| Ionizing radiation (IR) therapy is an important tool for the clinical treatment of cancer. Unfortunately, two to five percent of patients are hypersensitive to radiation and suffer severe adverse reactions including skin reactions, osteoradionecrosis and death.; IR creates DNA damage, including potentially cytotoxic DNA double strand breaks, which activates repair and signaling pathways in normal cells. We hypothesized that germline mutations in known IR response genes, including ATM, BRCA1 and BRCA2, could cause radiation hypersensitivity. To provide sensitive and accurate mutation detection, we employed a new technology, denaturing HPLC (DHPLC). Analysis revealed two radiosensitive patients with heterozygous mutations, one in BRCA2 and one in ATM. No mutations have been found in control patients who responded normally to radiotherapy. We conclude that at most, a small percentage of radiation reactions can be attributed to mutations in these genes.; Microarrays provide a complementary approach for analysis of the IR response in cancer patients, by measuring expression of thousands of genes simultaneously. Specific tools are required to mine the data for biologically relevant information. We developed Significance Analysis of Microarrays (SAM), a new tool for microarray data analysis. It allows identification of genes with statistically significant changes in gene expression and provides a measure of confidence for results. SAM can be applied to a wide variety of experimental formats.; Here, SAM is used to describe the transcriptional response to IR in lymphoblastoid cells. The wild-type response includes repression of genes involved in proliferation, and activation of death ligands and receptors, nucleotide excision repair genes, and genes involved in cell cycle arrest.; The transcriptional response to radiation is also examined in lymphoblastoid cell lines established from radiation sensitive and control patients, and in characterized mutant cell lines. Radiation sensitive patients display abnormal transcriptional responses, including a general blunting of IR-induced changes in gene expression. SAM can identify 86 genes which respond abnormally to IR in radiation sensitive patients. When this subset of genes is used to calculate the Pearson correlation between each sample and the average IR response in normal cells, almost all radiation sensitive patients can be differentiated from control patients. These results provide hope for development of a clinical test to predict adverse radiation reactions. |
