AutoCSA, an algorithm for high throughput DNA sequence variant detection in cancer genomes.

UNLABELLED: The undertaking of large-scale DNA sequencing screens for somatic variants in human cancers requires accurate and rapid processing of traces for variants. Due to their often aneuploid nature and admixed normal tissue, heterozygous variants found in primary cancers are often subtle and difficult to detect. To address these issues, we have developed a mutation detection algorithm, AutoCSA, specifically optimized for the high throughput screening of cancer samples. AVAILABILITY: http://www.sanger.ac.uk/genetics/CGP/Software/AutoCSA.

Mutations in the RSK2(RPS6KA3) gene cause Coffin-Lowry syndrome and nonsyndromic X-linked mental retardation.

We describe three families with X-linked mental retardation, two with a deletion of a single amino acid and one with a missense mutation in the proximal domain of the RSK2(RPS6KA3) (ribosomal protein S6 kinase, 90 kDa, polypeptide 3) protein similar to mutations found in Coffin-Lowry syndrome (CLS). In two families, the clinical diagnosis had been nonsyndromic X-linked mental retardation. In the third family, although CLS had been suspected, the clinical features were atypical and the degree of intellectual disability much less than expected. These families show that strict reliance on classical clinical criteria for mutation testing may result in a missed diagnosis. A less targeted screening approach to mutation testing is advocated.

COSMIC 2005.

The Catalogue Of Somatic Mutations In Cancer (COSMIC) database and web site was developed to preserve somatic mutation data and share it with the community. Over the past 25 years, approximately 350 cancer genes have been identified, of which 311 are somatically mutated. COSMIC has been expanded and now holds data previously reported in the scientific literature for 28 known cancer genes. In addition, there is data from the systematic sequencing of 518 protein kinase genes. The total gene count in COSMIC stands at 538; 25 have a mutation frequency above 5% in one or more tumour type, no mutations were found in 333 genes and 180 are rarely mutated with frequencies <5% in any tumour set. The COSMIC web site has been expanded to give more views and summaries of the data and provide faster query routes and downloads. In addition, there is a new section describing mutations found through a screen of known cancer genes in 728 cancer cell lines including the NCI-60 set of cancer cell lines.

Somatic mutations in human cancer: insights from resequencing the protein kinase gene family.

All cancers arise due to the accumulation of mutations in critical target genes that, when altered, give rise to selective advantage in the cell and its progeny that harbor them. Knowledge of these mutations is key in understanding the biology of cancer initiation and progression, as well as the development of more targeted therapeutic strategies. We have undertaken a systematic screen of all annotated protein kinases in the human genome for mutations in a series of cancers including breast, non-small-cell lung, and testicular cancer. Our results show a wide diversity in mutation prevalence within and between tumor types. We have identified a mutator phenotype in human breast previously undescribed. The results presented from sequencing the same 1.3 million base pairs through several tumor types suggest that most of the observed mutations are likely to be passenger events rather than causally implicated in oncogenesis. However, this work does provide evidence for the likely existence of multiple, infrequently mutated kinases.

The COSMIC (Catalogue of Somatic Mutations in Cancer) database and website.

The discovery of mutations in cancer genes has advanced our understanding of cancer. These results are dispersed across the scientific literature and with the availability of the human genome sequence will continue to accrue. The COSMIC (Catalogue of Somatic Mutations in Cancer) database and website have been developed to store somatic mutation data in a single location and display the data and other information related to human cancer. To populate this resource, data has currently been extracted from reports in the scientific literature for somatic mutations in four genes, BRAF, HRAS, KRAS2 and NRAS. At present, the database holds information on 66 634 samples and reports a total of 10 647 mutations. Through the web pages, these data can be queried, displayed as figures or tables and exported in a number of formats. COSMIC is an ongoing project that will continue to curate somatic mutation data and release it through the website.

Molecular classification of synovial sarcomas, leiomyosarcomas and malignant fibrous histiocytomas by gene expression profiling.

In this study, we have used genome-wide expression profiling to categorise synovial sarcomas, leiomyosarcomas and malignant fibrous histiocytomas (MFHs). Following hierarchical clustering analysis of the expression data, the best match between tumour clusters and conventional diagnosis was observed for synovial sarcomas. Eight of nine synovial sarcomas examined formed a cluster that was characterised by higher expression of a set of 48 genes. In contrast, sarcomas conventionally classified as leiomyosarcomas and MFHs did not match the clusters defined by hierarchical clustering analysis. One major cluster contained a mixture of both leiomyosarcomas and MFHs and was defined by the lower expression of a set of 202 genes. A cluster containing a subgroup of MFHs was also detected. These results may have implications for the classification of soft tissue sarcomas, and are consistent with the view that sarcomas conventionally defined as MFHs do not represent a separate diagnostic category.

Gene expression microarray analysis in cancer biology, pharmacology, and drug development: progress and potential.

With the imminent completion of the Human Genome Project, biomedical research is being revolutionised by the ability to carry out investigations on a genome wide scale. This is particularly important in cancer, a disease that is caused by accumulating abnormalities in the sequence and expression of a number of critical genes. Gene expression microarray technology is gaining increasingly widespread use as a means to determine the expression of potentially all human genes at the level of messenger RNA. In this commentary, we review developments in gene expression microarray technology and illustrate the progress and potential of the methodology in cancer biology, pharmacology, and drug development. Important applications include: (a) development of a more global understanding of the gene expression abnormalities that contribute to malignant progression; (b) discovery of new diagnostic and prognostic indicators and biomarkers of therapeutic response; (c) identification and validation of new molecular targets for drug development; (d) provision of an improved understanding of the molecular mode of action during lead identification and optimisation, including structure-activity relationships for on-target versus off-target effects; (e) prediction of potential side-effects during preclinical development and toxicology studies; (f) confirmation of a molecular mode of action during hypothesis-testing clinical trials; (g) identification of genes involved in conferring drug sensitivity and resistance; and (h) prediction of patients most likely to benefit from the drug and use in general pharmacogenomic studies. As a result of further technological improvements and decreasing costs, the use of microarrays will become an essential and potentially routine tool for cancer and biomedical research.

Breast cancer genetics: what we know and what we need.

Breast cancer results from genetic and environmental factors leading to the accumulation of mutations in essential genes. Genetic predisposition may have a strong, almost singular effect, as with BRCA1 and BRCA2, or may represent the cumulative effects of multiple low-penetrance susceptibility alleles. Here we review high- and low-penetrance breast-cancer-susceptibility alleles and discuss ongoing efforts to identify additional susceptibility genes. Ultimately these discoveries will lead to individualized breast cancer risk assessment and a reduction in breast cancer incidence.

Cancer and genomics.

Identification of the genes that cause oncogenesis is a central aim of cancer research. We searched the proteins predicted from the draft human genome sequence for paralogues of known tumour suppressor genes, but no novel genes were identified. We then assessed whether it was possible to search directly for oncogenic sequence changes in cancer cells by comparing cancer genome sequences against the draft genome. Apparently chimaeric transcripts (from oncogenic fusion genes generated by chromosomal translocations, the ends of which mapped to different genomic locations) were detected to the same degree in both normal and neoplastic tissues, indicating a significant level of false positives. Our experiment underscores the limited amount and variable quality of DNA sequence from cancer cells that is currently available.

Rapid detection of DNA sequence variants by conformation-sensitive capillary electrophoresis.

The identification of novel sequence variants, which may be either disease-causing mutations or silent polymorphisms, in large numbers of samples is becoming the rate-limiting step in associating diseases with specific genes. This is particularly true in light of the imminent arrival of the complete reference sequence of the human genome. A number of techniques have been developed to analyze DNA samples for sequence variants rapidly. We describe a new technique, capillary-based conformation-sensitive gel electrophoresis (capillary CSGE) that transfers mutation detection from acrylamide gel to capillary electrophoresis. Capillary CSGE was able to detect 7/7 short insertion/deletions and 16/22 base substitutions in a series of random single-nucleotide polymorphisms and known variants in the lipoprotein lipase and BRCA2 genes. This technique has the potential to screen many megabases of DNA in a single day.

Cancer classification with DNA microarrays is less more?

The dissection of cancer and the underlying molecular processes that are defective in cancer cells has become an important tool in the fight against this disease. DNA microarrays can provide detailed information of the expression pattern of thousands of genes in tumours. But how much of this data is useful and is some superfluous? Can array data be used to identify a handful of critical genes that will lead to a more-detailed taxonomy of tumours and can this or similar array data be used to predict clinical outcome? Primary tumours will give us the statistical power to draw these conclusions, but can cancer cell lines be used as models to point us in the right direction?

Sequence variants of the axin gene in breast, colon, and other cancers: an analysis of mutations that interfere with GSK3 binding.

Axin is a recently discovered component of a multiprotein complex containing APC, beta-catenin, GSK3, and PP2A, which functions in the degradation of the beta-catenin protein. As part of WNT signal transduction, the function of the Axin complex is inhibited, leading to the accumulation of beta-catenin. The inappropriate stabilization of beta-catenin has been implicated in a range of human tumors. Two oncogenic mechanisms leading to beta-catenin stabilization are the loss of the APC tumor suppressor protein and the mutational activation of beta-catenin, such that the Axin/APC complex can no longer regulate it. Studies in Drosophila and mammalian tissue culture showed loss of Axin function interfered with beta-catenin turnover and activated beta-catenin/TCF-dependent transcription. Based on these observations, Axin was screened for mutations in a range of human tumor cell lines and primary breast tumor samples. We identified two sequence variants causing amino acid substitutions in four colon cancer cell lines, a Ser-to-Leu at residue 215 in LS513 and a Leu-to-Met at residue 396 in HCT-8, HCT-15, and DLD-1. The Axin L396M mutation was selected for further study since it lay within a region that was shown to interact with glycogen synthase kinase-3. Biochemical and functional studies showed that the L396M change interfered with Axin's ability to bind GSK3. Interestingly, this mutation and a neighboring L392M change differentially altered Axin's ability to interfere with two upstream activators of TCF-dependent transcription, Frat1 and Disheveled.

Heterozygosity for mutations in the ataxia telangiectasia gene is not a major cause of radiotherapy complications in breast cancer patients.

Of patients being treated by radiotherapy for cancer, a small proportion develop marked long-term radiation damage. It is believed that this is due, at least in part, to intrinsic individual differences in radiosensitivity, but the underlying mechanism is unknown. Individuals affected by the recessive disease ataxia telangiectasia (AT) exhibit extreme sensitivity to ionizing radiation. Cells from such individuals are also radiosensitive in in vitro assays, and cells from AT heterozygotes are reported to show in vitro radiosensitivity at an intermediate level between homozygotes and control subjects. In order to examine the possibility that a defect in the ATM gene may account for a proportion of radiotherapy complications, 41 breast cancer patients developing marked changes in breast appearance after radiotherapy and 39 control subjects who showed no clinically detectable reaction after radiotherapy were screened for mutations in the ATM gene. One out of 41 cases showing adverse reactions was heterozygous for a mutation (insertion A at NT 898) that is predicted to generate a truncated protein of 251 amino acids. No truncating mutations were detected in the control subjects. On the basis of this result, the estimated percentage (95% confidence interval) of AT heterozygous patients in radiosensitive cases was 2.4% (0.1-12.9%) and in control subjects (0-9.0%). We conclude that ATM gene defects are not the major cause of radiotherapy complications in women with breast cancer.

Physical mapping of the CA6, ENO1, and SLC2A5 (GLUT5) genes and reassignment of SLC2A5 to 1p36.2.

Several human malignancies frequently exhibit deletions or rearrangements of the distal short arm of chromosome 1 (1p36), and a number of genetic diseases also map to this region. The carbonic anhydrase (CA6) and alpha-enolase (ENO1) genes, previously mapped to 1p36, were physically linked in yeast- and P1-artificial chromosome (YAC and PAC) contigs. PACs from the contig were mapped to 1p36.2 by fluorescence in situ hybridization. The ESTs D1S2068, D1S274E, D1S3275, and stSG4370 were also placed in the same contig. The physical map was integrated with the genetic map of chromosome 1 by assignment of genetic markers D1S160, D1S1615, and D1S503 to the contig. Sequencing of the EST clone representing D1S274E indicated that it was derived from the same transcript as D1S2068E and corresponded to the SLC2A5 (GLUT5) gene, previously assigned to 1p31. Reassignment of SLC2A5 to 1p36.2 was confirmed by somatic cell and radiation hybrid mapping panels and was consistent with previous EST mapping data. Sequencing of the EST clone for D1S274E revealed the presence of intronic sequences, suggesting that the clone was derived from an unprocessed message. The presence of unprocessed and/or alternatively spliced EST clones has potential ramifications for EST-based genomic projects. This information should facilitate the mapping of tumor suppressor and genetic disease loci that have been localized to this region.

Cancer risks in two large breast cancer families linked to BRCA2 on chromosome 13q12-13.

The penetrance of the BRCA2 gene on chromosome 13q12-13 has been estimated in two large, systematically ascertained, linked families, by use of a maximum-likelihood method to incorporate both cancer-incidence data and 13q marker typings in the families. The cumulative risk of breast cancer in female gene carriers was estimated to be 59.8% by age 50 years (95% confidence interval [95% CI] 25.9%-78.5%) and 79.5% by age 70 years (95% CI 28.9%-97.5%). The cumulative risk of breast cancer in male carriers was estimated to be 6.3% (95% CI 1.4%-25.6%) by age 70 years. There was no evidence of any risk difference between the two families. These results indicate that the lifetime breast cancer risk in BRCA2 carriers, for at least a subset of mutations, is comparable to that for BRCA1. A significant excess of ovarian cancer in gene carriers was observed (relative risk 17.69, based on three cases), but the absolute risk of ovarian cancer was less than that reported for BRCA1. Significant excesses of laryngeal cancer (relative risk 7.67, based on two possible carriers) and prostate cancer (relative risk 2.89, based on five possible carriers) were also observed. One case of ocular melanoma, as well as a second eye cancer of unspecified histology, occurred in obligate gene carriers.

Cloning, chromosomal mapping and expression pattern of the mouse Brca2 gene.

A proportion of human breast cancers result from an inherited predisposition to the disease. Mutations in the BRCA2 gene confer a high risk of breast cancer and are responsible for almost half of these cases. The recent cloning of the human BRCA2 gene has revealed that it encodes a large protein having little significant homology to known proteins. Here we describe the mouse Brca2 gene. The gene maps to mouse chromosome 5, consistent with its location on human chromosome 13q12. We have sequenced cDNA for the entire 3329 amino acid Brca2 protein and this has revealed that, like Brca1, Brca2 is relatively poorly conserved between humans and mice. Brca2 is transcribed in a diverse range of mouse tissues, and the pattern of expression is strikingly similar to that of Brca1. Taken together, our data highlight some intriguing similarities between two genes involved in inherited breast cancer susceptibility.

Screening for ESR mutations in breast and ovarian cancer patients.

In the present study, leukocyte DNA from 143 patients with familial clustering of breast and/or ovarian cancer and tumour DNA from 96 breast carcinomas were screened for base mutations in the estrogen receptor gene (ESR). Three patients with a family history of cancer were carrying a Gly160Cys germline substitution. This alteration was also detected in eight (four females and four males) of 729 controls (366 female, 363 males), indicating that the substitution probably represents a polymorphism. However, in the 229 female controls in whom family history of cancer was known, one of two who had a sister with breast cancer was carrying the variant allele. Hence, a possible clinical significance of the glycine into cysteine cannot be completely ruled out and should be further investigated. Somatic mutations were not detected in any of the tumours studied, and the present data do not provide support for somatic ESR base mutations as an important mechanism for hormonal therapy resistance in estrogen receptor-positive breast carcinomas.

The BRC repeats are conserved in mammalian BRCA2 proteins.

The breast cancer susceptibility gene BRCA2 encodes a protein of 3418 amino acids which does not exhibit substantial sequence similarity to any other protein in the public databases. A dot matrix comparison of BRCA2 with itself revealed an eight times repeated motif in the segment of the protein encoded by exon 11. As a preliminary test of the hypothesis that these motifs are functionally significant, we have sequenced exon 11 of BRCA2 in six mammals. An alignment of the predicted protein sequences shows that, overall, the motifs have been conserved while much of the intervening sequences has diverged. These data support the notion that the BRC motifs are important in BRCA2 function. There is, however, considerable interspecies variation within certain motif units, raising the possibility of redundancy and that not all of the repeats are required for the normal function of BRCA2.