|
|
 |
1997 Patient/Provider ConferenceResearch and VHL: Where do we stand, where are we going?
Abstracts
Saturday, May 3, 1997, Bethesda, Maryland
VHL, the Human Genome, and the Future
See "Sequencing the Human Genome," article in Hospital Practice, January 15, 1997. Related article online.
Functional Analysis of the VHL tumor suppressor protein (pVHL)
Presenting the work of the team of O. Iliopoulos, M. Ohn, K. Lonergan, W.G. Kaelin, Dana-Farber Cancer Institute and Harvard Medical School, Boston, Massachusetts
Our laboratory previously identified pVHL as a 30 kDa, primarily cytoplasmic phosphoprotein. Additionally, a second VHL isoform (pVHL-19) is generated by translational initiation at an internal ATG start codon.
We showed that reintroduction of wild type (wt), but not mutant pVHL, into VHL (-/-) RCC lines suppressed their growth as tumors in nude mice. To gain insight into the molecular mechanism(s) of tumor suppression, pVHL-binding proteins were sought. We and others showed that pVHL binds in vitro and in vivo to two transcription elongation factors, elongins B and C, through a domain which is a "hotspot" for naturally occurring mutations. These two factors bind to elongin A to form a transcriptional elongation factor complex (SIII). This complex enhances the processivity of RNA pol II-mediated transcription. pVHL competes with elongin A for binding to elongins B and C, at least in vitro, thereby inhibiting SIII activity and possibly regulating the expression of certain genes.
Indeed, pVHL negatively regulated hypoxia-inducible RNAs, some of which encode angiogenic peptides such as vascular endothelial growth factor (VEGF). RCC clones lacking wt pVHL overproduce mRNAs encoding VEGF, the glucose transporter GLUT 1, and the mutant pVHL into these cells specifically reduced the levels of these mRNAs and their corresponding proteins under normoxic conditions. VEGF inhibition may therefore be a mechanism of tumor suppression by pVHL and may explain the differential effect of pVHL on tumor cell growth in vitro and in vivo. The changes in mRNA abundance could be primarily attributed to changes in mRNA stability rather than changes in transcriptional elongation of hypoxia-inducible genes. This might suggest an additional role for the elongins in vivo: pVHL/elongin B/C complex may be directly involved in mRNA destabilization.
Recently we showed that pVHL/C/B complex interacts in vivo with cullin-2 (cul2), a member of a newly identified family of proteins (cullins). Cul2, a putative homologue of yeast cdc53, may act as an E 3 ubiquitin ligase by targeting specific proteins to destruction by the ubiquitin pathway. pVHL mutants unable to bind elongin C also fail to coimmunoprecipitate cul2 in vivo. The protein targets of cul2 containing complexes and the functional significance of pVHL-elongin C/B interaction with cul2 are currently under investigation.
Molecular Diagnosis of von Hippel-Lindau: Results from a clinical laboratory
Since localization and cloning of the VHL gene, there have been rapid advances in the ability to identify germline mutation in individuals with the disease. Localization of the gene to chromosome 3p25-26 led to the development of markers that could e used in linkage analysis in suitable families. Cloning of the gene provided a probe that could be used in direct mutation analysis for rearrangements in the gene present in 15-20% of patients. More recently, gene scanning techniques and DNA sequencing have been used to identify single base mutations in the gene.
The Genetic Diagnostic Laboratory at the University of Pennsylvania School of Medicine is a CLIA approved facility that has been performing molecular diagnostic testing for VHL since July of 1995. Testing consissts of Southern blot analysis for rearrangements of deletions in the gene, followed by DNA sequence analysis of all 3 exons (coding sequences), if necessary. To date, we have analyzed 100 probands and performed mutation analysis on 104 at risk family members. Of the probands diagnosed with VHL, 23% had rearrangements in the gene, 6% appeared to have a deletion of one allele, and 63% had point mutations (i.e., a substitution, deletion, or insertion of 1-8 nucleotides). The remainder of patients (8%) had no detectable mutation. Overall, mutations in the VHL gene were identified in 92% of individuals diagnosed with VHL.
A small number of samples from individuals that do not clearly fit the clinical criteria for VHL have been analyzed for VHL mutations. Although results may not be typical of a broader sampling of such patients, VHL gene mutations were identified in 1 out of 5 individuals with hemangioblastoma only (i.e., 20%), 2 out of 8 patients with retinal angioma only (i.e., 25%), and 1 out of 1 samples with pheochromocytoma only (i.e., 100%). None of the three samples submitted from patients with renal cell carcinoma only had VHL mutations.
While a detection rate of 92% is a considerable improvement over detection rates reported within the last three years, the question remains: where are (and what are) the mutations causing the remaining 8% of VHL? There is no evidence of another gene involved in VHL. Several possibilities exist: mutations affecting expression of the gene, mutations deep within the introns, mutations affecting mRNA stability, or mutations at some distance from the gene. Additional research may help to find these elusive causes of VHL.
Inactivation of the VHL gene by DNA Methylation
The von Hippel-Lindau gene is a classic tumor suppressor gene which is mutated in the germline of patients with von Hippel-Lindau disease, and in the tumors of patients with non-inherited clear cell renal carcinoma. Inactivation of VHL can also be caused by hypermethylation of the promoter region CpG island of this gene. This change leads to loss of RNA expression, which can be at least partially reversed by treatment with the demethylating drug 5-aza-2'-deoxycytidine. Hypermethylation of the VHL promoter is found in 20% of renal cell carcinoma. but thus far in no other tumor type, and not in the germlline of patients with the VHL disease. Methylation of the CpG island of VHL in these sporadic tumors appears to spread into this island from neighboring normally methylated sequences. The identification of VHL inactivation by DNA hypermethylation has led to the establishment of this mechanism of inactivation for other tumor suppressor genes, some of which may be altered in tumors harboring VHL mutations (for example the p16 gene). Finally, the normal methylation of the CpG dinucleotide in the coding region of VHL can account for many of the frequent transition mutations found in families with VHL mutations.
see "Silencing of the VHL tumor-suppressor gene by DNA methylation in renal carcinoma," article in Proc. Natl. Acad. Sci. USA, 19:9700-9704, October 1994.
|
| Contact Us | Copyright 1993-2008 |