Nbs1 potentiates ATP-driven DNA unwinding and endonuclease cleavage by the Mre11/Rad50 complex

Abstract

The Nijmegen breakage syndrome gene product (Nbs1) was shown recently to associate in vivo with the Mre11 and Rad50 proteins, which play pivotal roles in eukaryotic DNA double-strand break repair, meiotic recombination, and telomere maintenance. We show in this work that the triple complex of recombinant Nbs1, Mre11, and Rad50 proteins binds cooperatively to DNA and forms a distinct protein-DNA species. The Mre11/Rad50/Nbs1 complex displays several enzymatic activities that are not seen without Nbs1, including partial unwinding of a DNA duplex and efficient cleavage of fully paired hairpins. Unwinding and hairpin cleavage are both increased by the presence of ATP. On nonhairpin DNA ends, ATP controls a switch in endonuclease specificity that allows Mre11/Rad50/Nbs1 to cleave a 3'-protruding strand at a double-/single-strand transition. Mutational analysis demonstrates that Rad50 is responsible for ATP binding by the complex, but the ATP-dependent activities are expressed only with Nbs1 present.

Figure 1

Nbs1 alters the activities of Mre11 and Rad50. (A) Preparations of Mre11 (lane 1), M/R (lane 2), M/N (lane 3), and M/R/N (lane 4) were separated on an SDS–polyacrylamide gel and stained with Coomassie blue. Protein amounts varied between 0.5 and 2 μg, keeping the amount of Mre11 approximately constant at 0.5 μg. (m) The molecular mass markers are 116, 97, 66, and 55 kD. (B) Gel mobility-shift assays were performed with Mre11 (lanes 2–4), M/R (lanes 5–7), M/N (lanes 8–10), and M/R/N (lanes 11–13) as indicated. Approximately 37.5 ng of each protein was added in lanes 2, 5, 8, and 11; 75 ng each in lanes 3, 6, 9, and 12; and 150 ng each in lanes 4, 7, 10, and 13. Proteins were mixed with a ³²P-labeled double-stranded DNA substrate containing 3′ overhangs at each end, and incubated for 10 min at room temperature before electrophoresis in a 0.7% 1/2× TBE agarose gel.

Figure 2

M/R/N opens fully paired hairpins. (A) Nuclease assays were performed with 150 ng each of Mre11 (lane 2), M/R (lane 3), M/N (lane 4), and M/N/R (lane 5) in 1 mm MnCl₂ on a ³²P-labeled substrate containing a fully paired hairpin on one end, a 4-bp 3′ overhang on the other end, separated by a 50-bp duplex. Reactions were incubated for 30 min at 37°C before separation on a denaturing polyacrylamide gel. (*) The location of the ³²P label in the diagram of the substrate. The cleavage sites are numbered such that a cut exactly at the tip is “0”, sites 3′ of this are positive, and sites 5′ of this are negative. The predominant cut made by M/R/N is at +1, as indicated by the arrow in the diagram. (B) Nuclease assays were performed on four different fully paired hairpin substrates that varied in sequence within 10 bp of the hairpin tip. The sequence of each substrate near the tip is shown above the gel. The structure of the substrates is the same as in A, and the hairpin used in A is identical to hairpin I in B. The sequence at the tip of hairpin I is from the coding end of the mouse Jκ1 locus; hairpin II corresponds to the coding end of the JH1 locus; and hairpin III corresponds to the coding end of the JH2 locus. Reactions contained 150 ng M/R/N (as indicated) and were analyzed as in A. The asterisk indicates the location of the ³²P label in the diagram of the substrate, and the arrows indicate the predominant cleavage sites on each substrate.

Figure 2

M/R/N opens fully paired hairpins. (A) Nuclease assays were performed with 150 ng each of Mre11 (lane 2), M/R (lane 3), M/N (lane 4), and M/N/R (lane 5) in 1 mm MnCl₂ on a ³²P-labeled substrate containing a fully paired hairpin on one end, a 4-bp 3′ overhang on the other end, separated by a 50-bp duplex. Reactions were incubated for 30 min at 37°C before separation on a denaturing polyacrylamide gel. (*) The location of the ³²P label in the diagram of the substrate. The cleavage sites are numbered such that a cut exactly at the tip is “0”, sites 3′ of this are positive, and sites 5′ of this are negative. The predominant cut made by M/R/N is at +1, as indicated by the arrow in the diagram. (B) Nuclease assays were performed on four different fully paired hairpin substrates that varied in sequence within 10 bp of the hairpin tip. The sequence of each substrate near the tip is shown above the gel. The structure of the substrates is the same as in A, and the hairpin used in A is identical to hairpin I in B. The sequence at the tip of hairpin I is from the coding end of the mouse Jκ1 locus; hairpin II corresponds to the coding end of the JH1 locus; and hairpin III corresponds to the coding end of the JH2 locus. Reactions contained 150 ng M/R/N (as indicated) and were analyzed as in A. The asterisk indicates the location of the ³²P label in the diagram of the substrate, and the arrows indicate the predominant cleavage sites on each substrate.

Figure 3

M/R/N generates different hairpin cleavage products in the presence of ATP. (A) Nuclease assays were performed with 150 ng of M/R/N in 1 mm MnCl₂ with 0.5 mm ATP as indicated, on a 50-bp ³²P-labeled dumbbell substrate containing fully paired hairpins on each end. Reactions were incubated for 30 min at 37°C before separation on a denaturing polyacrylamide gel. The small amount of opened dumbbell in the control lane is because of spontaneous breakage during gel purification of the substrate. (B) M/R/N (150 ng) was incubated as in Fig. 2A with a single-hairpin substrate (hairpin II from Fig. 2B) in the presence of 0.5 mm ATP as indicated. Aliquots were taken at different time points as shown.

Figure 3

M/R/N generates different hairpin cleavage products in the presence of ATP. (A) Nuclease assays were performed with 150 ng of M/R/N in 1 mm MnCl₂ with 0.5 mm ATP as indicated, on a 50-bp ³²P-labeled dumbbell substrate containing fully paired hairpins on each end. Reactions were incubated for 30 min at 37°C before separation on a denaturing polyacrylamide gel. The small amount of opened dumbbell in the control lane is because of spontaneous breakage during gel purification of the substrate. (B) M/R/N (150 ng) was incubated as in Fig. 2A with a single-hairpin substrate (hairpin II from Fig. 2B) in the presence of 0.5 mm ATP as indicated. Aliquots were taken at different time points as shown.

Figure 4

M/R/N partially unwinds short DNA duplexes from a 3′ single-stranded overhang. (A) M/R/N (150 ng) was incubated in 5 mm MgCl₂ and 0.5 mm ATP as indicated with a DNA substrate containing a 17-bp duplex and either a 3′ or 5′ single-stranded overhang, with the 17-mer labeled at the 5′ end with ³²P. The reactions were incubated at 37°C for 30 min and were stopped with SDS and EDTA before separation in a native polyacrylamide gel. (Δ) Substrate that was heated to denature the duplex. The arrow indicates the position of the free labeled strand. (B) Mre11 (25 ng, lanes 2,3), M/R (60 ng, lanes 4,5), M/N (50 ng, lanes 6,7), and M/R/N (150 ng, lanes 8,9) were incubated with the same 3′ overhang substrate as in A in the presence of 0.5 mm ATP as indicated. The varying protein levels yield approximately the same amount of Mre11 (25 ng) in each reaction. (C) M/R/N (150 ng) was incubated as in A and B except with DNA substrates containing a 34-bp duplex region and either a 3′ or 5′ single-stranded overhang as shown.

Figure 5

Nonhydrolyzable ATP analogs do not substitute for ATP in strand unwinding. (A) Nuclease assays were performed with 150 ng of M/R/N on a ³²P-labeled single-hairpin substrate identical to hairpin II in Fig. 2B. Reactions contained 1 mm MnCl₂ and 0.5 mm nucleotides as indicated, and were incubated for 30 min at 37°C before separation on a denaturing polyacrylamide gel. (B) Unwinding assays were performed with a DNA substrate containing a 17-bp duplex and a 3′ overhang as described in Fig. 4A. The ratio of free labeled oligonucleotide to the total amount of labeled substrate in each reaction was determined by quantitation on a PhosphorImager, and is shown here with the standard deviation calculated from two different experiments.

Figure 5

Nonhydrolyzable ATP analogs do not substitute for ATP in strand unwinding. (A) Nuclease assays were performed with 150 ng of M/R/N on a ³²P-labeled single-hairpin substrate identical to hairpin II in Fig. 2B. Reactions contained 1 mm MnCl₂ and 0.5 mm nucleotides as indicated, and were incubated for 30 min at 37°C before separation on a denaturing polyacrylamide gel. (B) Unwinding assays were performed with a DNA substrate containing a 17-bp duplex and a 3′ overhang as described in Fig. 4A. The ratio of free labeled oligonucleotide to the total amount of labeled substrate in each reaction was determined by quantitation on a PhosphorImager, and is shown here with the standard deviation calculated from two different experiments.

Figure 6

ATP stimulates M/R/N cleavage of a 3′ overhang at the border of the duplex region. (A) Nuclease assays were performed with Mre11 (100 ng, lanes 2–5), M/R (150 ng, lanes 6–9) and M/R/N (150 ng, lanes 10–13) on a double-stranded DNA substrate with 3′ overhangs on each end. The 3′ end of the top strand was labeled with ³²P-labeled cordycepin, as diagrammed, which lengthens the 3′ overhang by one nucleotide. Reactions contained 1 mm MnCl₂ and 0.5 mm ATP as indicated, and aliquots were taken at 10- and 30-min time points. The locations of the cuts made by the proteins are shown as arrows in the diagram. (m) Size markers of 11 and 15 nucleotides are shown. (B) Nuclease assays were performed as in A with the same substrate, except that the top strand was labeled with ³²P at the 5′ end as shown in the diagram.

Figure 6

ATP stimulates M/R/N cleavage of a 3′ overhang at the border of the duplex region. (A) Nuclease assays were performed with Mre11 (100 ng, lanes 2–5), M/R (150 ng, lanes 6–9) and M/R/N (150 ng, lanes 10–13) on a double-stranded DNA substrate with 3′ overhangs on each end. The 3′ end of the top strand was labeled with ³²P-labeled cordycepin, as diagrammed, which lengthens the 3′ overhang by one nucleotide. Reactions contained 1 mm MnCl₂ and 0.5 mm ATP as indicated, and aliquots were taken at 10- and 30-min time points. The locations of the cuts made by the proteins are shown as arrows in the diagram. (m) Size markers of 11 and 15 nucleotides are shown. (B) Nuclease assays were performed as in A with the same substrate, except that the top strand was labeled with ³²P at the 5′ end as shown in the diagram.

Figure 7

M/R/N complexes containing Rad50 protein mutated in the Walker A or Walker B motifs do not respond to ATP. Nuclease assays were performed on a substrate containing a fully paired hairpin on one end and a 3′ overhang on the other end, identical to hairpin II in Fig. 2B. Reactions contained 150 ng each of wild-type M/R/N, M/R(D1231A)/N, or M/R(K42E)/N protein, 1 mm MnCl₂, and 0.5 mm ATP as indicated.

Figure 8

Ku and RPA inhibit nuclease digestion of 3′ overhangs by M/R/N. Nuclease assays were performed as in Fig. 6B with the addition of varying amounts of RPA (30 or 300 ng in lanes 3 and 4, respectively) or Ku protein (5 and 50 ng in lanes 6 and 7, respectively). (m) Size markers of 11, 15, 21, and 28 nucleotides are shown.