It is now well established that cancer is a genetic disease and that somatic mutations of oncogenes and tumour suppressor genes are the initiators of the carcinogenic process. The phosphatidylinositol 3-kinase signalling pathway has previously been implicated in tumorigenesis, and evidence over the past year suggests a pivotal role for the phosphatidylinositol 3-kinase catalytic subunit, PIK3CA, in human cancers. In this review, we analyse recent reports describing PIK3CA mutations in a variety of human malignancies, and discuss their possible implications for diagnosis and therapy.

The TP53 tumor suppressor gene is the most commonly mutated gene in human cancers. To evaluate the biological and clinical relevance of p53 loss, human somatic cell gene targeting was used to delete the TP53 gene in the non-tumorigenic epithelial cell line, MCF-10A. In all four p53−/− clones generated, cells acquired the capability for epidermal growth factor-independent growth and were defective in appropriate downstream signaling and cell cycle checkpoints in response to DNA damage. Interestingly, p53 loss induced chromosomal instability leading to features of transformation and the selection of clones with varying phenotypes. For example, p53-deficient clones were heterogeneous in their capacity for anchorage-independent growth and invasion. In addition, and of clinical importance, the cohort of p53-null clones showed sensitivity to chemotherapeutic interventions that varied depending not only on the type of chemotherapeutic agent, but also on the treatment schedule. In conclusion, deletion of the TP53 gene from MCF-10A cells eliminated p53 functions, as well as produced p53−/− clones with varying phenotypes possibly stemming from the distinct chromosomal changes observed. Such a model system will be useful to further understand the cancer-specific phenotypic changes that accompany p53 loss, as well as help to provide future treatment strategies for human malignancies that harbor aberrant p53.

Transforming growth factor-b type 1 (TGF-b) has been implicated as both a tumor suppressor and a tumor promoter in many solid epithelial cancers. We have previously demonstrated that the cyclin dependent kinase (CDK) inhibitor p21 acts as a molecular switch in determining a growth inhibitory versus growth proliferative response to TGF-b in the spontaneously immortalized human mammary epithelial cell line MCF-10A. We now demonstrate that this proliferative effect of TGF-b is mediated through the proinflammatory cytokine, interleukin-1a (IL-1a). Using gene expression array analysis, we identified IL-1a as a cytokine specifically upregulated only in cells lacking p21 and only upon TGF-b stimulation. Cell proliferation assays verified that recombinant IL-1a was capable of inducing a growth proliferative response in p21 null MCF-10A cells, while neutralizing antibodies against IL-1a prevented the growth proliferative effects of TGF-b. Mechanistically, both the CDK and proliferating cell nuclear antigen (PCNA) inhibitory functions of p21 were responsible for preventing TGF-b induced cell proliferation, but only PCNA inhibition by p21 regulated IL-1a gene expression. These studies demonstrate a novel role for IL-1a in mediating a proliferative response to TGF-b signaling, and suggest that therapies directed against IL-1a could abate the growth proliferative effects of TGF-b without compromising its tumor suppressive function.

The use of gene therapy to correct mutated or lost gene function for the treatment of human cancers has been an active, yet problematic area of biomedical research. Many technical difficulties, including efficient tissue-specific delivery, integration site specificity and general toxicity, are being addressed. Little is known, however, about the genetic and phenotypic stability that accompanies a successful gene-specific targeting event in a cancer cell. This question was addressed following the creation of a colon cancer cell line in which a mutated hMLH1 gene was corrected via targeted homologous recombination. This correction resulted in the expression of wild-type hMLH1 protein, restoration of the hPMS2 protein and mismatch repair (MMR) proficiency. One of two hMLH1-corrected clones, however, was found to retain defects in MMR activity. These cells continued to express the corrected hMLH1 protein, but had lost expression of another MMR protein, hMSH6. DNA sequence analysis of the hMSH6 gene revealed biallelic expansions of a cytosine repeat region in exon 5 that result in frameshifts leading to premature stop codons. These findings suggest that, similar to acquired drug resistance, the presence of genetically heterogeneous cancer cell populations or acquisition of compensatory mutations can result in 'resistance' to gene replacement therapy.