BackgroundSince the report on the association between FLG mutations and atopic dermatitis (AD) (1), genetic diversity of FLG mutations has been intensively investigated across ethnicities. However, except for well-established racial difference between European and Asian (2), detailed analysis of the genetic diversity and clinical implications of FLG mutations for East Asian ancestry was not performed yet.
Questions addressedThe purpose of this study was, first, to overview genetic landscape of FLG mutations across East Asia and to analyse geographic distribution from genealogic point of view, especially prehistoric human migration in East Asia. Second, we investigated the association of FLG mutations with clinical features of AD to evaluate the clinical significance of FLG mutations.
Experimental designWe collected data reporting FLG mutations among East Asian populations, including recently completed full sequencing of FLG for Koreans in our laboratory. In total, 1384 AD cases and 1031 normal controls were identified for three East Asian populations at six locations: Chinese Han/Singapore, Chinese Han/Shanghai, Chinese Han/Shanxi, Korean, Japanese/Kyushu and Japanese/ mainland (Table S1). From mutational diversity, genetic distances were calculated and visualized as phylogenetic trees. Ancestryrelated population structures were inferred by clustering software, STRUCTURE (3). To assess the significance of FLG mutations as clinical biomarkers, statistical tests were performed, and Review Manager (RevMan) (4) was employed for population pooled analysis and graphical presentation of results. Detailed information on 'Materials and Methods' can be found in the (Data S1).
ResultsThe geographic distribution of FLG mutations is summarized in Fig. 1a, with several notable patterns. First, c.3321delA was re-confirmed as a pan-Asian mutation, found in all populations. Meanwhile, some mutations showed south-to-north (or north-to-south) distribution gradient: p.K4022X, the most prevalent FLG mutation in northern China and Korea, declined in frequency moving southward; in contrast, c.6950del8 (p.Q2417X, p.E2422X, etc.) showed the reverse. Finally, several mutations (p.S2554X/p.S2889X/p.S329 6X/Q1701X) were Japanese-specific and not found elsewhere.Phylogenetic tree clearly showed separation of Asian from European (Irish) and branching of Japanese, Korean, northern/southern Chinese Han successively (Fig. 1b, far-left). STRUCTURE analysis revealed that at least three ancestries contributed to the current genetic make-up of East Asian AD patients: south-to-north, northto-south and Japanese-specific ancestries (Fig. 1b,. These three ancestries were replicated in other genetic data, such as autosomal (Fig. 1b Table S1 for details). Hypothetical human migration routes are depicted by blue arrows. data sets (Fig. 1b, far-right) (6). For the presence of Japanese-specific FLG mutations, so-called dual origin of Japanese could provide possible explanations. Historic/archaeological data suggested two ancestries for current Japanese: Jom...