On 2017 Jun 10, Akihiro Umezawa commented:

Thank you so much, Alistair. I also received the same comment from a chinese scientist. The authors totally agree with your comment. We will investigate phenotypes of the iPSC from the viewpoint of your comment 'Compound heterozygous mutations of the ERCC2 gene'. Some of the phenotypes could be linked to the gene functions.

On 2017 Jun 05, Alistair Pagnamenta commented:

Very interesting study and great that VCFs were freely available to review. Although my primary interest was to look at iPS induced artefacts (incl. UPD9), as the study “failed to find causal mutations in the XP-related genes”, I also searched for variants that might be responsible for the XP phenotype. Amongst the 300+ high-confidence, rare variants that were predicted potentially deleterious (https://variants.ingenuity.com/XP40OS), there were two in ERCC2, a gene which encodes a subunit of the TFIIH core complex helicase and which is linked to xeroderma pigmentosum complementation group D, XPD (OMIM *126340):

• a c.2048G>A; p.(Arg683Gln) in exon 22 which is listed in HGMD and ClinVar as a disease causing mutation (CM970443; RCV000248679.1). The variant is in gnomAD database present on 4/246,000 chromosomes.

• a 23bp deletion (c.2025_2046+1delCCTCATGGTCTTTGCCGACAAGG) which removes the end of the preceding exon. As indels aren’t robustly called from exome data, we are typically wary about reporting such variants without having viewed read alignments and/or having validated them with another method. However in this case, the deletion passes a number of confidence filters and is present in gnomAD on 1/246,108 chromosomes in an E Asian sample.

As these variants are both heterozygous, it remains to be confirmed that they are found in trans. However, given the loci are 153bp apart, the alleles can likely be phased by using Illumina read-pair information. Assuming compound-heterozygosity can be demonstrated, these variants represent plausible candidates to explain the condition.

The fibroblast cell line (XP40OS) was originally obtained from the JCRB cell bank catalogue which lists it as belonging to complementation group C and not group D. The stock cell line should be retested, either by the original polyethylene glycol-induced cell fusion based method (Sato K, 1982), or by molecular analysis of the above loci. Based on a review of submissions to major cell repositories, it has been estimated that 18-36% of cell lines are misidentified or contaminated (Hughes P, 2007). While in Okamura K, 2015, the mislabelling was relatively minor and doesn’t affect the overall conclusions, in other situations misidentified cell lines can lead to the inability to replicate scientific results and is a drain on scientific funding.

On 2017 Jun 05, Alistair Pagnamenta commented:

Very interesting study and great that VCFs were freely available to review. Although my primary interest was to look at iPS induced artefacts (incl. UPD9), as the study “failed to find causal mutations in the XP-related genes”, I also searched for variants that might be responsible for the XP phenotype. Amongst the 300+ high-confidence, rare variants that were predicted potentially deleterious (https://variants.ingenuity.com/XP40OS), there were two in ERCC2, a gene which encodes a subunit of the TFIIH core complex helicase and which is linked to xeroderma pigmentosum complementation group D, XPD (OMIM *126340):

• a c.2048G>A; p.(Arg683Gln) in exon 22 which is listed in HGMD and ClinVar as a disease causing mutation (CM970443; RCV000248679.1). The variant is in gnomAD database present on 4/246,000 chromosomes.

• a 23bp deletion (c.2025_2046+1delCCTCATGGTCTTTGCCGACAAGG) which removes the end of the preceding exon. As indels aren’t robustly called from exome data, we are typically wary about reporting such variants without having viewed read alignments and/or having validated them with another method. However in this case, the deletion passes a number of confidence filters and is present in gnomAD on 1/246,108 chromosomes in an E Asian sample.

As these variants are both heterozygous, it remains to be confirmed that they are found in trans. However, given the loci are 153bp apart, the alleles can likely be phased by using Illumina read-pair information. Assuming compound-heterozygosity can be demonstrated, these variants represent plausible candidates to explain the condition.

The fibroblast cell line (XP40OS) was originally obtained from the JCRB cell bank catalogue which lists it as belonging to complementation group C and not group D. The stock cell line should be retested, either by the original polyethylene glycol-induced cell fusion based method (Sato K, 1982), or by molecular analysis of the above loci. Based on a review of submissions to major cell repositories, it has been estimated that 18-36% of cell lines are misidentified or contaminated (Hughes P, 2007). While in Okamura K, 2015, the mislabelling was relatively minor and doesn’t affect the overall conclusions, in other situations misidentified cell lines can lead to the inability to replicate scientific results and is a drain on scientific funding.

On 2017 Jun 10, Akihiro Umezawa commented:

Thank you so much, Alistair. I also received the same comment from a chinese scientist. The authors totally agree with your comment. We will investigate phenotypes of the iPSC from the viewpoint of your comment 'Compound heterozygous mutations of the ERCC2 gene'. Some of the phenotypes could be linked to the gene functions.