| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Article |
Institute of Pathology (G.W.), Technische Universität München, D-81675 Munich; Wilex Biotechnology GmbH (T.W.), D-81675 Munich; Medical Department II (D.E.), Klinikum Grosshadern, University of Munich, D-81366 Munich; and Dr. Margarete Fischer-Bosch-Institute of Clinical Pharmacology (B.K., H.B.), D-70376 Stuttgart, Germany
Address all correspondence and requests for reprints to: Dr. Hiltrud Brauch, Dr. Margarete Fischer-Bosch-Institute of Clinical Pharmacology, Auerbachstrasse 112, D-70376 Stuttgart, Germany. E-mail: hiltrud.brauch{at}ikp-stuttgart.de.
Abstract
Von Hippel-Lindau disease (VHL) is a multitumor syndrome that develops on the basis of germline mutations in the VHL tumor suppressor gene. Genotype-phenotype correlations have helped to stratify the disease into VHL type 1 (without pheochromocytoma) and VHL type 2A, 2B, and 2C (with pheochromocytoma). VHL2C is characterized by a pheochromocytoma-only phenotype. We report on the P81S germline mutation in a German VHL2C family with the previously identified L188V mutation. The concurrent P81S mutation was identified by novel screening approaches including denaturing HPLC and sequencing. We show the co-segregation of these two mutations with the disease and discuss their possible impact on pVHL function and phenotype.
PHEOCHROMOCYTOMAS ARE CATECHOLAMINE-SECRETING tumors that develop from chromaffin cells of the neural crest. They are located mainly in the adrenal medulla and to a lesser extent at extraadrenal sites. About 10% of pheochromocytomas develop in the context of complex hereditary diseases, namely von Hippel-Lindau (VHL) disease (MIM 193300), multiple endocrine neoplasia type 2 (MIM 171400), and neurofibromatosis type 1 (MIM 162200). VHL is an autosomal dominant disease based on germline mutations in the VHL tumor suppressor gene (1, 2) and is characterized by a conspicuous pleiotropy. Most VHL patients are afflicted by cystic and tumorous lesions at multiple sites including eyes, brain, spinal cord, kidneys, adrenals, pancreas, and epididymis. Several hundred VHL germline mutations have been detected since the discovery of the disease gene, among which there are a limited number of hot spots, and it is, therefore, not surprising that the majority of VHL families carry unique mutations (2, 3, 4).
Extensive genotype-phenotype correlation efforts have helped to stratify the VHL disease into two main subtypes: VHL type 1 and VHL type 2 (2). Carriers of VHL type 2 mutations are at risk to develop pheochromocytoma, whereas carriers of VHL type 1 mutations are not. VHL type 2 has further been subdivided into mutation carriers with a low (2A) (5) or high risk (2B) (6) to develop renal cell carcinoma. A third group is characterized by the pheochromocytoma-only phenotype (2C) (7, 8).
Most VHL mutations are deleterious for the VHL protein (pVHL) and deprive target cells of the ability to orchestrate the expression of a diverse set of genes involved in some signal transduction pathways, cellular responses to hypoxia, and regulation of the compromised metabolic microenvironment in tumors (9). One of the best understood pVHL functions is the controlled degradation of 
-subunits of the heterodimeric transcription factor hypoxia inducible factor (HIF) (Fig. 1
; Ref. 10). In contrast to loss of function variants in VHL type 1, mutations predisposing to pheochromocytoma (VHL type 2) are mainly of the missense type predicted to give rise to conformationally changed pVHL (2). Such altered pVHL may still be able to exert some of its functions and/or may gain novel ones.
|
and ß (Fig. 2A). The -helical domain extends from residue 158 to 189 and binds Elongin C, which in turn binds Elongin B and Cul2 to form a ubiquitin ligase (11). The ß-sheet domain extends from residue 63 to 157 and binds to targets such as HIF. In the presence of oxygen, pVHL thus polyubiquitinates HIF1 and HIF2 subunits that are in turn degraded by the 26S proteasome, an event thought to regulate HIF abundance (12, 13). Interestingly, the two most frequent VHL germline mutations 712 C>T and 713 G>A (codon 167) are located in the -helical domain and are associated with pheochromocytoma and renal cell carcinoma (VHL2B) (2, 14). In contrast, the frequent VHL founder mutation 505 C>T affecting codon 98 is located within the ß-sheet domain and is associated with VHL2A (frequent pheochromocytoma and hemangioblastoma but rare renal cell carcinoma) (5). Likewise, other mutations identified in VHL type 2 families (Table 1) affect either the -helical domain or the ß-sheet domain. The pVHL is a tumor suppressor, and tumorigenesis is thought to occur whenever a single VHL germline mutation is passed down from a parent carrier to the offspring or occurs de novo, followed by loss of the homologous 3p-allele in tumor-prone tissues (1, 15). Here, we report on a family with the exceptional co-segregation of the pheochromocytoma-only phenotype (VHL2C) with two concurrent germline mutations in pVHL - and ß-domains.
|
|
Subjects
Our investigation included 16 members of two branches of a German family. Seven members had previously been diagnosed with pheochromocytomas; some patients had single (n = 2), and others had bilateral (n = 4) or unilateral and extraadrenal (n = 1) pheochromocytomas (7). The diagnosis of VHL disease had been established by the detection of a VHL 775 C>G (L188V) germline mutation in these patients (7, 8, 16). We subjected this family to a new investigation for two reasons. First, when optimizing the denaturing HPLC (DHPLC) method for VHL mutation screening (4), we noticed a novel mutation type elution profile of VHL exon 1 in the index patient. This patient and all other members of this family had formerly tested negative in single-strand conformation polymorphism screening of VHL exon 1, and, therefore, further analysis of this exon had not been performed (8, 16). Sequencing of the newly detected conspicuous PCR product now detected a 454 C>T transition, which is a missense mutation like the previously identified 775 C>G transversion. Second, in addition to this finding in the index patient, we also identified the exon 1 mutation in a new family member. Thus, to clarify whether the 454 C>T change represented a polymorphism or a mutation that co-segregated with the 775 C>G mutation and VHL2C phenotype, we subjected all 16 family members and 65 unrelated German individuals without VHL disease to this analysis. Informed consent was obtained from human subjects and the study was conducted in accordance with the Declaration of Helsinki and approved by the appropriate institutional review board.
Methods
DNA was isolated from whole blood by the phenol-chloroform method. Two independent mutation detection methods, i.e. the DHPLC technique and sequencing analysis, were applied. In conjunction with DHPLC, VHL exon 1 was amplified with the novel primers VHL13F (5'-CCGAGGAGGAGATGGAG-3') and VHL13R (5'-TATCGTCCCTGCTGGGTC-3'), which generate a fragment comprising nucleotides 362 to 553 + 41 (GenBank accession no. AF010238). Exon 3 was amplified as previously reported (16). Post-PCR analysis of PCR products was carried out by DHPLC as previously described (4). All exon 1 PCR products of 65 subjects without VHL disease and exon 1 and exon 3 PCR products of family members were sequenced by the dye terminator method on an ABI Prism 310 genetic analyzer (Applied Biosystems, Weiterstadt, Germany) following published procedures (4, 16).
Results and Discussion
Concurrent VHL mutations
DHPLC analysis detected heterozygous mutation type changes in exon 1 and exon 3 in 10 family members, nine patients, and a disease-free child. Figure 3 shows the larger branch of this family with six affected individuals (patients 2, 3, 5, 7, 8, and 9) and the disease-free mutation carrier (individual 11). All other family members exhibited wild-type chromatograms in DHPLC analysis (Fig. 3; individuals 1, 4, 6, and 10). Direct sequencing analyses confirmed the 454 C>T mutation (P81S) in exon 1 and the 775 C>G mutation (L188V) in exon 3 (Fig. 2B
). Both mutations are deposited at the genome variation database (HGVbase) under accession numbers SNP001494113 and SNP001494133 (http://hgvbase.cgb.ki.se). All patients and carriers of the 775 C>G (L188V) mutation were also carriers of the 454 C>T mutation (P81S). Thus, over three generations both mutations had co-segregated with the phenotype and were absent in nonaffected family members and spouses. Because of the fact that the unaffected young child (Fig. 3, individual 11) is a carrier of both mutations, this subject must be considered at high risk to develop pheochromocytoma.
|
The concurrence of two missense mutations in a VHL2C family raises the question of how each one of them contributes to the development of this remarkable phenotype. Advances in the understanding of the pheochromocytoma-only phenotype have been sparked by in vitro experiments. These studies showed that the 775 C>G (L188V) mutant, identified in a single VHL2C family, binds both HIF and Elongin C and is able to suppress growth of pVHL defective renal cell carcinoma cells in nude mice (17). These findings fit well into the concept that a minimal phenotype may originate from minimal impairment of protein function. Accordingly, the VHL2C phenotype may emerge from mutations with a low impact on pVHL core functions. The new finding of concurrent P81S and L188V mutations in this family now calls for a reinterpretation of the origin of the VHL2C phenotype. There is a possibility that the newly discovered P81S change may play a separate role for the development of VHL2C. This is supported by the fact that the P81S change is located in the functionally important ß-sheet domain of pVHL. Also, there is strong evidence that the 454 C>T (P81S) transition is not a single nucleotide polymorphism because it was absent in 65 unrelated subjects from Germany and has not been detected in more than 140 subjects in other studies (18, 19).
P81S mutation in other VHL families
In keeping with this finding are a few documented observations of the 454 C>T (P81S) mutation as the sole mutation in VHL families. With one exception, affected families exhibited a VHL type 1 phenotype and included one German (16), one Dutch (19), one Japanese (20), and two families from North America (19). Interestingly, the presence of the 454 C>T (P81S) mutation in these families was associated with low penetrance or mild phenotype. Low penetrance has previously been observed by us in the German VHL family with one affected member with central nervous system (CNS) hemangioblastoma and renal cell carcinoma as well as kidney and pancreatic cysts (16). Although the patient’s father was himself a mutation carrier, he remained free of disease until his death at the age of 89 yr. Likewise, a low penetrance of the 454 C>T allelotype has also been described in the Dutch family. In this family, a 44-yr-old patient with a solitary CNS hemangioblastoma was the only affected member among five mutation carriers (father, son, sister, and aunt, aged 77, 17, 43, and 66 yr, respectively) (19). The single patient of the Japanese family was reported having a CNS hemangioblastoma and a renal cell carcinoma (20). Two other 454 C>T (P81S) families were of North American origin. In one of these families, a single affected member had a CNS hemangioblastoma, whereas in the second family there was evidence for a VHL type 2 phenotype. In the latter, one patient was affected by retinal angioma and an islet cell tumor of the pancreas, and a second patient had a pheochromocytoma (19). In these five families with a 454 C>T allelotype the number of affected members is limited, which impedes a reliable genotype-phenotype correlation.
Nevertheless, it has been proposed that the 454 C>T transition may favor the development of CNS hemangioblastoma (19). In addition, the fact that most mutation carriers remained disease free throughout their lives supports at least the notion of a low penetrance of the 454 C>T (P81S) mutation. Independent evidence for a role of the 454 C>T transition (P81S) stems from observations in a subset of patients with sporadic clear cell renal cell carcinoma. On the somatic level, the 454 C>T (P81S) mutation was occasionally identified in nontumorous kidney parenchyma, which may be interpreted as manifestation of a low tumorigenicity (18).
Our findings in the German VHL2C family seem not to corroborate the assumption that the 454 C>T (P81S) mutation is associated with hemangioblastoma and renal cell carcinoma. Furthermore, all mutation carriers except for the disease-free child had developed pheochromocytoma at some point during their lives. The pheochromocytoma-only phenotype and the seemingly high penetrance both may be ascribed to the second mutation at nucleotide 775 C>G (L188V). One could further speculate that in this VHL2C family, the 775 C>G transversion overrides the effect of the 454 C>T transition with respect to the phenotype.
Putative impact of P81S and L188V mutations on pVHL function
It is worth noting that the 454 C>T transition (P81S) is located within the ß-sheet domain of pVHL and the 775 C>G transversion (L188V) within the -helical pVHL domain. Hence, a better understanding of putative functional alterations may be obtained from a review of the structural properties of pVHL. Stebbins et al. (11) assigned Elongin C binding mainly to the pVHL -helical domain including L188, and their data also show that P81 and R82 are located at the site of -pVHL, ß-pVHL, and Elongin C interactions. Later Hoffman et al. (17) provided evidence for the L188V mutation not to disturb Elongin C binding in vitro, but it is conceivable that the additional P81S mutation abrogates or weakens this interaction. In particular, replacement of proline by serine at residue 81 may disturb the native pVHL structure by destroying a rigid kink known to be associated with proline and to favor the formation of hydrogen bonds with neighboring polar amino acids. Because of the location of P81S within the ß-sheet domain of pVHL, we may infer that protein binding including HIF- is impaired. This effect may even be stronger in combination with the L188V mutation. Altogether there is a likelihood that both mutations influence protein-protein interactions vice versa at respective sites.
Considering its role in target capture, the P81S mutation may not only disturb HIF- regulation but may also affect fibronectin assembly. There have been somewhat conflicting observations in conjunction with VHL mutations associated with different phenotypes. For example, Clifford et al. (21) showed both loss of HIF- regulation and fibronectin binding for VHL mutants associated with hemangioblastomas including a mutation that replaced L188 by a glutamine (L188Q). In contrast, fibronectin binding was not affected when L188 was replaced by valine (L188V). However, Hoffman et al. (17) showed loss of intact fibronectin assembly in conjunction with the pheochromocytoma-associated L188V mutation from which we may infer loss of fibronectin capture function. When we assume that the L188V maintains wild-type pVHL functions with respect to fibronectin binding and HIF- regulation, there is a chance that P81S will disrupt these functions. On the other hand, if the VHL2C phenotype requires residual pVHL function, the interplay of L188V and P81S may result in a gain rather than a loss of function. Why this combination of germline mutations preferentially affects chromaffin cells is at present unclear. It will, therefore, be of interest to know the target molecules sensible to disturbances by the P81S mutation as well as its downstream effects in tumor tissue.
Acknowledgments
Footnotes
Abbreviations: CNS, Central nervous system; DHPLC, denaturing HPLC; HIF, hypoxia inducible factor; pVHL, von Hippel-Lindau disease protein; VHL, von Hippel-Lindau disease.
Received April 26, 2002.
Accepted July 26, 2002.
References
and HIF-2 under normoxic conditions in renal carcinoma cells by von Hippel-Lindau tumor suppressor gene loss of function. Oncogene 19:5435–5443[CrossRef][Medline]
regulation by disease-causing pVHL mutations correlate with patterns of tumourigenesis in von Hippel-Lindau disease. Hum Mol Genet 10:1029–1038This article has been cited by other articles:
 |
|
 |
|
  K. Knauth, E. Cartwright, S. Freund, M. Bycroft, and A. Buchberger VHL Mutations Linked to Type 2C von Hippel-Lindau Disease Cause Extensive Structural Perturbations in pVHL J. Biol. Chem., April 17, 2009; 284(16): 10514 - 10522. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. J. Hes, J. W. M. Hoppener, and C. J. M. Lips Pheochromocytoma in Von Hippel-Lindau Disease J. Clin. Endocrinol. Metab., March 1, 2003; 88(3): 969 - 974. [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
