• M.O.

HIF-2: A Novel Tumor Suppressor?

Imamura et al., pp. 762–770

As tumors develop, they quickly outgrow their blood supply, leaving regions of the tumors starved of oxygen. These hypoxic conditions trigger powerful adaptation mechanisms, including stimulation of new blood vessel growth. Two transcription factors play a key role in this process: hypoxia-inducible factors (HIF) 1 and 2. Both factors accumulate in hypoxic tumor cells and recognize identical DNA sequences in target gene promoters; however, their functions appear to be fundamentally different.

Imamura and colleagues selectively knocked down the regulatory alpha subunits of HIF-1 and 2 in SW480 colon cancer cells responsive to hypoxia. Loss of HIF-1α slowed cell proliferation and migration in vitro. After transplantation as xenografts in vivo, the tumor size was reduced. In contrast, knockdown of HIF-2α had no effect on in vitro growth, but xenografts deficient in HIF-2α formed tumors double the size of those from wild-type cells. This revealed an unexpected ‘tumor-suppressive’ function of HIF-2. Interestingly, the tumor-promoting properties of HIF-1 appeared dominant over HIF-2 protection since the phenotype in double-knockdown cells resembled the HIF-1 knockdown.

No difference in vascularization was observed between HIF-1 and -2 knockdown tumors; however, in gene expression profiling studies the authors identified novel target genes exclusively regulated by HIF-2 (cyclin G2, angiopoietin-like 4). The authors caution that HIF inhibitors developed to block tumor angiogenesis could predominantly inhibit HIF-2 and activate tumor growth instead of shrinkage.

Illustration 1.

Size of SW480 xenografts using wild-type (WT), HIF-1α knockdown (1K), HIF-2α knockdown (2K) or HIF1α/2α knockdown (1K2K) cells.

Lost in Translation: Exon Skipping Stops HER2 in Breast Cancer Cells

Wan et al., pp. 771–776

Human epidermal growth factor receptor 2 (HER2) is a cell membrane-bound receptor tyrosine kinase and an important target for cancer therapies. Approximately 25% of patients express abnormally high numbers of HER2 molecules on the surface of breast cancer cells. HER2 serves as the favorite dimerization partner for other HER family members, such as HER3, involved in signal transduction pathways controlling cell growth and differentiation. HER2+ cancers are considered highly aggressive, but are responsive to HER2-targeting therapeutics such as trastuzumab (Herceptin). In addition, natural splice variants of HER2 have been identified that function as soluble dominant negative inhibitors of HER2 and are regarded as promising anti-cancer drug candidates (Herstatin).

In this issue of IJC, Wan and colleagues employ an emerging new technology to experimentally engineer a new splice form of HER2 mRNAs. They target the 5' splice site of exon 15 in HER2 with specifically designed antisense oligonucleotides to induce skipping of exon 15 and premature termination of protein translation. Indeed, full-length membrane-bound HER2 expression is lost after treatment with antisense oligonucleotides while a soluble truncated Δexon15 protein is being produced. As a consequence, growth suppression and induction of apoptosis is observed in HER2-dependent SK-BR-3 cells but not in HER2-independent MCF7 cells. Treatment of SK-BR-3 cells with a synthetic Δexon15 HER2 protein also causes downregulation of HER2 expression and decreased HER3 transphosphorylation, mimicking the effects of the natural inhibitor Herstatin. The authors recommend that both technologies, splice switching oligonucleotides as well as recombinant Δexon15 proteins, be further explored as potential HER2-targeting therapeutics.

HPV Genotypes Predict Outcome in Women with Cervical Carcinoma

Cremoux et al., pp. 777–781

A new analysis of 515 cases of invasive cancer of the uterine cervix underlines the prognostic value of genotyping of human papillomaviruses (HPV). It is well established that HPVs are causative agents of cervical cancer; however, the influence of infections with different HPV types on the outcome of the disease remains controversial. Cremoux and colleagues performed a retrospective analysis of women afflicted with cervical cancer treated at the Institut Curie in Paris, France. They identified only 4% of cases in which no known HPV type could be detected. In 30% of cases, HPV types other than high-risk HPV16/18 types were found. These tumors developed preferentially in older women and were associated with better disease outcome than patients infected with high-risk types. These numbers are noteworthy since current vaccination strategies target mainly high-risk HPV types.

Tumors associated with intermediate HPV types (31, 33, 35, 39, 52, 53, 58, 59, 73) had a better prognostic value than tumors associated with high-risk types. However, no overall difference in the prognostic value of HPV18 versus 16 in long-term outcome was observed. HPV18-associated tumors preferentially relapsed early while late relapses were more frequent in HPV16-associated disease. Tumors associated with HPV16/18 developed on average 5.3 and 7.8 years earlier than intermediate type HPV-associated disease, supporting the model that early lesions associated with high-risk HPVs evolve more rapidly towards invasive carcinoma than those associated with other HPVs.