Differences in Allelic Distribution of Two Polymorphisms in the VHL ...

0021-972X/99/$03.00/0 The Journal of Clinical Endocrinology & Metabolism Copyright © 1999 by The Endocrine Society

Vol. 84, No. 9 Printed in U.S.A.

Differences in Allelic Distribution of Two Polymorphisms in the VHL-Associated Gene CUL2 in Pheochromocytoma Patients without Somatic CUL2 Mutations* EVA-MARIA DUERR, OLIVER GIMM†, DONNA S. NEUBERG, JENNIFER B. KUM, STEVEN C. CLIFFORD‡, SERGIO P. A. TOLEDO, EAMONN R. MAHER, PATRICIA L. M. DAHIA§, AND CHARIS ENG Clinical Cancer Genetics and Human Cancer Genetics Programs, Ohio State University Comprehensive Cancer Center (E.-M.D., O.G., J.B.K., P.L.M.D., C.E.), Columbus, Ohio 43210; Dana-Farber Cancer Institute, Harvard Medical School (E.-M.D., D.S.N., J.B.K., P.L.M.D.), Boston, Massachusetts 02115; University of Bonn School of Medicine (E.-M.D.), 53105 Bonn, Germany; the Section of Medical and Molecular Genetics, Department of Pediatrics and Child Health, University of Birmingham School of Medicine (S.C.C., E.R.M.), Birmingham B15 2TG, United Kingdom; the Endocrine Genetics Unit, University of Sao Paulo School of Medicine (S.P.A.T.), Sao Paulo 54199, Brazil; and the Cancer Research Campaign Human Cancer Genetics Research Group, University of Cambridge (C.E.), Cambridge, United Kingdom CB2 2QQ ABSTRACT Although the two major familial forms of pheochromocytomas, multiple endocrine neoplasia type 2 and von-Hippel-Lindau disease (VHL), have been associated with mutations of the RET and VHL genes, respectively, the molecular pathogenesis of sporadic pheochromocytomas is largely unknown. Recently, a putative tumor suppressor gene has been identified, CUL2, whose product has been shown to interact with the VHL tumor suppressor. To examine whether CUL2 plays a role in pheochromocytoma pathogenesis, we analyzed a series of 26 distinct tumor samples for mutations in the whole coding region of this gene. There were no somatic pathogenic mutations in

P

HEOCHROMOCYTOMAS are rare neoplasms embryologically derived from the neural crest (1). Most pheochromocytomas occur as sporadic tumors, but in about 10% of cases a familial form is observed. These familial tumors are usually benign and most frequently are associated with three different cancer syndromes: multiple endocrine neoplasia type 2 (MEN 2A and 2B), von Hippel-Lindau disease (VHL), Received March 1, 1999. Revision received May 17, 1999. Accepted May 24, 1999. Address all correspondence and requests for reprints to: Charis Eng, M.D., Ph.D., F.A.C.P., Human Cancer Genetics Program, Ohio State University Comprehensive Cancer Center, 690C Medical Research Facility, 420 West 12th Avenue, Columbus, Ohio 43210. E-mail: [email protected]. Or to: Patricia L. M. Dahia, M.D., Ph.D., Dana-Farber Cancer Institute, 44 Binney Street, SM820, Boston, Massachusetts 021156084. E-mail: [email protected]. * This work was supported in part by Grant P30CA16058 from the NCI (Bethesda, MD; to Ohio State University Comprehensive Cancer Center), the British Heart Association (London, UK; to E.R.M. and C.E.), and the Cancer Research Campaign (London, UK; to E.R.M.). † postdoctoral fellow of the Deutsche Forschungsgemeinschaft, Germany. ‡ Medical Research Council Research Fellow, United Kingdom. § Postdoctoral fellow of the Susan G. Komen Breast Cancer Research Foundation.

CUL2, except for 1 sporadic tumor that had a hemizygous gene deletion. We also found 3 novel polymorphisms in the gene. One of these variants, IVS5– 6C/T, as well as another previously described one, c.2057G/A, were overrepresented among the pheochromocytoma patients compared to that in a control population (P , 0.005 and P , 0.01, respectively). Although our findings suggest that CUL2 does not play a major role in the pathogenesis of pheochromocytomas, it is still unknown whether epigenetic mechanisms are involved in its inactivation in VHL-associated tumors. Furthermore, the potential role for the overrepresented alleles in the pheochromocytoma group requires further investigation. (J Clin Endocrinol Metab 84: 3207–3211, 1999)

and, rarely, neurofibromatosis type 1 (NF1). Even less commonly, an isolated familial form of pheochromocytoma without other associated clinical features occurs (2, 3). The molecular basis of some of the familial forms of pheochromocytomas has been recently unveiled. MEN 2, characterized by medullary thyroid carcinoma, pheochromocytoma, and hyperparathyroidism (MEN 2A) or mucocutaneous neuromas (MEN 2B), is caused by germline mutations in the RET protooncogene, encoding a receptor tyrosine kinase (4 – 8). VHL, which is characterized by a variety of tumors, including retinal angiomas, central nervous system hemangioblastomas, pheochromocytomas, clear cell renal carcinomas (RCC), pancreatic cysts, and endolymphic sac tumors (reviewed in Ref. 9), is associated with germline mutations in the VHL tumor suppressor gene (10, 11). In contrast to certain familial pheochromocytoma syndromes, somatic RET and VHL mutations are only rarely found in sporadic pheochromocytomas (12–14). Other candidate genes have been analyzed for their potential role in the pathogenesis of pheochromocytomas, such as p53 (15), p16 (16), and the gene encoding for the RET ligand, GDNF (17), but no specific mutations could be detected, suggesting that either another gene(s) or other gene-inactivating mecha-

3207

3208

JCE & M • 1999 Vol 84 • No 9

DUERR ET AL.

nisms play a bigger role in the pathogenesis of sporadic pheochromocytomas. Putative candidate genes include those that encode proteins that interact with the VHL gene product. Several such proteins have now been identified (9, 18). VHL has been found to inhibit transcription elongation by associating with elongin B and C, thereby preventing their binding to the catalytic subunit elongin A (19, 20). Recently, CUL2, a member of the cullin family, has been shown to bind to the VHL/ elongin B/elongin C complex (21, 22). Although its function has not been precisely characterized yet, CUL2 is homologous to a yeast protein, Cdc53, which has been associated with a complex that targets cell cycle proteins for ubiquitinmediated degradation (22). To examine whether CUL2 plays a role in the development of pheochromocytomas, we analyzed a series of 26 pheochromocytomas for sequence variations in the whole coding region and flanking intronic sequences of the CUL2 gene by a combination of single strand conformation polymorphism (SSCP) analysis, restriction digestion, and direct sequencing. Materials and Methods Samples DNA was obtained from 26 pheochromocytomas and also from peripheral blood of the patients using standard protocols. Clinical data from the patients have been described previously (15) and are summarized in Table 1. In brief, 4 tumors originated from patients with familial forms of pheochromocytoma: 1 VHL, 1 NF, and 2 siblings with germline VHL mutation-negative familial pheochromocytoma-only syndrome; the remaining were sporadic tumors, 4 of which were malignant. Informed consent was obtained from all subjects and/or guardians involved in this study.

In addition, peripheral blood DNA from 26 gender- and age-matched controls with the same ethnic background was analyzed.

PCRs and SSCP analysis Intronic primers flanking each exon of the CUL2 gene were used for PCRs as previously described (23) with minor modifications, as shown in Table 2. Each PCR product was diluted 10-fold (1:10) into a 95% formamide loading buffer, heat denatured, loaded onto nondenaturing polyacrylamide gels, and run overnight at 4 C or room temperature depending on the amplicon (see Table 2 for SSCP details). Bands were visualized after silver staining as described previously (23). Normal genomic DNA (Promega Corp., Madison, WI) were used as negative controls, and polymorphic variants (see below) were positive controls for the assays. All tumor samples with SSCP variants were resubjected to SSCP alongside with their matched germline DNA counterpart.

Restriction digestion and sequencing analysis All samples with variants were reamplified from genomic DNA and directly sequenced as previously described (24). Variants in exon 2, 5, and 6 were further digested with specific restriction enzymes according to the manufacturer’s instructions (New England Biolabs, Inc., Beverly, MA) for confirmation of SSCP and sequencing results (see Table 3, for enzymes used). In addition, three random tumor samples were entirely sequenced to exclude sequence variations missed by the SSCP screening. Differential restriction digestion of heterozygous samples was also used to characterize hemizygous deletions of the CUL2 gene. Paired blood and tumor DNA were amplified as described above and digested by the appropriate restriction enzyme (Table 3), and products were run on 6% polyacrylamide gels. Band intensities from captured gel images were analyzed by densitometry using ImageQuant software (Molecular Dynamics, Inc., Sunnyvale, CA). Loss of heterozygosity was defined by a ratio of normal to tumor DNA above 1.5 or by visual inspection of the restricted bands representing the distinct alleles, as previously described (24).

TABLE 1. Summary of clinical features of the pheochromocytomas used in this study Patient no.

Age (yr)

Gender

Tumor location

Histology

Other

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

17 32 38 32 36 32 32 41 56 18 11 9 35 37 42 33 52 43 55 40 64 34

F F F M F F F M M F M F F F F M F M F F M F

Retroperitoneum Adrenal Adrenal Adrenal Adrenal Adrenal Adrenal Adrenal Adrenal Retroperitoneum Adrenal Adrenal Adrenal Adrenal Adrenal Adrenal Retroperitoneum Adrenal Adrenal Adrenal Adrenal Adrenal

Malignant Benign Benign Benign Benign Benign Benign Malignant Benign Benign Benign Benign Benign Benign Benign Benign Benign Malignant Benign Benign Benign Benign

23 24 25 26

10 12 43 43

M M M F

Retroperitoneum Adrenal Adrenal Adrenal

Malignant Benign Benign Benign

Lymphnode metastasis, sporadic Neurofibromatosis type 1 Sporadic Sporadic Sporadic Sporadic Sporadic Sporadic Sporadic Sporadic Sporadic Sporadic Sporadic Sporadic Sporadic von-Hippel Lindau Sporadic Sporadic Sporadic Sporadic Sporadic Familial, bilateral pheochromocytoma (non-MEN2, non-VHL) Sporadic Sporadic Sporadic Familial, bilateral pheochromocytoma (non-MEN2, non-VHL)-sibling of patient 22

ALLELIC DISTRIBUTION OF POLYMORPHISMS IN CUL2 IN PHEOCHROMOCYTOMA

3209

TABLE 2. SSCP conditions for the 21 CUL2 amplicons analyzed in this study SSCP gel bisacrylamide:acrylamide ratio

Glycerol

Ampliconsa

Electrophoresis conditions

10% 1:60

No

100 –150V, o/n, RT

10% 1:37.5

No

12% 1:30

No

100V, o/n, RT 150v, o/n, 4C 100 –150V, o/n, RT

12% 1:37.5 14% 1:100

5% No

100 –150V, o/n, 4C 150 –200V, o/n, RT

2 (53), 15 (47), 16 (48), 18 (47), 20 (56), 21b (53) 14 (45) 19 (46) 6 (47), 8 (45), 10 (53), 11 (48), 13 (47), 17 (48), 21a (53) 3 (48), 4 (48), 5 (48), 7 (45) 9 (46), 12 (47)

a Numbers represent CUL2 exons; the annealing temperature used for each PCR is shown in parentheses. All fragments were amplified in a reaction containing 1.5 mmol/L MgCl2. o/n, Overnight; RT, room temperature.

FIG. 1. Representative SSCP gel of CUL2 amplicon 5 (A), showing the migration shift corresponding to the IVS4 –30G/A polymorphism, and amplicon 6 (B), spanning two new polymorphisms, IVS5– 6C/T and c.590 G/A, R148R. The lanes have been labeled with the respective haplotypes (see text and Table 3 for details).

TABLE 3. Frequency of heterozygosity of CUL2 polymorphisms in pheochromocytoma patients in comparison with ethnic-matched controls and a previous report on renal cell carcinomas (23) CUL2 exons

Type of polymorphism

Pheochromocytomas (%)

Controls (%)a

Clifford et al. (%)

Screening method

2 5 6 6 12 14 19 21

IVS1-43C/T IVS4-30G/A IVS5-6C/T R148R (c.590 G/A) T373T (c.1265G/A) IVS13-41G/A S637S (c.2057G/A) 39UTR (c.2538G/A)

73 23 27b 23 19 19 42 19

69 23 54b 46 11 25 50 11

35 21 0 0 0 26 36 12

SSCP 1 RD (NlaIII) SSCP 1 RD (DdeI) SSCP 1 RD (HpaII) SSCP SSCP SSCP SSCP SSCP

IVS, Intronic sequence; 39UTR, 39-untranslated region; amino acids are represented by standard single letter code; RD, restriction digestion. a Age- and gender-matched group of individuals of the same ethnic background as the patient group. b By one-tailed Fisher’s exact test, P , 0.05; by two-tailed Fisher’s test, P 5 0.08.

RT-PCR In four cases, including one VHL-derived pheochromocytoma, ribonucleic acid and complementary DNA (cDNA) were also available and were obtained as previously described (25). To determine whether CUL2 expression was detectable in these tumors, 2 mL cDNA were used in a 29-cycle duplex PCR containing 0.8 mmol/L of intron-spanning CUL2 primers (2CUL-3F, CAG CAA CCT TAC TCA GGA AAA CAT; 2CUL3R, CAG CGC TGA CAC TCA TAT CTG TA) and 0.1 mmol/L of primers for the housekeeping gene b-glucuronidase (GUSB3, ACT ATC GCC ATC AAC AAC ACA CTC ACC and GUSB5 GAC GGT GAT GTC ATC GAT GT). cDNA from peripheral leukocytes and from a neuronal cell line were used as controls.

Results

We analyzed 22 sporadic and 4 familial pheochromocytomas for the frequency of mutations in the entire coding

region and flanking intronic sequences of the CUL2 gene using the SSCP method. We did not find somatic CUL2 mutations in the DNA of any of the analyzed samples. Seven of the CUL2 exons, 2, 5, 6, 12, 14, 19, and 21, displayed variant SSCP banding patterns, five of which (Table 3) were described by Clifford et al. as polymorphisms in patients with RCC (23). The heterozyogosity frequencies of these five polymorphisms observed in our samples were similar to those described in the aforementioned study (Table 3). In addition, three new sequence variants, two within the coding region of CUL2 and one intronic, were detected in the pheochromocytoma samples (Table 3, Fig. 1). Neither of the two coding variants resulted in an amino acid change: the first one was a single

3210

JCE & M • 1999 Vol 84 • No 9

DUERR ET AL.

base pair change in exon 6, c.590G/A (R148R), and the second was found in exon 12, T373T (c.2057G/A). The third variant was an intronic single base change detected in intron 5, IVS5– 6C/T. Analysis of the germline DNA of each of these samples confirmed that the variants were not limited to the somatic tissue in all but one case. In this single case, only the tumor tissue showed heterozygosity for the c.590G/A, R148R SNP, whereas the corresponding germline DNA was homozygous for the G allele, indicating that the tumor had acquired the nucleotide change. To determine whether the frequencies of the single nucleotide polymorphisms (SNPs) were unique to the pheochromocytoma population, we examined germline DNA from an equal number of age-matched, gender-matched controls with the same ethnic background. We found that the polymorphism frequency was similar between the pheochromocytoma cases and the control population for most of the SNPs identified, except for 3 variants, 2 of which have been identified in the present study. In the IVS5– 6C/T SNP, the frequency of the C allele (combined CC and CT genotypes) was higher among pheochromocytoma cases (96%) than in the control population (80%). When only homozygosity for the C allele was analyzed (CC, compared to the CT or TT genotype), the differences were more striking; 16 of 26 pheochromocytomas were homozygous for the C allele, whereas only 7 of 26 controls had this genotype (61% vs. 27%, respectively; x2 5 9.84; P , 0.01). When C and T alleles were analyzed individually, of the total of 52 pheochromocytoma chromosomes tested, 43 (83%) carried the polymorphic variant C, whereas 9 (17%) harbored the allele T, thus corroborating the initial observation. In contrast, among 52 control chromosomes, 28 (54%) had the variant C, and 24 (46%) had the wild-type allele T (P , 0.005). The allelic distribution of another variant, the previously described SNP in exon 19, S637S (c.2057G/A), also differed between the 2 groups; pheochromocytoma patients carried the G allele more frequently (34 of 48 chromosomes) than the control population (22 of 48 chromosomes; P , 0.01). A third polymorphism, a novel SNP found at c.590G/A, R148R, was heterozygous in almost 50% of the controls, whereas less than one third of pheochromocytomas had heterozygosity at this site (Table 3). However, the difference between the two groups did not reach statistical significance. Of interest, all 4 malignant pheochromocytomas were homozygotes for the C allele at IVS5– 6C/T, for the G allele at S637S (c.2057G/A), and for the G allele at c.590G/A, R148R. The CUL2 SNPs were further used for loss of heterozygosity (LOH) analysis of these tumors. In one case, LOH was identified by restriction enzyme analysis of the IVS1– 43C/T polymorphism. This sample originated from a patient with a sporadic, benign, ACTH-secreting pheochromocytoma. None of the remaining tumors showed evidence of LOH in the CUL2 gene. Only two tumors, one benign and one malignant, were not informative for any of the SNPs. To exclude the possibility that some sequence variants were missed by SSCP, we directly sequenced the entire coding region of theCUL2 gene in three of our samples. No sequence variations were identified besides the SNPs described above. The duplex RT-PCR of four pheochromocytomas revealed

that CUL2 was expressed in all tumors at levels similar to those of one neuroectoderm-derived and one mesodermderived tissue (Fig. 2). Although only a small number of samples was available for such analysis, these results suggest that transcriptional silencing, e.g. via epigenetic mechanisms such as complete gene methylation, is unlikely to be a major mode of inactivation of CUL2 in pheochromocytomas. Discussion

We did not detect any somatic pathogenic mutations in the CUL2 gene in a comprehensive series of pheochromocytomas. To exclude the possibility that mutations were missed by the SSCP analysis, we directly sequenced the entire coding region of CUL2 in three of the tumor samples, and no mutations were detected. In contrast with the 24% frequency of LOH at the CUL2 locus at 10p11.1–11.2 in RCCs (23), we only found one pheochromocytoma with evidence of deletion of one copy of this gene. Taken together, the results of the mutation screening, sequencing, and LOH analysis suggest that it is unlikely that somatic high penetrance mutations in the CUL2 gene play a major role in the pathogenesis of pheochromocytomas. However, it is still not known whether CUL2 is inactivated by mechanisms other than mutations or deletions. It has been demonstrated that hypermethylation of CpG islands inactivate some tumor suppressor genes by reducing their transcription rates (e.g. VHL, p16) (reviewed in Ref. 26). The 59-end of the CUL2 gene contains many of the typical features of a CpG island (23). However, our expression data in a small number of pheochromocytomas have revealed that the levels of CUL2 transcription are similar to those of control tissues by RT-PCR, suggesting that hypermethylation of CUL2 is unlikely to play a major role in the pathogenesis of these endocrine tumors. Another potential mechanism of gene inactivation is posttranslational silencing. Mechanisms of inactivation of the CUL2 protein are not currently understood, and its status in tumor conditions is not known. It is possible that the function of CUL2 could be disrupted at this level. In this case, the pathogenic effects of this product might result not only from the absolute levels of the protein, but also from its interaction with the VHL product. Further studies are

FIG. 2. Ethidium bromide-stained agarose gel of duplex RT-PCR products. CUL2 and a housekeeping gene, GUSB, were coamplified from four pheochromocytomas (no. 1– 4), one leukocyte (no. 5), and one neuron-derived cell line (no. 6) cDNAs; no. 7 is a negative control for the PCR.

ALLELIC DISTRIBUTION OF POLYMORPHISMS IN CUL2 IN PHEOCHROMOCYTOMA

necessary to determine whether posttranslational inactivation plays a major part in CUL2 turnover. The finding of an apparent germline overrepresentation of certain alleles (the C allele in the IVS5–6C/T and the G allele in the exon 19 S637S polymorphism) in the pheochromocytoma patients compared with that in a control population with similar ethnic background is interesting, but its relevance is still unknown. Characterization of the true frequency of these variants in a larger number of pheochromocytoma cases from different ethnic backgrounds is required to confirm whether these alleles are nonrandomly associated with the pheochromocytoma phenotype in general or only those in a specific ethnic background. Overrepresentation of certain polymorphisms has previously been associated with disease phenotypes, such as in the RET gene and sporadic medullary thyroid carcinoma (27) or Hirschsprung disease (28), p53 and HPV-related cervical cancer (29), APC and familial colorectal cancer (30), the chemokine coreceptor CCR51, and human immunodeficiency virus (31). However, in some of these cases, e.g. the association between the p53 polymorphism, P72R, and cervical cancer, the initial finding has not been confirmed in further studies with larger populations (32) or in series with different ethnic backgrounds (33). Usually, such polymorphisms are located in coding regions of the respective genes and, with a single exception (27), result in amino acid change, unlike the SNPs studied in our series. However, they can also lie within noncoding areas (31). Due to its proximity to the exon-intron boundary, the IVS5– 6C/T CUL2 variant may affect splice efficiency at the nearby exon-intron boundary or may interfere with transcript stability. Unfortunately, ribonucleic acid was not available from the samples studied to investigate splice changes at this region. Despite their close proximity, we did not find any evidence for linkage disequilibrium between any alleles at these three SNPs. In summary, we have not identified somatic pathogenic mutations in the coding region of a VHL-associated molecule, CUL2, in a series of sporadic and familial pheochromocytomas. Although our findings suggest that this gene does not play a major role in the pathogenesis of such tumors, it is unknown whether epigenetic mechanisms might be involved in its inactivation in VHL-associated tumors. Furthermore, the potential role for the differences in the allelic frequencies of two CUL2 polymorphisms in the pheochromocytoma group requires further investigation. Acknowledgments We are grateful to Ricardo C. T. Aguiar for providing the control samples used in this study, and to Debbie J. Marsh for critical review of this manuscript. E.-M.D. thanks Andreas von Deimling for his support.

References 1. Bravo EL, Gifford RW. 1984 Pheochromocytoma: diagnosis, localization and management. N Engl J Med. 30:1682–1686. 2. Kalff V, Shapiro B, Lloyd R, Nakajo M, Sisson JC, Beierwaltes WH. 1984 Bilateral pheochromocytomas. J Endocrinol Invest. 7:387–392. 3. Irvin GLd, Fishman LM, Sher JA. 1983 Familial pheochromocytoma. Surgery. 94:938 –940. 4. Mulligan LM, Kwok JBJ, Healey CS, et al. 1993 Germline mutations of the RET proto-oncogene in multiple endocrine neoplasia type 2A. Nature. 363:458 – 460. 5. Donis-Keller H, Dou S, Chi D, et al. 1993 Mutations in the RET protooncogene are associated with MEN 2A and FMTC. Hum Mol Genet. 2:851– 856. 6. Eng C, Smith DP, Mulligan LM, et al. 1994 Point mutation within the tyrosine kinase domain of the RET proto-oncogene in multiple endocrine neoplasia type 2B and related sporadic tumours. Hum Mol Genet. 3:237–241.

3211

7. Carlson KM, Dou S, Chi D, et al. 1994 Single missense mutation in the tyrosine kinase catalytic domain of the RET protooncogene is associated with multiple endocrine neoplasia type 2B. Proc Natl Acad Sci USA. 91:1579 –1583. 8. Hofstra RMW, Landsvater RM, Ceccherini I, et al. 1994 A mutation in the RET proto-oncogene associated with multiple endocrine neoplasia type 2B and sporadic medullary thyroid carcinoma. Nature. 367:375–376. 9. Maher ER, Kaelin Jr WG. 1997 von Hippel-Lindau disease. Medicine. 76:381–391. 10. Crossey PA, Richards FM, Foster K, et al. 1994 Identification of intragenic mutations in the Von Hippel-Lindau disease tumour suppressor gene and correlation with disease phenotype. Hum Mol Genet. 3:1303–1308. 11. Zbar B, Kishida T, Chen F, et al. 1996 Germline mutations in the Von HippelLindau disease (VHL) gene in families from North America, Europe, and Japan. Hum Mutat. 8:348 –357. 12. Eng C, Crossey PA, Mulligan LM, et al. 1995 Mutations in the RET protooncogene and the von Hippel-Lindau disease tumour suppressor gene in sporadic and syndromic phaeochromocytomas. J Med Genet. 32:934 –937. 13. Hofstra RMW, Stelwagen T, Stulp RP, et al. 1996 Extensive mutation scanning of RET in sporadic medullary thyroid carcinoma and of RET and VHL in sporadic pheochromocytoma reveals involvement of these genes in only a minority of cases. J Clin Endocrinol Metab. 81:2881–2884. 14. Lindor NM, Honchel R, Khosla S, Thibodeau SN. 1995 Mutations in the RET protooncogene in sporadic pheochromocytomas. J Clin Endocrinol Metab. 80:627– 629. 15. Dahia PL, Aguiar RC, Tsanaclis AM, et al. 1995 Molecular and immunohistochemical analysis of P53 in phaeochromocytoma. Br J Cancer. 72:1211–1213. 16. Aguiar RC, Dahia PL, Sill H, Toledo SP, Goldman JM, Cross NC. 1996 Deletion analysis of the p16 tumour suppressor gene in phaeochromocytomas. Clin Endocrinol (Oxf). 45:93–96. 17. Dahia PL, Toledo SP, Mulligan LM, Maher ER, Grossman AB, Eng C. 1997 Mutation analysis of glial cell line-derived neurotrophic factor (GDNF), a ligand for the RET/GDNF receptor alpha complex, in sporadic phaeochromocytomas. Cancer Res. 57:310 –313. 18. Kaelin WG, Iliopoulos O, Lonergan KM, Ohh M. 1998 Functions of the von Hippel-Lindau tumour suppressor protein. J Intern Med. 243:535–539. 19. Duan DR, Pause A, Burgess WH, et al. 1995 Inhibition of transcription elongation by the VHL tumor suppressor protein. Science. 269:1402–1406. 20. Kibel A, Iliopoulos O, DeCaprio JA, Kaelin WG, Jr. 1995 Binding of the von Hippel-Lindau tumor suppressor protein to Elongin B and C. Science. 269:1444 –1446. 21. Lonergan KM, Iliopoulos O, Ohh M, et al. 1998 Regulation of hypoxia-inducible mRNAs by the von Hippel-Lindau tumor suppressor protein requires binding to complexes containing elongins B/C and Cul2. Mol Cell Biol. 18:732–741. 22. Pause A, Lee S, Worrell RA, et al. 1997 The von Hippel-Lindau tumorsuppressor gene product forms a stable complex with human CUL-2, a member of the Cdc53 family of proteins. Proc Natl Acad Sci USA. 94:2156 –2161. 23. Clifford SC, Walsh S, Hewson K, et al. Genomic organisation and chromosomal localisation of the human CUL2 gene and the role of von Hippel-Lindau tumour suppressor-binding protein (CUL2 and VBP-1) mutation and loss in renal cell carcinoma development. Genes Chrom Cancer. In press. 24. Dahia PL, Marsh DJ, Zheng Z, et al. 1997 Somatic deletions and mutations in the Cowden disease gene, PTEN, in sporadic thyroid tumors. Cancer Res. 57:4710 – 4713. 25. Dahia PLM, Aguiar RCT, Alberta J, et al. 1999 PTEN is inversely correlated with the cell survival factor Akt/PKB and is inactivated via multiple mechanismsin haematological malignancies. Hum Mol Genet. 8:185–193. 26. Baylin SB, Herman JG, Graff JR, Vertino PM, Issa JP. 1998 Alterations in DNA methylation: a fundamental aspect of neoplasia. Adv Cancer Res. 72:141–196. 27. Gimm O, Neuberg DS, Marsh DJ, et al. 1999 Over-representation of a germline RET sequence variant in patients with sporadic medullary thyroid carcinoma and somatic RET codon 918 mutation. Oncogene. 18:1369 –1373. 28. Borrego S, Saez ME, Ruiz A, et al. 1999 Specific polymorphisms in the RET proto-oncogene are over-represented in individuals with Hirschsprung disease and may represent loci modifying phenotypic expression. J Med Genet. In press. 29. Storey A, Thomas M, Kalita A, et al. 1998 Role of a p53 polymorphism in the development of human papillomavirus- associated cancer. Nature. 393:229–234. 30. Laken SJ, Petersen GM, Gruber SB, et al. 1997 Familial colorectal cancer in Ashkenazim due to a hypermutable tract in APC. Nat Genet. 17:79 – 83. 31. Dean M, Carrington M, Winkler C, et al. 1996 Genetic restriction of HIV-1 infection and progression to AIDS by a deletion allele of the CKR5 structural gene. Hemophilia Growth and Development Study, Multicenter AIDS Cohort Study, Multicenter Hemophilia Cohort Study, San Francisco City Cohort, ALIVE Study. Science. 273:1856 –1862. 32. Rosenthal AN, Ryan A, Al-Jehani RM, Storey A, Harwood CA, Jacobs IJ. 1998 p53 codon 72 polymorphism and risk of cervical cancer in UK. Lancet. 352:871– 872. 33. Minaguchi T, Kanamori Y, Matsushima M, Yoshikawa H, Taketani Y, Nakamura Y. 1998 No evidence of correlation between polymorphism at codon 72 of p53 and risk of cervical cancer in Japanese patients with human papillomavirus 16/18 infection. Cancer Res. 58:4585– 4586.