u B
`
`EQ 1039?‘. V‘
`Page 1
`" "
`
`BEQ 1039
`Page 1
`
`

`
`GENES
`THIRD EDITION
`
`BENJAMIN LEWIN
`Editor, Cell
`
`JOHN WILEY & SONS
`
`New York
`
`Chichester
`
`Brisbane
`
`Toronto
`
`Singapore
`
`BEQ 1039
`Page 2
`
`

`
`f
`
`' I
`
`Illustrations by John Balbalis with the assistance of the
`Wiley Illustration Department
`
`Copyright (9 1983, 1985, 1987 by John Wiley & Sons, Inc.
`
`All rights reserved . Published simultaneously in Canada.
`
`Reproduction or translation of any part of this work
`beyond that permitted by Section 107 or 1 08 of the
`1976 United States Copyright Act without the permission
`of the copyright owner is unlawful. Requests for
`permission or further information should be addressed to
`the Permissions Department, John Wiley & Sons, Inc.
`
`Library of Congress Cataloging-in-Publication Data:
`
`Lewin, Benjamin.
`Genes.
`
`Includes bibliographies and index.
`I. Title.
`1. Genetics.
`
`QH430.L487 1987
`ISBN 0-471-83278-2
`
`575.1
`
`86-18959
`
`Printed in the United States of America
`
`10987654321
`
`BEQ 1039
`Page 3
`
`

`
`CHAPTER 9
`RNA POLYMERASE-PROMOTER
`INTERACTIONS CONTROL INITIATION
`
`How did RNA come to be the intermediary between
`DNA and protein? Perhaps the first primitive cells made
`no distinction between types of nucleic acid, so that
`what passed for the genome was involved directly in
`both replication and translation. At some point, it may
`have become advantageous to separate translation from
`the genome, so that proteins were synthesized on
`messengers distinct from the genetic material itself.
`It is impossible to say how this separation related in
`time to the development of the other ribonucleic acid
`components involved in translation , but it is striking that
`RNA is present in the ribosome as well as constituting
`the tRNA adaptor. (It would not be surprising if the role
`of RNA in the ribosome was more prominent in the
`past than is apparent today.) Perhaps RNA was the
`original nucleic acid, and its ubiquitous presence today
`is merely a recollection of its former activity.
`All this speaks to the fact that RNA plays central
`roles in gene expression, not merely in constituting the
`messenger, but also in providing the means for its
`translation into protein . In a sense, these roles repre(cid:173)
`sent the various specific interests of an RNA conglom-
`
`erate generally concerned with gene expression , but
`with several different functions .
`The production of each type of RNA has a common
`origin: transcription of DNA. In the case of mRNA, the
`product is an intermediate whose function requires
`translation . In the case of tRNA and rRNA, the tran(cid:173)
`scriptional product itself fulfills the final function .
`To transcribe or not to transcribe: that is the ques(cid:173)
`tion? Transcription is the principal stage at which gene
`expression is controlled. The first (and sometimes the
`only) step in control is the decision on whether or not
`to transcribe a gene. In considering the various stages
`of transcription , we should therefore keep in mind the
`opportunities they offer for regulating gene activity.
`(Transcription is not the only means by which RNA
`can be synthesized. Viruses with RNA genomes specify
`enzymes able to synthesize RNA on a template itself
`consisting of RNA. Such reactions produce mRNAs
`coding for proteins needed in the infective cycle [RNA
`transcription] and provide genomic RNAs to perpetuate
`the infective cycle [RNA replication] . Yet a further reac(cid:173)
`tion is possible with the retroviruses, in which viral RNA
`
`183
`
`BEQ 1039
`Page 4
`
`

`
`CONTROLLING GENE EXPRESSION BY TRANSCRIPT~
`
`184
`
`Initiation
`RNA polymerase
`binds to
`duplex DNA
`
`.
`RNA synthesis.
`starts with unwmdmg
`of DNA
`
`Elongation
`RNA is synthesized
`by base pairing
`with one strand
`of DNA
`
`Unwound region
`moves along
`DNA
`
`Unwound region
`reaches end
`of gene
`
`5'
`
`Termination
`RNA is freed
`entirely and
`DNA duplex
`reform s
`
`?\.YAVJ~~YJ 00 ~~00\.YJ\.\IJ\.'U"YP"YJ\X/\YJ '"
`____ !_
`
`Figure 9.1
`RNA is synthesized by base pair(cid:173)
`ing with one strand of DNA in a
`region
`that
`is transiently un(cid:173)
`wound. As the region of unwind(cid:173)
`ing moves, the DNA duplex re(cid:173)
`forms behind it, displacing the
`RNA in the form of a single poly(cid:173)
`nucleotide chain.
`
`serves as a template for reverse transcription to pro(cid:173)
`duce a DNA complement.)
`
`TRANSCRIPTION IS CATALYZED
`BY RNA POLYMERASE
`
`Transcription involves synthesis of an RNA chain rep(cid:173)
`resenting one strand of a DNA duplex. By "represent(cid:173)
`ing" we mean that the RNA is identical in sequence
`
`with one strand of the DNA; it is complementary to the
`other strand, which provides the template for its syn(cid:173)
`thesis.
`Transcription takes place by the usual process of
`complementary base pairing, catalyzed by the enzyme
`RNA polymerase. The reaction can be divided into
`the three stages illustrated in Figure 9.1.
`
`• Initiation begins with the binding of RNA polymer(cid:173)
`ase to the double-stranded DNA. To make the tern-
`
`BEQ 1039
`Page 5
`
`

`
`CHAPTER 9 RNA POLYMERASE-PROMOTER INTERACTIONS CONTROL INITIATION
`
`185
`
`plate strand available for base pairing with ribonu(cid:173)
`cleotides, the strands of DNA must be separated.
`The unwinding is a local event that begins at the site
`bound by RNA polymerase. The initiation stage is
`completed when the first nucleotide is incorporated.
`The entire sequence of DNA needed for these
`reactions is called the promoter. The site at which
`the first nucleotide is incorporated is called the start(cid:173)
`site or startpoint.
`
`• Elongation describes the phase during which nu(cid:173)
`cleotides are covalently added to the 3' end of the
`growing polynucleotide chain. Successive bases are
`added to the RNA chain , forming an RNA-DNA hy(cid:173)
`brid in the unwound region .
`To continue synthesis, the enzyme moves along
`the DNA, unwinding the double helix to expose a
`new segment of the template in single-stranded con(cid:173)
`dition. As it moves, the RNA that was made previ(cid:173)
`ously is displaced from the DNA template strand,
`which pairs with its original partner to reform the
`double helix.
`Thus elongation involves the movement along DNA
`of a short segment that is transiently unwound, ex(cid:173)
`isting as a hybrid RNA-DNA duplex and a displaced
`single strand of DNA.
`
`• Termination involves recognition of the point at which
`no further bases should be added to the chain. To
`terminate transcription, the formation of phospho(cid:173)
`diester bonds must cease , and the transcription
`complex must come apart. When the last base is
`added to the RNA chain, the RNA-DNA hybrid is
`disrupted, the DNA reforms in duplex state, and the
`enzyme and RNA are both released from it. The
`sequence of DNA required for these reactions is called
`the terminator.
`
`Originally defined simply by its ability to incorporate
`nucleotides into RNA under the direction of a DNA
`template, the enzyme RNA polymerase now is seen
`as part of a more complex apparatus involved in tran(cid:173)
`scription. The ability to catalyze RNA synthesis defines
`the minimum component that can be described as
`RNA polymerase. It supervises the base pairing of the
`substrate ribonucleotides with DNA and catalyzes the
`formation of phosphodiester bonds between them.
`But ancillary activities may be needed to initiate and
`
`to terminate the synthesis of RNA, when the enzyme
`must associate with , or dissociate from , a specific site
`on DNA. The analogy with the division of labors be(cid:173)
`tween the ribosome and the protein synthesis factors
`is obvious. Sometimes it is difficult to decide whether
`a particular protein that is involved in transcription at
`one of these stages should be considered as part of
`the "RNA polymerase" or as an ancillary factor.
`All the components involved in elongation are nec(cid:173)
`essary for initiation and termination. Genes may differ
`in their dependence on additional polypeptides at the
`initiation and termination stages. Some of these addi(cid:173)
`tional polypeptides may be needed at all genes, but
`others may be needed specifically for the initiation or
`termination of particular genes. An additional polypep(cid:173)
`tide needed to recognize all promoters (or terminators)
`is likely to be classified as part of the enzyme. A poly(cid:173)
`peptide needed only for the initiation (or termination)
`of particular genes is likely to be classified as an an(cid:173)
`cillary control factor.
`With bacterial enzymes, it is possible to begin to
`define the roles of individual polypeptides in the stages
`of transcription. With eukaryotes, the enzymes are less
`well purified, and the actual enzymatic activities have
`yet to be completely resolved from the relatively crude
`preparations. Ironically enough, in eukaryotes we have
`begun to isolate ancillary factors needed to initiate or
`terminate particular genes, while the basic polymerase
`preparation itself remains rather poorly characterized .
`
`BACTERIAL RNA POLYMERASE CONSISTS
`OF CORE ENZYME AND SIGMA FACTOR
`
`A single type of RNA polymerase is responsible for all
`synthesis of mRNA, rRNA and tRNA in bacteria. The
`total number of RNA polymerase molecules present in
`an E. coli cell is ~ 7000. Many of them are actually
`engaged in transcription; probably between 2000 and
`5000 enzymes are synthesizing RNA at any one time ,
`the number depending on the growth conditions.
`The best characterized RNA polymerase is that of
`E. coli, but its structure is similar in all other bacteria
`studied. The complete enzyme or holoenzyme has
`a molecular weight of ~ 480,000 daltons. Its subunit
`composition is summarized in Table 9.1.
`
`BEQ 1039
`Page 6
`
`

`
`186
`
`Table 9.1
`E coli RNA polymerase has four types of subunit.
`
`CONTROLLING GENE EXPRESSION BY TRANSCRIPTION
`
`Subun it
`
`Gene
`
`Number
`
`Mass (d altons)
`
`Location
`
`Possible Fun c ti ons
`
`Q
`f3
`{3'
`(J
`
`rpoA
`rp oB
`rpoC
`rp oD
`
`2
`1
`1
`1
`
`40,000 each
`155,000
`160, 000
`85,000
`
`co re enzy me
`c ore enzyme
`core enzyme
`sig m a factor
`
`p romote r b inding
`nu cleotide bindi n g
`templ ate bind ing
`in itiation
`
`The a, f3, and f3' subunits have rather constant sizes in different bactenal spec1es; the a vanes more widely, from 32,000 to 92,000. The
`enzyme has a rather elongated structure, with a maximum dimension of 15 nm (one turn of the DNA double helix is 3.4 nm).
`
`The holoenzyme (a d3f3' cr) can be separated into two
`components, the core enzyme (a2~~ ' ) and the sigma
`factor (the cr polypeptide) . The names reflect the fact
`that only the holoenzyme can initiate transcription; but
`then the sigma "factor" is released, leaving the core
`enzyme to undertake elongation. Thus the core en(cid:173)
`zyme has the ability to synthesize RNA on a DNA tem(cid:173)
`plate, but cannot initiate transcription at the proper
`sites.
`Core enzyme starts transcription at the separated
`DNA strands of an initiation complex. As the enzyme
`moves along the template extending the RNA chain ,
`the region of local unwinding moves with it. The en(cid:173)
`zyme ?overs ~ 60 bp of DNA; the unwound segment
`compn~es only a small part of this stretch, < 17 bp
`accordmg to the overall extent of unwinding.
`. As the DNA unwinds to free the template, each of
`1ts strands probably enters a separate site in the en(cid:173)
`zyme str~cture . As Figure 9.2 indicates, the template
`~trand Will be free just ahead of the point at which the
`nbonucleotide is being added to the RNA chain and it
`will exist ~s a DNA-RNA hybrid in the region ' where
`R~A ha~ JUSt been synthesized. The length of the hy(cid:173)
`bnd reg1on may be a little shorter than the stretch of
`unwound DNA. Probably the RNA-DNA hybrid is ~ 12
`bp long.
`As the enzyme leaves the area, the DNA duplex
`reform~, and the RNA is displaced as a free poly(cid:173)
`nucleotide chain . About the last 50 ribonucleotides
`added to a growing chain are complexed with DNA
`and/or enzyme at any moment.
`. W~ still do not really understand the topology of un(cid:173)
`~;ndm~ _and rewinding during transcription, but the ability
`punfled RNA polymerase to transcribe double(cid:173)
`stranded DNA in vitro implies that the reaction depends
`
`on an intrinsic property of the enzyme. Unwinding and
`rewinding requires the strands of DNA to revolve about
`one another. One possibility is that the DNA revolves
`in the unwinding sense ahead of the enzyme, and re(cid:173)
`volves in the opposite sense beh ind it. This could re(cid:173)
`quire assistance in vivo from other enzymatic activities
`to adjust the topology of the DNA.
`
`Figure 9.2
`Bacterial RNA polymerase covers - 60 bp of DNA and has sev(cid:173)
`eral active centers.
`During elongation the reacting groups are held in two sites. The
`location occupied by the incoming nucleoside triphosphate is the
`elongation nucleotide site. The position of the last nucleotide added
`to the chain defines the primer terminus site.
`When transcription is initiated, of course, there is no primer ter(cid:173)
`minus, and the very first nucleotide (usually a purine) enters the
`initiation nucleotide site, which must largely overlap with the primer
`terminus site. The first nucleotide incorporated into the chain retains
`all its 5' triphosphate residues.
`The length of the region of unwinding is exaggerated ( - 4 fold) for
`the purposes of illustration.
`
`Direc ti on of enzyme movement
`
`10 bp
`->1
`'
`
`Primer term inus and
`initiati on nucleot ide si te
`
`BEQ 1039
`Page 7
`
`

`
`CHAPTER 9 RNA POLYMERASE-PROMOTER INTERACTIONS CONTROL INITIATION
`
`187
`
`5' end of
`polynucleotide
`cha in
`
`0
`I
`O= P - o-
`
`1
`0
`I
`O= P- 0-
`r--- o.--
`1
`o )&~ -'o-
`I
`o
`I
`CH2
`
`1
`1
`I
`I
`I
`o
`I
`I ____ _ _ _ ./
`H
`
`OH
`
`Figure 9.3
`Phosphodiester bond formation involves a hydrophilic attack
`by the 3'- 0H group of the last nucleotide of the chain on the
`5' triphosphate of the incoming nucleotide, with release of
`pyrophosphate.
`
`All nucleic acids are synthesized from nucleoside 5'
`triphosphate precursors. Figure 9.3 shows the con(cid:173)
`densation reaction between the 5' triphosphate group
`of the incoming nucleotide and the 3'-0H group of
`the last nucleotide to have been added to the chain .
`The incoming nucleotide loses its terminal two phos(cid:173)
`phate groups (-y and (3); its ex group is used in the
`phosphodiester bond linking it to the previous nucleotide.
`The core enzyme must hold the two reacting groups
`in the proper apposition for phosphodiester bond for(cid:173)
`mation ; then , once they are covalently linked, it moves
`one base farther along the DNA template so that the
`reaction can be repeated. The reaction rate is fast,
`- 40 nucleotides/second at 3JDC (see Chapter 8) .
`The acceptability of an incoming nucleotide is judged
`by its base pairing with the template strand of DNA,
`an action apparently supervised by the enzyme. Prob(cid:173)
`ably the site has a structure that allows phosphodiester
`bond formation to proceed only when the nucleotide is
`properly base paired with DNA. Presumably the nu-
`
`cleotide is expelled if its ability to base pair is deemed
`inadequate; then another can enter.
`Our knowledge of the topology of the core enzyme
`is really very primitive, and the best we can do at present
`is to make a diagrammatic representation of the sites
`defined by the various enzymatic functions , as illus(cid:173)
`trated in Figure 9.2. None of these sites has yet been
`physically located on the polypeptide subunits. How(cid:173)
`ever, there is some general information about the roles
`of individual subunits .
`Two types of antibiotic both act on the f3 subunit, as
`defined by the location of mutations conferring re(cid:173)
`sistance (see Table 9.3) . The rifamycins (of which
`rifampicin is the most used) prevent initiation , acting
`prior to formation of the first phosphodiester bond .
`Streptolydigins inhibit chain elongation . The f3 sub(cid:173)
`unit is the target for both types of antibiotic; also it is
`labeled by certain affinity analogs of the nucleoside
`triphosphates. Together these results suggest that the
`f3 subunit may be involved in binding the nucleotide
`substrates.
`Heparin is a polyanion that binds to the f3' subunit
`and inhibits transcription in vitro. Heparin competes
`with DNA for initially binding the polymerase. The f3'
`subunit is the most basic, which would fit with a role
`in template binding.
`The ex subunit has no known role. However, when
`phage T4 infects E. coli, the ex subunit is modified by
`ADP-ribosylation of an arginine. The modification is
`associated with a reduced affinity for the promoters
`formerly recognized by the holoenzyme, so the ex sub(cid:173)
`unit might play a role in promoter recognition .
`These assignments of individual functions are very
`primitive; probably each subunit contributes to the ac(cid:173)
`tivity of the core enzyme as a whole, and we cannot
`compartmentalize its actions.
`Why does bacterial RNA polymerase require a large,
`multimeric structure? The existence of much smaller
`RNA polymerases, comprising single polypeptide chains
`coded by certain phages, demonstrates that the ap(cid:173)
`paratus required for RNA synthesis can be much smaller
`than that of the host enzyme.
`These enzymes give some idea of the "minimum"
`apparatus necessary for transcription. They recognize
`a very few promoters on the phage DNA; and they
`have no ability to change the set of promoters to which
`they respond . Thus they are limited to the intrinsic abil-
`
`BEQ 1039
`Page 8
`
`

`
`188
`
`ity to recognize a very few specific DNA binding se(cid:173)
`quences and to synthesize RNA. How complex are
`they?
`The RNA polymerases coded by the related phages
`T3 and T7 are single polypeptide chains of -11,000
`daltons each. They synthesize RNA very rapidly (at
`rates of - 200 nucleotides/second at 37°C). The initi(cid:173)
`ation reaction shows very little variation.
`By contrast, the enzyme of the host bacterium can
`transcribe any one of many (> 1 000) transcription units.
`Some of these units are transcribed directly, with no
`further assistance. But many units can be transcribed
`only in the presence of further protein factors. Some
`of these factors are specific for a single transcription
`unit; others are involved in coordinating transcription
`from many units. Certain phages induce general
`changes in the affinity of host RNA polymerase, so that
`it stops recognizing host genes and instead initiates at
`phage promoters.
`So the host enzyme requires the ability to interact
`with a variety of host and phage functions that modify
`its intrinsic transcriptional activities. The complexity of
`the enzyme may therefore at least in part reflect its
`need to interact with a multiplicity of other factors, rather
`than any demand inherent in its catalytic activity.
`
`EUKARYOTIC RNA POLYMERASES
`CONSIST OF MANY SUBUNITS
`
`The transcription apparatus of eukaryotic cells is more
`complex and less well defined than that of bacteria.
`There are three nuclear RNA polymerases, occupying
`different locations, each with a complex subunit struc(cid:173)
`ture. Each enzyme is responsible for transcribing a
`different class of genes. Their general properties are
`defined in Table 9.2.
`
`Table 9.2
`Eukaryotic nuclei have three RNA polymerases.
`
`Enzyme
`
`Location
`
`Product
`
`CONTROLLING GENE EXPRESSION BY TRANSCRIPTION
`
`The most prominent RNA-synthesizing activity is the
`enzyme RNA polymerase I, which resides in the nu(cid:173)
`cleolus and is responsible for transcribing the genes
`coding for rRNA. It accounts for most cellular RNA
`synthesis.
`The other major enzyme is RNA polymerase II, lo(cid:173)
`cated in the nucleoplasm (the part of the nucleus ex(cid:173)
`cluding the nucleolus). It represents most of the rest
`of the cellular activity and is responsible for synthes(cid:173)
`izing heterogeneous nuclear RNA (hnRNA), the pre(cid:173)
`cursor for mANA.
`A minor enzyme activity is RNA polymerase Ill. This
`nucleoplasmic enzyme synthesizes tRNAs and many
`of the small nuclear RNAs.
`Inhibitors of transcription have been useful in distin(cid:173)
`guishing between the enzymes. Different inhibitors act
`on prokaryotic and eukaryotic enzymes. The properties
`of some common inhibitors are summarized in Table
`9.3.
`A major distinction between the eukaryotic enzymes
`is drawn from their response to the bicyclic octapeptide
`a -amanitln. In cells from origins as divergent as ani(cid:173)
`mals, plants, and insects, the activity of RNA poly(cid:173)
`merase II is rapidly inhibited by low concentrations of
`a-amanitin. In cells from all origins, the RNA polymer(cid:173)
`ase I enzyme is not inhibited. The response of RNA
`polymerase Ill to a-amanitin has not been so well con(cid:173)
`served; in animal cells it is inhibited by high levels, but
`in yeast and insects it is not inhibited.
`The crude enzyme activities all are large proteins,
`appearing as aggregates of 500,000 daltons or more.
`Their subunit compositions are complex. Each enzyme
`has two large subunits, generally one - 200,000 dal(cid:173)
`tons and one - 140,000 daltons. There are < 10 smaller
`subunits, ranging in size from 10,000 to 90,000 dal(cid:173)
`tons. We do not know whether any of the subunits
`found in the different enzymes are the same.
`
`RNA polymerase
`RNA polymerase
`RNA polymerase
`
`I
`II
`Ill
`
`nucleolus
`nucleoplasm
`nucleoplasm
`
`ribosomal RNA
`hnRNA
`small RNA
`
`Relative
`Activity
`
`50-70%
`20-40%
`- 10%
`
`r.t-Amanitin
`Sensitivity
`
`not sensitive
`sensitive
`species-specific
`
`BEQ 1039
`Page 9
`
`

`
`CHAPTER 9 RNA POLYMERASE-PROMOTER INTERACTIONS CONTROL INITIATION
`
`189
`
`Table 9.3
`Inhibitors of transcription act preferentially on particular enzymes.
`
`In hibitor
`
`Target Enzyme
`
`Inhibitory Action
`
`Rifamycin
`Streptolydigin
`Actinomycin 0
`a-Amanitin
`
`bacterial holoenzyme
`bacterial core enzyme
`eukar)'otic pol I
`eukaryotic pol II
`
`binds to (3 to prevent initiation
`binds to (3 to prevent elongation
`binds to DNA & prevents elongation
`binds to RNA polymerase II
`
`Do the enzyme preparations represent the basic
`transcription apparatus, essentially similar in all cells
`and subject to regulation by further protein factors? Or
`do they include such factors as well as a basic catalytic
`apparatus?
`Because it is not yet possible to reconstitute active
`RNA polymerase from the subunits of any of these
`enzymes, we have no evidence as to whether all of
`the protein subunits are integral parts of each enzyme.
`We do not know which subunits may represent cata(cid:173)
`lytic activities and whether others may be involved in
`regulatory functions.
`The route to investigating this question is to use iso(cid:173)
`lated enzyme preparations to transcribe defined tem(cid:173)
`plates in vitro. Several heterologous reactions have
`been characterized , in which an RNA polymerase II
`preparation from one cell type and species is used to
`transcribe a gene that is active in a different cell type
`and species. The success of such experiments indi(cid:173)
`cates that neither tissue- nor species-specific features
`are involved in promoter recognition per se.
`Of course, this conclusion does not exclude the pos(cid:173)
`sibility that further protein factors or other sequences
`are involved in modulating the reaction (especially in(cid:173)
`creasing its efficiency) in the natural situation. Factors
`have been found that are necessary to allow particular
`RNA polymerases to initiate transcription of a subset
`of their target genes. Sets of factors may be needed
`to transcribe particular groups of genes.
`The overall complexity of the eukaryotic transcription
`apparatus is being defined only system by system; all
`we can say at present is that multimeric enzymes are
`needed for each of the three classes of RNA synthesis.
`Whether all the components of these enzymes are es(cid:173)
`sential, and how many other proteins are needed, re(cid:173)
`mains to be seen.
`
`Because of these uncertainties, relatively crude, "dirty"
`systems may offer more chance of characterizing tran(cid:173)
`scription in vitro than purified "clean" systems; too much
`purification may remove the very factors that we need
`to characterize!
`The RNA polymerase activities of mitochondria and
`chloroplasts appear to be smaller and distinct from the
`nuclear enzymes. Of course, the organelle genomes
`are much smaller, the resident polymerase needs to
`transcribe only a few genes, and the control of tran(cid:173)
`scription is likely to be very much simpler (if existing
`at all). So these enzymes may be analogous to the
`phage enzymes that have a single fixed purpose and
`do not need the ability to respond to a more complex
`environment.
`
`BACTERIAL SIGMA FACTOR CONTROLS
`BINDING TO DNA
`
`The function of the sigma factor is to ensure that bac(cid:173)
`terial RNA polymerase binds stably to DNA only at
`promoters, not at other sites.
`The core enzyme itself has an affinity for DNA, in
`which electrostatic attraction between the basic protein
`and the acidic nucleic acid plays a major role. Probably
`this general ability to bind to any DNA, irrespective of
`its particular sequence, is a feature of all proteins that
`have specific binding sites on DNA (see Chapter 1 0).
`Any sequence of DNA that is bound by RNA poly(cid:173)
`merase in this general binding reaction is described as
`a loose binding site. The enzyme·DNA complex is
`described as closed, because the DNA remains strictly
`in the double-stranded form. A closed complex is sta(cid:173)
`ble; the half-life for dissociation of the enzyme from
`DNA is - 60 minutes.
`
`BEQ 1039
`Page 10
`
`BEQ 1039
`Page 10
`
`

`
`190
`
`Sigma factor introduces a major change in the affin(cid:173)
`ity of RNA polymerase for DNA. The holoenzyme has
`a drastically reduced ability to recognize loose bind(cid:173)
`ing sites-that is, to bind to any general sequence of
`DNA. The association constant for the reaction is re(cid:173)
`duced by a factor of - 10,000 and the half-life of the
`complex is < 1 second. Thus sigma factor destabilizes
`the general binding ability very considerably.
`But sigma factor also confers the ability to recognize
`specific binding sites. The holoenzyme binds to pro(cid:173)
`moters very tightly, with an association constant in(cid:173)
`creased from that of core enzyme by (on average) 1 000
`times and a half-life of several hours.
`The association constant can be quoted only as an
`average, because there is (roughly) a hundredfold var(cid:173)
`iation in the rate at which the holoenzyme binds to
`different promoter sequences; this is an important fac-
`
`Figure 9.4
`Initiating transcription requires several steps, during which a
`closed binary complex is converted to an open form and then
`into a ternary complex.
`
`Hol oenzyme
`
`I DNA T bmding
`
`~
`A.YA'YA.YAYAYA'0,_~YAYJW...'YA.YA'C'!: Open bi nary complex
`~
`I Phosphod iester
`ybond format ion
`
`A.YA'YA.YA'Vl\.~~YJW...YAYA.YA'C'!: Sigma factor released
`~ after first bond
`formation; core
`+
`enzyme continues
`(G)
`RNA synthesis
`
`CONTROLLING GENE EXPRESSION BY TRANSCRIPTION
`
`tor in determining the efficiency of a promoter in ini(cid:173)
`tiating transcription .
`Recognition of promoters by holoenzyme passes
`through two stages, illustrated in Figure 9.4. The holo(cid:173)
`enzyme·promoter reaction starts in the same way as
`the loose binding reaction, by forming a closed com(cid:173)
`plex. But then this complex is converted into an open
`complex by the "melting" of a short region of DNA
`within the sequence bound by the enzyme. The series
`of events leading to formation of an open complex is
`called tight binding.
`Both the closed and open associations of RNA po(cid:173)
`lymerase with DNA are described as binary com(cid:173)
`plexes. The next step is to incorporate the first two
`nucleotides; then a phosphodiester bond forms be(cid:173)
`tween them. This creates a ternary complex of po(cid:173)
`lymerase·DNA·nascent RNA. The ternary complex forms
`extremely rapidly when RNA polymerase finds a pro(cid:173)
`moter, so the binary complex has an exceedingly tran(cid:173)
`sient existence.
`Sigma factor is involved only in initiation. It is re(cid:173)
`leased from the core enzyme when RNA synthesis has
`been initiated.
`The core enzyme in the ternary complex is very tightly
`bound to DNA. It is essentially "locked in" until elon(cid:173)
`gation has been completed. When transcription ter(cid:173)
`minates, the core enzyme is released from DNA as a
`free protein tetramer. It must then find another sigma
`factor in order to undertake a further cycle of transcrip(cid:173)
`tion. The ratio of sigma factors to core enzymes is
`about one third.
`RNA polymerase may find promoters on DNA by the
`process of trial and error illustrated in Figure 9.5.
`The excess core enzyme exists largely in the form
`of closed loose complexes, because the enzyme en(cid:173)
`ters into them rapidly and leaves them slowly.
`By contrast, the holoenzyme very rapidly associates
`with, and dissociates from, loose binding sites. So it is
`likely to continue to make and break a series of closed
`complexes in an agitated manner until (by chance) it
`encounters a promoter. Then its recognition of the spe(cid:173)
`cific sequence will allow tight binding to occur by for(cid:173)
`mation of an open complex.
`Three steps are needed for RNA polymerase to move
`from one binding site to another on DNA. It must dis(cid:173)
`sociate from the first binding site, find the second site,
`and associate with it. Movement from one site to an-
`
`BEQ 1039
`Page 11
`
`BEQ 1039
`Page 11
`
`

`
`CHAPTER 9 RNA POLYMERASE-PROMOTER INTERACTIONS CONTROL INITIATION
`
`191
`
`...----------- ,
`I
`I
`I
`I
`I
`
`Fast l +
`t !Fast
`
`I Yery
`_.rast
`
`W..YAYAYAYAYAYAYAYAYAYAYJ\..YAYAYAYA~
`+
`d~
`W..YAYAYAYAYAYAYAYAYAYAYJ\..YAYAYAYA~
`~
`1 Very
`+fast
`
`:
`
`'W\Y/\_Y/\_Y/\_Y/\_Y/\_e
`
`YA_Y/\_Y/\_Y/\_Y/\_'0\
`
`moving about by leaving one binding site and diffusing
`to another, the enzyme is likely to take hold of one
`sequence of DNA, exchange it very rapidly for another,
`and continue to exchange sequences in this promis(cid:173)
`cuous manner until a promoter is found. Then the en(cid:173)
`zyme forms a stable, open complex, after which initi(cid:173)
`ation occurs. The search process becomes much faster
`because association and dissociation are virtually si(cid:173)
`multaneous, and time is not spent commuting between
`sites.
`The existence of a cycle in which sigma factor and
`core enzyme come together only temporarily solves
`the dilemma of RNA polymerase in reconciling its needs
`for initiation with those for elongation. It is a dilemma
`because initiation requires tight binding only to partic(cid:173)
`ular sequences (promoters), while elongation requires
`close association with a// sequences along which the
`enzyme must progress.
`Core enzyme has a high intrinsic affinity for DNA,
`which is increased by the presence of nascent RNA.
`But its affinity for loose binding sites remains too high
`to allow the enzyme to find promoters efficiently; the
`associations and dissociations involved in the trial and
`error of finding a tight binding site could take many
`hours.
`By reducing the stability of the loose complexes , sigma
`allows the process to occur much more rapidly ; and by
`stabilizing the association at tight binding sites, the fac(cid:173)
`tor drives the reaction irreversibly into the formation of
`open complexes. To avoid becoming paralyzed by its
`specific affinity for the promoter, the enzyme releases
`sigma, and thus reverts to a general affinity for all DNA,
`irrespective of sequence, that suits it to continue
`transcription.
`How does sigma change the enzyme so that pro(cid:173)
`moters are specifically recognized? As an independent
`polypeptide, sigma does not seem to bind DNA, but
`when holoenzyme forms a tight binding complex , O"
`contacts the DNA in the region of the initial melting .
`The inability of free sigma factor to recognize promoter
`sequences may be important: if cr could freely bind to
`promoters, it might block holoenzyme from initiating
`transcription . We do not know what role the core sub(cid:173)
`units play in promoter recognition ; sigma may change
`the conformation of core enzyme so that its ability to
`recognize DNA is altered, but the sequence specificity
`of this effect is not clear.
`
`'------:'"':~~~~-:---~
`
`Figure 9.5
`Sigma factor and core enzyme recycle at different points in
`transcription.
`Sigma factor is released as soon as a ternary complex has formed
`at an initiation site; it becomes available for use by another core
`enzyme. The core enzyme is released at termination; it must either
`find a sigma · and form a holoenzyme that can bind stably only at
`promoters or it