A physical and functional interaction between yeast Pol4 and Dnl4-Lif1 links DNA synthesis and ligation in nonhomologous end joining.

Genetic studies have implicated the Saccharomyces cerevisiae POL4 gene product in the repair of DNA double-strand breaks by nonhomologous end joining. Here we show that Pol4 preferentially catalyzes DNA synthesis on small gaps formed by the alignment of linear duplex DNA molecules with complementary ends, a DNA substrate specificity that is compatible with its predicted role in the repair of DNA double-strand breaks. Pol4 also interacts directly with the Dnl4 subunit of the Dnl4-Lif1 complex via its N-terminal BRCT domain. This interaction stimulates the DNA synthesis activity of Pol4 and, to a lesser extent, the DNA joining activity of Dnl4-Lif1. Notably, the joining of DNA substrates that require the combined action of Pol4 and Dnl4-Lif1 is much more efficient than the joining of similar DNA substrates that require only ligation. Thus, the physical and functional interactions between Pol4 and Dnl4-Lif1 provide a molecular mechanism for both the recruitment of Pol4 to in vivo DNA double-strand breaks and the coupling of the gap filling DNA synthesis and DNA joining reactions that complete the microhomology-mediated pathway of nonhomologous end joining.

Genetic studies have implicated the Saccharomyces cerevisiae POL4 gene product in the repair of DNA double-strand breaks by nonhomologous end joining. Here we show that Pol4 preferentially catalyzes DNA synthesis on small gaps formed by the alignment of linear duplex DNA molecules with complementary ends, a DNA substrate specificity that is compatible with its predicted role in the repair of DNA double-strand breaks. Pol4 also interacts directly with the Dnl4 subunit of the Dnl4-Lif1 complex via its N-terminal BRCT domain. This interaction stimulates the DNA synthesis activity of Pol4 and, to a lesser extent, the DNA joining activity of Dnl4-Lif1. Notably, the joining of DNA substrates that require the combined action of Pol4 and Dnl4-Lif1 is much more efficient than the joining of similar DNA substrates that require only ligation. Thus, the physical and functional interactions between Pol4 and Dnl4-Lif1 provide a molecular mechanism for both the recruitment of Pol4 to in vivo DNA double-strand breaks and the coupling of the gap filling DNA synthesis and DNA joining reactions that complete the microhomologymediated pathway of nonhomologous end joining.
Mechanisms for the repair of DNA double-strand breaks (DSBs) 1 can be divided into two classes based on the requirement for DNA sequence homology. In the major homologydependent pathway, repair involves an intact duplex that is homologous to the broken molecule. This is the major DSB repair pathway in the yeast Saccharomyces cerevisiae and is mediated by members of the RAD52 epistasis group that includes RAD50, RAD51, RAD52, RAD54, RAD55, RAD57, RAD59, MRE11, XRS2, and RDH54/TID1 (1). Alternatively, broken DNA ends are simply brought together, processed, and then ligated by repair mechanisms, known collectively as nonhomologous end joining (NHEJ) (2). Unlike the major recombinational repair pathway that faithfully restores the genetic information, nonhomologous end joining frequently causes genetic alterations that range from the loss or addition of a few nucleotides at the break site to gross rearrangements such as chromosomal translocations (2).
Many of the genetic studies and the biochemical study with purified NHEJ factors have focused on the joining of linear DNA molecules with short complementary single strands at their termini (3, 4, 7-10, 12, 13, 24). However, the majority of DSBs generated by agents such as ionizing radiation will have ends that are neither complementary nor ligatable, indicating that end processing reactions will be critical for the repair of in vivo DSBs by NHEJ. In fact analysis of DNA molecules repaired by NHEJ has revealed that a favored mode of end processing involves short tracts of DNA sequence homology, so-called microhomologies, close to the break site that presumably facilitate alignment of the DNA ends (14,29,30). Following alignment, the processing of DNA ends by nucleases and DNA polymerases to remove noncomplementary nucleotides and fill-in gaps is likely to be required to generate ligatable termini (31,32).
Interestingly, biochemical studies with purified human Mre11 have shown that this nuclease can act on DNA ends to expose and align microhomologies that can then be ligated (33,34). However, the efficiency of recircularization of linear plasmid DNA molecules with complementary single-strand ends in vivo is not affected by inactivation of yeast Mre11 nuclease activity (8,35). This observation suggests that, although the nuclease activity of Mre11 may not be required for nonhomologous end joining, the Rad50-Mre11-Xrs2 complex has another critical role in this repair pathway. Indeed, recent biochemical studies have shown that the Rad50-Mre11-Xrs2 complex has end bridging activity and functionally interacts with the Dnl4-Lif1 complex (24).
Recent genetic studies have shown that pol4 and rad27 strains have no defect in the recircularization of linear plasmid DNA molecules with complementary single-strand ends but exhibit reduced joining of linearized plasmid DNA molecules with noncomplementary termini (31,32). These observations, together with enzymatic properties of the DNA polymerase, Pol4 (32,36,37), and the flap endonuclease, Fen-1 (Rad27) (38,39), suggest that Pol4 and Fen-1 participate in microhomologymediated NHEJ events requiring gap-filling and nucleolytic processing steps prior to DNA joining. Pol4 is a member of the functionally diverse Pol X family of nucleotidyl transferases (see Fig. 1) (36, 37, 40 -43). Within this family, mammalian Pol ␤ catalyzes gap filling DNA synthesis in base excision repair (43,44), whereas terminal transferase adds nucleotides in a template-independent manner during V(D)J recombination (45). The cellular functions of Pol and Pol are less well understood (40 -42). In this study we describe a functional interaction between Pol4 and Dnl4-Lif1 that links the gapfilling and ligation steps of NHEJ.

MATERIALS AND METHODS
Plasmid Construction-The S. cerevisiae POL4 open reading frame was amplified from BJ5464 genomic DNA by the polymerase chain reaction. After the DNA sequence of the amplified product was verified, it was subcloned into the Escherichia coli expression plasmids pGSTag (46) and pET28b (Novagen) to generate the plasmids pGST-Pol4 and pET28b-Pol4, which express Pol4 as a glutathione S-transferase (GST) fusion and His-tagged polypeptide, respectively. Using a similar strategy we constructed the plasmids pGST-Pol4⌬BRCT and pET28b-Pol4⌬BRCT, which encode tagged versions of Pol4 lacking the N-terminal 112 amino acids encompassing the breast cancer susceptibility gene 1 C terminus (BRCT) domain (see Fig. 1) that was originally identified in the breast cancer susceptibility gene BRCA1 (47,48) and the plasmid pGST-BRCT that encodes the N-terminal 112-amino acid BRCT domain as a GST fusion protein.
Purification of His-tagged Pol4 and Pol4⌬BRCT-Overnight cultures (100 ml) of E. coli BL21 (DE3) cells harboring either pET28b-Pol4 or pET28b-Pol4⌬BRCT were inoculated into 2 liters of LB medium containing kanamycin (0.025 mg/ml) and chloramphenicol (0.034 mg/ ml) and grown at 37°C. At an absorbance at 600 nm of 0.5, isopropyl-␤-D-thiogalactopyranoside was added to a final concentration of 0.5 mM, and growth was continued at 25°C for 4 h. The cells were harvested by centrifugation, flash frozen, and stored at Ϫ80°C. Frozen cells were resuspended in 40 ml of lysis buffer (50 mM Tris-HCl, pH 7.5, 300 mM NaCl, 10% glycerol, 10 mM 2-mercaptoethanol, 0.1% Nonidet P-40, 1 mM phenylmethanesulfonyl fluoride, 1 mM benzamidine HCl, 1 g/ml leupeptin, 2 g/ml aprotinin, and 1 g/ml pepstatin) and lysed by sonication. After centrifugation at 15,000 rpm for 20 min at 4°C, the cleared lysate was supplemented with imidazole to a final concentration of 20 mM prior to incubation with 1 ml of nickel-nitrilotriacetic acid-agarose beads (Qiagen) for 2 h at 4°C. The beads were collected by centrifugation and then washed extensively with lysis buffer containing 40 mM imidazole. His-tagged Pol4 polypeptides were eluted with lysis buffer containing 250 mM imidazole and then further purified to near homogeneity by Resource Q and Resource S column chromatography. Approximately 40 g of Pol4 and 30 g of Pol4⌬BRCT were obtained from 2-liter cultures. Protein concentrations were measured by the Bradford assay (49) using bovine serum albumin as the standard.
Purification of GST Fusion Proteins-An overnight culture (1 liter) of E. coli BL21 (DE3) cells harboring either pGST-Pol4 or pGST-Pol4⌬BRCT were inoculated into 10 liters of LB medium containing ampicillin (0.1 mg/ml) and chloramphenicol (0.034 mg/ml) and grown at 37°C. When the absorbance at 600 nm reached 0.6, isopropyl-␤-Dthiogalactopyranoside was added to a final concentration of 0.5 mM, and growth was continued at 25°C for 2 h. The cells were harvested by centrifugation, flash frozen, and stored at Ϫ80°C. The frozen cells were resuspended in 100 ml of lysis buffer and lysed by sonication. After centrifugation, GST fusion proteins were purified from the cleared lysate by glutathione-Sepharose 4B affinity chromatography, gel filtration through a Superdex 75 column, and Resource S ion exchange chromatography. Approximately 30 g of nearly homogenous GST-Pol4 was obtained from the 10-liter culture. A similar quantity of GST-Pol4⌬BRCT was obtained, but this preparation also contained two proteolytic fragments of the fusion protein. GST-BRCT and GST were purified to near homogeneity from E. coli BL21 (DE3) cells harboring either pGST-BRCT or pGSTag by glutathione-Sepharose 4B affinity chromatography and gel filtration through a Superdex 75 column.
Purification of Dnl4-Lif1-His-tagged Lif1 and complexes containing either Dnl4 and Lif1 or His-tagged Lif1 and Dnl4 were purified from yeast cells as described previously (24).
Preparation of Yeast Cell Extracts-The lysates were prepared from the yeast strain BJ5464 and from the same strain harboring the plasmids pADH-Dnl4 and pYES-Lif1 (24). The cells from 100-ml cultures were resuspended in 5 ml of Buffer A (50 mM Tris-HCl, pH 7.5, 100 mM NaCl, 10 mM MgCl 2 , 10% glycerol, 1 mM phenylmethanesulfonyl fluoride, 1 mM benzamidine HCl, 1 g/ml leupeptin, 2 g/ml aprotinin, and 1 g/ml pepstatin) and lysed by mechanical shearing with glass beads. After centrifugation, the supernatant was used immediately for pulldown assays.
Coupled in Vitro Transcription and Translation-Labeled Dnl4 was synthesized from a pBKS-Dnl4 template using T7 RNA polymerase and [ 35 S]methionine (Amersham Biosciences) in the TNT Quick Coupled transcription/translation system (Promega). Dnl4 was partially purified by ammonium sulfate precipitation (50) and then resuspended in 50 l of Buffer A.
GST Pull-down Assays-Glutathione-Sepharose 4B beads (10 l; Amersham Biosciences) were incubated with GST-Pol4, GST-Pol4⌬BRCT, GST-BRCT, or GST (10 g of each) at 4°C for 2 h. After centrifugation, the supernatant was removed, and the beads were used for the pull-down assays.
To detect associations between Dnl4 and Pol4, beads with either GST-Pol4 or GST as the ligand were incubated with 1.5 ml of yeast lysate at 4°C for 4 h. After collection by centrifugation, the beads were washed extensively with Buffer A and then incubated for 15 min at 25°C in 20-l reaction mixtures containing 60 mM Tris-HCl, pH 8.0, 10 mM MgCl 2 , 5 mM dithiothreitol, 50 g/ml bovine serum albumin, and 0.5 Ci of [␣-32 P]ATP (3000 Ci/mmol; Amersham Biosciences). The reactions were stopped by the addition of SDS sample buffer (51).
To detect a direct interaction between purified Dnl4-Lif1 complex and Pol4, beads (10 l) with GST-Pol4, GST-Pol4⌬BRCT, or GST as the ligand were incubated with 0.5 g of purified Dnl4-Lif1 complex (24) in Buffer A containing 2% bovine serum albumin in a final volume of 20 l at 4°C for 4 h. After collection by centrifugation, the beads were washed extensively with Buffer A containing 150 mM NaCl and then incubated with [␣-32 P]ATP as described above. To further characterize the interaction between Dnl4 and Pol4, beads (10 l) with the indicated ligand were incubated with labeled in vitro translated Dnl4 supplemented with 2% bovine serum albumin (22 l) for 4°C for 4 h. After centrifugation, the supernatant was removed, and the beads were washed with Buffer A. The beads were resuspended in 20 l of SDS sample buffer to yield the eluate (E). After electrophoresis through a 7.5% SDS-polyacrylamide gel (51), labeled Dnl4 was detected using a PhosphorImager screen.
Equal amounts of the labeled and unlabeled duplexes (100 nM) were incubated with Pol ␤, Pol4, Pol4⌬BRCT, and Dnl4-Lif1 as indicated in reaction mixtures (10 l) containing 35 mM Tris-HCl, pH 7.5, 10 mM MgCl 2 , 0.05 mM of each of the four dNTPs, and 1 mM ATP at 25°C. The reactions were stopped by the addition of gel loading buffer (95% (v/v) formamide, 0.09% (w/v) bromphenol blue, and 0.09% (w/v) xylene cyanol). After separation by denaturing gel electrophoresis, labeled DNA molecules in the dried gel were detected and quantitated by Phosphor-Imager analysis.
Ligation Assay-Linear duplexes with complementary single-strand ends were constructed by annealing the following pairs of oligonucleotides; 50M annealed to 5Ј-phosphorylated 41M and 5Ј-phosphorylated 43M annealed to 5Ј 32 P-labeled 51M (5Ј-GTCTGTCTCACTATTAGAA-CCCTTTAGAGTCATGCGTCGCGAGGCAACGCA-3Ј, 51-mer). Alignment of the complementary single strands generates a ligatable nick in both the unlabeled and labeled strand.
Equal amounts of the labeled and unlabeled duplexes (100 nM) were incubated with Dnl4-Lif1 and, where indicated, Pol4 and Pol4⌬BRCT as described above. In assays to measure both DNA synthesis and ligation, the 5Ј termini of unlabeled oligonucleotides were phosphorylated.

RESULTS
DNA Synthesis Activity of Pol4 -Studies on the efficiency of in vivo recircularization of linear plasmid DNA molecules with noncomplementary single-strand ends have implicated Pol4 in a subset of DNA joining events that involve end processing (32). Full-length and truncated versions of Pol4 lacking the N-terminal BRCT domain ( Fig. 1) were expressed as His-tagged polypeptides in E. coli and then purified to near homogeneity ( Fig. 2A, lanes 1 and 2). The DNA polymerase activity of Pol4 was compared with that of Pol ␤. As shown in Fig. 2B, Pol4 and Pol ␤ have similar gap filling activity on a DNA substrate containing a single-nucleotide gap. In contrast, Pol4 was at least 4-fold more effective at filling in a single-nucleotide gap generated by the alignment of linear duplex DNA molecules with complementary ends than Pol ␤ (Fig. 2C, compare lanes 4  and 5 with lanes 6 and 7). Moreover, it appears that this substrate specificity is an intrinsic property of the Pol4 catalytic domain because deletion of the BRCT domain had no significant effect on DNA synthesis activity (Fig. 2C, compare  lanes 2 and 3 with lanes 4 and 5).
Next we examined the DNA synthesis activity of Pol4 on differently sized gaps generated by the alignment of linear duplex DNA molecules with complementary ends (Fig. 3). There was an inverse relationship between gap size and the amount of fully filled-in product (Fig. 3). The presence or absence of a phosphate group at the 5Ј end of the gaps did not effect Pol4 DNA synthesis activity (data not shown). DNA synthesis by Pol4 was distributive, with the enzyme preferentially adding a single nucleotide, even with larger gaps. Although a ladder of products corresponding to incompletely filled gaps were detectable, the fully filled-in product was the most abundant species. With gaps sizes greater than a single nucleotide, strand displacement synthesis of one to two nucleotides occurred (Fig. 3, compare lanes 2 and 3). Interestingly, the presence of a short 5Ј flap increased the amount of strand displacement DNA synthesis about 2-fold (Fig. 3, compare  lane 3 with lanes 5 and 6). In summary, the DNA substrate specificity of Pol4 is compatible with its in vivo role in DSB repair because small gaps, possibly with noncomplementary flaps, are predicted intermediates in the subpathway of NHEJ that involves microhomology-mediated alignment of DNA ends (31,32).
Pol4 Interacts the Dnl4-Lif1 Complex via the Dnl4 Subunit-To detect associations between Pol4 and NHEJ factors, we expressed and purified Pol4 as a GST fusion protein (Fig.  4A, lane 2). In affinity chromatography experiments with extracts from a wild type yeast strain, no specific binding of the core NHEJ factors, Hdf1-Hdf2, Rad50-Mre11-Xrs2, and Dnl4-Lif1, to GST-Pol4 beads was observed (data not shown). We suspected that associations were not detected because of the low endogenous levels of NHEJ factors. Therefore, we performed a similar experiment with an extract from a strain overexpressing Dnl4-Lif1 (24) and observed the specific binding of Dnl4 to the GST-Pol4 resin (Fig. 4B). To determine whether there is a direct interaction between Pol4 and Dnl4-Lif1, glutathione beads with GST-tagged full-length Pol4 (Fig. 4A, lane  2), GST-tagged Pol4 lacking the N-terminal BRCT domain (Fig.  4A, lane 4), or GST (Fig. 4A, lane 1) as the ligand were incubated with purified Dnl4-Lif1 that was then labeled by adenylation (24). The binding of Dnl4-Lif1 to the GST-Pol4 beads but not to either the GST-Pol4⌬BRCT or the GST beads (Fig. 4C) demonstrates that these protein factors interact directly and suggests that this interaction is mediated by the N-terminal BRCT domain of Pol4. In similar experiments, we did not observe specific binding of purified Lif1 to GST-Pol4 beads (data not shown), suggesting that the interaction is either mediated by Dnl4 or requires complex formation between Dnl4-Lif1. Because Lif1 is required for Dnl4 stability in yeast cells (9,24), we examined the interaction of labeled in vitro translated Dnl4 with Pol4. Dnl4 bound to glutathione beads with either GST-Pol4 (Fig. 4A, lane 2) or a GST fusion protein with only the N-terminal BRCT domain of Pol4 (Fig. 4A, lane 3) as the ligand but did not bind to glutathione beads with GST (Fig. 4A, lane 1) as the ligand (Fig. 4D). In similar experiments, in vitro translated Lif1 did not bind specifically to Pol4 beads (data not shown). Thus, we conclude that Pol4 interacts directly with the Dnl4 subunit of the Dnl4-Lif1 complex via its N-terminal BRCT motif.
Dnl4-Lif1 Specifically Stimulates the DNA Polymerase Activity of Pol4 -To elucidate the functional consequences of the interaction between Pol4 and Dnl4-Lif1, we examined the effect of Dnl4-Lif1 on gap filling DNA synthesis catalyzed by Pol4. As shown in Fig. 5A, Dnl4-Lif1 greatly stimulates DNA synthesis  by Pol4 on nonligatable gaps formed by the alignment of partial duplex oligonucleotides in a concentration-dependent manner. Notably, Dnl4-Lif1 is much more effective at stimulating fulllength Pol4 compared with either a truncated version of Pol4 lacking the BRCT motif (Fig. 5A) or Pol ␤ (data not shown). At a ratio of about 1:1, Dnl4-Lif1 increased Pol4-catalyzed DNA synthesis by 5-6-fold. When measured as a function of time, Dnl4-Lif1 increased both the rate and extent of Pol4-mediated DNA synthesis (Fig. 5B). Again the effect on DNA synthesis was dependent on the BRCT domain of Pol4. These results demonstrate that the stimulation of Pol4 DNA synthesis activity is mediated, at least in part, by the protein-protein interaction between Dnl4-Lif1 and Pol4. Dnl4-Lif1 also stimulated gap filling DNA synthesis by Pol4 on a linear duplex containing a single nucleotide nonligatable gap in a BRCT domaindependent manner (Fig. 5C). Because the stimulatory effect of Dnl4-Lif1 on DNA synthesis by Pol4 was similar whether the gap was within a linear DNA duplex or formed by the alignment of linear duplex DNA molecules with complementary ends (Fig. 5, A and C), it appears that the stimulation of Pol4 activity by Dnl4-Lif1 is mediated by mechanisms other than DNA end alignment. Finally, we examined whether the effect of Dnl4-Lif1 on the DNA synthesis activity of Pol4 was influenced by either increased gap size or the presence of a 5Ј flap. Dnl4-Lif1 stimulated the DNA synthesis activity of Pol4 on cohesive ended DNA molecules that when aligned form a threenucleotide gap either without (Fig. 6A) or with a three-nucleotide flap (Fig. 6B) but did not significantly alter the distribution of reaction products.
Pol4 Specifically Stimulates the DNA Joining Activity of Dnl4-Lif1-In assays with partial duplex oligonucleotides that when aligned form a ligatable structure, Dnl4-Lif1 exhibited a low activity that was stimulated by intact Pol4 (Fig. 7A) but not by the truncated version of Pol4 lacking the N-terminal BRCT domain. At a ratio of about 1:1, Pol4 increased Dnl4-Lif1catalyzed DNA joining by about 2-fold. These results demonstrate that the stimulation of DNA joining by Dnl4-Lif1 is mediated, at least in part, by the protein-protein interaction between Dnl4-Lif1 and Pol4, but the magnitude of this effect is less than that of Dnl4-Lif1 on Pol4 DNA synthesis activity.
These observations suggest that the interaction between Pol4 and Dnl4-Lif1 co-ordinates the gap filling DNA synthesis and ligation reactions that complete NHEJ. To provide support for this model, we compared the amount of ligated product produced by Pol4 and Dnl4-Lif1 in assays with oligonucleotide duplexes that when aligned either form a ligatable nick (Fig.  7A) or a single nucleotide gap (Fig. 7B). As shown in Fig. 7C, joining of the DNA substrate that requires both DNA synthesis and ligation (15 fmol) was significantly higher than the joining of the DNA substrate requiring only ligation (0.5-1 fmol). This synergistic effect indicates that the interaction between Pol4 and Dnl4-Lif1 not only enhances the catalytic activity of both these enzymes but co-ordinates their action, ensuring the efficient hand-over of pathway intermediates. DISCUSSION The preferred pathway for the repair of in vivo DSBs by NHEJ appears to involve microhomologies that are presumably revealed by nucleolytic digestion (14,29,30,34). After the alignment of short complementary sequences, it is likely that single-strand flaps are removed and gaps are filled in to gen-erate ligatable structures (31,32). Our biochemical studies have revealed differences in the DNA substrate specificity of the catalytic domain of Pol4 compared with Pol ␤, the prototypic member of the Pol X DNA polymerase family (Fig. 1) (36, 37, 40 -43) that are compatible with the predicted role of the POL4 gene product in end joining events that involve gap filling DNA synthesis (32). Specifically, Pol4 preferentially acts upon short gaps formed by the alignment of linear duplexes with complementary single-strand ends.
Hdf1-Hdf2, Rad50-Mre11-Xrs2, and Dnl4-Lif1 complexes have been identified as critical factors in the major NHEJ pathway in yeast (3)(4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14). Rad50-Mre11-Xrs2 stimulates DNA joining both by aligning DNA ends and by recruiting Dnl4-Lif1 via a direct interaction between Xrs2 and Lif1 (24). Furthermore, DNA end joining mediated by these factors is dependent on Hdf1-Hdf2 at physiological salt concentrations, suggesting that Hdf1-Hdf2 binds to DNA ends and facilitates the subsequent recruitment of Rad50-Mre11-Xrs2 and Dnl4-Lif1 (24). Although these in vitro studies involved DNA substrates with complementary single-strand ends, it seems likely that the same core factors will bring together noncomplementary DNA ends within a nucleoprotein complex and that the subsequent end-processing reactions to generate a ligatable structure will occur within the context of the nucleoprotein complex. Consistent with this idea, we have shown that the interaction between Pol4 and Dnl4-Lif1 not only increases the catalytic activity of both of these enzymes but also couples gap filling DNA synthesis with DNA joining. Because Dnl4-Lif1 stimulates Pol4 DNA synthesis on a gap within a linear duplex and the stimulatory effect of Pol4 on Dnl4-Lif1 is much lower than that of Rad50-Mre11-Xrs2 (24), it seems unlikely that Pol4 and Dnl4-Lif1 have robust end bridging activity. XRCC4, the human homolog of Lif1, preferentially binds to DNA ends and nicked DNA (52). Thus, it is possible that Lif1 binding to DNA nicks or gaps stabilizes the assembly of the DNA polymerase-DNA ligase complex in a manner analagous to the functional interaction between Pol ␤ and XRCC1, the partner protein of DNA ligase III␣, on nicked DNA (53,54).
Analysis of the assembly of NHEJ factors at in vivo DSBs by chromatin immunoprecipitation has shown that the recruitment of Dnl4 to DSBs is dependent upon Hdf1-Hdf2 and Lif1 (55). Because inactivation of POL4 has no effect on the repair of DNA breaks with cohesive ends (32), it appears that the Hdf1-Hdf2, Rad50-Mre11-Xrs2, and Dnl4-Lif1 factors can form a functional nucleoprotein complex in the absence of Pol4. Thus, the recruitment of Pol4 to in vivo DSBs may be mediated via its interaction with Dnl4. Our biochemical studies demonstrating that the BRCT domain of Pol4 is critical for the interaction with Dnl4-Lif1 but not for Pol4 DNA synthesis activity provide a molecular explanation for genetic studies showing that N-terminal deletions inactivate Pol4 function in vivo (32). Interestingly, several members of the Pol X family have an N-terminal BRCT domain ( Fig. 1) (36, 37, 40 -42, 45, 47, 48), suggesting that this may be a common mechanism for recruiting these enzymes to their in vivo substrates. This idea is supported by a recent report describing interactions between the human Pol X family members, Pol and terminal transferase, and DNA ligase IV-XRCC4 (56).
In summary, Pol4 efficiently fills in short gaps formed by the alignment of complementary single strands at the ends of duplex DNA and is specifically stimulated by Dnl4-Lif1. It will be interesting to determine whether end bridging by Rad50-Mre11-Xrs2 (24) will further stimulate the coupled DNA synthesis and ligation reaction mediated by Pol4 and Dnl4-Lif1. Moreover, genetic studies indicate that Pol4 participates in end joining events that require nucleolytic processing (32). Future studies are necessary to determine whether Pol4 specifically associates with and/or modulates the activity of nucleases such as Mre11 (33,34) and Fen-1 (31) within the nucleoprotein structure formed by NHEJ factors.