Genome-wide association meta-regression identifies stem cell lineage orchestration as a key driver of acne risk

Abstract

Over 85% of the population experience acne at some point in their lives, with its severity spanning a quantitative spectrum, from mild, transient outbreaks to more persistent, severe forms of the condition. Moderate to severe disease poses a substantial global burden arising from both the physical and psychological impacts of this highly visible condition. The analytical approach taken in this study aimed to address the impact of variation in the dichotomisation of acne case control status, driven by ascertainment and study design, on effect size estimates across independent genetic association studies of acne. Through a fixed intercept meta-regression framework, we combined evidence genome-wide for association with acne across studies in which case-control status had been ascertained in different settings, allowing for different severity threshold definitions. Across a combined sample of 73,997 cases and 1,103,940 controls of European, South Asian and African American ancestry we identify genetic variation at 165 genomic loci that influence acne risk. There is evidence for both shared and ancestry specific components to the genetic susceptibility to acne and for sex differences in the magnitude of effect of risk alleles at three loci. We observe that common genetic variation explains 13.4% of acne heritability on the liability scale. Consistent with the hypothesis that genetic risk primarily operates at the level of individual pilosebaceous units, a polygenic score derived from this case-control study of acne susceptibility is associated with both self-reported and clinically assessed acne severity in adolescence, further strengthening the link between genetic risk and disease severity. Prioritisation of causal genes at the identified acne risk loci, provides genetic validation of the targets of established and emerging acne therapies, including retinoid treatments. The identified acne risk loci are enriched for genes encoding downstream effectors of RXRA signalling, including SOX9 and components of the WNT and p53 pathways. Illustrating that the control of stem cell lineage plasticity and cellular fate are important mechanisms through which genetic variation influences acne susceptibility within the pilosebaceous unit.

Figure 1.

A) Approach to estimate ascertainment effects from the capture of different severity classes. Prevalence estimates of acne severity classes from epidemiological studies overlaying a normal distribution representing the acne liability distribution, with 50% of the population having a history of mild, moderate or severe acne, 14% moderate or severe, and 2% severe. Below the distribution, illustrates how the study-level parameter is determined for different ascertainment approaches. The areas representing cases are shaded in blue, while controls are shaded in white. The red bar indicates the mean liability for cases and controls. The difference in mean liability between cases and controls, which defines the ascertainment study-level parameter, is represented by the double headed green arrows. B) Systematic differences in effect sizes distributions for each of the 16 studies against the study’s respective ascertainment study-level parameter. The size of each point proportional to the sample size of the respective study. Systematic differences in effect sizes distributions for each study is measured by the slope of a Deming regression of the study’s effect size estimates on the fixed effect meta-analysis effect estimates of the lead variants at established acne loci (Methods). The R-squared (R²) value represents the proportion of variance in the systematic differences in effect sizes explained by the study-level parameter. C) Evidence of association of genetic variants across the genome with acne. Genetic variants ordered by chromosome and position along the x-axis with the evidence of association, −log10(P-value) from the association analysis using a two-sided Z-test, on the y-axis. The P-values are not adjusted for multiple comparisons, but the dashed pink line marks the genome-wide significance threshold (P = 5 × 10⁻⁸). Green points represent variants at established acne susceptibility loci, while points represent variants at novel risk loci.

Figure 2:. Mean of the polygenic risk score distributions by acne severity classifications in two independent cohorts.

The left-hand panel shows the PRS distributions in the PISA cohort, in which acne severity was assessed using a self-reported retrospective questionnaire, the right-hand panel the Generation R cohort, where acne severity was assessed by trained dermatologists. The x-axis denotes the mean PRS calculated using SBayesR from the FIMR results and the y-axis represents acne severity classification in each cohort. The bars demark the 95% confidence intervals for the mean.

Figure 3.. Sex specific genetics effects on acne risk.

Odd ratios of the allelic effects on acne risk in males, females and combined analysis. The two variants on the left are those identified with a genome-wide significant effect on acne risk in the sex stratified analyses, and the 17 on the right are the lead variants from the 162 acne risk loci identified in the combined analysis for which there is a nominally significant difference in magnitude between male and female effect sizes (P<0.05, two-sided Z-difference test). The x-axis lists the chromosome band where the lead variant is located, and the y-axis represents the odds ratios. The bars with demark the 95% confidence intervals. Variants whose evidence of a difference in effect between males and females surpasses a Bonferroni-adjusted significance threshold for 162 tests (P<3.08×10⁻⁴) are marked with a red asterisk.

Figure 4:. Prioritisation of candidate causal genes that are disrupted by rare alleles that cause Mendelian disorders or encode therapeutic targets.

Genes prioritised by local evidence are shown in blue, encompassing nearest gene to the lead variant of the association signal, coding, splice site or enhancer variants in high LD (r2 > 0.8) with the lead variant, eQTL colocalization and significant TWAS associations in sun exposed or not sun exposed skin. Red boxes indicate gene prioritised by methods that examine shared biological functions across loci with DEPICT and Polygenic Priority Score (PoPs). The 58 genes represented in the table are all located at acne susceptibility loci identified in this study for the first time and encode the molecular target of an approved drug therapy (green) and are disrupted by rare alleles in Mendelian disorders (yellow).

Figure 5.. Clustering of expression levels of selected genes across different cell types in facial skin and annotation of gene clusters.

The heatmap indicates scaled transcript levels within cell types: Keratinocytes, Macrophages, Dendritic Cells (DC), Mast Cells (MastC), Plasma Cells (PlasmaC), B Cells (BC), Innate Lymphoid Cells (ILC), T Cells (TC), Natural Killer cells (NK), Vascular Endothelial Cells (VEC), Lymphatic Endothelial Cells (LEC), Schwann Cells (SchwannC), Pericytes, Fibroblasts, and Melanocytes. Each row of the bottom plot indicates genes that have been clustered using evidence of direct and indirect (where a single intermediary gene serves as a linking bridge) gene-gene relationships from the STRING database. Clusters have a minimum size of four genes located across four acne loci. Black points indicate the prioritised putative causal genes identified at acne susceptibility loci, while grey points the intermediary genes in each cluster. Descriptions of the gene clusters are summaries of the three most enriched gene sets from KEGG, Reactome, GO Biological Processes or GO molecular function.