Common cis-regulatory variation modifies the penetrance of pathogenic SHROOM3 variants in craniofacial microsomia

Abstract

Pathogenic coding variants have been identified in thousands of genes, yet the mechanisms underlying the incomplete penetrance in individuals carrying these variants are poorly understood. In this study, in a cohort of 2009 craniofacial microsomia (CFM) patients of Chinese ancestry and 2625 Han Chinese controls, we identified multiple predicted pathogenic coding variants in SHROOM3 in both CFM patients and healthy individuals. We found that the penetrance of CFM correlates with specific haplotype combinations containing likely pathogenic-coding SHROOM3 variants and CFM-associated expression quantitative trait loci (eQTLs) of SHROOM3 expression. Further investigations implicate specific eQTL combinations, such as rs1001322 or rs344131, in combination with other significant CFM-associated eQTLs, which we term combined eQTL phenotype modifiers (CePMods). We additionally show that rs344131, located within a regulatory enhancer region of SHROOM3, demonstrates allele-specific effects on enhancer activity and thus impacts expression levels of the associated SHROOM3 allele harboring any rare coding variant. Our findings also suggest that CePMods may serve as pathogenic determinants, even in the absence of rare deleterious coding variants in SHROOM3 This highlights the critical role of allelic expression in determining the penetrance and severity of craniofacial abnormalities, including microtia and facial asymmetry. Additionally, using quantitative phenotyping, we demonstrate that both microtia and facial asymmetry are present in two separate Shroom3 mouse models, the severity of which is dependent on gene dosage. Our study establishes SHROOM3 as a likely pathogenic gene for CFM and demonstrates eQTLs as determinants of modified penetrance in the manifestation of the disease in individuals carrying likely pathogenic rare coding variants.

Figure 1.

Distribution of putative deleterious variants and common SNP associations in SHROOM3 among CFM patients and healthy Chinese. (A) The schematic delineates the PDZ, ASD1, and ASD2 domains of the SHROOM3 protein, color-coded in yellow, blue, and green, respectively. The upper panel enumerates the deleterious variants identified in healthy Chinese individuals, and the lower panel details those found in CFM patients, with variants from European/Hispanic/Australasian ancestry underlined. The count of deleterious variants within specific regions is indicated by numbered brown circles. Variants emphasized in red denote those predicted as pathogenic by the ESM1b and AlphaMissense algorithm. Likely pathogenic variants determined using combination patterns are marked by bold in CFM patients. (B) Manhattan plot of CFM-associated variants in SHROOM3. This plot showcases variants with a minor allele frequency above 0.05. The size of each dot indicates the odds ratio of the corresponding variant. The plot demarcates five linkage disequilibrium (LD) blocks within SHROOM3 in gray, with the color-coding reflecting r² values, signifying the correlation between the lead SNP and other variants within each LD block. Lead SNPs are marked with purple rhombus, and their rsID is shown with an orange background. eQTL sites are identified with rsIDs with a white background. (OR) Odds ratio.

Figure 2.

Patterns of haplotype combinations involving putatively pathogenic eQTLs and predicted pathogenic coding variants. (A) UpSet plot showing putatively pathogenic eQTLs fine-mapped using various methods, including conditional analysis, CAVIAR, FINEMAP, and PAINTOR. (B) Haplotype diagram illustrating the association between the putatively pathogenic eQTL rs10017322 and predicted pathogenic variants. The G and A alleles of rs10017322 correspond to high and low expression levels of SHROOM3, respectively, in the GTEx database (Supplemental Fig. 3). The A allele modifies the penetrance of the pathogenic L102V and R1936W alleles in healthy Chinese individuals, whereas other variants combined with the G allele are associated with the CFM phenotype in Chinese and European populations. (C) Missense score of identified coding variants in SHROOM3.

Figure 3.

Patterns of haplotype combinations involving putatively pathogenic eQTLs. (A) Conceptual diagram hypothesizing that the combination of risk alleles from SHROOM3 eQTLs correlates with the CFM phenotype. The greater the number of risk alleles carried, the higher the likelihood of being a CFM patient. (B,C) Risk allele combinations involving the putatively pathogenic eQTL rs10017322/rs344131 and other SHROOM3 eQTLs in 2009 Chinese CFM patients and 2625 healthy Chinese individuals. Carriers of four risk alleles show a significant enrichment (P = 0.0001983) in CFM patients. (D,E) Risk allele combinations as in B,C but analyzed in 361 CFM patients who strongly contribute to the SHROOM3 signal compared with 2625 healthy Chinese. Individuals carrying four risk alleles demonstrate a more significant enrichment (P = 2.253 × 10⁻¹⁰) in CFM patients. Ratio of risk allele proportion indicates the proportion of the risk allele compared with CFM patients and healthy Chinese.

Figure 4.

In vitro analysis of eQTL effects on enhancer activity. (A) Epigenetic landscape at SHROOM3 in human neural crest cells (hCNCCs). This panel presents an integrated epigenomic analysis, combining in vitro PCHi-C, ATAC-seq, and ChIP-seq data from hCNCCs (Rada-Iglesias et al. 2012; Xu et al. 2024). It displays the epigenetic markers H3K4me1, H3K4me3, and H3K27ac in hCNCCs. The promoter is highlighted with light orange shading. Enhancers are highlighted with blue shading, with eQTLs within these enhancers marked by black lines (darkness represents the significance level). In the PCHi-C section, differential chromatin interactions connect enhancers to the SHROOM3 promoter, with line colors indicating the strength of interaction. (B) Luciferase assays to assess the effects of eQTLs on SHROOM3 enhancers, including an empty vector as a control. Data from three independent experiments, each comprising three technical replicates, are depicted. (C) Enrichment analysis of risk allele combinations for eQTLs rs344131 and rs61090632 in CFM patients compared with control populations. Significant enrichment is observed in both the entire CFM cohort (P < 0.008884) and the SHROOM3 CFM subset (P < 1.71 × 10⁻²⁴) using chi-square tests. Control populations: (CASPMI) Chinese Academy of Sciences Precision Medicine Initiative, (WBBC) Westlake BioBank for Chinese, and (SG10K) Singapore 10K Genome Project. Patient cohorts: CFM (n = 2009) and SHROOM3 CFM (n = 361). (D) Luciferase assay results demonstrating the combined effect of alleles rs344131 and rs61090632 in HEK-293T cells. P-values are calculated using the Student's t-test.

Figure 5.

In vivo analysis of Shroom3 zygosity in mice is associated with microtia and facial asymmetry. (A) Lateral view of 3D rendered developing auricles (right side) from E14.25 wild-type (Shroom3^+/+), heterozygote (Shroom3^gt/+), and homozygote (Shroom3^gt/gt) littermates following optical projection tomography. Note the marked microtia in the homozygote and the subtle dorsal auricular deficiency in the heterozygote (both marked by yellow arrows in the respective images) compared with the wild-type littermate. (B–D) Landmark-based quantitative assessment of auricular, mandibular and maxillary length (top) and absolute asymmetry (bottom) in wild-type, Shroom3^gt/+, and Shroom3^gt/gt embryos between E14.25 and E15.25. In the graphs of the average lengths of the mandible, maxilla, and auricle, dotted lines “connect” the means of the different genotypes from the same litter (stage). Mean ± SD indicated for the average length measurements; the mean (large shape with horizontal line dissecting it) is shown for the absolute asymmetry measurements. Asterisks denote statistical significance (P < 0.05). (E) Lateral view of 3D rendered auricles (left side) from an E18.5 wild type (Shroom3^+/+) and two age-matched homozygote (Shroom3^{em1(IMPC)Bay/em1(IMPC)Bay}) embryos derived from microCT scans generated by the International Mouse Phenotyping Center (IMPC; https://www.mousephenotype.org). The first homozygote (central) shows the more typical presentation of a small pinna that is inwardly curved and posteriorly rotated (white arrow). The second homozygote had significantly asymmetric pinna phenotypes - the auricular presentation of the left ear (yellow arrow) was severe, whereas the right pinna was similar in presentation to the first homozygote. (F–H) Landmark-based quantitative assessment of mandibular, maxillary, and auricular length (top) and absolute asymmetry (bottom) in E18.5 wild-type (Shroom3^+/+) and Shroom3^{em1(IMPC)Bay/em1(IMPC)Bay} null embryos. The mean (large shape with horizontal line dissecting it) is shown for each. Asterisks denote statistical significance (P < 0.05).