Tuberous sclerosis complex (TSC) is an autosomal dominant disorder characterized by multiple dysplastic organ lesions and neuropsychiatric symptoms, caused by loss of function mutations in either TSC1 or TSC2. Genotype and phenotype analyses are conducted worldwide, but there have been few large-scale studies on Japanese patients, and there are still many unclear points. This study analyzed 283 Japanese patients with TSC (225 definite, 53 possible, and 5 genetic diagnoses). A total of 200 mutations (64 TSC1, 136 TSC2) were identified, of which 17 were mosaic mutations, 11 were large intragenic deletions, and four were splicing abnormalities due to deep intronic mutations. Several lesions and symptoms differed in prevalence and severity between TSC1 and TSC2 patients and were generally more severe in TSC2 patients. Moreover, TSC2 missense and in-frame mutations may attenuate skin and renal symptoms compared to other TSC2 mutations. Genetic testing revealed that approximately 20% of parents of a proband had mild TSC, which could have been missed. The patient demographics presented in this study revealed a high frequency of TSC1 patients and a low prevalence of epilepsy compared to global statistics. More patients with mild neuropsychiatric phenotypes were diagnosed in Japan, seemingly due to a higher utilization of brain imaging, and suggesting the possibility that a significant amount of mild TSC patients may not be correctly diagnosed worldwide.

DNA methylation of tumor suppressor genes in cancer is known to be a mechanism for silencing gene expression, but much remains unknown about its extent and relationship to somatic variants at the DNA sequence level. In this study, we comprehensively analyzed DNA methylation and somatic variants of all gene regions across the genome of the major tumor suppressor genes, APC, TP53, SMAD4, and mismatch repair genes in colorectal cancer using a novel next-generation sequencing-based analysis method. The Targeted Methyl Landscape (TML) shows that DNA hypermethylation patterns of these tumor suppressor genes in colorectal cancer are more complex and widespread than previously thought. Extremely high levels of DNA methylation were observed in relatively long regions around exon 1A of APC and exon 1 and surrounding region of MLH1. DNA hypermethylation occurred whether or not somatic DNA variants were present in the tumor. Even in tumors where the loss of heterozygosity has been demonstrated by somatic variants alone, additional methylation of the same gene can occur. Our data demonstrate that somatic variants and hypermethylation of these tumor suppressor genes were considered independent, parallel events, not exclusive of each other or having one event affecting the other.

DNA methylation of tumor suppressor genes in cancer is known to be a mechanism for silencing gene expression, but much remains unknown about its extent, degree, and rela-tionship to somatic variants at the DNA sequence level. In this study, we comprehensively analyzed DNA methylation and somatic variants of all gene regions across the genome of the major tumor suppressor genes, APC, TP53, SMAD4, and mismatch repair genes in colorectal cancer using a novel next-generation sequencing-based analysis method. The Targeted Methyl Landscape (TML) shows that DNA hypermethylation patterns of these tumor suppressor genes in colorectal cancer are more complex and widespread than pre-viously thought. Extremely high levels of DNA methylation were observed in relatively long regions around exon 1A of APC and exon 1 and surrounding region of MLH1. DNA hypermethylation occurred either alone in tumors without somatic variants or in tumors with somatic variants indicating loss of heterozygosity that were additionally methylated at the same genes. Our data demonstrate that somatic variants and hypermethylation of these tumor suppressor genes were considered independent, parallel events, not exclusive of each other or one event affecting the other.

RNA sequencing (RNA-Seq) is a powerful technique for the quantification of transcripts and the analysis of alternative splicing. Previously, our laboratory developed the targeted RNA long-amplicon sequencing (rLAS) method, which has the advantage of allowing deep analysis of targeted specific transcripts. The computational tools for analyzing RNA-Seq data have boosted alternative splicing research by detecting and quantifying splicing events. However, the performance of these splicing tools has not yet been investigated for rLAS. Here, we evaluated the performance of four splicing tools (MAJIQ, rMATS, MISO, and SplAdder) using samples with different types of known splicing events (exon-skipping, multiple-exon-skipping, alternative 5′ splicing, and alternative 3′ splicing). MAJIQ was able to detect all of the types of events, but it was unable to detect one of the exon-skipping events. On the other hand, rMATS was able to detect all of the exon-skipping events. However, rMATS failed to detect other types of events besides exon-skipping events. Both MISO and SplAdder were unable to detect any of the events. These results indicate that MAJIQ presents better performance for the different types of splicing events in rLAS and that rMATS shows better performance for exon-skipping splicing events.