Search

DNA mismatch repair (MMR) deficiency is associated with an increased mutational burden and predisposition to certain malignancies. Relatively little is known, however, about gene-specific mutation frequencies within MMR-deficient primary tumors. Thymic lymphomas from Msh2 −/− mice were thus analyzed by using a lacI -based transgenic shuttle-phage mutation detection system. All tumors exhibited greatly elevated lacI gene mutation frequencies, ranging from 3.2- to 17.4-fold above the ≈15-fold elevations present within normal Msh2 −/− thymi. In addition, lacI genes harboring multiple changes, including clusters of mutations, were found in thymic tumor DNA. The results suggest that an additional mutator activity, such as an error-prone DNA polymerase, leads to increased genomic instability in these MMR-deficient tumors.

The mutational response of mismatch repair-deficient animals to the alkylating agent N -methyl- N -nitrosourea was evaluated by using a transgenic lacI reporter system. Although the mutations detected in MSH2 heterozygotes were similar to those of controls, MSH2 −/− animals demonstrated striking increases in mutation frequency in response to this agent. G:C to A:T transitions at GpG sites, as opposed to CpG sites, dominated the mutational spectrum of both MSH2 +/+ and MSH2 −/− N -methyl- N -nitrosourea -treated animals. Extrapolating to humans with hereditary non-polyposis colorectal cancer, the results suggest that MSH2 heterozygotes are unlikely to be at increased risk of mutation, even when exposed to potent DNA methylating agents. In contrast, mismatch repair-deficient cells spontaneously arising within individuals with hereditary non-polyposis colorectal cancer would likely exhibit hypermutability in response to such mutagens, an outcome predicted to accelerate the pace of tumorigenesis.

Tumors derived from individuals with hereditary non-polyposis colorectal cancer syndrome frequently demonstrate mutations in both alleles of hMSH2, a key gene in DNA mismatch repair (MMR). Sporadic tumors also frequently exhibit MMR deficiency. In keeping with the role of MMR in the maintenance of genome integrity, mice deficient in MSH2 via gene targeting demonstrate a high incidence of thymic lymphomas and small intestinal adenocarcinomas. To investigate the effects of MSH2 deficiency in normal tissues, mice containing a retrievable transgenic lacI reporter gene for mutation detection were crossed with MSH2−/− mice. Mice homozygous for MSH2 deficiency revealed 4.8, 11.0 and 15.2-fold elevations in spontaneous mutation frequency in DNA obtained from brain, small intestine, and thymus, respectively, as compared to heterozygous or wild-type mice. Mutations most frequently recovered from MSH2−/− mice were single base substitutions (77%), particularly base transitions (64%). Frameshifts occurred less frequently (19%) and fell within very short (3 – 5 bp) mononucleotide runs. Thus the number of key growth control genes potentially impacted by MMR deficiency extends beyond those containing repetitive sequences. These results highlight the capacity for MSH2 deficiency to serve as a potent driving force during the multi-step evolution of tumors.

To assess DNA mutations in vivo, we have established a new transgenic mouse line, BC-1, carrying a lacI target gene for mutation detection within a bacteriophage shuttle-vector. The lacI gene was positioned within sequences derived from a rearranged murine immunoglobulin gene locus, a feature that distinguishes the BC-1 transgene from other shuttle vector systems. As mutations in lacI transgenes likely reflect mutations occurring throughout the genome, these systems have been successfully used to investigate spontaneous and induced mutations in a variety of tissues. An important additional application of the transgenic systems is the characterization of lacI mutations occurring in murine strains having specific DNA repair defects. For this study, scid (severe combined immunodeficiency) mice were selected as animals with this mutation have a defect in double-strand DNA break repair. To determine what impact the scid mutation might have on spontaneous mutation frequencies within DNA recovered from various tissues, these mice were crossed with the BC-1 line. Interestingly, mutation frequencies within BC-1/scid mouse DNA were not significantly different from those of BC-1 control mice. Furthermore, spontaneous lacI mutations obtained from BC-1 and from BC-1/scid liver DNA were similar in spectrum. As spontaneous BC-1 liver mutations were similar to those reported previously for other lacI systems, such as the Big Blue transgenic line, this suggested that the nature of the DNA sequences flanking the reporter gene did not modify lacI mutation rate or character.

A transcription map of the Huntington disease gene region was generated by a direct cDNA selection strategy using genomic DNA from the 4p16.3 region surrounding the D4S95 and D4S127 loci. A total of 58 cDNA fragments were obtained from cDNAs derived from fetal brain, frontal cortex, liver and bone marrow following hybridization to overlapping YACs from this region. These cDNA clones were aligned into transcription units by hybridization to specific mRNAs, by sequence overlap and by physical mapping onto overlapping YAC clones. Nine separate transcription units spanning approximately one megabase were detected by RNA hybridization. They represent a minimum number of genes in this region and do not include those genes expressed specifically in tissues not used for the hybridization. The transcription map that is provided by the cDNA segments will lead to the generation of a detailed gene map of this region.

A lack of DNA mismatch repair (MMR) is observed in approximately 20% of sporadic endometrial tumors, but few of these tumors have mutations in MSH2 or MLH1, the two major MMR genes. Promoter methylation is an important means of silencing transcription, and methylation of the MLH1 promoter has been reported as an important cause of MLH1 inactivation in endometrial cancers. Studies have shown that specific CpG sites within the MLH1 gene promoter are critical for gene expression, but other studies have shown that methylation of both more proximal and more distal sequences are important for MLH1 gene regulation. Here, we used a microsatellite instability assay and MLH1 immunohistochemistry to identify a subset of endometrial carcinomas of the endometrioid subtype lacking MMR. Sequencing of bisulphite-treated DNA from these tumors determined the methylation status of 42 CpG sites across the MLH1 promoter (spanning -204 to -702 bp upstream of the transcriptional start). Unlike the 4 normal endometrial samples that were unmethylated, 17 of 21 MMR-deficient samples showed complete or near-complete methylation and the remaining 4 MMR-deficient samples had a considerable degree of methylation (approximately 50% or greater). Five tumors demonstrated isolated unmethylated CpG sites, despite methylation throughout the rest of the promoter. This underscores the importance of examining the methylation status of at least several CpG sites within the promoter as methylation is not always consistent across DNA. Overall, our findings support the model that density of methylation of CpG sites across the MLH1 promoter is important in determining gene expression.

INTRODUCTION Disorders caused by triplet expansions present unique challengesfor understanding and correlating the genotype with the clinicalphenotype. Currently, there is some confusion over ranges fornormal and disease alleles and regarding understanding ofpenetrance and intermediate alleles. Various centres have adopteddifferent definitions, resulting in varying interpretations of allelesizes and their relationship to phenotype by a particular age.Furthermore, the observation in some trinucleotide repeat (TNR)diseases of new mutations arising from ‘intermediate’ sizedalleles has resulted in different interpretations over what is a‘normal’ and what is a ‘disease’ allele. This has also beenaccompanied by an emerging recognition that many factors,including genetic changes outside of the causative gene andpotentially even environmental features, might influence theseclassical ‘single gene’ disorders. We have collected data from >80peer-reviewed manuscripts on the seven disorders associated withexpansion of a CAG repeat, as well as consulted with experts inthe field on unpublished data, in order to define the normal anddisease range of alleles, determine the zone of reduced pene-trance, if any, and define the size range for intermediate sizedalleles proven molecularly to expand to a CAG size in the diseaserange. Here we present an approach for the interpretation of allelesizes and their relationship to a phenotype for this class ofdisorders.CAG SIZE RANGES ASSOCIATED WITH DISEASEThe ranges of allele sizes in the seven disorders is based on the totalnumber of individuals, normal and affected, that have been assessed.This in turn is influenced by the frequency of the disorder as well as,to a lesser extent, the length of time since the gene has been cloned(Table 1) (1–88). The affected range is defined as the range of CAGsdetected in persons with the clinical phenotype of a particulardisorder caused by mutations in the gene in question. The dataavailable to date are shown in Figure 1 (1–88). For spinocerebellarataxia type 1 and 2 (SCA1, SCA2), spinal and bulbar muscularatrophy (SBMA), dentatorubropallidoluysian atrophy (DRPLA) andespecially spinocerebellar ataxia type 6 (SCA6), the small numbersof affected chromosomes examined means that the ranges will likelychange at their margins, as more individuals are assessed (Table 1).Except for SCA6, the similarity between the ranges of disease(>40 CAGs) and normal size alleles at each of the seven TNRdiseases suggests a common mechanism of expansion. However,the interval between normal and disease range can also be usedto divide these seven disorders into two groups. For example,SBMA, SCA6, DRPLA and SCA3 clearly have mutuallyexclusive normal and disease range CAG alleles. On the otherhand, there is no obvious gap between normal and disease rangesfor Huntington’s disease (HD), SCA1 and SCA2 (Fig. 1).PENETRANCE, INDIVIDUALITY AND DISEASEDespite our extensive genetic variability and exposure to differentenvironments, it is remarkable that, for all these disorders, thatCAG size is associated so strongly with age of onset—a criteraitself subject to significant differences in interpretation. A diseasegene has reduced penetrance when a fraction of patients do notmanifest the disorder despite inheriting a known disease-causingmutation (89). However, for many individuals with TNRdiseases, the onset of illness is often late and highly variable. Thuspenetrance is age-dependent. In these instances it is appropriateto define reduced penetrance as occurring if persons haveinherited a mutation in the disease range but lived beyond theexpected lifespan without manifesting any signs of the disease.Reduced penetrance is the one end of the spectrum ofexpression of a disease-causing mutation and is clear evidence ofhow unique individuals accommodate to a particular mutation inthe context of the whole genome, environment and experience.The delineation of the molecular basis of these diseases hasallowed new and definitive approaches to assessment ofpenetrance in this class of disorders. For all of the CAG TNRdiseases combined, the reported instances of non-penetrance arefew, given the large number of individuals molecularly assessed(8,30,44,90–92).

An optical biosensor was used to monitor interactions between the Escherichia coli DNA mismatch repair molecule MutS and various immobilized oligonucleotides. While associating poorly with single-stranded DNA, MutS was capable of rapid association/dissociation from homoduplex DNA. The interaction of MutS with oligonucleotide 30-mers containing single site mismatches demonstrated that during the dissociation phase, MutS binding was greatest to a G-G mismatch, followed by G-T > A-A > C-T, A-C. Binding to A-G, T-T and C-C mispairs was marginally higher than that seen between MutS and homoduplex DNA. The ability of MutS to interact with 30-mers containing alkylated bases was also tested. While binding to O6-methyl-G-C, or to O4-methyl-T-A base pairs was similar to that of homoduplex DNA, strong binding was seen to a O6-methyl-G-T mispair. O4-methyl-T-G, however, was poorly recognized by MutS, with relative binding affinity similar to homoduplex DNA, predicting poor in vivo recognition of O4-methyl-T-G by MutS. Interestingly, MutS demonstrated a relatively high affinity for an 1,N6-etheno-A-T containing homoduplex. Thus, in allowing rapid evaluation of interactions between such molecules, the biosensor will be useful to structure-function analyses.

No detectable rearrangements involving chromosome 4p16.3 have been observed in patients with Huntington's disease (HD). New mutations for HD could involve structural alterations which might aid the localization of the defective gene. We have reinvestigated a well documented sporadic case of HD. DNA haplotyping with markers between D4S10 and the telomeric locus D4S141 reveals a recombination event in one chromosome of the sporadic HD patient. The site of recombination maps within a 50 kilobase (kb) region, about 700 kb from the 4p telomere. Based on the extremely low HD mutation rate and significantly decreased recombination in the distal region of 4p, we hypothesize a direct link between the site of the recombination and HD in this patient.

As part of the search for the Huntington disease (HD) gene we have cloned and sequenced 34 kb of genomic DNA containing the full-length gene for the beta-subunit of the human cGMP phosphodiesterase (beta-cGMP PDE). This gene is localized to 4p16.3 about 700 kb proximal to the 4p telomere and represents the most telomeric gene characterized on 4p to date. We show that this gene is comprised of 22 exons spanning approximately 43 kb of genomic DNA. We also provide 400 bp immediately 5' to the putative initiator methionine and 700 bp of 3' flanking sequences. Northern blot analysis of several human tissues revealed a highly abundant 3.5 kb transcript and a minor signal of 4.5 kb in retinal tissue. Alignment of the deduced amino acid sequence to the previously identified beta-subunits of the cGMP PDEs of mouse and cow demonstrates highly significant similarities and, therefore, confirms the identity of the cloned gene. A defect in the beta-subunit of the cGMP PDE gene has been shown recently to be the cause for the retinal degeneration in the rd mouse. The cloning of the human homolog and the knowledge of its genomic organization with exon/intron boundaries will allow rapid assessment of the role of this gene in the causation of human retinopathies.