The human Rad9–Rad1–Hus1 checkpoint complex stimulates flap endonuclease 1

Abstract

The toroidal damage checkpoint complex Rad9–Rad1–Hus1 (9-1-1) has been characterized as a sensor of DNA damage. Flap endonuclease 1 (FEN1) is a structure-specific nuclease involved both in removing initiator RNA from Okazaki fragments and in DNA repair pathways. FEN1 activity is stimulated by proliferating cell nuclear antigen (PCNA), a toroidal sliding clamp that acts as a platform for DNA replication and repair complexes. We show that 9-1-1 also binds and stimulates FEN1. Stimulation is observed on a variety of flap, nick, and gapped substrates simulating repair intermediates. Blocking 9-1-1 entry to the double strands prevents a portion of the stimulation. Like PCNA stimulation, 9-1-1 stimulation cannot circumvent the tracking mechanism by which FEN1 enters the substrate; however, 9-1-1 does not substitute for PCNA in the stimulation of DNA polymerase β. This suggests that 9-1-1 is a damage-specific activator of FEN1.

Flap endonuclease 1 (FEN1) is the primary nuclease involved in the removal of the RNA primers from Okazaki fragments (1). Deletion of the FEN1 gene in Saccharomyces cerevisiae produces temperature-sensitive growth and a phenotype common to DNA replication mutations (2–4). FEN1 is also a key nuclease in long-patch base-excision repair, a major pathway in S. cerevisiae (5–7). FEN1 cleaves a 5′ flap substrate produced by strand-displacement synthesis during replication or repair. FEN1 cleavage activity is stimulated by proliferating cell nuclear antigen (PCNA) (8–10). In eukaryotes, long-patch base-excision repair was found to be PCNA-dependent (11–13). PCNA encircles the double-stranded region of the flap substrate and improves FEN1 binding to the cleavage site at the base of the flap (10).

FEN1 has also been implicated in replication fork restart. A stalled fork can regress into a four-way junction called a chicken foot, which is structurally equivalent to a Holliday recombination junction (14, 15). The regression provides a mechanism of damage repair. Werner's protein unwinds this intermediate and creates a substrate for FEN1 (16). The Werner's protein–FEN1 interaction stimulates FEN1 to cleave the strands necessary to restore replication fork structure.

DNA damage evokes a cellular response that inhibits DNA replication but allows DNA repair (17). The damage response in eukaryotic cells involves activation of the ATM and ATR proteins. The ATM and ATR kinases activate checkpoint control proteins. ATM is activated in response to double-strand breaks, whereas ATR is activated in response to stalled replication forks and to a variety of damage that causes distortions and single strands (18). Rad9–Rad1–Hus1 (9-1-1) is a toroidal molecule that is loaded onto DNA by Rad17–RFC, a variation of the traditional clamp loader RFC (19). The 9-1-1 complex and ATR are recruited independently to damaged sites (20). The current model suggests that the 9-1-1 complex and ATR act as damage sensors and, therefore, participate in the activation of proteins that promote cell survival.

The damage response also induces the transcription factor p53, which increases the production of p21, causing growth arrest (17, 21). The p21 protein has two domains. The amino terminus inhibits cyclin-dependent kinases, and the carboxyl terminus binds and inhibits PCNA (22–26). Expression of the PCNA-binding domain alone is sufficient to cause arrest of cells in G₁ (21, 25, 27). This arrest is presumed to result from disruption of the interaction between DNA polymerase β (pol β) and PCNA that is necessary for highly processive DNA synthesis (26, 28). The p21 protein can also interfere with FEN1–PCNA interactions in vitro and in vivo (29). Furthermore, we have shown that concentrations of the PCNA-inactivating peptide that render pol β ineffective also prevent PCNA stimulation of FEN1 (30).

It is not clear how the inhibitory functions of p21 relate to DNA repair. Some results suggest that expression of p21 is compatible with, or even important for, DNA repair. For example, mice deficient in p21 are highly sensitive to radiation (31). Furthermore, cells expressing p21 have increased ability to repair damaged plasmids (32). If PCNA's stimulatory function is compromised by damage-related induction of p21, perhaps the residual functions of FEN1 are sufficient for repair. Alternatively, we hypothesized that another molecule may be a damage-specific substitute for PCNA. The 9-1-1 checkpoint complex and the Rad17–RFC resemble the RFC clamp loader and PCNA, both structurally and functionally (19, 33–36). We considered whether the 9-1-1 complex could be a PCNA substitute.

Materials and Methods

Substrate Preparation. Oligonucleotide substrates from Midland Certified Reagents (Midland, TX) or Integrated DNA Technologies (Coralville, IA) were designed to mimic DNA replication and repair intermediates with an upstream primer annealed to the 3′ end of a template and a downstream primer annealed to the 5′ end of a template. Sequences are listed in Table 1. The primer sequences are listed 5′ to 3′ and the template sequences are listed 3′ to 5′ to facilitate visual alignment. For any flap substrate, the unannealed region is measured from the 5′ end of the downstream primer. Upstream primer U₁ and downstream primers D₁ through D₅ were 5′-labeled by T4 polynucleotide kinase (Roche Molecular Biochemicals) by using [γ-³²P]ATP (PerkinElmer). Downstream primers D₆ and D₇ were 3′-labeled by annealing 20 pmol of primer to 50 pmol of template containing a G overhang at the 5′ end. [α-³²P]dCTP (PerkinElmer) was added to the 3′ end of the primer with the Klenow fragment of Escherichia coli DNA polymerase I (Roche Molecular Biochemicals). Labeled primers were purified by electrophoresis and isolation from a 15% polyacrylamide/7 M urea gel.

Table 1. Oligonucleotide sequences.

Primer	Length, nt	Oligonucleotide sequence
Upstream (5′ to 3′)
U1	20	CGCGGATGCGCTCTGGGTCC
U2	24	CGACCGTGCCAGCCTAAATTTCAA
U3	24	CGCCAGGGTTTTCCCAGTCACGAC
U4	25	CGACCGTGCCAGCCTAAATTTCAAT
U5*	25	CGCCAGGGTTTTCCCAGTCACGACC
U6	25	CCAGCACTGACCCAAAAGGGACCGC
U7	26	CGACCGTGCCAGCCTAAATTTCAATA
U8	26	CGCCAGGGTTTTCCCAGTCACGACCA
Downstream (5′ to 3′)
D1	18	GTAAAACGACGGCCAGTG
D2	19	AGTAAAACGACGGCCAGTG
D3	28	GCACCCGTCCACCCGACGCCACCTCCTG
D4	43	CGTACGGACGTAGAGCTGTTTCCAAGTAAAACGACGGCCAGTG
D5	55	AGGTCTCGACTAACTCTAGTCGTTGTTCCACCCGTCCACCCGACGCCACCTCCTG
D6	63	GGTCGTGACTGGGAAAACCACCCGTCCACCCGACGCCACCTCCTGCACTGGCCGTCGTTTTAC
D7	69	GACTCACGTAGAGTGTCTTTTTTTTTTTTTTTTTTTTTTTTCCACCCGTCCACCCGACGCCACCTCCTG
Template (3′ to 5′)
T1*	44	GCGGTCCCAAAAGGGTCAGTGCTGGCATTTTGCTGCCGGTCACG
T2	51	ACTGGCACGGTCGGATTTAAAGTTAGGGCAGGTGGACTGCGGTGGAGGACG
T3	54	GCTGGCACGGTCGGATTTAAAGTTAGGTGGGCAGGTGGGCTGCGGTGGAGGACG
T4	71	GCTGGCACGGTCGGATTTAAAGTTAATTGAGATCAGCAACAAGGTGGGCAGGTGGGCTGCGGTGGAGGACG
T5	73	GGTCGTGACTGGGTTTTCCCTGGCGCTGAGTGCATCTCACAGTTTGGTGGGCAGGTGGGCTGCGGTGGAGGAC
T6	110	GCGCCTACGCGAGACCCAGGACCGCATACGGATCAGTACTCGGATCTAGCTAGCAGTCTCGTTGCGTAGCCAGCTACCT CGGACGTTACATAGCGAGCTTAAGTCACGGG

^*The underlined nucleotide indicates biotin modification when noted in Materials and Methods.

To anneal substrates, primers were incubated at 95°C for 5 min, transferred to 70°C, and then slowly cooled to room temperature in annealing buffer containing 10 mM Tris·HCl (pH 8.0) and 1 mM EDTA. Substrate used for the pol β assay was annealed in a 1:2 ratio of labeled upstream primer to template. Substrates used for FEN1 assays were annealed in a 1:2:4 ratio of labeled downstream primer to template to upstream primer with the exception of the double-flap substrate D₆·T₂·U₇ and bubble substrate D₇·T₅·U_6, which were annealed in a ratio of 1:2:5 under conditions described in ref. 37.

Stimulation of FEN1 Cleavage. FEN1 (38), PCNA (39), pol β (39), and 9-1-1 (40) were prepared as previously described. Twenty-microliter reaction mixtures containing the indicated quantities of enzymes and 5 fmol of 5′ ³²P-radiolabeled DNA substrate were incubated at 37°C for 3 min, then the reactions were stopped with 20 μl of 2× termination dye [90% formamide (vol/vol) with 10 mM EDTA, 0.1% bromophenol blue, and 0.1% xylene cyanole]. The reaction buffer contained 30 mM Hepes–NaOH (pH 7.5), 0.1 mg/ml BSA, 4 mM MgCl₂, and 40 mM KCl. After termination, samples were heated at 95°C for 5 min and loaded onto a preheated denaturing (7 M urea) 15% polyacrylamide gel. After electrophoresis, the gel was dried and scanned by using a Molecular Dynamics PhosphorImager to detect products. Pixel densities were determined by using imagequant version 1.2 software (Molecular Dynamics).

Stimulation Time-Course Reaction. A control reaction mixture containing 11 fmol of FEN1 in 110 μl of reaction buffer [30 mM Hepes–NaOH (pH 7.5)/40 mM KCl/4 mM MgCl₂/0.1 mg/ml BSA/0.1% Nonidet P-40/0.1 M DTT] was initiated at time 0 by adding 110 μl of reaction buffer containing 5.5 fmol of the indicated substrate. Reaction temperature was 37°C. At times 24 sec, 36 sec, 48 sec, 1 min, 1.2 min, 1.8 min, 2.2 min, 2.8 min, and 3.6 min, 20 μl of reaction mixture was removed and mixed with 20 μl of 2× termination dye. Another reaction of 9-1-1 complex was carried out in parallel by mixing 110 μl of reaction buffer containing 11 fmol of FEN1 and 220 fmol of 9-1-1 complex with 110 μl of reaction buffer containing 5.5 fmol of the indicated substrate at time 0. At the same times noted above, 20 μl of reaction volume was removed and mixed immediately with 20 μl of 2× termination dye. After termination, samples were heated at 95°C for 5 min and loaded onto a preheated denaturing (7 M urea) 15% polyacrylamide gel. The gel was dried and scanned by PhosphorImager to detect products. Pixel densities were determined by using imagequant.

FEN1 Cleavage of a Blocked Flap Substrate. Substrate D₄·T₁·U₅ (10 fmol) with 5′ biotinylation on T₁ and U₅ was incubated with streptavidin (500 fmol) for 20 min on ice in reaction buffer containing 30 mM Hepes–NaOH (pH 7.5), 0.1 mg/ml BSA, 4 mM MgCl₂, and 40 mM KCl. Unblocked substrate (minus streptavidin) and blocked substrate (plus streptavidin) were then incubated with FEN1 (4 fmol) and either PCNA or 9-1-1 for 3 min at 37°C in the same reaction buffer. Reactions were stopped by the addition of 10 μlof2× termination dye and heated at 95°C for 5 min. Denatured reaction products were separated on a 12% polyacrylamide/7 M urea gel. The gel was scanned by using a PhosphorImager and analyzed with imagequant.

Coimmunoprecipitation of 9-1-1 and FEN1. Immunoprecipitation was performed according to standard methods. Briefly, protein G-Sepharose beads (Sigma) were equilibrated with cell-lysis buffer [10 mM Hepes (pH 7.6)/150 mM KCl/10 mM MgCl₂/50 μg/ml digitonin] (41) containing 0.1 mM PMSF, and a Protease Inhibitor Mixture tablet (Roche Diagnostics). RAD1 goat polyclonal antibody (Santa Cruz Biotechnology) was mixed with protein G-Sepharose beads in 500 μl of lysis buffer and rotated for 1 h at 4°C. Then, precleared HeLa cell lysate containing 1 mg of total protein was added and the mixture was rotated for 1 h at 4°C. After sedimentation, the supernatant fluid was discarded and the beads were washed three times with lysis buffer. The beads were then boiled in sample buffer [62.5 mM Tris·Cl (pH 6.8)/5% glycerol/2% SDS/0.02% bromophenol blue]. The immunoprecipitates were separated by SDS/10% PAGE. Western blot analysis was performed with anti-FEN1 polyclonal antibody (Novus Biologicals, Littleton, CO) and anti-RAD1 polyclonal antibody.

pol β Stimulation Assays. Substrate (5 fmol) was incubated with pol β (1.7 units) and either PCNA or 9-1-1 in buffer containing 40 mM Tris·HCl (pH 7.5), 0.2 mg/ml BSA, 1 mM DTT, 4 mM MgCl₂, 1 mM ATP, and 50 μM each dNTP for 8 min at 37°C. A control reaction with 2units of Klenow fragment of E. coli DNA polymerase I was incubated in the same buffer for 8 min at 37°C to produce and mark the position of full-length synthesis product. Reactions were stopped by the addition of 10 μl of 2× termination dye and incubation for 5 min at 95°C. Denatured reaction products were separated on a 10% polyacrylamide/7M urea gel. Gels were scanned by PhosphorImager and analyzed with imagequant.

Results

9-1-1 Checkpoint Complex Stimulates FEN1 Cleavage Uniformly Over Time on Substrates with Various Length Flaps. We hypothesized that the 9-1-1 complex is the repair-specific substitute for PCNA in stimulating FEN1. If true, 9-1-1 should stimulate FEN1 on a gamut of potential substrates present at repair sites. We tested the capacity of 9-1-1 to stimulate FEN1 on substrates with a range of flap lengths as well as on nicked and gapped substrates. 9-1-1 complex stimulated FEN1 cleavage on all of these substrates between 10- and 50-fold. Fig. 1A demonstrates that increasing 9-1-1 concentration stimulates FEN1 cleavage of a 27-nt flap composed of oligonucleotides D₅, T₃, and U₄. The difference between FEN1 alone (lane 3) and FEN1 with 500 fmol of 9-1-1 complex (lane 7) is between 40- and 50-fold. Fig. 1 B–E shows the quantified results of 9-1-1 stimulation on substrates with 27-, 25-, 10- and 1-nt flap lengths, respectively, and Fig. 1F shows 9-1-1 stimulation on a nicked and 1-nt gap substrate.

Fig. 1.

9-1-1 stimulates FEN1 activity on a variety of substrates. (A) FEN1 cleavage on a 27-nt nick flap (D₅·T₃·U_4, shown schematically at the top with * indicating ³²P) in the presence of increasing 9-1-1. Reaction mixtures with FEN1 contain 0.5 fmol of FEN1. Lane 2 contains 500 fmol of 9-1-1. Lanes 4–7 contain 50, 100, 200, and 500 fmol of 9-1-1. (B) FEN1 stimulation by 9-1-1 on a 27-nt flap with various upstream primers. Short-dashed line, 27-nt double flap (df) (D₅·T₃·U₇); solid line, 27-nt nick flap (nf) (D₅·T₃·U₄); long-dashed line, 27-nt gap flap (gf) (D₅·T₃·U₂). (C) FEN1 stimulation by 9-1-1 on a 25-nt flap with various upstream primers. Short-dashed line, 25-nt double flap (D₄·T₁·U₈); solid line, 25-nt nick flap (D₄·T₁·U₅); long-dashed line, 25-nt gap flap (D₄·T₁·U₃). (D) FEN1 stimulation by 9-1-1 on a 10-nt flap with various upstream primers. Short-dashed line, 10-nt double flap (D₅·T₄·U₇); solid line, 10-nt nick flap (D₅·T₄·U₄); long-dashed line, 10-nt gap flap (D₅·T₄·U₂). (E) FEN1 stimulation by 9-1-1 on a 1-nt flap with various upstream primers. Short-dashed line, 1-nt double flap (D₂·T₁·U₈); solid line, 1-nt nick flap (D₂·T₁·U₅); long-dashed line, 1-nt gap flap (D₂·T₁·U₃). (F) FEN1 stimulation on a nicked and gapped substrate. Solid line, nicked substrate (D₁·T₁·U₅); long-dashed line, gapped substrate (D₁·T₁·U₃).

To determine whether stimulation by 9-1-1 was uniform over the course of the reaction, we performed a FEN1-cleavage time course in the absence and presence of 9-1-1. Fig. 2A shows a scan of the polyacrylamide gel and Fig. 2B shows the quantified results demonstrating uniform stimulation throughout the 3.6 min course of the reaction.

Fig. 2.

9-1-1 uniformly stimulates FEN1 cleavage. (A) Stimulation of FEN1 cleavage was assayed in a time-course reaction (0–3.6 min) in the absence or presence of 9-1-1 on a 27-nt-nick flap substrate (D₅·T₃·U₄) as described in Materials and Methods. The starting material is 55 nt, and the 27-nt product marks the FEN1 cleavage of the flap. (B) Quantified results of stimulation time course. ▴, FEN1 alone control; ▪, FEN1 and 9-1-1 complex together. Stimulation at each time point was ≈10- to 15-fold.

9-1-1 Stimulates FEN1 Cleavage on Substrates with Different Configurations of Upstream Primer. FEN1 is known to be active on flap substrates with different upstream primer characteristics. We showed previously (42) that the preferred FEN1 substrate is a double-flap substrate containing a 1-nt tail on the 3′ end of the upstream primer in addition to the 5′ flap of the downstream primer. Nick flaps that do not contain a 1-nt 3′ tail on the upstream primer are also cleaved by FEN1. Substrates with a 1-nt gap (hereafter called “gap flaps”) between the upstream primer and the first annealed nucleotide of the downstream primer show greatly diminished cleavage activity. Previous work (10) showed that PCNA stimulates FEN1 activity on all three classes of flap substrates. We tested whether the 9-1-1 complex stimulates FEN1 activity similarly. Fig. 1 B–E demonstrates that 9-1-1 stimulates the activity of FEN1 on double flaps, nick flaps, and gap flaps of four distinct sequences and lengths.

Blocking the Double Strands Prevents a Portion of the Stimulation. PCNA can stimulate cleavage by FEN1 on a flap substrate with a linear double-stranded region because the toroidal PCNA molecule can slide over the end of the double strand. Blocking the ends of the double-strand region with biotin–streptavidin prevents stimulation by PCNA (43). When we used a biotin–streptavidin-blocked flap substrate, PCNA stimulation of FEN1 activity was substantially reduced (Fig. 3, lanes 5 and 6 compared with lanes 10 and 11). Furthermore, the same substrate reduced stimulation of FEN1 cleavage by 9-1-1 (Fig. 3, lanes 7 and 8 compared with lanes 12 and 13). A residual stimulation was seen with both rings and was greater with 9-1-1. Some residual stimulation probably occurs because the substrate is not totally biotinylated. At least some stimulation by 9-1-1 appears to occur by a mechanism independent of loading onto the double strands. Maximum stimulation by both PCNA and 9-1-1 requires substantial stoichiometric excess of the ring proteins with the linear substrates used here. This required excess is evidently because the rings must enter the substrate by sliding over the double-stranded ends, rather than through the clamp-loading process used in vivo. Overall results indicate that 9-1-1 may stimulate FEN1 by entering the double strands or through a direct binding mechanism.

Fig. 3.

Blocking 9-1-1 entry to the double strands prevents a portion of the stimulation. Stimulation of FEN1 cleavage was assayed on unblocked (lanes 1–8) and blocked (lanes 9–13) 25-nt nick-flap substrates (D₄·T₁·U₅) in the presence of increasing amounts of PCNA (3 or 6 pmol) or 9-1-1 (40 or 80 fmol) as described in Materials and Methods. The starting material is 43 nt, and FEN1 cleavage of the flap is marked by the 25- and 24-nt products. On unblocked substrate, the band densities of lanes 5–8 are 6×,9×,2×, and 8×, respectively, relative to FEN1 alone in lane 4. On blocked substrate, the band densities of lanes 10–13 are 2.5×,2×, 1.5×, and 4×, respectively, relative to FEN1 alone in lane 9.

As with PCNA, 9-1-1 Cannot Alter the FEN1 Tracking Mechanism. We previously demonstrated that, although PCNA could augment FEN1-cleavage activity, it could not subvert the requirement for FEN1 to enter flaps from the 5′ end and track to the point of cleavage (10). This inability was demonstrated by using bubble substrates in which the 5′ end of the flap was annealed to the template. Using a similar bubble substrate, we examined the ability of PCNA and 9-1-1 to stimulate FEN-1.

Both PCNA and 9-1-1 were able to stimulate FEN1 cleavage of a 27-nt flap substrate (Fig. 4, lanes 4 and 5 and lanes 14 and 15). The bubble substrate is refractory to cleavage by FEN1 (lanes 8 and 18). As shown previously, PCNA cannot overcome the inhibitory effects of the annealed 5′ end (lanes 9 and 10). Similarly, 9-1-1 cannot activate FEN1 to use the bubble substrate (lanes 19 and 20). In this manner 9-1-1 behaves similarly to PCNA, in that it cannot alter the substrate entry mechanism of FEN1.

Fig. 4.

9-1-1 does not stimulate FEN1 cleavage of a bubble substrate. Ten microliters of substrate (10 fmol) with a 37-nt flap or 24-nt bubble was mixed with 10 μl of enzymes containing 1 fmol of FEN1 and increasing amounts of PCNA (3 or 6 pmol) or 9-1-1 (0.5 or 1 pmol). The reactions took place at 37°C for 10 min. The ramps denote the different amounts of PCNA or 9-1-1. The * indicates the radiolabeled position. The sizes of substrate and cleavage product are indicated by arrows. The substrate shown in lanes 1–5 and 11–15 is derived from the annealing of D₆·T₃·U₇. The substrate shown in lanes 6–10 and 16–20 is from the annealing of D₇·T₅·U₆. The addition of PCNA to a flap substrate augments the cleavage product (lanes 4 and 5) compared with that of FEN1 only (lane 3). As with PCNA, more cleavage products are observed after addition of 9-1-1 (lanes 14 and 15) compared with those with FEN1 only (lane 13). The addition of PCNA (lanes 9 and 10) or 9-1-1 (lanes 19 and 20) to the bubble substrate does not lead to a visible cleavage product.

The 9-1-1 Complex Interacts with FEN-1. Because the 9-1-1 complex stimulated FEN1 activity, we anticipated that the two proteins would display a detectable affinity. We detected interaction by coimmunoprecipitation of 9-1-1 and FEN-1 from HeLa cell lysates. As shown in lane 2 of Fig. 5 Upper, immunoprecipitation with anti-hRad1 antibody successfully captured hFEN-1; however, we do not know whether this interaction is direct or mediated by one of the other proteins in the 9-1-1 ring.

Fig. 5.

9-1-1 interacts with FEN1. Immunoprecipitation was performed as described in Materials and Methods. HeLa cell extract was immunoprecipitated (IP) with anti-human (h)Rad1 polyclonal antibody. Lane 1, whole cell lysate mixed only with protein G-Sepharose beads (mock immunoprecipitation containing no antibody). Lane 2, immunoprecipitation with anti-hRad1 antibody. The weak band larger than hFEN1 in Upper and the weak band smaller than hRad1 in Lower are antibody IgG heavy chain and light chain, respectively. Lane 3, supernatant fluid of third wash after immunoprecipitation. Lane 4, whole cell lysate control. Lane 5, purified recombinant human FEN1 control. Lane 6, purified recombinant human Rad1 control. Upper is blotted with anti-hFEN1 polyclonal antibody; Lower is blotted with anti-hRad1 antibody. Arrows indicate the positions of endogenous hFEN1 and hRad1. The band indicated by an asterisk is likely partially degraded recombinant hRad1. Because purified recombinant hFEN1 contains a His₆ tag, and purified recombinant hRad1 is tagged with a Flag-cAMP kinase motif, the sizes of both proteins are larger than endogenous hFEN1 and hRad1 (19).

9-1-1 Fails to Stimulate pol β Synthesis. Specificity of the damage response would predict that inactivation of PCNA by p21 would prevent active and processive synthesis by pol β. Furthermore, 9-1-1 should not substitute for PCNA in promoting DNA synthesis. We tested whether the 9-1-1 stimulates replication by pol β in a manner similar to PCNA stimulation. DNA synthesis was measured on a primer–template substrate by observing extension of a labeled upstream primer in the presence of either PCNA or 9-1-1 (Fig. 6). The position of the full-length primer-extension product is marked by reaction of the substrate with an excess Klenow fragment of E. coli DNA polymerase I (Fig. 6, lane 1). Incubation of the substrate with pol β alone does not show synthesis to the end of the template (Fig. 6, lane 4). Yet, with addition of increasing amounts of PCNA, pol β extends some products to the end of the template (Fig. 6, lanes 5–9). Full-length synthesis is stimulated substantially by PCNA (Fig. 6, lanes 5–9); however, addition of increasing amounts of 9-1-1 has only a minimal effect on product size distribution (Fig. 6, lanes 11–15). Only at the highest concentration of 9-1-1 (Fig. 6, lane 15) is there a slight stimulation of pol β, generating about 50% more processive synthesis to the end of the template.

Fig. 6.

9-1-1 fails to stimulate pol β synthesis. Synthesis by pol β was assayed in the presence of increasing amounts of PCNA (10, 50, 100, 200, and 500 fmol) or 9-1-1 (10, 50, 100, 200, and 500 fmol) on a primer–template substrate (T₆·U₁) as described in Materials and Methods. The 20-nt band marks the labeled primer, and the 110-nt product marks pol β full-length extension of the upstream primer.

Discussion

We show that the 9-1-1 complex interacts with FEN1 and stimulates its cleavage activity. Because 9-1-1 has been characterized as a DNA damage sensor (17), this result suggests that the stimulation relates to roles of FEN1 in DNA damage repair. Moreover, 9-1-1 resembles PCNA in that it is a toroid that encircles double-stranded DNA. The 9-1-1 complex stimulates FEN1 on the full range of substrates that respond to PCNA. The 9-1-1 complex may stimulate by entering the double-strand region or by direct interaction with FEN1. Also, just as with PCNA, the 9-1-1 complex can augment FEN1-cleavage activity but not change the need for FEN1 to enter the substrate from the 5′ end of the flap. On the basis of these properties, we propose that 9-1-1 acts as a damage-specific substitute for PCNA. When might FEN1 use 9-1-1 in place of PCNA? Acting as a damage sensor, 9-1-1 may gather at damage sites in preference to PCNA. Then FEN1 would use the 9-1-1 molecules to improve its interaction with any damage site at which a 9-1-1 complex is residing. In this model, FEN1 uses a PCNA platform during replication and the 9-1-1 platform for repair.

Additionally, 9-1-1 may replace PCNA when the active PCNA concentration is low, as might occur during induction of p21. The ability of p21 to produce a G₁ arrest and to inhibit DNA replication by preventing PCNA from acting as the sliding clamp for pol β suggests a role of p21 in the inhibition of replication. The effects of p21 induction on DNA repair are much less clear. Some results suggest that p21 does not influence repair and include observations that p21 does not inhibit repair of UV damage in human colon carcinoma cells (44–47) and human bladder cells (47). In fact, in both normal (32) and tumor (44) cells, deletion of the p21 gene reduced nucleotide excision repair (NER) capacity. One report shows modest inhibition of NER after introduction by electroporation of the PCNA-binding domain of p21 (48). Some results in vitro also show that NER is insensitive to p21 (49, 50), whereas another result indicates sensitivity (51). Another indication of the complexity of the role of p21 is the recent observation (52) that p21 is degraded within several hours of irradiation via the ubiquitin/proteosome pathway. This observation suggests that cell-cycle arrest is maintained by a more involved process than the p21 pathway. It is unclear whether bound PCNA is also degraded with the p21; however, the disappearance of p21 would then allow for the restoration of PCNA.

PCNA is necessary for the structural integrity and function of replication forks. Uncoupling of the polymerase from the sliding clamp is sufficient to stop replication (53–55). The 9-1-1 ring did not activate pol β for processive DNA synthesis. This finding confirms other recent results (56) and supports the idea that inactivation of PCNA is sufficient to stop replication-fork progression. Therefore, 9-1-1 would not be able to reverse inactivation of DNA replication caused by PCNA depletion during the damage response. FEN1 is involved in both DNA replication and repair. The 9-1-1 protein can stimulate FEN1, but without processive pol β, the activated FEN1 will be confined to repair functions.

Our results indicate that 9-1-1 serves a dual role as a damage sensor and as a component of repair complexes. Stimulation of FEN1 by 9-1-1 raises the possibility that 9-1-1 could be a universal platform for DNA repair, just as PCNA serves such a role for replication. In this hypothesis, 9-1-1 would substitute for PCNA in all of its repair roles. For example PCNA is a component of the NER pathway (17). PCNA also stimulates the joining activity of DNA ligase I (57), thought to participate in long-patch base-excision repair. This concept suggests the likely value of determining the capacity for stimulation of a wider variety of repair proteins by 9-1-1.