Detection of single base substitutions in total genomic DNA

Certain single base substitutions causing genetic diseases or resulting in polymorphisms linked to mutant alleles, alter a restriction enzyme cleavage site and can therefore be detected in total genomic DNA using DNA blots. Many base substitutions do not lead to an altered restriction site, but these can be detected using synthetic oligonucleotides as hybridization probes if the DNA sequence surrounding the base substitution is known. In the case of beta-thalassaemia, where 22 different single base mutations have been identified, a large number of probes would be required for diagnosis. An approach which was used to detect mutations in viral DNA involves the S1 nuclease treatment of heteroduplexes formed between wild-type and mutant DNA. Although certain single base mismatches are cleaved by S1 nuclease (ref. 11 and T. Shenk, personal communication), many other mismatches examined by this procedure are not cleaved (B. Seed, personal communication; R.M.M., unpublished data). Heteroduplexes between mutant and wild-type subgenomic fragments of double-stranded reovirus RNA migrate slower than the corresponding homoduplexes in polyacrylamide gels containing 7 M urea, but it is not known whether this method is applicable to DNA heteroduplexes containing single base mismatches. Here we describe a procedure that involves the electrophoretic separation of DNA heteroduplexes in a well-characterized gel system. We show that four different human beta-thalassaemia alleles with known single base mutations can be detected with as little as 5 micrograms of total genomic DNA. The method should be useful in the localization and diagnosis of mutations associated with genetic diseases.     

A DNA insertion in the apolipoprotein A-I gene of patients with premature atherosclerosis

Apolipoprotein A-I (apo A-I) is the major protein constituent of high-density lipoprotein (HDL). The study of the apo A-I gene is of interest because plasma levels of HDL have been inversely correlated with the development of coronary artery disease and because polymorphisms related to this gene have been associated with hypertriglyceridaemia and premature atherosclerosis. We have recently isolated and characterized the human apo A-I gene and have shown that apo A-I and apolipoprotein C-III (apo C-III) genes are physically linked and that a polymorphism (of unknown frequency in the general population) of the apo A-I gene is inherited as a mendelian trait linked to premature atherosclerosis in an affected family (not the same polymorphism as has previously been reported to be associated with hypertriglyceridaemia). Here we report that this polymorphism is due an at least 6.5-kilobase (kb) DNA insertion in the coding region of the apo A-I gene and that there is no detectable alteration (as determined by limited genomic blotting analysis) of the apo C-III gene of these patients.     

Low variability in a Y-linked plant gene and its implications for Y-chromosome evolution

Sex chromosomes have evolved independently in several different groups of organisms, but they share common features, including genetic degeneration of the Y chromosome. Suppression of recombination between ancestral proto-X and proto-Y chromosomes is thought to have led to their gradual divergence, and to degeneration of the Y chromosome, but the evolutionary forces responsible are unknown. In non-recombining Y chromosomes, deleterious mutations may be carried to fixation by linked advantageous mutations ("selective sweeps"). Occurrence of deleterious mutations may drive "Muller's ratchet" (stochastic loss of chromosomes with the fewest mutations). Selective elimination of deleterious mutations, causing "background selection" may accelerate stochastic fixation of mildly detrimental mutations. All these processes lower effective population sizes, and therefore reduce variability of genes in evolving Y chromosomes. We have studied DNA diversity and divergence in a recently described X- and Y-linked gene pair (SLX-1 and SLY-1) of the plant Silene latifolia to obtain evidence about the early stages of Y degeneration. Here we show that DNA polymorphism in SLY-1 is 20-fold lower than in SLX-1, but the pattern of polymorphism does not suggest a selective sweep.     

gamma-beta-Thalassaemia studies showing that deletion of the gamma- and delta-genes influences beta-globin gene expression in man

In gamma-beta-thalassaemia, human gamma- and beta-globin gene expression is suppressed; this results in a severe anaemia in newborns which subsequently develops into a beta-thalassaemia syndrome in adult life. This hereditary disease is now shown to be the result of a deletion of at least 40,000 base pairs of the gammadeltabeta-globin gene locus. The gamma- and delta-globin genes are deleted in the affected chromosome but, surprisingly, the beta-globin gene is still present, together with a large segment of the DNA sequences flanking the gene on its 5'-side and the entire region on the 3'-side of the gene. Hence, a deletion of DNA far from the beta-globin gene results in the suppression of its activity.     

A novel alpha-globin gene arrangement in man

The human genome has two linked alpha-globin genes on chromosome 16. Deletion of one or more of them, as occurs in alpha-thalassaemia, leads to a reduced output of alpha-globin mRNA in proportion to the number of alpha-globin genes lost. In some racial groups deletion of one of the pair of alpha-globin genes may result from unequal crossing over between the genes on homologous chromosomes by a mechanism resembling that postulated for the formation of the delta beta fusion genes of the Lepore haemoglobins. By analogy, the opposite chromosome in this cross-over should have three alpha-globin genes just as the 'anti-Lepore chromosome has three non-alpha chain genes. We describe here a Welsh family in which three members have five alpha-globn increased alpha mRNA output and it may therefore produce the phenotype of mild beta-thalassaemia.     

Gene association study with SVM, MLP and cross-validation for the diagnosis of diseases

Gene association study is one of the major challenges of biochip technology both for gene diagnosis where only a gene subset is responsible for some diseases, and for the treatment of the curse of dimensionality which occurs especially in DNA microarray datasets where there are more than thousands of genes and only a few number of experiments (samples). This paper presents a gene selection method by training linear support vector machine (SVM)/nonlinear MLP (multilayer perceptron) classifiers and testing them with cross-validation for finding a gene subset which is optimal/suboptimal for the diagnosis of binary/multiple disease types. Genes are selected with linear SVM classifier for the diagnosis of each binary disease types pair and tested by leave-one-out cross-validation; then, genes in the gene subset initialized by the union of them are deleted one by one by removing the gene which brings the greatest decrease of the generalization power, for samples, on the gene subset after removal, where generalization is measured by training MLPs with leave-one-out and leave-four-out cross-validations. The proposed method was tested with experiments on real DNA microarray MIT data and NCI data. The result shows that it outperforms conventional SNR method in the separability of the data with expression levels on selected genes. For real DNA microarray MIT/NCI data, which is composed of 7129/2308 effective genes with only 72/64 labeled samples belonging to 2/4 disease classes, only 11/6 genes are selected to be diagnostic genes. The selected genes are tested by the classification of samples on these genes with SVM/MLP with leave-one-out/both leave-one-out and leave-four-out cross-validations. The result of no misclassification indicates that the selected genes can be really considered as diagnostic genes for the diagnosis of the corresponding diseases.     

Gene association study with SVM, MLP and cross validation for diagnosis of diseases

Gene association study is one of the major challenges of biochip technology both for gene diagnosis where only a gene subset is responsible to some diseases, and for treatment of curse of dimensionality which occurs especially in DNA microarray datasets where there are more than thousands of genes and only a few number of experiments (samples). This paper presents a gene selection method by training linear support vector machine (SVM)/nonlinear MLP (multi-layer perceptron) classifiers and testing them with cross validation for finding a gene subset which is optimal/suboptimal for diagnosis of binary/multiple disease types. Genes are selected with linear SVM classifier for the diagnosis of each binary disease types pair and tested by leave-one-out cross validation; then, genes in the gene subset initialized by the union of them are deleted one by one by removing the gene which brings the greatest decrease of the generalization power, for samples, on the gene subset after removal, where generalization is measured by training MLPs with leave-one-out and leave-4-out cross validations. The proposed method was tested with experiments on real DNA microarray MIT data and NCI data. The result shows that it outperforms conventional SNR method in separability of the data with expression levels on selected genes. For real DNA microarray MIT/NCI data, which is composed of 7129/2308 effective genes with only 72/64 labeled samples belonging to 2/4 disease classes, only 11/6 genes are selected to be diagnostic genes. The selected genes are tested by classification of samples on these genes with SVM/MLP with leave-one-out/both leave-one-out and leave-4-out cross validations. The result of no misclassification indicates that the selected genes can be really considered as diagnostic genes for the diagnosis of the corresponding diseases.     

Gene association study with SVM, MLP and cross validation for diagnosis of diseases

Gene association study is one of the major challenges of biochip technology both for gene diagnosis where only a gene subset is responsible for some diseases, and for the treatment of the curse of dimensionality which occurs especially in DNA microarray datasets where there are more than thousands of genes and only a few number of experiments （samples）. This paper presents a gene selection method by training linear support vector machine （SVM）/nonlinear MLP （multilayer perceptron） classifiers and testing them with cross-validation for finding a gene subset which is optimal/suboptimal for the diagnosis of binary/multiple disease types. Genes are selected with linear SVM classifier for the diagnosis of each binary disease types pair and tested by leave-one-out cross-validation; then, genes in the gene subset initialized by the union of them are deleted one by one by removing the gene which brings the greatest decrease of the generalization power, for samples, on the gene subset after removal, where generalization is measured by training MLPs with leaveone-out and leave-four-out cross-validations. The proposed method was tested with experiments on real DNA microarray MIT data and NCI data. The result shows that it outperforms conventional SNR method in the separability of the data with expression levels on selected genes. For real DNA microarray MIT/NCI data, which is composed of 7129/2308 effective genes with only 72/64 labeled samples belonging to 2/4 disease classes, only 11/6 genes are selected to be diagnostic genes. The selected genes are tested by the classification of samples on these genes with SVM/MLP with leave-one-out/both leave-one-out and leave-four-out cross-validations. The result of no misclassification indicates that the selected genes can be really considered as diagnostic genes for the diagnosis of the corresponding diseases.     

Correlation between a DNA restriction fragment length polymorphism and C4A6 protein

The fourth component of complement (C4) in man, is coded for by two separate but closely linked loci (C4A and C4B) within the major histocompatibility region (MHC), on the short arm of chromosome 6. Like class I and II loci of this region, the C4 genes are highly polymorphic with more than 30 alleles, including null alleles, assigned to the two loci. This extensive polymorphism, based mainly on electrophoretic mobility, provides a useful marker for studies of disease susceptibility. Several disorders, including systemic lupus erythematosus and type I diabetes, show associations with C4 phenotypes. We have used the technique of Southern with a C4 specific probe to examine the genomic DNA of individuals typed for C4 by protein electrophoresis. We have identified 10.7 and 3.8 kilobase (kb) BglII restriction fragments in each of 9 unrelated individuals with a C4A6 allele, and in none of 22 unrelated individuals in whom this allele was not expressed. This clear correlation of restriction fragment length polymorphism with C4 phenotype provides a precise basis for analysis of C4 polymorphism. It is likely to be of value in clinical investigations of autoimmune disease.     

Comparative chromosome mapping of a conserved homoeo box region in mouse and human

Specific genes are assumed to regulate pattern formation in the mammalian embryo, but as yet none has been identified unequivocally. It is possible that such genes in mammals may be identified by virtue of a conserved coding sequence, because many of the Drosophila melanogaster homoeotic and segmentation genes, which have crucial roles in the regulation of segmental pattern formation during embryonic development, contain a 180-base pair (bp) DNA sequence, the homoeo box, and that sequences homologous to the Drosophila homoeo box are also present in 6-10 copies in higher animals, including mammals. Although the assumption that the homoeo box identifies genes responsible for pattern formation in mammals remains to be validated, it is a particularly attractive hypothesis given the strong conservation of homoeo boxes over vast evolutionary distances. Here we report the localization of a human homoeo box region, previously cloned and shown to contain two homoeo boxes within a sequence of 5-kilobases (kb), to the long arm of chromosome 17. We show that two single-copy homoeo box-flanking probes derived from this region strongly hybridize to single-copy restriction fragments in mouse genomic DNA and that these conserved homoeo box-flanking sequences map to mouse chromosome 11. This may be significant as several genes that map to chromosome 17 in human also map to chromosome 11 in the mouse, implying that a segment of mouse chromosome 11 is homologous to a region of human chromosome 17. Taken together, these data suggest that the homoeo box region detected with our probes is highly conserved in human and mouse.     

Characterisation of deletions which affect the expression of fetal globin genes in man.

Deletions in the DNA of individuals with hereditary persistence of fetal haemoglobin (HPFH) and 8 beta-thalassaemia have been mapped as a means of identifying regulatory sequences involved in the switch from fetal to adult globin gene expression. The end points of these deletions have been precisely located with respect to restriction endonuclease cleavage sites within and surrounding the gamma-, delta- and beta-globin genes in normal human DNA and the deletion maps were used to obtain definitive evidence for the physical linkage of the fetal and adult beta-like globin genes in the order 5'Ggamma-Agamma-delta-beta 3'. Correlation of haematological data and the location of deletions in two cases of HPFH and one case of deltabeta-thalassaemia suggest that a region of DNA located near the 5'-end of the delta-globin gene may be involved in the suppression in cis of gamma-globin gene expression in adults. The interpretation of a second case of deltabeta-thalassaemia is complicated by the fact that the deletion removes the Agamma-gene in addition to the region near the 5'-end of the delta-globin gene.     

Patterns of somatic mutation in human cancer genomes

Cancers arise owing to mutations in a subset of genes that confer growth advantage. The availability of the human genome sequence led us to propose that systematic resequencing of cancer genomes for mutations would lead to the discovery of many additional cancer genes. Here we report more than 1,000 somatic mutations found in 274 megabases (Mb) of DNA corresponding to the coding exons of 518 protein kinase genes in 210 diverse human cancers. There was substantial variation in the number and pattern of mutations in individual cancers reflecting different exposures, DNA repair defects and cellular origins. Most somatic mutations are likely to be 'passengers' that do not contribute to oncogenesis. However, there was evidence for 'driver' mutations contributing to the development of the cancers studied in approximately 120 genes. Systematic sequencing of cancer genomes therefore reveals the evolutionary diversity of cancers and implicates a larger repertoire of cancer genes than previously anticipated.     

The origin of MHC class II gene polymorphism within the genus Mus

The I region of the major histocompatibility complex (MHC) of the mouse (H-2) contains a tightly-linked cluster of highly polymorphic genes (class II MHC genes) which control immune responsiveness. Speculation on the origin of this polymorphism, which is believed to be essential for the function of the class II proteins in immune responses to disease, has given rise to two hypotheses. The first is that hypermutational mechanisms (gene conversion or segmental exchange) promote the rapid generation of diversity in MHC genes. The alternative is that polymorphism has arisen from the steady accumulation of mutations over long evolutionary periods, and multiple specific alleles have survived speciation (trans-species evolution). We have looked for evidence of 'segmental exchange' and/or 'trans-species evolution' in the class II genes of the genus Mus by molecular genetic analysis of I-A beta alleles. The results indicate that greater than 90% (28 out of 31) of the alleles examined can be organized into two evolutionary groups both on the basis of restriction site polymorphisms and by the presence or absence of a short interspersed nucleotide element (SINE). Using this SINE sequence as an evolutionary tag, we demonstrate that I-A beta alleles in these two evolutionary groups diverged at least three million years ago and have survived the speciation events leading to several modern Mus species. Nucleotide sequence comparisons of eight Mus m. domesticus I-A beta alleles representing all three evolutionary groups indicate that most of the divergence in exon sequences is due to the steady accumulation of mutations that are maintained independently in the different alleles. But segmental exchanges between alleles from different evolutionary groups have also played a role in the diversification of beta 1 exons.     

Cloning of the breakpoint of an X;21 translocation associated with Duchenne muscular dystrophy

Duchenne muscular dystrophy (DMD) is an X-linked recessive disorder which affects approximately 1 in 3,300 males, making it the most common of the neuromuscular dystrophies. The biochemical basis of the disease is unknown and as yet no effective treatment is available. A small number of females are also affected with the disease, and these have been found to carry X; autosome translocations involving variable autosomal sites but always with a breakpoint within band Xp21 of the X chromosome (implicated by other kinds of genetic evidence as the site of the DMD lesion). In these female patients the normal X chromosome is preferentially inactivated, which it is assumed silences their one normal DMD gene, leading to expression of the disease. In one such affected female the autosomal breakpoint lies in the middle of the short arm of chromosome 21, within a cluster of ribosomal RNA genes. Here we have used rRNA sequences as probes to clone the region spanning the translocation breakpoint. A sequence derived from the X-chromosomal portion of the clone detects a restriction fragment length polymorphism (RFLP) which is closely linked to the DMD gene and uncovers chromosomal deletions in some male DMD patients.     

A potential donor gene for the bm1 gene conversion event in the C57BL mouse

The mammalian major histocompatibility complex (MHC; H-2 complex in mouse) is a large multigene complex which encodes cell-surface antigens involved in the cellular immune response to foreign antigens. Class I polypeptides expressed at the H-2K and H-2D loci of numerous mouse strains exhibit an unusually high degree of genetic polymorphism, which is assumed to be related to their function as primary recognition elements in the immune response. We suggested that this H-2 polymorphism may arise by gene conversion-like events between non-allelic class I genes. This is supported by our recent comparison of the DNA sequences of the normal H-2Kb gene sequence, from the C57BL/10 mouse, and a mutant form of this gene called H-2Kbm1: the mutant allele differs from the H-2Kb gene in seven bases out of a region of 13 bases in exon 3 of the class I gene (which encodes alpha 2 (C1) the second highly polymorphic protein domain), suggesting that this region of new sequence had been introduced into the H-2Kb sequence following unequal pairing of two class I genes in the genome of the C57BL mouse. Schulze et al. have obtained similar results. Here we report work identifying a potential donor gene in our library of 26 class I genes cloned from the C57BL/10 mouse.     

Preferential deletion of exons in Duchenne and Becker muscular dystrophies

Duchenne and Becker muscular dystrophy (DMD and BMD) genes are located in Xp21 on the short arm of the X chromosome. DMD patients display a much more severe clinical course than BMD patients, and yet about 10% of cases of each have been reported to have deletions for parts of the gene. Using a complementary DNA subclone of the DMD gene we have screened 66 DMD and BMD patients who had not previously shown deletions with the probes then available. Fifteen patients have a deletion of this part of the gene, indicating a higher deletion frequency in this region (22%). Exons were deleted in five severely affected DMD patients and in ten BMD patients. Significantly, most of these deletions begin in the same region of the cDNA, which implies that there is a common mechanism for the generation of many of these mutations. An apparently identical deletion in one family gave classical BMD in two brothers (presenting in their teens) and only very mild muscle weakness in their 86-year-old great-great-uncle. Taking these data together with data using the probes previously published, we are able to detect deletions directly in 40% of our families requiring antenatal diagnosis or carrier detection.     

Conservation and rearrangement of mitochondrial structural gene sequences

Mitochondria contain the simplest DNA molecules that are present in eukaryotes. Mitochondrial DNA (mtDNA) is easily purified, and is an important model system for studying eukaryote gene structure and basic molecular processes. The protein sequences of mitochondrial gene products have been shown to be conserved from yeast to man, and there are definite similarities at the DNA sequence level. In contrast, the overall organization of the mitochondrial genome is drastically different in these organisms. To understand this, we need to extend work on mtDNA to a wider range of species. We have chosen to study the mtDNA of Aspergillus nidulans because a particularly comprehensive analysis of this system can be achieved using genetics as well as biochemistry, and like most eukaryotes it is an obligate aerobe, whereas Saccharomyces cerevisiae is not. We have investigated whether defined pieces of particular yeast mitochondrial genes show enough homology to Aspergillus mtDNA fragments to enable the corresponding Aspergillus genes to be located on the physical map. The results reported here show that this is the case for all five genes tested, and present the first data on the physical organization of the structural genes in the mitochondrial genome of A. nidulans.