The biochemical requirements of DNA polymerase V-mediated translesion synthesis revisited

In addition to replicative DNA polymerases, cells contain specialized DNA polymerases involved in processes such as lesion tolerance, mutagenesis and immunoglobulin diversity. In Escherichia coli, DNA polymerase V (Pol V), encoded by the umuDC locus, is involved in translesion synthesis (TLS) and mutagenesis. Genetic studies have established that mutagenesis requires both UmuC and a proteolytic product of UmuD (UmuD'). In addition, RecA protein and the replication processivity factor, the beta-clamp, were genetically found to be essential co-factors for mutagenesis. Here, we have reconstituted Pol V-mediated bypass of three common replication-blocking lesions, namely the two major UV-induced lesions and a guanine adduct formed by a chemical carcinogen (G-AAF) under conditions that fulfil these in vivo requirements. Two co-factors are essential for efficient Pol V-mediated lesion bypass: (i) a DNA substrate onto which the beta-clamp is stably loaded; and (ii) an extended single-stranded RecA/ATP filament assembled downstream from the lesion site. For efficient bypass, Pol V needs to interact simultaneously with the beta-clamp and the 3' tip of the RecA filament. Formation of an extended RecA/ATP filament and stable loading of the beta-clamp are best achieved on long single-stranded circular DNA templates. In contrast to previously published data, the single-stranded DNA-binding protein (SSB) is not absolutely required for Pol V-mediated lesion bypass provided ATP, instead of ATPgammaS, activates the RecA filament. Further discrepancies with the existing literature are explainable by the use of either inadequate DNA substrates or a UmuC fusion protein instead of native Pol V.     

Characterization of Escherichia coli UmuC active-site loops identifies variants that confer UV hypersensitivity

DNA is constantly exposed to chemical and environmental mutagens, causing lesions that can stall replication. In order to deal with DNA damage and other stresses, Escherichia coli utilizes the SOS response, which regulates the expression of at least 57 genes, including umuDC. The gene products of umuDC, UmuC and the cleaved form of UmuD, UmuD', form the specialized E. coli Y-family DNA polymerase UmuD'2C, or polymerase V (Pol V). Y-family DNA polymerases are characterized by their specialized ability to copy damaged DNA in a process known as translesion synthesis (TLS) and by their low fidelity on undamaged DNA templates. Y-family polymerases exhibit various specificities for different types of DNA damage. Pol V carries out TLS to bypass abasic sites and thymine-thymine dimers resulting from UV radiation. Using alanine-scanning mutagenesis, we probed the roles of two active-site loops composed of residues 31 to 38 and 50 to 54 in Pol V activity by assaying the function of single-alanine variants in UV-induced mutagenesis and for their ability to confer resistance to UV radiation. We find that mutations of the N-terminal residues of loop 1, N32, N33, and D34, confer hypersensitivity to UV radiation and to 4-nitroquinoline-N-oxide and significantly reduce Pol V-dependent UV-induced mutagenesis. Furthermore, mutating residues 32, 33, or 34 diminishes Pol V-dependent inhibition of recombination, suggesting that these mutations may disrupt an interaction of UmuC with RecA, which could also contribute to the UV hypersensitivity of cells expressing these variants.     

The mutagenesis protein UmuC is a DNA polymerase activated by UmuD', RecA, and SSB and is specialized for translesion replication.

Replication of DNA lesions leads to the formation of mutations. In Escherichia coli this process is regulated by the SOS stress response, and requires the mutagenesis proteins UmuC and UmuD'. Analysis of translesion replication using a recently reconstituted in vitro system (Reuven, N. B., Tomer, G., and Livneh, Z. (1998) Mol. Cell 2, 191-199) revealed that lesion bypass occurred with a UmuC fusion protein, UmuD', RecA, and SSB in the absence of added DNA polymerase. Further analysis revealed that UmuC was a DNA polymerase (E. coli DNA polymerase V), with a weak polymerizing activity. Upon addition of UmuD', RecA, and SSB, the UmuC DNA polymerase was greatly activated, and replicated a synthetic abasic site with great efficiency (45% bypass in 6 min), 10-100-fold higher than E. coli DNA polymerases I, II, or III holoenzyme. Analysis of bypass products revealed insertion of primarily dAMP (69%), and to a lesser degree dGMP (31%) opposite the abasic site. The UmuC104 mutant protein was defective both in lesion bypass and in DNA synthesis. These results indicate that UmuC is a UmuD'-, RecA-, and SSB-activated DNA polymerase, which is specialized for lesion bypass. UmuC is a member of a new family of DNA polymerases which are specialized for lesion bypass, and include the yeast RAD30 and the human XP-V genes, encoding DNA polymerase eta.     

DNA polymerase V and RecA protein, a minimal mutasome

A hallmark of the Escherichia coli SOS response is the large increase in mutations caused by translesion synthesis (TLS). TLS requires DNA polymerase V (UmuD'2C) and RecA. Here, we show that pol V and RecA interact by two distinct mechanisms. First, pol V binds to RecA in the absence of DNA and ATP and second, through its UmuD' subunit, requiring DNA and ATP without ATP hydrolysis. TLS occurs in the absence of a RecA nucleoprotein filament but is inhibited in its presence. Therefore, a RecA nucleoprotein filament is unlikely to be required for SOS mutagenesis. Pol V activity is severely diminished in the absence of RecA or in the presence of RecA1730, a mutant defective for pol V mutagenesis in vivo. Pol V activity is strongly enhanced with RecA mutants constitutive for mutagenesis in vivo, suggesting that RecA is an obligate accessory factor that activates pol V for SOS mutagenesis.     

RecFOR proteins are essential for Pol V-mediated translesion synthesis and mutagenesis

When the replication fork moves through the template DNA containing lesions, daughter-strand gaps are formed opposite lesion sites. These gaps are subsequently filled-in either by translesion synthesis (TLS) or by homologous recombination. RecA filaments formed within these gaps are key intermediates for both of the gap-filling pathways. For instance, Pol V, the major lesion bypass polymerase in Escherichia coli, requires a functional interaction with the tip of the RecA filament. Here, we show that all three recombination mediator proteins RecFOR are needed to build a functionally competent RecA filament that supports efficient Pol V-mediated TLS in the presence of ssDNA-binding protein (SSB). A positive contribution of RecF protein to Pol V lesion bypass is demonstrated. When Pol III and Pol V are both present, Pol III imparts a negative effect on Pol V-mediated lesion bypass that is counteracted by the combined action of RecFOR and SSB. Mutations in recF, recO or recR gene abolish induced mutagenesis in E. coli.     

Lesion bypass by the Escherichia coli DNA polymerase V requires assembly of a RecA nucleoprotein filament

Translesion replication is carried out in Escherichia coli by the SOS-inducible DNA polymerase V (UmuC), an error-prone polymerase, which is specialized for replicating through lesions in DNA, leading to the formation of mutations. Lesion bypass by pol V requires the SOS-regulated proteins UmuD' and RecA and the single-strand DNA-binding protein (SSB). Using an in vitro assay system for translesion replication based on a gapped plasmid carrying a site-specific synthetic abasic site, we show that the assembly of a RecA nucleoprotein filament is required for lesion bypass by pol V. This is based on the reaction requirements for stoichiometric amounts of RecA and for single-stranded gaps longer than 100 nucleotides and on direct visualization of RecA-DNA filaments by electron microscopy. SSB is likely to facilitate the assembly of the RecA nucleoprotein filament; however, it has at least one additional role in lesion bypass. ATPgammaS, which is known to strongly increase binding of RecA to DNA, caused a drastic inhibition of pol V activity. Lesion bypass does not require stoichiometric binding of UmuD' along RecA filaments. In summary, the RecA nucleoprotein filament, previously known to be required for SOS induction and homologous recombination, is also a critical intermediate in translesion replication.     

Defining the position of the switches between replicative and bypass DNA polymerases

Cells contain specialized DNA polymerases that are able to copy past lesions with an associated risk of generating mutations, the major cause of cancer. Here, we reconstitute translesion synthesis (TLS) using the replicative (Pol III) and major bypass (Pol V) DNA polymerases from Escherichia coli in the presence of accessory factors. When the replicative polymerase disconnects from the template in the vicinity of a lesion, Pol V binds the blocked replication intermediate and forms a stable complex by means of a dual interaction with the tip of the RecA filament and the beta-clamp, the processivity factor donated by the blocked Pol III holoenzyme. Both interactions are required to confer to Pol V the processivity that will allow it synthesize, in a single binding event, a TLS patch long enough to support further extension by Pol III. In the absence of these accessory factors, the patch synthesized by Pol V is too short, being degraded by the Pol III-associated exonuclease activity that senses the distortion induced by the lesion, thus leading to an aborted bypass process.     

UV- and MMS-induced mutagenesis of lambdaO(am)8 phage under nonpermissive conditions for phage DNA replication

Mutagenesis in Escherichia coli, a subject of many years of study is considered to be related to DNA replication. DNA lesions nonrepaired by the error-free nucleotide excision repair (NER), base excision repair (BER) and recombination repair (RR), stop replication at the fork. Reinitiation needs translesion synthesis (TLS) by DNA polymerase V (UmuC), which in the presence of accessory proteins, UmuD', RecA and ssDNA-binding protein (SSB), has an ability to bypass the lesion with high mutagenicity. This enables reinitiation and extension of DNA replication by DNA polymerase III (Pol III). We studied UV- and MMS-induced mutagenesis of lambdaO(am)8 phage in E. coli 594 sup+ host, unable to replicate the phage DNA, as a possible model for mutagenesis induced in nondividing cells (e.g. somatic cells). We show that in E. coli 594 sup+ cells UV- and MMS-induced mutagenesis of lambdaO(am)8 phage may occur. This mutagenic process requires both the UmuD' and C proteins, albeit a high level of UmuD' and low level of UmuC seem to be necessary and sufficient. We compared UV-induced mutagenesis of lambdaO(am)8 in nonpermissive (594 sup+) and permissive (C600 supE) conditions for phage DNA replication. It appeared that while the mutagenesis of lambdaO(am)8 in 594 sup+ requires the UmuD' and C proteins, which can not be replaced by other SOS-inducible protein(s), in C600 supE their functions may be replaced by other inducible protein(s), possibly DNA polymerase IV (DinB). Mutations induced under nonpermissive conditions for phage DNA replication are resistant to mismatch repair (MMR), while among those induced under permissive conditions, only about 40% are resistant.     

All three SOS-inducible DNA polymerases (Pol II, Pol IV and Pol V) are involved in induced mutagenesis

Most organisms contain several members of a recently discovered class of DNA polymerases (umuC/dinB superfamily) potentially involved in replication of damaged DNA. In Escherichia coli, only Pol V (umuDC) was known to be essential for base substitution mutagenesis induced by UV light or abasic sites. Here we show that, depending upon the nature of the DNA damage and its sequence context, the two additional SOS-inducible DNA polymerases, Pol II (polB) and Pol IV (dinB), are also involved in error-free and mutagenic translesion synthesis (TLS). For example, bypass of N:-2-acetylaminofluorene (AAF) guanine adducts located within the NAR:I mutation hot spot requires Pol II for -2 frameshifts but Pol V for error-free TLS. On the other hand, error-free and -1 frameshift TLS at a benzo(a)pyrene adduct requires both Pol IV and Pol V. Therefore, in response to the vast diversity of existing DNA damage, the cell uses a pool of 'translesional' DNA polymerases in order to bypass the various DNA lesions.     

The effect of oxidative metabolism on spontaneous Pol zeta-dependent translesion synthesis in Saccharomyces cerevisiae

DNA lesions can stall or block high-fidelity polymerases, thus inhibiting replication. To bypass such lesions, low-fidelity translesion synthesis (TLS) polymerases can be used to insert a nucleotide across from the lesion or extend from a lesion:base mispair. When DNA repair is compromised in Saccharomyces cerevisiae, spontaneous DNA lesions can lead to a novel mutational event in which a frameshift is accompanied by one or more base pair substitutions. These "complex frameshifts" are dependent upon the TLS polymerase Pol zeta, and provide a mutational signature for mutagenic Pol zeta-dependent activity. In the current study, we have found that a specific subset of the Pol zeta-dependent mutational events requires oxidative metabolism. These results suggest that translesion bypass of spontaneously oxidized DNA bases can be a significant source of mutagenesis in repair compromised cells.     

DNA polymerases and SOS mutagenesis: can one reconcile the biochemical and genetic data?

Until recently, it had been concluded from genetic evidence that DNA polymerase III (Pol III, the main replicative polymerase in E. coli) was also responsible for mutagenic translesion synthesis on damaged templates, albeit under the influence of inducible proteins UmuD' and UmuC. Now it appears that these proteins themselves have polymerase activity (and are now known as Pol V) and can carry out translesion synthesis in vitro in the absence of Pol III. Here I discuss the apparent contradictions between genetics and biochemistry with regard to the role of Pol III in translesion synthesis. Does Pol V interact with Pol III and constitute an alternative component of the replication factory (replisome)? Where do the other three known polymerases fit in? What devices does the cell have to ensure that the "right" polymerase is used in a given situation? The debate about the role of Pol III in translesion synthesis reveals a deeper divide between models that interpret everything in terms of mass action effects and those that embrace a replisome held together by protein-protein interactions and located as a structural entity within the cell.     

PCNA trimer instability inhibits translesion synthesis by DNA polymerase η and by DNA polymerase δ

Translesion synthesis (TLS), the process by which DNA polymerases replicate through DNA lesions, is the source of most DNA damage-induced mutations. Sometimes TLS is carried out by replicative polymerases that have evolved to synthesize DNA on non-damaged templates. Most of the time, however, TLS is carried out by specialized translesion polymerases that have evolved to synthesize DNA on damaged templates. TLS requires the mono-ubiquitylation of the replication accessory factor proliferating cell nuclear antigen (PCNA). PCNA and ubiquitin-modified PCNA (UbPCNA) stimulate TLS by replicative and translesion polymerases. Two mutant forms of PCNA, one with an E113G substitution and one with a G178S substitution, support normal cell growth but inhibit TLS thereby reducing mutagenesis in yeast. A re-examination of the structures of both mutant PCNA proteins revealed substantial disruptions of the subunit interface that forms the PCNA trimer. Both mutant proteins have reduced trimer stability with the G178S substitution causing a more severe defect. The mutant forms of PCNA and UbPCNA do not stimulate TLS of an abasic site by either replicative Pol δ or translesion Pol η. Normal replication by Pol η was also impacted, but normal replication by Pol δ was much less affected. These findings support a model in which reduced trimer stability causes these mutant PCNA proteins to occasionally undergo conformational changes that compromise their ability to stimulate TLS by both replicative and translesion polymerases.     

The processing of a Benzo(a)pyrene adduct into a frameshift or a base substitution mutation requires a different set of genes in Escherichia coli

Replication through a single DNA lesion may give rise to a panel of translesion synthesis (TLS) events, which comprise error-free TLS, base substitutions and frameshift mutations. In order to determine the genetic control of the various TLS events induced by a single lesion, we have chosen the major N2-dG adduct of (+)-anti-Benzo(a)pyrene diol epoxide [(+)-anti-BPDE] adduct located within a short run of guanines as a model lesion. Within this sequence context, in addition to the major event, i.e. error-free TLS, the adduct also induces base substitutions (mostly G --> T transversions) and -1 frameshift mutations. The pathway leading to G --> T base substitution mutagenesis appears to be SOS independent, suggesting that TLS is most probably performed by the replicative Pol III holoenzyme itself. In contrast, both error-free and frameshift TLS pathways are dependent upon SOS-encoded functions that belong to the pool of inducible DNA polymerases specialized in TLS (translesional DNA polymerases), namely umuDC (Pol V) and dinB (Pol IV). It is likely that, given the diversity of conformations that can be adopted by lesion-containing replication intermediates, cells use one or several translesional DNA polymerases to achieve TLS.     

Translesion synthesis in Escherichia coli: lessons from the NarI mutation hot spot

Duplication of DNA containing damaged bases is a challenge to DNA polymerases that normally replicate with high speed, high accuracy and high processivity undamaged templates only. When a replicative DNA polymerase encounters a chemically altered base that it is unable to copy, a process called translesion synthesis (TLS) takes place during which the replicative polymerase is transiently replaced by a so-called specialized or lesion bypass polymerase. In addition to the central players that are the replicative and translesion DNA polymerases, TLS pathways involve accessory factors such as the general replication processivity factor (i.e. the beta-clamp in prokaryotes and PCNA in eukaryotes). In Escherichia coli, besides the beta-clamp, RecA plays a fundamental role as a co-factor of Pol V the major bypass polymerase in this organism. An integrated view of TLS pathways necessarily requires both genetic and biochemical studies. In this review we will attempt to summarize the insights into TLS gained over the last 25 years by studying a frameshift mutation hot spot, the NarI site. This site was initially discovered by serendipity when establishing a forward mutation spectrum induced by a chemical hepatocarcinogen, N-2-acetylaminofluorene (AAF). Indeed, this chemical carcinogen covalently binds to DNA forming adducts with guanine residues. When bound to G* in the NarI site, 5'-GGCG*CC-, AAF induces the loss of the G*pC dinucleotide at a frequency that is approximately 10(7)-fold higher than the spontaneous frequency. In vivo studies showed that the NarI mutation hot spot is neither restricted to the NarI sequence itself, nor to the carcinogen AAF. Instead, the hot spot requires a sequence containing at least two GpC repeats and any of a family of aromatic amides and nitro aromatic compounds that form a large class of human carcinogens. Genetic analysis initially revealed that the NarI frameshift pathway is SOS dependent but umuDC (i.e. Pol V) independent. More recently, DNA Pol II was identified as the enzyme responsible of this frameshift pathway. Concurrently the AAF adduct in the NarI site can be bypassed in an error-free way by Pol V. The NarI site thus offers a unique possibility to study the interplay between two specialized DNA polymerases, Pol II and Pol V, that can both extend replication intermediates formed when the replicative Pol III dissociates in the vicinity of the damage. Full reconstitution of the two pathways led us to highlight a key feature for TLS pathways, namely that it is critical the specialized DNA polymerase synthesizes, during the course of a single binding event, a patch of DNA synthesis (TLS patch) that is long enough as to "hide the lesion induced distortion" from the proofreading activity upon reloading of the replicative DNA polymerase (or any exonuclease that may get access to the primer when the specialized DNA polymerase detaches). The beta-clamp, to which all DNA polymerases bind, plays a critical role in allowing the specialized DNA polymerases to synthesize TLS patches that are long enough to resist such "external proofreading" activities.     

Analysis of the stimulation of DNA polymerase V of Escherichia coli by processivity proteins

Bypass of replication-blocking lesions in Escherichia coli is carried out by DNA polymerase V (UmuC) in a reaction that requires UmuD', RecA, and single-strand DNA-binding protein (SSB). The activity of this four-component basic bypass system is a low-fidelity and low-processivity activity. Addition of the processivity subunits of pol III, the beta subunit sliding DNA clamp, and the five-subunit gamma complex clamp loader increased the rate of translesion replication approximately 3-fold. This stimulation was specific to the lesion bypass step, with no effect on the initiation of synthesis by pol V. The beta subunit and gamma complex increased the processivity of pol V from 3 to approximately 14-18 nucleotides, providing a mechanistic basis for their stimulatory effect. Stimulation of bypass was observed over a range of RecA and SSB concentrations. ATPgammaS, which strongly inhibits translesion replication by pol V, primarily via inhibition of the initiation stage, caused the same inhibition also in the presence of the processivity proteins. The in vivo role of the processivity proteins in translesion replication was examined by assaying UV mutagenesis. This was done in a strain carrying the dnaN59 allele, encoding a temperature-sensitive beta subunit. When assayed in an excision repair-defective background, the dnaN59 mutant exhibited a level of UV mutagenesis reduced up to 3-fold compared to that of the isogenic dnaN(+) strain. This suggests that like in the in vitro system, the beta subunit stimulates lesion bypass in vivo.     

Selective disruption of the DNA polymerase III α-β complex by the umuD gene products

DNA polymerase III (DNA pol III) efficiently replicates the Escherichia coli genome, but it cannot bypass DNA damage. Instead, translesion synthesis (TLS) DNA polymerases are employed to replicate past damaged DNA; however, the exchange of replicative for TLS polymerases is not understood. The umuD gene products, which are up-regulated during the SOS response, were previously shown to bind to the α, β and ε subunits of DNA pol III. Full-length UmuD inhibits DNA replication and prevents mutagenic TLS, while the cleaved form UmuD' facilitates mutagenesis. We show that α possesses two UmuD binding sites: at the N-terminus (residues 1-280) and the C-terminus (residues 956-975). The C-terminal site favors UmuD over UmuD'. We also find that UmuD, but not UmuD', disrupts the α-β complex. We propose that the interaction between α and UmuD contributes to the transition between replicative and TLS polymerases by removing α from the β clamp.     

Lyase activities intrinsic to Escherichia coli polymerases IV and V

Escherichia coli DNA polymerase IV and V (pol IV and pol V) are error-prone DNA polymerases that are induced as part of the SOS regulon in response to DNA damage. Both are members of the Y-family of DNA polymerases. Their principal biological roles appear to involve translesion synthesis (TLS) and the generation of mutational diversity to cope with stress. Although neither enzyme is known to be involved in base excision repair (BER), we have nevertheless observed apurinic/apyrimidinic 5'-deoxyribose phosphate (AP/5'-dRP) lyase activities intrinsic to each polymerase. Pols IV and V catalyze cleavage of the phosphodiester backbone at the 3'-side of an apurinic/apyrimidinic (AP) site as well as the removal of a 5'-deoxyribose phosphate (dRP) at a preincised AP site. The specific activities of the two error-prone polymerase-associated lyases are approximately 80-fold less than the associated lyase activity of human DNA polymerase beta, which is a key enzyme used in short patch BER. Pol IV forms a covalent Schiff's base intermediate with substrate DNA that is trapped by sodium borohydride, as proscribed by a beta-elimination mechanism. In contrast, a NaBH(4) trapped intermediate is not observed for pol V, even though the lyase specific activity of pol V is slightly higher than that of pol IV. Incubation of pol V (UmuD'(2)C) with a molar excess of UmuD drives an exchange of subunits to form UmuD'D+insoluble UmuC causing inactivation of polymerase and lyase activities. The concomitant loss of both activities is strong evidence that pol V contains a bona fide lyase activity.     

跨损伤合成的DNA聚合酶——一类新的DNA聚合酶

细胞虽然拥有多种修复途径,但有些DNA损伤仍不可避免地会逃避修复而在基因组上保留下来,细胞跨损伤DNA合成的分子机制一直是DNA修复中主要的未解决问题之一.最近通过对一类结构相关性UmuC/DinB蛋白质超家族成员的研究发现它们具有DNA聚合酶功能.这类新发现的DNA聚合酶不同于经典的复制性DNA聚合酶,它们能以易误/突变(error-prone/mutagenic)或无误(error-free)方式进行跨损伤(translesion)DNA合成,并且从细菌到人在进化上功能保守.     

Nucleotide excision repair or polymerase V-mediated lesion bypass can act to restore UV-arrested replication forks in Escherichia coli

Nucleotide excision repair and translesion DNA synthesis are two processes that operate at arrested replication forks to reduce the frequency of recombination and promote cell survival following UV-induced DNA damage. While nucleotide excision repair is generally considered to be error free, translesion synthesis can result in mutations, making it important to identify the order and conditions that determine when each process is recruited to the arrested fork. We show here that at early times following UV irradiation, the recovery of DNA synthesis occurs through nucleotide excision repair of the lesion. In the absence of repair or when the repair capacity of the cell has been exceeded, translesion synthesis by polymerase V (Pol V) allows DNA synthesis to resume and is required to protect the arrested replication fork from degradation. Pol II and Pol IV do not contribute detectably to survival, mutagenesis, or restoration of DNA synthesis, suggesting that, in vivo, these polymerases are not functionally redundant with Pol V at UV-induced lesions. We discuss a model in which cells first use DNA repair to process replication-arresting UV lesions before resorting to mutagenic pathways such as translesion DNA synthesis to bypass these impediments to replication progression.     

Def1 Promotes the Degradation of Pol3 for Polymerase Exchange to Occur During DNA-Damage–Induced Mutagenesis in Saccharomyces cerevisiae

DNA damages hinder the advance of replication forks because of the inability of the replicative polymerases to synthesize across most DNA lesions. Because stalled replication forks are prone to undergo DNA breakage and recombination that can lead to chromosomal rearrangements and cell death, cells possess different mechanisms to ensure the continuity of replication on damaged templates. Specialized, translesion synthesis (TLS) polymerases can take over synthesis at DNA damage sites. TLS polymerases synthesize DNA with a high error rate and are responsible for damage-induced mutagenesis, so their activity must be strictly regulated. However, the mechanism that allows their replacement of the replicative polymerase is unknown. Here, using protein complex purification and yeast genetic tools, we identify Def1 as a key factor for damage-induced mutagenesis in yeast. In in vivo experiments we demonstrate that upon DNA damage, Def1 promotes the ubiquitylation and subsequent proteasomal degradation of Pol3, the catalytic subunit of the replicative polymerase δ, whereas Pol31 and Pol32, the other two subunits of polymerase δ, are not affected. We also show that purified Pol31 and Pol32 can form a complex with the TLS polymerase Rev1. Our results imply that TLS polymerases carry out DNA lesion bypass only after the Def1-assisted removal of Pol3 from the stalled replication fork.