Tumour suppressors normally inhibit cell proliferation, but one seems to act in the opposite manner. Chun-Ming Chen and Richard Behringer report in the February issue of Genes & Development that it is the loss of OVCA1 that causes proliferation defects — only when its loss is combined with that of p53 is its tumour-suppressive activity unleashed.

The genes that cause ovarian cancer are largely unknown, so the identification of a region of chromosome 17 that is frequently lost was an important find. The OVCA1 gene has since been shown to reside at that location, but it has yet to be shown to act as a tumour suppressor. The authors had previously identified the mouse orthologue, and continued their studies to understand the function of Ovca1.

They generated knockout mice, but these either died during development or shortly after birth. The embryos showed developmental defects and were generally smaller. Transplanting the ovaries into kidney capsules of wild-type mice allowed their development to be followed over time, and although smaller, development seemed normal and there was no evidence of carcinomas.

The growth defect could be caused by a decrease in proliferation or an increase in apoptosis, so Ovca1-null mouse embryonic fibroblasts (MEFs) were established to answer this question. The MEFs grew poorly, but there was no sub-G1 population, which would be indicative of apoptosis. Cell-cycle analysis revealed that fewer cells were in S phase, and this corresponded with a decrease in phosphorylated Rb, which would explain the inability of the cells to enter S phase.

But this is an unusual property for loss of a tumour suppressor, so might elimination of a checkpoint allow the cells to recover from this proliferation defect? The authors found that Ovca1−/−Trp53+/− MEFs also had a reduced S-phase population, but Ovca1−/−Trp53−/− MEFs grew normally. Loss of p53 could not rescue the developmental defects though — the mice still died soon after birth.

The next important question was whether Ovca1 actually did act as a tumour suppressor in vivo. Almost 60% of Ovca1+/− mice developed a range of tumours by two years, with an average latency of 92 weeks. This was shortened to 52 weeks in Ovca1+/−Trp53+/− mice, and the incidence was increased to 72%. Importantly, the loss of one Ovca1 allele increases the tumour incidence when compared with Trp53+/− mice, and although the tumour incidence in Ovca1+/−Trp53−/− and Trp53−/− mice is the same, more mice have numerous tumours when they have lost one Ovca1 allele. The tumour spectrum is also somewhat different in the Ovca1 heterozygous mice.

So, despite its role as a positive regulator of cell-cycle progression, it seems that OVCA1 is, indeed, a tumour suppressor. Further work is needed to understand exactly how it functions in normal cells, and how its loss accelerates tumour progression when p53 is also lost.