J Biol Chem, Vol. 275, Issue 11, 7451-7454, March 17, 2000

ACCELERATED PUBLICATION
Human DNA Damage Checkpoint Protein hRAD9 Is a 3' to 5' Exonuclease^*

Tadayoshi Bessho and Aziz Sancar

From the Department of Biochemistry and Biophysics, University of North Carolina School of Medicine, Chapel Hill, North Carolina 27599

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS Conclusion REFERENCES

Human RAD9 protein (hRAD9) is a homolog of the fission yeast Rad9 protein, one of the six so-called checkpoint Rad proteins involved in the early steps of DNA damage checkpoint response in Schizosaccharomyces pombe. It has been shown previously that, in vivo, a highly modified form of hRAD9 makes a ternary complex with two other checkpoint Rad proteins, hRAD1 and hHUS1 (Volkmer, E., and Karnitz, L. M. (1999) J. Biol. Chem. 274, 567-570; St. Onge, R. P., Udell, C. M., Casselman, R., and Davey, S. (1999) Mol. Biol. Cell. 10, 1985-1995). However, the function of this complex is not known at present. To help define the functions of checkpoint Rad proteins in humans, we expressed hRAD9 in Escherichia coli, purified the recombinant protein and characterized it. We found that hRAD9 is a 3' to 5' exonuclease and located the nuclease active site to the region between residues 51 and 91 of the 391-amino acid-long protein. Our results suggest that exonucleolytic processing of primary DNA lesion by hRAD9 may contribute to DNA damage checkpoint response in humans.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS Conclusion REFERENCES

DNA damage checkpoints are the biochemical pathways that stop cell cycle progression, when the DNA contains damage (1-3). The checkpoints help maintain genomic integrity by preventing replication of damaged DNA and inhibiting cell division before the generation of two genome complements. Conceptually, DNA damage checkpoints have three components, damage sensors, signal transducers, and effecter molecules. Genetic studies in Schizosaccharomyces pombe have identified six checkpoint Rad proteins (Rad1, Rad3, Rad9, Rad17, Rad26, and Hus1), which are required to regulate both the DNA replication and DNA damage checkpoints (3, 4). There are very little biochemical data available on these proteins at present. The human ATM (ataxia-telangiectasia mutated) and ATR (ataxia-telangiectasia and rad3⁺-related) proteins, which are phosphatidylinositol 3-kinase-related protein kinases, are considered to be the functional homologs of the S. pombe Rad3 protein (5) and to transmit DNA damage signal by phosphorylating downstream targets. Biochemical data on the other checkpoint Rad proteins are scarce. However, genetic data indicate that they function in the early steps of the checkpoint response (3, 4).

Recently, the human homologs of sprad1 (6-8), sprad9 (9), sprad17 (7), and sphus1 (10) have been cloned, which has greatly improved the prospect of biochemical analysis of DNA damage checkpoint response in humans. Preliminary in vivo biochemical studies have already revealed some important information; hRAD9¹ becomes hyperphosphorylated upon DNA damage and makes a ternary complex with hRAD1 and hHUS1 proteins (11, 12). A similar complex was detected in Saccharomyces cerevisiae as well (13), providing additional information about the early events in DNA damage checkpoint response reactions. In addition, hRAD1, which exhibits sequence homology to Ustilago maydis Rec1 exonuclease (14), was reported to have a 3' to 5' nuclease activity (7), although another study failed to detect this activity (6). In the present study, we have investigated the enzymatic properties of hRAD9. Surprisingly, we find that hRAD9, which does not have sequence motifs indicative of its function, possesses a 3' to 5' exonuclease activity. We suggest that this activity may be important in the damage sensing and signal transduction steps of the DNA damage checkpoint response reaction.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS Conclusion REFERENCES

Recombinant hRAD9-- The plasmid pcDNA3-AU1-hRAD9 (11, 12) was a kind gift from Dr. Larry M. Karnitz (Mayo Foundation, Rochester, MN). The hRAD9 gene was amplified from this plasmid by PCR using the primers 5'-CCGAGCTCGAGAATGAAGTGCCTGGTC and 5'-GATGCAAGCTTTCAGACTTCACCCTCA. The PCR product was digested with XhoI and HindIII and inserted into pRSET(b) from Invitrogen to obtain the His-tagged hRAD9 expression vector pRSET-hRAD9. Then, the BamHI-HindIII fragment carrying hRAD9 was isolated from pRSET-hRAD9 and inserted into pMalc-2 (New England Biolabs) to obtain the pMal-hRAD9 plasmid, which expresses the maltose-binding protein (MBP)-hRAD9 fusion protein. The various deletion derivatives of hRAD9 were generated from pMal-hRAD9 by PCR.

To purify MBP-hRAD9, cell extract from DR153/pMal-hRAD9 was loaded onto an amylose affinity column, and after extensive washing, as described by the manufacturer (New England Biolabs), MBP-hRAD9 was eluted with 10 mM maltose. The protein was further purified by loading onto a DEAE-Sepharose column equilibrated with buffer R (20 mM Hepes-KOH, pH 7.8, 0.1 mM dithiothreitol, 0.1 mM EDTA, and 10% glycerol) and eluting with a salt gradient of 50 mM to 400 mM KCl in buffer R. MBP-hRAD9 came off the column at around 160 mM KCl. The peak fractions were combined, dialyzed against buffer R containing 50 mM KCl, and then subjected to a second round of DEAE chromatography as described above. The peak fractions were dialyzed against buffer R and stored at 80 °C. The deletion derivatives of hRAD9 in the form of MBP fusion were purified by an amylose affinity column followed by chromatography on DEAE and phenyl-Sepharose. For phenyl-Sepharose chromatography, the MBP-hRAD9 fractions from the DEAE column were dialyzed against buffer C (10 mM Tris-HCl, pH 7.5, 0.1 mM EDTA, and 10% glycerol) containing 1 M ammonium sulfate and then loaded onto the phenyl-Sepharose column. After washing with buffer C containing 0.5 M ammonium sulfate, the proteins were eluted by a gradient of 0.5 M ammonium sulfate in buffer C to 30% ethylene glycol in buffer C. MBP-hRAD9 deletion derivatives came off the column at about 30% ethylene glycol. After dialysis against buffer R, the proteins were stored at 80 °C.

Factor Xa Treatment-- To establish that the exonuclease activity detected in MBP-hRAD9 was intrinsic to hRAD9, the purified MBP-hRAD9 was subjected to a variety of treatments. First, to obtain hRAD9 free of tag, 5 µg of MBP-hRAD9 was digested with 5 µg of Factor Xa in buffer R at 23 °C for 2 h. Then, 50 µl of amylose resin was added to the reaction mixture, which was incubated for another 2 h at 4 °C, the resin and bound proteins were removed by centrifugation, and the supernatant was used as the "hRAD9 fraction." Second, to remove MBP-hRAD9 fusion protein from the MBP-hRAD9 preparations, 5 µg of the purified fusion protein was mixed with 50 µl of amylose resin for 2 h at 4 °C, the resin and bound material were removed by centrifugation, and the supernatant was used as the "MBP-hRAD9 depleted fraction." Finally, the amylose resin-MBP-hRAD9 isolated in the previous procedure was resuspended in 25 µl of R buffer and digested with 5 µg of Factor Xa. The resin was removed by centrifugation, and the supernatant was used as the "hRAD9 recovered fraction." In all of the Factor Xa treatments only about 50% of MBP-hRAD9 was cleaved as determined by SDS-PAGE and silver staining.

DNA Substrates-- The substrate used for most of our experiments was a 30-mer oligodeoxynucleotide, 5'-TCAGGGTGGCCAGCTGGCGCAGATCTGGCT-3'. The oligomer was labeled at the 5' end using [-³²P]ATP and polynucleotide kinase and at the 3' terminus using [-³²P]ddATP and deoxynucleotidyl terminal transferase. To prepare a 3'-end-labeled substrate with a 3'-OH terminus, the oligomer was annealed to a complementary oligomer with a 34-nt extension in which the first non-complementary base was a G residue. Then, the 3' end was labeled with [-³²P]dCTP using the Klenow fragment for extension. The extended oligomer (31-mer) was purified on a 10% denaturing polyacrylamide gel. For double-stranded oligomer substrate, the 31-mer was annealed with a complementary oligomer. In addition to these substrates, M13 single- and double-stranded forms were used to test for endonuclease activity.

Exonuclease Assay-- The standard reaction mixture contained 20 mM Hepes-KOH, pH 7.8, 10 mM MnCl₂, 0.1 mM EDTA, and 10% glycerol (v/v), 0.3 pmol substrate, and 12 pmol (1 µg) of MBP-hRAD9. The reactions were performed at 37 °C and terminated by adding EDTA to 20 mM. Reaction products were separated on 10% denaturing polyacrylamide gels. Quantitative analyses of the data were conducted with a Storm 850 scanner and a PhosphorImager (Molecular Dynamics).

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS Conclusion REFERENCES

hRAD9 Is a 3' to 5' Exonuclease-- The sequence of hRAD9 does not contain any sequence motif indicative of an enzymatic activity. However, genetic analyses in S. pombe implicate its homolog in the "damage-sensing" step of the DNA damage checkpoint response (4, 11). Therefore, we wished to investigate the interaction of this protein with DNA. In preliminary experiments, we noticed that DNA incubated with hRAD9 preparations was degraded and decided to characterize this activity in more detail.

hRAD9 was expressed in Escherichia coli either in the form of (His)₆-hRAD9 or MBP-hRAD9. Both forms were overproduced at reasonable levels and were soluble. The nuclease activity was detected with both the His- and MBP-tagged forms. However, even after a four-column purification procedure including a nickel affinity column, the (His)₆-hRAD9 form contained minor contaminants (data not shown). Hence, we conducted most of our experiments with the MBP-hRAD9 recombinant protein. The protein was purified through an amylose affinity column and two ion exchange columns, and the peak fractions from the final step contained virtually pure MBP-hRAD9 as analyzed by SDS-PAGE and silver staining (Fig. 1).

View larger version (58K): [in this window] [in a new window]   Fig. 1.   Purification of recombinant hRAD9. The MBP-hRAD9 protein was purified through amylose affinity column followed by chromatography on two DEAE columns, respectively. One µg from the peak fraction of the second DEAE column along with molecular weight standards was separated on a 12% SDS-PAGE, which was visualized by silver staining.

When the purified protein was incubated with a 30-nt-long single-stranded DNA, the DNA was degraded exonucleolytically. With 5'-end-labeled substrate, the products were oligomers of gradually decreasing sizes (Fig. 2, lane 2), whereas with 3'-end-labeled DNA, the label was released in the form of a mononucleotide (Fig. 2, lane 4) indicating that hRAD9 is a 3' to 5' exonuclease. Because E. coli contains several 3' to 5' exonucleases, including exonuclease I, exonuclease VII, and RecJ, capable of degrading single-stranded DNA (15), we considered the possibility that the exonuclease activity in our MBP-hRAD9 preparation might be due to a contaminant. To demonstrate that the activity was intrinsic to hRAD9, we analyzed the individual fractions from the last purification column for protein composition and nuclease activity. As seen in Fig. 3 the MBP-hRAD9 protein (panel A) co-elutes with the 3' to 5' exonuclease activity tested with either 5'-end-labeled (panel B) or 3'-end-labeled (panel C) substrates. These data do not unambiguously prove, but strongly indicate, that hRAD9 is a 3' to 5' exonuclease. This conclusion was further strengthened by the following series of experiments.

View larger version (38K): [in this window] [in a new window]   Fig. 2.   hRAD9 is a 3' to 5' exonuclease. MBP-hRAD9 (12 pmol) was incubated with 0.3 pmol of single-stranded 30 nt-long oligomer, and the reaction products were analyzed on 10% denaturing polyacrylamide gels. A, exonuclease activity on 5'- or 3'-end labeled substrates. The sizes of the substrates and the reaction products are indicated by arrows and a bracket where appropriate. B, effect of 3' terminal structure on exonuclease activity. Substrates with a 3'-terminal nucleotide of dCMP (lanes 5-9) or ddAMP (lanes 11-15) were incubated with hRAD9 for the indicated times (0-90 min), and the products were separated on a 10% denaturing polyacrylamide gel. T4 is a control lane containing DNA digested with 10 units of T4 DNA polymerase 3' to 5' exonuclease for 5 min at 37 °C with 3'-OH substrate (lane 10) or 3'-dideoxy substrate (lane 16). Top, autoradiogram of the gel; bottom, quantitative analysis of data in the top panel.

View larger version (34K): [in this window] [in a new window]   Fig. 3.   3' to 5' exonuclease activity co-purifies with MBP-hRAD9. Column fractions from the second DEAE column were analyzed for protein composition and exonuclease activity. A, fractions analyzed by 10% SDS-PAGE and silver staining. B and C, exonuclease activity with 5' (B)- and 3' (C)-end-labeled substrates. The 3'-end-labeled substrate had dideoxy-AMP as the terminal nucleotide. T4 is a control lane containing DNA digested with 10 units of T4 DNA polymerase 3' to 5' exonuclease for 5 min at 37 °C.

First, the MBP-hRAD9-associated nuclease activity could be removed by amylose resin, showing that the responsible protein bound to the resin (Fig. 4, lane 3). The nuclease activity was released from this resin by treating the bound proteins with Factor Xa, which cleaves at the junction of MBP and hRAD9 (Fig. 4, lane 5). Second, if the purified protein was treated with Factor Xa and then amylose-bound material was removed, the released protein showed the exonuclease activity (Fig. 4, lane 4). These data, taken together, show that the exonuclease binds to the amylose resin and can be released from the resin by Factor Xa. Since the possibility of an E. coli exonuclease binding to amylose and being released by Factor Xa is remote, we consider the data in Fig. 4 as very strong evidence for hRAD9 being a 3' to 5' exonuclease.

View larger version (32K): [in this window] [in a new window]   Fig. 4.   The 3' to 5' exonuclease in MBP-hRAD9 is intrinsic to hRAD9. A, lane 1, DNA was incubated with the amylose resin; lane 2, DNA was incubated with MBP-hRAD9. Lane 3, MBP-hRAD9 depleted fraction; MBP-hRAD9 was mixed with amylose, then the amylose-bound proteins were removed by centrifugation, and the supernatant was incubated with the oligomer. Lane 4, hRAD9 fraction; MBP-hRAD9 was first incubated with Factor Xa, then amylose resin was added, and following centrifugation the supernatant was tested for nuclease activity. Lane 5, hRAD9 recovered fraction; MBP-hRAD9 was first bound to the amylose resin, then Factor Xa was added (asterisk), and the released hRAD9 in the supernatant was tested for nuclease activity. Lane 6, control reaction with Factor Xa and amylose resin; analysis by SDS-PAGE showed that under our conditions only 50% of MBP-hRAD9 was cleaved by Factor Xa (data not shown). This is an autoradiogram of a 10% denaturing polyacrylamide gel. B, quantitative analysis of the data in panel A and two other experiments conducted under identical conditions. The background in lane 1 was subtracted from the values in all other lanes before averaging the numbers. Bars 2-6 correspond to the average values obtained from three independent experiments of lanes 2-6 in panel A.

Localization of Exonuclease Domain of hRAD9-- The sequence of hRAD9 does not show an obvious homology to any known nuclease, hence the active site could not be assigned to a specific region by sequence inspection. To locate the active site, we generated N- and C-terminal deletions of hRAD9 and tested them for nuclease activity. The results are shown in Fig. 5. The hRAD9(1-91) fragment and other C-terminal deletion derivatives containing this region are nearly as active as the full-length protein. Analysis of N-terminal deletion derivatives confirmed this conclusion and helped narrow down further the nuclease active site; hRAD9(92-391) lacked nuclease activity as expected, whereas hRAD9(52-391) was active. Thus, the results from hRAD9(1-91) combined with that obtained with hRAD9(52-391) places the nuclease active site within a stretch of 40 amino acids spanning amino acids 52 through 91. 

View larger version (35K): [in this window] [in a new window]   Fig. 5.   Localization of exonuclease domain of hRAD9. MBP-hRAD9 derivatives (300 ng each) containing the indicated amino acids were tested for exonuclease activity. A, analytical gel; B, schematic summary of the data in panel A. The percentages of released end-labeled nucleotide were: hRAD9, 9.3%; amino acids 1-291, 8.6%; 1-191, 6.8%; 1-91, 13.0%; 1-51, <1%; 92-391, <1%; and 52-391, 8.1%.

Properties of hRAD9 3' to 5' Exonuclease-- The reaction conditions affecting the hRAD9 exonuclease are summarized in Table I.The enzyme is 2.5-fold more active on single-stranded DNA than on double-stranded DNA and has no endonuclease activity as tested by nicking assay using plasmid DNA. The nuclease requires a divalent cation for activity, with the Mn²⁺ being 2.5-fold more efficient than Mg²⁺. The activity is strongly reduced by moderate (100 mM) salt concentrations and product inhibited by high concentrations of a mononucleotide. Interestingly, the enzyme works more efficiently on substrates with 3'-OH termini than on substrates with 3'-dideoxy termini (Fig. 2B and Table I).

	Conclusion

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS Conclusion REFERENCES

Immunoprecipitation analyses show that the hRAD1, hRAD9, and hHUS1 proteins make a radioresponsive checkpoint complex (11, 12). At present the precise role of this complex in damage sensing, damage processing, and signal transduction in the DNA damage checkpoint response is not known. Genetic analyses in both S. pombe and S. cerevisiae suggest that the complex functions in the early stages of the DNA damage checkpoint response (4, 13, 16). However, direct biochemical evidence for such a role has been lacking. Our finding that hRAD9 has an intrinsic 3' to 5' exonuclease function constitutes strong biochemical evidence that the complex interacts with DNA and hence most likely plays a role in the damage sensing step of the DNA damage checkpoint pathway. Further work with hRAD1 and hHUS1 and with the ternary translationally modified hRAD9 will help define more precisely the roles of the individual components as well as of the complex in the DNA damage checkpoint response in humans.

	ACKNOWLEDGEMENT

We are grateful to Dr. Larry M. Karnitz at the Mayo Foundation for providing pcDNA3-AU1-hRAD9 DNA.

	FOOTNOTES

^* This work is supported by Grant GM32833 from the National Institutes of Health.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Dept. of Biochemistry and Biophysics, Mary Ellen Jones Building, CB 7260, University of North Carolina School of Medicine, Chapel Hill, NC 27599. Tel.: 919-962-0115; Fax: 919-843-8627; E-mail: Aziz-Sancar@med.unc.edu.

	ABBREVIATIONS

The abbreviations used are: hRAD9, human RAD9 protein; PCR, polymerase chain reaction; MBP, maltose-binding protein; PAGE, polyacrylamide gel electrophoresis; nt, nucleotide(s).

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS Conclusion REFERENCES