Survivin has been a controversial figure in cancer research over the past few years. It has been implicated in the control of cell division and cell death, but where does its function really lie? Survivin is also specifically upregulated in cancer cells, so could it be a useful drug target? Two papers from Dario Altieri's group, in the October issue of The Journal of Clinical Investigation, address some of these questions in vivo. They show that survivin can suppress apoptosis and that a survivin mutant might prove effective in cancer gene therapy.

In the first paper, Grossman et al. generated a transgenic mouse that selectively expressed survivin in the skin. Interestingly, survivin had no effect on the normal skin differentiation programme. It also did not enhance cell proliferation induced by UVB irradiation, indicating that survivin does not affect cell division. So, does survivin affect apoptosis? Exposure of the transgenic and wild-type mice to UVB irradiation, and the quantitation of apoptotic or 'sunburn' keratinocytes revealed that there was a 60% reduction in apoptosis when survivin was expressed. Survivin can also cooperate with loss of a Trp53 allele to further reduce apoptosis. In fact, survivin selectively inhibits the intrinsic, caspase-9-dependent apoptotic pathway, as keratinocytes isolated from the mice that express survivin are still susceptible to Fas-mediated apoptosis, which acts through the extrinsic, caspase-8-dependent apoptotic pathway.

With the role of survivin in suppressing apoptosis confirmed, the possibility of manipulating survivin in cancer cells — to increase apoptosis — now becomes a reality. In the second paper they achieve just that. Mesri et al. constructed a replication-deficient adenovirus that encoded a survivin mutant (pAd-T34A): the threonine residue of a cdc2 phosphorylation site is mutated to alanine, and the resulting mutant is thought to act as a dominant-negative by binding to cdc2 to prevent phosphorylation of endogenous survivin.

Infection of several cancer cell lines with pAd-T34A resulted in at least a 2–3-fold increase in apoptosis — breast carcinoma MCF-7 cells suffer a 7-fold increase — as visualized by DAPI staining of the nucleus and quantitiated by flow cytometry. pAd-T34A induces cytochrome c release, processing of caspase-9 and cleavage of caspase-3 — which corresponds with an increase in its activity — supporting the notion that survivin regulates the intrinsic apoptotic pathway.

When combined with chemotherapeutics, pAd-T34A enhanced the level of apoptosis induced by taxol, but adriamycin and pAd-T34A did not induce more apoptosis than pAd-T34A alone.

So, could pAd-T34A work in vivo to inhibit tumour cell growth? Immunodeficient mice were injected with MCF-7 cells that had previously been infected with either pAd-T34A or the vector alone. Only those that expressed pAd-T34A suppressed tumour growth — mitotic index was inhibited by 90% in these mice. A single injection of pAd-T34A also suppressed the growth of established tumours by 40%, and TUNEL staining of DNA fragmentation confirmed that this was due, at least in part, to apoptosis.

The debate over whether survivin is important for survival or cell division in higher organisms seems to be edging towards survival, but perhaps the most exciting prospects for cancer research are that, regardless of its function, survivin can be turned upon itself to block tumour growth.