Bases adjacent to mononucleotide repeats show an increased single nucleotide polymorphism frequency in the human genome.

Mononucleotide repeats (MNRs) are abundant in eukaryotic genomes and exhibit a high degree of length variability due to insertion and deletion events. However, the relationship between these repeats and mutation rates in surrounding sequences has not been systematically investigated. We have analyzed the frequency of single nucleotide polymorphisms (SNPs) at positions close to and within MNRs in the human genome. Overall, we find a 2- to 4-fold increase in the SNP frequency at positions immediately adjacent to the boundaries of MNRs, relative to that at more distant bases. This relationship exhibits a strong asymmetry between 3' and 5' ends of repeat tracts and is dependent upon the repeat motif, length and orientation of surrounding repeats. Our analysis suggests that the incorporation or exclusion of bases adjacent to the boundary of the repeat through substitutions, in which these nucleotides mutate towards or away from the base present within the repeat, respectively, may be another mechanism by which MNRs expand and contract in the human genome.

An expectation-maximization algorithm for the analysis of allelic expression imbalance.

A significant proportion of the variation between individuals in gene expression levels is genetic, and it is likely that these differences correlate with phenotypic differences or with risk of disease. Cis-acting polymorphisms are important in determining interindividual differences in gene expression that lead to allelic expression imbalance, which is the unequal expression of homologous alleles in individuals heterozygous for such a polymorphism. This expression imbalance can be detected using a transcribed polymorphism, and, once it is established, the next step is to identify the polymorphisms that are responsible for or predictive of allelic expression levels. We present an expectation-maximization algorithm for such analyses, providing a formal statistical framework to test whether a candidate polymorphism is associated with allelic expression differences.

Overexpression of aurora B kinase (AURKB) in primary non-small cell lung carcinoma is frequent, generally driven from one allele, and correlates with the level of genetic instability.

Aurora kinases are key regulators of chromosome segregation during mitosis. We have previously shown by microarray analysis of primary lung carcinomas and matched normal tissue that AURKB (22 out of 37) and AURKA (15 out of 37) transcripts are frequently over-represented in these tumours. We now confirm these observations in a second series of 44 carcinomas and also show that aurora B kinase protein levels are raised in the tumours compared to normal tissue. Elevated levels of expression in tumours are not a consequence of high-level amplification of the AURKB gene. Using a coding sequence polymorphism we show that in most cases (seven out of nine) tumour expression is predominantly driven from one AURKB allele. Given the function of aurora B kinase, we examined whether there was an association between expression levels and genetic instability. We defined two groups of high and low AURKB expression. Using a panel of 10 microsatellite markers, we found that the group showing the higher level of expression had a higher frequency of allelic imbalance (P=0.0012). Analysis of a number of other genes that are strongly and specifically expressed in tumour over normal lung, including SERPINB5, TERT and PRAME, showed marked allelic expression imbalances in the tumour tissue in the context of balanced or only marginally imbalanced relative allelic copy numbers. Our data support a model of early carcinogenesis wherein defects in the process of inactivation of lung stem-cell associated genes during differentiation, contributes to the development of carcinogenesis.

The alleles of the DNA repair gene O6-alkylguanine-DNA alkyltransferase are expressed at different levels in normal human lung tissue.

O6-Alkylguanine-DNA alkyltransferase (MGMT) confers resistance to many of the mutagenic and toxic effects of certain classes of alkylating agents by repairing the DNA lesions responsible. The levels of expression of this protein are of interest in relation to the prevention and treatment of cancer in man. They vary widely between individuals, and the basis of this variation is not understood. RT-PCR-RFLP analysis of mRNA from normal human lung tissue reveals that the two MGMT alleles are frequently expressed at different levels, indicating that there is a genetic component to inter-individual variation of MGMT levels and that at least some of this variation maps close to or within the MGMT locus.

A phylogenetic approach to assessing the significance of missense mutations in disease genes.

The identification of deleterious mutations within candidate genes is a crucial step in the elucidation of the genetic bases of human disease. However, the significance of any base or amino acid change within a gene is unknown until detailed structural and functional analysis has been carried out. A potentially rapid way of identifying functionally important sites within a gene is to identify evolutionarily conserved regions. Mutations affecting such sites are assumed to be deleterious for the carrier. In this communication we generalize this approach and present a formal framework to assess whether a specific mutation is deleterious given sequence data from a set of homologues. We propose a score that takes into account the nature of the mutation, the conservation of the affected residue among the different species, and their phylogenetic relationships. Its performance is examined using published TP53 mutations and frequent polymorphic variants.

The comparative genomics of polyglutamine repeats: extreme differences in the codon organization of repeat-encoding regions between mammals and Drosophila.

Polyglutamine repeats within proteins are common in eukaryotes and are associated with neurological diseases in humans. Many are encoded by tandem repeats of the codon CAG that are likely to mutate primarily by replication slippage. However, a recent study in the yeast Saccharomyces cerevisiae has indicated that many others are encoded by mixtures of CAG and CAA which are less likely to undergo slippage. Here we attempt to estimate the proportions of polyglutamine repeats encoded by slippage-prone structures in species currently the subject of genome sequencing projects. We find a general excess over random expectation of polyglutamine repeats encoded by tandem repeats of codons. We nevertheless find many repeats encoded by nontandem codon structures. Mammals and Drosophila display extreme opposite patterns. Drosophila contains many proteins with polyglutamine tracts but these are generally encoded by interrupted structures. These structures may have been selected to be resistant to slippage. In contrast, mammals (humans and mice) have a high proportion of proteins in which repeats are encoded by tandem codon structures. In humans, these include most of the triplet expansion disease genes.

A role for selection in regulating the evolutionary emergence of disease-causing and other coding CAG repeats in humans and mice.

The evolutionary expansion of CAG repeats in human triplet expansion disease genes is intriguing because of their deleterious phenotype. In the past, this expansion has been suggested to reflect a broad genomewide expansion of repeats, which would imply that mutational and evolutionary processes acting on repeats differ between species. Here, we tested this hypothesis by analyzing repeat- and flanking-sequence evolution in 28 repeat-containing genes that had been sequenced in humans and mice and by considering overall lengths and distributions of CAG repeats in the two species. We found no evidence that these repeats were longer in humans than in mice. We also found no evidence for preferential accumulation of CAG repeats in the human genome relative to mice from an analysis of the lengths of repeats identified in sequence databases. We then investigated whether sequence properties, such as base and amino acid composition and base substitution rates, showed any relationship to repeat evolution. We found that repeat-containing genes were enriched in certain amino acids, presumably as the result of selection, but that this did not reflect underlying biases in base composition. We also found that regions near repeats showed higher nonsynonymous substitution rates than the remainder of the gene and lower nonsynonymous rates in genes that contained a repeat in both the human and the mouse. Higher rates of nonsynonymous mutation in the neighborhood of repeats presumably reflect weaker purifying selection acting in these regions of the proteins, while the very low rate of nonsynonymous mutation in proteins containing a CAG repeat in both species presumably reflects a high level of purifying selection. Based on these observations, we propose that the mutational processes giving rise to polyglutamine repeats in human and murine proteins do not differ. Instead, we propose that the evolution of polyglutamine repeats in proteins results from an interplay between mutational processes and selection.

Amino acid reiterations in yeast are overrepresented in particular classes of proteins and show evidence of a slippage-like mutational process.

Long amino acid repeats are often observed in eukaryotic proteins. In humans, several neurological disorders are caused by proteins containing abnormally long polyglutamines. However, no systematic analysis has attempted to investigate the relationship between reiterations of particular amino acids and protein function, the possible mechanisms involved in the generation of these regions, or the contribution of selection in restricting their genomic distribution, in a large collection of wild-type proteins. We have used baker's yeast open reading frames to study these questions. The most abundant amino acid repeats found in yeast proteins are repeats of glutamine, asparagine, aspartic acid, glutamic acid, and serine. Different amino acid repeats are concentrated in different classes of proteins. Acidic and polar amino acid repeats are significantly associated with transcription factors and protein kinases, while serine repeats are significantly associated with membrane transporter proteins. In most cases the codon structures encoding the repeats at the gene level show a significant bias toward long tracts of one of the possible codons, suggesting that trinucleotide slippage has played an important role in generating these reiterations. However, many, particularly those encoding serine repeats, do not show evidence of slippage. The distributions of codon repeats within proteins and between coding and noncoding regions of the genome, and of amino acids between proteins with different functions, suggest that repeats of these kinds are subject to strong selection.

Germ-line mutations of TP53 in Li-Fraumeni families: an extended study of 39 families.

We have previously reported on the analysis of TP53 coding mutations in 12 classic Li-Fraumeni syndrome (LFS) families plus 9 families that were Li-Fraumeni-like (LFL) families (J. M. Birch et al., Cancer Res., 54: 1298-1304, 1994). Mutations were found in 6 of 12 LFS families and in 1 of 9 LFL families. We have now extended these studies to include an additional nine LFS and nine LFL families, and TP53 mutations have been detected in eight of nine LFS families and in three of nine LFL families. Six of the new mutations described here are the same as those previously identified in other Li-Fraumeni families and are missense mutations at codons 245, 248, and 273 (in two families); a nonsense mutation at codon 209; and a mutation at the splice donor site in exon 4. The other five mutations are novel germ-line mutations and include missense mutations at codons 136 and 344, a 2-bp deletion within codon 191, a splice acceptor mutation in intron 3, and a 167-bp deletion of part of exon 1 and intron 1. In addition, we have detected a codon 175 mutation in a family previously reported as TP53 negative. To summarize all of the data from the families we have studied in this and our previous report (J. M. Birch et al., Cancer Res., 54: 1298-1304, 1994), mutations have been detected in 15 of 21 LFS families (71%) and in 4 of 18 LFL families (22%). These figures are somewhat higher than those previously reported by us and others for the frequency of TP53 mutations in LFS and LFL families. This could reflect our analysis of all 11 exons of TP53, including noncoding regions, as well as the use of direct sequencing rather than other less-sensitive mutation detection methods.

An extended Li-Fraumeni kindred with gastric carcinoma and a codon 175 mutation in TP53.

We present an extended family with Li-Fraumeni syndrome characterised by gastric and breast carcinoma, glioma, sarcoma, and leukaemia. This family showed strong evidence of linkage to TP53, and three of four tumours analysed showed loss of the wild type allele. A codon 175 missense mutation was identified in exon 5 in all available affected subjects. Counselling, screening, and issues surrounding presymptomatic testing are discussed.

Mapping of gene loci in the Q13-Q15 region of chromosome 12.

The gene loci CDK4, GLI, CHOP and MDM2 have been mapped to the q13-q15 region of chromosome 12. Using fluorescence in situ hybridization onto simultaneously DAPI-banded metaphase chromosomes and interphase nuclei, we have more precisely mapped and ordered these loci, together with a number of Genethon microsatellite markers. GLI and CHOP localize to 12q13.3-14.1, CDK4 to 12q14 and MDM2 to 12q14.3-q15, and the gene order is cen-GLI/CHOP-CDK4-MDM2. The Genethon microsatellites D12S80 and D12S83 flank MDM2.

Determination of the gene order of the three loci CD2, NGFB, and NRAS at human chromosome band 1p13 and refinement of their localisation at the subband level by fluorescence in situ hybridisation.

The three loci NRAS, NGFB, and CD2 map to human chromosome band 1p13. Using fluorescence in situ hybridisation (FISH) to simultaneously DAPI-banded metaphase chromosomes, we have further refined the localisation of these three genes to specific subbands. NRAS localises to subband 1p13.2 and CD2 and NGFB to 1p13.1. Also, with the use of multicolour FISH, we have determined the order and orientation of the three loci in relation to the centromere. The order is cen-CD2-NGFB-NRAS.

Structural organisation and chromosomal mapping of the human Id-3 gene.

The helix-loop-helix (HLH) family of transcription factors plays a central role in the regulation of cell growth, differentiation and tumourigenesis. Members of the Id (inhibitor of DNA binding) class of these nuclear proteins are able to heterodimerise with and thereby antagonise the functions of other transcription factors of this family. We report here on the genomic organisation of the human Id3 (HLH 1R21/heir1) gene. Comparison with the two other mammalian Id genes, Id1 and Id2, reveals a highly conserved protein coding gene organisation consistent with evolution from a common, ancestral Id-like gene. In addition, by using a yeast artificial chromosome (YAC) clone of Id3, we have fine-scale mapped the gene to chromosome band 1p36.1 by fluorescence in situ hybridisation (FISH) and, using the same FISH technique, we have detected heterogeneity in tumour-associated 1p36 chromosome translocations.

Evidence for the simultaneous expression of alternatively spliced alkylpurine N-glycosylase transcripts in human tissues and cells.

We have isolated a novel human alkylpurine N-glycosylase (APNG) cDNA from a placental library by screening with an oligonucleotide based on the published sequence of the human liver cDNA encoding this protein. The nucleotide sequences of the two cDNAs were essentially identical, but the 5' untranslated region of the new sequence was truncated and the 5'-terminal 92 nucleotides of the novel cDNA were different, indicating the possibility of alternative transcripts. This region included a portion of the open reading frame, so that the predicted protein was truncated and the seven N-terminal amino acids differed from the published sequence for APNG. PCR amplification of reverse transcribed mRNA, using 5' primers unique to the two cDNAs and a common 3' primer showed that the alternative transcripts can be co-expressed in the same cells and tissues.

Linkage studies in a Li-Fraumeni family with increased expression of p53 protein but no germline mutation in p53.

We report a family with the Li-Fraumeni syndrome (LFS) in whom we have been unable to detect a mutation in the coding sequence of the p53 gene. Analysis of linkage to three polymorphic markers within p53 enabled direct involvement of p53 to be excluded. This is the first example of a LFS family in whom exclusion of p53 has been possible. Four affected members of the family with sarcoma or premenopausal breast cancer showed increased expression of p53 protein in their normal tissues as detected by immunohistochemistry. It therefore appears that the LFS phenotype has been conferred by an aberrant gene, showing a dominant pattern of inheritance, which may be acting to compromise normal p53 function rather than by a mutation in p53 itself. In order to try to determine the chromosomal location of this putative gene, we have carried out studies of linkage to candidate loci. By these means we have excluded involvement of Rb1 and BRCA1 on chromosomes 13q and 17q respectively. The MDM2 oncogene on chromosome 12q was considered to be the prime candidate as MDM2 is amplified in sarcomas and the MDM2 product binds to p53. Furthermore, p53 mutation and amplification of MDM2 have been shown to be mutually exclusive events in tumour development. Linkage analysis to two polymorphic markers within MDM2 yielded a three-point LOD score of -5.4 at a recombination fraction theta equal to zero. Therefore MDM2 could be excluded. It is possible that the gene which is responsible for cancer susceptibility in this family, possibly via interaction with p53, will be important in the histogenesis of breast cancer in general. We are now carrying out further studies to locate and identify this gene.

Rare variant (Glu to Lys) in the leucine zipper region of CHOP (GADD153).

CHOP (GADD153) has been shown to be a dominant negative inhibitor of specific transcription factors. Direct sequencing of the gene, amplified from the DNA of a Li-Fraumeni family index case (liposarcoma, breast cancer) revealed a constitutional variant within the coding region. This alteration, though not responsible for the Li-Fraumeni phenotype, resulted in a glutamic acid to lysine switch within the leucine zipper domain, at a residue conserved between CHOP and its potential target molecules and between the human and hamster sequences. The variant created a Taq I restriction fragment length polymorphism (RFLP) facilitating screening. Analysis of 159 breast tumour DNA samples detected two encoding variant alleles (tumour and constitutional DNA).

The use of loss of constitutional heterozygosity data to ascertain the location of predisposing genes in cancer families.

A method is described to investigate the inheritance of disease predisposition in cancer families. It is an extension of classic genetic linkage analysis, which enables information on loss of constitutional heterozygosity (LOCH) to be incorporated into the model. This adapted model treats LOCH data as additional observations on the disease phenotype. One of the major benefits of this approach is that isolated parent-offspring pairs are now potentially informative for linkage analysis. Examples are presented.