. 3%:

v

 

 

 

 

n:
fink".-. ..
.. ... a
2.... an...
wait

viva”...
.5 .c

..
.
‘CXI‘
8. r. {a Hum
:11? #18... i?!
".3. Agkflu...

1.0-1..."

mm...

o. . 21.,

5.... .30?!

:4 .15).? 1. I!
k .55u! .«fifiitnnwu.

fufa .1! h: 02“ Eli .
.v 2. 5......
31.1.. v . ,

.. . 7.. 1.95 .4...?l
. v1.9: . .L .U\
.750}

‘51 1'.

I. \I.t .
(I

. b
. $.21“ 1.1L...

 

 

unmfiwfi w. 3.4%?

 

 

THEQS

 

llllllllll

 

 

 

uamv Imllllygl\ll‘llu
Michigan .;tate
University

 

 

 

This is to certify that the

. , , dissertation entitled
npurlflcatlon of a prOPhage encoded EF—Tu-specific protease

and studies of its mechanism of activation"

presented by
Stephen Imasuen ‘Nbsakhare Ekunue

has been accepted towards fulfillment
of the requirements for

Ph. D. degree in_MJ'.CJ:obiology

WM;

Major professor

 

 

Date 8A 3’1/7 8’

MSU is an Affirmative Action/Equal Opportunity Institution 0-12771

 

PLACE IN RETURN BOX to remove this checkout from your record.
TO AVOID FINE return on or before date due.
MAY BE RECALLED with earlier due date if requested.

 

DATE DUE

DATE DUE

DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

W mu

PURIFICATION OF A PROPHAGE ENCODED EF—TU-SPECIPIC
PROTEASE AND STUDIES OF ITS MECHANISM OF ACTIVATION

By

Stephen Imasuen Nosakhare Ekunwe

A DISSERTATION
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
DOCTOR OF PHILOSOPHY
Department of Microbiology

1998

ABSTRACT

PURIFICATION OF A PROPHAGE ENCODED EF-TU- SPECIFIC
PROTEASE AND STUDIES OF ITS MECHANISM OF ACTIVATION

By

Stephen Imasuen Nosakhare Ekunwe

Bacteria use a number of mechanisms to defend themselves against
bacteriophage. The death of the infected cells to halt the spread of infection
to healthy cells is one such survival mechanism. This survival mechanism is
called phage exclusion. One of the best understood phage exclusion systems
is e14 exclusion of T4. In this system, a 34 kDa Escherichia coli (E. coli)
protease (Lit) encoded by the cryptic DNA element, e14 plays a crucial role.
Upon infection of E. coli by phage T4, this protease cleaves the 43 kDa
translation elongation factor Tu (BF-Tu), disrupting the translation machinery
needed for phage growth and propagation. This proteolysis is activated by a
29 amino acid long phage polypeptide detemtinant (Gol) internal to the major
head protein.

One of the structural genes for EF-Tu, tqu, was cloned into an
expression vector in such a way that the gene was over-expressed and the
protein was fused to an affinity tag to facilitate its purification. Similarly, the
PCR engineered gene for Lit protease was cloned into a vector that has a

different affinity tag, over-expressed, and the fusion protein was affinity

purified. The 29 amino acid long phage polypeptide determinant (601) was
chemically synthesized. When these three components were mixed together,
the 43 kDa EF-Tu was cleaved to yield 37 kDa and 6 kDa fragments.
Therefore, only three components are required for the cleavage reaction: Lit
protein, EF-Tu and Gol peptide.

There are three possible mechanisms of Go] peptide activation of the
proteolysis of EF-Tu by Lit protein: (a) Gol peptide may bind to EF-Tu and
Lit protein recognizes this Gel-bound EF-Tu as a substrate for proteolysis; (b)
G01 peptide may bind to Lit protein to make Lit protein active so that it can
proteolyze EF-Tu; or ( c) 601 peptide may bind both Lit protein and EF-Tu to
bring about the proteolysis of EF-Tu. Results fiom this research work appear
to favor (a) above.

Another question is what type of protease is Lit protein. Some
evidence suggests that it may belong to the zinc metalloprotease superfamily.
First, it has the signature sequence of zinc metalloproteases, HEXXH.
Second, it is sensitive to inhibitors of metal enzymes and not to inhibitors of
other types of proteases. Plasma emmission spectroscopy also shows the
presence of some zinc in Lit protein. Finally, site-directed mutagenesis shows
that the residues H and E in the IIEXXH sequence could play a role in the

conformation and or activity of Lit protein.

To The Blessed One At Whose Feet We Lay Our Burden
and

My Parents On Whose Shoulders I Have Stood

iv

ACKNOWLEDGEMENT

First and foremost, I give Him thanks Who made this day. May His
Name be on every tongue for His mercies are tender and limitless. All praise
and glory are His forever.

I would like to thank my mentor, Dr. Loren Snyder, who provided the
financial support and the intellectual environment for my graduate develop-
ment. 1 thank him for his patience in consistently shaping and re-shaping this
diamond in the rough to the point where its luster can begin to shine forth I
am also very grateful to the members of my guidance committee: Drs. Sue
Conrad, Robert Hausinger, Julius Jackson and Arnold Revzin for their
patience, forthrightness and numerous suggestions as to the best way to
tackle the many obstacles encountered during the course of this Ph. D. work.
Their words of understanding and encouragement when the going was rough
made it easier for me to persevere. I am grateful.

I am especially grateful to my very good friend, Dr. Ronald Patterson. I
was blessed the day I met you. Without your wise counsel both in Giltner
Hall and on the squash court, I would not be in the position to write this
acknowledgement.

I wish to thank my colleagues and dear friends in the Snyder laboratory
and the Microbiology Department: Angel Lake, Veronica Prush, Cedric
Buckley, Debbie Hogan, Lakeisha Price, Katrina Linning, Todd Anderson
and Mark Johnson.

l".

Finally, I thank my parents, Gabriel E. O. and Mary E. Ekunwe, who
admonished me to pursue education to its highest level, and who taught me
never to quit; my sisters: Ohuimumwen, Remi and Esther; my brothers:
Dada, Festus, Patrick, Ogieva; my cousin Mr. Samuel omorOdiom‘ Without
whose initial financial support I would not have made it to America; my
wife, Lynette and my children: Nosakhare, Ekhuemuenogiemwen, Kannyn,
and Adesuwa. I could not have made it without your selfless support,
unconditional love and unshakable faith in me. Thank you all.

TABLE OF CONTENTS

List ofTables...
List ofFigures...... ..

Introductron

Chapter 1: Literature Review...
Overview of Bacteriophage T4
Early Period...

e14 DNA Element...
PhageExclusion... .
T4 r11 Mutant Exclusion by km rex..

T4 Exclusion by the prr Element...

T4 Exclusionbyel4 Element
.....16
18
......19

.......20
. .. 23

Zinc Metalloproteases...

Activation of Proteases... . ..

Role of Elongation Factor Tu in Translation

Structure and Function of Elongation Factor Tu
Bibliography... .. .. .. ... .... .. ... . .. ..

Chapter 2: Purification and Preliminary Characterization of... ....

e14 Encoded Lit Protein, Preliminary Results
Suggesting that Lit Protein may be a Zinc Metalloprotease
Abstract...

Introduction

Materials and Methods

Overproduction and Purification of EF-Tu
Overproduction and Purification of Lit...

Lit Activity Assays...... ..
Results...

vii

p.—

#
O‘OOOqM-b

. .....12
...12

14

.32

33
34
.. 34

34

35

37
37

TABLE OF CONTENTS (cont’d.)

Effects ofTemperature on Lit Stability

viii

.37

Effect ofInhibitors on Activity ofLit Protein..................... 38
AnalysisofMetalContentofLitProtein.......................... 39
Site-directedMutagenesis of ...... Mot1f39
Drscussron 41
References 44
Chapter3: ThePeptideencodedbyaShortRegulatory.......................68
Region of Bacteriophage T4 Binds to Translation
Elongation Factor Tu
Introduction... .. 70
Results... 73
Equimolar Amounts of Col Peptide and EF-Tu... ..... 73
are Required for Complete Cleavage of EF-Tu.
Lit Protease is not Activated by a Preincubation. .. .74
with Go] Peptide.
Direct Evidence that Gol Peptide Binds to EF-Tu... .... 76
Discussion... .. 79
Experimental Procedures.. 85
Purification of Lit Protein... .. 85
Assay of Lit Protease... . 86
Synthesis and Isolation of G01 Peptide .. 86
Binding Experiments on Affinity Columns .. 87
References 88

LIST OF TABLES

Chapter 2
1. Metal analysis of Lit protein using inductively coupled ...... 62
plasma emission spectroscopy
2. The characteristics and references of bacterial strains... ..64
plasmid constructs and phage mutants used in this article.

Chapter 3
1. The characteristics and references of bacterial strains ...... 103
and plasmid constructs used in this article.

LIST OF FIGURES

Chapter2
1. Purification ofEF-Tu 4.6
2. Cloningoflitgene.. 48
3. Cloningoflitgenecont’d... 50
4. Purification of Lit protein/activity assay” . . ...52
4-1. Effect of dilution on ability of Lit protein to cleave EF—Tu ..54
5. Effect of temperature on Lit protein stability... .56
6. Consensus sequence of zinc-binding region .58
7. Effect of typical protease inhibitors on activity of. .. ..60
Lit protein.
8. Site-directed mutagenesis of the DNA encoding the... ..66

160His-“5lGlu-Xaa-Xaa-“‘“His motif.

Chapter 3
1. The G01 peptide must be present at equimolar or higher ....... 91
concentrations with EF-Tu for efficient cleavage of EF-Tu.
1-1. Concentration of Gol peptide needed for cleavage of EF-Tu..93

2. Can the Gol peptide activate Lit protease during .. 95
a pre-incubation?

3a Binding ofGol peptide to EF-Tu .... 97

3b. The peptide binds to EF-Tu through its 601 sequence... ..99

4. Model for the cis-inhibition of translation by Lit protease. . .. .101

INTRODUCTION

The study of the biology of bacteriophages, especially T4 and A, has
led to the discovery of numerous basic principles concerning replication,
transcription, regulation and recombination. To give some examples, there is
the classic work that provided the evidence that deoxyribonucleic acid is
genetic material Hershey and Chase (195 2); recombination within genes was
demonstrated in the r11 genes of phage T4 Benzer (195 5); bacteriophages
and their host bacteria were used to show that the genetic code works as
redundant unpunctuatcd triplets Crick et al. (1961). The discovery of many
phage enzymes in the early 19603, which included such enzymes as
polynucleotide kinase, DNA ligases, polymerases, phosphatases, and
endonucleases, led to the rapid development of new technologies in
molecular genetics and molecular biology.

Bacteriophages are viruses that infect bacteria Although they are able
to persist on their own, they are incapable of replication except within a
bacterial cell. They are usually composed of DNA or RNA surrounded by a
coat of protein. Their presence is usually demonstrated by the plaques they
form on a lawn of bacteria which represent regions where bacteria have been
killed and lysed by the infecting phage.

Bacteria have mechanisms in place for protecting themselves against
bacteriophage attack. These mechanisms include phage exclusion systems
in which the host cell is killed to prevent multiplication of the phage.

l

2
Although phage exclusion mechanisms have been extensively studied over

the past fifty years, most are not well understood. This dissertation deals with
one of the best understood phage exclusion mechanisms, the e14 exclusion of
phage T4.

While searching for Escherichia coli mutants that restrict the growth
of T4 mutants that are deficient in the production of polynucleotide 5 ’-
kinase, 3’- phosphatase, Cooley et al. (1979) isolated E. coli K12 mutants
that were unable to support the late gene expression of T4. They called these
mutants lit mutants (for late inhibitor of 14 development). However, certain
rare T4 mutants are able to grow and form plaques on these E. coli K12 lit
mutants Champness and Snyder (1982). These rare T4 mutants were named
gal mut-ants (for grow on lit). The inhibition of late gene expression in wild
type T4 was found to be due to the interaction between the Lit protein from
the E. coli K12 lit mutant and either the RNA or peptide or both encoded by
the gal region internal to the major head protein gene of T4, gene 23
Bergsland et a1. (1990). The molecular basis for the inhibition of gene
expression became apparent when Yu and Snyder (1993) found that extracts
of cells in which Gol peptide had been induced in the presence of Lit protein
were defective in translation and showed that this defect was due to the
cleavage of EF-Tu. The cleavage site is between 59Glycine and 60Isoleucine in
the sequence: Arg - Gly - Ile - Thr - Ile a 5 amino acid sequence conserved
in translation elongation factors fiom prokaryotes to eukaryotes.

Delineation of the participants involved in a reaction in which crude
cell extracts were used is not possible. Therefore, to be able to determine the
components involved in the in vitro cleavage experiments of Yu, it is

necessary to use purified proteins. The goal of my project was to purify the

3

proteins in the in vitro cleavage experiments and use these to determine the
components involved in the Gel-induced Lit cleavage of EF-Tu. Experiments
were also conducted to determine how Gol peptide activates the proteolysis of
EF-Tu. Since Lit protein has the HEXXH sequence characteristic of zinc
metalloproteases, experiments were done to answer the question: Is Lit
protein a zinc metalloprotease? Experiments to determine what residues may
be required for activity of Lit protein were conducted. From these
experiments, a model is proposed for how the cleavage of EF-Tu by Lit
protein is activated.

CHAPTER 1

LITERATURE REVIEW

Overview of Bacteriophage T4

Bacteriophage T4, a representative of the T-even phages, has been
extensively studied. Entire books have been published on what is known
about T4. Therefore this overview is merely a highlight of some of the more
basic information about this phage.

Although the structure of T4 is more complex than those of most other
bacteriophages, T4 is still basically the same as other viruses. T4 has an
icosehedral head in which the large T4 DNA molecule is housed. This head
is attached to a tubular tail that has fibers at its free end. T4 contains one of
the largest known phage chromosomes. Its genome is a large, linear, duplex
DNA molecule (165,000 base pairs). This large genome consists of about
200 genes, organized into extensive functional groups, some of which are
referred to as “essential genes” and others, “non-essential genes”. As the
name suggests, essential genes are required for phage multiplication in
laboratory strains of E. coli. Therefore, a mutation in an essential gene is
usually deleterious to the phage. The large DNA molecule is packaged into
the protective icosehedral protein head by the headful mechanism whereby
enormously long DNA molecules, concatemers, resulting from multiple
rounds of DNA replication and recombination are packaged into the head
until the head is completely filled. The concatemer is then cut. The cutting
and packaging of the concatemers is mediated by the products of terminase
genes 16 and 17 Franklin and Mosig (1996). For these large DNA molecules
to be made quickly, as required by the phage, actively metabolizing bacteria

6
are required. This leads to the question of how T4 infection of E. coli is
achieved.

Since bacteriophages are not capable of independent multiplication,
they require a bacterial host for growth. E. coli is the host for T4. In an
overview of the T4 developmental program, Mathews (1994) lucidly lays out
the events, factors and known mechanisms of phage infection. The first step
in the infection process is phage attachment to the host cell, E. coli. T4
adsorbs to host bacteria very rapidly and efficiently by utilizing its tail fibers,
the distal tips of which are made up of gp37 protein. Together with gp38,
gp37 is a “host-range cassette” functional unit Snyder and Wood (1989). T4
tail fibers bind to specific receptor molecules in the bacterial outer
membrane. In E. coli B, this receptor molecule is lipopolysaccharide
(containing diglucosyl residues), whereas in E. coli K12, it is the OmpC
protein. According to Bayer (19683), the short tail fibers, made up of gplZ,
extending from the base plate “pin” the phage to the cell outer membrane.
The large DNA molecule of phage T4 is then ejected from the phage head
into the interior of the host cell. The mechanism of this ejection is not well
understood, but it is believed to be brought about by the interaction of several
proteins: gpl8, gpl9, gpS and perhaps gp25. Within the cytoplasm of host E.
coli, phage DNA has to be protected from endonuclease activity of E. coli
enzymes. T4 protects the ends of the injected DNA fiom exonucleolytic
degradation by binding gp2. T4 DNA is further protected from degradation
by host cell enzymes by using a double modification of its DNA nucleotide,
cytidine. First, cytidine is modified to hydroxymethyl-dCMP, and then it is
glucosylated to B-D-glucosyl-hydroxymethyl—dCMP. Nucleases that degrade
cytosine-containing DNA will not degrade this modified phage DNA.

7
The T4 developmental program is broadly divided into an early period
and a late period.

Early Period

Immediately after infection of the E. coli host cell by phage T4, some
E. coli genes are transcribed. The others are shut down, and phage T4 begins
to use the host RNA polymerase (RNAP) to transcribe its own early genes.
Phage T4 is able to use host RNAP in this manner because it modifies the
host RNA polymerase core enzyme by ribosylating the alpha units of RNAP
with the phage proteins, gpalt, gpmod, and an RNA polymerase-binding
phage protein called prA. Some of the early phage genes to be transcribed
are the nuclear disruption, ndd, genes. The product of these genes disrupts
the chromosome of the host cell. Along with other T4 encoded nucleases, the
product of the ndd gene causes E. coli DNA to be degraded to
mononucleotides. These mononucleotides serve as precursor molecules for de
novo synthesis of phage DNA during the replication stage which starts about
six minutes after infection. The phage proteins that participate in the
replication of phage DNA include DNA polymerase (gp43), proteins that
enhance the processivity of DNA polymerase (gp45 and gp44/62), single-
strand DNA binding protein (gp32), topoisomerase (gp39, gpSZ and gp60),
DNA ligase (gp30), RN aseH, and a host of helicases (Nossal and Alberts,
1983; Nossal, I994). Phage DNA replication usually increases in rate for
several minutes in preparation for the late stage.

Late Period

This period ushers in the transcription of T4 late genes whose
translation products are required for making the structural components of the
phage: the head, the tail and its fibers. The work of Epstein et a1. (1963) and
of Edgar and Wood (1966) showed that the late genes were not sequentially
expressed but rather simultaneously expressed. However, the interaction of
the gene products was sequential. The transcription of phage T4 late genes
differs from that of E. coli genes in several ways. First, it occurs from late
promoters that have only the -10 sequence, TATAAATA, upstream of the
transcriptional start site as opposed to the usual -35 to -10 sequence of the 07°
or sigA promoters of E. coli Kassavetis and Geiduschek (1982), Christensen
and Young (1982) and Elliott and Geiduschek (1984). Another difference is
that transcription of T4 late genes requires replicating DNA because it is
enhanced by components of the T4 DNA replication apparatus Herendeen et
a]. (1989). Finally, it is different because it requires a modified host RNAP
Herendeen et a1. (1990, 1992). So how are T4 late genes turned on?

The gp44 and gp62 proteins are clamp-loaders that load the gp45
sliding clamp on the DNA. The gp45 clamp binds to DNA polymerase and
tracks with it along the DNA where it increases its processsivity. If the gp45
encounters RNAP bound to a late promoter, it activates transcription. Thus,
the replication apparatus serves as a sort of “mobile enhancer” activating
transcription from late promoters as it moves along the DNA, and making

late transcription replication dependent.

9
Late transcription also requires that host RNAP be modified by

binding two phage proteins, gp33 and gpSS. The gp33 protein serves as a
bridge between the gp45 sliding clamp and RNAP. The gpSS is a sigma
factor, specific for the late promoters, that confers promoter specificity on the
modified RNAP by out-competing host 07°. Therefore, the modified RNAP
recognizes only T4 late promoters Williams et a1. (1987) and Malik and
Goldfarb (1988).

The products of many of the late genes of the phage are involved in
making the T4 phage particle. About 20 proteins are involved in making the
head and another 30 in making the tail and associated structures. In all, more
than 25% of the entire genome is devoted to assembly of the phage particle.
Of the 11 proteins that make up the phage head, ng3 is made in larger
amounts than any of the other late proteins with about 1000 copies of the
protein in each head. The interaction of a region of this protein with a

protease encoded by a defective prophage is the subject of this thesis.

e14 DNA Element

DNA elements are foreign DNAs integrated into the chromosome.
They can be transposons, integrated plasmids or prophages. A prophage is a
stable, inherited, noninfectious provirus form of temperate phage in which
the phage DNA has become incorporated into and replicates with the host
bacterial DNA. Some examples of prophages are: 71. prophage, E. coli prr,
and the e14 element. Some of these prophages can be defective in that they
lack some essential genes to make a phage. The e14 DNA element of E. coli
K12 is an example of a defective prophage. This element was discovered
when Greener and Hill (1980) observed that the induction of the SOS repair

10
process in E. coli cells that have been exposed to damaging ultraviolet

radiation caused a 14.4 kb fragment of covalently closed circular DNA to be
excised fi'om the E. coli chromosome. They named this DNA e14 because it
contains 14 kilobases of DNA. Hill et a1. (1989) found that this e14 DNA
element was integrated into the E. coli K12 isocitrate dehydrogenase
structural gene (icd) between purB and umuC at 25 min in the E. coli
chromosome. This integration is site-specific like the integration of it and P4
phage. All of the prokaryotic DNA elements that integrate by site-specific
recombination use enzymes that belong to the integrase family Campbell
(1993). DNA element e14 is not found in either E. coli B or C.

Many gene products are now known to be encoded by the e14 element
including a restriction modification system directed against methylated DNA,
a DNA invertase similar to the invertase responsible for phase shift in
Salmonella, and tail fiber genes related to those of phage 1.. It is the presence
of these tail fiber genes that has led to the conclusion that e14 is a defective
prophage, which has lost some of the genes required to make phage particles.
Another of the genes carried by e14 is lit which encodes the protease

involved in phage exclusion and which is the subject of this dissertation.

Phage Exclusion

Usually, extra-chromosomal elements like prophages and plasmids
confer on their host cells some advantage. It may be antibiotic resistance,
increased ability to compete for scarce nutrients or simply improved survival
rate in a hostile environment. Resistance to phage is one such advantage

often conferred on the bacterial host that harbours a DNA element.

11

Bacteria use a number of mechanisms to defend themselves against
bacteriophage. One such survival mechanism is called phage exclusion. All
phage exclusions are superficially similar. A gene product made by the
prophage or other DNA element kills the cell upon infection by another type
of phage, thereby preventing the multiplication of the infecting phage and its
spread to other cells. In all the known exclusions, the infection begins
normally, but then a catastrophe occurs, gene expression stops and the cell
dies. In most cases it is not known what causes the cessation of gene
expression and what aspect of phage infection triggers it. Many phage
exclusion systems have been studied over the past forty years, and they
include:

(a) P2 old gene exclusion of phage 1.

(b) F plasmid pif exclusion of wild type T7 and mutant T3
(c) 2. rex exclusion of T4 111 mutants and many other phages
(d) prr element exclusion of T4 rli and pnk’ mutants

(e) e14 element exclusion of T4

Most phage exclusion systems are not well understood. For example, A,
phage is excluded in E. coli carrying P2 phage because of the interaction
between the products of gam and red genes of A and the product of P2 old
gene although the reason the cell dies is not known. Similarly, wild type T7
and mutant T3 are known to be excluded by the product of the pif gene of
the conjugative plasmid, F, but again, the reason the cell dies is not known
Molineux (1991).

12

T4 111 mutant exclusion by k rex

One of the most extensively studied exclusion mechanisms is that due
to the rex genes of A prophage. Although 1. rex exclusion of T4 r11 mutants
was the first to be described Benzer (1955), the mechanism of exclusion still
remains unclear. The products of two genes, rexA and rexB genes of A.
phage are required for the exclusion of T4 rII mutants Matz et al. (1982),
Landesmann et al. (1982). After 7» lysogens are infected by T4 rII mutants,
the infection proceeds normally until about the time T4 DNA replication
starts. At this point, a severe loss of membrane potential occurs, accompanied
by a drop in cellular ATP levels Sekiguchi (1966), Snyder and McWilliams
(1989), Panna et al. (1992). The loss of cellular energy is most severe if there
are monovalent cations in the medium and if the pH is below 7 Garen (1961),
Sekiguchi (1966), Brock (1965), Ames and Ames (1965). RexB protein has
been proposed to be a membrane ion channel that allows the passage of
monovalent cations. Its activation by the already activated RexA protein
would reverse the membrane ATPase leading to a loss of cellular ATP Parma
et al. (1992). While T4 rII mutants would be excluded under such
conditions, wild type T4 would not be so affected. However, Shinedling et al.
(1 98 7) observed that the overproduction of rex gene products will cause even
wild type T4 to be excluded.

T4 Exclusion by the prr element

One of the best understood phage exclusion systems is the prr
system. The prr element is not found in most laboratory strains of E. coli. It
was found in a clinical isolate of E. coli, CT196, where it is integrated at 29

min in the genetic map of the E. coli chromosome Depew and Cozzarelli

13
(1974), Abdul-Jabbar and Snyder (1984), Levitz et al. (1990). It was
subsequently transduced into laboratory strains of E. coli Abdul-Jabbar and
Snyder (1984). Wild type T4 can multiply in E. coli carrying the prr element
because it has polynucleotide kinase and RNA ligase. T4 mutants lacking
these enzymes cannot multiply in E. coli carrying the prr element. This was
how the element was found Depew and Cozzarelli (1974), Sirotkin et al.
(1978), Runnels et al. (1982), Abdul-Jabbar and Snyder (1984). Work done
by the Kaufmann group and their collaborators led to the discovery of the
molecular basis of the prr exclusion of T4. The prr element carries a gene,
prrC, which encodes a ribonuclease specific for host lysine tRNA, tRNAL”.
This ribonuclease cleaves tRNALys in the anticodon loop after T4 infection.
Amitsur et al. (198 7) have shown that the cleavage occurs immediately 5’ of
the anticodon in the anticodon loop, thereby leaving 5 ’ hydroxyl and cyclic
2’:3’ P04 ends. This cleavage of RNA“ blocks translation and stops phage
development. T4 that has polynucleotide kinase and RNA ligase can repair
this cut and multiply in the cells possessing the prr element, while mutant T4
that lack these repair enzymes would not survive because translation has been
blocked and so phage development is halted. If such mutants possess a
second mutation that suppresses the original mutation, then the mutant T4
can multiply in cells that possess the prr element. Such mutations, termed
suppressor of three prime phosphatase, stp, lie in a gene that encodes a small
peptide that is 26 amino acids long. The Stp peptide is not essential for
normal growth of the phage. So why does the phage make the peptide?
Levitz et al. (1990) discovered that prrC gene which encodes the
ribonuclease, is associated with a restriction-modification system. It was
found that the prrC gene was sandwiched between the hst and hsaS genes

14

of the restriction-modification system. The prrC gene product is physically
associated with the products of the hde, hst and hsaS genes. Amitsur et
al. (1992) showed that in the presence of the restriction proteins, the activity
of the ribonuclease appears to be masked. If exogenous Stp was added to this
mix of restriction proteins and ribonuclease, the activity of the ribonuclease
was restored. Synthesis of the ribonuclease from a clone containing only the
prrC gene yielded an active but very unstable ribonuclease Morad et al.
(1993). The Stp peptide also inactivates this type of restriction in vivo. Based
upon these findings, Penner et al. (1995) proposed that the Hsd proteins
mask the activity of the PrrC ribonuclease by binding to it, and in so doing,
increases the stability of the protein so that any free PrrC ribonuclease is
rapidly inactivated and is not available to cleave any RNA”. In Type I
restriction-modification systems, both restriction and modification activities
are contained in the same multisubunit complex. The Stp peptide may
inactivate this complex by causing dissociation. Therefore when Stp peptide
binds the Hsd proteins complex, it may dissociate the complex thereby
removing the masking of PrrC by the restriction-modification proteins. PrrC
is activated and can then cleave tRNAL” thereby blocking translation and
halting phage propagation.

Work in Snyder’s laboratory has also focused on an exclusion system
similar to the prr element exclusion system. It is the T4 exclusion by e14
DNA element of E. coli K12 which is the subject of this thesis.

T4 Exclusion by e14 element
As mentioned, e14 is integrated at 25 min in the genetic map of E. coli
K12 chromosome Greener and Hill (1980). E. coli K12 mutants that were

15
deficient in their ability to support the late gene expression of T4 were

isolated. The mutations were named lit (for late inhibition of T4) mutants
Cooley et al. (1979). Kao et a1. (1987) cloned and characterized the E. coli
K12 lit gene and showed that the inability to plate T4 was due to the
constitutive expression of this gene, named the Lit(Con) phenotype. Small
fragments of the DNA element, e14, were found to confer the Lit(Con)
phenotype on cells. These clones were sequenced and found to contain an
open reading frame (ORF) that encoded a 34 kDa protein, the Lit protein.
The inhibition of late gene expression in T4 was found to be due to the
interaction between the Lit protein (the product of the lit gene of the e14
element) and the Gal peptide, a small portion of the major head protein of T4
gene 23 Bergsland et al. (1990). The gal region was found through isolation
of rare T4 mutations that overcame this defect in late gene expression. These
rare mutations allowed T4 to form plaques in the presence of Lit protein,
hence they were named gal (for grow on lit), Champness and Snyder (1982).

To determine the reason for the block in late gene expression,
Bergsland et al. (1990) measured the rate of DNA, RNA and protein
synthesis after induction of synthesis of the Gal peptide in the presence of Lit
protein. They found that protein synthesis was specifically inhibited. DNA
replication and RNA synthesis continued unabated. The question remained:
what is happening to the translation apparatus when the G01 peptide is
synthesized in the presence of Lit protein?

Yu and Snyder (1993) answered this question when they prepared
extracts of cells in which translation had been inhibited and discovered that
elongation factor Tu (BF-Tu) was cleaved. They could also show that
cleavage of EF-Tu was probably solely responsible for the inhibition of

16
translation. The cleavage of EF-Tu occurs between Gly’9 and Ile60 giving rise

to 37 kDa and 6 kDa fragments.

Zinc Metalloproteases

Four major classes of proteases are known and are designated by the

principal functional group in their active site. They are:

(a) Serine proteases: which have a serine residue in their active sites.

(b) Thiol proteases: which have a cysteine thiol group in their active sites.

(0) Carboxyl (acid) proteases: which have acidic residues in their active
sites.

(d) Metalloproteases: which have a Group II metal in their active sites.

Since some evidence suggests that Lit protein might be a member of
the zinc metalloprotease superfamily, I shall review the properties of such
proteases in more detail. Metalloproteases that have Zn+2 in their active sites
are called zinc metalloproteases. Although metalloproteases that have Cu+2
and Co+2 in their active sites exist, those that have Zn‘i2 in their active sites
occur more frequently in nature. Zinc has been found to be an essential and
integral component in over 300 enzymes. Zinc plays a role in both enzyme
catalysis and structure. It plays a structural role when it stabilizes the
structure of proteins and nucleic acids.

In zinc metalloproteases, zinc plays a catalytic role in proteolysis. One
of the identifying characteristics of zinc metalloproteases is that they possess
the HEXXH signature sequence. The number of zinc metalloproteases/pepti-
dases identified in recent years has increased greatly Hooper (1994),
Jongeneel et al. (1989). Using a scheme based on the comparison of the

sequences around the HEXXH motif of the zinc-binding site, this super-

17
family has been classified into five distinct sub-families: (1) Thermolysin (2)

Astacin (3) Serratia (4) Matrixin and (5) Reprolysin (snake venom)
metalloproteinases. Two histidines and a glutamate are the zinc ligands in the
thermolysin sub-family while three histidines and one tyrosine are the zinc
ligands in the other four sub-families Jiang and Bond (1992). The non-
thermolysin sub-families differ from one another in unique ways. For
example, members of the astacin sub-family possess a glutamate following
the third histidine which is used to form a salt bridge with the N-terminus of
the mature enzyme. There is also a glycine residue that is important for
secondary structure Bode et al. (1992). The serratia sub-family possesses a
proline instead of the second glutamate of the astacin sub-family, while the
reprolysin sub-family has an aspartic acid residue at position 12 away fi'om
the first histidine of the HEXXH motif. This sequence has been found in all
clostridial neurotoxins whose sequences are available Schiavo et al. (1992),
elastase of Pseudomonas aeroginosa Kawarnoto et al. (1993), lethal factor of
anthrax toxin of Bacillus anthracis Klimpel et al. (1994), insulin degrading
enzyme in which the sequence is HXXEH, a mirror image of the standard
HEXXH motif Perlrnan and Rosner (1994) and deformylase, encoded by E.
coli fins gene Meinnel et al. (1995). In all these enzymes, this motif is
required for the binding of the metal ion that is needed for the activity of the
enzyme. The role zinc plays in catalysis has been studied in many zinc
metalloproteases. Vallee and Auld (1990) reviewed the results of such
studies. They suggest the following as a possible general mechanism for the
catalytic activity of metalloproteases. In all zinc metallopeptidases for which
crystal structures are known, the catalytic zinc atom is coordinated to three

amino acid residues of the protein and an “activated” water molecule. A

18

combination of any three residues of His, Glu, Asp, or Cys creates a
tridentate active zinc site although histidine is the most frequent ligand. An
“activated” water molecule fills and completes the coordination sphere of the
active zinc site. From the work of Vallee et al. (1983) and Auld and Vallee
(1987) on carboxypeptidase A, Quiocho and Lipscomb (1971) and Schmid
and Heniott (1976) on the bovine carboxypeptidase A and B enzymes, and
T an et al. (1980) on carboxypeptidase M, it is now believed that 2 His and a

Glu are ligands to the zinc atom at the active site.

Activation of Proteases

Some proteases can be activated from inactive precursor proteins by
processes that are not related to metal incorporation. Vallee and Auld (1990)
proposed a model for how zinc metalloproteases could be activated fi'om
these longer precursors called the “Velcro model”. In this model, the inactive
precursor peptide is proteolyzed to remove a peptide from one end, resulting
in the conversion of the zinc coordinated to four amino acid residues
(tetradentate) to zinc that is coordinated to only three residues (tridentate)
through the removal of a cysteine ligand, which is then replaced by water.
Other models for the activation of proteases involve an exchange reaction or
a conformational change. An example of an exchange reaction activating a
protease is demonstrated by the adenovirus protease, Ad2, a thiol protease
that is required for virus maturation. For this protease to be active, it requires
an 11 amino acid long peptide determinant, GVQSLKRRRCF, derived from
the C-terminus of the structural protein, pVI. Using its cysteine residue, this
peptide participates in a disulphide exchange reaction with Ad2 to activate
Ad2 by exposing the Ad2 active site cysteine Webster et al. (1993). The

l9

cleavage of the largest subunit, p220, of eukaryotic translation initiation
factor 4F (eIF-4F) in the presence of a functional poliovirus protease 2Apm is
another example of a thiol protease being involved in proteolysis. This
proteolytic event requires the presence of eukaryotic translation initiation
factor 3 (eIF-3). Wyckoff et al. (1990) proposed that eIF-3 - eIF-4F complex
presents p220 in a proper conformation to be a substrate for cleavage by
protease 2A9”. An example of conformational change as a regulator of
protease activity is dimerization. Human immunodeficiency virus (HIV)
protease, an aspartyl protease, is activated by dimerization Navia et al.
(1989) and Krausslich (1991). The active HIV protease is composed of two
COpies of the HIV protease monomer linked to each other. The active
protease proteolyses HIV polyprotein substrates. If dimerization occurs
before the mature virus particle is formed, the processing of viral
polyproteins will prevent particle formation and infectivity.

Role of Elongation Factor Tu in Translation

Because Lit protein cleaves EF-Tu I shall briefly review what is known
about EF-Tu. The information contained in messenger RNA, mRN A, is
converted or “translated” into protein. EF-Tu is one of the components of the
protein translation machinery. Although translation may be divided into three
steps: initiation, elongation and termination, I shall focus only on the
elongation step. This step is brought about by the elongation factors,
primarily, EF-Tu, EF-Ts and EF-G. There are two forms of EF-Tu: the
active GTP bound form and the inactive GDP bound form. In its active GTP
bound state, EF-Tu recognizes, transports and positions the codon-specified
arninoacyl-tRN A onto the acceptor (A) site of the 708 ribosome that is

20
tracking along the mRNA Miller and Weissbach (1977). After the tRNA has

bound and the GTP bound to EF-Tu is cleaved to GDP, the EF-Tu is
released. EF-Tu is recycled when EF-Ts dissociates the EF-Tu-GDP complex
and GTP binds again to EF-Tu. EF-G translocates the tRNA to the peptidyl
(P) site to make room for another charged tRN A at the A site. Two tRNAs
are, at this stage, attached to the ribosome, with the one at the P site having a
growing polypeptide chain attached to it. The enzyme peptidyl transferase
catalyzes the formation of a peptide bond between the amino acid attached to
the A site tRNA and that at the P site. The P site tRNA then releases the
polypeptide chain which is now attached to the A site. EF-G then translocates
the tRNA and the polypeptide chain which is now one amino acid longer to
the P site so that the A site is once again empty. The ribosome moves along
the mRNA in the 3’ direction to the next three nucleotides, 3 charged tRNA
is once again brought to the A site by GTP bound EF-Tu and the process is
repeated. In eukaryotes, the counterparts of EF-Tu and EF-Ts are eEF -1(1 and

eEF-l B, respectively and the analog of EEG is EF-2.

Structure and Function of Elongation Factor Tu

Elongation factor Tu (BF-Tu) has been the subject of extensive
investigation since it was first discovered Lucas-Lenard and Lipman (1966).
It is one of the most abundant proteins in the bacterial cell and accounts for
5-10% of the protein mass Van der Meide et al. (1980) and Jacobson and
Rosenbusch (1976). E. coli EF -Tu is encoded by two almost identical
structural genes tqu and tufB Jaskunas et al. (1975). The amino acid
sequences encoded by the two genes differ only by one amino acid: Tqu has
a glycine at its carboxyl terminus while TufB has a serine Arai et al. (1980)

21

and Jones et al. (1980). The tqu gene is the distal gene in the str operon
which also contains the genes for ribosomal proteins S7 and 812 and EF-G
Zengel and Lindahl (1990). The tufB gene is in an operon for tRNAs. It is not
known why there are two genes for EF-Tu, but the two genes may be
regulated differently. The tqu gene is expressed 50% more than the tufB
gene under normal growth conditions. The tujB gene is regulated by ppGpp.

EF-Tu is a 43 kDa monomeric-protein-flrat'contains '393“amino'acid
residues arranged in three domains: I, II, and III. Domain I is the nucleotide
binding domain of EF-Tu. The crystal structure of EF-Tu revealed that
domain I of EF-Tu is similar to that of other nucleotide binding proteins,
especially those that have GTPase activity such the ras oncogene proteins,
p21 Jurnak (1985), Wittinghofer (1993), Weijland et al. (1992) and Sprinzl
(1994). It is made up of five B-sheets and six a-helices. Within this domain,

there is a region that connects the first a-helix to the second B-sheet. This
region was named the “effector region”, L2, since its structure changes
between GDP-bound and GTP-bound forms of EF-Tu. Domains H and III are
made up of anti-parallel B-sheets exclusively, forming two B-barrels. In EF-
Tu-GTP, the active form of EF-Tu, all three domains are tightly packed,
whereas in the inactive form, EF-Tu-GDP, domain H is separated from
domain I Sprinzl (1994), apparently to release the bound tRNA.

The cleavage of EF-Tu by the e14-encoded Lit protein occurs between
Gly’9 and Ile60 in the highly conserved sequence Arg-Gly-Ile-Thr-Ile found in
elongation factors of both prokaryotes and eukaryotes. This sequence is in the
carboxyl terminal part of the L2 region within the nucleotide binding domain.
In EF-Tu, the arginine residue of this sequence is a trypsin-hypersensitive site
while the threonine residue coordinates the Mg2+ ion and the y-phosphate of

22
the EF-Tu bound GTP Berchtold et al. (1993). In all regulatory GTPases,

this region is involved in binding the GTPase activating protein. In the case
of EF-Tu, the ribosome fimctions as the GTPase activating protein Georgiou
et al. (1998) and Zeidler et al. (1996). Ziedler et al. (1996), in studies on the
effector region of EF-Tu of T hermus thermophilus, showed that the structural
integrity of the L2 region of EF-Tu around the arginine in the sequence Arg-
Gly-lle-Thr-Ile, is important for the control of the GTPase activity by
ribosomes. This structural integrity is not maintained when EF-Tu is cleaved
by the Lit protein. Thus the cleavage may abolish the ability of ribosomes to
stimulate GTPase activity.

BIBLIOGRAPHY

23

BIBLIOGRAPHY

Abdul Jabbar, M., and L. Snyder. 1984. Genetic and physiological studies of
an Escherichia coli locus that restricts polynucleotide kinase and
RNA ligase-deficient mutants of bacteriophage T4. .1. Virol. 51:
522-529.

Ames, C. F., and B. N. Ames. 1965. The multiplication of T4 rII phage in

E. coli K420.) in the presence of polyamines. Biochem. Biophys.
Res. Commun. 18: 639-647.

Amitsur, M., I. Morad, D. Chapman-Shirnshoni and G. Kaufmann. 1992.
Hsd restriction-modification proteins partake in latent anticodon
nuclease. EMBO J. l 1: 3129-3134.

Amitsur, M., R. Levitz and G. Kaufinann. 1987. Anticodon nuclease, poly-
nucleotide kinase and RNA ligase reprocess the host lysine tRN A.
EMBO J. 6: 2499-2503.

Arai, K., B. F. C. Clark, L. Duffy, M. D. Jones, Y. Kaziro, R. A. Laursen,
J. L’Italien, D. L. Miller, S. Nagarkatti, S. Kanamura, K. M. Nielson,
T. E. Petersen, K. Takahashi and M. Wade. 1980. Primary structure

of elongation factor Tu from Escherichia coli. Proc. Natl. Acad. Sci.
USA 77: 1326-1330.

Auld, D. S., and B. L. Vallee. 1987 in Hydrolytic enzymes (Neuberger, A.,
and Brocklehurst, K., Eds.) pp. 201-255, Elsevier, New York.

Bayer, M. E. 1968a Adsorption of bacteriophage to adhesions between
cell wall and membrane of Escherichia coli. J. Viral. 2: 346-356.

Benzer, S. 1955. Fine structure of a genetic region in bacteriophage. Proc.
Natl. Acad. Sci. USA 41: 344-354.

24

Bergsland, K. J ., C. Kao, Y.-T. N. Yu, R. Gulati and L. Snyder. 1990. A site
in the T4 bacteriophage major head protein gene that can promote the
inhibition of all translation in Escherichia coli. J. Mol. Biol. 213: 477
-496.

Berchtold, H., L. Reshetnikova, C. O. A. Resier, N. K. Schirmer, M. Sprinzl
and R. Hilgenfeld. 1993. Crystal structure of active elongation factor
Tu reveals major domain rearrangements. Nature 365: 126-132.

Bode, W., F. X. Gomis-Ruth, R. Huber, R. Zwilling and W. Stacker. 1992.
Structure of astacin and implications for activation of astacins and
zinc-ligation of collagenases. Nature 358: 164-167.

Brock, M. L. 1965. The effects of polyarnines on the replication of T4 r11
mutants in Escherichia coli K420»). Virol. 26: 221-227.

Campbell, A. M. 1993. Thirty years ago in genetics: prophage insertion into
bacterial chromosomes. Genetics 133: 433-438.

Champness, W. C., and L. Snyder. 1982. The gal site: a cis-acting bacterio-
phage T4 regulatory region that can affect expression of all the T4
late genes. J. Mol. Biol. 155: 395-407.

Christensen, A. C., and E. T. Young. 1982. T4 late transcripts are initiated
near a conserved DNA sequence. Nature 299: 369-371.

Cooley, W., K. Sirotkin, R. Green and L. Snyder. 1979. A new gene of
Escherichia coli K-12 whose product participates in T4 bacterio-
phage late gene expression: interaction of lit with the T4-induced
polynucleotide 5’-kinase 3’-phosphatase. J. Bacterial. 140: 83-91.

Crick, F. H. C., L. Barnett, S. Brenner and R. J. Watts-Tobin. 1961. General
nature of the genetic code for protein. Nature 192: 1227-1232.

Depew, R. E., and N. R. Cozzarelli. 1974. Genetics and physiology of bacte-
riophage T4 3 ’-phosphatase: evidence for involvement of the enzyme
in T4 DNA metabolism. J. Virol. 13: 888-897.

25

Edgar, R. S., and W. B. Wood. 1966. Morphogenesis of bacteriophage T4 in
extracts of mutant-infected cells. Proc. Natl. Acad. Sci. USA 55: 498
-505.

Elliott, T., and E. P. Geiduschek. 1984. Defining a bacteriophage T4 late pro-
moter: absence of a “-35” region. Cell 36: 211-219.

Epstein, R. H., A. Bolle, C. M. Steinberg, E. Kellenberger, E. Boy De La
Tour, R. Chevalley, R. S. Edgar, M. Susman, G. H. Denhardt and A.
Lielausis. 1964. Physiological studies of conditional lethal mutants of
bacteriophage T4D. Cold Spr. Harb. Symp. Quant. Biol. 28: 375-
392.

Franklin, J. L., and G. Mosig. 1996. Expression of the bacteriophage T4
DNA terminase genes 16 and 17 yields multiple proteins. Gene 17 7:
179-189.

Garen, A. 1961. Physiological effects of r11 mutations in bacteriophage T4.
Virol. 14: 151-163.

Georgian, T., Y.-T. N. Yu, S. Ekunwe, M. J. Buttner, A.-M. Zuurrnond, B.
Kraal, C. Kleanthous and L. Snyder. 1998. Specific peptide-activated

proteolytic cleavage of Escherichia coli elongation factor Tu. Proc.
Natl. Acad. Sci. USA 95: 2891-2895.

Greener, A., and C. W. Hill. 1980. Identification of a novel genetic element
in Escherichia coli K-12. J. Bact. 171: 4083-4084.

Herendeen, D. R., G. A. Kassavetis, J. Barry, B. M. Alberts and E. P. Geidu-
schek. 1989. Enhancement of bacteriophage T4 late transcription by
components of the T4 DNA replication apparatus. Science 245: 952-
958.

Herendeen, D. R., K. P. Williams, G. A. Kassavetis and E. P. Geiduschek.
1990. An RNA polymerase-binding protein that is required for

communication between an enhancer and a promoter. Science 248:
57 3-5 7 8.

26

Herendeen, D. R., G. A. Kassavetis and E. P. Geiduschek. 1992. A Trans-
criptional enhancer whose fimction imposes a requirement that pro-
teins track along DNA. Science 256: 1298-1303.

Hershey, A. D., and M. Chase. 1952. Independent functions of viral proteins

and nucleic acid in growth of bacteriophage. J. Gen. Physiol. 36:
39-56.

Hill, C. W., J. A. Gray and H. Brody. 1989. Use of the isocitrate dehydrogen-
ase structural gene for attachment of e14 in Escherichia coli K-l2. J.
Bacteriol. 171: 4083-4084.

Hooper, N. M. 1994. Families of zinc metalloproteases. FEBS Lett 354: l-6.

Jacobson, G. R., and J. P. Rosenbusch. 1976. Abundance and membrane
association of elongation factor Tu in E. coli. Nature 261: 23-26.

Jaskunas, S. R., L. Lindahl, M. Nomura and R. R. Burgess. 1975. Identifica-
tion of two copies of the gene for the elongation factor EF-Tu in E.
coli. Nature 257: 458-462.

Jiang, W. and Bond, J. S. 1992. Families of metalloendopeptidases and their
relationships. FEBS Lett 312: 110-114.

Jones, M. D., T. E. Petersen, K. M. Nielsen, S. Magnusson, L. Sottrup-
Jensen, K. Gausing and B. F. C. Clark. 1980. The complete amino
acid sequence of elongation factor Tu from Escherichia coli. Eur.
J. Biochem. 108: 507-526.

Jongeneel, C. V., J. Bouvier and A. Bairoch. 1989. A unique signature ident-
ifies a family of zinc-dependent metallopeptidases. F EBS Lett 242:
21-214.

Jumak, F. 1985. Structure of the GDP domain of EF-Tu and location of the
amino acids homologous to ras oncogene proteins. Science 230: 32-
36.

Kao, C., E. Gumbs and L. Snyder. 1987. Cloning and characterization of the

27

Escherichia coli lit gene which blocks bacteriophage T4 late gene
expression. J. Bacteriol. 169: 1232-1238.

Kassavetis, G. A., and E. P. Geiduschek. 1982. Bacteriophage T4 late pro-
moters: mapping 5’ ends of T4 gene 23 mRNAs. EMBO J. 1: 107-
1 l4.

Kawarnoto, S., Y. Shibano, J. Fukushima, N. Ishi, K. Morihara and K.
Okuda 1993. Site-directed mutagenesis of Glu-141 and His-223 in
Pseudomonas aeroginosa elastase: catalytic activity, processing and

protective activity of the elastase against Pse udamanas infection.
Infect. Immun. 61: 1400-1405.

Klimpel, K. R., N. Arora and S. H. Leppla 1994. Anthrax toxin lethal factor
contains a zinc metalloprotease consensus sequence which is required
for lethal toxin activity. Mol. Microbiol. 13(6): 1093-1100.

Krausslich, H.-G. 1991. Human immunodeficiency virus proteinase dimer as

a component of the viral polyprotein prevents particle assembly and
viral infectivity. Proc. Natl. Acad. Sci. USA 88: 3213-3217.

Landesmann, J ., M. Kroger and G. Hobom. 1982. The rex region of bacterio-
phage lambda: two genes under three-way control. Gene 20: 11-24.

Levitz, R., D. Chapman, M. Amitsur, R. Green, L. Snyder and G. Kaufrnann.
1990. The optional E. coli prr locus encodes a latent form of phage
T4- induced anticodon nuclease. EMBO J. 9: 1383-1389.

Lucas-Lenard, J ., and F. Lipmann. 1966. Separation of three microbial amino
acid polymerization factors. Proc. Natl. Acad. Sci. USA 5521562-
1566.

Malik, S., and A. Goldfarb. 1984. The effect of a bacteriophage T4-induced
polypeptide on host RNA polymerase interaction with promoters. J.
BioI.Chem. 259: 13292-13297.

Matz, K., M. Schmandt and G. N. Gussin. 1982. The rex gene of bacterio-
phage A is really two genes. Genetics 102: 319-327.

28

Mathews, C. K. 1994. Overview of the T4 development program pp. 1-8.
Bacteriophage T4 , American Society for Microbiology, Washington,
DC.

Meinnel, T., C. Lazennec and S. Blanquet. 1995. Mapping of the active site
zinc ligands of peptide deforrnylase. J. Mol. Biol. 254: 175-183.

Miller, D. L., and H. Weissbach. 1977. In Molecular mechanisms of protein
biosynthesis, S. Petska and H. Weissbach, (Eds) p. 323, Academic
Press, New york.

Molineux, I. J. 1991. Host-parasite interactions: recent developments in the
genetics of abortive phage infections. New Biol. 3: 230-236.

Morad, I., D. Chapman-Shimshoni, M. Amitsur and G. Kaufinann. 1993.
Functional expression and properties of tRNALys-specific core
anticodon nuclease encoded by Escherichia coli prrC. J. Biol. Chem.
268: 26842-26849.

Navia, M. A., P. M. D. Fitzgerald, B. M. McKeever, C.-T. Leu, J. C. Heirn-
bach, W. K. Herber, I. S. Sigal, P. L. Darke and J. P. Springer. 1989.

Three-dimensional structure of aspartyl protease from human immun
-odeficiency virus HIV-1. Nature 337: 615-620.

Nossal, N. G. 1994. The bacteriophage T4 DNA replication fork, pp. 45-53.
J .D. Karam, J .W. Drake, K.N. Kreuzer, G. Mosig, D.H Hall, F.A.
Eiserling, L.W. Black, E.K. Spicer, E. Kutter, K. Carlson and ES.
Miller (Eds), Molecular Biology of Bacteriophage T4 . American
Society for Microbiology, Washington, DC.

Nossal, N. G., and BM Alberts. 1983. Mechanism of DNA replication cata-
lyzed by pmified T4 replication proteins, pp.71-8 1. CK. Mathews,
E.M.Kutter, G. Mosig, and RB. Berget (Eds), Bacteriophage T 4.
American Society for Microbiology, Washington, DC.

Parma, D. H., M. Snyder, S. Sobolevski, E. Brody and L. Gold. 1992. The
Rex system of bacteriophage it: tolerance and altruistic cell death.

29
Genes Dev. 6(3): 497-510.

Penner, M, I. Morad, L. Snyder and G. Kaufrnann. 1995. Phage T4-encoded
Stp: double-edged effector of coupled DNA and tRNA-restriction
systems. J. Mol. Biol. 249: 857-868.

Perlman, R. K., and M. R. Rosner. 1994. Identification of zinc ligands of the
insulin-degrading enzyme. J. Biol. Chem. 269: 33140-33145.

Quiocho, F. A., and W. N. Lipscomb. 1971. Carboxypeptidase A: A protein
and an enzyme. Adv. Prat. Chem. 25: 1-58.

Runnels, J. M., D. Soltis, T. Hey and L. Snyder. 1982. Genetic and physio-
logical studies of the role of RNA ligase of bacteriophage T4. J. Mol.
Biol. 154: 273-286.

Schiavo, G., O. Rossetto, A. Santucci, B. R. DasGupta and C. Montecucco.

1992. Botulinum neurotoxins are zinc proteins. J. Biol. Chem. 267:
23479-23483.

Schmid, M. F., and J. R. Herriott. 1976. Structure of carboxypeptidase B at
2.8 Angstrom resolution. J. Mol. Biol. 103: 175-190.

Sekiguchi, M. 1966. Studies on the physiological defect in r11 mutants of
bacteriophage T4. J. Mol. Biol. 16: 503-522.

Shinedling, S., D. Panna and L. Gold. 1987. Wildtype bacteriophage T4 is
restricted by A. rex genes. J. Virol. 61: 3790-3794.

Sirotkin, K., W. Cooley, J. Runnels and L. Snyder. 1978. A role in true late
gene expression for T4 bacteriophage 5 ’ polynucleotide kinase 3 ’phos
-phatase. J. Mol. Biol. 123: 221-233.

Snyder, L., and K. McWilliams. 1989. The rex genes of bacteriophage 7. can
inhibit cell function without phage superinfection. Gene 81: 17-24.

Snyder, L., and W. Champness. 1997. Transposition and Nonhomologous
recombination, p.195. Molecular Genetics of Bacteria. American
Society for Microbiology, Washington, DC.

3O

Snyder, M., and W. B. Wood. 1989. Genetic definition of two functional
elements in bacteriophage T4 host-range “cassette”. Genetics 122:
47 1-479.

Sprinzl, M.. 1994. Elongation factor Tu: a regulatory GTPase with an
integrated effector. T183. 19: 245-250.

Tan, F., S. J. Chan, D. F. Steiner, J. W. Schilling and R. A. Skidgel. 1989.
Molecular cloning and sequencing of the cDNA for human mem-
brane-bound carboxypeptidase M. J. Biol. Chem. 264: 13165-13170.

Vallee, B. L., and D. S. Auld. 1990. Zinc coordination, function, and struc-
ture of zinc enzymes and other proteins. Biochemistry 29: 5647 -
5659.

Vallee, B. L., A. Galdes, D. S. Auld and J. F. Riordan. 1983. In Zinc
enzymes, pp.26-75. T.G. Spiro (Ed.) Wiley, New York.

Van der Meide, P. H., T. H. Barman, A. M. A. Van Kimmenade, P. Van De
Putte and L. Bosch. 1980. Elongation factor Tu isolated from Escheri
-chia coli mutants altered in tqu and tufB. Proc. Natl. Acad. Sci.
USA 77: 3922-3926.

Webster, A., R. T. Hay and G. Kemp. 1993. The Adenovirus protease is
activated by a virus-encoded disulphide-linked peptide. Cell 72: 97-
104.

Weijland, A., K. Harmark, R. H. Cool, P. H. Anborgh and A. Parmeggiani.
1992. Elongation factor Tu: a molecular switch in protein biosynthe-
sis. Mol. Microbial. 6(6): 683-688.

Williams, K. P., G. A. Kassaveti and E. P. Geiduschek. 1987. Interactions of
the bacteriophage T4 gene 55 product with E. coli RNA polymerase:
competition with E. coli 070 and release from late T4 transcription
complexes following initiation. J. Biol. Chem. 262: 12365-12371.

Wittinghofer, A.. 1993. From EF-Tu to p21” and back again. Curr. Biol.

31
3(12): 874-876.

Wyckoff, E. E., J. W. B. Hershey and E. Ehrenfeld. 1990. Eukaryotic initia-
tion factor 3 is required for polio virus 2A protease-induced cleavage
of the p220 component of initiation factor 4F. Proc. Natl. Acad. Sci.
USA 87: 9529-9533.

Yu, Y.-T. N., and L. Snyder. 1994. Translation elongation factor Tu cleaved

by a phage-exclusion system. Proc. Natl. Acad. Sci. USA 91: 802-
806.

Zeidler, W., N. K. Schirmer, C. Egle, S. Ribeiro, R. Kreutzer and M. Sprinzl.
1996. Limited proteolysis and amino acid replacements in the effector

region of T hermus thermophilus elongation factor Tu. Eur. J.
Biochem. 239: 265-271.

Zengel, J. M., and L. Lindahl. 1990. Mapping of two promoters for elonga-
tion factor Tu within the structural gene for elongation factor G.
Biochim. Biophys. Acta. 1050: 317-322.

CHAPTER 2

PURIFICATION AND PRELIMINARY CHARACTERIZATION OF e14

ENCODED LIT PROTEIN: PRELIMINARY RESULTS SUGGESTING
THAT LIT PROTEIN MAY BE A ZINC METALLOPROTEASE.

32

33

ABSTRACT Some strains of Escherichia coli harbour genes that trigger
cell death upon infection by bacteriophage. The death of the infected cells
halts the spread of infection to healthy cells ensuring their survival. This
survival mechanism is called phage exclusion. One of the best understood
phage exclusion systems is the e14 exclusion of T4 caused by the interaction
of the e14 encoded protein, Lit, and a short polypeptide sequence encoded by
gal from within the major head protein gene of T4. This interaction caused
severe inhibition of all translation. Using purified proteins, we have shown
here that translation inhibition is due to the activation of the proteolysis of the
43 kDa translation elongation factor Tu (EF-Tu) by the 34 kDa E. coli
protease (Lit) encoded by the cryptic DNA element e14. The proteolysis is
activated by the 29 amino acid long phage polypeptide determinant (G01)
internal to the major head protein of bacteriophage T4. This Lit-specific Gol
peptide-activated proteolysis of EF-Tu leads to the disruption of the protein
biosynthetic machinery needed for phage growth and propagation. A
comparison of the sequence of Lit protein to known sequences shows that Lit
protein has a sequence similar to the consensus sequence of zinc binding
region of zinc metalloproteases. Here we present some evidence suggesting
that Lit protein may be a member of the zinc metalloprotease superfamily,
and that the consensus sequence His-Glu-Xaa-Xaa-His might play a role in

the activity, stability or conformation of Lit protein.

34

Introduction

The phenomenon of phage exclusion has been known and studied for
over 40 years and several phage exclusion systems are now well understood
and characterized. Among the best understood of these systems is the e14
exclusion of bacteriophage T4. The defective prophage e14 is a DNA
element integrated into the the isocitrate dehydrogenase gene in the
chromosome of many E. coli K-12 strains (1). It encodes a 34 kDa protein,
Lit (iate inhibitor of T4) which proteolyses translation elongation factor Tu
(BF-Tu) after infection by bacteriophage T4. Using purified proteins, we
were able to show the cleavage of EF-Tu in a system involving Lit protein,
chemically synthesized Gol peptide and EF-Tu (2). Also using purified
proteins, we present evidence here that Lit protein may belong to the zinc
metalloprotease superfamily, and that the characteristic zinc metalloprotease
sequence His-Glu-Xaa-Xaa-His it possesses could play a role in its activity,
stability or its conformation. In addition, we show that Lit protein is active
even after being pre-incubated over a range of temperatures, and is inhibited

by inhibitors of metalloproteases.

MATERIALS AND METHODS
Overproduction and Purification of EF-Tu. E. coli DH5a harbouring
pGEXZT-tqu (3 ), an EF-Tu over-producing clone in which the tqu gene of
EF-Tu was cloned downstream of the glutathione-S -transferase gene in the
pGEX-2T vector (a gift from the Parrneggianni laboratory), was grown in
500 ml Luria Bertani (LB) broth supplemented with ampicillin (50 ug/ml) at

35

30°C with shaking. EF-Tu was overproduced by inducing the culture with
isopropyl B-D-thiogalactoside (IPTG) (0.2 mM final cone.) for 3 hrs at mid-
log phase during growth. Cells were harvested by centrifugation, re-
suspended in buffer A: 50 mM Tris.Cl, pH 7.5/ 150 mM KCl/ 5 mM MgC12/
1 mM DTT and disrupted by sonication. The suspension was then
centrifuged at 12,000 x g for 1 hr at 4°C. The supernatant was loaded onto
pre-washed Redipack glutathione column (Pharmacia) and allowed to stand
at 4°C for 30 min. The column with the bound GST-EF-Tu fusion protein
was washed several times with ice cold buffer A. The EF-Tu was eluted by
developing the column with re-suspended thrombin according to
manufacturer’s instructions (Pharmacia). The glutathione-S-transferase
remaining on the column was eluted with reduced glutathione. The protein
was stored in 15% glycerol at -70°C. The EF-Tu was >90% pure as judged
by sodium dodecyl sulfate/polyacrylamide gel electrophoresis (SDS/PAGE)
(Fig. 1). Protein concentration was determined by the Bradford method (4),
using bovine serum albumin as the standard. The yield was typically 25-50
mg/liter of culture.

Overproduction and Purification of Lit. To facilitate purification of Lit
protein, the lit gene was cloned into the pET-3 0b vector (N ovagen). This
vector expresses cloned genes from the powerful promoter and the highly
efficient ribosome binding site (rbs) from the phage T7 major capsid protein
gene and also fuses the protein to a string of six histidines (His-tag) and an S-
tag to allow its detection. To clone the lit gene in flame, flanking BamHI sites
were engineered by PCR onto the ends of a 998 bp lit gene fragment lacking

its own promoter and rbs. The amplicon was cloned first into the BamI-II site

36

of pUC8 vector (Fig. 2). Colorless colonies were selected. The lit gene was
then sub-cloned into the EcoRI/Sall sites of pET-3 0b immediately
downstream of the cleavable N-terminal His-tag/ S -protein tag. The resulting
plasmid was called pEKS (Fig. 3). The His-tag allows for a one-step protein
purification using the nickel affinity column (N ovagen) and the S-tag allows
its detection. The pEKS plasmid was used to transform competent
JM109DE3 lit°pLysS E. coli strain. Transforrnants were tested for over-
production of Lit protein. E. coli strain JM109DE31it°pLysS containing the
pEKS plasmid was grown in 120 ml of LB broth supplemented with
kanarnycin (50 ug/ml) at 37°C in a shaker water bath to mid-log phase. The
culture was shifted to 23°C for 30 min and then induced with isopropyl-B-D-
thiogalactoside (IPTG) (0.1 mM final cone.) to over-produce the fusion Lit
protein. After 3 hrs., cells were harvested by centrifugation, re-suspended in
1X binding buffer: 20 mM Tris.HCl, pH 7.9/0.5 M NaCl/5 mM irnidazole,
and disrupted by sonication. The suspension was centrifuged at 12,000 x g
for 40 min at 4°C. The supernatant was filtered through a 0.45 pm filter and
then layered on a pre-equilibrated 2.5 ml nickel affinity column and allowed
to stand for 20 min. The column with the bound fusion protein was washed
with 1X wash buffer: 20 mM Tris-HCl, pH 79/05 M NaCl/60 mM
imidazole. The fusion protein was eluted from the column by developing the
column with 3 ml of IX Wash buffer for 8 hrs. The fusion protein was
stored in 15% glycerol in -70°C. The Lit protein was >90% pure as judged by
SDS/PAGE (Fig. 4). Protein concentrations were determined as above.
Typically, protein yield was 10-25 mg from 1 liter of cell culture.

37

Lit Activity Assays. Standard EF- Tu in-vitro cleavage assay. Lit protein,
EF-Tu and chemically synthesized Gol peptide were combined in a total
volume of 45111 at final concentrations of 0.466 11g, 2.55 11g and 0.2 mM
respectively. This reaction mixture was incubated at 30°C for 30 min. after
which it was stopped by addition of sodimn dodecyl sulfate (SDS) gel
loading buffer followed by boiling. The reaction products were analyzed on a
sodium dodecyl sulfate-polyacrylarnide gel (SDS-PAG) containing 13%
acrylarnide. The gel was electrophoresed at 60 volts for 3 hrs, fixed in 12%
nichloroacetic acid and stained with 0.1% Coomassie blue. The depletion of
the 43 kDa EF-Tu band and the appearance of a 36 kDa EF-Tu“ band was
indicative of Lit protease activity. The bands were quantified in the AMBISS

laser densitometer.

RESULTS.
Effect of Temperature on Lit Stability. We have previously reported the
cleavage of EF-Tu in a purified three-component system containing Lit
protein, chemically synthesized G01 peptide and EF-Tu (2). Those studies
also had shown that Lit protein functions enzyrnatically. One of the factors
that affect enzyme function is temperature. We studied the effect of
temperature on the stability of Lit protein by first incubating the Lit protein to
be used for the in vitro cleavage assay at different temperatures for 15 min,
transferring to 30°C for 2 min before performing the cleavage assay at 30°C
for 30 min as previously described. The results were expressed as the
percentage of EF-Tu cleaved. From these results shown in Figure 5, it is
clearly evident that while temperature has an effect on the stability of Lit

38

protein, the protein is thermostable. Lane 4 shows that with incubation at
30°C, virtually all EF-Tu in the reaction was cleaved under these conditions.
Although activity of Lit protein was reduced as the temperature of incubation
was increased, as shown in Lanes 5 - 7, the reduction only became
substantial from 65°C and higher. This suggests that Lit protein is
thermostable.

Effect of Inhibitors an Activity of Lit Protein. Lit protein has the
consensus zinc metalloprotease motif: His-Glu-Xaa-Xaa-His shown in Figure
6, that may suggest membership within zinc metalloprotease superfamily (5-
10). If Lit protein is a zinc metalloprotease, it should be sensitive to inhibitors
of metalloproteases such as ethylenediaminetetraacetic acid (EDTA) and
1,10-phenanthroline, while phenylrnethylsulfonyl-fluoride (PMSF) a serine
protease inhibitor, Leupeptin, a serine/thiol protease inhibitor, and Pepstatin
A, an acid protease inhibitor, ought not inhibit its activity. However, these
compounds do not inhibit all enzymes within the indicated classes and
alternate inhibition are known. The effect of each of these different protease
inhibitors on the activity of Lit protein was studied. Lit protein was first
incubated at room temperature for 10 min with the following protease
inhibitors, separately : EDTA (5 mM); 1,10-phenanthroline (0.1 mM); PMSF
(0.5 mM); Leupeptin (50 uM); and Pepstatin A (50 uM). These samples
were then used in the standard in vitro EF-Tu cleavage assay as previously
described. Results shown in Figure 7 demonstrate inhibition of activity of Lit
protein by the metal chelator, EDTA and 1,10-phenanthroline. Surprisingly,
Lit protein activity was also inhibited by PMSF, an inhibitor of serine

proteases. This would seem to suggest that Lit protein is a serine protease. It

39
is not clear why PMSF would have this effect on Lit protein. It seems

unlikely that Lit protein could be a serine protease because it is not known to
possess a catalytic triad. However, there is a case reported of a different
metalloprotease being inhibited by PMSF. The urease isolated fi'om
Staphylococcus xylosus is a metalloprotease that contains threonine instead of
cysteine, an amino acid located within a highly conserved domain of 17
amino acids, that is supposed to be a part of the enzyme active site (11).
Neither Leupeptin nor Pepstatin A, inhibitors of thiol and acid proteases
respectively, significantly inhibited Lit protein, suggesting that Lit protein is

neither a thiol nor acid protease.

Analysis of Metal content of Lit Protein. If Lit protein is a zinc
metalloprotease, it should contain zinc. The result of metal analysis using
inductively coupled plasma emission spectroscopy is shown in Table 1. Less
than 1% of the Lit protein contain a zinc atom. It is still possible that Lit
protein is a zinc metalloprotease but , if so, it may have lost zinc during the
purification based on this result. We have not made careful measurements of

the percentage of Lit protein which is active when purified by this method.

Site-directed Mutagenesis of the l6°His-l°’Glu-Xaa-Xaa-‘°"His motif. If Lit
protein is a zinc metalloprotease, then changing the zinc metalloprotease
motif would be expected to affect its activity. Using the PCR-based
QuikChange site-directed mutagenesis method of Stratagene, the mutations:
H160A (cat160gct), E161A (gaa161gca), were made. The template DNA
used in both mutations is DNA of the over-expressing clone of lit, pEKS.
The PCR products were used to transform the competent E. coli strain,

40
JM109DE3lit°pLysS. 40% of all transformants tested lit° (do not make Lit

protein). These were selected. From these, “mutant” Lit protein was
produced, purified and tested for activity in the standard in vitro EF-Tu
cleavage assay as previously described. Figure 8 shows the result of the SDS-
PAGE analysis. Lane 3 shows cleavage of EF-Tu as expected of the “wild
type” Lit protein from the non-mutated over-expressing clone. In Lane 6,
“mutant” Lit protein from E161A does not cleave EF-Tu. The mutation
giving rise to E161A Lit inactivates the protease, consistent with the
interpretation that Lit protein is a zinc metalloprotease. The mutation that
yields Hl60A protein appears to significantly reduce the yield of Lit protein.
Figure 8 shows dot blot of wild type Lit protein and mutant Lit proteins from
H160A and E161A. From this figure, it is clear that more of the mutant Lit
protein fi'om H160A mutation is in the pellet, suggesting that it has gone into
inclusion bodies. This suggests that the H160A mutation changes the
conformation of Lit protein and makes it difficult to determine whether the
l°°His is required for its protease activity. Also from this figure, it is clear
that more of the wild type Lit protein and the mutant Lit protein from E161A
mutation are in the soluble fraction. More changes e.g. (Hl60R, H160K,
E161D, E161Q, H164R, and H164K) would be needed to fully study the
effects of mutating this motif before a definite role for these residues can be
ascribed. However, based on what is known of the chemistry of active sites
of proteins that possess this motif, l°°H and 1“H may serve as metal ligands,
while 1“B may fimction as a catalytic base (7-10).

41

DISCUSSION
We have purified both the translation elongation factor Tu (BF-Tu) and Lit
protein (product of e14 element of E. coli K12) fi'om clones that over-express
these genes. The purification of these proteins was facilitated by cloning their
genes into cloning vectors that introduce affinity tags. This allows the
resulting fusion proteins to bind preferentially to affinity columns and so
permit a one-step purification.

Using the purified proteins in an in vitro cleavage assay reaction, we were
able to show that Lit protein remained active as the temperature of pre-
incubation was raised from 30°C to 55°C. The reduction in activity became
substantial from 65°C and higher. This suggests that Lit protein is quite
thermostable.

We were also able to show that Lit protein is inhibited by EDTA and
1,10-phenanthroline. EDTA is a strong metal chelator as is 1,10-
phenanthroline which has a strong affinity for zinc and has been used in
many cases to remove zinc from an enzyme to produce the apoenzyme (12-
13). We were unable to restore activity to Lit protein by adding back zinc.
Although the activity of Lit protein was inhibited by PMSF, a serine protease
inhibitor, we have no compelling evidence to indicate that Lit protein is a
serine protease (Lit protein does not have the catalytic triad characteristic of
serine proteases). It is not clear why PMSF has this effect on Lit protein.
However, there is a case reported in which the activity of a known
metalloprotein, urease, isolated from Staphylococcus aylossus was inhibited
by PMSF (11). The activity of Lit protein was not inhibited by inhibitors of
both acid and thiol proteases. Although less than 1% of Lit protein contains

42
zinc as shown in the results from metal analysis of Lit protein, it may still be

a zinc metalloprotein. It could have lost most of its zinc during purification.
If the metal analysis result is taken in conjunction with the inhibitor studies,
the site-directed mutagenesis of E161A and the presence of the His-Glu-Xaa-
Xaa-His in what is presumably the active site of the protein, it could suggest
that Lit protein may be a member of the zinc metalloprotease superfamily.
Structural studies would be required to identify the active site of Lit protein
and to definitely classify Lit protein as a zinc metalloprotease.

The His-Glu-Xaa-Xaa-His motif has been used to identify zinc-binding
sites in metalloendopeptidases (15). If this conserved consensus sequence
which is present in the Lit protein plays a similar role as it does in zinc
metalloproteases, then mutating the invariant residues could shed some light
on the role of those residues in the activity of the Lit protein. We successfully
mutated H160A and E161A and tested the resulting “mutant” Lit proteins.
Mutation H160A appears to have changed the conformation of Lit protein
because most of the protein is found in the pellet, suggesting that the protein
went into inclusion bodies. This makes it difficult to determine whether this
residue plays a similar role in Lit protein as it does in the canonical His-Glu-
Xaa-Xaa-His sequence of zinc enzymes where it is required to ligand zinc (7,
14-16). Figure 4a shows that the mutant Lit protein from E161A is made and
more of it is in the soluble fraction. However, this mutant Lit protein is not
active. In a majority of metalloproteases studied so far, the glutamic acid
residue serves as a catalytic base (15). So the loss of function due to the
E161A mutation in which alanine is unable to firlfill the catalytic base
firnction of glutamic acid may be responsible for the lack of activity observed
with mutant Lit protein fiom E161A and not zinc binding since the glutamic

43
acid residue is generally not used for zinc binding. There are cases where the
activity of the protein is only slightly or not affected by such a change. In
these proteins, glutamic acid in that conserved sequence may be replaced by
asch acid with little or no loss of activity (10). Therefore fi'orn these
preliminary results, l°°His and 161Glu may play some role in the activity of Lit
protein or at least its stability or conformation. More detailed structural
studies will be needed to assign definite roles in the activity of Lit protein to

these residues.

9.

44

REFERENCES

. Hill, C. W., Gray, .1. A. & Brody, H. (1989) J. Bacteriol. 171, 4083-4084.

Georgiou, T., Yu, Y.-T. N., Ekunwe, S., Buttner, M. J., Zuurmond, A.-
M., Kraal, B., Kleanthous, C., & Snyder, L. (1998) Proc. Natl. Acad. Sci.
USA 95, 2891-2895.

. Cetin, R., Anborgh, P. H., Coal, R. H., & Parmeggianni, A. (1998)

Biochemistry 37, 486-495.

Bradford, M.M. (1976) Anal. Biochem. 72, 248-254.

. Schiavo, G., Rossetto, O., Santucci, A., DasGupta, B. R., & Montecucco,

C. (1992).]. Biol. Chem. 267, 23479-23483.

Klimpel, K. R., Arora, N., & Leppla, S. H. (1994) Mol. Microbiol. 13,
1093-1100.

Meinnel, T., Lazennec, C., & Blanquet, S. (1995) J. Mol. Biol. 254, 175-
183.

. Vazeux, G., Wang, J ., Corvol, P., & Llorens-Cortes, C. (1996) J. Biol.

Chem. 271, 9069-9074.

Vallee, B. L., & Auld, D. S. (1993) Biochemistry 32, 6493-6500.

10.Cha, J ., & Auld, D. S. (1997) Biochemistry 36, 16019-16024.
11.Jose, J., Schafer, U. K., & Kaltwasser, H. (1994) Arch. Microbial. 161,

384-392.

12.Drum, D. E. & Vallee, B. L. (1970) Biochemistry 9, 4078-4086.

45
13.Giedroc, D. P., Keating, K. M., Williams, K. R., Konigsberg, W. H.,
Coleman, J. E. (1986) Proc. Natl. Acad. Sci. USA 83, 8452-8456.

14.Vallee, B. L. & Auld, D. S. (1990) Proc. Natl. Acad. Sci. USA 87, 220-
224.
15.Jiang, W., & Bond, J. S. (1992) FEBS Lett 312, 110-114.

16.Le Moual, H., Roques, B. P., Crine, P., & Boileau, G. (1993) FEBS Lett
324, 196-200.

46

Figure Captions

Figure 1: Purification of EF-Tu. This figure shows the purification of EF-
Tu from the overexpression clone, pGEXZT-tqu. Lane 1 has cell extracts of
the overexpressed cell culture; Lane 2 shows GST-EF-Tu fusion protein
eluted hour the glutathione affinity column; Lanes 3 and 5 represent EF-Tu
cleaved off GST-EF-Tu fusion protein with thrombin and Lane 4 is GST
alone.

 

of El-
racts if
protet
L El-ll
is 651

h

 

1.39
”.1. :13
”.1 33139“

48

Figure 2: Cloning of lit gene. PCRlit3 is a PCR-amplified 998 bp lit
gene fragment that was designed such that it lacked its own promoter and
ribosome binding site, but possessed BamHI sites engineered onto its ends
to facilitate its cloning into the BamHI site of pUC8 cloning vector.

 

49

 

50

Figure 3: Cloning of lit gene (continued). The lit gene fragment from
pTAH was subcloned between the EcoRI and Soil sites of pET3 0b cloning
vector. The resulting plasmid, called pEKS, overexpresses the lit gene.

 

51

pET30b
5.422 kb

   

(if
VII

 

 

 

52

Figure 4: Purification of Lit protein/Activity assay. Lane 1 was loaded
with cell extracts of uninduced cell culture from E. coli strain
JM109DE3lit°pLysS containing the pEKS plasmid. Lane 2 has cell extracts
of induced cell culture fi'om the same E. coli strain as in Lane 1. Lane 3
contained column flow-through fraction after induced cell extracts were
loaded on the Ni-affinity column. Lane 4 was loaded with purified His-
tag/S-tag Lit fusion protein. A: Lane 5 represents purified EF-Tu. Lanes 6
and 7 represent Lit protein activity assay. In Lane 6, 4.66 ug Lit protein and
2.55 ug EF-Tu were combined in a total volume of 45 111. Lane 7 is the
same as Lane 6 except that 2 mM chemically synthesized Gol peptide was
added. Contents of Lanes 5-7 were incubated at 30°C for 30 minutes.
Activity of Lit protein is shown by the cleavage of EF-Tu. This is seen in
Lane 7 where the band corresponding to EF-Tu is missing and a band, EF-
Tu“, representing the cleavage fragment appears.

 

 

 

kDa

68——
43—

29—

 

 

54

Figure 4-1: Effect of dilution on ability of Lit protein to cleave EF-Tu.
Panel A: Lanes a, b and c were loaded with 0.466 pg of purified Lit protein,
1.3 pg of purified EF-Tu and a mixture of purified Lit protein and EF-Tu at
concentrations of 0.466 pg and 1.3 pg respectively. Lane d was the same as
Lane 0 except for the addition of 0.2 mM Gol peptide. The reaction volume
of each set up was 45 pl. The reaction was done at 30°C for 30 minutes.
Lane d shows the cleavage of EF-Tu with the resulting cleavage fragment,
EF-Tu“. Lanes 1-5 contained the same reactants as Lane d except that the
concentrations of Lit protein were 4.66 pg, 0.466 pg, 0.233 pg, 0.155 pg
and 0.093 pg respectively. These results show that as the concentration of
Lit protein is reduced, less EF-Tu is cleaved.

Panel B: This is a plot of results of Lanes 1-5 showing the amount of EF-Tu
cleaved in percent versus the concentration of Lit protein.

 

55

 

 

 

 

°/o E F-Tu deaved

Mabcd12345

 

 

 

     

kDa
43——— .- -' .- .. __ 4‘7“
29 ‘ a. -- - - ... — , _EF.W
t
1
Elect of dilution on abilityot Lit prohin b deava EF-Tu
m
90 —.
80 _.
70 ~
60 4
50 .
40 .
30
2O
10
I" [L - . .
L: 391:" 71 155 O 233 O 466
A'Mvivmll 01 Li! protein 111 reaction mixture [pg]

B

 

56

Figure 5: Effect of temperature on Lit protein stability.

Panel A: shows SDS-PAGE. The standard Lit activity assay was modified
by using Lit protein that had been incubated at different temperatures before
assay was conducted. In all cases, the Lit protein was incubated at the
chosen temperature for 15 minutes. Lane a contained 0.466 pg Lit protein;
Lane b was loaded with 0.466 pg Lit protein and 1.3 pg EF-Tu. These two
lanes were controls. All other lanes were loaded as in Lane b except that
0.2mM G01 peptide was added to each one. The incubation temperatures of
the Lit protein used were 30, 37,45, 55, 65, 80 and 90°C. Lit protein
showed stability up to 55°C as evidenced by the amount of EF-Tu cleaved.
Panel B: is a graphical representation of Panel A.

57

 

 

Ma b$3745556580410°€

2.— 1*..- -- - —EF—Tu'

Eflect of temperature on the stabilityot Lit prohin

 

70q

60...

50—4

40*

% EF-Tu

30“

20

 

 

 

3C 7 4‘; 55 65 BO 90

7 {finger mutt-3 loll

B

 

58

Figure 6: Consensus sequence of zinc-binding region. Comparison of Lit

protein sequence with the signature motif of zinc peptidases to show region
of similarity.

 

59

 

 

r: 9.085” -350 be: am 3mm... 0:. 5 m9. Em...

N: woman—omen 1:0 X X Em GE :0 m9. Em

m2. <3 .2...
>5 go» >5
5: he:
.2: 25

.5

 

ll in
air

 

60

Figure 7: Effect of typical protease inhibitors on activity of Lit protein.
Panel A: Standard assay: 0.002 mM Lit + 0.04 mM EF-Tu + 0.2 mM Gol
peptide in a total volume of 45 p1. Incubated at 30°C for 30 minutes.
Stopped by adding SDS loading buffer followed by boiling. Lane 1 is EF-
Tu alone. Lane 2 is EF-Tu + Lit protein, while Lane 3 is EF-Tu + Lit
protein + Gol peptide. Lanes 4-8 are the same as Lane 3 except that the Lit
protein has been incubated with a different protease inhibitor for 10 minutes
before being assayed. The protease inhibitors used at the final
concentrations shown were EDTA (5 mM), 1,10-phenanthroline (0.1 mM),
PMSF (0.5 mM), Leupeptin (50 pM) and Pepstatin A (50 pM) respec-
tively. PMSF was dissolved in isopropanol. In Lanes 4-6, Lit protein was
inhibited by the typical protease inhibitors in those lanes, whereas inhibition
was only partial in Lanes 7 and 8.

Panel B: graphical representation of Panel A.

61

M 1 2 3 4 5 6 7 8
11011
n
68—— M
43—— . —- - ——EF-Tu

—-EF-Tu‘

29— .

 

100a

80_.

70*

60—

50*

40“

% Uncleaved EF-Tu

30*

20 r

 

 

 

Lane 2 Lane 3 Lane 4 Lane 5 Lane 6 Lane 7 Lane 8
Protease Inhibitors

62

Table

Table 1: Metal analysis of Lit protein using inductively coupled plasma
emission spectroscopy.

The values shown in the table are (ppm). The concentration of Lit protein is
4 mg/ml. The blank is 1X binding buffer: 20 mM Tris-HCl, pH 79/05 M
NaCl/S mM imidazole. 1 mM metal solutions were made by dissolving the
metal salts in 1X binding buffer. The samples were analyzed in the
Chemical Analysis Laboratory of University of Georgia, Athens, GA.

Note: Lit protein was purified on a Ni-affinity column.

muoaOJn

Du
D—
be
0

0-
DO
U»
00
no
no
D1
Ur.
WI
1

In
33
30
2|
2»
U

DO
ma
w.
my
0....
no.1
4»
C

<

t

N!

'1);

.0.000000
0.000000
0.0000
0.0000
.0.000000
.0.000000
0.000000
0.00.00
0.000000
.0.000000
.0.000000
Ooopwbfiu
0..0_00
.0.000000
0.000000
0.000;.»
0.00000.
0000.0
20.000000
0..»000
20.000000
A0.000000
0..0000
.o.oooooo
.u.»rpu
0.000000
.0.000000
.o.0ooooo
«0.000000
.0.000000
0.005400

00

.0. 000000
0.000000
.0.0&0N00
0.u0000
.0. 000000
0.00wn00
0.0).fifl0
.0.000000
0.0»0000

0:

.0.000000
0.000000
.0.000000
0.00000m
.0.000000
.0.000000
0.00000
0.000000
0.004000

00.400. 0.0Nuflflfl

.0. 000000
0. 00080
.0. 000000
0. N000

.0. 000000
0. 2000p
0. 00nd:
u. "Duu

0. Oflaflo
.0.oooooo
0. u ~00)
.0. 000000
0. ougwu
.0. 000000
A0. 000000
.0. 000000
.0. 000000
0. 00400
0. 000000
.0. 000000
.0. 000000

0. 00—0—0

“we..‘0

.0. 000000
N. 0000

0. 0 u u #0.
0. 000» .0
.0. 000000
0. 00000
.0. 000000
0. 00000
0. 0000» n
50.. 000000
A0. 000000
Ao. 000000
no. 000000
50. 000000
0. 000000
0 . u u «00
AO. 000000
“0. 000000
0. 00:00

II

.0. 000000
0.. 000000
.0. 000000
0. 000000
0. 8000p
.0. 000000
«0. 000000
0. 0000.
0. 0N~0Ofl
0. OnNdUD
0. 00.0%»
0. ouuou

O). 50

0. ON:

0. 00» p00
0. Oflgflo
0. 80.000
.0. 000000
0. 00.»:
A0. 000000
0. 000000
0. 0000p”
0. 00000»
.0. 000000
0. 0 ”0000
0. 0000p“
.0. 000000
0. Oflnuu
0. 0a #000
«0. 000000
0. 00030

N)

.0.000000
0.000000
.0.000000
0.000000
.0.000000
.0.000000
.0.000000
0.0000.»
0.00.0.0
0.0.00.0
.0.000000
0.060000
0.000000
0.0000
.0.000000
0.00»00»
0.0:0000
0.0.0000
.0.000000
.0.000000
0.0mmouc
0.00000
0.000000
.0.000000
20.000000
0.000.00
.0.000000
0.0.000
0.00.0~m
0.00000

'“00”..

Z»

.0. 000000
0. 000000
.0. 000000
0. 000000
.0. 000000
.0. 000000
0. .0000
0. 000000
0. 0— ”00%
0. ONOO» w
.0. 000000
0. 00400.

.0. 000000 .

D. 0.03
O. Oflmgfi
O. OUOQOV
0. 8800
A0. 000000
"a0flqwd
no. 00080
O. OOONOO
.0. 000000
0. 00 u #09
«0. 000000
A0. 8008
O. 0000 p N
“0.000000
0. @N r @O
O. COVOOQ
AO- 00080
O. N w CON

rm.

.o.000000
0.000000
n. 0000
0.00040
8. 0808
.0.000000
.o.oooooo
o. 0flupfl
o.opouou
.0.000000
.0.000000
.0. 000000
0.000».
.o.oooooo
0. can» 00
0.004000
0. 03.50
0000.0
0.09400
o.»qmm.
8. 000000
3. 000000
o.menqo-
20.000000.
0.0»ou
o.opu~ow
.0. 000000
.0.000000
3. 000000
.0.000000
0. aged

Table 2: The characteristics and references of bacterial strains,
plasmid constructs and phage mutants used in this article.

65

mg
0300.

 

ace

0 _ o
Eézug
Eourwmm

3:535
3mX~a¢§>
0C0 x

0m... mod
35m

.0me

2:50
0.0030

 

maugkmocaoamewog.$025380.»—
025 _ $5.693- _ 8...:

ES _ 3.003. 056800000 5. .0885.
235... 0000.. 3.03 3. C.<

.938. m. 00: W 008.. 0m0... 00:00.5 0. 0mm.
£3. 3 W2... 00530800 mama .5000 5030
033080 0.0.. 90 3 €800.50 2003....

3%.. «0:0 08:00 58 Hummus... 0.03.5 <88.

$25.5: Emu“ mason—0 wag: $038100

.30.

.90 00.. main... .03.
W00 2:. mes—00. .03.
.25.. £03..

000.00mb... 500880..

Em: 003. 0:302 0.003... 90. 9.00.8.0 9.00820- 7.80% .0qu 203500.. 0. a; $me <58?

80:82.0: .0 038.0108 .80..

083.5 <88. 0050.5 .5 2.83.3». Eméfi
Ecacmamémfiaaafiammo 00000000. 000650
390.20 3 0330.00.

wow :0 «0:0 mamas: 08:00 38 ESE. 0:0
0.. .060 m.

VOW Rm «000 mam—50:. 0x. 3...... 0:70—0:00
002.00: M003 8... me: 0:00 0.. cm... we? 96?
089.000 5. .0885.

0000800000 808:. $5000 3:82.00 0:05 urn—ma
0825 .0 5.003 0.350.

.0030: 0. w... Emu.
29.0%"?

.35 200...

...—am $01..

0.533000. .2. BE Mada—u ... .05.

66

Figure 8: Site-directed mutagenesis of the DNA encoding the 1”His-
161Glu-Xaa-Xaa-mflis motif. Panel A shows the standard Lit protein
activity assay done with a mutant Lit protein. Lanes 1-3 show wild type Lit
protein: Lane 1 is Lit protein alone, Lane 2 is Lit protein + EF-Tu and Lane
3 is Lit protein + EF-Tu + G01 peptide. Lanes 4-6 show E161A mutant Lit
protein: Lane 4 is mutant Lit protein alone, Lane 5 is mutant Lit protein +
EF-Tu and Lane 6 is mutant Lit protein + EF-Tu + 601 peptide. Panel B
shows dot blots of soluble fractions and pellets of wild type Lit, mutant
H160A and E161A protein. Going from left to right, the dots are undiluted,
1:5, 1:10,]:15 and 1:20 dilution respectively. The pellet was resuspended in
4 ml 1X Binding buffer. 3 pl of each boiled protein solution was spotted on
nitrocellulose and the protein was detected using the S-tag detection Kit

(N ovagen).

67

 

 

kDa
43 - a a —EFOTU
"3 EFCTU.
29—
Wt—
H160A—

Ei61 A—

 

CHAPTER 3
THE PEPTIDE ENCODED BY A SHORT REGULATORY REGION OF
BACTERIOPHAGE T4 BINDS TO TRANSLATION ELONGATION
FACTOR Tu.
To Be Submitted to Cell
STEPHEN I.N. EKUNWE AND LARRY SNYDER“
DEPARTMENT OF MICROBIOLOGY

MICHIGAN STATE UNIVERSITY
EAST LANSING, NH 48824-1101

*To whom reprint requests should be addressed.

68

69

Summary

The defective prophage e14 in E. coli K12 excludes T4 and other T-even
bacteriophages by encoding a protease, called Lit, that cleaves translation
elongation factor Tu (BF-Tu) afier phage infection, thereby blocking
translation and preventing the multiplication of the infecting phage and its
spread to other e14-containing cells. The proteolysis of EF-Tu by Lit protease
is activated by the appearance in the cell of the short Gol peptide internal to
the major head protein of the infecting phage, and the cleavage of EF-Tu
occurs when the region of the major head protein gene encoding this peptide,
the gal region, is transcribed and translated in the late stage of infection. The
cleavage of EF-Tu has been demonstrated in a purified in vitro system
containing only Lit protein, EF-Tu, and a chemically synthesized 29 amino
acid peptide that is encoded by the gal region. No detectable cleavage of EF-
Tu occurs unless the G01 peptide is added. The Gol peptide might activate the
proteolysis of EF-Tu by Lit protein in one of two ways. Either the G01
peptide binds to the Lit protein and activates an otherwise dormant protease
activity, or the G01 peptide binds to EF-Tu and somehow creates the unique
substrate for an already active Lit protease. In this paper, we present a
number of lines of evidence in support of the second possibility: the G01
peptide binds to EF-Tu and this binding is required to convert EF-Tu into a
substrate for the Lit protease.

70

Introduction

Important insights and research tools have come from studies of how cells
protect themselves against viruses. One method of protection is through
programmed cell death, in which the infected cell kills itself to prevent
multiplication of the virus and its spread to other cells. Even single-celled
bacteria use this strategy in the form of phage exclusion systems (Snyder,
1995; Yarmolinsky, 1995). The two phage exclusion systems that are best
understood (e14 and prr exclusions of T4) are directed against T-even
bacteriophage and use remarkably similar strategies although their molecular
bases are very different. In both these exclusions, an enzyme encoded by the
indigenous prophage is constitutively synthesized. After infection, due to the
appearance in the cell of a peptide encoded by the infecting phage, the
enzyme then specifically cleaves an evolutionarily highly conserved
component of the translational apparatus, thereby blocking translation and
multiplication of the infecting phage. In one such exclusion system, the
defective prophage e14 constitutively synthesizes a protease, Lit (for Late
inhibitor of 14) which is highly specific for translation elongation factor Tu
(EF-Tu) (Kao and Snyder, 1989; Yu and Snyder, 1994). The proteolysis of
EF-Tu is activated by a short polypeptide determinant, the G01 peptide,
internal to the major head protein of the infecting phage (Champness and
Snyder, 1984; Bergsland et al., 1990; Georgiou et al., 1998). Cleavage of EF-
Tu occurs site-specifically between glycine 59 and isoleucine 60 in the highly

conserved effector region of EF-Tu responsible for triggering the cleavage of

71
GTP when EF-Tu with tRN A bound enters the A site of the ribosome. The

cleavage of EF-Tu causes a air-dominant inhibition of translation of mRNA
sequences downstream of the gal region and a strong cis-polarity effect on
downstream transcription (Bergsland et al., 1990), thereby preventing the
synthesis of the T4 major head protein and blocking the development of the
infecting phage. Overproduction of the Lit protease in the presence of Go]
peptide can cause the cleavage of all the cellular EF-Tu, totally blocking
cellular translation (Bergsland et al., 1990; Yu and Snyder, 1994). Note the
similarity between the mechanism used by these phage exclusions and other
types of viral defense mechanisms such as those induced by interferon in
which infected cells are killed by specific enzymes that target evolutionarily
highly conserved cellular constituents, ofien components of the translation
apparatus.

The Go] peptide, so named because mutations in this region of the major
head protein gene allow the phage to grow on Lit protein containing bacteria
(gal mutations, for grow on Lit), is less than 29 amino acids long, and
approximately one-fifth of the way in from the N terminus of the protein. The
amino acid sequence of the Gal peptide, extending from amino acid 94 to
amino acid 122 of the major head protein prior to processing, is almost
identical in the four members of this diverse family of phages in which it has
been sequenced (Monod et al., 1997). Furthermore, a test of twenty wild type
isolates of T-even phages revealed that all are vulnerable to the Lit exclusion
(our unpublished observations). Therefore, the Lit exclusion system seems to
play an important role in defending cells containing the e14 element against
the large and ubiquitous family of T-even phages.

72
The activation of cleavage of EF-Tu has been demonstrated in a purified

in vitro system by adding the chemically synthesized 29 amino acid Gol
peptide to EF-Tu and Lit protein (Georgiou et al., 1998). Therfore no other
cellular components, either phage or host encoded, are required for the
activation of cleavage of EF-Tu. Furthermore, the Gal peptide seems to be
absolutely required for the activation of cleavage since no cleavage of EF-Tu
has been detected if the Gal peptide is omitted from the reaction. Mutations
in the gal region, which reduce the ability of the mutant Gol peptide to
activate the cleavage of EF-Tu also alleviate the cis-dominant inhibition of
downstream translation, indicating that even limited cleavage of EF-Tu can
have a strong effect on translation in cis.

We imagine two general possibilities for how the G01 peptide might
activate the proteolysis of EF-Tu by Lit. One possibility is that the Gal
peptide binds to the Lit protein and somehow activates a dormant protease
activity, for example by furnishing a needed amino acid for the active center.
Another possibility is that the peptide does not activate Lit at all, but rather
binds to EF-Tu and somehow converts EF-Tu into a unique substrate for the
Lit protease. In this paper we present a number of lines of evidence in
support of the latter possibility: that the peptide binds to EF-Tu, somehow
creating the substrate for the protease.

73

Results

Equimolar amounts of Col peptide and EF-Tu are required for
complete cleavage of EF-Tu.

As mentioned in the Introduction, there are two general possibilities for how
the G01 peptide activates the proteolysis of EF-Tu. Either the peptide binds to
the Lit protease and somehow activates it, or the peptide binds to EF-Tu and
somehow creates the unique substrate for the already active Lit protease.
Clues to which general mechanism is correct may come from measurements
of the molar ratios of peptide to Lit and EF-Tu required for complete
cleavage of EF-Tu. Lower amounts of peptide may be required if the peptide
binds to Lit and activates it since Lit acts enzymatically (Georgiou et al.,
1998), and each molecule of activated Lit protease should be able to cleave
many molecules of EF-Tu. If, however, the peptide binds to EF-Tu and
creates the substrate for Lit proteolysis, higher amounts of Gol peptide may
be required because then one molecule of peptide might have to bind to each
molecule of EF-Tu before it can be cleaved.

To determine the molar ratios of Gol peptide to EF-Tu and Lit protein
required for complete cleavage of EF-Tu, we mixed the Lit protein and EF-
Tu at a molar ratio of about 1 to 20 and added varying amounts of Gol
peptide. In Figure 1 is shown the extent of cleavage of EF-Tu at each
concentration of Go] peptide. Figure 1-1 shows percent EF-Tu cleaved at
each concentration of Go] peptide and the molar ratio of peptide to EF-Tu

74
and peptide to Lit at that concentration. It can be seen that significant

cleavage of EF-Tu occurred only when the molar ratio of Gol peptide to EF-
Tu was 1:1 or higher, even though at these concentrations of Gol peptide
there was still a great molar excess of peptide over Lit protein. The fact that
such high concentrations of Gol peptide are required for efficient cleavage of
EF-Tu suggests that one molecule of peptide must bind to each molecule of
EF-Tu that is cleaved, rather than each molecule of peptide binds to and
activates a molecule of Lit protein.

In interpreting the results of the above experiment, we have assumed that
most of the peptide we added was active. The concentration of Gol peptide
was determined solely by measuring the optical absorbance of the solution of
the peptide (ex. Chiron Inc. ) and it is conceivable that the mixture contained
many inactive peptides, that were either, for example, too short or had been
chemically damaged during synthesis or storage. If only a small fraction of
the Gal peptide added was active, then the actual amount of active Gol
peptide present when significant amounts of EF-Tu were cleaved might have
been much less than estimated, and more commensurate with the lower

amount of Lit protein in the reaction.

Lit protease is not activated by a preincubation with Gol Peptide.

Another way to test the hypothesis that G01 peptide must bind to EF-Tu to
activate the proteolysis is to test a prediction of the alternative hypothesis:
that the G01 peptide binds to Lit and thereby somehow activates its protease
activity. If the Lit protease activity is activated by Go] peptide, then it should
be possible to activate the Lit protease by incubating Lit with Go] peptide at
high concentrations. Each molecule of activated Lit protease should then be

75
able to cleave many molecules of EF-Tu, since Lit protease acts catalytically.

If, however, the Gal peptide must bind to EF-Tu to activate the proteolysis,
the Lit protease will not have been activated during the preincubation and
amounts of G01 peptide commensurate with the amount of EF-Tu may still
have to be added to see significant cleavage of EF-Tu. Figure 2 shows the
results of such an experiment. In Lane C, 0.2 pg Lit protein was preincubated
with 0.2 mM Gol peptide. The mixture was diluted 1 to 20 and added to 1.5
pg EF-Tu in a total volume of 45 pl to initiate the cleavage reaction. Very
little EF-Tu was cleaved, inconsistent with the interpretation that the protease
had been activated during the preincubation. In Lane D is a control indicating
that the Lit protease is not inactivated by the preincubation. In this control,
the Lit protein was preincubated by itself, diluted 1 to 20 and then added to
0.2 mM Gol peptide and 1.5 pg EF-Tu in a total volume of 45 pl. Now,
substantial amounts of EF-Tu was cleaved. The last lane (Lane E) contains
an additional control that shows that preincubating Lit with Gol peptide is not
deleterious to the activity of Lit and that the G01 peptide is not significantly
inactivated by the preincubation. This lane was the same as Lane C except
for the extra Gol peptide added after preincubation and dilution. More EF-Tu
was cleaved. Lane B contains a control of undiluted components. Thus one
prediction of the alternative hypothesis, that the Gal peptide can activate Lit
during a preincubation, is not fulfilled, offering indirect support to the
hypothesis that the G01 peptide does not activate the Lit protease but rather
binds to EF-Tu and thereby creates the substrate for proteolysis.

76

Direct Evidence that Gol Peptide Binds to EF—Tu

If Gol binds to EF-Tu, it may be possible to detect this binding directly. One
way to demonstrate binding between two protein molecules is to fix one
protein on a column and then pass the second protein through the column,
usually accompanied by other proteins (cf. Formosa et al., 1991). If the
second protein binds to the first protein affixed to the column, its passage
through the column may be retarded relative to the other proteins. It may also
be that the second protein fails to elute altogether, or only elutes under
conditions in which the first protein elutes.

To determine whether the G01 peptide binds to EF-Tu, either EF-Tu or the
G01 peptide could have been bound to a colmnn and the other component
passed through the column. However, EF-Tu often behaved anomalously on
columns, eluting after most other proteins, even if no G01 peptide had been
attached to the column. This anomalous behaviour might reflect the tendency
of EF-Tu to polymerize under some conditions (cf. Beck et al., 1978).
Therefore, it proved more feasible to pass the Gal peptide through a column
to which EF-Tu had been batmd.

To determine if Gol peptide was retained on a column to which EF-Tu
had been attached, we needed a way to bind EF-Tu to a column matrix as
well as a way to detect the G01 peptide, since the latter is too small to
visualize on SDS PAGE gels. To fix EF-Tu to a column, we expressed the
EF-Tu protein from a plasmid pGEXtqu (Cetin et al., 1998) in which the
tqu gene encoding EF-Tu is translationally fused to the gst gene encoding

77
glutathione-S-transferase (GST). The GST portion of the resultant fusion

protein binds tightly to a glutathione column, and can be eluted with an
excess of reduced glutathione. To allow detection of the Gal peptide, we
expressed the G01 peptide from a plasmid pET30PZ1 in which the gal coding
sequence has been translationally fused to an S tag (see Experimental
Methods). The Gal peptide expressed from this plasmid is also attached to a
string of six consecutive histidines (His-tag) which allows its purification on
nickel affinity columns, which will be important for later experiments.

To investigate the ability of Gol peptide to bind to EF-Tu, extracts were
prepared separately from cells in which synthesis of the GST-EF-Tu fiision
protein and S-tagged Gol peptide had been induced (see Experimental
Procedures). The extracts were then mixed in a 1:1 ratio, incubated, and
layered on a glutathione column as in Experimental Procedures. The columns
were washed with an equal volume of the loading buffer followed by a buffer
containing reduced glutathione to elute the GST-EF-Tu fusion protein.
Fractions were collected during the loading as well as during the washes and
after elution with reduced glutathione. Aliquots of these fi’actions were then
analyzed by SDS-PAGE and protein staining to follow the elution of the
GST-EF-Tu fusion protein as well as other proteins in the extract and
analyzed by S-tag dot blots to detect elution of the G01 peptide (see
Experimental Procedures). Figure 3A shows the results of such an
experiment. Most of the GST-EF-Tu fusion protein, indicated by the arrow,
was stripped from the extracts as expected, and did not elute until the column
was washed with reduced glutathione. Most of the S-tagged Gol peptide
flowed through the column unimpeded and could be detected in the flow
through fractions (see dots below the lanes). However, some of the S-tagged

78
Gal peptide was retained on the column, and did not elute until the GST-EF-
Tu fusion protein eluted.

The retention of some of the S-tagged Gol peptide on the column to which
the GST-EF-Tu fusion protein was fixed and its subsequent elution with the
GST-EF-Tu fusion protein suggested, but did not prove, that some G01
peptide bound to the EF-Tu attached to the column. For example, the G01
peptide may be binding to the GST portion of the fusion protein, or binding
to some column component through either its S tag or His tag rather than its
Gol peptide sequence. If the Gal peptide was retained by virtue of its binding
to EF-Tu, it should not be retained if only the GST protein with no EF-Tu
attached is fixed to the column, or if the peptide being passed through the
column does not contain the G01 peptide sequence, but contains only the His
and S tags. The same experiment was repeated with the S-tagged peptide
made from the cloning vector pET3 Oc without the gal sequence cloned into
it, so the peptide did not contain the internal Gol peptide sequence although it
was identical at both the N and C termini. The GST-EF-Tu fusion protein
was retained on the column. Very little S-tagged peptide was retained and
eluted with the GST-EF-Tu fusion protein indicating that the S-tagged
peptide is being retained on the column because of its Gol peptide sequence
rather than some other sequence on the peptide (Figure 3b). We conclude that
at least some of the Gal peptide is being retained on the column through its
binding to EF-Tu that is attached to the column. Note that the peptide that
flows through the column unimpeded in the experiment in Figure 3A may
have been bound to the chromosomally-encoded EF-Tu (represented by the
strong band at 43 kDa in the flow through) which does not contain a GST tag
so it is not retained on the glutathione affinity column. Based on the results of

 

 

 

79
the binding experiments, a model for the activation of the proteolytic
cleavage of EF-Tu by Lit protease is proposed as shown in Figure 4.

Discussion

We have presented evidence that a small region of the major head protein of
T4 phage, the Gal peptide, binds to translation EF-Tu during the normal
course of phage infection. This binding was revealed because the peptide-
bound EF-Tu is the target of the specific Lit protease, encoded by the
defective prophage e14. Cleavage of EF-Tu by the protease blocks T4 and
other T-even phage development, causing phage exclusion. We have
presented separate lines of evidence that the Gal peptide binds to EF-Tu.
First, the amount of Gol peptide required for complete cleavage of EF-Tu is
more commensurate with the amount of EF-Tu in the cleavage reaction than
it is with the amount of Lit protease. Second, the Lit protease is not activated
by a preincubation with the G01 peptide, as might be expected if the G01
peptide activates the Lit protease by binding to the Lit protein. Finally, the
binding has been demonstrated directly even in the presence of all other
cellular proteins by showing that the Gal peptide is retained on an EF-Tu
affinity column and elutes when the EF-Tu elutes.

The binding of the Gal peptide to EF-Tu is not likely to be an in vitro
artifact since it is substantiated by both genetic and physiological evidence.
For example, T4 mutations that prevent the cleavage of EF-Tu after T4
infection or allow transformation of cells containing the Lit protein by
plasmids containing clones of the major head protein gene, T4 gene 23, all lie
in the Gal peptide coding region of T4 gene 23 (Champness et al., 1984;

80
Bergsland et al., 1990). At least some of these same gal mutations also

reduce the ability of the peptide to activate cleavage of EF-Tu in a purified in
vitro system (Georgiou et al., 1998). It will be interesting to determine if
some of these same gal mutations prevent or weaken the binding of the
peptide to EF-Tu, providing a test of this model for how they prevent the
cleavage. Our previous work had also indicated that only the G01 peptide is
required to activate cleavage of EF-Tu by Lit protease in viva and in vitro
(Y u and Snyder, 1994; Georgiou et al., 1998). Finally, the experiments
reported here indicate that the binding of the Gal peptide to EF-Tu is required
to activate cleavage of EF-Tu by the Lit protease in a purified in vitro system,
and it seems very likely that the mechanism of activation of cleavage of EF-
Tu occurs by the same mechanism in viva.

The results reported here and in earlier publications offer a satisfying
explanation for how the defective prophage e 1 4 excludes T-even phages.
Late in infection by a T-even phage, when the synthesis of late proteins
including the major head protein of the phage has begun, the G01 peptide

internal to newly synthesized major head proteins binds to EF—Tu. This
- binding is part of the normal course of infection, even in cells that are not
lysogenic for e14. However, if the cells are lysogenic for e14, as are most
laboratory E. coli K12 strains, EF-Tu with Gol peptide bound provides a
unique substrate for the e14-encoded Lit protease which is made
constitutively but has no substrate in the uninfected cell. The Lit protease
then cleaves EF-Tu, blocking the multiplication of the infecting phage and its
spread to other neighboring E. coli cells which, because they are often
siblings, also contain the e14 prophage. In this way, the e14 prophage

81
protects the population of host bacteria, and thereby itself, fiom phage

contagion.

The normal situation is probably somewhat more complicated than this,
however. Normally, not enough Lit protease is made from a wild type e14
prophage to cleave all the EF-Tu in the cell, and as much as 50% of the EF-
Tu remains uncleaved after T4 infection of wild type E. coli K12 strains (our
unpublished observations). These levels of uncleaved EF-Tu are sufficient to
maintain almost normal levels of total protein synthesis so total translation is
only partially blocked when T4 infects wild type E. coli K12 strains.
Nevertheless, T4 phage production in wild type E. coli K12 strains is severely
delayed, especially at lower temperatures, due to a severe delay in the
synthesis of the major head protein. This inhibition of the translation of genes
containing the gal sequence is apparently strongly ctr-acting, as has been
demonstrated in a number of ways. One demonstration of the cis-effect came
from coinfecting Lit protease-containing cells with wild type T4 and T4 with
a gal mutation (Champness and Snyder, 1984). The major head protein was
almost exclusively synthesized from the phage genomes with the gal
mutation. Another demonstration came from measuring the expression of
reporter genes fused translationally downstream to the gal region (Bergsland
et al., 1990). Very little expression of the downstream reporter gene
occurred, even though cell growth and expression of most of the other genes
in E. coli were not significantly affected. Assuming this cis-effect is due to
cleavage of EF-Tu, which seems likely considering that it is affected by the
same gal mutations and, as far we know, has all the same requirements as the
global inhibition of translation due to cleavage of all the cellular EF-Tu, it
appears that those EF-Tu proteins involved in translating the gal region are

82

preferentially cleaved and then act to block translation from ribosomes on the
same messenger RNA. But how could a cleaved EF-Tu due to a G01 peptide
exiting a ribosome act retroactively to block translation by that ribosome?
According to the textbook picture of the role EF-Tu plays in translation, EF-
Tu binds to GTP and arninoacylated tRN A and the ternary complex enters the
A site of the ribosome, promoting proper pairing between the codon in the
mRNA and the anticodon in the tRNA. The GTP is then cleaved and EF-Tu
exits the ribosome. Perhaps the newly synthesized Gal peptide activates
cleavage of EF—Tu and then remains attached to the cleaved EF-Tu, tethering
the cleaved EF-Tu to the ribosome from which the Gal peptide was
translated, and limiting its diffusion to other ribosomes. The cleaved EF-Tu
then preferentially enters the A site of the ribosome to which it is tethered
and blocks translation, perhaps because the defective EF-Tu cannot cleave
GTP and exit the ribosome (Georgiou et al., 1998). However, there are recent
indications that this textbook picture of the role of EF-Tu in translation may
be too simple and it may not be necessary to invoke tethering to explain the
air-effect. Some evidence suggests that two EF-Tu molecules rather than only
one may be involved in each amino acid addition on the ribosome (W eij land
and Parmeggianni, 1993; Ehrenberg et al., 1990). It seems possible that this
second EF-Tu molecule could be making contact with the peptidyl tRNA at
the P site. Furthermore, nonsense suppression by some mutant forms of EF-
Tu is affected by the nature of the penultimate amino acid in the polypeptide
chain, suggesting that EF-Tu may be in contact with the peptidyl tRN A at the
P site of the ribosome (Mottagui-Tabar and Isaksson, 1996).

While it seems clear that the Gal region of the major head protein of T4

binds to EF-Tu during the normal course of infection, lefi unanswered is the

83
question of why a protein would bind to EF-Tu in the first place. The

evidence suggests the G01 region of the major head protein probably binds to
EF-Tu during its synthesis on the ribosome and not afterwards when the
mature protein has been synthesized and folded into heads where it is
probably not accessible for binding. From the available evidence, it is not
clear whether the G01 region of the major head protein is exposed on the
surface of heads where it could bind to EF—Tu (cf. Black et al., 1994).
However, the ctr-inhibition of translation is much easier to explain if the G01
peptide activates cleavage of EF-Tu- durmg-the-synthesis-ef-thehead-protein.
Moreover, inhibition of translation after T4 infection of Lit-containing cells is
much more severe when the cells are infected by T4 with an amber mutation
downstream of the gal region in gene 23 than when they are infected by wild
type T4 (cf. Bergsland et al., 1990). In retrospect, we interpret this to mean
that there is more exposed Gol peptide to activate proteolysis of EF-Tu after
infection by the amber mutants when the free amber fragments accumulate,
than afier wild type T4 infection, when the newly synthesized ng3 is
immediately assembled into heads.

Possible reasons why the Gal region of the major head protein of T-even
phages may bind to EF-Tu during synthesis can be divided into three
categories. In the first category are those hypotheses in which the binding has
nothing to do with the normal function of EF-Tu in translation, for example
by playing a scaffolding role in the assembly of phage heads or a chaperone
role in the folding of the major head protein. There are precedents for
proteins being used as scatfoldings. For example, the RNA ligase of T4
bacteriophage is also used as a scaffolding in tail fiber assembly and the two

roles for this protein seem unrelated (Snopeck et al., 197 7). A chaperone role

84
for EF-Tu in head protein folding is also not far-fetched. The head protein of

T4 certainly needs chaperones to assist in its folding and T4 even encodes a
special co-chaperonin dedicated to the folding of this protein alone.
Moreover, EF-Tu has recently been shown to play a chaperone role in the
folding of some proteins (Kudlicki et al., 1997; Caldas et al., 1998). The Go]
peptide might be expected to bind to the GTPase activating region, since this
is where the cleavage by Lit occurs and the GTPase cleavage and recycling
functions of EF-Tu may be important for its chaperon activity on some
proteins, perhaps by promoting cycling on and off the protein during folding
(Kudlicki et al., 1997).

In the second category are those hypotheses based on the translational role
of EF-Tu. One attractive hypothesis within this category has the binding of
the G01 region to EF-Tu temporarily halting the translation of the major head
protein until it can be fed into the growmg phage head, in analogy to the
stoppage of translation until the signal sequences of membrane-destined
proteins enter the rough endoplasmic reticulum in eukaryotes. It has been
known for some time that the uncontrolled synthesis of major head protein, in
the absence of head assembly, leads to the formation of largely insoluble
structures. Perhaps EF-Tu with the peptide bound can enter the A site on the
ribosome, thereby blocking translation, invoking the same arguments used to
explain the air-blocking of translation when EF-Tu is cleaved by the Lit
protease. Perhaps the hypothesized blockage of translation due to Gol peptide
binding to EF-Tu is only temporary and normally reversed by other cellular
components, unlike the blockage due to the cleaved form of EF-Tu which
may be irreversible but otherwise similar. A pause in translation could be

difficult to detect if it is too transient.

85
The third general category of reasons for why the peptide binds to EF-Tu

proposes that the binding of the Gal peptide to EF-Tu affects some other,
heretofore unknown, function of EF-Tu. However, this must await the
discovery of another function for EF-Tu other than in translation and protein
folding.

Experimental Procedures

Purification of Lit protein

Experiments to determine the mechanism of activation of the Lit protease
require a source of purified Lit protein. Earlier studies used Lit protein
purified by solubilization from inclusion bodies afier overexpression from a
T7-based expression vector (Georgiou et al., 1998). The Lit protein obtained
by this method was quite pure but the purification procedure was laborious
and most of the Lit protein obtained by this method was in the form of
inactive aggregates from which active monomers had to be separated. To
facilitate the purification of Lit protein, we PCR amplified a 998 bp fragment
extending from nucleotide 262 to 1260 in the lit gene sequence. This
fragment had neither a promoter sequence nor a ribosome binding site.
Flanking BamHI sites were introduced at its ends by PCR so that it could be
cloned into the vector pET-3 0b (N ovagen) in such a way that the AUG of the
Lit protein is fused in flame to a His tag of six consecutive histidines. The
fusion protein could then be purified to near homogeneity on nickel affinity
columns (N ovagen). The fusion protein with the His tag attached was still
active in the proteolysis of EF-Tu in the presence of Gol peptide making it

86
adequate for studying the mechanism of activation of proteolysis by the

peptide.

Assay of Lit Protease

In the standard assay, the following final concentrations of 0.466 pg
Lit protein, 2.55 pg EF-Tu and 0.2 mM Gol peptide were combined in a total
volume of 45 pl. This reaction mixture was incubated for 30 min at 30°C. The
reaction was stopped by addition of sodium dodecyl sulfate (SDS) gel
loading buffer followed by boiling. The reaction products were analyzed on a
sodium dodecyl sulfate-polyacrylamide gel (SDS-PAG) containing 13%
acrylarnide. The gel was run at 60 volts for 3 hrs, fixed in 12% trichloroacetic
acid and then stained with 0.1% Coomassie blue. The depletion of the 43
kDa EF-Tu band and the appearance of a 36 kDa EF-Tu“ band was
indicative of Lit protease activity. The bands were quantified in the AMBISS

laser densitometer.

Synthesis and isolation of Go] peptide

The chemically-synthesized 29 amino acid Gol peptide was purchased
from Chiron Inc., California To synthesize the Gal peptide in viva the gal
coding sequence from the PstI site in gene 23 to the HindHI site at the Al
deletion (Bergsland, et al., 1990) was cloned between the PstI and HindIII
sites of the plasmid pET-30b (Novagen). In the presence of T7 RNA
polymerase, the resultant plasmid, pET-3 OPZI , directed the synthesis of an
approximately 12 kDa peptide containing a 53 amino acid sequence from the
major head protein gene of T4, including the 29 amino acid Gol peptide
sequence and an S-Tag and His-Tag on its N terminus. That this plasmid

87
made active Gol peptide was confirmed by the cleavage of EF-Tu in crude

extracts of cells containing the induced plasmid to which purified Lit protein
was added.

Binding Experiments on Affinity Columns

Extracts containing the S-tagged Gol peptide or the GST-EF-Tu fusion
protein were prepared from E. coli JMlO9DE3/pET3OPZl or DH5/pGEX2T-
tqu, respectively. 500 ml of cells were grown to early log phase at 37°C and
then shifted to 23°C for one hour before isopropyl-[SD-thiogalactoside
(IPTG) was added to 0.1 mM final concentration. After 3 hrs., the cells were
harvested, resuspended in 10 ml Buffer A (50 mM Tris pH 7.5, 150 mM
KCl, 5 mM MgC12) and sonicated to lyse. They were centrifuged for 30 min
at 25,000 g and the supematants were stored at -70°C. Almost all the G01
peptide and approximately 30% of the GST-EF-Tu fusion protein remained
soluble in the supernatant.

To perform the binding experiments, the extracts were mixed and
incubated at 30°C for 30 min. They were then filtered through 0.45 pm
Gelman filters and layered on a glutathione Redipack column (Pharmacia).
The column was washed with 7.5 ml of Buffer A and the GST-EF-Tu fusion
protein was eluted with reduced glutathione according to the manufacturer’s
instructions. Individual 1.5 ml fractions were assayed for the S-tag on the Gal
peptide by spotting 3 p1 on nitrocellulose and developing the filter with the S-
tag kit (Novagen). Aliquots of the fractions were boiled in SDS loading
buffer (final cone. 1% SDS, 10% glycerol plus a color indicator dye) and
applied to 10% acrylamide gels (Laemmli, 1970).

88

REFERENCES

. Snyder, L. (1995). Phage-exclusion enzymes: a bonanza of biochemical
and cell biology reagents? Mal. Microbial. 15, 415-420.

. Yarmolinsky, M. B. (1995). Programmed cell death in bacterial popula-
tions. Science 267, 836-837.

. Kao, C. & Snyder, L. (1988). The lit gene product which blocks bacte-
riophage T4 late gene expression is a membrane protein encoded by a
cryptic DNA element, e14. J. Bacterial. 170, 2056-2062.

. Yu, Y-T. N. & Snyder, L. (1994). Translation elongation factor Tu

cleaved by a phage-exclusion system. Proc. Natl. Acad. Sci. USA 91,
802-806.

. Champness, W. C. & Snyder, L. (1984). Bacteriophage T4 gal site:
sequence analysis and effects of the site on plasmid transformation. J.
Viral. 50, 555-562.

. Bergsland, K. J., Kao, C., Yu, Y-T. N., Gulati, R. & Snyder, L. (1990). A
site in the T4 bacteriophage major head protein gene that can promote the
inhibition of all translation in Escherichia coli. .1. M01. Biol. 213, 477-
494.

. Georgiou, T., Yu, Y.-T. N., Ekunwe, S., Buttner, M. J., Zuurmond, A.-
M., Kraal, B., Kleanthous, C. & Snyder, L. (1998). Specific peptide-acti-
vated proteolytic cleavage of Escherichia coli elongation factor Tu. Proc.
Natl. Acad. Sci. USA 95, 2891-2895.

. Monod, C., Repoila, F., Kutatelodze, M., Tetort, F. & Krisch, H. M.
(1997). The genome of pseudo T-even phages, a diverse group that
resembles T4. .1. Mal. Biol. 267, 237-239.

89

9. Formosa, T., Barry, J ., Alberts, B. M. & Greenblatt, J. (1991). Using
protein affinity chromatography to probe structure of protein machines.
Methods in Enzymolagy 208, 24-45.

10.Beck, B. D., Arscott, P. G. & Jacobson, A. (1978). Novel properties of
bacterial elongation factor Tu. Proc. Natl. Acad. Sci. USA 75, 1250-1254.

11.Cetin, R., Anborgh, P. H., Cool, R. H. & Parmeggiani, A. (1998).
Functional role of the noncatalytic domains of elongation factor Tu in the
interactions with ligands. Biochemistry 37, 486-495.

12.Weijland, A. & Parmeggiani, A. (1993). Toward a model for the
interaction between elongation factor Tu and the ribosome. Science 259,
1311-1314.

13.Ehrenberg, M., Rojas, A. M., Weiser, J. & Kurland, C. G. (1990). How
many EF-Tu molecules participate in aminoacyl-tRNA binding and
peptide bond formation in Escherichia coli translation? J. Mol. Biol. 211,
739-750.

14.Mottagui, S. & Isaksson, C. A. (1996). Influence of the last amino acid in
the nascent peptide on EF-Tu during decoding. Biachimie. 78, 953-958.

15.Black, C. W., Showe, M. K. & Steven, A. C. (1994). Morphogenesis of
the T4 head: In “Molecular Biology of Bacteriophage T4 ”, Karam, J. D.
(ed.) ASM Press, pp. 218-258.

16.Snopeck, T. J ., Wood, W. B., Conley, M. P., Chen, P. & Cozzarelli, N. R.
(1977). Bacteriophage T4 RNA ligase is gene 63 product, the protein that
promotes tail fiber attachment to the baseplate. Proc. Natl. Acad. Sci.
USA 54, 158-165.

17.Kudlicki, W., Coffman, A., Kramer, G. & Hardesy, B. (1997).
Renaturation of rhodanase by translational elongation factor (EF) Tu.
protein refolding by EF-Tu flexing. J. Biol. Chem. 272, 32206-32210.

 

9O

l8.Caldas, T. D., Yaagoubi, A. E. & Richarme, G. (1998). Chaperone
properties of bacterial elongation factor Tu. J. Biol. Chem. 273, 1147 8-
11482.

19.Laemmli, U. K. (1970). Cleavage of the structural proteins during the
assembly of the head of bacteriophage T4. Nature 227, 680-685.

 

91

Figure Captions

Figure l: The Go] peptide must be present at equimolar or higher
concentrations with EF-Tu for efficient cleavage of EF-Tu. In this
experiment, increasing dilutions of Go] peptide were added to pmified Lit +
EF-Tu, at a molar ratio of 1 Lit to 20 EF-Tu. The extent of cleavage of EF-
Tu after a 30 min incubation at 30°C was determined by SDS PAGE. A
Coomasie blue-stained gel is shown. EF-Tu‘ is the large cleavage fragment
of EF-Tu, extending from 60Ile to the carboxy terminus. The other cleavage
fi'agment from the N terminus to ”Gly is only about 6 kDa and ran off the
gel. Lane A: EF-Tu + Lit protease molar ratio of 20 to 1, no Gol peptide.
Lane B: same amounts of EF-Tu and Lit protease but with Gol peptide
added to a molar ratio of Go] peptide to EF-Tu of 5:1. The following lanes
have progressively lower ratios of Gol peptide to EF-Tu. Lane C: 1:1. Lane
D: 1:2. Lane E: 1:4. Lane F: 1:6. Lane G 1:10. Lane H: 1:20. Note that there
is very little cleavage of EF-Tu in the reaction in Lane H even though the
molar ratio of Go] peptide to Lit protease is still 1:1.

92

mmaclll ...:

m3...

.3

 

> m 0 U m m0 I

.60
lllmm

93

Figure 1-1: Concentration of Gol peptide needed for cleavage of EF-
Tu. This figure is a graphical representation of the experiment in Figure 1.
It shows that at as low a concentration of G01 peptide as 0.02 mM, EF-Tu is
cleaved. As the concentration of Gol peptide increases, more and more EF-
Tu is cleaved. At a concentration of 2 mM of Go] peptide, over 90% of EF-
Tu is cleaved. At this concentration, the molar ratio of GoleF-Tu is 5:1,
and of GolzLit protease is 100:1 which suggests a very tight binding
between Gol peptide and EF-Tu.

94

 

 

 

 

 

 

95

Figure 2: Can the Gal peptide activate Lit protease during a
preincubation? In this experiment, the standard assay was done with these
modifications. Gal peptide (2 mM) was incubated with 0.23 pg Lit protein
at 30°C for 30 minutes. The mixture was then diluted 1:20, added to 1.3 pg
EF-Tu and re-incubated at 30°C for 30 minutes. Lane C shows such an
experiment. Very little EF-Tu was cleaved. In contrast, substantial EF-Tu
cleavage occurred in the control experiment shown in Lane D in which 0.23
pg Lit protein was incubated alone, at 30°C for 30 minutes, diluted 1:20,
added to 1.3 pg EF-Tu + G01 peptide (2 mM) and re-incubated at 30°C for
30 minutes. Apparently, the Gal peptide did not activate the Lit protease
during preincubation, at least not irreversibly, and a molar excess of Go]
peptide to EF-Tu must still be present for efficient cleavage. Lane E was the
same as Lane C + extra Gol peptide (2 mM). More EF-Tu was cleaved in
this lane than in Lane C, presumably because of the extra Gol peptide.
Lanes A and B are standard assay controls. Lane A contains 1.3 pg EF-Tu +
0.23 pg Lit protein, while Lane B is the same as Lane A + 2 mM Gol

peptide.

mat...

 

mun—w.

:00

mm

aw

no

97

Figure 3A: Binding of Col peptide to EF-Tu. Two extracts, one from
cells in which synthesis of the G01 peptide had been induced fused to an S-
tag (S-tag Gol- peptide) and the other in which 'EF-Tu had been induced
fused to glutathione-S-transferase (GST-EF-Tu) were mixed and applied to
a glutathione column (Pharmacia). Afier washing, the GST-EF-Tu fusion
protein was eluted with glutathione and the fractions applied to an SDS-
PAGE gel. The fractions were also spotted on nitrocellulose paper and the
presence of the S-tag detected using an S protein alkaline phosphatase
conjugate (N ovagen). Some Gol peptide passed through the column but
substantial amounts were retained and eluted with the GST-EF-Tu fusion
protein. Lane A: Mixed extracts. Lane B; A flowthrough fraction. Lanes C,
D, E and F: First 4 fi'actions eluted with reduced glutathione. The
expression vector fusing GST to EF-Tu was the kind gift of Andrea
Parmeggiani.

 

- -EF-Tn

 

. . ‘3 . . . s Tag-Col Peptide

99

Figure 3b: The peptide binds to EF-Tu through its Gol Sequence. Same
as experiment in Figure 3A except that cells contain only the cloning vector
so they make the S-tag peptide without the Gal portion. Much less S-tagged
peptide is retained than in Figure 3A, even though more GST-EF-Tu fusion
protein was retained in this experiment than in the one in Figure 3A.
Apparently, the peptide is binding to GST-EF-Tu through its Gol peptide
sequence. In a similar experiment, much less S-tagged Gol peptide was
retained by a column on which only the GST portion of the fusion protein
had been retained, indicating that the G01 peptide is binding to EF-Tu
portion of the GST-EF-Tu firsion protein (result not shown).

100

A B __C%MD_ E-

 

..- . - - GST-EF-Tn

“
——

. ' . a O STag— Peptide

101

Figure 4: Model for the ctr-inhibition of translation by Lit protease.

RP stands for RNA polymerase; Tu is translation elongation factor, EF-Tu;
A is the A-site (acceptor site); P is the P-site (peptidyl site) of ribosome; and
the purple colored structure is ribosome.

 

103

Table

Table 1: The characteristics and references of bacterial strains and
plasmid constructs used in this article.

STRAIN
JM109DE3

DH5

PLASMID
pET30PZl

pGEXZT-nng

104

JM109, E. coli K carrying it DE3 with
T7 RNA polymerase under LacUVS
promoter.

supE44hstl 7recA1endA1 gyrA96
thi-lrelAl.

gal coding sequence from the PstI site in
gene 23 to the Hindm site at the Al deletion
cloned between the PstI and HindIII sites of
pET30b.

trng gene cloned into pGEXZT under too
promoter

Promega, Inc.

Low 1968;
Meselson and
Yuan 1968;
Hanahan 1983

This work
(Snyder, L.)

Parmeggiani
laboratory.