Genome-wide association study of multiethnic non-syndromic orofacial cleft families identifies novel loci specific to family and phenotypic subtypes

Orofacial clefts (OFCs) are among the most common craniofacial birth defects and constitute a high public health burden around the world. OFCs are phenotypically heterogeneous, affecting only the lip, only the palate, or involving both the lip and palate. Cleft palate alone is demonstrably a genetically distinct abnormality from OFCs that involve the lip, therefore, it is common to study cleft lip (CL) in combination with cleft lip plus cleft palate (CLP) as a phenotypic group (i.e. cleft lip with or without cleft palate, CL/P), usually considering CLP to be a clinically more severe form of CL. However, even within CL/P, important genetic differences among subtypes may be present. The Pittsburgh Orofacial Cleft (Pitt-OFC) multiethnic study is a rich resource for the study of non-syndromic OFC, comprising a large number of families (~12,000 individuals) from multiple populations worldwide: US and Europe (whites), Central and South America (mixed Native American, European and African), Asia, and Africa. In this study we focused on the CL/P families from this resource grouped into three non-overlapping family types: those with only CL affected members, only CLP affected members, or both CL and CLP. In all, seven total subtypes besides the combined CL/P phenotype, were defined based on the cleft type(s) that were present within pedigree members. The full sample for these analyses includes 2,218 CL and CLP cases along with 4,537 unaffected relatives, as well as 2,673 pure controls with no family history of OFC. Genome-wide association analyses were conducted within each subset, as well as the combined sample. Five novel genome-wide significant associations were observed: 3q29 (rs62284390, p=2.70E-08), 5p13.2 (rs609659, p= 4.57E-08), 7q22.1 (rs6465810, p= 1.25E-08), 19p13.3 (rs628271, p=1.90E-08) and 20q13.33 (rs2427238, p=1.51E-09). In addition, five significant and four suggestive associations confirmed regions previously published as OFC risk loci - PAX7 , IRF6 , FAM49A , DCAF4L2 , 8q24.21, ARID3B, NTN1 , TANC2 and the WNT9B:WNT3 gene cluster. At each of these loci, we compared effect sizes of associated SNPs observed across subtypes and the full sample, and found that certain loci were associated with a specific cleft type, and/or specific family types. Our findings indicate that risk factors differ between cleft and family types, but each cleft type also exhibits some degree of genetic heterogeneity.