Vaccinia virus transcript release requires the vaccinia virus protein A18 and a host cell factor

MISSING IMAGE

Material Information

Title:
Vaccinia virus transcript release requires the vaccinia virus protein A18 and a host cell factor
Physical Description:
ix, 148 leaves : ill. ; 29 cm.
Language:
English
Creator:
Lackner, Cari Aspacher, 1972-
Publication Date:

Subjects

Subjects / Keywords:
Research   ( mesh )
Vaccinia virus -- genetics   ( mesh )
Vaccinia virus -- physiology   ( mesh )
Viral Proteins   ( mesh )
Transcription, Genetic   ( mesh )
Gene Expression Regulation   ( mesh )
Department of Molecular Genetics and Microbiology thesis Ph.D   ( mesh )
Dissertations, Academic -- College of Medicine -- Department of Molecular Genetics and Microbiology -- UF   ( mesh )
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph.D.)--University of Florida, 2000.
Bibliography:
Bibliography: leaves 133-147.
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Cari Aspacher Lackner.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 025683877
oclc - 53073010
System ID:
AA00022099:00001

Full Text










VACCINIA VIRUS TRANSCRIPT RELEASE REQUIRES THE VACCINIA VIRUS
PROTEIN A18 AND A HOST CELL FACTOR

















By

CARl ASPACHER LACKNER


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

UNIVERSITY OF FLORIDA


2000

























This work is dedicated to the memory of my grandfather, Joseph Vavrik.















ACKNOWLEDGMENTS

I have many people to thank for their support and contributions to this

dissertation. First, I must thank my mentor, Rich Condit, for his patience, his guidance,

and most importantly for his cheerleading. Without his encouragement this project may

have never left the ground. I want to thank my committee, Dick Moyer, Jim Resnick, and

Tom Yang for their guidance through the tough spots. I also want to thank David Price,

an excellent Outside examiner, who greatly reassured me that we were on the right page,

even if he did encourage me to proceed with the purification of the cellular factor. I owe

a huge debt to my pseudo mentor, Penni Black, who turned this "monster" over to me,

taught me about life in science, encouraged me through the really trying moments, and

always answered my stupid questions. I want to thank Jackie whose expert technical

assistance enabled me to be able to finish this project. I also want to recognize the many

members of the Condit lab, past and present, with whom I have enjoyed many memorable

experiences and who have supported me through the difficult years on this project.

Special thanks are extended to the Muzyczka lab, especially Bill McDonald, who made

everything in the lab available to me and who taught me about protein purification. I also

owe special thanks to Joyce Connors who always helped me meet the deadlines.

I must thoroughly thank my parents, Harley and Gina Aspacher, who have always

supported everything I have done and whose love and encouragement have enabled me to

achieve my goals. They raised me to believe that I was capable of doing anything I

wanted to do as long as I worked hard. I want to thank Nanny and Papa who were always









there to encourage me. Papa saw me begin this journey and I hope he is with me in spirit

as I complete it.

Finally, I must thank my husband, Dan. His love and support through the last five

years gave me the stability to stay the course. He's not only my best friend but also a

great scientific advisor.

I thank everyone who has so greatly affected my life. I am a better person

because of all of them.















TABLE OF CONTENTS

paMe

ACKNOWLEDGMENTS......................................................................1i

ABSTRACT ..................................................................................................................... viii

CHAPTERS

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

Overview of Eukaryotic and Prokaryotic Gene Expression ........................................... 1
RNA Polym erase ................................................................................................... 2
Transcription Initiation .......................................................................................... 7
Prokaryotic transcription initiation ...................................................................... 7
Eukaryotic chrom atin rem odeling ...................................................................... 8
Eukaryotic pre-initiation com plex assembly ................................................... 10
Eukaryotic initiation ........................................................................................... 11
Transcription Elongation ..................................................................................... 12
Prom oter clearance ............................................................................................. 12
Current model of the structure of the RNAP ternary complex ......................... 14
Backtracking of the ternary com plex ............................................................... 18
Elongation factors ............................................................................................ 21
Transcription Term ination ................................................................................... 28
Transcription Antiterm ination .............................................................................. 30
Vaccinia Virus Biology ............................................................................................. 31
Vaccinia Virus Early Gene Transcription ............................................................. 36
Vaccinia Virus Interm ediate Gene Transcription ................................................. 38
Vaccinia Virus Late Gene Transcription ............................................................... 40
Identification and Characterization of Vaccinia Virus Transcription Elongation and
Term ination Factors ............................................................................................... 40
The A 18 Protein .................................................................................................... 41
The G2 Protein ...................................................................................................... 42
The J3 Protein ........................................................................................................ 43
Sum m ary ....................................................................................................................... 43

2 M ATERIA LS AND M ETHODS ............................................................................... 45

Eukaryotic Cells, Viruses, and Bacterial Hosts ........................................................ 45
Plasm ids ........................................................................................................................ 45









Infected Cell Extracts for Transcription ................................................................... 47
Immobilized DNA Templates .................................................................................... 47
In Vitro Transcript Release Assay ............................................................................ 48
Induction and Preparation of Extract from E. coli ................................................... 50
His-bind Column and Phosphocellulose Column ................................................... 50
Western Blot Analysis .............................................................................................. 51
Preparation of Nuclear and Cytoplasmic Fractions of HeLa Cells ........................... 52
Chromatography and Fractionation .......................................................................... 52
Crude Fractionation of Wt or Cts23 Extract ........................................................ 52
HQ Purification ...................................................................................................... 53
Hydroxyapatite Purification ................................................................................. 53
Phosphocellulose Purification ............................................................................... 54

3 R E SU L T S ....................................................................................................................... 55

Objectives and Specific Aims ................................................................................... 55
Specific Aim 1: Develop An Assay to Determine the Biochemical Activity of Al 8,
G 2, and/or J3 ........................................................................................................... 56
Specific Aim 2: In Vitro Analysis of the A18 Phenotype ..................................... 58
Specific Aim 3: Characterization of the Cellular Factor ..................................... 59
Specific Aim 4: Characterize Al 8/CF-Dependent Release From All Vaccinia
Prom oters ................................................................................................................ 59
Specific Aim 1: Develop An Assay to Determine the Biochemical Activity of A18, G2,
and/or J3 ...................................................................................................................... 60
Formation of Paused Transcription Complexes .................................................... 60
Sarkosyl Stability of Elongation and Termination ............................................... 61
Salt Stability of Transcription Elongation Complexes ........................................ 67
In Vitro Transcription Is Specific for the Viral Promoter .................................... 70
Specific Aim 2: In Vitro Analysis of the A18 Phenotype ........................................ 74
Release Does Not Require the Presence of Al 8R during Initiation ..................... 74
Transcript Release Is Time and Concentration Dependent ................................... 77
Transcript Release Is Complemented by Crude Fractions from Wt Extract ...... 80
Release Occurs From a Stalled Elongation Complex and Can Be Complemented by
His-A18 and a Cellular Factor ............................................................................ 85
Release Requires ATP Hydrolysis ........................................................................ 89
Specific Aim 3: Characterization of the Cellular Factor .......................................... 93
Cellular Factor is not Human Factor 2 ................................................................. 93
Cellular Factor Is Present in HeLa Cell Nuclear and Cytoplasmic Fractions ..... 96
Cellular Factor Activity Is Inactivated by Heat ................................................... 99
Purification of the Cellular Factor ........................................................................ 99
Specific Aim 4: Characterize Al 8/CF-Dependent Release From All Vaccinia
Prom oters .................................................................................................................. 107
A 18-Dependent Transcript Release Occurs from All Vaccinia Promoters ............ 107
CF Enhances Release of Terminated Transcripts Initiated from an Early Promoter
................................................................................................................................. 1 10

4 DISCUSSION ............................................................................................................... 117









Transcript Release Requires A 18 and a Cellular Factor ............................................. 118
M echanistic Requirements for Transcript Release ..................................................... 119
Biochemical Characterization of the Cellular Factor .................................................. 121
Role of Al 8/CF-Dependent Release Throughout Infection ....................................... 125
Future Directions ........................................................................................................ 127
Summary ..................................................................................................................... 129

APPENDIX TABLE OF ABBREVIATIONS .............................................................. 131

REFERENCES ................................................................................................................ 133

BIOGRAPHICAL SKETCH ........................................................................................... 148















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

VACCINIA VIRUS TRANSCRIPT RELEASE REQUIRES THE VACCINIA VIRUS
PROTEIN A18 AND A HOST CELL FACTOR

By

Can Aspacher Lackner

December 2000

Chairman: Richard C. Condit, Ph.D.

Major Department: Molecular Genetics and Microbiology

Elongation and termination have proven to be important stages in the regulation

of both prokaryotic and eukaryotic transcription. Vaccinia virus is an extremely useful

model for the study of gene expression because the virus encodes much of its own

enzymatic transcription machinery. Genetic studies have identified several viral proteins

that are believed to be involved in vaccinia virus transcription elongation and

termination. The in vivo analysis of viruses containing mutations in the genes G2R and

J3R indicate that the transcripts synthesized from intermediate and late genes are 3'

truncated as compared to a wild type (Wt) infection. We hypothesize that G2 and J3

function as positive transcription elongation factors. Prior phenotypic analysis of a

vaccinia virus gene A18R mutant, Cts23, showed the synthesis of longer than Wt length

viral transcripts during the intermediate stage of infection, indicating that the A 18 protein

may act as a negative transcription elongation factor. The overall goal of the research

described here is to provide a biochemical characterization of the regulation of vaccinia









virus transcription elongation and/or termination. Pulse-labeled transcription complexes

established from intermediate viral promoters on bead-bound DNA templates were

assayed for elongation and transcript release during an elongation step that contained

nucleotides and various proteins. The addition of Wt extract during the elongation phase

resulted in release of the nascent transcript as compared to extract from Cts23- or mock-

infected cells that were unable to induce release. The lack of release following addition

of Cts23 extract suggests that A 18 is involved in the release of nascent RNA. By itself,

purified polyhistidine-tagged A18 protein (His-A18) was unable to induce release;

however, release did occur in the presence of purified His-A 18 protein plus extract from

Cts23- or mock-infected cells, suggesting that an additional factor(s) is present in

uninfected cells. These data taken together indicate that A18 is necessary but not

sufficient for release of nascent transcripts. To identify the cellular factor(s), purification

using conventional chromatography was initiated. We conclude that A18 and an as yet

unidentified cellular factor(s) are required for the in vitro release of nascent RNA from a

vaccinia virus transcription elongation complex.














CHAPTER 1
INTRODUCTION


Overview of Eukaryotic and Prokaryotic Gene Expression

The synthesis of messenger RNA is regulated at all stages from pre-initiation to

termination and 3'-end processing. This control ensures the appropriate expression of the

multitude of genes necessary during the life cycle of any organism. Prokaryotic gene

expression is regulated at the level of initiation by activators and repressors, such as in

the lac operon, and during promoter clearance, elongation, and termination. Eukaryotic

gene expression also is regulated at many levels including chromatin remodeling,

promoter activation, transcription complex assembly, promoter clearance, elongation,

termination, and mRNA processing. Recent work in the field of transcription identified

many higher order complexes that indicate an interaction among all of the factors

regulating transcription.

This introduction includes descriptions of the stages and factors involved in gene

expression of prokaryotes, yeast, and metazoans. The bacterial polymerase and its

associated factors are highly amenable to both in vitro biochemistry using purified

components and genetic analysis. The prokaryotic system is much simpler than its

eukaryotic counterparts and for this reason is subject to a high degree of study. Yeast, the

model eukaryote, is still a relatively simple system subject to genetic analysis but with a

more complex regulation than prokaryotes. The detailed knowledge ascertained from

studies with prokaryotes and yeast can be applied to metazoans that have a much more









complex biochemistry and limited genetics. Vaccinia virus synthesizes eukaryotic-like

mRNA but due to its cytoplasmic life cycle has evolved less complicated transcription

machinery. This affords us an ideal system for the study of mechanisms involved in the

regulation of transcription elongation and termination. The state of the art in

transcription is summarized here with a specific focus on the mechanisms and factors

involved.


RNA Polymerase

The RNA polymerase is the primary component of all transcription complexes on

which the other factors build. The core RNA polymerase is defined as the minimal set of

elements required for promoter-independent transcription on a DNA template in vitro.

The bacterial core enzyme is composed of two (x subunits, and a single [P and 3' subunit

(oc2pI3'). Using an assay for promoter-specific transcription, the a factor was discovered.

The a factor associates with the core RNA polymerase in the absence of promoter DNA

to form a holoenzyme complex (a2 p'a) that is now capable of promoter recognition and

transcription initiation (19,72). Several forms of a factor have been identified, although

a70 is the principal factor used by most promoters. The alternative a factors direct the

RNA polymerase to structurally distinct promoters and control genes for specialized

functions such as the heat shock response, expression of flagellar and chemotaxis genes,

and control of nitrogen metabolism (56-58,75). A high-resolution crystal structure of the

Thermus aquaticus (Taq) RNA polymerase will be discussed later in this introduction.

In eukaryotes, three DNA-dependent RNA polymerases (designated I, 1I, and III)

transcribe ribosomal genes (rRNA), protein-coding genes (mRNA), and genes coding for

tRNA and other small RNAs, respectively. This description concentrates on RNA









polymerase II (RNAPII), where most of the work on transcription has focused, although

important insights are also derived from work on RNA polymerase I (RNAPI) and RNA

polymerase III (RNAPIII) and are discussed briefly. Similar to prokaryotes, RNAPII was

first purified using promoter-less template transcription assays (136). Both yeast and

human RNAPII are composed of 12 similar subunits, among which there is extensive

structural conservation. These 12 subunits comprise the equivalent of the prokaryotic

core enzyme. The two largest subunits, Rpbl and Rpb2, are the most highly conserved

and are homologous to the P' and 3 subunits, respectively, of bacterial RNA polymerase

(Fig. 1). The Rpb3 subunit is related to the a subunit of bacterial RNA polymerase.

Although none of the RNAPII subunits are closely related to the a subunit, the general

transcription factors (GTFs) of RNAPII are the functional counterparts (111). The GTFs

are discussed in more detail below. A unique feature of the largest RNAPII subunit,

Rpbl, is a highly conserved domain consisting of 26 to 52 repeats (depending on the

species) of the consensus sequence YSPTSPS at the carboxy-terminus (CTD). The CTD

is not present in the prokaryotic 3' subunit, the related subunit of RNAPI or RNAPIII, or

the RPO147 subunit of vaccinia virus RNA polymerase. The deletion of most or all of

the CTD in yeast is lethal, demonstrating that the domain is essential in vivo. An RNAP

containing a hypophosphorylated CTD is recruited to the pre-initiation complex and at

some point during the transition from initiation to elongation the CTD becomes highly

phosphorylated. Several cellular kinases are implicated in this event including the

initiation factor TFIIH and the positive elongation factor P-TEFb. The role of these

factors will be described in more detail below.





























Fig. 1. RNAP subunit composition from vaccinia virus, E. coli, and S. cerevisiae.
The cartoon represents the separation of RNAP subunits after SDS polyacrylamide gel
electrophoresis indicating the apparent molecular size of each subunit. The size in kDa is
indicated at the left. Sequence, amino acid, and/or functional homologies between
subunits of different species are indicated by similar fill patterns. The subunits shown in
black do not have significant homology. This cartoon was adapted from Woychik and
Young (158).













kDa

200



92

69



46



30


22

14


E. Coli


Vaccinia












-


070


RP0147
RP0132











RP035
RPO30


RP022
-RPO19
"RP018

RPO7


m1


Rpbl
Rpb2







Rpb3


Rpb4
Rpb5
Rpb6

Rpb7
Rpb8
- Rpb9
Rpbl1
SRpbl0
Rpb12


Yeast


IK::









There is increasing evidence for an RNAPII "holoenzyme" as indicated by the

isolation of various multiprotein complexes interacting with core RNAPII. For

recognition and promoter-specific initiation, core RNAPII requires a set of additional

proteins known as the general transcription factors (GTFs) similar to the requirement for

a factor in prokaryotes. The GTFs include TFIID, TFIIB, TFIIE, TFIIF, and TFIIH. In

yeast and mammals, the five GTFs together comprise a total of 23 polypeptides.

Although purified RNAPII and the GTFs are sufficient for promoter-specific initiation,

they fail to respond to activators in vitro, implying the necessity of another factor. First,

several proteins designated as Srb proteins were identified during a selection for second-

site suppressors of a partial truncation of the CTD in yeast Rpb 1. Second, a multi-protein

complex from yeast termed "mediator" was identified based on its ability to mediate

activation (45). A less complex "core" Srb-mediator complex recently was purified (105)

and includes a subset of Srb proteins, a subset of mediator proteins denoted MEDs, and

several polypeptides previously identified as positive and negative effectors of

transcription (102). The exact number and composition of the polypeptides of the Srb-

mediator complex differs based on the method of purification and functional

requirements imposed. For example, in addition to the aforementioned polypeptides, a

subcomplex consisting of four Srb proteins is required for repression at some repressors

(101). Regardless of the precise composition, the function of the Srb-mediator complex

is mediated through interactions with the RNAPII CTD (102). An additional complex,

isolated using an antibody to the CTD of RNAPII, lacks Srb and mediator subunits but

includes a subset of the GTFs. It is possible that the Srb-mediator interaction with the

CTD was disrupted by the CTD antibody (141). Two attempts at purification of human









RNAPII complexes using conventional chromatography have resulted in different

complex composition as well. Analysis of polypeptides that co-elute with RNAPII

identified a complex containing chromatin-remodeling activities, including the SWI/SNF

and histone acetyltransferases CBP and PCAF, but was lacking the GTFs. This complex

was not assayed for transcription activity (25). The second purification involved

chromatography fractions that were assayed on naked DNA templates. This resulted in

the isolation of a complex lacking the chromatin-remodeling factors but including a

subset of Srb-mediator proteins and GTFs (86). These data support a model of RNA

holoenzyme that contains many subcomplexes important for mediating a number of

regulatory events. A model for transcription complex formation based on these data is

discussed below.


Transcription Initiation

For the purpose of this dissertation, descriptions of several pre-initiation events

are grouped with those involved in initiation. The mechanism of prokaryotic

transcription initiation is less complex than that of eukaryotes due to the fact that the

prokaryotic template is not packaged in nucleosomes and there are fewer protein factors

involved. Specific and accurate eukaryotic transcription initiation is more complex and

requires chromatin remodeling, promoter activation, pre-initiation complex (PIC)

assembly, and initiation.

Prokaryotic transcription initiation

Prokaryotic transcription initiation is accomplished in several steps. The core

RNAP (a23f3') first associates with the ay factor in the absence of DNA to form the RNAP

holoenzyme. The holoenzyme binds to the promoter DNA to form an RNAP-promoter









closed complex. The a70 factor binds to each of the core RNA polymerase subunits and

can only bind DNA when complexed with the core enzyme. The structure of the

polymerase suggests that the "jaw-like" clamp of the RNAP holoenzyme is capable of

clamping down on the promoter DNA to yield an RNAP-promoter intermediate complex.

The double-stranded DNA strands are melted to reveal 12 to 14 nt of open DNA

surrounding the transcription start site. The formation of this RNAP-promoter open

complex allows access to the genetic information in the template strand of the DNA

(40,95). Both repressors and activators control the regulation of transcription initiation.

Repressors function either by physically blocking the RNA polymerase or by forming a

repressosome structure. Activators stimulate transcription by direct interaction with at

least three of the subunits of the RNA polymerase (or, ;, 03'). Activators can either

increase the association of the polymerase with the promoter or stimulate the RNA

polymerase activity (148).

Eukaryotic chromatin remodeling

The coiling of DNA around a histone octamer to form a nucleosome provides a

major stage of transcriptional regulation in eukaryotes. A nucleosome consists of two

copies each of four histone proteins, H2A, H2B, H3, and H4, which interact with the

DNA to form a core particle. Each histone also possesses an amino-terminal tail that

extends outside of the core particle, a prime target of interaction for higher-ordered

coiling and gene activation. The result is a highly ordered structure that is capable of

repressing transcription. This was demonstrated by depletion of histone H4 in yeast,

which prevented the formation of intact nucleosomes and led to activation of several

promoters including PH05 (53).









How does the transcription machinery deal with the chromatin template?

Nucleosomes prevent binding of TBP, a subunit of TFIID, to the TATA promoter

element in vitro (74), and yet have only a slight inhibitory effect on the ability of a

variety of activator proteins to bind their target sites. Eukaryotic activators could

enhance the recruitment of RNAPII to its promoter by either a direct interaction of

activators with components of the polymerase or an indirect recruitment of RNAPII by

alteration of the chromatin structure. Recent work delineating the RNAP holoenzyme

revealed the presence of activator proteins that interact with proteins possessing catalytic

activities directed at the histones. As previously described, one purified human

holoenzyme complex contained SWI/SNF, a chromatin remodeling complex, and CBP, a

histone acetylase. Histone acetylation is a characteristic of transcribed chromatin (25).

Several chromatin-remodeling complexes were identified from different organisms and

include SWI!SNF from Drosophila, yeast, and humans; RSC (remodels the structure of

chromatin) from yeast; and CHRAC (ckomatin accessibility complex ) from Drosophila

(20). Recently the use of an in vitro abortive initiation assay on a chromatin template

revealed two additional human chromatin-remodeling complexes; RSF (remodeling and

spacing factor) and ACF (ATP-utilizing chromatin assembly and remodeling factor) that

promote initiation only in the presence of an activator (79,80). These two complexes are

independent yet both contain the hSNF2h protein. The larger subunits are unique to each

complex and must be responsible for specificity. The combination of activators and a

chromatin-remodeling complex may open the chromatin template near the promoter to

achieve transcription initiation. However, this does not necessitate the removal of all the

nucleosomes for a given transcribed gene. The polymerase complex must interact further









with the chromatin template as elongation continues and this is addressed in a subsequent

section.

Eukaryotic pre-initiation complex assembly

The necessary components of the promoter region were defined by mutational

analysis. The structure of eukaryotic promoters is divided into two portions: the core

promoter of approximately 50-bp adjacent to the transcription start site and a more distant

enhancer region (129,154). The core promoter elements are defined as "the minimal

DNA elements that are necessary and sufficient for accurate transcription initiation by

RNAPII in reconstituted cell-free systems" (106). The core promoter consists of a TATA

box located near position -30 to -25 and a pyrimidine-rich initiator (Inr) region located

near the transcription start site (position +1) (106). The enhancer is important for

interaction with activator proteins and can be located either upstream or downstream of

the core promoter (148).

Order of addition experiments demonstrated that purified general transcription

factors (GTFs) assemble at the promoter in a stepwise manner in vitro. Transcription

complex assembly at the promoter is initiated by TFIID via the TBP subunit binding to

the TATA element of the promoter, followed by binding of TFIIB that in turn recruits

RNAPII-TFIIF, TFIIE and TFIIH (111). This work was essential for establishing a basic

understanding of the interactions between the GTFs. However, this assay does not

necessarily mimic the in vivo situation in terms of factor ratios and preassembled

complexes.

Using an immobilized template assay and nuclear extract, which more closely

resembles the in vivo situation, two pre-initiation complex (PIC) intermediates were

isolated (122). A new model of PIC assembly is based on these data. The first step in









assembly is TBP binding to the TATA box along with TFIIA. The transcription factor

TFIIA was shown in vivo and in vitro to "encourage" the interaction between TFIID and

the promoter (30). The second intermediate is composed of RNAPII, TFIIB, the Srb4

protein (a component of the Srb-mediator complex), TFIIE, and TFIIH (122). The

identification of this intermediate suggests that the polymerase enters the PIC bound to

the GTFs (except TFIID) and to the Srb-mediator complex.

Eukaryotic initiation

The initiation of transcription is signified by formation of the first phosphodiester

bond. The functions of the yeast GTFs in initiation were studied by analysis of mutant

subunits in vitro and in vivo. The GTF TFIIB directly interacts with TBP and the DNA

sequence surrounding the TATA element to recruit the RNAPII holoenzyme complex to

the promoter (104) and also affects transcription start site selection (81). Genetic and

biochemical studies indicate that TFIIF interacts directly with TFIIB and helps stabilize

RNAPII at the promoter. The TFIIH factor has an ATP-dependent DNA helicase activity

that is required for promoter melting, and a kinase activity that was shown to

phosphorylate a number of targets including the CTD of RNAPII (151). The

phosphorylation state of the CTD is linked to the transition from initiation to elongation.

The transition from a closed to an open complex is ATP-dependent and thought to be

preceded by phosphorylation of various sites in the initiation complex including the CTD.

After promoter clearance, TBP remains at the promoter, TFIIF remains bound to the

polymerase, and the other GTFs are released. There is also an indication of an exchange

of the Srb-mediator complex for another multisubunit complex called Elongator, a step

that may be linked to the phosphorylation state of the RNAPII CTD (114).









Transcription Elongation

Elongation is the process by which RNAP catalyzes the successive

polymerization of nucleotide monophosphates into an RNA transcript based on their

complementarity to bases in a DNA template. The central component of elongation is the

ternary complex composed of the RNAP, the template DNA, and the nascent transcript.

Most of what is known about elongation derives from studies of bacterial RNAP. The

known biochemical data derived from E. coli are supported by the recent elucidation of a

high-resolution crystal structure for Taq RNAP. This is an exciting time for the

elongation field, as many hypotheses regarding conformational changes that the

polymerase undergoes to achieve elongation can begin to be supported by structural data.

This dissertation provides a summary of these data and a discussion of the stages of

transcription elongation. These stages include promoter clearance, the structure of the

RNAP ternary complex, a description of backtracking, and a summary of some of the

known elongation factors.

Promoter clearance

Entry into the stage of transcription defined as elongation requires that the

transcription complex clear the promoter. In prokaryotes, the transition from initiation to

elongation is characterized by three biochemical changes in the RNAP complex: the

RNAP undergoes the first translocation that displaces the polymerase relative to the

promoter, the a factor is released, and the ternary complex becomes tightly associated

with both the DNA template and the nascent RNA resulting in a very stable complex

(155). The polymerase also undergoes a process defined as abortive initiation that

generates a set of nested RNA transcripts that are less than 12-nt in length. This process

may reflect multiple attempts of the RNAP to direct the 5'-end of the growing RNA chain









to the RNA binding site of the polymerase. The placement of the RNA 5'-end is not

understood but it is recognized as key to rendering the complex fully stable. In E. coli

this process coincides with release of the a subunit after the synthesis of between 4 and

10-nt and may reflect a conformational change of the RNAP to an elongation competent

form. The conformational change on transition from initiation to elongation is supported

by a nearly two fold decrease in the footprint size of the polymerase (107).

Eukaryotes have analogous reactions for promoter clearance. The transition from

initiation to elongation requires breaking the initial ties with the promoter, the GTFs, and

the accessory factors, as well as conversion of RNAP to an elongation competent form.

Promoter clearance is also plagued by abortive initiation and arrest. The RNAPII

complexes containing transcripts less than 9-nt in length are unstable and likely to abort

(59,66). Dissociation of the GTFs occurs after the synthesis of 10- to 15-nt of RNA. The

transition from initiation to elongation is accompanied by an increase in the stability of

the ternary complex. This stability is necessary for the processivity of the RNAP based

on the fact that if the polymerase releases the nascent RNA prior to complete synthesis it

is unable to rebind and continue transcription. The synthesis of new RNA must begin

with reinitiation at the promoter. Most elongation complexes are stable in high

concentrations of salt, survive purification by gel filtration or precipitation with an

antibody, and can be stored in the absence of nucleotides at 40C for days without

significant loss of activity (155).

Many proteins are implicated in the regulation of eukaryotic early elongation.

The general transcription factor TFIIF is required for initiation and stimulation of

elongation and may act to decrease abortive initiation by increasing the rate of nucleotide









addition (167). As previously described, the transition from initiation to elongation is

accompanied by opening of the DNA template and phosphorylation of the CTD of

RNAPII, events that could be accomplished by the TFIIH ATPase, helicase, and kinase

activities. Additional positive and negative transcription elongation factors (P-TEFs and

N-TEFs) are postulated to regulate promoter clearance in a DRB-sensitive manner

(26,89). Several of these factors were identified including P-TEFb, NELF, DSIF, and

Factor 2. P-TEFb is the positively acting factor that is composed of Cdk9 and cyclin TI.

Although the kinase activity of P-TEFb could have more than one target for

phosphorylation, evidence suggests that the CTD of RNAPII is a physiologically

important target (88). The phosphorylation of the CTD by P-TEFb is required to prevent

arrest by the elongating RNAPI. The NELF and DSIF factors interact with an RNAPII

containing a hypophosphorylated CTD to negatively regulate elongation (166). Factor 2

is responsible for the release of short transcripts from early elongation complexes (165).

These negative activities may be overcome by the positive action of P-TEFb (156,166).

Current model of the structure of the RNAP ternary complex

Analysis using X-ray crystallography has revealed the structure of Thermus

aquaticus (Taq) RNA polymerase (RNAP) at 3.3 A resolution (169). The Taq RNAP is

similar in size and shape and has a high degree of sequence similarity to the E. coli

RNAP defined by low-resolution electron crystallography (36,37). This allows for the

comparison of the biochemical data elucidated from work with E. coli and the structural

data from Taq to construct a structure-function model of the transcription complex. The

RNAP has a "crab-claw-like" shape where the "jaws" are separated by several channels

leading to the Mg 2 active site (Fig. 2). It is suggested that the "jaws" close around the

downstream duplex DNA in the transition from initiation to elongation. One channel









encloses the double-stranded DNA downstream of the transcription bubble, while a

second channel accommodates the upstream DNA resulting in a 900 bend of the DNA.

An upstream element called the "rudder" protrudes from the floor of the active site and is

positioned such that it could separate the DNA template strand and the RNA transcript

thus allowing the two strands of DNA to reanneal. Crosslinking studies predict another

channel for the passage of the RNA transcript opposite the DNA. An additional channel

called the secondary channel is postulated to recruit nucleotides and may be blocked by

the RNA 3'-end when in the backtracked conformation (169).

Three sites in the elongation complex were characterized functionally and

structurally: the double-stranded DNA binding site, the RNA-DNA heteroduplex binding

site, and the single-stranded RNA binding site (Fig. 2) (107). The DNA binding site was

mapped to 9-bp of double-stranded DNA just downstream of the 18-nt transcription

bubble (108). The RNA-DNA hybrid is composed of 8-bp as determined by chemical

footprinting and RNA-DNA chemical cross-linking (109). The heteroduplex-binding site

is a region of weak ionic interactions between the protein and the first six basepairs of the

RNA-DNA hybrid. The RNA binding site was defined using photoreactivated RNA

probes that showed tight RNA contacts with the RNAP and nine nucleotides of RNA

spanning from -8 to -17 next to the hybrid-binding site. Footprint analysis shows

protection of 40- to 50-bp of DNA (124) and 18-nt of RNA (49) in the transcription

elongation complex.























Fig. 2. Model of the paused transcription elongation complex. The cartoon represents the structure-function model of the
prokaryotic polymerase based on structural data from Taq and biochemical data from E. coli. The cartoon is adapted from the meeting
notes from 'Post-initiation Activities of RNA Polymerase', Fall 2000 meeting, Mountain Lake, Virginia and Mooney, Artsimovitch,
and Landick (95).









Transcription bubble 18nt 9bp DNA binding site
RNA:DNA hybrid 1 8bp







.Flap AUGUGUGC -- -_......-NTP


Active site channel
RNA binding site









Backtracking of the ternary complex

The study of the mechanistic aspects of transcription elongation is confined to

work in prokaryotes due to the simplicity of the RNAP enzyme and the difficulty of the

elongation assays. There are three blocks to transcription elongation at which factors

theoretically can act to effect elongation: pause, arrest, and termination. Transcription

pause and arrest can be a result of intrinsic signals (interactions between the RNAP and

sequence in the RNA and DNA), a response to DNA binding proteins that physically

block progression of the RNAP, or a response to artificial conditions such as the absence

of one of the four nucleotides. Pausing is a temporary delay in RNA chain elongation

and is a precursor to arrest (complete halting without dissociation) and dissociation of the

ternary complex at p-independent and p-dependent terminators (126). However, not all

pauses are termination precursors (155).

The relationship between translocation of the RNAP and synthesis of each

phosphodiester bond of the RNA transcript is a subject of great debate. Several models

were proposed over the past seven years including the classical and revisionist models.

In the classical model, the RNAP moves along the template monotonically, i.e., the

RNAP moves synchronously with the addition of each nucleotide (46). In the revisionist

model, or the inchworm model, the RNAP is proposed to move in a two-step cycle. The

RNA is synthesized while the RNAP is in a static position, and movement of the RNAP

occurs in short bursts, or jumps, which allows the DNA and RNA to be threaded through

the enzyme (21). The inchworm model was based on three lines of experimental

evidence: irregular DNA footprints of elongation complexes halted at successive sites









(73), formation of arrested transcription complexes (3), and cleavage of internal RNA

from defined complexes resulting in the loss of 3' RNA fragments (150).

Recent data indicate that the inchworm model of contraction and expansion of the

RNAP was a misinterpretation. Footprinting experiments have now demonstrated that a

stalled E. coli RNAP translocates backwards relative to the catalytic site and that the

translocation can be suppressed by hybridization of oligonucleotides upstream of the

RNAP (69). This activity is described as backtracking, or the lateral oscillation of the

RNAP ternary complex. Backtracking also was demonstrated using nucleotide analogs

that either strengthen or weaken the RNA-DNA hybrid (109). Pause, arrest, and

termination signals all appear to slow RNAP due to unstable basepairing in the hybrid

that displaces the RNA 3'-end from the active site of the RNAP. This instability may

allow the backward sliding (backtracking) of RNAP into a more stable complex. The

backtracking model may provide an explanation for transcriptional fidelity and control of

the rate of transcription elongation.

Recent work from Landick and co-workers has demonstrated that the elongating

RNAP can adopt open and closed conformations that dictate slow and fast elongation by

the RNAP (39). These conformational changes are suggested by the structure of the

RNAP, kinetic studies of RNAP, and formation of a pause RNA hairpin in prokaryotes.

The flexibility and structure of the RNAP suggests that the jaws of the RNAP may close

around the DNA, locking the downstream double-stranded DNA within the DNA binding

site. In addition, a flap of the RNAP is predicted to close over the exiting transcript thus

creating the RNA binding site (169). These two modifications of the polymerase create a

closed conformation that is capable of rapid elongation. This idea is supported by the









kinetics of elongation observed on single RNAP molecules that reveal dynamics that are

averaged out in bulk RNAP experiments. The kinetics suggest both a fast and slow state

of incorporation of nucleotides (39).

RNA hairpins that form as the transcript emerges from the ternary complex are

integral parts of some pause signals and are required at p-independent terminators where

they induce dissociation of the ternary complex. RNA hairpins can also function as

antiterminators which alter the ternary complex to block recognition of both pause and

termination sites. The role of hairpin formation in termination and antitermination is

discussed in more detail below. Mutational analysis of the pause hairpin indicates that

the structure and not the specific sequence within the hairpin are important for pausing

(23). It should be noted that hairpin formation is not sufficient to signal a pause, and not

all pauses contain stable RNA hairpins (82). DNA and RNA sequences between the base

of the hairpin and the RNAP active site also affect pausing and this distance may, in part,

distinguish p-independent termination from transcription pause sites (23). The use of

crosslinking agents demonstrated that the loop region of an RNA hairpin makes contacts

with the RNAP flap, the same structure that is predicted to close over the exiting RNA

(Fig. 2). The interaction with the RNA hairpin may open the grasp of the RNAP flap on

the exiting RNA. Based on the structure of the RNAP, the flap is linked to the 3' subunit

which forms the base of the channel contacting the RNA-DNA hybrid. This link may

explain how formation of an RNA hairpin can lead to disruption of the catalytic activity

of the RNAP. These data together support the model of fast and slow elongation

characterized by closed and open conformations, respectively, of the RNAP (5).









The Landick model also suggests that at every template position the RNAP can

fluctuate between normal elongation and a state susceptible to pausing, arrest, or

termination. The position of the RNA 3'-end may vary between several different

positions including backtracked (RNA extending downstream of the active site), frayed

(RNA 3'-end is separated from the template DNA strand), pretranslocated (RNA blocking

the nucleotide binding site), active (RNA primed for nucleotide addition), and

hypertranslocated (RNA 3'-end pulled out of the active site) (Fig. 3) (5). The RNAP

switches between the active and pretranslocated conformations during rapid elongation

and may engage in the other conformations at pause and arrest sites (5). Rescue from

these conformations may be spontaneous or the result of regulation by specialized

proteins. There are two highly studied types of pause signals that are characterized by the

structure of the paused complex. These signals are designated as class I and class II

pauses (5). A class I pause site is characterized by the interaction between an RNA

hairpin and the RNAP but is also dependent on the 11-nt distance between the base of the

hairpin and the 3'-end. These pauses are found in the leader regions of several bacterial

amino acid biosynthetic operons. The interaction between the RNA hairpin and the

RNAP induces the RNA 3'-end to adopt the frayed or hypertranslocated position (Fig. 3)

(4,24). A class II pause is characterized by a weak RNA-DNA hybrid that induces

backtracking of the RNAP (Fig. 3). These pause sites have been characterized in vitro at

arrest or termination sites and in the early transcribed region of E. coli to recruit the

antitermination factor RfaH.

Elongation factors

The factors that regulate transcription elongation can be divided into at least three

functional classes based on their ability to prevent arrest of RNAP, to regulate the rate of























Fig. 3. Position of the RNA 3' end at various positions. The cartoon represents the possible positions of the RNA 3' end (UOH)
during active elongation and pausing. The active site of the RNAP is represented by the circles. The template DNA is indicated as a
black line and the RNA as a gray line. This cartoon was adapted from Artsimovitch and Landick (5).










Pretranslocated Active


=3 U1-


Frayed


UOH


Backtracked


Hypertranslocated


0









RNAP through chromosomal templates, and to increase the catalytic rate through

suppression of pausing. Factors that prevent arrest of the RNAP include the prokaryotic

factors GreA and GreB, and the eukaryotic factors P-TEFb and SII (Table 1). The

prokaryotic Gre factors and eukaryotic factor SII share functional but not sequence or

structural similarities. These factors interact with their respective RNAPs to activate the

endoribonucleolytic cleavage activity intrinsic to the polymerase at sites of DNA-specific

arrest. Induction of the cleavage activity in a backtracked ternary complex (Fig. 3)

removes the unpaired Y-end of the nascent transcript and repositions the RNA in the

polymerase active site for continued elongation (13,90,110). Unlike GreA, GreB and SlI

can act on a ternary complex that has arrested in the absence of the factor. GreA, on the

other hand, must be associated with the ternary complex prior to arrest in order to activate

the cleavage activity.

In eukaryotes, the production of full-length runoff transcripts in vitro and

functional mRNA in vivo is sensitive to the drug 5,6-dichloro-l-beta-D-

ribofuranosylbenzimidazole (DRB). The Drosophila and human factor P-TEFb rescues

DRB-sensitive arrest in early eukaryotic elongation complexes. It is composed of two

subunits, a cyclin (TI, T2a, or T2b) and a cyclin-dependent kinase (Cdk9) (118,119).

The P-TEFb factor can phosphorylate the CTD of RNAPII, an action that is thought to

regulate the transition from abortive initiation to elongation. The action of P-TEFb also

is postulated to counteract the actions of N-TEFs, negative transcription elongation

factors, including DSIF and NELF (156,166).

Several eukaryotic factors were identified based on their ability to promote

transcription through chromatin templates. An ATP-dependent activity found in fractions









Table 1: Select Prokaryotic and Eukaryotic Transcription Elongation and Termination
Factors
Name System Function Properties
DSIF Eukaryotes, Induce arrest Binds RNAPII in vitro, works with NELF to
yeast arrest RNAPII, activity is counteracted by P-
TEFb
ELL Eukaryotes Stimulate Inhibits transcription initiation by competing
elongation for RNAPII, suppress transient pausing,
suppress backtracking?, implicated in
oncogenesis
Elongin Eukaryotes Stimulate Implicated in oncogenesis, suppress
elongation backtracking?
FACT Eukaryotes, Histone Stimulates transcription through chromatin
yeast chaperone
Factor2 Eukaryotes Transcript DNA-dependent ATPase
release factor
GreA Prokaryotes Prevent arrest Cleavage stimulatory factor, interacts with
RNAP P subunit, must be associated with
complex prior to arrest
GreB Prokaryotes Prevent arrest Cleavage stimulatory factor, 35% aa identity
with GreA, acts on arrested complex
N Prokaryotes Anti- RNA binding protein, binds RNAP, induces
termination read through of termination factor dependent
and independent sites on X genome
NELF Eukaryotes Induce arrest Works with DSIF to arrest RNAPII, activity
is counteracted by P-TEFb
NusG Prokaryotes Accelerates Interacts with core RNAP, stimulates escape
elongation from class II pause sites, inhibits
backtracking?
P-TEFb Eukaryotes Prevent arrest Stimulates the production of long transcripts,
ATP-dependent, phosphorylates RNAPII
CTD, counteracts N-TEFs
PTRF Eukaryotes RNAPI Induces blocked murine RNAPI to terminate
termination
Q Prokaryotes Anti- DNA binding protein, promotes read through
termination of termination signals
Reblp Yeast RNAPI DNA binding protein, blocks elongating
termination yeast RNAPI
Rho Prokaryotes Termination RNA binding protein, ATPase, RNA-DNA
helicase
SII Eukaryotes, Prevent arrest Stimulates NTP incorporation, binds RNAPII
(TFIIS) yeast in vitro, can act on arrested complex
TFIIF Eukaryotes, Increase Promotes read through of some blocks to
yeast elongation elongation, GTF
TI'F-l Eukaryotes RNAPI Site specific DNA binding protein, blocks
I I termination RNAPI transcription









containing SWI/SNF promotes elongation downstream of the Drosophila hsp7O promoter

by remodeling nucleosomes downstream of the promoter (14). The GTFs and purified

human RNAPII can form preinitiation complexes and initiate transcription on a

promoter-proximal chromatin-remodeled template but cannot undergo productive

transcription elongation. A novel factor termed FACT (facilitates chromatin

transcription) then was purified and identified as a heterodimeric complex composed of

the human homolog of the S. cerevisiae Spt 16 and the HMG- 1-like protein SSRP 1. The

FACT is not ATP-dependent and is thought to function as a histone chaperone (112,113).

Several factors were identified for their ability to increase the catalytic rate and

processivity of the RNAP. NusG is a factor from E.coli that stimulates escape from

pausing at class II (hairpin-less) pause sites by a mechanism that inhibits backtracking

(5). The eukaryotic factors that increase the rate of elongation include TFIIF, Elongin,

and ELL. The general transcription factor TFIIF is not only required for transcription

initiation but remains associated with the polymerase for stimulation of elongation and

read through of some blocks to elongation (67,153). TFIIF is phosphorylated by TFIIH

and P-TEFb, although the functional significance of this phosphorylation is not known

(43,120). In addition, TFIIF also was shown to partially inhibit Factor 2, a termination

factor important for the release of early elongation complexes (117).

Elongin and ELL are both implicated in oncogenesis. Elongin is a heterotrimeric

complex of A, B, and C subunits. Elongin A is the catalytic subunit and capable of in

vitro elongation stimulatory activity which is stimulated by association with Elongin B

and C (6,47,152). The Elongin BC complex also can interact with the product of the von

Hippel-Lindau (VHL) tumor suppressor gene (68). Mutation of the VHL gene









predisposes affected individuals to a variety of cancers including clear-cell renal

carcinoma, multiple endocrine neoplasias, and renal hemangiomas (77). A vast number

of naturally occurring VHL mutations show reduced binding to the Elongin BC complex

(68). It was thought initially that the binding of Elongin BC to VHL and Elongin A

represented two independent and mutually exclusive events. However, recent data

indicate that Elongin BC is 100- to 1000-fold more abundant than Elongin A and VHL in

cell extract (31).

The product of the human ELL gene stimulates the rate of elongation by RNAPII

by suppressing transient pausing of the polymerase at many sites along the DNA

template. This stimulation occurs using purified core RNAPII on a promoter-less

template, indicating that stimulation occurs through interactions with RNAPII, the

template DNA, or the nascent transcript (63,143). The ELL factor also is capable of

inhibiting transcription initiation by binding RNAPII and by preventing its entry into the

preinitiation complex (143). Acute myeloid leukemia is associated with translocations of

the human ELL gene and the MLL gene. It is unknown how this fusion protein results in

acute myeloid leukemia. The ELL factor recently was purified as a complex with three

other proteins which was termed the "Holo-ELL" complex (142). Unlike ELL, however,

Holo-ELL does not negatively regulate the polymerase in transcription initiation. A

model was proposed where one of the associated proteins in the Holo-ELL complex

regulates the transcription inhibitory activity of ELL and deletion of this domain (such as

in the MLL-ELL translocation) overrides this regulation (142).

The current model of transcription elongation proposes that the protein-DNA

contacts downstream of the RNAP are responsible for the stability of the elongation









complex. The model also proposes that closing of the RNAP around the DNA locks the

enzyme into a transcriptionally processive conformation (140). This complex responds to

both DNA and RNA sequences that may open the transcription complex and slow the rate

of nucleotide addition. The role of some elongation factors may be to stabilize the

polymerase in the closed conformation, creating a pause- or arrest-resistant enzyme that

is therefore more processive.


Transcription Termination

Termination signals cause the release of RNA and DNA, as well as the RNAP,

and can be regulated both positively and negatively. There are three types of prokaryotic

transcription termination signals: intrinsic or p-independent terminators, p-dependent

terminators, and persistent RNA-DNA hybrid terminators. Intrinsic terminators require

stable RNA hairpin formation followed by 7- to 9-nt of U residues and are independent of

extrinsic factors. Pause and termination hairpins probably affect the transcription

complex in different ways. Neither the his nor the trp operon pause RNA hairpin extends

to within 10-nt of the transcript 3'-end. The p-independent terminator hairpin reaches to

within 7- to 9-nt of the RNA 3'-end. This suggests that a termination hairpin may

destabilize the ternary complex by disruption of key contacts in the complex that are

unaffected by pause hairpin formation, perhaps within the RNA-DNA hybrid (22). The

model of intrinsic termination proposes that the U-rich sequence induces a pause that

allows time for the formation of the termination hairpin. The hairpin interacts with the

RNAP inducing the open, less stable conformation and ultimately results in release of the

nascent transcript (155). Rho-dependent termination requires Rho, an RNA-binding

protein that possesses both ATPase and RNA-DNA helicase activities. Rho loads onto









the nascent transcript in an unstructured region of the RNA upstream of the terminator

and translocates along the nascent RNA. When Rho "catches up with" the paused

transcription complex at a termination site, it induces the release of the RNA polymerase

and the nascent transcript by destabilizing the RNA-DNA hybrid (11,102,163).

In prokaryotes, most steady state transcript 3'-ends are formed by genuine

termination events. In eukaryotes, on the other hand, almost all of the steady state

RNAPII transcript 3'-ends are generated by processing of the primary transcript and not

by termination of transcription. The processing events are a result of cleavage and

polyadenylation sequences within the nascent RNA. Termination of the ternary complex,

however, occurs downstream by an unknown mechanism generating heterogeneous

transcripts. The transcripts for several RNAPII genes were analyzed by nuclear run-off

analysis. The length of these transcripts was shown to range from 100- to 4000-nt

downstream of the polyadenylation site (29,51,121). The stability of the elongation

complex as demonstrated by its resistance to challenge with sarkosyl or high

concentrations of salt, provides a significant question as to the mechanism of termination.

There must be some extrinsic signal to induce the extremely stable ternary elongation

complex to cease RNA synthesis and terminate. The mechanism of eukaryotic

transcription termination may parallel post-replicative transcription in vaccinia virus.

Vaccinia transcripts are heterogeneous in length and are not generated by cleavage and

polyadenylation. A further discussion of vaccinia virus transcription is found below.

Several transcription termination factors have been identified in eukaryotes.

Termination by RNAPI uses a two-component system. One protein binds the DNA at a

specific sequence and serves as a block to the elongating polymerase, while the other









protein dissociates the stalled complex. In mice, TTF-l (Transcription Termination

Factor for Pol I) blocks the elongating RNAP and PTRF (_pol I Transcript Release Factor)

is responsible for termination of the complex (62,91,123). In yeast, Reblp is a DNA

binding protein that blocks RNAPI. An unidentified element is responsible for induction

of termination (123). A second mechanism of eukaryotic termination is proposed to be a

result of ATP hydrolysis by DNA and/or RNA polymerase binding proteins. Factor 2,

identified in Drosophila and humans, is a DNA-dependent ATPase that acts on early

RNAPII complexes to induce transcription termination. This activity is counterbalanced

by positive acting factors such as P-TEFb (84,165). Transcription termination of vaccinia

virus early genes is accomplished by two viral factors, VTF/CE and NPH-I. The early

termination signal may be recognized by VTF/CE and termination induced by the DNA-

dependent ATPase NPH-I (41,42).


Transcription Antitermination

Lambda phage is most extensively studied for its regulation of elongation and

termination. Work by Jeff Roberts on X phage demonstrated the first example of anti-

termination in the positive control of transcription elongation (127). Two proteins are

involved in X anti-termination, XN and XQ, and they function in different ways.

Synthesis of a hairpin structure in the nascent transcript recruits N protein. The binding

of N to the RNA is stabilized by the assembly of the Nus proteins from E. coli. The

binding of N to the RNA 5'-end is transmitted to the ternary complex through interaction

of N with the RNAP. The N protein becomes a stable polymerase subunit during

transcription and the nascent transcript loops. The result is stimulation of transcription

elongation and transcription through pause and arrest sites (38,155). The Q protein, on









the other hand, is a DNA binding protein that interacts with RNAP paused at a specific

site near the promoter. This results in a transcription complex that can read through

downstream termination signals. Both N and Q regulated read through allows the phage

to complete the lytic portion of its life cycle.


Vaccinia Virus Biology

Vaccinia virus historically served as a superb model for transcription. The

prototypic Orthopoxvirus has a linear double-stranded DNA genome of 192,000 base

pairs, which it replicates in the cytoplasm of the infected host cell. Because of the

cytoplasmic site of infection, the virus encodes most of the enzymatic machinery

necessary for both viral RNA and DNA metabolism. Many of the viral-encoded enzymes

have structural and functional similarities to the host cell enzymes. The viral RNA

polymerase is eukaryotic-like, composed of 2 large subunits (RPO147 and RPO132) with

approximately 30% identity to the Rpbl and Rpb2 subunits of S. cerevisiae and 6 small

subunits, one of which (RPO7) is homologous to Rpbl2 (Fig. 1) (98). In addition, the

viral RPO30 subunit shares 23% amino acid identity and is structurally homologous to

mouse SII, the mammalian transcription elongation and cleavage stimulatory factor (1).

During infection, viral genes are expressed in a transcriptional cascade encompassing

three stages as follows: early, intermediate, and late. Each stage requires trans-acting

factors for transcription initiation that are synthesized in the previous stage thus providing

the basis for sequential regulation. Biochemical and biological experiments during the

past few years showed that elongation and termination of all three transcriptional stages

are also regulated events.









The vaccinia virion is composed of a biconcave core containing the viral genome

surrounded by a lipid bilayer. Infection of the host cell involves membrane fusion and

internalization of the core (Fig. 4). The transcriptional cascade commences with early

mRNA synthesis using the enzymes and factors present within the core. The early

mRNA is extruded into the cytoplasm and translated on host cell polysomes. These early

mRNAs encode the proteins required for DNA replication, the RNA polymerase, and the

intermediate transcription factors. Early mRNA synthesis is followed by DNA

replication. Intermediate gene transcription then generates the transcription factors

necessary for late mRNA synthesis. Late proteins include the early transcription factors

to be packaged within the virion for the next round of infection, as well as the structural

proteins (98).

Each of the three gene classes is regulated by its own set of cis-acting elements.

This is the framework for regulating the timing of gene expression. The specific and

distinct critical sequences required for initiation of stage-specific transcription are

essential for recognition by different trans-acting factors (99). These factors are

discussed below. Early, intermediate, and late-stage promoter sequences are each

approximately 30-bp in length, divided into core and initiator regions that were defined

by saturation mutagenesis. The initiator region includes the site of transcription

initiation, designated as +1. The core region is approximately 15-bp in length and is

located from -15 to -30. The intervening DNA, between the core and initiator regions, is

defined as the spacer region and is insensitive to mutation. The stringency of the critical























Fig. 4. Vaccinia virus life cycle. Vaccinia virus enters the host cell and undergoes early transcription, followed by DNA replication,
intermediate transcription, and late transcription. Virions are assembled, packaged, and released for the next round of infection. This
figure was a generous gift from Richard C. Condit.








IMV membrane
lateral body

core
IMV :









sequence regions differs among the three classes. Early promoters contain a conserved

critical core region, from nucleotides -13 to -27, and a less stringent initiator region

where the only requirement is a purine at +1. The intermediate promoter core region

resembles that of early promoters in A+T-richness but differs in specific sequence. The

core region spans from nucleotides -14 to -24. The initiator region differs from early

promoters and is defined by a tetranucleotide sequence (TAAA) that is more similar to

late initiator regions. Late promoters have a less stringent A+T-rich core region spanning

from nucleotides -10 to -15 separated by a 6-bp spacer region from the TAAAT initiator

(96,97). Vaccinia virus transcription apparently does not require enhancer elements, the

sequences found upstream of eukaryotic promoters that are important for activated

transcription initiation. Most viruses do not respond to environmental signals and

therefore do not evolve an unnecessary level of sophisticated regulation. The vaccinia

genome also differs from eukaryotic genomes by the absence of chromatin packaging

although there may be viral proteins bound to the DNA.

In addition to the differences in promoter sequence, the three gene classes also

differ in the formation of the transcript 5'- and 3'-ends. Early mRNAs usually contain a

short 5'-untranslated region that is capped, but otherwise unmodified. Termination

occurs downstream of a sequence specific termination signal producing early mRNA of

discrete length that is 3'-polyadenylated (98). Intermediate and late mRNAs initiate

within the AAA element of their core promoters but the resulting RNAs contain

additional 5'-A residues incorporated by slippage of the polymerase. This results in a

"poly(A) head" that is 30- to 50-nt in length and capped (2,116,139). At intermediate and

late times during infection the RNA polymerase does not recognize the early termination









signal and synthesizes 3'-heterogeneous transcripts that are polyadenylated (98). This

implies that the mechanism for post-replicative gene 3'-end formation is different from

termination of early genes.


Vaccinia Virus Early Gene Transcription

Early transcription differs from the post-replicative stages of transcription in that

it occurs mainly in the virion, as opposed to the infected cell cytoplasm. An in vitro early

transcription system was developed using purified viral transcription factors isolated from

the virion (48,98). Early gene transcription requires the vaccinia early transcription

factor (VETF), the RNA polymerase-associated protein RAP94, (the product of the H4L

gene), and the viral RNA polymerase for promoter-specific initiation (Table 2) (16,170).

VETF binds the core region of the early promoter, as well as DNA downstream of the

RNA start site, and alters the conformation of the DNA template. The DNA-dependent

ATPase of VETF is not required for promoter binding but is essential for transcription.

The ATPase activity may be a requirement for promoter clearance (15,18,83). It is

important to emphasize the requirement for a vaccinia RNA polymerase that contains the

RAP94 subunit. There are clearly two forms of the RNAP present in infected cells. The

RNAP-RAP94 complex is necessary for initiation at early vaccinia promoters and may

allow the polymerase to carry a "memory" of the class of the initiating promoter.

Intermediate and late promoters recruit RNAP molecules lacking RAP94. The

importance of this subunit will be more apparent in the discussion of transcription

termination of the three classes of transcripts. A stable ternary complex is formed after

the synthesis of a 7- to 9-nt transcript similar to the prokaryotic polymerase complex.

The 5'-cap is synthesized by the time the nascent RNA is 31-nt long, although









Table 2: Vaccinia Virus Transcription Factors
Common Name Vaccmia Transcription Properties/Activity
Gene Stage
RNA Pol All Multisubunit RNA polymerase
RPO 147 J6R -Homologous to Rpb I
RPO 132 A24R -Homologous to Rpb2
RP035 A29L
RPO30/VITF-I E4L -Intermediate initiation factor,
homologous to Euk. TFIIS
RP022 J4R
RPO19 A5R
RPO18 D7R
RPO7 G5.5R -homologous to Rpb 12?
RAP94 H4L Early Early promoter specificity factor
VETF A7L Early Early promoter binding, DNA-
D6R dependent ATPase, Early initiation
factor
CENTF DIR All/Early Early, intermediate, and late capping
D1 2L enzyme, Early termination factor,
Intermediate initiation factor
NPH-I Dl lL Early DNA-dependent ATPase, RNA
helicase, Early termination factor
VITF-2 Cellular Intermediate Intermediate initiation factor
YY1 Cellular Intermediate Binds intermediate promoters
VLTF-1 G8R Late Late initiation factor
VLTF-2 AlL Late Late initiation factor
VLTF-3 A2L Late Late initiation factor
VLTF-4 H5R Late Late transactivator
VLTF-X Cellular Late Late initiation factor
A18R All DNA helicase, DNA-dependent
ATPase, Early, intermediate, and late
transcript release factor
G2R Intermediate Intermediate and late elongation factor
and Late
Poly(A) Pol All Poly(A) polymerase (PAP)
J3R -PAP stimulatory subunit, 2'0-
methyl transferase, Intermediate
and late elongation factor
El L -PAP catalytic subunit









stable association of the capping enzyme with the complex does not occur until the

nascent transcript is 51-nt in length (52).

Early gene mRNA 3'-ends are formed by termination and not endonucleolytic

cleavage (130,145). The newest model for early termination evolved from recently

published data demonstrating a protein-protein interaction between RAP94 and NPH-1

(94). Since only RNA polymerase containing RAP94 is capable of initiating

transcription from early promoters, the specificity of the early transcription termination

system may be explained by the physical interaction between RAP94 and NPH-I. The

interaction suggests that RAP94 functions as a transcription termination cofactor,

recruiting NPH-I to the transcription complex (94). In the absence of RAP94, as in

intermediate and late transcription complexes, NPH-I is not recruited to the ternary

complex and recognition of the termination signal does not occur. NPH-I requires single-

stranded DNA to activate its ATPase activity and the ATPase activity is necessary for

termination. The most obvious source of single-stranded DNA in the transcription

complex is the nontemplate strand in the transcription bubble. The model proposes that

the vaccinia termination factor (CEIVTF) is poised to scan the RNA for the termination

signal, UUUUUNU. Recognition of the U5NU signal by CE!VTF may induce

conformational changes to make the single-stranded DNA available to NPH-I. The

activation of the ATPase activity of NPH-I results in termination and release of the

nascent transcript 20- to 50-nt downstream from the termination signal (28,42,94).


Vaccinia Virus Intermediate Gene Transcription

The intermediate stage of vaccinia gene transcription can be reconstituted in vitro

by the use of hydroxyurea-treated infected cell extracts. Several proteins were shown to









be required for intermediate transcription initiation although initiation has not been

reconstituted from purified factors. These include the RNA polymerase (-RAP94),

capping enzyme (CENTF), VITF-1 (E4L/RPO30), an unidentified cellular factor found

in the nucleus of uninfected HeLa cells and distributed between the cytoplasm and the

nucleus of infected cells (VITF-2), and a two-subunit enzyme, VITF-3, composed of the

protein products from open reading frames A8R and A23R (Table 2) (1,131-134). The

use of a cellular factor, VITF-2, for intermediate initiation may be a regulatory

mechanism between the early and post-replicative stages of the virus life cycle by

indicating whether a cell has been activated for optimal replication (98). It is

hypothesized that VITF-3 also could be responsible for regulation of post-replicative

gene transcription. The VITF-3 subunits are synthesized from early genes and the

mRNAs are not detected after 6 hours post infection. Therefore the regulation could be

due to a cessation of synthesis of these viral transcripts or competition of more abundant

late transcription factors (134). The cellular transcription factor, YY 1, is the first cellular

factor identified for its role in vaccinia transcription. YY1 was thought to bind the late

gene promoter IlL but recent evidence indicates that the IlL promoter belongs to the

intermediate class (Steven Broyles, personal communication) (17). The YY1 protein

activates transcription from the intermediate protein in vitro and requires its DNA

binding domain (17). The intermediate RNA polymerase complex does not recognize

early termination signals and synthesizes a heterogeneous family of intermediate

transcripts that differ at the 3'-end (98). This implies that if there are cis-acting

termination sequences in the DNA they are likely to be ubiquitous and/or highly

degenerate.









Vaccinia Virus Late Gene Transcription

Late stage vaccinia mRNA transcription is reconstituted in vitro by the use of

infected cell cytoplasmic extract. Several factors required for late gene transcription

initiation were identified, however, additional factors are still being sought as

transcription cannot be reproduced from purified factors alone (Table 2). Three

intermediate proteins encoded by the open reading frames of AlL (VLTF-2), A2L

(VLTF-3), and G8R (VLTF-1), are necessary for late gene transcription initiation

(64,65,160,171). Both Al and A2 are zinc binding proteins (65) and G8 interacts with

itself and Al, as demonstrated by the yeast two-hybrid system (92). An additional viral

factor, VLTF-4 encoded by the HSR open reading frame, is synthesized early and late

during infection and stimulates late gene transcription (70,71). A cellular factor, VLTF-

X, was also described as necessary for in vitro transcription of late genes and is an RNA

binding protein (Cynthia Wright, personal communication) (50,159). Similar to

intermediate transcription, the late transcription complex does not recognize early

termination signals that are frequently present within the coding region of late genes and

generates long transcripts with heterogeneous 3'-ends (98).

Identification and Characterization of Vaccinia Virus Transcription Elongation and

Termination Factors

The power of vaccinia virus lies in the ability to genetically manipulate the

genome and study gene expression in vivo. During the 1970s and 1980s several groups

isolated several collections of vaccinia virus temperature-sensitive mutants. Genetic

characterization of these mutants revealed some mutants that have noticeable effects on

the transcript Y-ends. Two notable complementation groups are represented by the G2R









and A18R mutants. These mutant viruses were chosen for further study of vaccinia virus

transcription elongation and termination.


The A18 Protein

Cts23, a temperature sensitive virus containing a mutation in the gene A18R,

shows an abortive late phenotype. Viruses designated as abortive late show a defect in

protein synthesis and do not produce progeny virions under the nonpermissive conditions.

At the nonpermissive temperature the A18R mutant viruses show a drastic decrease in the

level of late steady state RNA (8). Transcriptional analysis of several vaccinia genes

using northern blots, RNase protection, and RT-PCR analysis determined that mutations

in the gene A18R result in readthrough transcription from intermediate promoters into

downstream genes (115,163). These transcripts are longer than those in a Wt infection.

The vaccinia genome contains open-reading frames that are transcribed in both rightward

and leftward directions. Therefore, in some regions of the genome readthrough

transcription results in the synthesis of complementary strands of RNA. The elevated

levels of double-stranded RNA induce the cellular 2'-5'A pathway resulting in the

degradation of late viral messages and accounts for the abortive late phenotype (8). The

vaccinia A 18R gene encodes a 56-kDa protein that is expressed throughout infection and

packaged in virions (146). The A18 protein is both a 3'-5' DNA helicase and a DNA-

dependent ATPase (9,147). Based on the phenotypic analysis of Cts23 and the data

presented in this dissertation, the A 18 protein is a transcript release factor and possibly a

transcription termination factor.

The treatment of Wt virus with the anti-poxviral drug isatin-p-thiosemicarbazone

(IBT), results in the synthesis of longer than Wt transcripts at intermediate and late times









during infection, similar to the effect of the A18R gene mutation. This implies that IBT

promotes readthrough transcription in a Wt virus. The exact mechanism of IBT action is

not known. We hypothesize that the target of IBT is involved in transcription

termination. This is supported by the isolation of IBT-dependent mutants that have

phenotypes in post-replicative transcription elongation and termination.


The G2 Protein

The in vivo phenotypic analysis of the G2R gene was enabled with the use of two

conditional lethal mutants: Cts56 and G2A. Cts56 is a temperature-sensitive mutant that

requires the anti-poxviral drug IBT for growth at the non-permissive temperature (400C)

and is IBT-resistant at the permissive temperature (310C). G2A is an IBT-dependent

deletion mutant that plaques only in the presence of IBT at 370C (93). The G2R mutants

appear to have normal initiation of all three gene classes and early mRNA structure is

unaffected. However, intermediate and late mRNAs are reduced in size as a result of

truncation from the 3'-end, suggesting an effect on transcription elongation (11). The

G2R gene is expressed early and predicted to encode a 26-kDa protein.

The G2A mutant virus is not only IBT-dependent but is also an extragenic

suppressor of A18R mutants. Theoretically, the reduced elongation seen in a G2R mutant

virus is compensated by the readthrough transcription that results from an A 18R mutation

or IBT treatment. Alternatively, the enhanced elongation in A18R mutants and IBT

treatment is compensated by a mutation in the G2R gene (35).

The viral H5R gene product was shown to associate directly with the G2 protein

(12). The H5 protein is an abundant phosphoprotein found associated with virosomes

(10), and it was shown to stimulate late viral transcription in vitro (70). We believe the









G2 protein functions in a Wt infection by enhancing transcription elongation at

intermediate and late times during infection. Together, the evidence suggests that the

A18, G2, and H5 proteins are all associated either directly or indirectly as a complex in

vivo (12).


The J3 Protein

The J3 protein was previously characterized as a bifunctional (nucleoside-2'-O-)-

methyltransferase and as a processivity factor for the heterodimeric viral poly(A)

polymerase (138). This 39-kDa protein is expressed throughout infection and packaged

in virions (100,103). Isolation of aJ3R mutation as an extragenic suppressor of the A18R

mutation has led to the hypothesis that J3, like G2, functions as a positive transcription

elongation factor. In fact, several J3R mutations were isolated by selecting for IBT-

dependent mutants, all of which were null mutations that synthesize no detectable J3

protein (78). Northern blot and structural analysis of the F1 7R gene indicate that J3R

mutant viruses produce intermediate and late transcripts that are specifically Y-end

truncated consistent with the reduction in large proteins late during infection. Analysis of

two J3R mutant viruses, which retain or lack the poly(A)-stimulatory activity,

demonstrate that the poly(A)-stimulatory activity of J3 is separable from the elongation

activity (78,162).


Summary
The goal of this dissertation is to provide a biochemical characterization of the

regulation of vaccinia virus transcription elongation and termination. The in vivo

analysis of several vaccinia virus mutants in the genes A 18R, G2R, and J3R, provided the

initiative for our hypothesis. We propose that these three viral proteins, in conjunction









with other viral or cellular factors, regulate vaccinia transcription elongation and

termination at post-replicative times during infection. These proteins work as positive

and negative factors to balance the synthesis of mRNA of the correct length. We further

propose, based on recent data from studies of vaccinia early transcription termination,

that the vaccinia transcription machinery may exist as a holoenzyme, similar to those

demonstrated for both prokaryotic and eukaryotic systems. This holoenzyme is

composed of the factors necessary for processing of mRNA 5'- and 3'-ends in addition to

the machinery for transcription initiation, elongation, and termination. This hypothesis is

supported by the physical recycling of viral proteins for different functions during the

transcription cascade. For example, the viral capping enzyme is involved in 5'-cap

formation of all 3 stages of transcription, as well as serving as an early termination factor

and an intermediate initiation factor. The J3 protein is another example of a recycled

protein, as it has activity both in 5'-cap formation and 3'-end polyadenylation, as well as

elongation. There may be two forms of this holoenzyme in the cell, as early promoters

clearly are selective in recruiting RNA polymerase molecules containing RAP94,

whereas intermediate and late promoters recruit the RAP94(-) polymerase. The data

presented in this dissertation demonstrates that at least one of these proteins, A18, is

directly involved in post-replicative transcription termination. We have also identified a

cellular factor that appears to participate in transcription termination.














CHAPTER 2
MATERIALS AND METHODS


Eukaryotic Cells, Viruses, and Bacterial Hosts

A549 cells, wild type vaccinia strain WR, and A 18R temperature-sensitive mutant

Cts23, and the conditions for their growth, infection, and plaque assay have been

described previously (32-34). Escherichia coli DE3 pLysS contains an isopropyl-1-thio-

P-B-inducible chromosomal copy of the bacteriophage T7 RNA polymerase gene (149).


Plasmids
All plasmids used for transcription are based on pC2AT19 (135) containing a 375-

nt G-less cassette cloned into pUC13 with the total size approximately 3-kb. pG8G,

pVGFG, and pCFWlO contain upstream of the 375-nt G-less cassette promoters from the

intermediate vaccinia gene G8R, the early vaccinia gene C11R, and the late vaccinia gene

F1 7R (32,161), respectively. pSB24 contains a synthetic early promoter upstream from

the 375-nt G-less cassette (85). pG8GX is a derivative of pG8G that contains the

vaccinia gene G8R intermediate promoter upstream of a 3'-truncated, 94-nt G-less

cassette derived from pC2AT19 (76). pSB23term contains a synthetic early promoter

upstream of a 540-nt G-less cassette and contains the early termination signal UsNU

(28,32).

pG8GI is a derivative of pG8G that contains the vaccinia gene G8R intermediate

promoter upstream of a 3-truncated, 37-nt G-less cassette derived from pC2AT19. The

G8R promoter and the 5' 37-nt of the pC2AT19 cassette were PCR-amplified from pG8G

using an upstream primer that hybridized approximately 270-nt upstream of the G8R
45









promoter flanked with a SacI site and a Sal site and a downstream primer that contained

nucleotides 18 to 37 of the G-less cassette flanked with a Sinai site and a BamHI site.

The PCR-amplified fragment was cleaved with SacI (upstream) and BamHI

(downstream) and cloned into the vector pGEM3ZF-, which had also been cleaved with

SacI and BamHI. The SinaI site at the 3' end of the resulting truncated G-less cassette

serves to efficiently arrest transcription of the G-less cassette, and the upstream Sall site

was used for identification of the desired clone. Accurate transcription of the pGSGI G-

less cassette should yield RNA of approximately 37-nt in length.

pG8G4a and pG8fe are derivatives of pG8GI that contain the vaccinia gene G8R

intermediate promoter upstream of a Y-truncated, 37-nt G-less cassette derived from

pC2AT19. The pG8G4a plasmid contains the vaccinia late gene AJOL which was PCR-

amplified from purified Wt vaccinia virus DNA using an upstream primer that hybridized

to nucleotides 1 to 22 corresponding to the initiating ATG of AlOL flanked by a Sinai site

and a downstream primer that hybridized to the 3' 19-nt of the AJOL gene flanked by a

HindIII site. The PCR-amplified fragment was cleaved with SmiaI and HindlII and

cloned into pG8GI that had also been cleaved with SinaI and HindI. The resulting

plasmid contains the 37-nt G-less cassette followed by the coding sequence of the AIOL

gene. The pG8fe plasmid contains the vaccinia late genes F1 7R and EJL that were PCR-

amplified from purified Wt vaccinia virus DNA using an upstream primer that hybridized

to nucleotides 1 to 23 corresponding to the initiating ATG of FI7R flanked by a SinaI site

and a downstream primer that hybridized to nucleotides 1 to 20 corresponding to the

initiating ATG of EJL flanked by a PstI site. The PCR-amplified fragment was cleaved

with SmaI and PstI and cloned into pG8GI that had also been cleaved with SmaI and PstI.









The resulting plasmid contains the 37-nt G-less cassette upstream of the coding sequence

of the F17R and ElL genes.

p1 6A 18 (9) contains the vaccinia virus gene A 18R coding sequence inserted in

frame downstream from an amino-terminal polyhistidine tag in the vector pET16b

(Novagen).


Infected Cell Extracts for Transcription

Confluent 100-mm dishes of A549 cells were either mock-infected or infected

with vaccinia virus with a multiplicity of infection of 15 and incubated at 40'C for 16 h in

the presence of 10 mM hydroxyurea or in the absence of drug. Extracts were prepared as

described (32). Briefly, vaccinia-infected cell monolayers were permeabilized with

lysolecithin, harvested, treated with micrococcal nuclease, clarified by centrifugation, and

stored at -700C. Total protein concentration was determined by the Bradford protein

assay (Bio-Rad).


Immobilized DNA Templates

All templates used for transcription were immobilized by binding linearized

plasmid DNA to paramagnetic beads. One set of immobilized templates, including

NpG84a, NpG8G, NpG8fe, NpSB24, and NpCFW1O, were generated by linearization

with NdeI, which cleaves the DNA template 220-nt upstream from the promoter. The

resulting templates contain a 375-nt G-less cassette and approximately 2400-nt DNA

downstream from the G-less cassette. Two additional shorter templates, N/VpG8G and

N/VpG8GX, were constructed by restriction digest with NdeI and VspI. The resulting

templates contain 220-bp DNA upstream from the G8R intermediate promoter and either

540- or 260-bp downstream for transcription (Fig. 8B). In all cases the cleaved DNA

fragments were end-filled with Klenow, dCTP, dGTP, and dATP, and biotin-16-dUTP









(Roche Molecular Biochemicals). The biotinylated DNA was separated from the free

nucleotides using the High Pure PCR Product Purification Kit (Roche Molecular

Biochemicals). The DNA was eluted from the column in 100 W of TE (10 mM Tris-HCl,

pH 8.0, 1 mM EDTA) and adjusted to 1 M NaC1. DNA samples were then incubated

with streptavidin-conjugated Dynabeads M280 (Dynal) in 1 M NaCI/TE for 30 min at

42C to generate bead-bound templates. Beads with bound DNA were concentrated

using a magnet and washed twice in 1 M NaCi/TE, followed by two washes in TE. The

bead-bound DNA was stored in TE at 4C.


In Vitro Transcript Release Assay

The purpose of this dissertation was to develop an in vitro system to characterize

the A18, G2 and/or J3 proteins. The reaction described here represents the final

conditions of this assay as used to measure transcript release. Variations on this assay

were used during the development and are described in the text of Chapter 3.

Transcription reactions were performed in three phases, initiation, pulse, and

chase. Reactions (25 W11) contained a final concentration of 25 mM HEPES, pH 7.4, 4.5%

glycerol, 80 mM KOAc, 5 mM MgC12, 1.6 mM DTT, 1 mM AT?, 5 W.1 of bead-bound

DNA template, and 15 1 of extract from hydroxyurea-treated wild type vaccinia-infected

cells. Reactions were incubated at 30'C for 10 min to form initiation complexes. The

pulse phase was initiated by adding 3 W1 of a solution containing 11 mM ATP, 11 mM

GTP, 6 mM UTP, and 6 tCi of [a-P32] CTP (-3000 Ci/mmol stock) such that the final

concentration is 2.1 mM ATP, 1.1 mM GTP, 0.6 mM UTP, 22.3 mM HEPES, pH 7.4,

4% glycerol, 71.4 mM KOAc, 4.5 mM MgC12, and 1.4 mM DTT in a total of 28 .1.

These reactions were then incubated at 30'C for 30 s. The reactions were stopped by









placing the tube on a magnet on ice. The pellets were washed with 1 to 1.5 pulse reaction

volumes of high salt transcription buffer (5 mM MgC12, 25 mM HEPES, pH 7.4, 1.6 mM

DTT, 1 M KOAc, and 7.5% glycerol), followed by three washes in 1 to 1.5 pulse reaction

volumes of low salt transcription buffer (5 mM MgC12, 25 mM HEPES, pH 7.4, 1.6 mM

DTT, 80 mM KOAc, 200 g.g/ml bovine serum albumin, and 7.5% glycerol). The chase

phase was done by adding to the resuspended complexes a mixture of NTPs, extract, and

proteins in a final volume of 25 p1 containing 25 mM HEPES, pH 7.4, 4.5% glycerol, 80

mM KOAc, 5 mM MgC12, 1.6 mM DTT, 600 pM ATP, 600 pM GTP or 10 pM 3'-

OMeGTP, 600 VM UTP, 1.2 mM CTP, 20 units RNasin, and purified protein or extract

as indicated. Chase reactions were performed at 30'C for various times. The beads were

concentrated using a magnet, and the 25 Wl supernatant was removed to a separate tube.

One hundred seventy five microliters of "PK mix" (114 mM Tris-HC1, pH 7.5, 14 mM

EDTA, 150 mM NaC1, 1.14% SDS, 40 jig of glycogen, 230 jig/ml proteinase K) was

added, and reactions were incubated at 37C for 30 min. Reactions were extracted once

with 175 Wl of phenol/chloroform. Nucleic acids were precipitated by addition of 50 1

10 M ammonium acetate and 150 1 isopropyl alcohol, incubation at room temperature

for 30 min, and centrifugation for 20 min. Pellets were washed once with 70% ethanol,

dried, and resuspended in 10 .1 of formamide loading buffer. Samples were denatured at

90'C for 3 min and loaded on a 6% 8 M urea-PAGE. Gels were fixed, dried, and

analyzed by autoradiography and phosphorimagery. Released transcripts were expressed

as a percentage derived by dividing the quantity of transcripts in the supernatant by the

total quantity of transcripts in both the supernatant and associated with the beads.









Induction and Preparation of Extract from E. coli

An overnight culture of pLysS cells harboring the p16A18 plasmid was used to

inoculate 1 liter of L-broth, containing 50 gtg/ml ampicillin and 34 pgg/ml

chloramphenicol. The culture was incubated at 37C to an A600 of 0.5. Isopropyl-l-thio-

P-D-Galactopyranoside was added to a final concentration of 1 mM, and the culture was

incubated at 370C for 4 h. The cells were pelleted and stored at -700C overnight. All

subsequent procedures were performed at 40C. The thawed bacterial pellet was

resuspended in 50-ml of lysis buffer (50mM Tris, pH 7.5, 0.15M NaC1, 10% sucrose)

plus a final concentration of 50 gig/ml lysozyme and 0.1% Triton X- 100. The cells were

sonicated at 4C for eight sequences consisting of 15 s on and 45 s off. Insoluble material

was removed by centrifugation for 30 min at 18,000 rpm in a Sorvall SS34 rotor at 40C.

For purification of the soluble Al 8R protein, the supernatant was then chromatographed

on a His-Bind (Novagen) column and phosphocellulose column as described below.


His-bind Column and Phosphocellulose Column

The supernatant was mixed for 1 h with 2-ml of nickel-nitrilotriacetic acid-

agarose resin (Quiagen) that was equilibrated with lysis buffer. The slurries were poured

into a column and washed sequentially with 20-ml of lysis buffer, 20-ml of binding

buffer (5mM imidazole, 0.5 M NaC1, 20 mM Tris-HC1, pH 7.9, 5% glycerol), and 20-ml

of wash buffer 1 (60 mM imidazole, 0,5 M NaCl, 20 mM Tris-HCl, pH 7.9, 5% glycerol).

Bound proteins were eluted with 20-ml of wash buffer 2 (200 mM imidazole, 0.5 M

NaCl, 20 mM Tris-HC1, pH 7.9, 5% glycerol) collecting 1-ml fractions. Peak fractions

were identified using the Bradford protein assay (Bio-Rad), pooled, and dialyzed

overnight against 1 liter of Buffer A (25 mM Tris-HC1, pH 7.5, 1 mM EDTA, 0.01%

Nonidet P-40, 1 mM DTT, 10% glycerol, 0.1 mM phenylmethylsulfonyl fluoride, 0.5









tg/gl leupeptin, and 0.7 tg/gl pepstatinA). The dialysate was applied to a 2-ml column

of phosphocellulose that had been equilibrated with Buffer A. The column was washed

with 5-ml of Buffer A containing 250 mM NaCi. Bound proteins were eluted with 10-ml

of Buffer A containing 500 mM NaC1 collecting 0.5-ml fractions. Peak fractions were

identified using the Bradford protein assay, pooled and dialyzed overnight against 4

changes, 1 liter each, of a solution containing 40 mM Tris-HCI, pH 8, 20 mM KCI, and

40% glycerol. The enzyme was stored at -20'C. The His-A 18R protein preparation was

greater than 90% pure as judged by PAGE and displayed DNA-dependent ATPase

activity of 10,000 nmol of ATP hydrolyzed per min per ptg of protein, equivalent to

previously reported preparations (9).

Vaccinia virus J3R protein containing both a polyhistidine- and thioredoxin-tag,

was prepared in a fashion similar to His-Al 8R (162).

Polyhistidine-tagged human factor 2, prepared as described (84), was a gift from

Dr. David Price (University of Iowa).


Western Blot Analysis

Samples were separated by electrophoresis on 10% SDS-PAGE. The proteins

were transferred to nitrocellulose in 25 mM Tris-HC1, 192 mM glycine, 20% methanol at

4'C overnight. Nitrocellulose filters were incubated with monoclonal anti-A 18 primary

antibody (1:10,000) (12), and the bound antibody was detected using polyclonal anti-

mouse horseradish peroxidase-conjugated antibody (1:5000, Amersham Pharmacia

Biotech) and enhanced chemiluminescence. Western blotting reagents (Amersham

Pharmacia Biotech) were used as described by the manufacturer.









Preparation of Nuclear and Cytoplasmic Fractions of HeLa Cells

HeLa cells grown in suspension culture to a density of 5 x 105 cells per ml (a

generous gift from Brian O'Donnell) and were harvested for extraction. All subsequent

procedures were performed at 4C. Cell pellets were resuspended in Buffer A at a ratio

of 5-ml of buffer per ml of packed cell pellet. Cells were allowed to swell on ice for 10

min followed by centrifugation at 1000 x g for 10 min. The cell pellets were resuspended

in Buffer A at a ratio of 2-ml of buffer per ml of packed cell pellet. The cells were

ruptured by Dounce homogenization using a tight-fitting pestle. The lysate was

centrifuged at 1000 x g for 15 min to pellet nuclei. The nuclear pellet was saved for

extract preparation. The supernatant was centrifuged again at 10,000 x g for 15 min and

the resulting supernatant was saved and labeled as cytoplasmic extract (HCE). The

nuclear pellet was subjected to additional centrifugation at 25,000 x g for 20 min. The

pellet was resuspended in 3-ml Buffer C (20 mM HEPES, pH 7.9, 25% glycerol, 0.42 M

NaCl, 1.5 mM MgC12, 0.2 mM EDTA, 0.5 mM PMSF, and 0.5 mM DTT) for every 1 x

l09 cells. The nuclei were triturated by Dounce homogenization using a tight-fitting

pestle. The homogenate was mixed using a stir bar for 30 min at 40C, centrifuged at

25,000 x g for 30 min, and dialyzed overnight against Buffer P (20 mM HEPES, pH 7.9,

10% glycerol, 1.5 mM MgC12, 0.1 mM EDTA, 1 mM DTT, 0.1 mM PMSF). The

dialysate was centrifuged at 25,000 x g for 20 min and the resulting supernatant was

stored at -70C as nuclear extract (HNE).


Chromatography and Fractionation

Crude Fractionation of Wt or Cts23 Extract

Extract from Wt- or Cts23-infected A549 cells was chromatographed on 2-ml

columns of phosphocellulose (Whatman) or Q-Sepharose (Amersham Pharmacia









Biotech) equilibrated in Buffer A. All steps were performed at 40C. Extract was loaded

on the column, and the column was washed in 4-ml of Buffer A, and 0.5-ml flow-through

fractions were collected. Bound proteins were eluted stepwise with 4-ml each of Buffer

A containing 0.25, 0.5, and 1 M NaC1, and 0.5-ml fractions were collected. Peak

fractions were identified using the Bradford protein assay, pooled, and dialyzed overnight

against Buffer A containing 50 mM NaC1. The fractions were stored at -20'C.


HQ Purification

Cytoplasmic extract from uninfected HeLa spinner cells was chromatographed on

Porus 20 HQ (PerSeptive Biosystems) equilibrated in bis-Tris propane, pH 6. The

column was run using the BioCAD Perfusion Pump (provided by the Protein Chemistry

Core Facility, Biotechnology Program, University of Florida) at room temperature.

Extract was loaded on the column, and the column was washed in five column volumes

of bis-Tris propane, pH 6, and the flow-through fraction was collected and placed on ice.

Bound proteins were eluted using a gradient of bis-Tris propane, pH 6 from 0 M to 0.5 M

NaCI followed by a wash with 2 M NaC1, and 1-ml fractions were collected and placed

on ice. Peak fractions were identified based on absorbance at 280 nm. The fractions

were stored at -700C.


Hydroxyapatite Purification

The hydroxyapatite purification was performed subsequent to purification over Q-

Sepharose. Cytoplasmic extract from uninfected HeLa spinner cells was

chromatographed on a 2-ml column of Q-Sepharose equilibrated in Buffer P. All steps

were performed at 40C. Extract was loaded on the column, and the column was washed

in 4-ml of Buffer P, and 0.5-ml fractions were collected. Bound proteins were eluted

using a continuous gradient of Buffer P from 0 M to 1 M NaC1, and 0.5-ml fractions were









collected. Peak fractions were identified using the Bradford protein assay and the in vitro

transcript release assay. Q-Sepharose fractions 21 to 35 were pooled and dialyzed against

Buffer D (20 mM HEPES, pH 7.4, 0.1 mM EDTA, 1 mM DTT, 10% glycerol, 0.7 g.g/ gl

pepstatin A, 0.1 mM PMSF, 0.5 gtg/ .1l leupeptin) containing 0.01 M phosphate. The

dialyzed fractions were chromatographed on a 1.5-ml hydroxyapatite column equilibrated

in 0.01 M Buffer D, washed in 4-ml 0.01 M Buffer D collecting 0.5-ml fractions. Bound

protein was eluted using a continuous gradient of Buffer D from 0.01 M to 0.4 M

phosphate, and 0.5-ml fractions were collected. Peak fractions were determined using the

Bradford protein assay. The fractions were stored at -700C.


Phosphocellulose Purification

Cytoplasmic extract from uninfected HeLa spinner cells was chromatographed on

a 2-ml column of DE-52 equilibrated in Buffer P to remove the majority of nucleic acid

prior to fractionation of phosphocellulose. The column was washed in 4-ml Buffer P and

bound protein was eluted in Buffer P containing 0.5 M NaC1. Peak fractions were

determined using the Bradford protein assay and fractions 15 to 19 were pooled and

dialyzed against Buffer P. The dialyzed DE-52 fraction was subjected to gradient

fractionation on phosphocellulose from 0 M to 1 M NaCl and 0.5-ml fractions were

collected. Peak fractions were determined by Bradford protein assay. The fractions were

grouped and dialyzed against Buffer P. The fractions were stored at -70C.













CHAPTER 3
RESULTS


Objectives and Specific Aims

The overall goal of my research is to provide a biochemical characterization of

the regulation of vaccinia virus transcription elongation and/or termination. The in vivo

analysis of the viruses containing mutations in the genes G2R and J3R indicates that the

transcripts synthesized from intermediate and late genes are 3' truncated as compared to a

Wt infection (11,78,162). We therefore hypothesize that G2 and J3 function as positive

transcription elongation factors. Through the use of Northern blots, RNase protection,

and reverse transcriptase-PCR analysis it was determined that virus containing a

temperature sensitive mutation in the gene A18R synthesize transcripts that are longer

than those synthesized during a Wt infection (163). We therefore hypothesize that A 18

functions as a negative transcription elongation factor or a termination factor. The goal

was to develop and characterize assays for elongation and termination and to

simultaneously screen for activity of any or all of the aforementioned proteins. This was

a huge undertaking, as there were no established assays for pausing, pause suppression,

or termination for the post-replicative genes of vaccinia virus. Therefore, the assays

varied until an in vitro phenotype was discovered for A18. The subsequent assays

focused on characterizing the A 18 protein using the new transcript release assay. Clearly

there are other experiments that could be pursued to characterize G2 and J3 using the









knowledge gained from the preliminary elongation assays described in this dissertation

and those experiments are discussed further in Chapter 4.


Specific Aim 1: Develop an Assay to Determine the Biochemical Activity of A18, G2,
and/or J3

The first aim of this project was to characterize the transcription reaction by

testing variables important for elucidating elongation and termination factors in other

systems. We developed various transcription assays based on a previously described

crude system for the study of vaccinia early, intermediate, and late gene transcription

initiation (32). Previous experiments showed that crude extract prepared from cells

infected under normal conditions is competent for transcription of early, intermediate,

and late gene promoters. Since intermediate and late viral gene expressions are coupled

to viral DNA replication, treatment of infected cells with a DNA replication inhibitor

such as hydroxyurea permits synthesis of only early gene products, including

intermediate transcription factors. Thus extracts prepared from cells infected in the

presence of hydroxyurea are competent for transcription of intermediate promoters only

(32). For most experiments, we chose to use hydroxyurea-treated, intermediate

promoter-specific extract for two reasons. First, the best evidence that the A 18, G2, or J3

proteins have elongation factor activity is based on in vivo studies of intermediate genes

(11,78,162,163). Second, we wished to prepare extract from A18R mutant infections

under non-permissive conditions while at the same time circumventing undesirable

pleiotropic effects of the A18R mutation. Readthrough transcription from convergent

intermediate promoters during A18R mutant infections causes double-stranded RNA

accumulation, induction of the cellular 2'-5'A pathway, and ultimately activation of

RNase L (146,147,163). Hydroxyurea treatment prevents 2'-5'A pathway activation by









preventing intermediate transcription. Cells infected with A18R mutant virus at the non-

permissive temperature produce less than 10% of the normal amount of A 18 protein due

to instability of the mutant protein (146). Thus preparation of extracts from A18R

mutant-infected cells at the non-permissive temperature provides an A18 protein-

deficient extract that is otherwise comparable to extract from cells infected with Wt virus

under identical conditions.

Due to the fact that the transcription assay was under development and the

experiments are not necessarily presented in chronological order, the exact details of the

assay as presented in this dissertation may differ between experiments. The variation in

the basic steps of the transcription assay can be summarized through the description of a

general formula. The general formula for the transcription assay includes: 1)

transcription complex formation, 2) 32p-labeling of the nascent RNA, 3) washing of the

isolated complex, 4) elongation during a chase step, and 5) RNA isolation. Transcription

complexes are assembled on a paramagnetic bead-bound DNA template in one of two

fashions; either during a separate incubation of viral extract with the DNA template or

concurrent with the pulse reaction containing the viral extract, DNA template, and

nucleotides. In both cases initiation and p32-labeling of the nascent RNA occurs during

the pulse reaction. The pulse-labeled ternary complexes are isolated using a magnet and

washed in transcription buffer containing either 80 mM KOAc (low salt wash) or 1 M

KOAc (high salt wash). Transcription elongation continued during the chase step in

which the nucleotide concentration and protein composition are varied. The final step is

isolation of the RNA. RNA released from the ternary complex is isolated using a magnet

to separate bead and supernatant fractions, representing bound and released RNA,









respectively. Alternatively, the total RNA synthesized during the reaction is isolated by

not separating the beads from the supernatant. The specific variations used in each

experiment will be described relative to this general formula.

Based on the in vivo data regarding mutations in the G2, J3, and A18 proteins, we

hypothesized that these proteins were positive and negative elongation factors. To test

this hypothesis, intermediate promoter-specific transcription complexes established on

bead-bound DNA templates were assayed for different patterns of elongation during a

chase step that contained limiting nucleotides and Wt or mutant extract or purified

protein. This assay is based on the idea that nucleotide starvation enhances pausing of

the RNAP thus allowing us to assay the effect of potential transcription elongation factors

at various pause sites in vitro. To date, we have not defined the activities of either G2 or

J3 using the in vitro elongation assay and additional experiments are in progress.

Experiments to assess the stability of transcription elongation in this assay were

performed with the use of sarkosyl or NaC1 during the elongation step. The ternary

complexes were fairly resistant to challenge with either agent indicative of the stability of

the complex during elongation. During this period of experimentation we discovered that

in vitro the A18 mutant virus, Cts23, did not induce release of nascent RNA at levels

similar to the Wt extract. This discovery resulted in the subsequent specific aims and

focused on characterization of the A 18 protein using the transcript release assay.


Specific Aim 2: In Vitro Analysis of the A18 Phenotype

The second specific aim focuses on the characterization of transcript release and

the importance of A18 in that process. Intermediate promoter-specific, pulse-labeled

transcription complexes established on bead-bound DNA templates were assayed for









transcript release during an elongation step that contained nucleotides and various

proteins. Release was analyzed by comparing transcripts present in the supernatant to

transcripts in the bead-bound fraction. Extract from Wt-infected cells, but not mock- or

Cts23-infected cells, stimulated transcript release. The presence of A18 protein is an

absolute but not the sole requirement for transcript release. We also demonstrate that an

additional activity, in combination with A 18, is necessary for efficient transcript release.


Specific Aim 3: Characterization of the Cellular Factor

Transcript release is achieved using a combination of extract from Cts23- or

mock-infected cells plus purified A18 protein. These data suggest that the additional

activity necessary for intermediate transcript release is provided by a cellular factor (CF).

With the ultimate goal of identifying the cellular factor in mind, we first tested a known

cellular termination factor for activity in the vaccinia in vitro system. This protein,

Factor 2, did not substitute for A 18 or CF in the vaccinia transcript release assay. We

therefore continued experiments to identify CF using conventional chromatography. The

CF is present in both nuclear and cytoplasmic extracts from uninfected HeLa cells.

Cytoplasmic extract fractionated over several chromatography resins demonstrated that

CF binds to DEAE, Q-Sepharose or HQ, and hydroxyapatite and does not bind to

phosphocellulose.

Specific Aim 4: Characterize A18/CF-Dependent Release from All Vaccinia

Promoters

The fourth specific aim of my research was to determine the promoter specificity

of Al 8-dependent transcript release. Transcription complexes were formed on early and

late vaccinia promoters and assayed for the ability to induce transcript release in the









absence or presence of A18 protein and CF. The combination of A18 and CF is active in

inducing release from all three classes of promoters. In addition, CF also enhances

release of transcripts terminated by recognition of the vaccinia early gene-specific

termination signal.


Specific Aim 1: Develop an Assay to Determine the Biochemical Activity of A18, G2,
and/or J3

Formation of Paused Transcription Complexes

Extract from hydroxyurea-treated vaccinia-infected cells was used to assay

elongation from linear bead-bound DNA templates containing a vaccinia intermediate

promoter as follows. Transcription complexes were assembled and initiated during a 30-

minute pulse reaction containing Wt extract, bead-bound template, [a-32p] CTP, ATP,

UTP, and 3'-OMeGTP. The ternary complexes, consisting of the DNA template, the

transcription apparatus, and the radiolabeled nascent RNA, were purified from

nonspecifically bound proteins and unincorporated nucleotides during two washes in a

low salt (80 mM KOAc) transcription buffer. The elongation assay was performed by

addition of a "chase" mixture containing NTPs, extract, and proteins as indicated. Total

labeled RNA was analyzed on a denaturing polyacrylamide gel.

One of the predominant characteristics of identified pause sites is the presence of

T-rich sequences in the nontemplate strand (155). To artificially enhance pausing at

these sites in vitro we used a reduced quantity of UTP in the chase mixture.

Transcription was initiated on the NpG84a bead-bound template that contains the late

vaccinia gene AIOL downstream from the vaccinia G8R intermediate promoter and a 37-

nt G-less cassette. Purified ternary complexes were chased in the presence of 1 mM









ATP, CTP, GTP, and 2 pM UTP to reveal several paused transcription complexes (Fig. 5,

Lane 1). We then tested the effect of purified proteins during transcription elongation

with reduced UTP concentration (Fig. 5, Lanes 3-5). We detected no change in the

elongation pattern due to the presence of A18, G2 or J3 proteins. We also tested the

effect of either Wt or mutant extract (G2A or Cts23) in the presence and absence of

purified A18, G2, or J3 proteins (Fig. 5, Lanes 6-20). There is an increase in the length

of transcripts recovered in the presence of either Wt or G2A extract, although there are

still discernible pause bands (Fig. 5, Lanes 6-15). This increase in length of the paused

transcripts probably represents the presence of contaminating UTP in the Wt and G2A

extracts. There is again no difference with the presence or absence of purified proteins

combined with Wt or mutant extract (Fig. 5, compare Lane 6 with Lanes 8-10, Lane 11

with Lanes 13-15, and Lane 16 with Lanes 18-20). We did not titrate the concentration

of the proteins used in the assay, so it is possible that an effect would be seen using

higher concentrations of purified protein.


Sarkosyl Stability of Elongation and Termination

Sarkosyl is a detergent used to strip factors responsible for elongation from the

transcription complex. We tested the effect of sarkosyl on vaccinia virus transcription

elongation using a modified in vitro transcription protocol. First, transcription complexes

were assembled during a separate pre-incubation reaction containing Wt extract and the

NpG8fe template (contains the vaccinia G8R intermediate promoter, a 37-nt G-less

cassette, and the vaccinia F17R and ElL open reading frames). Transcription was

initiated and the nascent transcript radiolabeled with the addition of [0a-32p] CTP, ATP,

GTP, and UTP during a 30-second pulse reaction. Nonspecifically bound proteins and























Fig. 5. Formation of paused ternary complexes. Figure shows an autoradiogram of an
in vitro elongation assay. Transcription complexes were assembled and initiated in a
mixture containing Wt extract, immobilized NpG8G4a DNA containing the vaccinia G8R
intermediate promoter, 3.5 jiCi [a-32p] CTP (-3000 Ci/mmol stock), 1 mM ATP, 1 mM
UTP, 10 pM 3'-OMeGTP for 30 min at 300C. The labeled complexes were isolated using
a magnet and washed three times in low salt transcription buffer (B+W, lane 1).
Elongation was continued in the absence (B+ W+inc, lane 2) or presence of 1 mM ATP,
CTP, GTP, and 2 IM UTP alone (NTP, lane 3) or with additional 3W protein buffer (DB)
75 ng vvHis-A18 (A18), 75 ng vvHis-G2 (G2), 1 jtg J3 (J3), 37.5 jig Wt extract (W),
18.75 gg G2A extract (G), or 7.5 jig Cts23 extract (C) as indicated by the (+) for 10 min
at 30C. The transcripts were analyzed by 4% 8 M urea-PAGE. Sizes, in nt, are shown
on the left.
















u
Cl
a) +
'h3 :oa


+


330 nt -


















100 nt -


1 4 6 8 1


S 1 14 16 1 20 z
12 14 16 18 20 22


A18
DB
Extract


+ I 1. 1+1 1 +1 1 ~J3
1+ 11 11 G2









unincorporated nucleotides were removed from the ternary complex during a single wash

in low salt (80 mM KOAc) transcription buffer. The isolated complexes were

resuspended in a solution containing nucleotides, Wt extract or buffer, and increasing

concentrations of sarkosyl and incubated for 20 minutes. After the elongation reaction

the beads were concentrated using a magnet, the supernatant was removed to a separate

tube, and the labeled RNA in each fraction was analyzed on a denaturing polyacrylamide

gel. Released transcripts were expressed as a percentage derived by dividing the quantity

of transcripts in the supernatant by the total quantity of transcripts in both the supernatant

and associated with the beads.

In the absence of sarkosyl and Wt extract low levels of transcripts are released

into the supernatant (Fig. 6, Lanes 1 and 2). With the addition of Wt extract the amount

of transcripts released into the supernatant is increased by at least 2-fold (Fig. 6, compare

Lanes 1 and 2 with Lanes 3 and 4). This indicates that there may be additional factors

necessary for transcript release provided by Wt extract but not associated with the

washed transcription complex. The low level of transcript release in the absence of Wt

extract may be a result of the single low salt wash that may not efficiently remove non-

specifically bound proteins. There was no effect on Wt extract-dependent transcript

release with the addition of sarkosyl from 0.01% to 0.025% (Fig. 6, Lanes 5-12).

Concentrations of sarkosyl from 0.05% to 0.3% inhibited transcription elongation and

resulted in release of nascent RNA regardless of the presence of Wt extract, indicating

that these complexes are not stable to high concentrations of sarkosyl (Fig. 6, Lanes 13-

28).























Fig. 6. Instability of elongation complex to high concentrations of sarkosyl.
Transcription complexes were assembled as described in Fig. 3 on immobilized NpG8fe
DNA containing the vaccinia G8R intermediate promoter and the F17R and EJL open
reading frames. Isolated ternary complexes were chased in a mixture containing 0.6 mM
ATP, 0.6 mM GTP, 0.6 mM UTP, and 1.2 mM CTP alone (N, lanes 1 and 2, 5 and 6, 9
and 10, 13 and 14, 17 and 18, 21 and 22, and 25 and 26) or in the presence of 15 jig Wt
extract (W, lanes 3 and 4, 7 and 8, 11 and 12, 15 and 16, 19 and 20, 23 and 24, and 27
and 28). Increasing concentrations of sarkosyl were used in the presence of only NTPs
(N) or NTPs and Wt extract (W) as follows: 0.01% (lanes 5-8), 0.025% (lanes 9-12),
0.05% (lanes 13-16), 0.1% (lanes 17-20), 0.2% (lanes 21-24), 0.3% (lanes 25-28).
Elongation was continued for 20 min at 300C and the bead-bound RNA (B) was separated
from released RNA (S) using a magnet. The transcripts were analyzed by 6% 8 M urea-
PAGE. Percent transcript release is indicated in the table below the autoradiogram.
Sizes, in nt, are shown on the right.

















0.01% 0.025%

N W N W N W
2BSBSBSBSBSBS


-4-'


0




so
40


0.05% 0.1%

N W N W
BSB SBSBS


1 3 5 7 9 11 13 15 17 19


2 4 6 8 10 12


0.2%

N W
BSBS


0.3%

N W*
BSBS2


III
21 23 25 27


14 16 18 20 22 24 26 28


117 40 1 3 x o 1 0 0 1 85 189 88 97 82 78 75 1%transcriptrelease


Sarkosyl






-800 nt



- 350 nt










Salt Stability of Transcription Elongation Complexes

To further characterize the elongation complexes we tested the salt sensitivity of

elongation under conditions of non-limiting nucleotides. Salt, similar to sarkosyl, also

dissociates elongation factors during an elongation reaction and can therefore reveal the

presence or absence of other factors important for elongation. Transcription complexes

were assembled and initiated during a 30-minute pulse reaction containing Wt extract,

bead-bound template, [a-3P] CTP, ATP, UTP, and 3'-OMeGTP, similar to Figure 5.

The NpG8G template contains the vaccinia G8R intermediate promoter, a 375-nt G-less

cassette, and approximately 2.5-kB additional downstream DNA. The presence of 3'-

OMeGTP halts the elongation complex at the end of the G-less cassette where the first

GTP would be incorporated resulting in the synthesis of an approximately 400-nt

transcript. The pulse-labeled ternary complexes were isolated, washed twice in low salt

transcription buffer, resuspended in a mixture containing nucleotides, Wt extract, and

various concentrations of NaCl, and allowed to continue elongation during a 20-minute

chase. After the elongation reaction the beads and supernatant were separated and

quantified using a phosphorimager as described in Figure 6.

During the pulse reaction a significant number of transcripts are released into the

supernatant (Fig. 7, compare Lanes 1 and 2) due to the long incubation and presence of

Wt extract. The bead-bound complexes were removed from the supernatant containing

released transcripts and then chased in the absence or presence of additional nucleotides

and extract. A 20-minute incubation in the absence of nucleotides and extract (Fig. 7,

compare Lanes 3 and 4) or in the presence of only nucleotides (Fig. 7, compare Lanes 5

and 6) results in minimal transcript release. The addition of Wt extract and nucleotides to


















Fig. 7. Salt stability of transcription elongation complexes. Transcription complexes were assembled and initiated on immobilized
NpG8G template containing the vaccinia G8R intermediate promoter as described in Fig. 1 (Pulse, lanes 1 and 2). The isolated,
labeled ternary complexes were washed twice in low salt transcription buffer and resuspended in a mixture containing only buffer
(P+inc, lanes 3 and 4), or 1 mM ATP, GTP, CTP, and 0.6 mM UTP either alone (P+chase, lanes 5 and 6) or in the presence of 15 mg
Wt extract (P+C+(Wt), lanes 7 and 8) and increasing concentrations of NaCl as follows: 25 mM (lanes 9 and 10), 50 mM (lanes 11
and 12), 100 mM (lanes 13 and 14), 200 mM (lanes 15 and 16), 300 mM (lanes 17 and 18), 350 mM (lanes 19 and 20), 400 mM (lanes
21 and 22), 450 mM (lanes 23 and 24), 500 mM (lanes 25 and 26). Elongation continued for 20 min at 30oC. The bead-bound RNA
(B) was separated from released RNA (S) using a magnet. The transcripts were analyzed by 6% 8 M urea-PAGE. Bound and
released transcripts were quantitated using a Phosphorlmager for the whole lane; the quantity of transcripts in the supernatant was
divided by the quantity of transcripts on both the beads and in the supernatant and expressed as a percentage in the table below the
autoradiogram. X, indicates an empty lane. Sizes, in nt, are shown on the right.












cc~
0 Cu
U) C) .r
-) T 00+ NaCI W"
n~ n~ n~-
a. a. -a.
BSBSBSBS BSBS BSBSBSBSBSBSBSM


800nt




.. ..... . .. 3 5 0 n t




1 2 3 4 5 6 7 8 9 1011 121314151617181920212223242526

191 418 4,1 39.- 41 j43 1 35 1 47 14 1 56 54 159 % transcript release









the bead-bound complexes promotes the release of additional transcripts (Fig. 7, compare

Lanes 7 and 8). These data indicate that halted ternary complexes isolated using 3'-

OMeGTP are capable of continued elongation when complemented with GTP and

additional nucleotides. Additionally, the isolated complexes are stable during continued

incubation and do not release the nascent RNA until supplemented with extract from Wt-

infected cells, indicating that release may be dependent on additional factors not present

within the isolated ternary complex. We then tested the salt stability of this reaction by

including a titration of NaCl during the chase phase, in addition to the Wt extract and

nucleotides (Fig. 7, Lanes 9-26). Two observations are of note: release of the nascent

RNA in the presence of Wt extract occurs regardless of the concentration of NaC1 and

increasing NaCl concentration impairs transcription elongation. The release of nascent

RNA does appear to increase slightly with addition of NaC1 from 300 mM to 500 mM

(Fig. 7, Lanes 17-26). This increase in release is accompanied by a decrease in the

elongation potential as evidenced by the appearance of multiple bands representing

transcripts shorter than the full-length template (Fig. 7, Lanes 21-26). These data could

indicate that the increasing concentration of NaCl is inhibiting the interaction between a

positive transcription elongation factor and the ternary complex. In the absence of this

factor increased pausing and release of the transcripts may occur.


In Vitro Transcription Is Specific for the Viral Promoter

The aforementioned in vitro transcription reaction was again modified to decrease

the background levels of release seen in the absence of Wt extract. Additionally, we

tested the fidelity of the pre-incubation step in the in vitro system by proving that the

intermediate promoter was accurately recognized. Transcription complexes were









assembled during the pre-incubation reaction containing Wt extract, bead-bound

template, and ATP. Transcription was then initiated and the nascent transcript

radiolabeled by the addition of [ox-32P] CTP, ATP, GTP, and UTP during a short, 30-sec

pulse reaction. The ternary complexes were stripped of nonspecific proteins and

unincorporated nucleotides during three washes in high salt transcription buffer (1 M

KOAc) followed by three washes in low salt transcription buffer (80 mM KOAc). The

elongation reaction was performed with the addition of a chase mixture containing NTPs,

extract, and proteins. Following the elongation reaction the beads were concentrated

using a magnet, the supernatant was removed to a separate tube, and the labeled RNA in

each fraction was analyzed on a denaturing polyacrylamide gel.

To prove that the intermediate promoter was accurately recognized, two bead-

bound templates were designed such that transcription from the G8R promoter to the

downstream end of the template would generate either 260-nt or 540-nt of RNA (Fig.

8B). Pulse-labeled elongation complexes were established and analyzed on a denaturing

polyacrylamide gel (Fig. 8A, Lanes 1 and 10). The transcripts were approximately 100-

nt in length and were cut off on the autoradiograph shown. Elongation was continued on

addition of ribonucleotides during the chase phase and the transcripts synthesized from

each template were of the appropriate length, either 260-nt or 540-nt (Fig. 8A, Lanes 2

and 11). At the end of the chase phase, the bead-bound template was separated from the

supernatant using a magnet. Comparison of Lanes 2 and 3 and Lanes 11 and 12, Fig. 8A,

indicate that transcripts synthesized during a nucleotides-only chase reaction are not

released into the supernatant but remain associated with the bead-bound template. This

newest protocol for generating elongation complexes used extensive washing with 1 M
























Fig. 8. Transcription is promoter-specific. A, autoradiogram of in vitro transcript
release assay. Transcription complexes were formed from Wt extract on immobilized
NiVpG8GX or N/VpG8G DNA that contain the vaccinia G8R intermediate promoter.
Following a 30-sec pulse reaction (Pulse), labeled complexes were washed in
transcription buffer, and elongation was continued in the presence of 0.6 mM ATP, 0.6
mM GTP, 0.6 mM UTP, and 1.2 mM CTP alone (NTP) or with additional 7.5 jig mock
extract (Mock), Wt extract (Wt), or Cts23 extract (Ts23) for 20 min. The bead-bound
RNA (B) was separated from released RNA (S) using a magnet. These transcripts were
analyzed by 6% 8 M urea-PAGE. Sizes, in nt, are shown on the left. B, diagram of the
DNA templates used for transcription. The DNA template (line) contains a biotinylated
ATP incorporated at both the 5' and 3' end, which anchors the DNA to a streptavidin-
coated magnetic bead (circles). The bead is anchored 220 nt from the promoter at the 5'
end of the template. The transcription unit consists of the G8R intermediate promoter
(arrow) fused to either 260 or 540 nt of downstream DNA. C, graphic representation of
the percent transcript release for each reaction in A.














B SB SB S


cv,
o CJ
cn
_ I-

BSB SBS


350 nt- a 1


123456789


NN pG8GX


220nt 260 nt


45
40
35

30
u 25

~20
* 15


5


10 111213 14 1516 17 18


NN pG8G


n220 t 540 nt






NTP
D- Mock
]]fwt
0 TS23


NN pG8GX NN pG8G


0-
z
B S


800 nt- o


0. z
B BS


I









salt and we questioned whether additional proteins could act on the isolated elongation

complexes to induce release of the nascent transcript from the bead-bound template as

shown in the salt and sarkosyl stability experiments. Chase reactions were therefore

performed in the presence of nucleotides plus extract from mock-, Wt-, or the A18R

mutant Cts23-, infected cells. The addition of extract from Wt-infected cells resulted in

the release of transcripts during a 20-minute chase from either template (Fig. 8A,compare

Lanes 6 and 7, and Lanes 15 and 16). The percent transcript release was analyzed by

phosphorimagery (Fig. 8C). Extract from neither mock-infected nor Cts23-infected cells

was capable of generating a significant amount of released transcripts (Fig. 8A, Lanes 4

and 5, 8 and 9, 13 and 14, and 17 and 18, Fig. 8C). In summary, these experiments show

that initiation in vitro occurs specifically at the viral intermediate promoter. These data

also suggest that transcript release in Wt extract is due to the presence of A 18 protein,

which is absent in Cts23 extract.


Specific Aim 2: In Vitro Analysis of the A18 Phenotype

Release Does Not Require the Presence of A18R during Initiation

In Figures 6 and 8, Wt extract was used to generate the transcription complexes

formed during the pre-incubation step. To determine whether factors specific to a Wt

extract and present in the washed elongation complex contributed to release, we

compared transcription complexes formed using either Wt or Cts23 extract during the

pre-incubation and pulse steps (Fig. 9A, Wt or Cts23 PIC). Transcription complexes

were formed on linearized bead-bound NpG8G, a template that contains approximately 3

kb of sequence downstream from the G8R promoter. After initiation with the addition of

nucleotides and a thorough wash in high salt and low salt transcription buffers, these





























Fig. 9. A18 is not required for initiation in vitro. A, transcription complexes were
formed on immobilized NpG8G DNA and extract from either Wt- (Wt PIC) or Cts23
(Ts23 PIC)-infected cells. Transcription was performed as described in Fig. 3 and
released transcripts (S) were separated from bound transcripts (B) and analyzed by 6% 8
M urea-PAGE. Sizes in nt are shown at the right. B, graphic representation of the
percent transcript release for each reaction in A.













wt PIC


0


BS


_ z

B SBS B S


Ts23 PIC


0

B S


LO

BS BSm


1 2 34




40
: 35
30
_25
=,20
.
15

10o


W 2652 nt

"800 nt







5 6 7 8 9 1011 1213141516








NTP
D Mock

DBwt
STs23


wt PIC Ts23 PIC


U)

z
B S









complexes were chased in the presence of unlabeled ribonucleotides, or nucleotides plus

mock, Wt, or Cts23 extract (Fig. 9A). Both complexes show similar levels of transcript

release in response to the addition of Wt extract (Fig. 9A, compare Lanes 5 and 6, 13 and

14, Fig. 9B). Therefore, Wt or Cts23 extracts are equally competent for transcription

complex assembly and initiation. Therefore, Wt extract was used to generate

transcription complexes for all release assays.


Transcript Release Is Time and Concentration Dependent

To determine the kinetics of release, we performed a time course of elongation.

Pulse-labeled elongation complexes were formed and samples were taken at various time

points during elongation. Similar kinetics of elongation were observed with the addition

of ribonucleotides alone, or in combination with Wt or Cts23 extract (Fig. 1 OA). Release

is detected with the addition of Wt extract (Fig. 10A, Lanes 13-24) and the level of

release increases linearly as a function of time (Fig. 10B). Longer incubation times do

not result in more than 60% release. Cts23 extract also resulted in a linear increase in

release activity with time that was measurably above the nucleotides-only control but

significantly less than Wt (Fig. 10A, Lanes 25-36, Fig. 10B). The lower level of release

observed with addition of Cts23 extract could represent non-specific release or result

from the lower level of A18 protein in Cts23 extract. In summary, release activity is

significantly diminished in a Cts23 extract throughout a time course substantiating our

hypothesis that release is specific to the presence of A 18 protein.

We then titrated the concentration of extract included in the elongation step to

determine the optimal quantity of extract for efficient release. Transcription complexes

were formed from Wt extract during the pre-incubation step, initiated with the addition of





















Fig. 10. Time course of elongation in a chase reaction. A, pulse-labeled transcription elongation complexes were formed on
NpG8G bead-bound template using extract from Wt-infected cells. Complexes were washed in 1 M transcription buffer, and
transcription was continued in the presence of 0.6 mM ATP, 0.6 mM GTP, 0.6 mM UTP, and 1.2 mM CTP alone (NTPs) or in
addition to 30 gg of Wt extract (Wt) or Cts23 (Ts23) for 1, 2.5, 5, 10, 15, and 20 min. Released transcripts in the supernatant (S) and
bound transcripts associated with the bead-bound template (B) were separated and analyzed by denaturing 6% 8 M urea-PAGE. B,
graphic representation of the percent transcript release for each reaction in A.









A NTPs Te WtTime Ts23
~Time .,...] ,.

BSBSBSBSBSBSBSBSB SBSBSBSBSBSBSBSBSBS
oo-2652 nt
i i i -800 nt




0 350 nt
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 1819 20 21 22 23 24 25 26 27 28 29 3031 32 33 34 35 36


60 ,.

*50

" 40
c 40- 4- NTPs
30, M--
--Ts23

120.

10

0 5 10 15 20 25
Time (min)









nucleotides, washed in high and low salt transcription buffers, and then assayed for

elongation and transcript release using increasing concentrations of mock, Wt, or Cts23

extract and ribonucleotides during a 20-minute chase reaction. Increased transcript

release occurred as the quantity of Wt extract was increased (Fig. 11A, Lanes 12-19,

Fig. 11 B), however, no effect on release was observed with increasing quantities of either

mock or Cts23 extract (Fig. 1 A, Lanes 4-11 and Lanes 23-30, Fig. 1 IB). These results

further support the hypothesis that A 18 is important for transcript release.


Transcript Release Is Complemented by Crude Fractions from Wt Extract

In an attempt to correlate the release activity achieved by the addition of Wt

extract with the presence of A18 protein, a crude fractionation protocol was employed.

Extracts were prepared from either Wt- or Cts23-infected cells and fractionated on

phosphocellulose and Q-Sepharose columns separately. Columns were eluted step-wise

with 0.25 M, 0.5 M, and 1 M NaCl. Each Wt extract fraction was analyzed by SDS-

PAGE (data not shown) and by western blot analysis using an anti-A18 monoclonal

antibody (Fig. 12B and C). As demonstrated by western blot, A18 protein fractionated

into the 0.5 M phosphocellulose fraction and the 0.25 M Q-Sepharose fraction (Fig. 12B,

0.5 M and Fig. 12C, 0.25 M). Each fraction was assayed for its ability to induce

transcript release using the protocol described in Figure 11 (Fig. 12A). As controls,

elongation reactions containing ribonucleotides alone, or ribonucleotides plus mock, Wt,

or Cts23 extract were performed (Fig. 12A, Lanes 1-8, 13 and 14). As previously shown,

only the addition of Wt extract is capable of inducing transcript release (Fig. 12A, Lanes

5 and 6 and Lanes 13 and 14, Fig. 9D). Two of the column fractions were capable of

inducing release, the 0.5 M phosphocellulose fraction and the 0.25 M Q-Sepharose



























Fig. 11. Add-back extract titration. A, elongation complexes were generated as
detailed in Fig. 10, washed in 1 M transcription buffer, and transcription was continued
for 20 min in the presence of 0.6 mM ATP, 0.6 mM GTP, 0.6 mM UTP, and 1.2 mM
CTP alone (NTPs). Other reactions were supplemented with increasing concentrations of
mock extract (Mock), Wt extract (Wt), or Cts23 extract (Ts23), as follows: 0.5 Rg (lanes 4
and 5, 12 and 13, and 23 and 24), 3 gg (lanes 6 and 7, 14 and 15, and 25 and 26), 15 jig
(lanes 8 and 9, 16 and 17, and 27 and 28), 30 jg (lanes 10 and 11, 18 and 19, and 29 and
30). B, bound; S, supernatant. B, percent transcript release plotted against the quantity of
mock, Wt, or Cts23 extract.















A
z ,-- Mock ,-- Wt
B SB SBSB SBS BSBS BSBS

ti!r!w


C)
i-.
z T s 2 3
BSBSBSBS BSM
l y ~ I *-2652 nt

800 nt


* 350 nt


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28


-4-NTPs
-0-Mock

-A,-Ts23


0 5 10 15 20 25 30 35
ug extract



















Fig. 12. Wt extract fractionation. A, pulse-labeled elongation complexes were generated as detailed in Fig. 10. Transcript release
was assayed with the addition of 0.6 mM ATP, GTP, UTP, and 1.2 mM CTP (NTPs) or NTPs and 30 gg of mock (Mock), Wt (Wt and
E080798), or Cts23 (Ts23) extract, 1.32 jig of vaccinia virus (vv) His-A18 protein (A18), or 5 Rig of each fraction from the
phosphocellulose and Q-Sepharose columns during a 20-min chase reaction. E080798 was the extract fractionated on the
phosphocellulose and Q-Sepharose columns. B, bound; S, supernatant. B and C, Western blot analysis. Monoclonal a-A18 antibody
was used to probe a 10% SDS-PAGE containing 3.125 jtg of each sample from the phosphocellulose and Q-Sepharose columns, 7.5
jg either Wt extract (E080798) or Cts23 extract (E122297), and 0.3 jig of purified vHis-A18 protein. D, graphic representation of
the percent transcript release for each sample in A.









Phosphocellulose Q-Sepharose
column column


MN t
CO o


BS


BSBS


BS


LO
d 6

BSBS


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36


Phosphocellulose
column
CO 0)-
0 C D


- on w


i m nwt


1 40
IL
20


Q-Sepharose
column



do Ch LO
wC 04 LO
0o S -.66


NT23A18
D 7iS


131uffer A
HP-FT
P25
B3PI
O -FT
Q25
*O1


0-


0
a-
z
BS


I


Goa
o M
S,,,S


C4

B S


B SB


-CU
BS


- 2652 nt
- 800 nt


R









fraction (Fig. 12A, Lanes 23 and 24 and Lanes 31 and 32). These same fractions contain

A18 protein as judged by western blot analysis (Fig. 12B, 0.5M and Fig. 12C, 0.25M).

The phosphocellulose wash fraction, Fig. 12A, Lanes 19 and 20, also showed release in

this experiment. This result was not reproducible in subsequent release experiments done

with the same material. For comparison, Cts23 extract was also fractionated by the same

protocol (data not shown). A18 protein was not detected by western blot in extract from

Cts23-infected cells (Fig. 12B, E122297), nor any Cts23 extract fractions from the

phosphocellulose or Q-Sepharose columns (data not shown). In addition, no significant

release was detected with the addition of fractions from Cts23 extract (data not shown).

The fractionation protocol described here provides circumstantial evidence for the role of

A18 protein in transcript release. However, these are crude fractions that contain many

more proteins than just A18. Conclusive evidence for the role of A18 must be obtained

with a purified fraction or purified protein.

Release Occurs From a Stalled Elongation Complex and Can Be Complemented by
His-A18 and a Cellular Factor

In all of the experiments described above release occurs predominately at the

downstream end of the template where the template is joined to a paramagnetic bead. In

order to eliminate the possibility that the observed release is an artifact due to the

presence of the bead, we conducted experiments designed to promote release in the

middle of a DNA template. We refer to this protocol as a "mid-template" assay. This

assay is designed to reflect the situation in vivo where a transcription complex will

terminate despite the presence of additional template downstream. We accomplished this

by arresting transcription at the end of a 375-nt G-less cassette downstream from the

intermediate G8R promoter present within the 3-kB NpG8G template. Transcription









complexes were assembled on NpG8G during the pre-incubation reaction, pulse-labeled,

washed in high salt transcription buffer, and elongated either in the absence of GTP (with

all other nucleotides present) (data not shown) or in the presence of 3'-OMeGTP and all

other ribonucleotides (Fig. 13A) with additional proteins provided as indicated. The

addition of 3'-OMeGTP arrests the elongation complex at the end of the G-less cassette

where the first GTP is incorporated (Fig. 13A, Lane 1) resulting in the synthesis of an

approximately 400-nt transcript. Addition of Wt extract during the chase reaction

resulted in release of the transcript at the end of the G-less cassette (Fig. 13A, Lanes 5

and 6). Release did not occur with mock or Cts23 extract (Fig. 13A, Lanes 3 and 4,

Lanes 7 and 8). Similar results were obtained when the complex was elongated in the

absence of GTP (data not shown). In other experiments not shown, we attempted to

induce release by first elongating to the end of the G-less cassette in the absence of added

extract and then adding extract to the arrested complex for an additional incubation. We

also tried to induce mid-template release by slowing elongation using reduced

concentrations of UTP. In neither protocol did we observe significant mid-template

release. These results show definitively that release can be induced in the middle of the

template but strongly suggest that release can only be accomplished on a complex that is

stalled. Furthermore, the results indicate that in order to observe release, release factors

need to be present during elongation, before the polymerase stalls.

In order to determine definitively whether A18 is required for transcript release,

we attempted to complement the defect in release activity observed in Cts23 extracts with

the addition of purified His-A18 protein. Pulse-labeled elongation complexes were

formed and assayed for transcript release during an elongation step using purified His-





















Fig. 13. Release occurs from a stalled elongation complex and can be complemented by His-A18 and a cellular factor. A,
pulse-labeled transcription elongation complexes were formed on NpG8G bead-bound template using extract from Wt-infected cells.
Complexes were washed in 1 M transcription buffer, and transcription elongation was continued to the end of the G-less cassette using
0.6 mM ATP, UTP, 1.2 mM CTP, and 0.01 mM 3'-OMeGTP alone (NTPs), or in addition to 30 jig of mock-infected extract (Mock),
Wt extract (Wt), or Cts23 extract (Ts23). Transcripts synthesized in the presence of 3'-OMeGTP are approximately 400 nt in length.
Purified recombinant His-A18 protein was used at 300 ng either alone (A18) or in combination with Cts23 or mock extract. DB, A18
protein storage buffer; B, bound; S, supernatant. B and C, graphic representation of the percent transcript release for each sample in A














U, ~ +
a.. oCf) Ts23 +
S 0 ** i (D Go N


BSBSBSBSBSBSBSBSBSBSBSBS


+ ~Mock +

A18B

BSB SB SB SB SB S


10

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36


0 100 200 300 400 500 600
ng bHisA18 protein


B
30




is,
10



51


- 400 nt
- 350 nt


-&- T23
-e- Mock


NTIs
Mock



G bHilA18


0


AJ


M14"









A 18 protein. His-A18 was expressed in E. coli and purified over nickel and

phosphocellulose columns as described in Chapter 2. The addition of ribonucleotides,

Cts23 extract, or purified His-A18 protein alone to the chase was not sufficient for

transcript release (Fig. 13A, Lanes 11 and 12, Fig. 13B). Addition of increasing amounts

of His-A18 protein to the Cts23 extract resulted in increasing release equivalent to the

levels of His-A18 protein (Fig. 13A, Lanes 15-24, Fig. 13C). As a control, a similar

titration of purified His-J3 protein (J3 is the vaccinia 2'-O-methyltransferase and poly(A)

polymerase processivity factor) expressed in E. coli was tested in combination with Cts23

extract (data not shown). The transcription complexes did not release the nascent RNA

in the presence of His-J3 protein. These results demonstrate that the release defect

observed in Cts23 extract can be complemented by purified A 18 protein.

The results described above show that purified A18 protein is necessary but not

sufficient for transcript release. To determine whether the additional factors required for

release are viral or cellular in nature, extract from mock-infected cells was tested in the

release assay. Mock extract alone does not produce a significant level of released

transcripts (Fig. 13A, Lanes 3 and 4). A titration of His-A 18 in combination with mock

extract induced more efficient release than His-A18 plus Cts23 extract (Fig. 13A,

compare Lanes 27-36 and Lanes 15-24, Fig. 13C). The simplest explanation for these

observations is that a cellular factor(s) is needed in addition to A 18 for transcript release.


Release Requires ATP Hydrolysis

It was shown previously that A18 possesses a DNA-dependent ATPase activity

and that the enzyme can readily use dATP as a substrate rather than ATP (9). We

therefore hypothesize that any stage of transcription that requires A18 would also be









ATP-dependent. Assessing the role of ATP hydrolysis in transcription is complicated by

the requirement for ATP as a substrate for the polymerase during elongation. Therefore,

we examined the ATP-dependence of the release activity by replacing the ATP in the

elongation step of the mid-template assay with the non-hydrolyzable ATP analog,

AMPPNP. AMPPNP can be used as a substrate for the vaccinia RNA polymerase and

substitution results in efficient synthesis of long transcripts (Fig. 14A, compare Lanes 1

and 3). Substitution of ATP with dATP, a hydrolyzable ATP analog that cannot be

efficiently used for synthesis, yielded transcripts that are much shorter in length (Fig.

14A, compare Lanes 1 and 5). Transcription elongation in the presence of dATP can be

rescued with the provision of AMPPNP (Fig. 14A, Lane 7). The combination of dATP

and AMPPNP satisfies the energy requirement and provides a nucleotide capable of

being incorporated into the nascent RNA chain. We then assayed the effect of AMPPNP

substitution on release in combination with mock extract (Fig. 14A, Mock), Wt extract

(Fig. 14A, Wt), or mock extract plus His-A18 protein (Fig. 14A, Mock+A18). As

controls, the level of release in response to a given extract was assayed using ATP or

dATP alone or the combination of AMPPNP and dATP, and quantified as previously

described (Fig. 14A, Lanes 9 and 10, 15 and 16, 17 and 18, 23 and 24, 25 and 26, and 31

and 32, Fig. 14B). Since the extract added during the elongation step contains some

endogenous ATP, substitution of ATP with dATP in these controls did not restrict

elongation as much as elongation in the presence of nucleotides alone. Substitution of

ATP with AMPPNP did not have an effect on the low level of release detected in the

presence of mock extract (Fig. 14A, compare Lanes 9 and 10 and Lanes 11 and 12, Fig.

14B). On the other hand, replacing ATP with AMPPNP severely inhibits transcript





























Fig. 14. Transcript release requires ATP hydrolysis. A, ternary complexes were
formed and elongated as described in Fig. 9. The standard elongation reaction included
0.6 mM ATP, UTP, 0.01 mM 3'-OMeGTP, and 1.2 mM CTP (A, C, G, L). In other
reactions, adenosine analogs AMPPNP (AMPPNP) and dATP (dA or dATP) replaced
ATP as indicated, each at 0.6 mM concentration. Released transcripts (S) were separated
from bound transcripts (B) and analyzed as described previously. B, graphic
representation of the percent transcript release for each sample in A.