Unraveling ADHD: genes, co-occurring traits, and developmental dynamics

Abstract

Attention-deficit/hyperactivity disorder (ADHD) is a heterogeneous neurodevelopmental condition with a high prevalence of co-occurring conditions, contributing to increased difficulty in long-term management. Genome-wide association studies have identified variants shared between ADHD and co-occurring psychiatric disorders; however, the genetic mechanisms are not fully understood. We integrated gene expression and spatial organization data into a two-sample Mendelian randomization study for putatively causal ADHD genes in fetal and adult cortical tissues. We identified four genes putatively causal for ADHD in cortical tissues (fetal: ST3GAL3, PTPRF, PIDD1; adult: ST3GAL3, TIE1). Protein-protein interaction databases seeded with the causal ADHD genes identified biological pathways linking these genes with conditions (e.g., rheumatoid arthritis) and biomarkers (e.g., lymphocyte counts) known to be associated with ADHD, but without previously shown genetic relationships. The analysis was repeated on adult liver tissue, where putatively causal ADHD gene ST3GAL3 was linked to cholesterol traits. This analysis provides insight into the tissue-dependent temporal relationships between ADHD, co-occurring traits, and biomarkers. Importantly, it delivers evidence for the genetic interplay between co-occurring conditions, both previously studied and unstudied, with ADHD.

Figure 1.. Two-sample Mendelian randomization identifies three putatively causal ADHD genes in the fetal cortical tissue and two putatively causal ADHD genes in the adult cortical tissue.

Only those statistically significant (P < 0.05) after the Bonferroni correction are shown. Odds ratios (OR) are calculated based on the SNP–gene pair. The difference in odds ratios between ST3GAL3 in the fetal and adult tissue is due to the opposing allelic fold changes (i.e., the decreased allelic fold change from rs7529895 causes an increased risk of ADHD, whereas the increased allelic fold change from rs3011217 causes a decreased risk of ADHD).

Figure S1.. Overview of the methods.

The CoDeS3D pipeline (Fadason et al, 2018) was used to generate both an adult and fetal cortical tissue gene regulatory network by integrating SNP–gene chromatin interactions from Hi-C data with eQTL–gene association data. These gene regulatory networks were used as exposure data for two-sample Mendelian randomization (2SMR) (Hemani et al, 2018) analyses (fetal and adult separately). ADHD GWAS data were used as outcome data. Causal genes identified in these 2SMR analyses and ADHD-associated SNPs from GWAS were fed into the Multimorbid3D (Golovina et al, 2023) algorithm (separately) where protein interactions were used to expand the networks and identify associated GWAS traits (see the Materials and Methods section).

Figure 2.. Multimorbid3D identifies traits associated with ADHD in the fetal cortical tissue.

(A) Bubble plot showing statistically significant traits (bootstrap P < 0.05) found in the analysis of ADHD GWAS variants in the fetal cortical tissue that were also identified in the adult cortical tissue. The size of the bubbles refers to the number of eQTLs associated with that particular trait. Bubble colors refer to the −log(P-value) after bootstrapping. The y-axis trait names come from the GWAS Catalog, and brain/mood-related traits are highlighted gray. (B) is the same as (A) for traits identified in the fetal analysis only. (C) Heatmap outlining which genes are regulated by variants associated with the traits from (A, B) in the fetal cortical tissue. The color is based on the number of eQTLs associated with the trait and gene. Clustering is based on the numbers of eQTLs. The first subsection is genes at chr17q21.31, the second subsection is FADS1/FADS2, and the third subsection is genes related to cognition, including those at chr3p21.1. The full heatmap is shown in Fig S1.

Figure S2.. Genes associated with ADHD co-occurring traits identified in the fetal cortical tissue.

Full version of Fig 2C. The heatmap shows genes regulated by eQTLs associated with traits listed on the y-axis. Genes and traits listed are on level 0 of the fetal cortical tissue network. The heatmap has been clustered based on the number of eQTLs.

Figure S3.. Genes associated with ADHD co-occurring traits identified in the adult cortical tissue.

The heatmap shows genes regulated by eQTLs associated with traits listed on the y-axis. Genes and traits listed are on level 0 of the adult cortical tissue network. The heatmap has been clustered based on the number of eQTLs.

Figure S4.. Developmental enrichment analysis identifies the dorsolateral prefrontal cortex as enriched with ADHD-associated genes.

The graph showing the family-wise error rate for the dorsolateral prefrontal cortex being enriched with the adult or fetal gene set from our network analysis at five developmental timepoints. Those below the red dashed line (family-wise error rate = 0.05) are statistically significantly enriched.

Figure 3.. Multimorbid3D on causal genes identifies traits co-occurring with ADHD.

(A) Bubble plot showing statistically significant traits (bootstrap P < 0.05) found in the analysis of causal ADHD genes in the adult cortical tissue. The size of the bubbles refers to the number of eQTLs associated with that particular trait. Bubble colors refer to the −log(P-value) after bootstrapping. Y-axis trait names come from the GWAS Catalog, and brain/mood-related traits are highlighted gray. (B) is the same as (A) for causal ADHD genes in the fetal cortical tissue. Purple traits in (A, B) are related to the networks in (C, D, E). (C, D) highlights the protein network linking the adult causal gene TIE1 to lipoprotein (a) levels, whereas (D) links the fetal causal gene PTPRF to lymphocyte counts; both potential biomarkers. (E) shows the protein pathways linking the fetal causal gene PIDD1 to co-occurring conditions related to the eye and rheumatoid arthritis. In (C, D, E), the red bubble is for proteins encoded for by causal genes, green is for proteins in the network, and purple is for proteins encoded for by genes regulated by SNPs associated with the listed trait.

Figure 4.. ADHD causal genes associated with cholesterol traits in the liver GRN.

(A) Bubble plot of statistically significant traits (bootstrap P < 0.05) identified in the analysis of causal ADHD genes in the liver tissue. The size of the bubbles refers to the number of eQTLs associated with that particular trait. Bubble colors refer to whether the trait comes from the analysis of genes from the fetal causal genes (ST3GAL3, PTPRF, PIDD1) or the adult causal genes (ST3GAL3, TIE1), or whether it was present in both analyses. Asterisk traits were present in the cortical tissue analysis. (B) shows the network genetically linking the causal gene ST3GAL3 to cholesterol traits highlighted purple in (A). Coloring of (B) matches Fig 3C–E.