Common variants in DLG1 locus are associated with non-syndromic cleft lip with or without cleft palate.

Non-syndromic cleft lip with or without cleft palate (nsCL/P) is a common craniofacial anomaly with a complex and heterogeneous aetiology. Knowledge regarding specific genetic factors underlying this birth defect is still not well understood. Therefore, we conducted an independent replication analysis for the top-associated variants located within the DLG1 locus at chromosome 3q29, which was identified as a novel cleft-susceptibility locus in our genome-wide association study (GWAS). Mega-analysis of the pooled individual data from the GWAS and replication study confirmed that common DLG1 variants are associated with the risk of nsCL/P. Two single nucleotide polymorphisms (SNPs), rs338217 and rs7649443, were statistically significant even at the genome-wide level (Ptrend = 9.70E-10 and Ptrend = 8.96E-09, respectively). Three other SNPs, rs9826379, rs6805920 and rs6583202, reached a suggestive genome-wide significance threshold (Ptrend < 1.00E-05). The location of the strongest individual SNP in the intronic sequence of the gene encoding DLG1 antisense RNA suggests that the true causal variant implicated in the risk of nsCL/P may affect the DLG1 gene expression level rather than structure of the encoded protein. In conclusion, we identified a novel cleft-susceptibility locus at chromosome 3q29 with a DLG1 as a novel candidate gene for this common craniofacial anomaly.

Confirming genes influencing risk to cleft lip with/without cleft palate in a case-parent trio study.

A collection of 1,108 case-parent trios ascertained through an isolated, nonsyndromic cleft lip with or without cleft palate (CL/P) was used to replicate the findings from a genome-wide association study (GWAS) conducted by Beaty et al. (Nat Genet 42:525-529, 2010), where four different genes/regions were identified as influencing risk to CL/P. Tagging SNPs for 33 different genes were genotyped (1,269 SNPs). All four of the genes originally identified as showing genome-wide significance (IRF6, ABCA4 and MAF, plus the 8q24 region) were confirmed in this independent sample of trios (who were primarily of European and Southeast Asian ancestry). In addition, eight genes classified as 'second tier' hits in the original study (PAX7, THADA, COL8A1/FILIP1L, DCAF4L2, GADD45G, NTN1, RBFOX3 and FOXE1) showed evidence of linkage and association in this replication sample. Meta-analysis between the original GWAS trios and these replication trios showed PAX7, COL8A1/FILIP1L and NTN1 achieved genome-wide significance. Tests for gene-environment interaction between these 33 genes and maternal smoking found evidence for interaction with two additional genes: GRID2 and ELAVL2 among European mothers (who had a higher rate of smoking than Asian mothers). Formal tests for gene-gene interaction (epistasis) failed to show evidence of statistical interaction in any simple fashion. This study confirms that many different genes influence risk to CL/P.

The MSH2 c.388_389del mutation shows a founder effect in Portuguese Lynch syndrome families.

The MSH2 c.388_389del mutation has occasionally been described in Lynch families worldwide. At the Portuguese Oncology Institute in Porto, Portugal, we have identified 16 seemingly unrelated families with this germline mutation. To evaluate if this alteration is a founder or a recurrent mutation we performed haplotype analysis in the 16 Portuguese index cases and 55 relatives, as well as in four index cases and 13 relatives reported from Germany, Scotland, England, and Argentina. In the Portuguese families we observed a shared haplotype of approximately 10 Mb and all were originated from the north of Portugal. These results suggest that this alteration is a founder mutation in Portugal with a relatively recent origin. In the reported families outside Portugal with this mutation different haplotype backgrounds were observed, supporting the hypothesis that it occurred de novo on multiple occasions. We also conclude that the high proportion of families with the MSH2 c.388_389del mutation indicates that screening for this alteration as a first step may be cost-effective in the genetic testing of Lynch syndrome suspects of Portuguese ancestry, especially those originating from the north of Portugal.

Evidence for a polygenic contribution to androgenetic alopecia.

BACKGROUND: Male pattern baldness (androgenetic alopecia, AGA) is a highly heritable trait and the most common form of hair loss in humans. Eight genome-wide significant risk loci for AGA have been identified. OBJECTIVES: To determine whether a polygenic component contributes to the genetic risk for AGA. METHODS: This study used a German case-control sample for AGA, which comprised 581 severely affected patients and 617 controls, to determine the contribution of polygenic variance to AGA risk. The sample was divided evenly into discovery and test samples. An additive polygenic risk score was calculated from risk alleles with increasingly liberal P-values in the discovery dataset, which was then used to test for the enrichment of AGA risk score alleles in the independent test samples. RESULTS: The polygenic score analysis provided significant evidence for a polygenic contribution to AGA where the amount of variance explained was 1·4-4·5%. CONCLUSION: This study provides evidence for the specific contribution of a polygenic component to the overall heritable risk for AGA. To some degree, the polygenic architecture of AGA might reflect the complexity of the biological pathways involved. Further analyses and strategies that complement conventional genome-wide association studies are needed to identify these factors. These may include pathway-based analyses, the analysis of functional candidate genes and tests for epistatic effects with known loci.

MiRNA-149 as a Candidate for Facial Clefting and Neural Crest Cell Migration.

Nonsyndromic cleft lip with or without palate (nsCL/P) ranks among the most common human birth defects and has a multifactorial etiology. Human neural crest cells (hNCC) make a substantial contribution to the formation of facial bone and cartilage and are a key cell type in terms of nsCL/P etiology. Based on increasing evidence for the role of noncoding regulatory mechanisms in nsCL/P, we investigated the role of hNCC-expressed microRNAs (miRNA) in cleft development. First, we conducted a systematic analysis of miRNAs expressed in human-induced pluripotent stem cell-derived hNCC using Affymetrix microarrays on cell lines established from 4 unaffected donors. These analyses identified 152 candidate miRNAs. Based on the hypothesis that candidate miRNA loci harbor genetic variation associated with nsCL/P risk, the genomic locations of these candidates were cross-referenced with data from a previous genome-wide association study of nsCL/P. Associated variants were reanalyzed in independent nsCL/P study populations. Jointly, the results suggest that miR-149 is implicated in nsCL/P etiology. Second, functional follow-up included in vitro overexpression and inhibition of miR-149 in hNCC and subsequent analyses at the molecular and phenotypic level. Using 3'RNA-Seq, we identified 604 differentially expressed (DE) genes in hNCC overexpressing miR-149 compared with untreated cells. These included TLR4 and JUNB, which are established targets of miR-149, and NOG, BMP4, and PAX6, which are reported nsCL/P candidate genes. Pathway analyses revealed that DE genes were enriched in pathways including regulation of cartilage development and NCC differentiation. At the cellular level, distinct hNCC migration patterns were observed in response to miR-149 overexpression. Our data suggest that miR-149 is involved in the etiology of nsCL/P via its role in hNCC migration.

MRPL53, a New Candidate Gene for Orofacial Clefting, Identified Using an eQTL Approach.

A valuable approach to understand how individual and population genetic differences can predispose to disease is to assess the impact of genetic variants on cellular functions (e.g., gene expression) of cell and tissue types related to pathological states. To understand the genetic basis of nonsyndromic cleft lip with or without cleft palate (NSCL/P) susceptibility, a complex and highly prevalent congenital malformation, we searched for genetic variants with a regulatory role in a disease-related tissue, the lip muscle (orbicularis oris muscle [OOM]), of affected individuals. From 46 OOM samples, which are frequently discarded during routine corrective surgeries on patients with orofacial clefts, we derived mesenchymal stem cells and correlated the individual genetic variants with gene expression from these cultured cells. Through this strategy, we detected significant cis-eQTLs (i.e., DNA variants affecting gene expression) and selected a few candidates to conduct an association study in a large Brazilian cohort (624 patients and 668 controls). This resulted in the discovery of a novel susceptibility locus for NSCL/P, rs1063588, the best eQTL for the MRPL53 gene, where evidence for association was mostly driven by the Native American ancestry component of our Brazilian sample. MRPL53 (2p13.1) encodes a 39S protein subunit of mitochondrial ribosomes and interacts with MYC, a transcription factor required for normal facial morphogenesis. Our study illustrates not only the importance of sampling admixed populations but also the relevance of measuring the functional effects of genetic variants over gene expression to dissect the complexity of disease phenotypes.

Candidate Genes for Nonsyndromic Cleft Palate Detected by Exome Sequencing.

Nonsyndromic cleft palate only (nsCPO) is a facial malformation that has a livebirth prevalence of 1 in 2,500. Research suggests that the etiology of nsCPO is multifactorial, with a clear genetic component. To date, genome-wide association studies have identified only 1 conclusive common variant for nsCPO, that is, a missense variant in the gene grainyhead-like-3 ( GRHL3). Thus, the underlying genetic causes of nsCPO remain largely unknown. The present study aimed at identifying rare variants that might contribute to nsCPO risk, via whole-exome sequencing (WES), in multiply affected Central European nsCPO pedigrees. WES was performed in 2 affected first-degree relatives from each family. Variants shared between both individuals were analyzed for their potential deleterious nature and a low frequency in the general population. Genes carrying promising variants were annotated for 1) reported associations with facial development, 2) multiple occurrence of variants, and 3) expression in mouse embryonic palatal shelves. This strategy resulted in the identification of a set of 26 candidate genes that were resequenced in 132 independent nsCPO cases and 623 independent controls of 2 different ethnicities, using molecular inversion probes. No rare loss-of-function mutation was identified in either WES or resequencing step. However, we identified 2 or more missense variants predicted to be deleterious in each of 3 genes ( ACACB, PTPRS, MIB1) in individuals from independent families. In addition, the analyses identified a novel variant in GRHL3 in 1 patient and a variant in CREBBP in 2 siblings. Both genes underlie different syndromic forms of CPO. A plausible hypothesis is that the apparently nonsyndromic clefts in these 3 patients might represent hypomorphic forms of the respective syndromes. In summary, the present study identified rare variants that might contribute to nsCPO risk and suggests candidate genes for further investigation.

Novel IRF6 Mutations Detected in Orofacial Cleft Patients by Targeted Massively Parallel Sequencing.

Common variants in interferon regulatory factor 6 ( IRF6) have been associated with nonsyndromic cleft lip with or without cleft palate (NSCL/P) as well as with tooth agenesis (TA). These variants contribute a small risk towards the 2 congenital conditions and explain only a small percentage of heritability. On the other hand, many IRF6 mutations are known to be a monogenic cause of disease for syndromic orofacial clefting (OFC). We hypothesize that IRF6 mutations in some rare instances could also cause nonsyndromic OFC. To find novel rare variants in IRF6 responsible for nonsyndromic OFC and TA, we performed targeted multiplex sequencing using molecular inversion probes (MIPs) in 1,072 OFC patients, 67 TA patients, and 706 controls. We identified 3 potentially pathogenic de novo mutations in OFC patients. In addition, 3 rare missense variants were identified, for which pathogenicity could not unequivocally be shown, as all variants were either inherited from an unaffected parent or the parental DNA was not available. Retrospective investigation of the patients with these variants revealed the presence of lip pits in one of the patients with a de novo mutation suggesting a Van der Woude syndrome (VWS) phenotype, whereas, in other patients, no lip pits were identified.

The p53 codon 72 variation is associated with the age of onset of hereditary non-polyposis colorectal cancer (HNPCC).

The polymorphic variants at codon 72 of the p53 gene were shown to be functionally distinct in vitro, whereby the arginine (arg) variant induces apoptosis more efficiently than the proline (pro) variant. From the evidence that the DNA mismatch repair system and p53 interact to maintain genomic integrity, we hypothesized that the codon 72 variation may influence the age of onset of disease in HNPCC patients. We tested 538 patients for p53 codon 72 variants, including 167 unrelated patients with pathogenic germline mutations in MSH2 or MLH1 and colorectal carcinoma as first tumour, 126 patients with sporadic microsatellite stable colorectal cancers, and 245 healthy controls. The median age of onset was 41, 36, and 32 years for MSH2 or MLH1 mutation carriers with arg/arg, arg/pro, and pro/pro genotypes, respectively. The log rank test revealed significant differences in the age of onset between arg/arg and pro/pro individuals (p = 0.0002) and in arg/pro versus arg/arg and pro/pro individuals (p = 0.0026 and p = 0.0217, respectively). A Cox regression model indicated an additive mode of inheritance. No significant differences in age of onset were observed among different genotype carriers with microsatellite stable tumours. Our results suggest that p53 codon 72 genotypes are associated with the age of onset of colorectal carcinoma in a mismatch repair deficient background in a dose dependent manner. These findings may be relevant for preventive strategies in HNPCC.

"Second hit" in sebaceous tumors from Muir-Torre patients with germline mutations in MSH2: allele loss is not the preferred mode of inactivation.

Muir-Torre syndrome is an autosomal-dominant inherited disorder predisposing to both sebaceous skin tumors and internal neoplasms. In a significant proportion of Muir-Torre syndrome patients skin tumors exhibit microsatellite instability as a hallmark of hereditary nonpolyposis colorectal cancer. Most individuals predisposed to hereditary nonpolyposis colorectal cancer harbor a germline mutation in the DNA mismatch repair genes MSH2 or MLH1. In Muir-Torre syndrome the vast majority of germline mutations have been identified in MSH2. Microsatellite instability in tumor tissue develops after somatic inactivation of the corresponding second mismatch repair allele ("second hit"). So far, the mechanisms of somatic inactivation of the second allele in microsatellite instability positive tumors from patients with known mismatch repair germline mutations are not well understood. We examined whether allele loss (loss of heterozygosity) is a frequent mechanism for inactivation of the second MSH2 allele in a sample of nine microsatellite instability positive skin tumors from eight unrelated Muir-Torre patients with known MSH2 germline mutations. Loss of heterozygosity was determined using microsatellite markers or heteroduplex analysis, respectively. Only one of the nine skin tumors exhibited loss of heterozygosity at the MSH2 locus. Thus, we could show in a sample of sebaceous tumors from patients with genetically proven Muir-Torre syndrome that loss of heterozygosity most probably is not the preferred mode of somatic inactivation of the second MSH2 allele.

DHPLC mutation analysis of the hereditary nonpolyposis colon cancer (HNPCC) genes hMLH1 and hMSH2.

Denaturing high-performance liquid chromatography (DHPLC) is an efficient method for detection of mutations involving a single or few numbers of nucleotides, and it has been successfully used for mutation detection in disease-related genes. Colorectal cancer is one of the most common cancers, and mutations in the genes for hereditary nonpolyposis colon cancer (HNPCC), hMLH1 and hMSH2, also involve mainly point mutations. Sequence analysis is supposed to be a screening method with high sensitivity; however, it is time-consuming and expensive. We therefore decided to test sensitivity and reproducibility of DHPLC for 71 sequence variants in hMLH1 and hMSH2 initially found by sequence analysis in DNA samples of German HNPCC patients. DHPLC conditions of the PCR products were based on the melting pattern of the wild-type sequence of the corresponding PCR fragments. All but one of the 71 mutations was detected using DHPLC (sensitivity of 97%). Running time per sample averaged only 7 min, and the system is highly automated. Thus DHPLC is a rapid and sensitive method for the detection of hMLH1 and hMSH2 sequence variants.

Strong association of variants around FOXE1 and orofacial clefting.

Nonsyndromic orofacial clefting (nsOFC) is a common, complex congenital disorder. The most frequent forms are nonsyndromic cleft lip with or without cleft palate (nsCL/P) and nonsyndromic cleft palate only (nsCPO). Although they are generally considered distinct entities, a recent study has implicated a region around the FOXE1 gene in both nsCL/P and nsCPO. To investigate this hypothesis, we analyzed the 2 most strongly associated markers (rs3758249 and rs4460498) in 2 independent samples of differing ethnicities: Central European (949 nsCL/P cases, 155 nsCPO cases, 1163 controls) and Mayan Mesoamerican (156 nsCL/P cases, 10 nsCPO cases, 338 controls). While highly significant associations for both single-nucleotide polymorphisms were obtained in nsCL/P (rs4460498: p Europe = 6.50 × 10(-06), p Mayan = .0151; rs3758249: p Europe = 2.41 × 10(-05), p Mayan = .0299), no association was found in nsCPO (p > .05). Genotyping of rs4460498 in 472 independent European trios revealed significant associations for nsCL/P (p = .016) and nsCPO (p = .043). A meta-analysis of all data revealed a genomewide significant result for nsCL/P (p = 1.31 × 10(-08)), which became more significant when nsCPO cases were added (p nsOFC = 1.56 × 10(-09)). These results strongly support the FOXE1 locus as a risk factor for nsOFC. With the data of the initial study, there is now considerable evidence that this locus is the first conclusive risk factor shared between nsCL/P and nsCPO.

5q31 Microdeletions: Definition of a Critical Region and Analysis of LRRTM2, a Candidate Gene for Intellectual Disability.

Microdeletions including 5q31 have been reported in only few patients to date. Apart from intellectual disability/developmental delay (ID/DD) of varying degrees, which is common to all reported patients, the clinical spectrum is wide and includes short stature, failure to thrive, congenital heart defects, encephalopathies, and dysmorphic features. We report a patient with a 0.9-Mb de novo deletion in 5q31.2, the smallest microdeletion in 5q31 reported thus far. His clinical presentation includes mild DD, borderline short stature, postnatal microcephaly, and mild dysmorphic signs including microretrognathia. Together with data from 7 reported overlapping microdeletions, analysis of our patient enabled the tentative delineation of a phenotype map for 5q31 deletions. In contrast to the mild phenotype of small microdeletions affecting only 5q31.2, carriers of larger microdeletions which also include subbands 5q31.1 and/or 5q31.3 seem to be more severely affected with congenital malformations, growth anomalies, and severe encephalopathies. A 240-kb smallest region of overlap in 5q31.2 is delineated which contains only 2 genes, CTNNA1 and LRRTM2. We propose LRRTM2 as the most promising candidate gene for ID/DD due to its expression pattern, function as a key regulator of excitatory development, and interaction with Neurexin 1. However, sequence analysis of LRRTM2 in 330 patients with ID/DD revealed no relevant alterations, excluding point mutations in LRRTM2 as a frequent cause of ID/DD in patients without microdeletions.

Coexisting somatic promoter hypermethylation and pathogenic MLH1 germline mutation in Lynch syndrome.

Somatic epimutations in the MLH1 promoter mimic the phenotype of Lynch syndrome. To date, no somatic hypermethylation of the MLH1 promoter in the carrier of a pathogenic MLH1 germline mutation has been identified, prompting the recommendation that a germline mutation in MLH1 should only be sought in the absence of tumour tissue methylation. We aimed to determine whether methylation of the MLH1 promoter may coexist in carriers of a pathogenic germline mutation in MLH1. We examined the methylation status of the MLH1 promoter in 123 tumour tissue samples, demonstrating high microsatellite instability and loss of expression of a mismatch repair protein (60 cases with MLH1 germline mutation, 25 cases without mutation, 38 cases with MSH2 mutations), using combined bisulphite restriction analysis (COBRA) and SNaPshot analysis. Methylation of the MLH1 promoter was found in two patients with pathogenic germline mutations, one a carrier of a MLH1 mutation and the other a carrier of a MSH2 mutation. Our results demonstrate that methylation of the MLH1 promoter region does not exclude the presence of a germline mutation in a mismatch repair (MMR) gene. Hypermethylation of the MLH1 promoter may be present in most cases of sporadic colorectal cancers, but this does not exclude a diagnosis of Lynch syndrome.

Location and type of mutation in the LIS1 gene do not predict phenotypic severity.

BACKGROUND: Lissencephaly is a neuronal migration disorder leading to absent or reduced gyration and a broadened but poorly organized cortex. The most common form of lissencephaly is isolated, referred as classic or type 1 lissencephaly. Type 1 lissencephaly is mostly associated with a heterozygous deletion of the entire LIS1 gene, whereas intragenic heterozygous LIS1 mutations or hemizygous DCX mutations in males are less common. METHODS: Eighteen unrelated patients with type 1 lissencephaly were clinically and genetically assessed. In addition, patients with subcortical band heterotopia (n = 1) or lissencephaly with cerebellar hypoplasia (n = 2) were included. RESULTS: Fourteen new and seven previously described LIS1 mutations were identified. We observed nine truncating mutations (nonsense, n = 2; frameshift, n = 7), six splice site mutations, five missense mutations, and one in-frame deletion. Somatic mosaicism was assumed in three patients with partial subcortical band heterotopia in the occipital-parietal lobes or mild pachygyria. We report three mutations in exon 11, including a frameshift which extends the LIS1 protein, leading to type 1 lissencephaly and illustrating the functional importance of the WD domains at the C terminus. Furthermore, we present two patients with novel LIS1 mutations in exon 10 associated with lissencephaly with cerebellar hypoplasia type a. CONCLUSION: In contrast to previous reports, our data suggest that neither type nor position of intragenic mutations in the LIS1 gene allows an unambiguous prediction of the phenotypic severity. Furthermore, patients presenting with mild cerebral malformations such as subcortical band heterotopia or cerebellar hypoplasia should be considered for genetic analysis of the LIS1 gene.

Frequency of hereditary non-polyposis colorectal cancer among unselected patients with colorectal cancer in Germany.

INTRODUCTION: Hereditary non-polyposis colorectal cancer (HNPCC) is a major form of familial colorectal cancer (CRC). It is diagnosed when either the Amsterdam criteria (AC) are fulfilled or mutations in one of the mismatch repair (MMR) genes have been identified. This project aims at estimating the proportion of HNPCC among unselected patients with CRC. PATIENTS AND METHODS: During a period of 2 years, a total of 351 non-selected patients with CRC were registered prospectively. 92 patients met the Bethesda criteria (9 of them fulfilled the AC) and 259 did not. 348 tumours were examined for microsatellite instability (MSI) and expression of MMR proteins. RESULTS: MSI-H and MSI-L were identified in 17 and 6%, respectively. Loss of MSH2 or MLH1 was found in 1.5 and 8.8%, respectively. Based on the results of tumour tissue analyses, 80 patients with MSI and/or loss of MSH2 or MLH1 expression were identified as candidates for germline mutation screening. DNA of 40/80 patients was available. These patients were screened for MSH2 and MLH1 mutations; 19/40 patients with MSI and normal MSH2 or MLH1 expression were screened for mutations in MSH6. Three patients had relevant MMR gene mutations and six variants of unknown functional relevance were detected. CONCLUSIONS: After adjusting for the cases not evaluable for germline mutations, 1.7% of the CRC patients had HNPCC proven by molecular genetics.