control after background

subtraction. The antibiotic solutions were made fresh prior

to each assay and assays were conducted in subdued ambient

lighting to prevent any photolytic decomposition of the

antibiotics during the assay.

Apparent Dissociation Constant (K'd) Calculations

As a measure of how well the CAP analogs could interact

with the ribosomes, an apparent dissociation constant

(K'd) was calculated from the PTase inhibition data

using the method of Denslow and O'Brien (34). The method

transformed the data into a Scatchard format so that the

K'd could be evaluated from the slope (68). The formula

of transformation is: i/[I] = -(i/K'd) + (n/K'd),

where i is the fraction of inhibition defined as l-(percent

of control/100), [I] = antibiotic concentration, n = number








of binding sites, and K'd = apparent dissociation

constant. From the plot of i/[I] versus i, the negative of

the reciprocal slope (-l/slope) is the K'd and n, the

abscissa intercept is the number of binding sites. The

slope was evaluated by linear regression using a Texas

instrument SR-51-II hand-held calculator.

Structure-Activity Relationship of CAP Analogs

Previously, CAP analogs have been examined for a

structure to activity relationship, primarily by medicinal

chemists (29, 35, 36, 37, 38). In some instances, the Hansch

equation has been used to correlate the structure to the

activity and in others a measure of a physical feature of the

structure has been plotted versus the measured activity to

examine for correlations (36, 38, 69, 70). I chose the

latter route and found literature values for the properties

of the para substituents. The Hammett sigma values were used

as a measure of the electronegativity and were 0.78 for the

nitro, 0.15 for the azido, 0.12 for the nitroso, -0.34 for

the hydroxylamino, and -0.66 for the amino groups (71). The

shift in the infrared absorbance frequency upon hydrogen bond

formation was used as a measure of the hydrogen bond
-l
strength; these values were 346 cm- for the nitro, 240

cm- for the nitroso, 175 cm- for the amino, and 40

cm-1 for the azido groups (72, 73). The infrared

frequency shifts (i.e. hydrogen bond strengths) had been

measured in a manner which would be similar to the situation

involved when CAP is in its binding site. The activities










of the ribosomes were computed from the K'd values

obtained from the PTase assays using the formula: log Act =

log [(K'd of CAP/K'd of CAP analog) x 100], where

K'd of CAP is 38 pM and 9.8 iM for the mitochondrial

and bacterial ribosomes, respectively, and K'd of CAP

analog represents the calculated K'd of the CAP analog

tested. This equation is a transformation of an equation

used previously to calculate CAP analog activities (38).

Labelling Reactions

The mitochondrial ribosomes were retrieved from the

-700C freezer and suspended in Buffer L, which lacks

2-mercaptoethanol (2ME). The ribosomes were centrifuged at

30,000 rpm for 16 hours in a Beckman type 65 rotor over a

20% sucrose cushion of Buffer L to remove traces of 2ME

prior to the labelling reactions. The supernatant was

discarded and the ribosomal pellets were resuspended into

Buffer L. A small aliquot was removed for an A260
260
measurement and the ribosome concentration was adjusted to

give a final concentration of 1 A260/10 pl after the

addition of the labelling reagent. The labelling reagent

[3H]IAP (100 mCi/mmole), [1C]NOCAP (14.9 mCi/mmole)
14
or [14C]iodoacetamide (14.7 mCi/mmole), was added to the

ribosome solution in 95% ETOH to give a final ETOH

concentration of 10%. In the reactions including CAP as the

competitive blocker, one-half of the ethanol added contained

CAP and the other half contained the labelling reagent.









N-Acetyl-AA-tRNA and AA-tRNA were a gift from Nancy Denslow

for use in the [3H]IAP affinity labelling studies and

were added to the labelling reaction solution in water. The

AA-tRNA may contain some uncharged tRNA and the

N-acetyl-AA-tRNA was greater than 95% acetylated, so that

AA-tRNA may not be fully confined to the A site, but the

N-acetyl-AA-tRNA would be expected to bind mainly to the P

site. Regardless, they will still be binding to a

restricted region of the ribosomal surface. The labelling

reactions were terminated by the addition of a large excess

of 100 mM 2ME in Buffer L (7 ml) or in 95% ETOH (1 ml).

Reactions terminated with 100 mM 2ME in Buffer L were

centrifuged at 30,000 rpm for 16 hours in a Beckman type 65

rotor to concentrate the ribosomes. The reactions

terminated by the addition of 100 mM 2ME in 95% ETOH were

chilled, and the ribosome precipitate was collected by

centrifugation for 20 minutes in a Beckman Microfuge. The

ribosomes were resuspended into an appropriate buffer for

the extraction of the ribosomal proteins for polyacrylamide

gel electrophoresis.

Sucrose Density Gradient Separation of Labelled
Ribosomal Subunits

Subsequent to the [3H]IAP labelling of the 55S

ribosomal monosome, a separation and isolation of the

individual subunits were required. The labelled ribosomal

pellets were resuspended in 2.0 ml Buffer G to bring about

monosome dissociation into subunits. The resuspended








ribosomes were loaded on a linear 10% to 30% sucrose

gradient of Buffer G in a Beckman SW 27 rotor and

centrifuged at 23,000 rpm for 18 hours. The gradient was

fractioned under continuous UV monitoring at 260 nm by a

Gilford 2400 UV spectrophotometer. The fractions containing

subunits were pooled and centrifuged at 30,000 rpm (Beckman

type 65 rotor) for 22 hours to concentrate the ribosomal

subunits. The ribosomal pellets were resuspended in 50 Vl

of Buffer G and an aliquot was removed for an absorbance

measurement. Equal amounts of ribosomal subunits were then

prepared for electrophoresis.

Polyacrylamide Gel Electrophoresis (PAGE)

One-Dimensional SDS PAGE

Unidimensional PAGE was conducted using a modified

Laemmli system (74, 75). The modifications consisted of

doubling the separator gel buffer (Buffer M) concentration

and the stacking gel buffer (Buffer N) concentration, which

helped to prevent streaking artifacts of the migrating

proteins. The polyacrylamide concentration in the separator

gel was 15.6% T 3.3% C and the stacking gel was 10% T 2.6%

C, to provide resolution of proteins from approximately

8000 MW to 70,000 MW. The slab gels measured 130 mm x 240

mm x 1.5 mm and were electrophoresed on a two-slab

laboratory made unit. Ammonium persulfate was added at

0.03% and 0.1% to initiate polymerization of the separator

and stacking gels, respectively. The separator gel was

overlayed with isobutanol after the polymerization was








initiated to provide an even surface for the stacking gel.

Residual isobutanol was washed away prior to the addition of

the stacking gel. The separator gel was allowed one full

hour to polymerize before the stacking gel was added, which

was also allowed to polymerize with a ten-well comb in place

for one hour.

The ribosomal pellets were suspended in Buffer O and

heated at 600C for at least 30 minutes prior to loading

onto the gels. Tank buffer (Buffer P) was used to flood the

wells of the stacking gel and the samples consisting of

20 pl to 50 pl each were underloaded through the tank

buffer via micropipets and a hand-held micro-pipetting

device. The PAGE was conducted at an initial constant

current of 15 mA per gel (30 mA for two gels) and was

increased to 20 mA per gel (40 mA total constant current)

after the Bromophenol Blue tracking dye had fully entered

the gel. The voltage routinely began at about 70 volts

which plateaued at about 220 volts near the termination

time. The gels were electrophoresed for 16 to 24 hours or

for about two hours after the tracking dye had migrated

off the bottom of the gels. The gels were removed after

termination and placed in a fix bath of 25% isopropanol

and 10% acetic acid overnight to remove the SDS.

Subsequently, the gels were stained with 0.25% Coomassie

Brilliant Blue R-250 in 50% ethanol and 7.5% acetic acid

for four to six hours. The gels were destined by

repeated changes of a destain solution of 5% ethanol and









10% acetic acid. After sufficient destaining, the gels

were prepared for fluorography (see below). The outside

lanes were used to resolve molecular weight standards,

which had been obtained from Sigma and Calbiochem. They

consisted of 1) bovine serum albumin, 68,000 MW; 2)

ovalbumin, 44,500 MW; 3) human gammaglobulin, 50,000 MW

and 23,500 MW; 4) carbonic anhydrase, 30,000 MW;

5) myoglobin, 17,200 MW; 6) lysozyme, 14,400 MW; and 7)

bovine cytochrome C, 12,500 MW, which were stored in Buffer

0. Molecular weights of ribosomal proteins were calculated

from a plot of the migration position of these standards

versus their molecular weights (76).

Two-Dimensional PAGE

Bidimensional PAGE was conducted using the Leister and

Dawid system as has been previously modified in our

laboratory (13, 77, 78, 79). The ribosomal pellets obtained

from the labelling reaction were resuspended in 50 pl of

Buffer L and 65 mg of high purity urea (Beckman), 15.3 mg of

LiC1, 2 il of 1 N HC1, and 1 pl of 10% butylatedhydroxy-

toluene in ethanol were added after which the volume was

adjusted to 120 pl. The final extraction solution

consisted of 9 M urea and 3 M LiC1 with a pH of about

three. The extraction mixtures were allowed to stir

overnight (about 16 hours) in small conical vials at 40C.

The rRNA was removed from the extraction solution by

centrifugation at 50,000 rpm in a Beckman type 65 rotor

for one hour. The rRNA precipitate was re-extracted in








80 il of 8 M urea, 3 M LiCI, pH 3.5 for about ten hours

to dislodge any ribosomal proteins that had become

entrapped in the rRNA precipitate. The combined

supernatants were dialyzed overnight against sample buffer

(Buffer Q).

The first dimension gel was designed to resolve basic

proteins and was formed in a tube 2 mm in diameter and

120 mm long. The gel consisted of a separator gel (4.6% T

3.2% C), stacking gel (4.0% T 3.5% C), and a sample gel

(4.7% T 3.5% C) of Buffer R, Buffer S, and Buffer T,

respectively. The separator gel was polymerized by the

addition of 0.1% total anmoniaT persulfate and was

pre-electrophoresed in Buffer Q at 0.2 mA per gel overnight,

before the addition of the stacking and sample gels, which

contained riboflavin (0.001%) in addition to ammonium

persulfate (0.02%) so that the acrylamide could be induced

to photo-polymerize. Pyronin Y (0.02%) was used as the

tracking dye and the gels were electrophoresed under

constant current towards the cathode at 0.2 mA per gel until

the dye was one centimeter into the separator gel whereupon

the current was increased to 0.5 mA per gel. The tank

buffer was buffer U. The run was terminated when the

tracking dye reached the bottom of the tubes.

The second dimension gels were slab gels (180 mm x 200 mm x

1.5 mm) consisting of a separator (Buffer V) and stacking

(Buffer W) gels of 10% T 3.5% C and 4% T 3.5% C acrylamide

concentrations, respectively. The first dimension tube gel








was extruded from the glass tube by water pressure, placed

on the top of the slab gel, and allowed to polymerize into

the stacking gel. Molecular weight marker proteins were

placed in wells on both sides of the tube gel along with an

agarose plug of Bromophenol Blue tracking dye. The second

dimension gels were electrophoresed at 40 to 45 mA, in

Buffer X (tank buffer), until the dye reached the bottom of

the gels, generally after about 24 hours. Upon termination,

the gels were fixed and stained as described above.

Fluorography

The stained polyacrylamide gels were originally

prepared for fluorography by the method of Bonner and Laskey

(80). The gels were dehydrated by two washes of 1.2 liters

DMSO for 30 minutes each, then 230 ml of a 20% solution of

2,5-diphenyloxazole (PPO) in DMSO was added for three hours

to impregnate the gels with fluor. Finally, the fluor was

precipitated in the gels by a one hour water wash. The

fluor-impregnated gels were dried under heat and vacuum on

filter paper (Whatman 3MM) by a Biorad gel dryer for 1.5

hours. More recently the gels were impregnated with sodium

salicylate as the fluor (81). The gels were washed in water

for about 30 minutes to remove the destain solution. Next

the gels were soaked in ten volumes of 1 M sodium salicylate

pH 6-7 for 30 minutes after which they were dried as above.

The dried, fluor-impregnated gels were placed directly

against Kodak XRP-5 medical x-ray film which had been pre-








flashed with a photographic strobe as described by Laskey

and Mills (82). The cassettes containing the gels and

film were stored at -700C for exposure. After an

appropriate time interval the film was removed and

developed in an automated x-ray developer.

Densitometry

The fluorograms obtained from the labelled ribosomal

protein gels were scanned by a light transmission

Densitometer (EC 910), while the tracings were recorded on a

Heathkit strip-chart recorder.

Cutting and Digestion of Protein Bands or
Spots from Polyacrylamide Gels

The stained protein bands or spots were cit from the

dried gels using a small pair of scissors. The paper

backing was removed and the dry plugs were placed in

screw-type scintillation vials. Hydrogen peroxide (200 il)

was added to each vial to hydrate and partially digest the

gel plugs. After one or two days of hydration at room

temperature the vials were frozen and 500 l of NCS

(Amersham) was added to solubilize the labelled protein from

the residual of the gel plugs. After one or two more days

of soaking at room temperature, the gel plugs were usually

totally digested and free of color. Once again the vials

were frozen and then 10 ml ACS (Amersham) liquid

scintillation cocktail was added to each vial. The samples

were allowed to become quiescent before liquid scintillation

counting by storage in the dark for two or three days.









Radioactive standards were added to vials containing gel

plus of the molecular weight marker protein to examine for

the loss of radiolabel during the digestion procedure, which

was generally less than 5%.

Liquid Scintillation Counting (LSC)

Liquid scintillation counting was conducted in a

Beckman LS8000 liquid scintillation counter using the

preselected programs. The counting efficiency was

generally about 20% for tritium and about 75% for

carbon-14.
















CHAPTER III

RESULTS





It has been known for some time that chloramphenicol can

inhibit protein synthesis within mitochondria, but very little

has been reported about the interaction of CAP with the

ribosomes from mammalian mitochondria (83, 84, 85, 86). The

goal of the present study was to examine the structural and

functional features of the CAP binding site from these

ribosomes. Two approaches were employed in the examination of

the CAP binding site and the results are reported in two

parts. First, I report the initial examination of CAP

analogs as inhibitors of the PTase activity of mitochondrial

ribosomes in comparison to the inhibition of the PTase

activity of bacterial ribosomes. This report includes the

first usage of the CAP reduction intermediates and products as

inhibitors of the PTase activities of bacterial and

mitochondrial ribosomes. In addition, novel photo-activatable

CAP analogs are introduced and tested in the PTase assay. A

new structural and functional relationship of CAP and its

ribosomal binding site was developed and is reported at the

conclusion of the first half of the Results.








The second half of the Results reports the first use

of the electrophilic CAP affinity analogs for the

identification of proteins at the CAP binding site and

within the domain of the Peptidyl Transferase Center of

mammalian mitochondrial ribosomes.

Functional Studies of the Chloramphenicol Binding
Site Through the Use of Chloramphenicol Analogs

One of the approaches employed to gain information

about the functional features of the CAP binding site was

through the use of analogs of chloramphenicol where specific

changes have been made to the parent compound and the

resultant changes in binding were then measured. In this

manner, a functional understanding of the CAP binding site

could be gained without having to dissect out the individual

components of the ribosome. The constraints on

chloramphenicol and its binding site could be determined,

along with the types of interactions between CAP and its

binding site.

Chloramphenicol was modified in either the dichloro-

acetate chain or the p-nitro moiety to produce the CAP

analogs and the peptidyl transferase assay was employed to

determine how well the analogs were bound to bacterial

ribosomes and mammalian mitochondrial ribosomes.















CI
CI

HN


o OH

tmEuHmNIcOL (CAP)


Cl

-=0
HN

HOH
I HN OH

0P-AIDENZY-CHO (FP)
P-AZIDOBENZOYL-C&LoRAMPHENIcoL (FTO)


HN

%7^N OH
-NQ OH
CHLWPHENICL-P-AZIDOBZMTE (P F)


CI

C IO
>- 0
HN

NN-N O OH
OH


P-AZIDO-CHLORPHENICOL (PAP)



Figure 1. Comparison of the structures of chloramphenicol
and the photo-activatable chloramphenicol analogs.
(Hydrogens have been left off for clarity.) Notice that
the phenylazido substituent of PAF is larger than the
dichloroacetate of CAP.












NO2
1 2


H CI H CI
HO NCI | HO&. cC I
HO 0 HO 0
CHLORAMPHENICOL (CAP) P-HYDROXYLAMINO CHLORAMPHENICOL (HACAP)

N-O NH2

0 0
H CI CI
HO HNACI HO N ACI
I I I I
HO 0 HO 0
P-NITROSO CHLORAMPHENICOL (NOCAP) P-AMINO CHLORAMPHENICOL (ACAP)



Figure 2. Comparison of the structures of chloramphenicol
and the chloramphenicol reduction analogs. (Hydrogens
have been left off for clarity.) Only the para position
should bring about an alteration in the binding of these
analogs since the rest of the molecule is left intact.








Chloramphenicol Analogs Modified in the Para-Nitro Moiety

CAP analogs with modifications in the para-nitro

moiety belonged to two groups in this study. Two of the

CAP analogs were potential photo-activatable affinity

probes (Figure 1) by having aromatic azide substitutions

onto CAP and the three other CAP analogs were the closely

related family of nitro-reduction products (Figure 2),

from chloramphenicol as the parent compound.

Chloramphenicol (CAP)

Mitochondrial ribosomes and bacterial ribosomes were

subjected to CAP (0.01 mM to 1.0 mM) inhibition of their

PTase activities. Details of the PTase assay are given in

the Materials and Methods chapter. Briefly, the ribosomes

are aliquoted (0.5 to 1.0 A260) and the antibiotic is

added at the proper concentration. The reaction is

initiated by the addition of N-acetyl-[3H]leucyl tRNA

and after 10 minutes the reaction is terminated by the

addition of 10 N KOH. The reaction mixture is neutralized

with sodium phosphate buffer and then extracted with ethyl

acetate. The ethyl acetate phase contains the N-acetyl

3 H]leucyl-puromycin adduct which is counted by LSC (see

Materials and Methods). The control values are obtained

from the puromycin-adduct formation by ribosomes in the

absence of antibiotic. The data is presented as a percent

of the untreated control value.

CAP inhibits the PTase activities of bacterial and

mitochondrial ribosomes very well (Figure 3). Using the





















100






50

0*


0
5 I 4 I0 3 i-2

Antibiotic concentration (M)





Figure 3. Comparison of the peptidyl transferase
activities of mitochondrial ribosomes and bacterial
ribosomes in the presence of chlorainphenicol (CAP).
Mitochondrial 39S subunits (* ) and bacterial 50S
subunits (A ) were treated with CAP at the indicated
concentrations and assayed for their peptidyl transferase
activities. The activities are reported as a percentage
of the activities of ribosomes assayed in the absence of
antibiotic. From Scatchard analysis of the data, a
K' of 38 iM and 9.8 pM were calculated for the
mitochondrial and bacterial ribosomes, respectively.

















TABLE II


Apparent Dissociation Constants
for Chloramphenicol and
the Chloramphenicol Analogs


K'd (pM)

Ribosome Type

Mitochondrial Bacterial


Analog


CAP

ACAP

FAP

HACAP

IAP

NOCAP

PAF

PAP


38*

362

970

113

147

190

n.a.

n.a.


9.8

103

860

98

n. d.

83

n. a.

n. a.


*all values have units of pM

n.a. not active

n.d. not determined

The apparent dissociation constants were calculated
using a Scatchard format as described in the
Materials and Methods.








inhibition data transformed into the Scatchard format, an

apparent dissociation constant (K'd) for CAP of 9.8 pM

was calculated for bacterial ribosomes and a K'd of

38 pM, for mitochondrial ribosomes (Table II) (34). In

general, bacterial ribosomes are more sensitive to

peptidyl transferase inhibitors than mitochondrial

ribosomes (34). This relationship holds true for

chloramphenicol and the CAP analogs tested in the present

study.

p-Amino-chloramphenicol (ACAP)

ACAP was the fully reduced analog of CAP where an amino

group replaced the p-nitro moiety (Figure 2). The PTase

activities with ACAP as an inhibitor were tested with

concentrations of ACAP ranging from 0.01 mM to 5.0 mM (see

Materials and Methods). ACAP was the weakest inhibitor of

the CAP reduction products (Figure 4c and Table II), yet it

remained an effective inhibitor of the PTase activities.

Soon after the discovery of chloramphenicol, it was

suggested that ACAP would not inhibit protein synthesis in

bacteria (87). This followed the mistaken notion that the

bacterial nitro-reductase activity would detoxify CAP by

reducing the nitro moiety to an inactive amine; however,

ACAP itself had not been tested for activity by the

investigators (87). As shown here, the reduced

chloramphenicol did inhibit the PTase activities of both

bacterial and mitochondrial ribosomes.
















Figure 4.
100 A Comparison of the
peptidyl
transferase
activities of
mitochondrial
ribosomes and
50 bacterial
ribosomes in the
presence of the
chloramphenicol
reduction analogs.
Mitochondrial 39S
0 subunits (* ) and
bacterial 50S
100 subunits (A) were
treated with
p-hyroxylamnino-
chlorampheni ol
(HACAP) (A),
Sp-nitroso-
50 chloramphenicol
S(NOCAP) (3), or
S\ \p-amino-
chloramphenicol
(ACAP) (C), at the
indicated
0 rI I I I concentrations and
assayed for their
0oo \ c peptidyl
transferase
activities. The
activities are
reported as a
Percentage of the
activities of
ribosomes assayed
\ in the absence of
antibiotics.
A
0
oc nc* Im ITM
Antaotic concem rtlIon (M)








p-Hydroxylamino-chloramphenicol (HACAP)

HACAP was a partially reduced CAP analog where the

hydroxylamino group replaced the p-nitro moiety (Figure 2).

HACAP was tested for the inhibition of the ribosomal PTase

activities over the concentration range 0.01 mM to 5.0 mM

(see Materials and Methods) and found to be a very effective

inhibitor of the PTase activities of bacterial and

mitochondrial ribosomes (Figure 4a). Although HACAP did not

bind to ribosomes as well as CAP, it was the best CAP analog

that I have synthesized in regard to its interaction with

the mitochondrial ribosomes (Table II).

p-Nitroso-chloramphenicol (NOCAP)

NOCAP (Figure 2) was synthesized with a twofold

purpose: 1) as a reduction product analog of CAP and a

reversible inhibitor of the PTase activities of the

bacterial and mitochondrial ribosomes and 2) as a potential

electrophilic affinity probe for the CAP binding site.

NOCAP was used likewise from 0.01 mM to 5.0 mM to

inhibit the PTase activities of the ribosomes (see Materials

and Methods) and NOCAP was found to be a good inhibitor of

the PTase activities of both the bacterial ribosomes and the

mitochondrial ribosomes (Figure 4b and Table II). With the

criterion of "interaction" with the ribosome met, the

continued testing of NOCAP for use as an affinity probe was

warranted. (The use of NOCAP as an affinity probe is

discussed below.)

























I00






S50
.0




0
0 0 10 0I,
Io6 Io4 Io3 Io2

Antibiotic concentration (M)







Figure 5. Comparison of the peptidyl transferase
activities of mitochondrial ribosomes and bacterial
ribosomes in the presence of p-azido-chloramphenicol
(PAP). Mitochondrial 39S subunits ( *) and bacterial 50S
subunits (A) were treated with PAP at the indicated
concentrations and assayed for their peptidyl transferase
activities. The activities are reported as a percentage
of the activities of ribosomes assayed in the absence of
antibiotic. From the lack of effect on the peptidyl
transferase activity, PAP was assigned a biological
activity of zero for the structure-activity analysis.


A I I ~ I a I
A
A r~










p-Azido-chloramphenicol (PAP)

PAP (Figure 1) had previously been synthesized for use

as a photo-affinity probe for the CAP binding site of

bacterial ribosomes (64, 65). The authors presented data

which suggested that there was some inhibitory activity

towards the PTase Center of bacterial ribosomes (65). In

the present study, PAP was found to provide no inhibition of

the PTase activities of bacterial or mitochondrial ribosomes

when tested over a broad concentration range (0.01 mM to

18.5 mM) (Figure 5). (The PTase assay was conducted in the

dark to prevent photo-activation of the azido function of

PAP, as described in the Materials and Methods.)

Since PAP does not bind to either ribosome, it was

removed from consideration as an affinity probe and was

assigned a biological activity of zero for the

structure-activity relationship to be discussed below.

p-Azidobenzoamido-chloramphenicol (FAP)

FAP was the second of the photo-activatable CAP analogs

modified in the p-nitro position (Figure 1). (The PTase

assay involving FAP was also carried out in the dark, see

Materials and Methods.) The added phenylazide provided the

photo-activatable nucleus, but did not prevent FAP from

inhibiting the PTase activities of the ribosomes (Figure 6

and Table II). It had been shown that CAP could accept

bulky substituents from the p-nitro position without losing

its ability to inhibit the PTase activities of bacterial


















SI I I '


100





A *
50






0 II II I
io"5 I 4 1c10 Ic2
Antibiotic concentration (M)






Figure 6. Comparison of the peptidyl transferase
activities of mitochondrial ribosomes and bacterial
ribosomes in the presence of p-azidobenzoamido-
chloramphenicol (FAP). Mitochondrial 39S subunits ( )
and bacterial 50S subunits (A) were treated with FAP at
the indicated concentrations and assayed for their
peptidyl transferase activities. The activities are
reported as a percentage of the activities of ribosomes
assayed in the absence of antibiotic.








ribosomes. With this fact in mind, FAP had been synthesized

as a novel photo-affinity analog of CAP.

Chloramphenicol Analogs Modified in the Dichloroacetate
Chain

Idoamphenicol (IAP)

Iodoamphenicol (Figure 7) had previously been used as

an affinity probe for the CAP binding site in bacterial

ribosomes (59, 61). IAP was therefore selected for use as

an affinity probe for mammalian mitochondrial ribosomes.

(Details of the affinity labelling experiments to be

discussed below.) To be useful as an affinity probe, an

antibiotic must be shown to interact with the ribosomes.

IAP (0.01 mM to 5.0 mM) was allowed to inhibit the PTase

activity of the mitochondrial ribosomes under conditions

where irreversible inactivation of the Ptase activity was

thought to be negligible (see Materials and Methods).

IAP is an effective reversible inhibitor of the PTase

activity of the mitochondrial ribosomes (Figure 8). From

Scatchard analysis of the inhibition data, IAP has a K'd
d
of 147 pM where CAP has a K'd of 38 pM.

Chloramphenicol-p-azidobenzoate (PAF)

PAF (Figure 1) with p-azidobenzoate substituting for

the dichloroacetate, was synthesized as a potential

photo-activatable affinity probe for the CAP binding site.

PAF has been independently synthesized and tested for the

inhibition of the PTase activity of bacterial ribosomes by

another group of investigators and their data presented


















NO2 CAP NO, APN


6 0
HOC HO



H N I
HO 6 HO1











Figure 7. Comparison of the structures of
chloramphenicol, iodoamphenicol, and iodoacetamide.
Chloramphenicol (CAP), iodoamphenicol (IAP), and
iodoacetamide (IAM) are compared, where IAP has an
iodoacetate instead of the dichloroacetate. IAP can be
considered a substituted IAM, both contain the reactive
alkyl halide, but IAM does not contain the active site
directing p-nitrophenylpropanediol moiety.




















100




O 50- -
0

O 0
0
0
105 10-4 13 l0-2

Antibiotic concentration (M)





Figure 8. Comparison of the peptidyl transferase
activities of mitochondrial ribosomes in the presence of
chloramphenicol (CAP) or iodoamphenicol (IAP).
Mitochondrial 39S subunits were treated with CAP (0) or
IAP ( ) at the indicated concentrations and assayed for
their peptidyl transferase activities. The activities are
reported as a percentage of the activities of ribosomes
assayed in the absence of antibiotic.





63

















100-

*



S50
0




0 I I II I
10l a4 10S 10-2
Antibiotic concentration (M)








Figure 9. Comparison of the peptidyl transferase
activities of mitochondrial ribosomes and bacterial
ribosomes in the presence of chloramphenicol-p-
azidobenzoate (PAF). Mitochondrial 39S subunits (0) and
bacterial 50S subunits (A) were treated with PAF at the
indicated concentrations and assayed for their peptidyl
transferase activities. The activities are reported as a
percentage of the activities of ribosomes assayed in the
absence of antibiotic. From the lack of effect on the
peptidyl transferase activities, PAF was assigned a
biological activity of zero for the structure-activity
analysis.








little promise that PAF would be a useful CAP analog (88).

In the present study, PAF was tested for its ability to

inhibit the PTase activities of mitochondrial and bacterial

ribosomes over a range of concentrations from 0.01 mM to 5.0

mM (the assay was performed in the dark, see Materials and

Methods). PAF was found not to inhibit the PTase activities

of either kind of ribosome (Figure 9); consequently PAF was

discarded from consideration as an affinity probe and was

assigned a biological activity of zero for the

structure-activity relationships (discussed below).

Structure-Activity Relationships of Chloramphenicol
Analogs

Two major concepts were developed concerning the CAP

binding site of bacterial ribosomes and mitochondrial

ribosomes, involving two domains of the CAP molecule, the

dichloroacetate chain and the para-nitro moiety. The

dichloro group has a size constraint, so that an analog such

as IAP is still capable of binding to the CAP site, since

the iodo substitution has a smaller size than the dichloro

moiety. On the other hand, PAF cannot, since the

phenylazido substitution has a much larger bulk than the

dichloro group of CAP.

Using bacterial growth as an indication of the

inhibitory activity of CAP analogs, it has been reported

previously that the molar volume and electronegativity of

the dichloro position are important for the binding of CAP

to the CAP binding site (35, 36, 37, 89, 90, 91). Even








though only two CAP analogs with dichloroacetate chain

substitutions were tested in this study, it is apparent from

these data that the molar volume of this position is an

important determinant as to whether the CAP analog will bind

to mitochondrial ribosomes, as well.

Several studies have examined CAP analogs with p-nitro

substitutions and the resultant inhibition of bacterial

growth as an indicator of the binding properties of the

analogs. From the studies, the electronegative property of

the para position had been suggested to be the major

contribution of the p-nitro moiety for the binding of CAP to

the CAP binding site of bacterial ribosomes (36, 38, 39,

91).

In the present study, a "family" of closely related CAP

analogs, the nitro reduction products and PAP, have been

tested for their abilities to bind to bacterial and

mitochondrial ribosomes. To examine the relationship of the

electronegativity of the para position to the ability of the

CAP analogs to bind to the ribosomes, the Hammett sigma

value, (as a measure of electronegativity), was plotted

versus the activity of the CAP analog as measured in the

PTase assay. A poor correlation was observed (Figure 10).

ACAP and HACAP are much more active than their respective

electronegativities would predict, whereas PAP, which is

inactive, would be predicted to bind to the ribosomes about

as well as NOCAP.




















1.0 -
,ph NO2
0.5 -
N
0.0 N
/ A NHOH
-05 -
/ NH
/2
-1.0 -- -- --- -- --- -
400 -
SNO2
300 -
Yy A r NO
200 0
NH2


J "3
0 I I I 3
0.0 0.5 1.0 15 2.0
log Act.





Figure 10. Structure-activity relationships of chloram-
phenicol analogs with para-nitro substitutions. In the
upper panel is plotted the Hammett sigma value ( ') to
activity relationship, which exhibits a low correlation.
In the lower panel is plotted the shift in infrared
aborbance frequency (4y), as a measure of hydrogen bond
strength, to activity, which exhibits a high degree of
correlation. The physical parameters were found in (71,
72, 73) and the activities for the mitochondrial ribosomes
( ) and bacterial ribosomes (A ) were determined as
reported in the Materials and Methods. The para substitu-
ents are listed on the right; NO2 (CAP), N3 (PAP),
NO (NOCAP), NHOH (HACAP), and NH2 (ACAP). The dashed
lines represent a direct correction of the physical
parameters of the para positions to the activity of the
antibiotic analog. The open circle and open triangle
represent the hydrogen bond strength to activity
relationship of HACAP for mitochondrial and bacterial
ribosomes, respectively.








Besides being extremely electronegative, another

striking property of nitro groups is their ability to form

hydrogen bonds (H-bonds) with suitable donors (72, 73, 92).

Using the shift in the infrared absorbance frequency of

donor hydroxyl groups forming hydrogen bonds with the para

substituents, a good correlation between the H-bond strength

and the inhibitory activities of the CAP analogs was found

(Figure 10). PAP, with the azido substitution, binds poorly

to the ribosomes and is a poor H-bond acceptor (72, 73, 92).

Electronegativity has been loosely correlated with

hydrogen bond strength, but features such as the

directionality of the H-bond are not involved in the

correlation (92). The previous suggestion of the para-nitro

position requiring a high electronegativity for good CAP

analog binding may have been because the substituents used

had a high degree of correlation between the H-bond

strengths and the electronegativities.

I propose that the essential feature of the para-nitro

position of CAP is its ability to act as a hydrogen bond

acceptor to a hydrogen bond donor in the CAP binding site,

and that this arrangement plays a role in the anchoring of

CAP to the CAP binding site of bacterial ribosomes and

mitochondrial ribosomes.








Affinity Labelling for Structural (and Functional?)
Proteins of the Peptidyl Transferase Center
and the Chloramphenicol Binding Site of Mammalian
Mitochondrial Ribosomes

Affinity labelling is the technique with the highest

potential for allowing the identification of structurally

important ribosomal proteins in the CAP binding site, which,

in addition, may be functionally important for the peptidyl

transferase activity, as well as for the binding of CAP.

Verification of the functional importance of the affinity

labelled proteins can be obtained through the use of other

techniques such as salt stripping and functional

reconstitution (40, 41, 42).

Two electrophilic CAP analogs were used in this study

to probe the CAP binding site by affinity labelling the

ribosomal proteins therein. Iodoamphenicol was the

dichloroacetate substituted analog of CAP and p-nitroso-

chloramphenicol was the p-nitro substituted analog of CAP

used, where the distal positions of the parent molecular

structure of CAP have the electrophilic modifications. In

this manner, it was expected that proteins associated with

the binding of the distal portions of CAP might be

affinity labelled, if indeed more than one protein

comprises the CAP binding site, and if the proteins

contain the appropriate nucleophiles.








Iodoamphenicol as the Affinity Probe

lodoamphenicol was among the first affinity labelling

reagents used to probe for proteins of the CAP binding site

of bacterial ribosomes (59, 61). Therefore, IAP was

selected to be used to affinity label proteins from the CAP

binding site of mammalian mitochondrial ribosomes. Dr. Olaf

Pongs had donated a small sample of unlabelled IAP to our

laboratory for the initial examination of the utility of IAP

as an affinity probe for the mitochondrial ribosomes. I

synthesized a fresh stock of unlabelled IAP (see Materials

and Methods) to complete the tests of the interaction of IAP

with the mitochondrial ribosomes. For the affinity

labelling experiments, I synthesized [ H]IAP from

[3H]iodoacetic acid and CAP base as the starting

materials (see Materials and Methods).

Effect of iodoamphenicol on the peptidyl transferase
activity of mitochondrial ribosomes

Reversible effect. Iodoamphenicol reversibly

inhibits the PTase activity of mitochondrial ribosomes in a

manner similar to CAP (Figure 8). The interaction of IAP

(K'd 147 JM) is about fourfold lower than the

interaction of CAP (K'd 38 pM) with the mitochondrial

ribosomes. Bacterial ribosomes have the same relationship

of a fourfold difference, with K' 's for IAP and CAP of

15 pM and 4 pM, respectively (59).

Irreversible effect. Since IAP binds reversibly to

the mitochondrial ribosomes, it was expected that a








prolonged incubation of the ribosomes in the presence of IAP

would allow IAP to covalently attach to the CAP binding site

of the ribosomes and thereby irreversibly inhibit the PTase

activity associated with the ribosomes. It had been

determined in our laboratory that an incubation of the

mitochondrial ribosomes with iodoamphenicol would bring

about some irreversible inactivation of the PTase activity

of these ribosomes (Mary Conde, personal communication).

The irreversible inactivation of the PTase activity

saturates at a plateau of about 25% (Figure 11).

The level of irreversible inactivation of the PTase

activity of mitochondrial ribosomes is lower than that

observed when IAP was used to irreversibly inhibit

bacterial ribosomes (25% and 75% respectively) (59, 61).

This seemingly low level of irreversible inhibition can be

interpreted to mean that the amino acid nucleophile being

alkylated by IAP may not be precisely oriented within the

CAP binding site to allow the covalently attached IAP

molecule to completely obscure the site and thereby prevent

the PTase activity. The CAP binding site can be thought of

as being occupied 25% of the time or as being 25% blocked

when IAP has been covalently attached to the CAP binding

site.

Iodoamphenicol can be thought of as being an

N-substituted IAM (Figure 7). IAP and IAM would react with

the same proteins, but IAM lacks the necessary active-

site-directing p-nitrophenyl-1,3-propanediol moieties.


























20






0 10

S 0.1 0.5 1.0
IAP concentration (mM)








Figure 11. Peptidyl transferase activity of mitochondrial
ribosomes after prolonged incubations with iodoamphenicol
(IAP). Mitochondrial ribosomes were incubated with IAP at
the indicated concentrations for one hour at 340C,
dialyzed extensively overnight, centrifuged 100,000 xg
through a 20% sucrose wedge, and then assayed for their
peptidyl transferase activity. A saturation of the
inactivation is observed for concentrations above 0.1 mM
IAP at a level of about 25%. (Data from Mary Conde)








Under the conditions where IAP could irreversibly

inhibit the mitochondrial ribosomal PTase activity, IAM was

found not to inactivate the ribosomes and CAP was binding in

a fully reversible manner (Mary Conde, personal

communication) (Table III).

The irreversible inactivation observed for

mitochondrial ribosomes after incubation with IAP is

probably not simply due to the alkylation of an essential

amino acid in the PTase Center (since IAM does not cause

inactivation), but rather that the covalently attached IAP,

because of it's extra bulk, is able to prevent the binding

of the substrates required for the PTase activity, in the

region of the CAP binding site.

Through the reversible binding of IAP to the CAF

binding site of mitochondrial ribosomes and the irreversible

inactivation of the PTase activity, a functional interaction

with the ribosomes had been established. The next step was

to use radiolabelled IAP as an affinity probe for proteins

of the CAP binding site of mitochondrial ribosomes.

1-D PAGE analyses of [ H]iodoamphenicol labelling
of ribosomal large subunit proteins

[3H]Iodoamphenicol was synthesized from chloram-

phenicol base and [3H]iodoacetic acid as described in

the Materials and Methods. The [ H]IAP was judged to

be greater than 99% pure from TLC analyses (see Materials

and Methods).


















TABLE III



Peptidyl Transferase Activity of Mitochondrial
Ribosomes Treated with Iodoamphenicol,
Chloramphenicol or lodoacetamide


Agent Present During
Incubation


IAP (1 mM)

CAP (1 mM)


Percent of Control Activity
Before Dialysis After Dialysis


IAM (4 mM)


Mitochondrial ribosomes were incubated with the reagent
for one hour at 340C, dialyzed overnight, and
concentrated by centrifugation. The activities were
determined using the peptidyl transferase assay and are
reported as a percentage of the activity of ribosomes
incubated in the absence of reagent. (Data from Mary
Conde)








The affinity labelling conditions were selected from

the consideration of several criteria. The buffer was

developed to be compatible with the mitochondrial ribosomal

subunits (200 mM KC1 and 10 mM Mg Acetate) and to be free

from nucleophilic components that would compete with the

affinity labelling reaction (triethanolamine instead of Tris

and no 2-mercaptoethanol). (Buffer L: 200 mM KC1, 10 mM Mg

Acetate, 25 mM triethanolamine HC1, pH 7.5). The ribosomal

concentration was chosen to be approximately

1 A260/10 il, which is 5.5 pM for the 39S large
ribosomal subunit. This provides for a concentrated

ribosomal solution with a small volume to allow less of t

radiolabelled probe to be used for each experiment so that

several experiments could be accomplished. The

concentrations of [3H]IAP selected for the affinity

labelling experiments extends through the response range of

IAP with the ribosomes as determined from the reversible and

irreversible PTase assays, and provides for [3H]IAP to

ribosome ratios of approximately 1 to 1 through 200 to 1.

Mitochondrial ribosomal large subunits were incubated

for one hour at 340C with [3H]IAP at concentrations of

a) 0.005 mM, b) 0.01 mM, c) 0.05 mM, d) 0.1 mM, e) 0.5 mM,

and f) 1.0 mM. Subsequently, the ribosomal proteins were

extracted and separated by one-dimensional SDS PAGE (see

Materials and Methods). One protein migrating at Band 4 in

the SDS unidimensional gels (Figure 12, panel A, stained

electropherogram) labels at low concentrations of



















(a) (b)
abcde f a bcdef



E ------
S10C,1-r m;a
13-- .












F gure 12. One-dimensional PAGE analyses of the
[ H]iodoamphenicol ([ H]IAP) labelling of mito-
chondrial large subunit proteins. Mitochondrial 39S
sibunits were incubated for one hour at 340C with
[ H]IAP at (a) 0.005 mM, (b) 0.01 mM, (c) 0.05 mM, (d)
0.1 mM, (e) 0.5 mM, and (f) 1.0 mM. The ribosomal
proteins were extracted and separated by 1-D PAGE as shown
in the electropherogram (panel A). The fluorogram (four
day exposure) of the dried gel is shown in panel B and
from its inspection, a protein in Band 4 can be observed
to label at low concentrations of probe and be the most
intensely labelled protein at higher concentrations.




















S4
SID -

10,11

<0: 24

13

0 0.1 0.5 ID
.. H]IAP concentration (mM)






Figure 13. Comparison of label incorporated into
individual protein bands of [ H]iodoamphenicol
([ HIIAP) labelled mitochondrial ribosomal large
subunit proteins separated by 1-D PAGE. Individual
protein bands were cut from the dried gel (shown in
Figure 12), digested, and their label content was
determined by liquid scintillation counting. Only
Band 4 exhibits a labelligg profile which saturates at a
stoichio-netric level of [ H]IAP incorporation into a
protein. The saturation effect is observed for
concentrations of [ H]IAP above 0.1 mM, which
parallels the saturation of irreversible inactivation
observed previously (Figure 11).








[3H]IAP (Figure 12, panel B, fluorogram), where label

does not appear in the other protein bands. After

fluorography, the protein bands from the stained gel were

cut, digested, and counted by liquid scintillation counting

(see Materials and Methods). As was observed in the

fluorogram (Figure 12, panel B),the protein in Band 4 has

the greatest amount of label at the low concentrations and

the labelling pattern saturates at one mole [ H]IAP-

adduct per mole of ribosome at 0.1 mM [ H]IAP

(Figure 13). The other ribosomal proteins do not label well

at low concentrations of [3H]IAP (no visible bands by

fluorography, Figure 12, panel B), and as shown in Figure

13, the labelling patterns of some of these ribosomal

proteins approach stoichiometry in a non-saturating

manner only at a high [3H]IAP concentration (1.0 mM).

The appearance of label on a ribosomal protein at low

[ H]IAP concentrations and the saturation of the

labelling pattern on this protein at high [ H]IAP

concentrations, suggest that this ribosdmal protein was

affinity labelled. Band 4 has a molecular weight of

approximately 45,000 (see Materials and Methods) and was

tentatively assigned as ribosomal protein L2.

l-D PAGE analyses of [ C]iodoacetamide labelling of
ribosomal large subunit proteins

Lacking the CAP binding site directing groups,
14
S14C]IAM should produce a different labelling pattern on

the [3H]IAP affinity labelled ribosomal proteins. The








[3H]IAP affinity labelled proteins would be expected to

label less well with [1C]IAM.

Mitochondrial ribosomes were incubated with
14
[14 CIAM under conditions similar to those used for the

incubation of [3H]IAP with the ribosomes (see Materials

and Methods). The ribosomal proteins were extracted and

separated by unidimensional SDS PAGE (Figure 14, panel A).

Fluorography of the dried gel (see Materials and Methods),

revealed a labelling pattern that was different from the

[3H]IAP labelling of the ribosomal proteins at low

concentrations of labelling reagent (Figure 14, panel B),
14
although at high concentrations of [ 14C]IAM and

[3H]IAP the labelling patterns appeared to be more

similar (compare Figure 12, panel B to Figure 14, panel B).

The protein bands of the dried gel were cut, digested, and

counted by LSC (Figure 15) (see Materials and Methods), as

had been the bands from the IAP gel. In contrast to the

labelling of ribosomal proteins by [3H]IAP, [ C]IAM

labels all the ribosomal proteins in a strictly

concentration-dependent manner (Figure 13 compared to Figure

15). The saturation of labelling on a protein is not
14
observed when [ 14C]IAM labels the ribosomes and

stoichiometric labelling occurs only at a high concentration

of labelling reagent (Figure 15). Several of the protein

bands label similarly whether [3H]IAP or [14C]IAM is

used as the labelling reagent (compare Bands 10, 11, 8 and

13 from Figures 13 and 15). Of special interest is the fact




































Flgure 14. One-dimen sonal PAGE analyses of the
[ C]iodoacetamide ([ C]IAM) labelling of
mitochondrial ribosomal large subunit proteins.
Mitochondria 39S subunits were incubated for one hour at
340C with [ C]IAM at (a) 0.004 mM, (b) 0.01 mM,
(c) 0.04 mM, (d) 0.1 mM, (e) 0.4 mM, and (f) 1.0 mM. The
ribosomal proteins were extracted and separated by I-D
PAGE as shown in the electropherograin (panel A). The
fluorogram (four day exposure) of the dried gel is shown
in panel B, where several protein bands show the
incorporation of label at high concentrations .but none
exhibit labelling at a low concentration of [ C]IAM.


(a) (b)
abcdef abcdef


4- ONi" -|I

8- III

- 13-


I I


I
































0.5
['4C]IAM concentration (mM)


Figure 15. Comparison of label incorporated into
individual protein bands of [ C]iodoacetamide
([ C]IA1) labelled mitochondrial ribosomal large
subunit proteins separated by 1-D PAGE. Individual
protein bands were cut front the dried gel (shown in
Figure 14), digested, and their label content was
determined by liquid scintillation counting. None of
the protein bands exhibit a saturating labelling profile.
Some of the bands arelqnly approaching stoichiometric
labelling at 1.0 mM [ C]IAM. The labelling profile
observed for Band 4 appears to be only concentration-
dependent, like that observed for the other bands, in
contrast to its labelling with [ H]IAP.








that the ribosomal protein in Band 4 labels like other

ribosomal proteins in a non-saturating manner with

[14CIAM, in contrast to its [3H]IAP labelling

profile (Figures 13 and 15).

The ribosomal protein in Band 4 appears to be

specifically affinity labelled by the CAP affinity analog,

[ H]IAP, when comparing the labelling patterns of

ribosomal proteins with [ H]IAP and [ C]IAM.

Effect of chloramphenicol on the labelling of ribosomal
large subunit proteins by [ H]iodoamphenicol

Chloramphenicol was used to competitively block the

labelling reaction of [ H]IAP with the mitochondrial

ribosomes. It was expected thaL CA-.P could effectively

reduce the amount of [3H]IAP that would be available to

bind to the CAP binding site and thereby reduce the amount

of [3H]IAP-adduct formation with ribosomal proteins at

the CAP binding site.

The mitochondrial ribosomal large subunits were in-

cubated in the presence of 0.1 mM [ H]IAP and 4 mM CAP,

to act as a competitive blocker of the labelling reaction,

or the subunits were incubated only with 0.1 mM [ H]IAP

(see Materials and Methods). The ribosomal proteins were

extracted and separated by unidimensional SDS PAGE, after

which the stained protein bands were cut, digested, and

counted by LSC (see Materials and Methods). There were 33

resolved bands cut for LSC (Figure 16, A). As can be

observed from a histogram of the results (Figure 16, B),














A B

400
2- U control
S- +CAP

a 2200
10I
12-


33o
s9 5 10 15 20 25 30
222 Ba2d ruTBber

26 2
27
28
3029
32 31
33






Figure 16. Comparison of label incorporated into
individual protein bands separated by 1-D PAGE, from
m tochondrial ribosomsl large subunits labelled by
[ H]iodoamphenicol (i H]IAP) in the absence and
presence of chloramphenicol (CAP). M tochondrial 39S
subunits were incubated with 0.1 mM [ HIIAP in the
absence and presence of 4.0 mM CAP for one hour at 340C.
The ribosomal proteins were extracted and separated by
1-D PAGE, where there were 33 resolvable bands (A). The
bands of the two gels were cut, digested, and counted by
liquid scintillation counting in parallel. As can be
observed from the histogram (B), only Band 4 labels well
with [ H]IAP (filled boxes) and has the extent of
labelling reduced to the level observed for the other
proteins in the presence of CAP (slashed boxes). Absence
of CAP, filled boxes and presence of CAP, slashed boxes.









only one protein band labels well with [ H]IAP and has

the amount of label greatly reduced (by 60%) when the

competitive blocker, CAP is present (Figure 16, B). This

Band 4 protein, tentatively named L2 of the mitochondrial

ribosomal large subunit, was assigned to the region of the

CAP binding site of these ribosomes on the basis of the

labelling patterns observed with [3H]IAP and

14CIIAM, and by the CAP competition of the [3H1IAP

labelling of this protein.

Effect of chlpamphenicol on the labelling of ribosomal
proteins by [ C]iodoacetamide

To detect if the binding of CAP to the ribosomes

induces a structural change, which. alters the accessibility

of the nucleophile in Band 4 to labelling reagents,

mitochondrial ribosomal large subunits were incubated with

[14CIIAM in the presence and absence of CAP (see Mater-

ials and Methods). Following the incubation, the ribosomal

proteins were extracted, separated by unidimensional SDS

PAGE, and the gels were prepared for fluorography (see

Materials and Methods). From the densitometric scans of

the fluorograms (see Materials and Methods) (Figure 17),
14
virtually no difference was observed in the [14 CIAM

labelling pattern of ribosomal proteins in the presence or

absence of CAP. In particular, the region of Band 4 is

identical whether CAP was present to block the labelling

reaction or was not present.





















Figure 17. Comparison of
the tracings of
densitometric scans from
the fluorograms (16 day
A exposure) of 1-D PAGE
separated mitochondrial
ribosomal large subunit
pjteins labelled with
[ VC]iodoacetamide
([ C]IAM), where
"hloramphenicol (CAP) was
included or not included
in the labelling
reactions. Mitochondrial
39S subunits were
E incubated for one hour at
S34C with 1.0 mM
E [ C]IAM in the
absence (A) and presence
(B) of 4.0 mM CAP. The
ribosomal proteins were
extracted, separated by
1-D PAGE, and analyzed by
fluorography. The
fluorograms were
densitometrically scanned
(the top of the gel is on
the left and the bottom on
the right), revealing that
CAP does not exKt an
effect on the [ C]IAM
E labelling of ribosomnal
a proteins. This is in
Scntrast to the
[ H]IAP labelling of
ribosonal proteins,
suggesting that CAP can
block the specific
affinity labelling by
[ H]IAP.