BackgroundFace morphology is strongly determined by genetic factors. However, only a small number of genes related to face morphology have been identified to date. Here, we performed a two-stage genome-wide association study (GWAS) of 85 face morphological traits in 7569 Koreans (5643 in the discovery set and 1926 in the replication set).ResultsIn this study, we analyzed 85 facial traits, including facial angles. After discovery GWAS, 128 single nucleotide polymorphisms (SNPs) showing an association of P < 5 × 10− 6 were selected to determine the replication of the associations, and meta-analysis of discovery GWAS and the replication analysis resulted in five genome-wide significant loci. The OSR1-WDR35 [rs7567283, G allele, beta (se) = −0.536 (0.096), P = 2.75 × 10− 8] locus was associated with the facial frontal contour; the HOXD1-MTX2 [rs970797, A allele, beta (se) = 0.015 (0.003), P = 3.97 × 10− 9] and WDR27 [rs3736712, C allele, beta (se) = 0.293 (0.048), P = 8.44 × 10− 10] loci were associated with eye shape; and the SOX9 [rs2193054, C allele, beta (se) (ln-transformed) = −0.007 (0.001), P = 6.17 × 10− 17] and DHX35 [rs2206437, A allele, beta (se) = −0.283 (0.047), P = 1.61 × 10− 9] loci were associated with nose shape. WDR35 and SOX9 were related to known craniofacial malformations, i.e., cranioectodermal dysplasia 2 and campomelic dysplasia, respectively. In addition, we found three independent association signals in the SOX9 locus, and six known loci for nose size and shape were replicated in this study population. Interestingly, four SNPs within these five face morphology-related loci showed discrepancies in allele frequencies among ethnic groups.ConclusionsWe identified five novel face morphology loci that were associated with facial frontal contour, nose shape, and eye shape. Our findings provide useful genetic information for the determination of face morphology.Electronic supplementary materialThe online version of this article (10.1186/s12864-018-4865-9) contains supplementary material, which is available to authorized users.

A variant in the PRDM16 locus has been correlated with QRS duration in an electrocardiogram genome-wide association study, and the deletion of PRDM16 has been implicated as a causal factor of the dilated cardiomyopathy that is linked to 1p36 deletion syndrome. We aimed to determine how a null mutation of Prdm16 affects cardiac function and study the underlying mechanism of the resulting phenotype in an appropriate mouse model. We used cardiac-specific Prdm16 conditional knockout mice to examine cardiac function by electrocardiography. QRS duration and QTc interval increased significantly in cardiac-specific Prdm16 knockout animals compared with wild-type mice. Further, we assessed cardiomyopathy-associated features by trichrome staining, densitometry, and hydroxyproline assay. Prdm16-null hearts showed greater fibrosis and cardiomyocyte hypertrophy. By quantitative real-time PCR, Prdm16-null hearts upregulated extracellular matrix-related genes ( Ctgf, Timp1) and α-smooth muscle actin ( Acta2), a myofibroblast marker. Moreover, TGF-β signaling was activated in Prdm16-null hearts, as evidenced by increased Tgfb1–3 transcript levels and phosphorylated Smad2. However, the inhibition of TGF-β receptor did not reverse the aberrations in conduction in cardiac-specific Prdm16 knockout mice. To determine the underlying mechanisms, we performed RNA-seq using mouse left ventricular tissue. By functional analysis, Prdm16-null hearts experienced dysregulated expression of ion channel genes, including Kcne1, Scn5a, Cacna1h, and Cacna2d2. Mice with Prdm16-null hearts develop abnormalities in cardiac conduction and cardiomyopathy-associated phenotypes, including fibrosis and cellular hypertrophy. Further, the RNA-seq findings suggest that impairments in ion homeostasis (Ca2+, K+, and Na+) may at least partially underlie the abnormal conduction in cardiac-specific Prdm16 knockout mice. NEW & NOTEWORTHY This is the first study that describes aberrant cardiac function and cardiomyopathy-associated phenotypes in an appropriate murine genetic model with cardiomyocyte-specific Prdm16-null mutation. It is noteworthy that the correlation of PRDM16 with QRS duration is replicated in a murine animal model and the potential underlying mechanism may be the impairment of ion homeostasis.

No abstract

PR interval is the period from the onset of P wave to the start of the QRS complex on electrocardiograms. A recent genomewide association study (GWAS) suggested that GAREM1 was linked to the PR interval on electrocardiograms. This study was designed to validate this correlation using additional subjects and examined the function of Garem1 in a mouse model. We analyzed the association of rs17744182, a variant in the GAREM1 locus, with the PR interval in 5646 subjects who were recruited from 2 Korean replication sets, Yangpyeong (n = 2471) and Yonsei (n = 3175), and noted a significant genomewide association by meta-analysis (P = 2.39 × 10). To confirm the function of Garem1 in mice, Garem1 siRNA was injected into mouse tail veins to reduce the expression of Garem1. Garem1 transcript levels declined by 53% in the atrium of the heart (P = 0.029), and Garem1-siRNA injected mice experienced a significant decrease in PR interval (43.27 ms vs. 44.89 ms in control, P = 0.007). We analyzed the expression pattern of Garem1 in the heart by immunohistology and observed specific expression of Garem1 in intracardiac ganglia. Garem1 was expressed in most neurons of the ganglion, including cholinergic and adrenergic cells. We have provided evidence that GAREM1 is involved in the PR interval of ECGs. These findings increase our understanding of the regulatory signals of heart rhythm through intracardiac ganglia of the autonomic nervous system and can be used to guide the development of a therapeutic target for heart conditions, such as atrial fibrillation.