Lennarz W.J., Lane M.D. (eds.) Encyclopedia of Biological Chemistry. Four-Volume Set . V. 4

Подождите немного. Документ загружается.

Translation Termination

and Ribosome Recycling

Nadja Koloteva-Levin and Mick F. Tuite

University of Kent, Canterbury, UK

The decoding of genetic information stored in an mRNA

molecule is initiated by a ribosome locating the translation

start codon AUG. The codons that deﬁne the encoded

polypeptide chain are then read sequentially by the ribosome

with the amino acids being delivered to the ribosome by

tRNAs. The chain of amino acids is therefore assembled as

directed by the order of codons in the mRNA which in turn are

recognized by the anticodon sequence of the tRNA. Trans-

lation termination is the last stage in this process and results in

the release of the completed polypeptide chain from the

ribosome once the end of the coding sequence has been

reached. The trigger for the termination process is the

appearance of one of the three possible stop codons at

the decoding center of the ribosome. Following termination

the ribosome dissociates from the mRNA and is recycled to

take part in a new protein synthesis cycle.

Translation Termination Apparatus

in Prokaryotes and Eukaryotes

The facilitators of the ﬁnal stage of the translation cycle

are the protein release factors (RFs). Translation

termination occurs when one or other of the three stop

codons—UAA, UAG, or UGA—at the end of the

mRNA’s coding sequence are positioned at the riboso-

mal “acceptor” (A) site, i.e., at the decoding site within

the small ribosomal subunit (Figure 1, step 1). The stop

codon is recognized by an individual RF or by an RF

complex that then triggers hydrolysis of peptidyl-tRNA

from the nascent polypeptide chain positioned at the

adjacent ribosomal “peptidyl” (P) site (Figure 1, step 2).

This is mediated by the peptidyl-transferase center

(PTC) of the large ribosomal subunit transferring the

peptidyl group to water rather than to an amino-

acyl tRNA.

RFs fall into one of two classes: class-I RFs recognize

and bind to the stop codon while class-II RFs facilitate

the release of the class-I factor after peptidyl-tRNA

hydrolysis (Figure 1, step 3). In the ﬁnal stage, a

ribosome recycling factor (RRF) triggers the disassembly

of the terminated ribosome from the mRNA and the

release of the deacylated tRNA from the P ribosome

(Figure 1, step 4). Table I summarizes the known protein

factors involved in translation termination.

The ribosome also plays an active role in the process

of termination: interaction of the small ribosomal

subunit RNAs and ribosomal proteins with the stop

codon and a class-I RF ensures that the stop codon is

accurately decoded at the A site, whereas the interaction

of large ribosomal subunit RNAs with the RF or the RF

complex facilitates the catalysis of peptidyl-tRNA

hydrolysis at the P site via the ribosomal PTC.

Release Factors

PROKARYOTIC POLYPEPTIDE CHAIN

RELEASE FACTORS

In bacteria, translation termination is controlled by

three different RFs (Table I). Two class-I protein release

factors, RF1 and RF2, each decode two of three stop

codons, UAA or UAG (RF1) and UAA or UGA (RF2).

Recognition of the stop codon by the RFs is mediated via

a conserved tripeptide motif: Pro-Ala-Thr (PAT) in RF1

and Ser-Pro-Phe (SPF) in RF2. These tripeptides are

referred to as peptide anticodons. The other important

functional domain of class-I RFs contains the highly

conserved amino acid motif Gly-Gly-Gln (GGQ) and it

is this domain that triggers hydrolysis of the protein-

tRNA bond. An understanding of how RFs trigger the

release of the completed polypeptide from the tRNA at

the P site has come from a study of RF2 and its

interaction with the terminating ribosome. When RF2

binds to the ribosome with a stop codon positioned at

the A site, RF2 changes its three-dimensional confor-

mation such that the domain with the conserved

“peptide anticodon” (SPF) interacts with the mRNA at

the decoding center, and the GGQ-containing

domain comes in the contact with the ribosomal PTC

to trigger hydrolysis of the peptidyl-tRNA linkage

(Figure 2).

The single class-II RF in bacteria, namely RF3, is a

GTP-binding protein that accelerates dissociation of

either RF1 or RF2 from the ribosomal A-site after

release of the completed polypeptide chain from the

ribosome. RF3 bound to GDP accesses the ribosome

which, in complex with RF1 or RF2, acts as guanine

nucleotide exchange factors (GEFs) and triggers dis-

sociation of GDP from RF3. This leads to the formation

of a stable ribosome-RF1 (or RF2-) RF3 complex.

Hydrolysis of the peptidyl-tRNA linkage triggered by

RF1 or RF2 allows GTP binding to RF3. This induces an

altered RF3 conformation with a high afﬁnity for the

ribosome and leads to rapid dissociation of RF1 or RF2

from the termination complex. To leave the ribosome,

RF3 requires GTP hydrolysis, which converts it to the

GDP-bound form of RF3 which has a lower afﬁnity for

the ribosome. Once RF3 leaves the ribosome, it is ready

to enter the next translation cycle.

EUKARYOTIC CHAIN RELEASE FACTORS

In contrast to bacteria, in eukaryotic cells translation is

terminated by a single heterodimer consisting of two

different RFs, eRF1 and eRF3. eRF1 is a class-I RF that

decodes all three stop codons and triggers peptidyl-

tRNA hydrolysis by the ribosome to release the nascent

polypeptide. In eRF1, the stop codon recognition site is

located close to the amino terminus of the protein

molecule in a region that contains an evolutionarily

conserved tetrapeptide sequence, Asn-Ile-Lys-Ser

(NIKS). This sequence may be functionally equivalent

to the bacterial RF peptide anticodon. The GGQ

motif found in bacterial RF1 and RF2 is located in a

FIGURE 1 An overview of the mechanism of translation termination. The four key steps of the process triggered by the arrival of a stop codon at

the ribosomal A site are shown. The peptidyl tRNA is positioned at the ribosomal peptidyl P site. The major protein factors associated with each

step are indicated: RF1, class-I release factor; RF2, class-II release factor; RRF, ribosome recycling factor.

TABLE I

Protein Factors Involved in Translation Termination

Organisms RF Function

Archaebacteria

aRF1 Class-I RF, recognizes UAA/UAG/UGA

Prokaryotes

RF1 Class-I RF, recognizes UAA/UAG in

mRNA

RF2 Class-I RF, recognizes UAA/UGA in

mRNA

RF3 Class-II RF, GTPase, accelerates dissociation

of termination complex from the ribosome

RRF Ribosome recycling factor

Eukaryotes

eRF1 Class-I RF, recognizes UAA/UAG/UGA

eRF3 Class-II RF, GTPase, increases termination

efﬁciency, precise function unknown

No RRF homologue has yet been described for eukaryotic cells

other than a mitochondrially encoded homologue.

TRANSLATION TERMINATION AND RIBOSOME RECYCLING 243

separate domain of the protein to the NIKS sequence.

The carboxy-terminal part of the eRF1 binds eRF3.

The crystal structure of human eRF1 shows that it is a

Y-shaped molecule that resembles the structure of a

tRNA (Figure 3). Since both bacterial and eukaryotic

class-I RFs carry out essentially the same function, one

would expect them to interact similarly with ribosomal

A site. As with RF2 (see Figure 2), eRF1 must also

undergo a conformational change after binding to the

ribosome in order that it can interact with both the

ribosomal decoding site and the ribosomal PTC.

The eukaryotic class-II RF, eRF3, forms a complex

with eRF1, and via this interaction enhances the

efﬁciency of the translation termination although its

function remains to be fully deﬁned. The GTPase

activity of eRF3, which is triggered by stop codons, is

both eRF1- and ribosome dependent. The carboxy-

terminal half of the eRF3 molecule shows signiﬁcant

amino acid identity to the translation elongation factor

eEF1A that is responsible for bringing the aminoacyl

tRNAs to the ribosome during polypeptide chain

elongation. This suggests that eRF3 may act in a manner

analogous to eEF1A, a protein factor bringing the RF

complex to the ribosome. A number of functions have

been attributed to eRF3, e.g., it may displace eRF1 from

the ribosome (i.e., the function assigned to prokaryotic

RF3), or proofread the eRF1:stop codon interaction, or

stimulate ribosome disassembly (i.e., the function

assigned to prokaryotic RRF; see below), although

direct experimental proof for any one speciﬁc function

is lacking. In mammalian cells, eRF3 binds to the

poly(A)-binding protein (PABP), a protein associated

with the poly(A) tails located at the 3

end of the

majority of eukaryotic mRNAs. This interaction might

be important for the regulation of both mRNA stability

and translation initiation, since disruption of the

eRF3::PABP interaction suppresses the recycling of the

ribosome on the same mRNA.

In Bakers’ yeast (Saccharomyces cerevisiae), the eRF3

protein (also known as Sup35p) has an additional

remarkable property; it is a prion protein that acts as

the protein determinant of a non-Mendelian genetic

element called [PSI

]. Like the mammalian prion

protein PrP, eRF3 can exist in one of two different

conformations with the aggregation-prone prion con-

former catalyzing the conversion of the normal soluble

form, to the prion form. In [PSI

] cells the majority of

eRF3 is found as an inactive, high molecular weight

complex, and thus [PSI

] cells have a defect in

translation termination albeit without detriment to

yeast cell growth. This mechanism may represent a

novel means of regulating the efﬁciency of translation

termination in yeast. There is no evidence that mamma-

lian eRF3 is a prion.

FIGURE 2 The three-dimensional conformation of a release factor

changes once it is bound to the ribosome. (A) The RF2:bacterial

ribosome complex indicating the points of contact between RF2 and the

decoding center and the PTC of the ribosome. (B) The three-

dimensional structures of the unbound and bound forms of bacterial

RF2 indicating the location of the “peptide anticodon” sequence

Ser-Pro-Phe (SPF) and the conserved Gly-Gly-Gln (GGQ) amino acid

motif.

FIGURE 3 Many of the protein factors involved in translation

elongation and termination have three-dimensional shapes that mimic

a tRNA molecule. (A) Three-dimensional structure of a tRNA showing

the location of the anticodon and the site to which the amino acid is

covalently attached. (B) The three-dimensional structures of the

mammalian RF eRF1 (left) and the bacterial translation elongation

factor EF-G (right). The position of the conserved NIKS and GGQ

amino acid motifs are indicated on eRF1, together with the domain to

which the class-II factor eRF3 binds.

244 TRANSLATION TERMINATION AND RIBOSOME RECYCLING

In Archaebacteria, the single class-I RF found (aRF1)

shares signiﬁcant amino acid sequence and structural

homology with eRF1, but not with either RF1 or RF2.

Like eRF1, aRF1 is able to decode all three stop codons.

No eRF3 homologue has been identiﬁed in any

Archaebacterial genome.

Molecular Mimicry

RFs decode stop codons at the ribosomal A site and in

that respect can functionally be compared to aminoacyl

tRNAs which decode sense codons of the mRNA when

positioned at the ribosomal A site during translation

elongation. In so doing, RF1 and RF2 physically interact

with the stop codons in the mRNA. That a common

ribosomal binding site for tRNAs, translation

elongation factors, and RFs exists has led to the

suggestion that these protein and RNA molecules may

also have similar three-dimensional structures, the so-

called “molecular mimicry” model. The structural

similarity—discovered by Clark, Nyborg, and col-

leagues in 1995—between bacterial elongation factor

EF-G and the tRNA-EF-Tu-GTP complex provided

direct evidence for this model. One of the structural

domains of the EF-G protein strongly mimics the

structure of the T-stem and anticodon loop of a tRNA

(Figure 3). The “molecular mimicry” model can also be

extended to include the RFs and conﬁrmation of this

came from the solution of the crystal structure of human

eRF1, whose overall shape and dimensions resemble a

tRNA molecule with the three clearly identiﬁable

structural domains corresponding to the anticodon

loop, the aminoacyl acceptor stem, and T-stem of a

tRNA molecule (Figure 3). There is also signiﬁcant

amino acid sequence similarity between several RF

domains and the carboxy-terminal portion of elongation

factor EF-G, the region of the molecule that interacts

with the ribosome.

Posttermination Events

Once the translation of an mRNA has been terminated

and the completed polypeptide chain released, the

ribosome must dissociate from the mRNA in order to

be able to participate in a new round of protein

synthesis. In bacteria, another protein factor, the RRF,

is involved in this posttermination step together with the

elongation factor that mediates ribosomal translocation,

i.e., EF-G. The three-dimensional structure of RRF, like

eRF1, also mimics a tRNA molecule which, in turn, is

important for its activity in the dissociation of the

posttermination ribosome. After RF3 catalyzes the

release of RF1/RF2 from the ribosome, RRF enters

the A-site of the ribosome, and it is then translocated by

EF-G to the P site, similarly to the translocation of the

peptidyl-tRNA during translation elongation (Figure 1,

step 4). This releases the deacylated tRNA from the P

site which then moves to the ribosomal exit (E) site from

where it rapidly dissociates from the ribosome. The

release of the mRNA from the ribosome is accompanied

by the release of RRF and EF-G. The ﬁnal step before the

ribosome can re-enter the translation cycle is dis-

sociation of the ribosome into its two subunits: an

event that requires an additional protein factor, the

translation initiation factor 3 (IF3).

In eukaryotes, the posttermination step still remains

to be characterized. A gene encoding a eukaryotic RRF-

like protein is found in most eukaryotic genomes but

encodes an RRF that only participates in translation

in the mitochondria. It is conceivable that eRF3 fulﬁlls

this function.

Bypassing Stop Codons: The Causes

and Consequences

The accurate recognition of a stop codon as a

termination signal is important for cell viability,

although cells can tolerate a certain level of stop codon

translation by nonsense suppressor tRNAs. However,

the efﬁciency with which RFs mediate translation

termination at a stop codon can be affected by several

factors. These include the nucleotide context of the stop

codons in the mRNA (i.e., the identity of the bases

directly preceding or following the termination codon),

the cellular concentration of RFs, and the presence of

naturally occurring, nonmutant tRNAs, which are able

to recognize a stop codon as well as their usual normal

sense codon(s) albeit at much lower efﬁciency. For

example, in yeast the tRNA that decodes the CAA codon

as a glutamine residue can also recognize the related stop

codon UAA at low efﬁciency.

A number of nonstandard translational events are

known to compete with the termination machinery in

order to facilitate the translational bypassing of speciﬁc

stop codons. Such a bypass can extend the decoding

properties of a single gene and is an event that can be

regulated. For example, certain animal viruses can cause

ribosomes to shift into a different reading frame just

before a stop codon is reached thereby generating a

second extended form of the translation product – which

in this case is the viral gag-pol protein. In certain plant

viruses, e.g., Tobacco Mosaic Virus (TMV), host-

encoded tRNAs can decode a viral stop codon as a

sense codon if that stop codon is present in the

appropriate nucleotide context.

A further example where stop codons are treated as

sense codons is in the translation of genes encoding

certain selenoproteins, i.e., contains one or more

TRANSLATION TERMINATION AND RIBOSOME RECYCLING 245

selenocysteine residues. In these mRNAs the UGA codon

is used both as a signal for termination and as a signal

for selenocysteine incorporation. How the UGA codon

is read is deﬁned by both a speciﬁc translation factor

(e.g., the SelB protein in bacteria) and a structural

element within the mRNA molecule that is being

translated (e.g., the SECIS element in eukaryotic

selenoprotein-encoding mRNAs).

FURTHER READING

Bertram, G., Innes, S., Minella, O., Richardson, J., and Stansﬁeld, I.

(2001). Endless possibilities: Translation termination and stop

codon recognition. Microbiology UK 147, 255–269.

Ehrenberg, M., and Tenson, T. (2002). A new beginning of the end of

translation. Nat. Struct. Biol. 9, 85–87.

Kisselev, L., Ehrenberg, M., and Frolova, L. (2003). Termination of

translation: Interplay of mRNA, rRNAs, and release factors? Eur.

Molecul. Biol. Orgn. J. 22, 175– 182.

Klaholz, B. P., Pape, T., Zavialov, A. V., Myasnikov, A. G., Orlova,

E. V., Vestergaard, B., Ehrenberg, M., and van Heel, M. (2003).

Structure of the Escherichia coli ribosomal termination complex

with release factor 2. Nature 421, 90–94.

Nakamura, Y., and Ito, K. (2003). Making sense of mimic in

translation termination. Trends Biochem. Sci. 28, 99– 105.

Nissen, P., Kjeldgaard, M., Thirup, S., Polekhina, G., Reshetnikova, L.,

Clark, B. F., and Nyborg, J. (1995). Crystal structure of the ternary

complex of Phe-tRNAPhe, EF-Tu, and a GTP analog. Science 270,

1464–1472.

Rawat, U. B., Zavialov, A. V., Sengupta, J., Valle, M., Grassucci, R. A.,

Linde, J., Vestergaard, B., Ehrenberg, M., and Frank, J. (2003). A

cryo-electron microscopic study of ribosome-bound termination

factor RF2. Nature 421, 87–90.

BIOGRAPHY

Mick Tuite is a Professor of Molecular Biology at the University of

Kent and the former Head of the Department of Biosciences at this

institution. He has had a long-standing interest in the mechanism and

regulation of translation termination using the yeast S. cerevisiae as a

model system. This has recently led to a study of the prion-like

properties of one of the release factors eRF3 (Sup35p). He has a D.Phil.

from the University of Oxford and undertook postdoctoral research

at both Oxford and the University of California, Irvine before joining

the faculty at Kent. He has published over 150 research papers

and reviews.

Nadja Koloteva-Levin is a Senior Postdoctoral Researcher in Mick

Tuite’s laboratory working on the prion behavior of eRF3 (Sup35p).

She received her Ph.D. from University of Manchester Institute of

Science and Technology (UMIST) in the United Kingdom and followed

this with 4 years of postdoctoral research on posttranscriptional gene

regulation at Tel Aviv University in Israel.

246 TRANSLATION TERMINATION AND RIBOSOME RECYCLING

Translesion DNA

Polymerases, Eukaryotic

Alexandra Vaisman and Roger Woodgate

National Institutes of Health, Bethesda, Maryland, USA

Translesion DNA polymerases are enzymes capable of copying

templates containing DNA-distorting lesions. Unlike the highly

processive and accurate replicative polymerases, translesion

polymerases are relatively distributive and lack intrinsic 3

–5

exonuclease activity and are, therefore, unable to proofread

any errors made during replication. As a consequence, they

have low base substitution and frameshift ﬁdelity on unda-

maged DNA templates. Translesion DNA polymerases are also

characterized by an ability to extend mismatched and

misaligned primer termini.

Lesion bypass is believed to occur in two mechanistically

discrete steps, both of which can be either relatively accurate or

highly error-prone. The ﬁrst step is nucleotide incorporation

opposite the damaged base(s) itself, and the second step is

extension of the nascent chain beyond the damage. For some

lesions, translesion replication is facilitated by a single enzyme,

but for certain lesions two polymerases are required; one

enzyme for incorporation opposite the lesion and another for

extension. The polymerase that gains access to the lesion

terminus largely determines the ﬁdelity of translesion synthesis,

i.e., whether it is accurate or mutagenic. After lesion bypass is

completed, the inherently distributive translesion polymerases

dissociate from the DNA template so as to allow replicative

polymerases to resume chromosomal duplication.

Y-Family DNA Polymerases: pol

;

pol

; pol

and Rev1

Most of the enzymes implicated in replication of

imperfect DNA belong to the Y-family of DNA

polymerases. Phylogenetic analysis of this family

suggests that it can be subdivided into several discrete

branches consisting of UmuC-, DinB-, Rev1-, and

Rad30-like proteins. The DinB subfamily is the most

diverse and is found in prokaryotes, archaea, and

eukaryotes. Members of the UmuC subfamily are

found exclusively in bacteria, and proteins from the

Rad30 and Rev1 branches are found only in eukaryotes.

Multiple Y-family orthologues are often present in one

organism. Humans, for example, have four Y-family

DNA polymerases, two RAD30 paralogues, pol

and

pol

, a DinB homologue pol

, and Rev1.

BIOCHEMICAL PROPERTIES

Enzymatic characterization of the Y-family DNA poly-

merases in vitro reveals that despite numerous simi-

larities, each member of the family exhibits distinctive

properties during DNA synthesis. For example, pol

truly unique among eukaryotic DNA polymerases in its

efﬁciency and accuracy of replication past UV-induced

cyclobutane pyrimidine dimers (CPDs). In contrast, Pol

;

a homologue of Rad30, differs from other Y-family DNA

polymerases in that it has a 5

-deoxyribose phosphate

lyase activity, similar to X-family pols

and

: A most

unusual biochemical feature of pol

is, however, its

sequence context ﬁdelity when replicating undamaged

DNA templates. Remarkably, Pol

prefers making

dGMP·T and dGMP·U wobble mispairs – three- to

tenfold greater than the correct dAMP·T and dAMP·U

base pairs. Yet at the same time, pol

discriminates

against certain other mismatches quite well. Thus, errors

made by pol

at template A occur up to 100-fold less

frequently than that made by human pol

: Pol

the other hand, readily catalyzes extension of misaligned

primer termini resulting in -1 frameshift mutations

and is considerably more proﬁcient at extending

terminal mismatches compared to pol

and pol

Finally, unlike other Y-family DNA polymerases that

utilize all four deoxynucleotide triphosphates (dNTPs) in

a template-dependent manner, Saccharomyces cerevisiae

and human Rev1 proteins appear to be specialized

deoxycytidyl transferases, since they utilize the three

other dNTPs only very sparingly.

STRUCTURAL FEATURES

Sequence alignment of the Y-family DNA polymerases

shows a high degree of conservation in the N-terminal

region containing the catalytic domain, while the

C-terminal part is unique for each family member.

Even though the amino acid sequences of these

polymerases show no similarity to classical DNA

polymerases, the general structure of polymerases

from all families appears to be remarkably conserved.

Crystallographic data have shown that like poly-

merases from the A-, B-, and X-families, the Y-family

polymerases adopt a right-hand structure with charac-

teristic ﬁnger, thumb, and palm domains. The Y-family

polymerases differ, however, from classical poly-

merases in having a fourth small domain that has

been termed the “little ﬁnger.” DNA is held in a cleft

between the thumb and the little ﬁnger domains

(Figure 1). The active site of Y-family polymerases is

more solvent accessible and, due to shorter ﬁngers and

thumb domains, forms a more open structure than

that of replicative polymerases. Therefore, if the

geometric constraints on the active site of Y-family

polymerases are less stringent, DNA lesions and

mispairs can be more readily accommodated within

their active site. Such observations thereby help

explain their proﬁciency at lesion bypass and a

concomitant decrease in ﬁdelity on undamaged DNA.

B-Family DNA Polymerase: pol

The ability to accommodate distorted DNA is not

just limited to Y-family polymerases. Pol

; a eukaryotic

B-family polymerase, is another important enzyme

involved in the replication of damaged DNA. Catalyti-

cally active pol

comprises a heterodimer made from the

REV3 and REV7 proteins. The ﬁdelity of pol

undamaged DNA is signiﬁcantly higher than that of

Y-family polymerases (Table I). However, pol

extremely efﬁcient at the extension of aberrant DNA

primer ends, including lesion-containing sites and

misaligned primer-template termini.

X-Family DNA Polymerases: pol

;

pol

; and pol

Although much attention in the last few years has

focused on the Y-family polymerases, earlier studies

raised the possibility that DNA pol

may play a role in

translesion synthesis, even though its major function is

in base excision repair. Pol

is a distributive enzyme in

its mode of action, however, it is more processive than

most Y-family DNA polymerases and the inherent

ability of pol

to select correct rather than incorrect

dNTPs for incorporation is considerably higher than

that of the Y-family polymerases (Table I).

Lesion bypass with varying efﬁciencies can also be

achieved in vitro by other X-family polymerases, such as

the recently discovered pols

and

: Pol

is closely

related to terminal deoxynucleotidyl transferase (TdT),

but in contrast to TdT, it is a partially processive

template-directed DNA polymerase. In vitro, pol

highly error-prone and makes 2 1 frameshift errors in

repetitive sequences. Pol

is a close homologue of pol

and like pol

is more accurate than pol

and possesses

an intrinsic deoxyribophosphate lyase activity (Table I).

Lesion Bypass Speciﬁcity

DNA molecules are intrinsically unstable and are

constantly subjected to the damaging effects of endogen-

ous or exogenous agents. DNA synthesis past a wide

variety of DNA lesions has been studied using both

biological and biochemical approaches. Among the

most common lesions that have been investigated are

UV-induced cis-syn CPDs and the 6-4 pyrimidine-

pyrimidone photoproducts (6-4PPs); abasic sites occur-

ring spontaneously or from the action of DNA

glycosylases; cisplatin adducts (CP) formed at adjacent

guanines by anticancer drugs; and benzo[a]pyrene diol

epoxide (BaP DE) adducts, a naturally occurring meta-

bolite of carcinogens commonly found in cigarette

FIGURE 1 Crystal structure of Dpo4, an archaeal Y-family

polymerase, in a ternary complex with a DNA template and an

incoming nucleoside triphosphate. The surface of the enzyme is color-

coded to indicate the four subdomains found in all Y-family

polymerases. These consist of the palm (red), thumb (green), ﬁnger

(blue), and little ﬁnger (purple) domains. The DNA molecule is drawn

as sticks. Unlike replicative DNA polymerases, the catalytic cleft in

Dpo4 is cavernous and solvent exposed, which explains in part why

Y-family polymerases exhibit low-ﬁdelity synthesis on undamaged

DNA and can accommodate bulky DNA adducts/mismatched bases

in their active sites.

248 TRANSLESION DNA POLYMERASES, EUKARYOTIC

smoke. The accuracy and efﬁciency of translesion

replication varies not only on the chemical nature of

the lesion and speciﬁc distortion that it imposes upon

DNA, but also on the local sequence context within

which the lesion is located and on the speciﬁc

polymerase or combination of polymerases involved in

a given translesion pathway. For example, in both

S. cerevisiae and humans, pol

is the most proﬁcient and

accurate enzyme at replicating past thymine –thymine

CPDs. Pols

; pol

and to a lesser extent pol

and pol

are also able to bypass the lesion but with a much lower

efﬁciency and ﬁdelity. By comparison, both pol

and

pol

are proﬁcient at incorporating a nucleotide

opposite the 3

base of a thymine-thymine 6-4PP, and

pol

can readily extend the resulting primer termini.

Similarly, translesion replication past an abasic site

might involve pols

;

; or Rev1 to incorporate a

base opposite the lesion and pol

for its subsequent

extension. Pol

and pol

are both very efﬁcient and

accurate when replicating past a CP lesion, whereas

bypass by pol

is highly error-prone since it generally

results in 2, 3, and 4 nucleotide deletions, pol

; pol

and pol

are completely blocked by CP adducts in

template DNA.

The ability of the polymerase to facilitate lesion

bypass also depends on the nucleotide at which a

speciﬁc adduct is formed. For example, in human cells,

efﬁcient and accurate bypass of BaP DE adducts

formed at deoxyadenosine (dA) can be achieved

through the sequential action of pol

and pol

; but

not by pol

or pol

: In contrast, the ability of

translesion polymerases to bypass the same BaP DE

adduct formed at deoxyguanosine (dG) base is quite

different. Whereas human pol

and pol

bypass BaP

DE dG adducts inefﬁciently and inaccurately, pol

bypasses the lesion with high efﬁciency and predomi-

nantly incorporates the correct base, dCMP, opposite

the adducted dG.

Protein Partners

Despite extensive biochemical characterization of the

polymerases involved in translesion synthesis, it is still

unclear how the process physically occurs in living cells.

It has been shown that at least three of the translesion

polymerases, pol

; pol

; and Rev1 are associated with

the replication machinery in an undamaged cell and that

upon DNA damage the polymerase accumulates at sites

of stalled replication forks.

The catalytic activity of pols

;

; and

undamaged and lesion-containing DNA is substantially

increased through protein– protein interactions with

accessory proteins normally utilized by replicative

polymerases. These include proliferating cell nuclear

antigen (PCNA), which serves as a sliding clamp that

helps recruit polymerases to their site of action, and a

clamp loader, replication factor C (RFC). Given the

sixfold symmetry of the PCNA complex, it is possible

that multiple DNA polymerases connect simultaneously

to a replicative single clamp. However, how a cell

chooses between the various translesion polymerases it

has at its disposal and how the switch from the

replicative polymerase to a translesion polymerase and

then back to the replicase occurs in vivo still remains to

be determined.

Cellular Role(s)

Pol

is clearly responsible for the accurate and robust

replication of CPD-containing DNA in vivo. Although

the lack of pol

in human cells does not confer a

substantial hypersensitivity to killing by UV light,

humans with a defective pol

do exhibit the Xeroderma

Pigmentosum Variant (XP-V) syndrome and are

predisposed to the manifestation of skin cancers.

TABLE I

Eukaryotic DNA Polymerases Known to Participate in Translesion Replication

DNA

polymerase Gene or synonym Family

Error rate on

undamaged DNA Additional proposed functions

; beta POLB X10

–10

Base excision repair, double-strand break repair, synapsis and

recombination, meiosis, neurogenesis

; zeta POLZ (REV3L/REV7L) B10

–10

Somatic hypermutation, mismatch extension, developmental

cell proliferation

; eta POLH (RAD30A, XPV) Y10

–10

Somatic hypermutation

; iota POLI (RAD30B) Y10

–10

Base excision repair, somatic hypermutation

; kappa POLK (DINB1) Y10

–10

Spontaneous mutagenesis

; lamda POLL (POL4) X10

Base excision repair, repair during meiosis

; mu POLM X ? Nonhomologous end-joining pathway, lymphoid formation

Rev1 REV1L Y

TRANSLESION DNA POLYMERASES, EUKARYOTIC 249

To date, XP-V is the only known example of a human

disorder caused by a defect in translesion replication.

Apart from a clear role for pol

in the bypass of CPDs,

there is indirect evidence suggesting the participation of

other enzymes in translesion replication in vivo. For

example, cells with a reduced level of the deoxycytidyl

transferase Rev1 show lower levels of UV-induced

mutagenesis. In vitro assays suggest that pol

may

play an important role in accurate bypass of certain

lesions generated by polycyclic aromatic hydrocarbons

and recent studies with transgenic mice devoid of pol

support such hypotheses.

Just as translesion replication often requires multiple

DNA polymerases, so a particular translesion poly-

merase might have evolved to participate in one or

more functional tasks in a cell. Indeed, various lines of

evidence support a role for translesion polymerases

and

in the somatic hypermutation of rearranged

immunoglobulin genes. While S. cerevisiae strains with

deletions in their REV3 or REV7 genes (encoding pol

)

are viable, disruption of murine REV3 unexpectedly

leads to embryonic lethality in mice, suggesting that, in

addition to lesion bypass, mammalian pol

may play

additional role(s) in cellular development. Genetic and

biochemical studies have shown that S. cerevisiae

Rev1, in addition to participating in the bypass of

UV-induced lesions and abasic sites, also has a second

function that is independent of its dCMP transferase

activity. Pol

whose primary function is in base

excision repair, has also been implicated in double-

strand-break repair, meiotic events associated with

synapsis and recombination and in neurogenesis. On

the other hand, owing to an associated 5

-deoxyribose

phosphate lyase activity, it has been hypothesized that

pol

and pol

might possibly substitute for pol

certain forms of specialized base excision repair.

Finally, pol

’s main cellular role may actually be to

participate in a non-homologous end-joining pathway

that repairs DNA double-strand breaks and its ability

to traverse certain DNA lesions may, in fact, be rarely

utilized in vivo.