Genetics in medicine 07: von Hippel-Lindau

These are my notes from week 7 of Harvard’s Genetics 228: Genetics in Medicine from Bench to Bedside course, held on March 13, 2015. Lecture delivered by Dr. Othon Iliopoulos.

Dr. Othon Iliopoulos: Clinical and molecular aspects of hereditary RCC disease

Here is an overview of hereditary renal cancers.

We still do not not understand the reasons why mutations in different genes give rise to tumors of different cell types.

Today’s talk will focus on VHL. It is genetically recessive but exhibits autosomal dominant inheritance - in other words, a homozygous loss-of-function mutation in VHL is what causes tumorigenesis, but a heterozygous mutation transmitted in the germline will almost certainly be homozygosed or compound heterozygosed with a somatic mutation, in some cell at some point in your life, thus you get almost 100% penetrance. The incidence of VHL is 1 in 36,000 live births. Thanks to good treatment options, they have almost as high a life expectancy as the general population does, so the prevalence is higher than the incidence would suggest. Age of onset of first tumor can be as early as age 18, but is most often in the 30-35 range. Sporadic renal cell carcinomas, by contrast, are usually at age 50 or older.

Treatment has evolved over time. In the 1980s, when a tumor was observed, doctors would remove the whole kidney. This fell out of favor because many patients eventually get tumors in both kidneys and you don’t want to take out both. In the 1990s they would remove just the tumor. Today they just monitor tumors <3cm and when they get large enough, they will treat laproscopically or some other minimally invasive way. They do still take out the kidneys and turn to transplantation or dialysis occasionally, but only as a last resort. In VHL patients who come to their checkups regularly so that tumors can be closely monitored, kidney cancers are very unlikely to be life-threatening.

The bigger challenge is hemangioblastomas, which sometimes strike earlier, even during childhood, and often on the retina. Sometimes you can remove them with laser treatment, but not if it gets into the optic nerve.

Pheochromocytomas are sufficiently rare that anytime somebody presents with one, they immediately test for germline mutations. Almost all cases are genetic, only a few are sporadic.

VHL is a great example of the difference between penetrance and expressivity. All identified mutations are nearly 100% penetrant, but they have widely varying expressivity, meaning they result in a different set of phenotypes. Nonsense, deletions and a few missense variants predicted to completely disrupt protein folding cause “type I” VHL with hemangioblastoma and renal carcinomna but never pheochromocytoma. Missense mutations predicted to be hypomorphic or partial LoF cause “type II” VHL, associated with pheochromocytoma and sometimes also hemangioblastoma or renal carcinoma.

The VHL clinic at MGH recommends that patients be referred to them if they have even a single pheochromocytoma or hemangioblastoma with no family history, or if they have ≥2 other tumor types associated with VHL, or if they have one tumor type associated with VHL plus a family history.

The VHL protein inhibits a transcription factor called HIFa, which stands for hypoxia-inducible factor A. HIFa’s activity is also regulated by hydroxylation of its prolines by another protein called EGLN, as part of a larger regulatory pathway such that tumor hypoxia actually increases HIFa activity, which in turn supports tumor formation by promoting angiogenesis. This seems like a terrible decision on evolution’s part, but it is speculated that this pathway is needed because ordinary, non-cancerous cells need nutrients too, and so when they’re short on oxygen they need to be able to recruit blood vessels. In healthy humans, HIFa is almost never expressed - it is expressed a little bit in the kidney cortex, and if-and-only-if you have coronary artery disease or peripheral circulation problems, then you’ll get expression in the affected tissues.

VHL^-/- renal cell carcinoma tumors can be completely suppressed by expressing wild-type VHL [Iliopoulos 1995] or by inhibiting HIF2a (couldn’t find the citation for this), even if the tumor has later acquired mutations in other tumor suppressors or activated other oncogenes. Dr. Iliopoulos’ lab has worked on small molecule HIF2a inhibitors [Zimmer 2008]. This sort of small molecule would probably be contraindicated for people with coronary artery disease since they need HIFa expression, but in healthy people, it is expected to be well tolerated. Dr. Iliopoulos has just helped to launch a Phase I clinical trial of an unrelated candidate drug, a glutaminase inhibitor, here at MGH [NCT02071862].

People with FH mutations often present first with a seemingly minor, cosmetic leiomyoma, and then get referred by dermatologists for germline genetic testing. Birt-Hogg-Dube is caused by mutations in FLCN which encodes folliculin. These patients often present first with fibrofolliculomas (lots of skin tags on the nose) and pneumothorax in the lung which leads to collapsed lung. The tumors in these patients may be of any histologic subtype.

Q&A

Q. Are there a lot of variants of unknown significance found in renal cell carcinoma patients?

A. Yes, but mostly in novel or recently discovered RCC genes. In VHL we almost exclusively see known mutations where pathogenicity is clearly established.

Q. Have there ever been patients with two hits in the germline?

A. A few cases have been reported. There was a family where the parents each had a different FH mutation, and one of their kids was compound heterozygous - she died in chidlhood. There’s another RCC gene where homozygous mutations cause mental retardation and death in early childhood before any tumors develop.

Journal club

The paper for this week was In vivo HIF-mediated reductive carboxylation is regulated by citrate levels and sensitizes VHL-deficient cells to glutamine deprivation [Gameiro 2013]. For additional background, see [Zimmer 2008, Scheuermann 2013, Masson & Ratcliffe 2014].