Proc. Nati. Acad. Sci. USA Vol. 85, pp. 2339-2343, April 1988 Medical Sciences Immunological lesions in human uracil DNA glycosylase: Association with Bloom syndrome (monoclonal antibodies/DNA repair/human genetic disorders/marker enzyme/clinical diagnosis)

GITA SEAL, KILIAN BRECH, SETH J. KARP, BARBARA L. COOL, AND MICHAEL A. SIROVER Fels Research Institute and the Department of Pharmacology, Temple University School of Medicine, Philadelphia, PA 19140 Communicated by Sidney Weinhouse, November 23, 1987

ABSTRACT Three monoclonal antibodies that react with syndrome could be potentially characterized by a unique uracil DNA glycosylase of normal human placenta were tested reaction to one monoclonal antibody (23). Accordingly, we to determine whether one of the antibodies could be used as a examined whether altered reactivity of the glycosylase with negative marker for Bloom syndrome. As dermed by enzyme- this monoclonal antibody, 40.10.09, has the potential to be linked immunosorbent assay, monoclonal antibody 40.10.09, used as a negative marker for Bloom syndrome. We report which reacts with normal human glycosylase, neither recog- here the immunological properties of the uracil DNA gly- nized nor inhibited native uracil DNA glycosylase from any of cosylase from 24 diverse human sources. The uracil DNA five separate Bloom syndrome cell strains. Immunoblot anal- glycosylases isolated from 5 separate Bloom syndrome cell yses demonstrated that the denatured glycosylase protein from strains display an identical immunological aberration to all five Bloom syndrome cell strains was immunoreactive with antibody 40.10.09 that is unique to Bloom syndrome. This the 40.10.09 antibody. Further, each native enzyme was singular lack of immunoreactivity identifies this monoclonal immunoreactive with two other anti-human placental uracil antibody as a negative marker to provide potentially a means DNA glycosylase monoclonal antibodies. In contrast, ELISA by which to identify individuals with Bloom syndrome prior reactivity was observed with all three monoclonal antibodies in to the onset of clinical symptoms. reactions of glycosylases from 5 normal human cell types and 13 abnormal human cell strains. These results experimentally MATERIALS AND METHODS verify the specificity of the aberrant reactivity of the Bloom syndrome uracil DNA glycosylase. The possibility arises that Purification of Human Uracil DNA Glycosylase. Normal determination of the lack of immunoreactivity with antibody human skin fibroblasts (CRL 1222) were purchased from the 40.10.09 may have value in the early diagnosis of Bloom American Type Culture Collection. Bloom syndrome fibro- syndrome. blasts (GM 1492, GM 2548, GM 3402, GM 3498, and GM 3510) were purchased from the Human Genetic Cell Repos- itory (Camden, NJ). Each Bloom syndrome cell strain was Bloom syndrome is an autosomal recessive human genetic derived from a different patient in the Bloom Syndrome disease that clinically presents low body weight at birth, Registry (Registry nos. 44, 71, 9, 87, and 47, respectively). stunted growth, cutaneous rash, and immunological defi- Cell strains were grown in humidified 5% C02/95% air in ciency (1, 2). Individuals with Bloom syndrome are predis- Dulbecco's modified Eagle's medium (GIBCO) supple- posed to infection and are cancer prone (3, 4). Currently, mented with 2 mM glutamine, 10%o fetal calf serum, and 100 individuals with this disease cannot be readily identified units of penicillin and 100 mg of streptomycin per ml (13, 14). prior to the appearance of clinical symptoms. Bloom syn- Cells were harvested at confluence and sonicated at 60 W for drome cells are characterized by their high rates of chromo- 20 sec. Cell debris was pelleted by centrifugation. Uracil somal aberration (5, 6), spontaneous hypermutability (7-9), DNA glycosylase from each cell strain was then purified by hypersensitivity to environmental agents (10-12), and a gradient elution through DEAE-cellulose and then phospho- unique series of temporal alterations in the proliferative- cellulose column chromatography (22). dependent regulation of DNA repair (13-15). In particular, Monoclonal Antibodies. Monoclonal antibodies against the cells fail to enhance DNA repair pathways prior to the normal human uracil DNA glycosylase were prepared as initiation of DNA replication during the cell cycle. As a described (16). The antigen used was human placental uracil result, miscoding lesions remaining in the DNA are evi- DNA glycosylase chromatographed sequentially through denced by hypermutability and chromosomal aberrations DEAE-cellulose, phosphocellulose, and hydroxylapatite. that contribute to the debilitation in this human genetic The glycosylase monoclonal antibodies 37.04.12, 40.10.09, syndrome. and 42.08.07 were chosen for further study because of their To begin to examine the regulation of the involved genes ELISA reactivity with the normal placental enzyme. Each in Bloom syndrome at a molecular level, a series of mono- was further analyzed in detail by enzyme immunoprecipita- clonal antibodies were prepared against the uracil DNA tion with a second antibody and by glycerol gradient sedi- glycosylase of normal human placenta (16). The uracil DNA mentation with only the monoclonal antibody. With the glycosylase excises uracil from DNA during base-excision homogeneous human placental enzyme, ELISA showed that repair (17-19). Uracil can arise from the utilization of dUTP each monoclonal antibody recognized determinants on the instead of TTP during DNA synthesis (20) and by cytosine normal human placental glycosylase molecule (22). The deamination (21). Detailed analyses have shown that three of control monoclonal antibody, 1.05, was prepared from a these antibodies recognize different determinants on the spontaneous hybridoma that was isolated from a fusion with placental glycosylase molecule (22). An early finding raised spleen cells from unimmunized mice. Antibodies were puri- the possibility that the uracil DNA glycosylase in Bloom fied from culture medium by ammonium sulfate precipitation followed by DEAE-cellulose chromatography. Protein con- The publication costs of this article were defrayed in part by page charge centrations were determined by the Bradford method (24). payment. This article must therefore be hereby marked "advertisement" ELISA. Enzyme samples (10-50 p.l containing 10-50 ng of in accordance with 18 U.S.C. §1734 solely to indicate this fact. protein) were added to 96-well polyvinyl chloride microtiter

Downloaded by guest on October 2, 2021 2339 2340 Medical Sciences: Seal et al. Proc. Natl. Acad. ScL USA 85 (1988) plates. Each plate was incubated at 370C for 2 hr and then at 40C for 48-72 hr. The wells were washed twice with Dul- becco's phosphate-buffered saline (PBS) followed by a 30- min incubation with 1% (wt/vol) bovine serum albumin in PBS at 370C (200 ul per well). Each well was rinsed twice with PBS. For storage, 100 gl ofPBS was added to each well and removed just prior to analysis. Monoclonal antibodies (1 ng/pul) were added in aliquots of 50 jul and incubated for 2 hr 0 V1.0- B at 370C. Each well was rinsed with washing buffer (New 0 England Nuclear; 10 mM Tris*HC1, pH 8.0/0.05% Tween 40. 20). Sheep alkaline phosphatase-conjugated F(ab')2 fragment 0 too anti-mouse IgG antiserum (New England Nuclear) as second 0.5 antibody was then applied (50 1.l of a 1:250 dilution) for 2 hr at 370C; the wells were washed once with washing buffer 0 followed by two washes with deionized-distilled water. Immunoreactivity with the second antiserum was detected 1.0- C by incubation with 0.05 M p-nitrophenyl phosphate as sub- strate in the dark for 16 hr at 250C. The reaction was ended by addition of 50 1.l of 1 M NaOH. Colorimetric determina- 0.5- tions were performed at 405 nm with a Uniskan plate reader. Uracil DNA Glycosylase Assay. Uracil-containing polynu- cleotide substrate was prepared by using Escherichia coli 0* DNA polymerase I and [3H]dUTP as precursor (25). Uracil 0 25 50 DNA glycosylase was determined in a reaction mixture (final PROTEIN (ng) volume, 100 1,u) that contained 100 mM Tris*HCl (ph 8.0)/10 mM FIG. 1. Abnormal ELISA reactivities of Bloom syndrome uracil K2EDTA/5 mM dithiothreitol/1 ,ul of poly(dA).poly DNA glycosylases. Human uracil DNA glycosylases were purified ([3H]dU) (specific activity, 5,000-15,000 dpm/pmol), and sequentially through DEAE-cellulose and phosphocellulose as de- 1-2 ,ug of purified enzyme or cell extract. The mixture was scribed (22). ELISA was performed in triplicate with each anti- incubated for 30 min at 37°C. The reaction was terminated by human placental uracil DNA glycosylase monoclonal antibody over the addition of 300 ,ul of 95% ethanol (- 20°C), 60 ,u1 of 2 M the indicated range of protein concentrations. (A) ELISA with NaCl, and 100 ,u1 of 1 mg/ml heat-denatured calf thymus antibody 37.04.12. (B) ELISA with antibody 42.08.07. (C) ELISA DNA. The ethanol precipitate was collected by centrifuga- with marker antibody 40.10.09. o, Normal human fibroblast (CRL tion at 2300 x g for 10 min at 4°C. Uracil DNA glycosylase 1222); A, Bloom syndrome cell line GM 1492; e, Bloom syndrome activity was measured by the release of [3H]uracil into the GM 2548; A, Bloom syndrome GM 3402; r, Bloom syndrome GM ethanol supernatant. 3498; m, Bloom syndrome GM 3510. could not be detected (16). In contrast, this second peak of RESULTS activity could be observed by using an anti-DNA polymerase Immunoreactivity ofBloom Syndrome Uracil DNA Glycosyl- a monoclonal antibody that recognizes the glycosylase when ase. The ELISA reactivity of the glycosylase of each Bloom it is bound to the polymerase (16, 26). syndrome cell strain was examined with all of the monoclo- The ability of each monoclonal antibody to inhibit the nal antibodies raised against the normal enzyme. Each Bloom syndrome glycosylase (GM 2548) was then ascer- Bloom syndrome enzyme reacted with antibodies 37.04.12 tained. Significant enzyme inhibitions occurred when the and 42.08.07 (Fig. 1 A and B) in a concentration-dependent Bloom syndrome enzyme was preincubated with either manner, identical to that of the glycosylase purified from monoclonal antibodies 37.04.12 or 42.08.07 (Fig. 2 D and E). normal human fibroblasts (Fig. 1 A and B) and the homoge- The extent of inhibition was virtually identical to that neous human placental enzyme (22). In contrast, the gly- observed in their reactions with the normal human fibroblast cosylase from all five Bloom syndrome cell strains failed to enzyme. In contrast, marker monoclonal antibody 40.10.09 react with the marker monoclonal antibody, 40.10.09 (Fig. failed to inhibit the Bloom syndrome glycosylase (Fig. 2F). 1C). Further, this lack of recognition ofthe Bloom syndrome The level of enzyme activity and the position of the glycos- glycosylase was observed in ELISA, even when 10 ,ug ylase in the reaction mixture after the glycerol gradient (100-fold excess) of the Bloom syndrome glycosylase prep- centrifugation matched those of the simultaneously assayed arations was used. However, ELISA immunoreactivities to control antibody, 1.05. The identical lack of inhibition with this same marker antibody (40.10.09) were observed with the marker antibody 40.10.09 was observed in separate experi- normal human fibroblast enzyme (Fig. 1C) and with the ments with the uracil DNA glycosylase purified from each of placental enzyme (22). the other four Bloom syndrome cell strains (results not The reactivity of each Bloom syndrome glycosylase was shown). also studied in terms of enzyme inhibition by glycerol if the of the Bloom gradient sedimentation analysis (16). First, normal human To determine altered reactivity syn- fibroblast glycosylase was preincubated with each type of drome glycosylase was due to a loss of the primary amino anti-glycosylase antibody or with the control antibody, 1.05 acid sequence that was detected by marker antibody (Fig. 2, A-C). In each instance, uracil DNA glycosylase 40.10.09, each glycosylase was denatured and examined by activity sedimented near the top of the gradient. Preincuba- immunoblot analysis. Notice that all Bloom's syndrome tion ofthe normal enzyme with each glycosylase monoclonal enzymes were then recognized by marker antibody 40.10.09 antibody resulted in a loss of activity. Noteworthy is the fact (Fig. 3, lanes 1-5). Further, only one immunoreactive spe- that the most significant inhibitions occurred when the cies of Mr 37,000 was detected, corresponding to the molec- normal enzyme was preincubated with either antibody ular size of the normal human fibroblast glycosylase (Fig. 3, 37.04.12 (Fig. 2A) or the marker antibody, 40.10.09 (Fig. lane 7) and the homogeneous human placental enzyme (Fig. 2C). As each antibody inhibited catalysis, the glycosyl- 3, lane 6). These findings suggest that each Bloom syndrome ase-antibody complex that sediments at a higher density glycosylase shares a common aberration in its native struc- Downloaded by guest on October 2, 2021 Medical Sciences: Seal et al. Proc. Natl. Acad. Sci. USA 85 (1988) 2341

360- A D -300

80- - 150 - 92,000

I - 360-_ B E -300 aJ - 66,000 U) 80 w -j w: 180' - 150 -j - 45,000 92 . ~~~~~4ftb 49

Cc F - 36,000 300 -300 -29,000 - 24,000 150 -II - 150 ji 1 2 3 4 5 6 7 FIG. 3. Immunoblot analyses of human uracil DNA glycosylases.

0 20 0 20 Human uracil DNA glycosylases were purified as described. Equiv- T B T B alent amounts of protein of each sample (20 /ug) were used. FRACTION NaDodSO4 gel electrophoresis and immunoblot analyses were per- FIG. 2. Inhibition of glycosylase activity by anti-human placental formed as described (22). Molecular weight standards were electro- DNA phoresed in an adjacent lane. The position of each standard is uracil glycosylase monoclonal antibodies. Enzyme inhibition indicated in the right-hand margin. Lanes: 1-5, GM 2548, GM 1492, of each monoclonal antibody was examined by glycerol centrifugal of normal GM 3498, GM 3402, and GM 3510, respectively; 6, human placenta; gradient analysis (16). Eighty-two micrograms purified 7, CRL 1222. human glycosylase (CRL 1222) (A-C) or purified Bloom syndrome glycosylase (GM 2548) (D-F) was preincubated with either 82 ,ug of control antibody 1.05 (o) or 80 ug of anti-human glycosylase immunoreactivity with all three monoclonal antibodies. The monoclonal antibodies (o) at 4°C for 120 min. The enzyme-antibody xeroderma pigmentosum cells included different comple- complex was then sedimented through a 10-35% glycerol gradient. mentation groups (C and D) as well as the xeroderma Fractions were collected from the top of each gradient. Uracil DNA pigmentosum variant. Progeria is characterized by prema- glycosylase activity was then determined. Glycosylase preincuba- ture aging (30) and possibly by altered capacity of DNA tion was with monoclonal antibody 37.04.12 (A and D), monoclonal repair (31, 32). Normal immunoreactivity was observed antibody 42.08.07 (B and E), or marker monoclonal antibody when using the glycosylase from two different progeroid cell 40.10.09 (C and F). B, bottom of gradient; T, top of gradient. strains. ture that precludes recognition by the marker antibody, ELISA immunoreactivities of the glycosylases from three 40.10.09. other human genetic syndromes were determined. Bloom Specificity of Immunoreactivity of Bloom Syndrome Gly- syndrome and Tay-Sachs disease arise with frequencies of cosylase. To determine the uniqueness of this loss of immu- approximately 1/120 and 1/30 in the Ashkenazic Jewish noreactivity in Bloom syndrome, ELISA was performed on population, respectively (33, 34). The Tay-Sachs uracil the glycosylase in cell extracts derived from a wide variety DNA glycosylase displayed normal immunoreactivity (Table of human sources. Their immunoreactivities with all three 1). ELISA was also performed on two genetic syndromes monoclonal antibodies against human placental uracil DNA characterized by other metabolic disorders. Normal reactiv- glycosylase were compared with those of the glycosylase ity was observed with each of the three anti-human glycosy- from a normal human fibroblast cell strain over the protein lase monoclonal antibodies. concentration range of 10-50 ng. First, we evaluated the ELISA reactivities of the glycosyl- DISCUSSION ase from the following five different normal human cell types: two different human skin fibroblast cell strains, one embry- This report provides evidence that raises the possibility of a onic lung fibroblast cell strain, human lymphocytes, and method by which to identify Bloom syndrome prior to the several different placental preparations. The extent of recog- onset of clinical symptoms. The altered immunoreactivity of nition was at all times comparable to that observed for the the Bloom syndrome glycosylase with antibody 40.10.09 normal human skin fibroblast enzyme. Similarly, a simian demonstrates that this antibody may serve as a negative virus 40-transformed human cell line exhibited normal ELISA marker of this human genetic disease. In common with reactivity, even though viral transformation can dramatically Tay-Sachs disease, Bloom syndrome heterozygotes do not affect the capacity for human DNA repair (27, 28). appear to manifest clinical symptoms. However, in Tay- We also evaluated the ELISA immunoreactivities of the Sachs disease, heterozygotes may be identified by the glycosylase of fibroblasts from a variety of human genetic expression of the mutant allele. Bloom syndrome cell strains disorders. Like individuals with Bloom syndrome, those derived from such individuals are currently unavailable. As with ataxia telangiectasia and xeroderma pigmentosum are the uracil DNA glycosylase is easily analyzed in blood cancer prone (19). Further, xeroderma pigmentosum cells of samples (35), family studies of Bloom syndrome individuals complementation groups A, C, and D fail to enhance nucle- can be performed to examine the expression of the altered otide-excision repair during the cell cycle (29). However, glycosylase in heterozygotes. each complementation group regulates in a normal fashion Although antibody-antigen interactions depend on the both base-excision repair and the uracil DNA glycosylase. amino acid sequence that specifies the primary determinant, The uracil DNA glycosylase protein from two different immunoreactivity may also depend on the secondary or ataxia telangiectasia cell strains and three different xero- tertiary structure of the protein (36). Further, independent of derma pigmentosum cell strains displayed normal ELISA the primary antigenic determinant, immunoreactivity may be Downloaded by guest on October 2, 2021 2342 Medical Sciences: Seal et al. Proc. Natl. Acad. Sci. USA 85 (1988) affected by alterations of epitope structure, including protein observed in five Bloom syndrome cell strains, the findings folding (37). The results presented in this study demonstrate suggest that each strain contains the identical lesion in the that the native glycosylase from five separate Bloom syn- human uracil DNA glycosylase gene. drome cell strains was not recognized by the marker mono- This laboratory has previously reported that Bloom syn- clonal antibody, 40.10.09. In contrast, each denatured drome may be characterized by several alterations in the Bloom syndrome enzyme displayed normal immunoreactiv- proliferation-dependent regulation of DNA repair, including ity with that antibody in immunoblot analyses. Therefore the temporal displacement of nucleotide-excision repair after each protein must contain the primary amino acid sequence UV irradiation, base-excision repair after exposure to that comprises the 40.10.09 antigenic determinant. The de- methylmethane sulfonate, and abnormal regulation of the tected alteration in the Bloom syndrome enzyme must reside uracil DNA glycosylase and the hypoxanthine DNA gly- in a folding that changes the structural configuration of the cosylase enzymes (13, 14). In each instance, DNA repair molecule to preclude recognition of the primary determi- capacity was not enhanced prior to DNA replication but was nant. At present, it is unknown whether the epitope detected increased coordinately with DNA synthesis. Further, no by antibody 40.10.09 is structurally continuous or inter- induction of the 06-methylguanine methyltransferase was rupted-i.e., discontinuous-within the primary sequence of observed (15). These regulatory defects were postulated to the glycosylase. Further, the proximity of a putative struc- result in the aberrant replication of the cell's genetic infor- tural alteration to the 40.10.09 determinant in the Bloom mation before there is the opportunity to correct the DNA or syndrome enzyme is similarly unknown. The present find- to remove accumulated DNA lesions. This altered regulation ings provide a system by which to examine the effect of of the uracil DNA glycosylase was subsequently reported in secondary or tertiary conformation of a protein on antibody a cell strain from a Japanese Bloom syndrome patient (38). recognition. As the identical immunological alteration was In addition, recent analysis demonstrated a 50% decrease in Table 1. Immunoreactivities of human uracil DNA glycosylases Immunoreactivities of monoclonal antibodies relative to normal human skin fibroblasts as 1.00 Source 37.04.12 42.08.07 40.10.09 Normal human cells Skin fibroblasts (CRL 1222) 1.00 1.00 1.00 Human placenta 0.92 ± 0.09 0.92 ± 0.19 0.90 ± 0.10 Lung fibroblasts (CCL 75) 1.02 ± 0.10 0.89 ± 0.16 0.94 ± 0.12 Skin fibroblasts (GM 5879) 0.98 ± 0.04 1.28 ± 0.07 0.91 ± 0.11 Lymphocytes 1.32 ± 0.23 1.04 ± 0.08 1.19 ± 0.26 Transformed cells SV40-transformed lung fibroblasts (CCL 75.1) 0.92 ± 0.14 1.12 ± 0.12 1.14 ± 0.21 Bloom syndrome fibroblasts GM 1492 1.02 ± 0.12 1.15 ± 0.07 0.00 GM 2548 1.08 ± 0.09 1.11 ± 0.08 0.00 GM 3402 1.01 ± 0.03 1.13 ± 0.03 0.00 GM 3498 0.99 ± 0.07 0.94 ± 0.10 0.00 GM 3510 0.81 ± 0.05 1.23 ± 0.22 0.00 Ataxia telangiectasia fibroblasts GM 0367 1.17 ± 0.23 0.91 ± 0.24 0.94 ± 0.18 GM 2052 0.85 ± 0.22 0.96 ± 0.08 1.03 ± 0.19 Xeroderma pigmentosum (XP) fibroblasts CRL 158 (XP-C) 0.98 ± 0.07 1.08 ± 0.20 0.81 ± 0.09 CRL 1258 (XP-D) 1.11 ± 0.25 0.93 ± 0.05 0.93 ± 0.04 GM 3614 (XP variant) 0.82 ± 0.04 0.84 ± 0.04 0.99 ± 0.04 Tay-Sachs disease fibroblasts GM 2968 1.20 ± 0.25 1.00 ± 0.06 0.98 ± 0.24 GM 4863 0.80 ± 0.08 1.07 ± 0.08 0.99 ± 0.18 Familial hypercholesterolemia fibroblasts GM 1355 1.07 ± 0.15 1.01 ± 0.08 1.04 ± 0.30 GM 2408 0.65 ± 0.13 1.09 ± 0.06 0.75 ± 0.13 Galactosemia fibroblasts GM 1209 0.89 ± 0.13 0.89 ± 0.19 1.16 ± 0.08 GM 1908 1.09 ± 0.13 1.05 ± 0.07 1.10 ± 0.26 Progeroid fibroblasts AG 3911 0.99 ± 0.09 0.98 ± 0.09 1.08 ± 0.17 AG 6917 1.03 ± 0.12 1.14 ± 0.07 1.01 ± 0.06 The concentration-dependent recognition of the uracil DNA glycosylase in the ELISA by the anti-human uracil DNA glycosylase monoclonal antibodies was determined over a protein range of 10-50 ng. ELISA reactivity was determined by alkaline phosphatase activity as measured by absorbance at 405 nm. Ratios of reactivities were calculated by dividing the A4,5 of the cell strain at each concentration by the A405 of the normal human fibroblast cell strain (CRL 1222) at that concentration. Ratios were determined at a minimum of four separate protein concentrations (performed similarly to the concentration curves shown in Fig. 1). Absorbances at each protein concentration were performed in duplicate. Cells were cultured as described herein; cell extracts were prepared as previously described (13, 14). SV40, simian virus 40. Downloaded by guest on October 2, 2021 Medical Sciences: Seal et al. Proc. Natl. Acad. Sci. USA 85 (1988) 2343 DNA ligase I in Epstein-Barr virus-transformed Bloom 14. Dehayza, P. & Sirover, M. A. (1986) Cancer Res. 46, syndrome cell lines (39, 40). Each of these alterations in 3756-3761. capacity appears to be separate and indepen- 15. Kim, S., Vollberg, T. M., Ro, J.-Y., Kim, M. & Sirover, DNA repair M. A. (1986) Mutat. Res. 173, 141-145. dent. However, genetic studies demonstrate that Bloom 16. Arenaz, P. & Sirover, M. A. (1983) Proc. Natl. Acad. Sci. syndrome is an autosomal recessive human genetic disease, USA 80, 5822-5826. presumably localizing the molecular defect within a single 17. Lindahl, T. (1982) Annu. Rev. Biochem. 51, 61-88. gene. Thus, it would appear that each defect in DNA repair 18. Teebor, G. W. & Frenkel, K. (1983) Adv. Cancer Res. 38, capacity, including the results presented in this report, may 23-59. the Bloom 19. Friedberg, E. (1985) DNA Repair (Freeman, New York). represent alterations characteristic of syndrome 20. Bessman, M. J., Lehman, I. R., Adler, J., Zimmerman, S. B., phenotype but secondary to the primary genetic defect. At Simms, E. S. & Kornberg, A. (1958) Proc. Natl. Acad. Sci. present, that single defect remains unknown. Alternatively, USA 44, 633-640. these series of repair-enzyme regulatory alterations may 21. Shapiro, R. & Klein, R. S. (1966) Biochemistry 5, 2358-2362. indicate a corecessive pattern of inheritance for Bloom 22. Seal, G., Arenaz, P. & Sirover, M. A. (1987) Biochim. syndrome (41). In either instance, molecular analysis of Biophys. Acta 925, 226-233. DNA-repair structural genes may identify one or more 23. Vollberg, T. M., Seal, G. & Sirover, M. A. (1987) Carcinogen- genomic aberrations that can result in separate changes that esis 8, 1725-1729. 24. Scopes, R. (1982) Protein Purification, Principles and Practice include the alteration of the amino acid sequence of the (Springer, New York), p. 266. uracil DNA glycosylase protein. 25. Sirover, M. A. (1979) Cancer Res. 39, 2090-2095. 26. Seal, G. & Sirover, M. A. (1986) Proc. NatI. Acad. Sci. USA The generous counsels of Drs. Sidney Weinhouse, Sam Sorof, 83, 7608-7612. and Hope Punnett are gratefully appreciated. This study was funded 27. Heddle, J. A. & Arlett, C. F. (1980) Mutat. Res. 72, 118-125. by a grant to M.A.S. from the National Cancer Institute (CA-29414) 28. Teo, I. A., Lehmann, A. R., Muller, R. & Rajewsky, M. F. and by grants to the Fels Research Institute from the National (1983) Carcinogenesis 4, 1075-1077. Cancer Institute (CA-12227) and from the American Cancer Society 29. Gupta, P. K. & Sirover, M. A. (1984) J. Cell Biol. 99, (SIG-6). 1275-1281. 30. DeBusk, F. L. (1972) J. Pediatr. 80, 697-724. 1. Bloom, D. (1966) J. Pediatr. 68, 103-113. 31. Epstein, J., Williams, J. R. & Little, J. B. (1973) Proc. Natl. 2. German, J. (1969) Am. J. Hum. Genet. 21, 196-227. Acad. Sci. USA 70, 977-981. 3. German, J., Bloom, D. & Passarge, E. (1977) Clin. Genet. 12, 32. Regan, J. D. & Setlow, R. B. (1974) Biochem. Biophys. Res. 162-168. Commun. 59, 858-864. 4. German, J., Bloom, D. & Passarge, E. (1984) Clin. Genet. 25, 33. German, J., Bloom, D., Passarge, E., Fried, K., Goodman, 166-174. R. M., Katzenellenbogen, I., Laron, Z., Legum, C., Levin, S. 5. Chaganti, R. S. K., Schonberg, S. & German, J. (1974) Proc. & Wahrman, J. (1977) Am. J. Hum. Genet. 29, 553-562. Nati. Acad. Sci. USA 71, 4508-4512. 34. Kolodny, E. H. (1979) in Genetic Diseases Among Ashkenazi 6. Bryant, E. M., Hoehn, H. & Martin, G. M. (1979) Nature Jews, eds. Goodman, R. M. & Motulsky, A. G. (Raven, New (London) 279, 795-796. York), pp. 217-229. 7. Gupta, R. S. & Goldstein, S. (1980) Mutat. Res. 73, 331-338. 35. Koistinen, P. & Vilpo, J. A. (1986) Mutat. Res. 159, 99-102. 8. Warren, S. T., Schultz, R. A., Chang, C.-C., Wade, M. H. & 36. Geysen, H. M., Tainer, J. A., Rodda, S. J., Mason, T. J., Trosko, J. E. (1981) Proc. Nati. Acad. Sci. USA 78, 3133- Alexander, H., Getzoff, E. D. & Lerner, R. A. (1987) Science 3137. 235, 1184-1190. 9. Vijayalaxmi, Evans, H. J., Ray, J. H. & German, J. (1983) 37. Blond, S. & Goldberg, M. (1987) Proc. Natl. Acad. Sci. USA Science 221, 851-853. 84, 1147-1151. 10. Selsky, C. A., Henson, P., Weichselbaum, R. R. & Little, 38. Yamamoto, Y. & Fujiwara, Y. (1986) Carcinogenesis 7, J. B. (1979) Cancer Res. 39, 3392-33%. 305-310. 11. Krepinsky, A. B., Rainbow, A. J. & Heddle, J. A. (1980) 39. Willis, A. E. & Lindahl, T. (1987) Nature (London) 325, Mutat. Res. 69, 357-368. 355-357. 12. Hook, G. J., Kwok, E. & Heddle, J. A. (1984) Mutat. Res. 40. Chan, J. Y. M., Becker, F. F., German, J. & Ray, J. M. (1987) 131, 223-230. Nature (London) 325, 357-359. 13. Gupta, P. K. & Sirover, M. A. (1984) Proc. Natl. Acad. Sci. 41. Lambert, W. C. & Lambert, M. W. (1985) Mutat. Res. 145, USA 81, 757-761. 227-234. Downloaded by guest on October 2, 2021