Rhabdomyosarcoma in patients with constitutional mismatch-repair-deficiency syndrome.

BACKGROUND: Biallelic germline mutations in the mismatch repair genes MLH1, MSH2, MSH6 or PMS2 cause a recessive childhood cancer syndrome characterised by early-onset malignancies and signs reminiscent of neurofibromatosis type 1 (NF1). Alluding to the underlying genetic defect, we refer to this syndrome as constitutional mismatch repair-deficiency (CMMR-D) syndrome. The tumour spectrum of CMMR-D syndrome includes haematological neoplasias, brain tumours and Lynch syndrome-associated tumours. Other tumours, such as neuroblastoma, Wilm tumour, ovarian neuroectodermal tumour or infantile myofibromatosis, have so far been found only in individual cases. RESULTS: We analysed two consanguineous families that had members with suspected CMMR-D syndrome who developed rhabdomyosarcoma among other neoplasias. In the first family, we identified a pathogenic PMS2 mutation for which the affected patient was homozygous. In family 2, immunohistochemistry analysis showed isolated loss of PMS2 expression in all tumours in the affected patients, including rhabdomyosarcoma itself and the surrounding normal tissue. Together with the family history and microsatellite instability observed in one tumour this strongly suggests an underlying PMS2 alteration in family 2 also. CONCLUSION: Together, these two new cases show that rhabdomyosarcoma and possibly other embryonic tumours, such as neuroblastoma and Wilm tumour, belong to the tumour spectrum of CMMR-D syndrome. Given the clinical overlap of CMMR-D syndrome with NF1, we suggest careful examination of the family history in patients with embryonic tumours and signs of NF1 as well as analysis of the tumours for loss of one of the mismatch repair genes and microsatellite instability. Subsequent mutation analysis will lead to a definitive diagnosis of the underlying disorder.

RNA-based mutation analysis identifies an unusual MSH6 splicing defect and circumvents PMS2 pseudogene interference.

Heterozygous germline mutations in one of the mismatch repair (MMR) genes MLH1, MSH2, MSH6, and PMS2 cause hereditary nonpolyposis colorectal cancer (HNPCC) or Lynch syndrome, a dominantly inherited cancer susceptibility syndrome. Recent reports provide evidence for a novel recessively inherited cancer syndrome with constitutive MMR deficiency due to biallelic germline mutations in one of the MMR genes. MMR-deficiency (MMR-D) syndrome is characterized by childhood brain tumors, hematological and/or gastrointestinal malignancies, and signs of neurofibromatosis type 1 (NF1). We established an RNA-based mutation detection assay for the four MMR genes, since 1) a number of splicing defects may escape detection by the analysis of genomic DNA, and 2) DNA-based mutation detection in the PMS2 gene is severely hampered by the presence of multiple highly similar pseudogenes, including PMS2CL. Using this assay, which is based on direct cDNA sequencing of RT-PCR products, we investigated two families with children suspected to suffer from MMR-D syndrome. We identified a homozygous complex MSH6 splicing alteration in the index patients of the first family and a novel homozygous PMS2 mutation (c.182delA) in the index patient of the second family. Furthermore, we demonstrate, by the analysis of a PMS2/PMS2CL "hybrid" allele carrier, that RNA-based PMS2 testing effectively avoids the caveats of genomic DNA amplification approaches; i.e., pseudogene coamplification as well as allelic dropout, and will, thus, allow more sensitive mutation analysis in MMR deficiency and in HNPCC patients with PMS2 defects.

Extensive in silico analysis of NF1 splicing defects uncovers determinants for splicing outcome upon 5' splice-site disruption.

We describe 94 pathogenic NF1 gene alterations in a cohort of 97 Austrian neurofibromatosis type 1 patients meeting the NIH criteria. All mutations were fully characterized at the genomic and mRNA levels. Over half of the patients carried novel mutations, and only a quarter carried recurrent minor-lesion mutations at 16 mutational warm spots. The remaining patients carried NF1 microdeletions (7%) and rare recurring mutations. Thirty-six of the mutations (38%) altered pre-mRNA splicing, and fall into five groups: exon skipping resulting from mutations at authentic splice sites (type I), cryptic exon inclusion caused by deep intronic mutations (type II), creation of de novo splice sites causing loss of exonic sequences (type III), activation of cryptic splice sites upon authentic splice-site disruption (type IV), and exonic sequence alterations causing exon skipping (type V). Extensive in silico analyses of 37 NF1 exons and surrounding intronic sequences suggested that the availability of a cryptic splice site combined with a strong natural upstream 3' splice site (3'ss)is the main determinant of cryptic splice-site activation upon 5' splice-site disruption. Furthermore, the exonic sequences downstream of exonic cryptic 5' splice sites (5'ss) resemble intronic more than exonic sequences with respect to exonic splicing enhancer and silencer density, helping to distinguish between exonic cryptic and pseudo 5'ss. This study provides valuable predictors for the splicing pathway used upon 5'ss mutation, and underscores the importance of using RNA-based techniques, together with methods to identify microdeletions and intragenic copy-number changes, for effective and reliable NF1 mutation detection.

Identification of the Ly49L protein: evidence for activating counterparts to inhibitory Ly49 proteins.

Previous studies have indicated that NK cells from different strains of inbred mice may express distinct Ly49 repertoires. Screening of NK cells from the CBA/J mouse for inhibitory and activating Ly49s revealed a novel DAP12-associated receptor that was immunoprecipitated with the Ly49G-specific mAb 4D11. Degenerate primers were designed to amplify and clone Ly49 cDNAs from CBA/J NK cells. A novel activating Ly49 cDNA was identified, which bears strong homology to the partially sequenced Ly49l gene found in C57BL/6 mice. Transfection of Ly49l into a DAP12+ cell line and subsequent immunoprecipitation experiments showed that Ly49L is likely the activating Ly49 detected by the 4DD11 antibody in CBA/J NK cells. Antibody-mediated cross-linking of Ly49L induced DAP12 phosphorylation, providing evidence that Ly49L is a functional activating receptor. Comparison of the extracellular domains of Ly49 family members indicates that all known activating members have an inhibitory counterpart with a highly related extracellular region.

Cellular distribution and pharmacological sensitivity of low Km cyclic nucleotide phosphodiesterase isozymes in human cardiac muscle from normal and cardiomyopathic subjects.

Cyclic nucleotide phosphodiesterase (PDE) isozymes isolated by DEAE-Sephacel or Mono-Q High Performance Liquid Chromatography from cardiac left ventricular tissue of normal subjects and patients with end-stage heart failure have been compared. With both separation techniques, four major peaks of PDE activity were evident in the soluble fractions; only one peak of activity was present in particulate fractions. The specific activity of the particulate PDE from myopathics was approximately 30-50% of that of normals while the specific activity of a soluble form of this PDE (peak IIIa) was reduced by 30% in myopathics. No differences in comparison of the other peaks of PDE activity were evident. The particulate PDE isozyme has a low Km for cAMP (0.27-0.29 microM), is inhibited by cGMP (60-80% at 1 microM), is sensitive to inhibition by submicromolar concentrations of CI-930 but not rolipram, and is competitively inhibited by milrinone (Kj = 0.3 microM). The first soluble peak of PDE activity hydrolyzes both cAMP and cGMP and is stimulated by calmodulin while cyclic AMP hydrolysis by peak II PDE is stimulated by cGMP. The other soluble peak III fractions (IIIa and IIIb) hydrolyze cAMP; peak IIIa is inhibited by cGMP or by CI-930 and milrinone, whereas peak IIIb is also inhibited by rolipram when the cardiotonic sensitive PDE is inhibited by CI-930. Thus, cardiotonic-sensitive, cGMP-inhibitable, low Km cAMP PDE is present in both the soluble and particulate fractions of human cardiac left ventricular muscle of hearts from normal and cardiomyopathic subjects while the rolipram-sensitive PDE is present in the soluble fraction. The major differences in PDE activity of myopathic relative to normal left ventricular tissue are a reduced specific activity and Vmax of particulate PDE and one of the soluble peak III PDEs.

Reduced cardiac myofibrillar Mg-ATPase activity without changes in myosin isozymes in patients with end-stage heart failure.

In this study we tested the hypothesis that reduced myofibrillar ATPase activities in end-stage heart failure are associated with a redistribution of myosin isozymes. Cardiac myofibrils were isolated from left ventricular free wall from normal human hearts and hearts at end-stage heart failure caused by coronary artery diseases, cardiomyopathy or immunological rejection. The hearts had been excised in preparation for a heart transplant. Myofibrillar Ca2(+)-dependent Mg-ATPase and myosin Ca2(+)- and K+EDTA-ATPase activities were compared. Possible changes in myosin isozyme distribution in the diseased heart were investigated using polyacrylamide gel electrophoresis of native myosin in the presence of pyrophosphate. Significant reduction in myofibrillar Ca2(+)-dependent Mg-ATPase with no changes in the sensitivity of the myofibrils to Ca2+ was observed in heart with coronary artery diseases (25.2 to 27.1% at pCa 5.83 to pCa 5.05), cardiomyopathy (21.1 to 25.5% at pCa 5.41 to pCa 5.05), and in the immunologically rejected heart (18.4 to 22.8% at pCa 5.41 to pCa 5.05). Significantly lower myosin Ca2(+)-ATPase was observed with coronary artery diseases only and myosin K-EDTA activities did not differ in diseased and normal hearts. Polyacrylamide gel electrophoresis of native myosin from the normal and three models of end-stage heart failure revealed two distinct bands in the human left ventricle and one diffuse band in the human right atria. No apparent differences in myosin isoenzyme pattern were observed between the normal and diseased hearts. Further evaluation is needed to clarify the ATPase nature of the two bands.