All solid cancers contain regions with very low cellular oxygen concentrations. These are known as hypoxic regions, and this hypoxia can either be acute or chronic. In acute hypoxia, temporary changes in oxygen flow lead to transient, exceptionally low oxygen concentrations; chronic, long-lasting hypoxia, which is not as severe, occurs in parts of the tumour that are located far from blood vessels. Hypoxia in cancer cells is associated with DNA damage and metastasis, and tumours with constant high levels of hypoxia have a poor prognosis.

Recent studies have shown that either severe or acute hypoxia can induce defects in the homologous recombination (HR) pathway, which is involved in the repair of double-stranded DNA breaks. This is the DNA repair pathway that is compromised in tumours that have mutations in the BRCA1/BRCA2 genes and that are sensitive to the class of investigational drugs that inhibit the enzyme poly (ADP-ribose) polymerase-1 (PARP). A group of researchers at Princess Margaret Hospital, Toronto, Canada, led by Robert Bristow and working with collaborators in Oxford, UK and Armilla, Spain, has now investigated the sensitivity of hypoxic cancer cells to PARP inhibition [1].

Hypoxia is known to decrease the expression of a number of genes involved in the HR pathway, including RAD51, RAD52, RAD54 and the BRCA genes. Bristow and his co-workers first explored the effect of hypoxia on gene expression in tumour cells by testing RAD51 expression in immortalized murine embryonic fibroblasts (MEFs) under conditions that mimic both the acute (16 hours at 0.02% O₂) and the chronic (72 hours at 0.2% O₂) hypoxic conditions found in tumours. Decreased RAD51 expression was observed under both hypoxic conditions, independent of the level of expression of the hypoxia inducible factor HIF-1a. Furthermore, MEF cells in which the PARP gene had been deleted were found to be unable to adapt to hypoxic conditions.

Bristow and his co-workers then tested proliferating cells from several human tumour types exposed to moderate chronic and severe acute hypoxia with PARP inhibitors including Abbott’s investigational compound, ABT-888. PARP inhibitors were found to be lethal to hypoxic cells, with cells under acute hypoxia being more severely affected. When the cell cycles of the tested cells were synchronised, they were found to be most sensitive in S-phase (while undergoing DNA replication). This fits in with the understanding that inhibiting PARP during DNA replication causes a build-up of the double-stranded DNA breaks that require the HR pathway for their repair.

Tumours with mutations in the BRCA genes, and thus with defects in the HR pathway, are exquisitely sensitive to PARP inhibitors. These results suggested that a similar HR-deficient phenotype was induced in hypoxic cells. To test this hypothesis further, mice bearing human tumour xenografts were treated with ABT-888 and then irradiated. Cells in hypoxic areas of the tumours were found to be selectively killed. Taken together, these results suggest that hypoxia induces a “BRCA-like” phenotype of synthetic lethality to PARP inhibition in tumours. So far, although PARP inhibitors have shown significant promise in clinical trials, their use has been limited to tumours with BRCA1 and BRCA2 mutations. Both the study authors and Teresa Villanueva of Nature Reviews Cancer [2] suggest that this induced sensitivity of hypoxic cells to PARP inhibitors may “expand the horizons” for these drugs in therapy to include otherwise resistant, hypoxic tumours.

References

[1] Chan, N., Pires, I.M., Bencokova, Z. and 9 others (2010). Contextual synthetic lethality of cancer cell kill based on the tumor microenvironment. Cancer Res. 70(20): 8045-54 DOI: 10.1158/0008-5472.CAN-10-2352

[2] Villanueva, T. (2010). Expanding the horizons of PARP inhibitors. Nature Rev. Drug Disc. 9(12): 919.