VHL localization depends on an intact microtubule network. VHL30 (green) co-localizes with the microtubule network (β-tubulin; red) in HeLa cells (top panel). Co-localization is shown in yellow. This localization of VHL30 is disrupted when microtubules are depolymerized with colcemid (bottom panel).

Inactivation of the von Hippel–Lindau ( VHL ) tumour-suppressor gene is linked to the development of several different tumour types in humans, including tumours of the kidney, retina, central nervous system and the adrenal gland. Despite identification of the gene almost 10 years ago, we are only now beginning to understand how VHL functions in the cell and how its mutation leads to tumour development. Reporting in the January issue of Nature Cell Biology, Krek and colleagues have now uncovered a novel function for VHL — microtubule stabilization — and show that disruption of this function is linked to the development of a specific subtype of VHL disease.

So far, the best-characterized function of VHL has been as a component of an E3 ubiquitin ligase complex, which mediates degradation of the hypoxia-inducible factor (HIF) under normoxic conditions. Although there are some links between this and particular subtypes of VHL disease, the extent to which this function of VHL contributes to tumour progression is unclear. More recently, VHL has also been implicated in extracellular-matrix formation and cell-cycle progression.

The importance of the new work from Krek and colleagues is that they identify a novel cytoskeletal function, which is specific to an isoform of VHL that has not previously been linked to VHL tumour development. VHL exists as two isoforms: the longer VHL30 isoform and the shorter VHL19 isoform that results from internal initiation at methionine 54. By raising antibodies that are specific for each isoform, Krek and colleagues revealed that, whereas the shorter isoform localizes predominantly to the nucleus, the longer VHL30 isoform co-localizes with the cytoplasmic microtubule network and depends on an intact microtubule network for this.

To address the functional significance of this localization, the authors tested the effect of VHL binding on microtubule dynamics and found that it mediates microtubule stabilization, protecting microtubules against nocodazole treatment. One important distinction is that this function of VHL seems to be independent of its ability to form an active E3 ligase complex. So what is the relevance, if any, of this role to the tumour-suppressor function of VHL? To test this, the authors looked at the effects of different VHL mutations associated with each disease subtype on microtubule stabilization. Intriguingly, only mutations associated with type 2A VHL disease (and one associated with type 2C disease), which is characterized by a high risk of developing adrenal-gland tumours and cerebellar haemangioblastomas, were abrogated in microtubule stabilization.

How this function of VHL might contribute to tumour development remains to be seen, but from this work a new model for VHL function is beginning to emerge. The authors propose that each of the two VHL isoforms has a distinct role. Whereas the shorter isoform resides in the nucleus and is required as part of an E3 ligase complex to regulate HIF under normoxic conditions, the longer isoform has a novel E3-independent function in the cytoplasm, mediating microtubule stability. Exactly how loss of microtubule stabilization by VHL contributes to tumour progression remains to be seen.