Exome Sequencing Identifies A Causal Gene Of Disseminated Superficial Actinic Porokeratosis

Background DSAP is the most common subtype of porokeratosis, characterized by multiple small, annular, anhidrotic, keratotic lesions predominantly on sun-exposed areas of the skin. An atrophic centre and prominent peripheral ridge, which is recognized as a compact column of keratin called the cornoid lamella, a histopathological hallmark for this group of disorders. Some studies showed that genetics, UVA and immunodepression are involoved in the pathogenesis of porokeratosis. DSAP is transmitted in an autosomal dominant mode. Five linkage loci for DSAP (12q23.2-24.1,12q24.1-q24.2,15q25.1-26.1,1p31.3-p31.1and16q24.1-24.3) were reported previously. By candidate gene sequencing, although sequence variants in SSH1and SART3have been deected in DSAP families, the findings was not supported by subsequent studies in additional patients with DSAP. The genetic basis and the pathogenesis of the DSAP remain unclear.Objective (1) To identify the causal gene of DSAP by exome sequencing in2cases and1control in one DSAP family combining with step by step filtering and additional DSAP validation.(2) To explore the possible pathogenesis of DSAP by function studies.Methods (1) We collected all the cases and controls from the Chinese Han population through collaboration with other two hospitals. After quality controls,115DSAP cases,31controls in DSAP families,11cases with other clinical forms of porokeratisis were selected in our study.(2) We performed exome capture by Agilent SureSelect Human All Exon Kit followed by massively parallel sequencing (Hiseq2000) in two affected individuals and one unaffected individual from a family with DSAP (Family1). Raw image files were processed by Illumina basecalling Software1.7. Sequence reads in each individual were aligned to human reference genome (NCBI build36.3, hgl8) using SOAPaligner2.20.(3) The consensus genotypes in target regions were called by SOAPsnp (v1.03).A consensus genotype with Phred-like quality of at least20and at least four coverage depth was considered as the high-confident genotype. The genotypes that were different from the reference were extracted as candidate SNPs. For indels calling, we switched to GATK pipline.(4) The variants were functionally annotated using an in-house pipeline and categorized into missense, nonsense, readthrougth, splice-site mutations and coding indels, synonymous and non-coding mutations. Based on these annotations, variants were filtered first for the nonsysnonymous (NS), splice acceptor and donor site (SS) mutations and coding indels (I), and then filtered against available public databases (dbSNP129, eight HapMap exomes and1000Genome variants databases) and the control step by step. At last the variants which were shared between two cases but were neither present in the public databases nor in the control individual of the family were considered to be the candidate variants. Then, we selected possible DSAP causal variants using ANNOVAR and GERP to assess these variants for likely functional impact combing with the reported linkage regions of DSAP.(5) We sequenced the exons containing possible DSAP causal variants in all available individuals in familyl (6case and3controls)(6) We further sequenced all the exons and exon-intron boundaries of the gene containing the variant which was cosegregated with DSAP in famiy1in additional DSAP individuals.(7) We filtered all the mutations against676unrelated, ethnically matched controls.(8) To further investigate whether MVK mutation was associated with the other clinical subtypes of PK, we sequenced all the exons and exon-intron boundaries of MVK in5individuals with porokeratosis of Mibelli,2individuals with linear porokeratosis and4individuals with disseminated superficial porokeratosis (9) We performed immunohistochemistry to explore the expression of MVK on skin.(10) KCs were transiently transfected with the empty vector (pcDNA) or the wild MVK eukaryotic expression (Wild MVK) or MVK shRNA vector (MVK shRNA) or Non-targeting shRNA vector (Non-targeting control) for48h to build cell model with MVK knockdown and overexpression. Then, we detected the influence of MVK gene expression on KCs proliferation, differentiation and apoptosis by Flow Cytometry, Real-time Quantitative PCR and Western-blotting.(11) In order to explore whether there is secondary somatic mutations involved in the development of DSAP, we obtained KCs from the lesions of5DSAP patients by microdissection and sequenced all exons and exon-intron boundaries of MVK.Results (1) After exome capture, massively parallel sequencing, mapping and filtering step by step, We selected104candidate variants which were shared by the two affected individuals but not present in the non-affected individual from the family and available variants databases. Among these, only one heterozygous variant, found in MVK (c.764T>C), was located within the regions linked to DSAP identified by previous linkage studies and to be probably damaging as predicted by both two methods (ANNOVAR and GERP).(2) All the six affected individuals were heterozygous for this mutation which was absent in the three unaffected family members in family1.(3) We identified13additional heterozygous mutations in MVK in22DSAP patients (18probands and4sporadic cases). None of these mutations was detected in676unrelated, ethnically matched controls, which provided evidence that they were not polymorphisms.(4) No mutations within MVK were found in11individuals with other clinical subtypes of porokeratosis.(5) We found the cytoplasmic expression of MVK protein in the epidermis keratinocytes.(6) We found that the overexpression of wild type MVK protein was associated with the increased expression of KCs differentiation marker Kl and involucrin. Consistently, the knockdown of MVK expression by shRNA in KCs showed a corresponding decrease in the expression of K1and involucrin. The expression of K5did not show any significant change. The PI staining FACS results suggested that the proportion of cells in S phase decreased slightly after MVK knockdown and increased slightly after MVK overexpression, but not statistically significant. The caspase3activity significantly increased after MVK knockdown and significantly decreased after MVK overexpression which was consistent with the results from early apoptosis FACS. The overexpression of MVK could protect KCs from UVA-caused apoptosis.(7) All the germline heterogenous mutations were confirmed in5lesions, but no secondary somatic mutation was found.Conclusion (1) We identified a causal gene of DSAP by exome sequencing in3individuals from one DSAP family and additional DSAP samples validation.(2) The expression of MVK was associated with KCs differentiation and apoptosis, but no significant correlation with KCs proliferation.(3) The results should help advance our understanding on the pathogenesis of DSAP and extend the pleiotropic functional spectrum of the MVK on disease development.