Does SNAI2 mutation cause human piebaldism and Waardenburg syndrome?

To the Editor: We and others previously reported that rare cases of human piebaldism (OMIM #172800) (Sánchez-Martín et al., 2003) and Waardenburg syndrome (WS2D; OMIM #608890) (Sánchez-Martín et al., 2002) can result from deletions of the SNAI2 gene. Here, we call into question those conclusions, which were based on older technologies and do not stand up to re-assessment. Both piebaldism andWaardenburg syndromes are genetically determined neurocristopathies that include pigmentation anomalies such as congenital patchy leukoderma and poliosis (Spritz & Hearing, 2013). SNAI2, also known as SLUG, was suggested as a candidate gene based on its identification as a downstream molecular target of the KITsignaling pathway and the phenotype of patchy hypopigmentation in SNA-mutant mice (Pérez-Losada et al., 2002). In humans, piebaldism is usually caused by pathogenic variants in the KIT protooncogene (Spritz & Hearing, 2013). The association of SNAI2 with human piebaldism (OMIM #172800) is based on a single, 17-year-old report in this journal, which included our patient (Sánchez-Martín et al., 2003). His childhood diagnosis of “vitiligo” was corrected to piebaldism at age 22 on the basis of his characteristic skin phenotype and two affected brothers. He was included in an early research study (Case 13) (Ezoe et al., 1995) and clinical report (Case 3) (Ward, Moss, & Sanders, 1995) of patients with classical piebaldism, and was one of the few in whom no KIT mutation was found by SSCP/heteroduplex screening of amplified PCR products. His DNA was then sent for analysis of the SNAI2 gene; quantitative Southern blotting and quantitative RT-PCR indicated a heterozygous SNAI2 deletion in our patient and two others from a series of 17 unrelated patients with piebaldism and no apparent pathogenic variants in KIT (Sánchez-Martín et al., 2003). Subsequently, our patient went on to have four children, three of whom were affected with piebaldism, and in 2013 the genetic result was checked for counseling purposes by a commercial laboratory. No SNAI2 mutation or deletion was found by MLPA. Finally, our patient's family was recruited to the UK 100,000 Genomes Project (Genomics England, 2020), and in 2019 whole-genome sequencing identified a novel heterozygous KIT (NM_000222) variant c.2665A>G, p.Met889Val in our patient and his affected children, which was subsequently confirmed in the extended family. Based on reported genotype–phenotype correlations, this variant is likely pathogenic (Spritz & Hearing, 2013). To our knowledge, the two other piebaldism patients previously thought to have SNAI2 deletions have not been retested. In a 2014 report in this journal (Yang et al., 2014) attributed a piebald-like phenotype in a Chinese girl to a heterozygous mutation in SNAI2. However, both the phenotype and this molecular attribution seem questionable. The patient is female, carrying a clinical diagnosis of severe XLHED, and she has a pathogenic mutation in EDA1. Although she has a small segment of asymmetric ventral chest leukoderma, she lacks other typical features of piebaldism, including poliosis and other regions of leukoderma with hyperpigmented macules within hypopigmented areas. No details were given of family history, age-of-onset or clinical course, and a diagnosis of segmental vitiligo cannot be excluded. DNA sequencing additionally demonstrated two adjacent variants in the 50 UTR of SNAI2. One is a rare nonpathogenic variant, rs1289888558 (NC_000008.11:g.48921413C>A), which is not associated with piebaldism, and the other is apparently unique, NC_000008.11: g.48921414G>T. The parents were not investigated; thus, phase of these two variants cannot be ascertained. The authors' assertion that they might affect RNA splicing is inconsistent with DNA strand orientation. Thus, pathogenicity of even the novel SNAI2 variant is not clear in this patient. A homozygous deletion of SNAI2 was also reported in two patients with Waardenburg syndrome (Sánchez-Martín et al., 2002), now termed WS2D (OMIM #608890), also based on quantitative Southern blotting and quantitative RT-PCR. However, Dr. Andrew Read (personal communication) is no longer convinced that this result was a correct finding and not a technical artifact. Finally, recently a patient was reported with a heterozygous 1.7 Mb 8q11.1q11.21 deletion that includes the entire SNAI2 gene (ClinVar accession VCV000662275), detected by whole-exome sequencing and MLPA. The clinical phenotype in this case does not include pigmentation abnormalities, deafness, or dysmorphia that would be suggestive of either piebaldism or Waardenburg syndrome (unpublished data). Thus, this case would seem to contradict the assertions that deletion of SNAI2 causes either piebaldism or Waardenburg syndrome. It took more than 25 years to discover the KIT mutation in our patient with classical piebaldism. Over that period, there were remarkable advances in our ability to detect and define pathogenic variants, deletions, and other genomic rearrangements. Conclusions based on long-obsolete technology require reconsideration. Based on the above evidence we now doubt the claimed associations of SNAI2 deletions with piebaldism and Waardenburg syndrome. Received: 1 August 2020 Revised: 1 September 2020 Accepted: 5 September 2020