Hostname: page-component-76fb5796d-dfsvx Total loading time: 0 Render date: 2024-04-27T13:17:38.345Z Has data issue: false hasContentIssue false
  • English
  • Français

Distributed genetic effects of the corpus callosum subregions suggest links to neuropsychiatric disorders and related traits

Published online by Cambridge University Press:  24 August 2023

Megan L. Campbell Open the ORCID record for Megan L. Campbell [Opens in a new window] ,
Shareefa Dalvie Open the ORCID record for Shareefa Dalvie [Opens in a new window] ,
Alexey Shadrin ,
Dennis van der Meer Open the ORCID record for Dennis van der Meer [Opens in a new window] ,
Kevin O’Connell ,
Oleksander Frei ,
Ole A. Andreassen Open the ORCID record for Ole A. Andreassen [Opens in a new window] ,
Dan J. Stein  and
Jaroslav Rokicki Open the ORCID record for Jaroslav Rokicki [Opens in a new window]
Megan L. Campbell*
Affiliation:
Department of Psychiatry and Mental Health, University of Cape Town, Cape Town, South Africa Global Initiative for Neuropsychiatric Genetics Education in Research (GINGER), Harvard, T.H. Chan School of Public Health, Boston, MA, USA
Shareefa Dalvie
Affiliation:
Division of Human Genetics, Department of Pathology, University of Cape Town, Cape Town, South Africa
Alexey Shadrin
Affiliation:
NORMENT Centre for Psychosis Research, Division of Mental Health and Addiction, University of Oslo, and Oslo University Hospital, Oslo, Norway
Dennis van der Meer
Affiliation:
NORMENT Centre for Psychosis Research, Division of Mental Health and Addiction, University of Oslo, and Oslo University Hospital, Oslo, Norway School of Mental Health and Neuroscience, Faculty of Health, Medicine and Life Sciences, Maastricht University, Maastricht, Netherlands
Kevin O’Connell
Affiliation:
NORMENT Centre for Psychosis Research, Division of Mental Health and Addiction, University of Oslo, and Oslo University Hospital, Oslo, Norway
Oleksander Frei
Affiliation:
NORMENT Centre for Psychosis Research, Division of Mental Health and Addiction, University of Oslo, and Oslo University Hospital, Oslo, Norway
Ole A. Andreassen
Affiliation:
NORMENT Centre for Psychosis Research, Division of Mental Health and Addiction, University of Oslo, and Oslo University Hospital, Oslo, Norway
Dan J. Stein
Affiliation:
Department of Psychiatry and Mental Health, University of Cape Town, Cape Town, South Africa
Jaroslav Rokicki
Affiliation:
NORMENT Centre for Psychosis Research, Division of Mental Health and Addiction, University of Oslo, and Oslo University Hospital, Oslo, Norway Centre of Research and Education in Forensic Psychiatry (SIFER), Oslo University Hospital, Oslo, Norway
*
Corresponding author: Megan L. Campbell; Email: mcampbel@broadinstitute.org
Rights & Permissions [Opens in a new window]

Abstract

Background:

The corpus callosum (CC) is a brain structure with a high heritability and potential role in psychiatric disorders. However, the genetic architecture of the CC and the genetic link with psychiatric disorders remain largely unclear. We investigated the genetic architectures of the volume of the CC and its subregions and the genetic overlap with psychiatric disorders.

Methods:

We applied multivariate genome-wide association study (GWAS) to genetic and T1-weighted magnetic resonance imaging (MRI) data of 40,894 individuals from the UK Biobank, aiming to boost genetic discovery and to assess the pleiotropic effects across volumes of the five subregions of the CC (posterior, mid-posterior, central, mid-anterior and anterior) obtained by FreeSurfer 7.1. Multivariate GWAS was run combining all subregions, co-varying for relevant variables. Gene-set enrichment analyses were performed using MAGMA. Linkage disequilibrium score regression (LDSC) was used to determine Single nucleotide polymorphism (SNP)-based heritability of total CC volume and volumes of its subregions as well as their genetic correlations with relevant psychiatric traits.

Results:

We identified 70 independent loci with distributed effects across the five subregions of the CC (p < 5 × 10−8). Additionally, we identified 33 significant loci in the anterior subregion, 23 in the mid-anterior, 29 in the central, 7 in the mid-posterior and 56 in the posterior subregion. Gene-set analysis revealed 156 significant genes contributing to volume of the CC subregions (p < 2.6 × 10−6). LDSC estimated the heritability of CC to (h2SNP = 0.38, SE = 0.03) and subregions ranging from 0.22 (SE = 0.02) to 0.37 (SE = 0.03). We found significant genetic correlations of total CC volume with bipolar disorder (BD, rg = −0.09, SE = 0.03; p = 5.9 × 10−3) and drinks consumed per week (rg = −0.09, SE = 0.02; p = 4.8 × 10−4), and volume of the mid-anterior subregion with BD (rg = −0.12, SE = 0.02; p = 2.5 × 10−4), major depressive disorder (MDD) (rg = −0.12, SE = 0.04; p = 3.6 × 10−3), drinks consumed per week (rg = −0.13, SE = 0.04; p = 1.8 × 10−3) and cannabis use (rg = −0.09, SE = 0.03; p = 8.4 × 10−3).

Conclusions:

Our results demonstrate that the CC has a polygenic architecture implicating multiple genes and show that CC subregion volumes are heritable. We found that distinct genetic factors are involved in the development of anterior and posterior subregions, consistent with their divergent functional specialisation. Significant genetic correlation between volumes of the CC and BD, drinks per week, MDD and cannabis consumption subregion volumes with psychiatric traits is noteworthy and deserving of further investigation.

Keywords

Corpus callosumGeneticsPsychiatric disordersHeritabilityT1-weighted magnetic resonance imagine
Type
Original Article
Information
Acta Neuropsychiatrica , First View , pp. 1 - 8
Check for updates
Creative Commons
Creative Common License - CCCreative Common License - BYCreative Common License - NC
This is an Open Access article, distributed under the terms of the Creative Commons Attribution-NonCommercial licence (http://creativecommons.org/licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original article is properly cited. The written permission of Cambridge University Press must be obtained prior to any commercial use.
Copyright
© The Author(s), 2023. Published by Cambridge University Press on behalf of Scandinavian College of Neuropsychopharmacology

Significant outcomes

We discovered novel loci not previously associated with the corpus callosum and provide evidence for unique genetic aetiology of the subregions. Additionally, we provide the first large-scale genetic evidence of heritability of the subregions and provide evidence to suggest overlap with psychiatric disorders.

Limitations

This study consisted only of individuals from the UK Biobank, and future studies should focus on including diverse populations and those with a lower mean age.

Introduction

The corpus callosum (CC) is the largest white matter tract in the brain and serves as the major interhemispheric commissure, integrating information from the two cerebral hemispheres to facilitate language, affective and sensorimotor function (Wang et al., Reference Wang, Meng and Yuan2020). The CC is commonly segmented into five subregions: anterior, mid-anterior, central, mid-posterior and posterior, with each subregion having distinct axonal projections from cortical regions (Hofer and Frahm, Reference Hofer and Frahm2006). For example, projections from the auditory and somatosensory cortices primarily cross the central subregion, while those from the visual cortex cross the anterior subregion (Hofer and Frahm, Reference Hofer and Frahm2006). Although the exact boundaries of these subregions are challenging to delineate using magnetic resonance imaging (MRI) (Hofer and Frahm, Reference Hofer and Frahm2006), their functional importance in various cognitive processes such as language, sensory processing and emotional regulation is well established (Cover et al., Reference Cover, Herrera, Bento, Appenzeller and Rittner2018).

Differences in volume of the CC have been associated with several psychiatric disorders, including mood, substance use and developmental disorders (Rosenbloom et al., Reference Rosenbloom, Sullivan and Pfefferbaum2003; Newbury and Rosen, Reference Newbury and Rosen2012; Abramovic et al., Reference Abramovic, Boks and Vreeker2018; Prunas et al., Reference Prunas, Delvecchio and Perlini2018; de Souza et al., Reference de Souza, Rosa, Escobar, Gasparetto and Nakamura-Palacios2019). Volumetric reductions in the CC have been observed in bipolar disorder (BD) patients, with their relatives showing intermediate volumetric reductions between controls and BD probands (Walterfang et al., Reference Walterfang, Malhi and Wood2009; Francis et al., Reference Francis, Mothi and Mathew2016). There is evidence to suggest that alcohol consumption affects CC morphometry (de Souza et al., Reference de Souza, Rosa, Escobar, Gasparetto and Nakamura-Palacios2019). These effects appear to be subregion-specific, with volume of the anterior CC subregion displaying a negative association with alcohol consumption, irrespective of an alcohol use disorder diagnosis (de Souza et al., Reference de Souza, Rosa, Escobar, Gasparetto and Nakamura-Palacios2019). Frontal and superior subregions of the CC appear to be most affected by alcohol consumption, with the inferior and posterior CC subregions remain relatively preserved (de Souza et al., Reference de Souza, Rosa, Escobar, Gasparetto and Nakamura-Palacios2019). CC volume reductions are also present in autism spectrum disorder (ASD), with the anterior and mid-anterior subregions seemingly most affected (Frazier and Hardan, Reference Frazier and Hardan2009; Valenti et al., Reference Valenti, Pino, Mazza, Panzarino, Di Paolantonio and Verrotti2020). Differences in volume of the CC subregions have also been noted in attention-deficit hyperactivity disorder (ADHD), major depressive disorder (MDD) and schizophrenia (SCZ), although the pattern of alteration in these disorders is less clear (Connaughton et al., Reference Connaughton, Whelan, O'Hanlon and McGrath2022; Zhao et al., Reference Zhao, Lau and Wang2022; Zhou et al., Reference Zhou, Wang and Wang2022).

Twin studies have demonstrated a high heritability of CC size, estimated to 0.67 for total CC and 0.48–0.62 for the subregions (Scamvougeras et al., Reference Scamvougeras, Kigar, Jones, Weinberger and Witelson2003; Woldehawariat et al., Reference Woldehawariat, Martinez and Hauser2014). Although twin and family studies have established the heritability of the CC, there has been a lack of large-scale genetic investigations aimed at identifying the contribution of common genetic variants to the morphology of CC. Recent advances in methods for exploring the genetic architecture of the brain have provided insight into the subregion-specific genetic aetiology of multiple brain regions (van der Meer, Rokicki, et al., Reference van der Meer, Rokicki and Kaufmann2020; Zhao et al., Reference Zhao, Li and Yang2020; Mufford et al., Reference Mufford, van der Meer and Kaufmann2021; Bahrami et al., Reference Bahrami, Nordengen and Shadrin2022; Ou et al., Reference Ou, Wu and Ge2023). This work has been valuable in identifying shared and specific genetic loci and pathways across subregions, as well as providing insight into differential genetic overlap with psychiatric disorders. Accordingly, we aimed to determine the genetic influences on CC subregions and total CC volume, assess the heritability of this structure, and investigate the genetic correlation with psychiatric traits and disorders.

Methods

Processing of imaging and genetic data

We accessed raw, T1-weighted MRI data from 35,592 genotyped, Caucasian individuals of British ancestry (age range: 45–81 years; mean age: 64.3 years, SD: 7.5; 52.3% female) from the UK Biobank (Bycroft et al., Reference Bycroft, Freeman and Petkova2018) (accession code 27412) for our main analysis, and 5302 individuals of non-Caucasian ancestry for our generalisation analyses conducted in independent data (age range: 45 to 81 years; mean age: 62.9, SD: 7.7 years; 46.3% female). Two different methods were used to segment the CC: FreeSurfer v7.1, which is based on the Witelson segmentation (Witelson, Reference Witelson1989), and the division by Hofer and Frahm (Reference Hofer and Frahm2006) implemented with C8 software (Herron et al., Reference Herron, Kang and Woods2012). Both methods divide the CC into five subregions, and the differences are illustrated and described in more detail in the supplementary material (Supplementary Material, Figure 1). However, the main difference being that FreeSurfer extracts volume of the CC and each subregion while the C8 software extracts thickness measurements along the medial line in the sagittal plane. Given the widespread use of FreeSurfer for subcortical segmentation, we chose to focus on this method and use the approach proposed by Hofer and Frahm (Reference Hofer and Frahm2006) to validate our findings. To account for possible confounders, neuroCombat was used to pre-residualise for age, age squared, sex, scanning site, Euler score (a proxy of image quality), total intracranial volume and the first 20 genetic principal components for each CC subregion and total CC volume (Fortin et al., Reference Fortin, Cullen and Sheline2018). These variables were included as covariates in the univariate genome-wide association study (GWAS). The pre-residualisation step was performed for both segmentation methods and for both the European and non-European participants. These residuals were used as the outcome variables in subsequent multivariate analyses.

Standard quality control (QC) procedures were used to process the UK Biobank version 3, imputed genetic data: Single nucleotide polymorphisms (SNPs) with an imputation score of less than 0.5, a minor allele frequency of less than 0.005, a missingness rate of more than 10% and those that failed the Hardy–Weinberg test of equilibrium (p < 1 × 10−9) were removed. A total of 35,593 individuals remained in the European cohort after QC and an additional 5286 in the non-European cohort.

Genome-wide association analyses: multivariate and univariate

The pre-residualised CC subregions were used as the outcome in the multivariate analysis. This was conducted using the multivariate omnibus statistical test (MOSTest), which has been previously applied to a variety of brain morphology traits (van der Meer, Frei, et al., Reference van der Meer, Frei and Kaufmann2020; Bahrami et al., Reference Bahrami, Nordengen and Shadrin2022). MOSTest implements permutation testing to identify genetic effects across multiple phenotypes, resulting in a multivariate GWAS summary statistic across all five CC subregions. Replication of the MOSTest results was conducted using MOSTest-PolyVertex Score (MOSTest-PVS) (Loughnan et al., Reference Loughnan, Shadrin and Frei2022) in the non-European sample of 5210 individuals.

To identify the distinct genetic contribution of every CC subregion, an individual univariate GWAS of each subregion and total CC were performed using Plink v2 (Chang et al., Reference Chang, Chow, Tellier, Vattikuti, Purcell and Lee2015). This approach allows for the determination of the effect direction and enables further follow-up analyses. Additionally, univariate GWAS of the subregions from the Hofer and Frahm segmentation was conducted as an additional replication step. Significant SNPs were defined using a significant threshold of p < 9.6 × 10−9 (Bonferroni, accounting for five subregions and total CC). Independent significant genetic variants are identified as variants with p < 8.3 × 10−9 and LD r 2 < 0.6 with each other. For each independent significant variant, all candidate variants are identified as variants with LD r 2 ≥ 0.6 with the lead variant. For a given independent significant variant, the borders of the genomic locus are defined as min/max positional coordinates over all corresponding candidate variants. Loci are then merged if they are separated by less than 250 kb. A subset of independent significant variants with LD r 2 < 0.1 are then selected as lead variants. If a locus contains several lead SNPs using this definition, then the one with the most significant p-value is selected as a lead SNP representing the locus. In order to identify the unique genetic contribution to each CC subregion, additional univariate GWAS of each subregion was conducted, including total CC volume as a covariate. An overview of the GWAS is provided in the supplementary material (Supplementary Material, Figure 2).

Functional annotation

Functional Mapping and Annotation of Genome-wide Association Studies (FUMA, platform v1.3.7, accessed September 2022) was used to annotate the GWAS results, using the 1000GPhase3 EUR as reference panel and default parameters (Watanabe et al., Reference Watanabe, Taskesen, van Bochoven and Posthuma2017).We conducted genome-wide gene-based association and gene-set analyses using MAGMA v.1.08 (http://ctg.cncr.nl/software/magma) in FUMA. All variants in the major histocompatibility complex (MHC) region (GRCh37: 6:26,000,000–33,000,000) were excluded before running the MAGMA analyses. MAGMA performs multiple linear regression to map the input SNPs to 18,091 protein-coding genes and estimates the significant value of each gene. Genes were considered significant if p < 2.76 × 10−6 (Bonferroni correction). These genes were investigated for enrichment of biological processes, tissue and cell types. In addition to FUMA, we used Open Target Genetics which integrates functional and biological data from multiple sources to highlight functionally implicated genes and conduct phenome-wide association studies and tests for association with previous GWAS (Mountjoy et al., Reference Mountjoy, Schmidt and Carmona2021).

SNP-based heritability and genetic correlation

SNP-based heritability (h 2 SNP) and genetic correlation were calculated using Linkage Disequilibrium Score Regression (LDSC) (B. K. Bulik-Sullivan et al., Reference Bulik-Sullivan, Loh and Finucane2015). All variants in the MHC region (GRCh37: 6:28,477,797–33,448,354) were removed prior to the analysis. Heritability estimates were calculated for subregions from the FreeSurfer segmentation and the Hofer and Frahm segmentation. Genetic correlations were calculated for total CC volume, each subregion and relevant traits which had indicated CC alterations in prior imaging studies and which had publicly available GWAS. Genetic correlations were calculated for: ADHD (Stergiakouli et al., Reference Stergiakouli, Hamshere and Holmans2012), ASD (PGC ASD, 2017), BD (Mullins et al., Reference Mullins, Forstner and O’Connell2021), MDD (Wray et al., Reference Wray, Ripke and Mattheisen2018), SCZ (Liu et al., Reference Liu, Li and Li2021), the PGC cross-disorder GWAS (Pouget et al., Reference Pouget, Wu and Mignot2019), cigarettes per day, drinks per week, smoking initiation, smoking cessation (Liu et al., Reference Liu, Jiang and Wedow2019) and cannabis use (Pasman et al., Reference Pasman, Verweij and Gerring2018).

Results

Genome-wide association analyses

We found 70 unique genetic loci with distributed associations across the CC, using MOSTest. The polygenic architecture of the CC is shown in Fig. 1A, illustrating the results of the multivariate analysis. The univariate statistics indicate subregion-specific genetic signals for each of the CC subregions (Fig. 1). The loci that showed strongest significant effects from the multivariate analysis showed distributed effects with varying patterns, often exhibiting a gradient from posterior to anterior across CC subregions (Fig. 1B). A list of significant SNPs from the univariate GWAS of total CC volume and the volume of each CC subregion can be found in the supplementary material (Supplementary Material, Table 1). Figure 2A displays the results from the univariate GWAS for the volume of each subregion. The results for the GWAS co-varying for total CC volume can be found in the supplementary material (Supplementary Material, Figure 3).

Figure 1. Miami plot of the −log10(p) statistics from the multivariate (top) and univariate (bottom) GWAS of total CC volume. (A) Top: The multivariate analysis was conducted using MOSTest and illustrate the distributive genetic architecture of the CC. Bottom: Individual univariate GWAS were conducted for total CC volume and each subregion, indicating both a shared and unique genetic aetiology of the CC subregions. (B) Top SNPs from the multivariate GWAS show subregion-specific effects across the CC. The top 10 SNPs from the multivariate GWAS were mapped to the CC subregions using the Z-scores from the respective univariate subregion GWAS. The results indicate a gradient of effect across the posterior and anterior subregions.

Figure 2. (A) Manhattan plots from univariate GWAS of each CC subregion. (B) Quantile-Quantile plot (QQ plot) of observed versus expected −log10(p-values) the GWAS of CC subregions. The shaded regions represent the expected null distribution of −log10(p-values) under the assumption of no association. (C) Volumetric correlations of CC subregions, each square represents the r-squared values.

Replication of the multivariate results was conducted using MOSTest-PVS in the 5286 individuals of non-European ancestry, for the FreeSurfer segmentation. Of the 70 significant, lead SNPs in the multivariate analysis of the European sample, only 33 were available in the individuals of non-European ancestry. Of these 33 SNPs, all had the same direction of effect in the group of non-European ancestry. Twenty-two of the 33 SNPs are nominally significant (p < 0.05) in the replication and six remained significant after correction for multiple testing (Bonferroni, p < 1.51 × 10-3) (Supplementary Material, Table 2). The results for the multivariate analysis of the Hofer and Frahm segmentation can be found in the supplementary material (Supplementary Material, Figure 4).

Independent, lead SNPs from the univariate, European GWAS were extracted, and a linear regression was conducted in the non-European individuals. Of the 195 lead SNPs from the univariate GWAS of total CC and the univariate GWAS of each subregion, 140 were present in the non-European replication sample. Fifty-four of the SNPs are nominally significant (p < 0.05), and five SNPs remained significant after correction for multiple testing (Bonferroni, p < 0.05/140 = 3.57 × 10-4). Consistent directions of allelic effect were observed in the non-European analysis when compared to the European lead SNPs, with 89% of the SNPs having the same direction of effect (Supplementary Material, Table 2).

Functional annotation

Significant gene sets for the GWAS of each subregion and total CC are listed in the supplementary material (Supplementary Material, Table 3). Gene sets identified for the anterior, mid-anterior and central subregion overlapped significantly with previous GWAS of BD, alcohol use disorder, Alzheimer’s dementia, loneliness and neuroticism (Nagel et al., Reference Nagel, Jansen and Stringer2018; Turley et al., Reference Turley, Walters and Maghzian2018; Jansen et al., Reference Jansen, Savage and Watanabe2019; Kranzler et al., Reference Kranzler, Zhou and Kember2019; Stahl et al., Reference Stahl, Breen and Forstner2019). Gene sets identified for the posterior subregion overlapped with multiple hippocampal phenotypes and implicated cytoskeleton organisation (van der Meer, Rokicki, et al., Reference van der Meer, Rokicki and Kaufmann2020). Manhattan plots from the gene-based association tests, conducted through FUMA, can be found in the supplementary material (Supplementary Material, Figure 4). PheWAS, conducted with Open Target, implicated kidney and liver function (Supplementary Material, Table 4).

SNP-based heritability and genetic correlation

SNP-based heritability was estimated using LDSC for both the FreeSurfer and the Hofer and Frahm segmentation (B. Bulik-Sullivan et al., Reference Bulik-Sullivan, Finucane and Anttila2015; B. K. Bulik-Sullivan et al., Reference Bulik-Sullivan, Loh and Finucane2015). The heritability of total CC volume was estimated to be 0.38 (SE = 0.03) in the FreeSurfer segmentation, with the posterior subregion having the highest heritability estimate (h 2 SNP = 0.37, SE = 0.03) and the mid-posterior subregion having the lowest (h 2 SNP = 0.22, SE = 0.03). The heritability estimates for the Hofer and Frahm segmentation showed the mid-anterior subregion having the highest heritability (h 2 SNP = 0.32, SE = 0.03) and the mid-posterior subregion having the lowest (h 2 SNP = 0.26, SE = 0.02). All heritability estimates for the FreeSurfer segmentation are illustrated in Fig. 3A, and a full table of heritability estimates for the Hofer and Frahm segmentation can be found in the supplementary material (Supplementary Material, Table 5).

Figure 3. (A) SNP-based heritability estimates of total CC volume and volume of the subregions. LDSC was used to estimate the SNP-based heritability of total CC volume and the five subregions. A schematic of the CC and the five subregions, extracted using FreeSurfer, based on the Witelson segmentation. The CC and subregions have been overlaid onto an anatomical reference image. (B) Heat map of genetic correlations between CC subregions and relevant psychiatric and substance use traits each square represents the r-squared values, and significant correlations are denoted by *

Using LDSC, we found significant genetic correlations between total CC volume, BD (r g = −0.09, SE = 0.03; p = 5.9 × 10−3) and drinks per week (r g = −0.09, SE = 0.02; p = 4.8 × 10−4). Additionally, we found the significant genetic correlations between volume of the mid-anterior subregion BD (r g = −0.12, SE = 0.02; p = 2.5 × 10−4), MDD (r g = −0.12, SE = 0.04; p = 3.6 × 10−3), drinks per week (r g = −0.13, SE = 0.04; p = 1.8 × 10−3) and cannabis (r g = −0.09, SE = 0.03; p = 8.4 × 10−3) (Fig. 3B). A full table of these results can be found in the supplementary material (Supplementary Material, Table 6).

Discussion

Several key findings emerged from this work. First, we identified 70 unique genetic loci contributing to CC volume, with distributed associations across the structure, indicating a polygenic architecture. Second, subregions of the CC also have different genetic architecture, providing evidence that the CC and its subregions have both overlapping and unique genetic contributions. Third, we note significant genetic overlap of CC and CC subregions with neuropsychiatric disorders, including BD, MDD, ASD, as well as cannabis and alcohol consumption.

We identified 70 novel loci that were associated with CC volume, and these loci had distributed effects across the five subregions of the CC. This finding suggests that there is a similar biology underlying the development of these different subregions, and that they may share common genetic pathways. By using multivariate analysis, we were able to boost genetic discovery and gain a more comprehensive understanding of the genetic architecture of the CC and its subregions.

Our univariate GWAS indicates that the five subregions making up the CC have both a shared and distinct genetic aetiology, with the loci associated with whole CC volume only partially overlapping with those of the individual CC subregions. We identified a number of novel loci associated with CC subregional volume (33 significant loci in the anterior subregion, 23 in the mid-anterior, 29 in the central, 7 in the mid-posterior and 56 in the posterior subregion). Although there were no SNPs that overlapped with all subregions, rs755857 was significant in each GWAS except for that of the mid-posterior. Although rs755857 has not been previously associated with any other trait, it is associated with gene PINX1. Alterations in PINX1 have been associated with gamma glutamyl transferase and serum alkaline phosphatase levels both of which are used as clinical markers of liver function (Barton et al., Reference Barton, Sherman, Mukamel and Loh2021).

The SNP-based heritabilities of the CC subregions ranged from 0.22 (mid-posterior) to 0.37 (posterior), with a SNP-based heritability of 0.38 for total CC volume. These findings indicate that the CC and its subregions have a higher SNP-based heritability compared to subcortical brain volumes such as the amygdala and hippocampus (Satizabal et al., Reference Satizabal, Adams and Hibar2019; van der Meer, Frei, et al., Reference van der Meer, Frei and Kaufmann2020; Mufford et al., Reference Mufford, van der Meer and Kaufmann2021; Bahrami et al., Reference Bahrami, Nordengen and Shadrin2022; Ou et al., Reference Ou, Wu and Ge2023). Our heritability estimates are consistent with previous family reports indicating that the size of the CC and its subregions are under strong genetic influence (Woldehawariat et al., Reference Woldehawariat, Martinez and Hauser2014). The anterior and posterior subregions have the highest heritability estimates (h 2 SNP = 0.32 and h 2 SNP = 0.37, respectively), consistent with previous work Woldehawariat et al. (Reference Woldehawariat, Martinez and Hauser2014), who found the anterior and posterior subregions of the CC to have the highest heritability estimates. Our findings corroborate this pattern of high heritability estimates in the anterior and posterior subregions, and lower estimates in the mid-anterior and mid-posterior subregions.

Gene-set analyses conducted through FUMA indicated a degree of overlap between the subregions, with the exception of the mid-posterior subregion. The mid-posterior subregion has a distinct set of significant SNPs, with minimal genetic overlap with the other subregions and no gene sets in common with the other subregions. Additionally, the mid-posterior subregion had the lowest number of significant loci and the smallest heritability estimate. This is in line with previous work which shows that the mid-posterior subregion is difficult to examine; perhaps due to its smaller size upon extraction, but to date there are no studies that have assessed its heritability in humans. Studies looking at its heritability in other primates suggest that it has a lower heritability when compared to larger subregions, such as the anterior subregion (Hopkins et al., Reference Hopkins, Westerhausen, Schapiro and Sherwood2022). Although we found no evidence to suggest a relationship between increasing subregion volume and the number of significant loci detected, this remains a possibility. The subregion has projections into the posterior parietal and superior temporal cortex which also have lower heritability estimates (h 2 SNP = 0.07, SE = 0.02; h 2 SNP = 0.12, SE = 0.02, respectively) when compared to other cortical regions, which may explain the lower heritability seen for this region (Grasby et al., Reference Grasby, Jahanshad and Painter2020).

We found significant genetic overlap with previous GWAS of BD, alcohol use disorder and ASD. This is consistent with evidence of CC involvement in these conditions (Prunas et al., Reference Prunas, Delvecchio and Perlini2018; de Souza et al., Reference de Souza, Rosa, Escobar, Gasparetto and Nakamura-Palacios2019; Valenti et al., Reference Valenti, Pino, Mazza, Panzarino, Di Paolantonio and Verrotti2020). Our genetic correlation results also echoed findings from previous imaging studies. We found significant negative genetic correlation of BD with total CC volume, as well as volumes of the mid-anterior and central subregions. Consistently, BD patients have been shown to have significantly smaller CC subregions in comparison with healthy controls (Arnone, et al., Reference Arnone, Abou-Saleh and Barrick2006; Gifuni et al., Reference Gifuni, Olié and Ding2017; Prunas et al., Reference Prunas, Delvecchio and Perlini2018). Volumetric MRI studies have consistently demonstrated reduced CC volume in individuals with BD and their healthy co-twins, implying that these abnormalities may be related to genetic factors (Thompson et al., Reference Thompson, Narr, Blanton and Toga2003; Bearden et al., Reference Bearden, van Erp and Dutton2011). Further, volumetric reductions in the CC have been observed in relatives of patients with BD, showing intermediate volumetric reductions when compared to both controls and BD probands (Walterfang et al., Reference Walterfang, Malhi and Wood2009; Francis et al., Reference Francis, Mothi and Mathew2016). Our findings add additional evidence to previous work, indicating that altered CC volumes in BD may be due to a shared genetic aetiology. Reductions in total CC volume have also been reported in ASD (Frazier and Hardan, Reference Frazier and Hardan2009; Valenti et al., Reference Valenti, Pino, Mazza, Panzarino, Di Paolantonio and Verrotti2020). Additionally, there appears to be a gradient of volumetric reduction whereby the anterior subregions show more pronounced effect and the posterior comparatively less affected (Frazier and Hardan, Reference Frazier and Hardan2009). Our genetic correlation results are in line with these findings: there was a significant correlation between ASD and the central subregion.

Previous work has also shown that alcohol consumption affects CC morphometry, irrespective of an alcohol use disorder diagnosis (de Souza et al., Reference de Souza, Rosa, Escobar, Gasparetto and Nakamura-Palacios2019). Here, we find significant genetic correlations between the number of drinks per week and total CC, mid-anterior and central subregion volumes. Since alterations in the CC are consistently noted in individuals who consume alcohol, our findings suggest that there may be a genetic underpinning to these changes. The exact nature of this relationship warrants further investigation. The top SNP from our multivariate analysis, rs11245366, is associated with the gene EEF1AKMT2. PheWAS has found this SNP to be associated with cystatin C, creatine and urate levels (Barton et al., Reference Barton, Sherman, Mukamel and Loh2021). These phenotypes are related to kidney function and are markers of kidney dysfunction. Abnormalities in renal function have been consistently noted in individuals with agenesis of the CC; additionally, alterations in white matter structure are present in individuals with chronic kidney disease (Franco et al., Reference Franco, Kogan and Fisher1993; Liu et al., Reference Liu, Wu and Wu2020). Here, we provide preliminary evidence for a shared genetic aetiology between these phenotypes. The genetic relationship between the CC and renal phenotypes provides an interesting new path for future investigation.

Some limitations deserve emphasis. First, our study was limited to individuals residing in the United Kingdom and consisted largely of individuals of European ancestry. We did, however, replicate our findings in an independent sample of non-European ancestry. Future genetic work may benefit additional analyses conducted in more diverse cohorts. Second, the age range of individuals in the UK Biobank is limited to those above 45 years old. Therefore, it would be important to replicate these findings in a younger sample to ensure generalisability. Third, the comparison of our results with the literature is challenging due to differences in approach to subregion divisions. Although we implemented two segmentation protocols with consistent results, a robust CC segmentation is needed to improve replicability of research in this field.

In conclusion, our results demonstrate that the CC has a polygenic architecture with multiple genes contributing to CC volume but also suggest that distinct genetic factors are involved in the development of anterior and posterior subregions, consistent with their divergent functional specialisation. The pheWAS findings emphasise the importance of liver and kidney function in CC volume. We provide the first SNP-based evidence that CC and subregion volumes are heritable. Genetic correlations of CC and subregions volumes with specific psychiatric traits are noteworthy and deserv further investigation.

Supplementary material

The supplementary material for this article can be found at https://doi.org/10.1017/neu.2023.32

Author contribution

All authors contributed to the conceptualisation of the work. ML Campbell, A Shadrin and J Rokicki were involved in the analysis. All authors contributed to the draft of the manuscript.

Financial support

This work was supported in part by the NIMH (DS, 1R01 – MH120642)

Competing interests

The authors declare none.

Note in support

The work in this paper was presented at the World Congress of Psychiatric Genetics (Florence, 2022), the American Society of Human Genetics (Los Angeles, 2022) and at the International Congress of Human Genetics (Cape Town, 2023) (Campbell et al., Reference Campbell, Dalvie and Shadrin2022). While in the process of uploading this paper for submission, we noted a paper by Chen et al., Reference Chen, Wu and Ge2023 (Chen et al., Reference Chen, Wu and Ge2023). This publication takes a similar approach, albeit with some differences in methods and results, providing additional support for our findings.

References

Abramovic, L, Boks, MPM, Vreeker, A, et al. (2018) White matter disruptions in patients with bipolar disorder. European Neuropsychopharmacology 28, 743–751. DOI: 10.1016/j.euroneuro.2018.01.001.CrossRefGoogle ScholarPubMed
Arnone, D, Abou-Saleh, MT and Barrick, TR (2006) Diffusion tensor imaging of the corpus callosum in addiction. Neuropsychobiology 54(2), 107–113.CrossRefGoogle ScholarPubMed
Bahrami, S, Nordengen, K, Shadrin, AA, et al. (2022) Distributed genetic architecture across the hippocampal formation implies common neuropathology across brain disorders. Nature Communications 13(1), 3436. DOI: 10.1038/s41467-022-31086-w.CrossRefGoogle ScholarPubMed
Barton, AR, Sherman, MA, Mukamel, RE and Loh, P-R (2021) Whole-exome imputation within UK Biobank powers rare coding variant association and fine-mapping analyses. Nature Genetics 53(8), 1260–1269. DOI: 10.1038/s41588-021-00892-1.CrossRefGoogle ScholarPubMed
Bearden, CE, van Erp, TGM, Dutton, RA, et al. (2011) Mapping corpus callosum morphology in twin pairs discordant for bipolar disorder. Cerebral Cortex 21(10), 2415–2424. DOI: 10.1093/cercor/bhr030.CrossRefGoogle ScholarPubMed
Bulik-Sullivan, B, Finucane, HK, Anttila, V, et al. (2015) An atlas of genetic correlations across human diseases and traits. Nature Genetics 47(11), 1236–1241.CrossRefGoogle ScholarPubMed
Bulik-Sullivan, BK, Loh, P-R, Finucane, HK, et al. (2015) LD Score regression distinguishes confounding from polygenicity in genome-wide association studies. Nature Genetics 47(3), 291–295.CrossRefGoogle ScholarPubMed
Bycroft, C, Freeman, C, Petkova, D, et al. (2018) The UK Biobank resource with deep phenotyping and genomic data. Nature 562(7726), 203–209.CrossRefGoogle ScholarPubMed
Campbell, M, Dalvie, S, Shadrin, A, et al. (2022) The genetic architecture of the corpus callosum and its subregions. European Neuropsychopharmacology 63, e142. DOI: 10.1016/j.euroneuro.2022.07.263.CrossRefGoogle Scholar
Chang, CC, Chow, CC, Tellier, LC, Vattikuti, S, Purcell, SM and Lee, JJ (2015) Second-Generation PLINK: rising to the challenge of larger and richer datasets. Gigascience 4, s13742-015.CrossRefGoogle Scholar
Chen, S-J, Wu, B-S, Ge, Y-J, et al. (2023) The genetic architecture of the corpus callosum and its genetic overlap with common neuropsychiatric diseases. Journal of Affective Disorders 335, 418–430. DOI: 10.1016/j.jad.2023.05.002.CrossRefGoogle ScholarPubMed
Connaughton, M, Whelan, R, O'Hanlon, E and McGrath, J (2022) White matter microstructure in children and adolescents with ADHD. NeuroImage: Clinical 33, 102957. DOI: 10.1016/j.nicl.2022.102957.CrossRefGoogle ScholarPubMed
Cover, GS, Herrera, WG, Bento, MP, Appenzeller, S and Rittner, L (2018) Computational methods for corpus callosum segmentation on MRI: a systematic literature review. Computer Methods and Programs in Biomedicine 154, 25–35. DOI: 10.1016/j.cmpb.2017.10.025.CrossRefGoogle ScholarPubMed
de Souza, RSM, Rosa, M Jr., Escobar, TDC, Gasparetto, EL and Nakamura-Palacios, EM (2019) Anterior to midposterior corpus callosum subregions are volumetrically reduced in male alcoholics but only the anterior segment is associated to alcohol use. Frontiers in Psychiatry 10, 196. DOI: 10.3389/fpsyt.2019.00196.CrossRefGoogle ScholarPubMed
Fortin, J-P, Cullen, N, Sheline, YI, et al. (2018) Harmonization of cortical thickness measurements across scanners and sites. NeuroImage 167, 104–120. DOI: 10.1016/j.neuroimage.2017.11.024.CrossRefGoogle ScholarPubMed
Francis, AN, Mothi, SS, Mathew, IT, et al. (2016) Callosal abnormalities across the psychosis dimension: bipolar schizophrenia network on intermediate phenotypes. Biological Psychiatry 80(8), 627–635. DOI: 10.1016/j.biopsych.2015.12.026.CrossRefGoogle ScholarPubMed
Franco, I, Kogan, S, Fisher, J, et al. (1993) Genitourinary malformations associated with agenesis of the corpus callosum. The Journal of Urology 149(5), 1119–1121. DOI: 10.1016/s0022-5347(17)36313-9.CrossRefGoogle ScholarPubMed
Frazier, TW and Hardan, AY (2009) A meta-analysis of the corpus callosum in autism. Biological Psychiatry 66(10), 935–941. DOI: 10.1016/j.biopsych.2009.07.022.CrossRefGoogle ScholarPubMed
Gifuni, AJ, Olié, E, Ding, Y, et al. (2017) Corpus callosum volumes in bipolar disorders and suicidal vulnerability. Psychiatry Research: Neuroimaging 262, 47–54. DOI: 10.1016/j.pscychresns.2017.02.002.CrossRefGoogle ScholarPubMed
Grasby, KL, Jahanshad, N, Painter, JN, et al. (2020) The genetic architecture of the human cerebral cortex. Science 367(6484). DOI: 10.1126/science.aay6690.CrossRefGoogle ScholarPubMed
Herron, TJ, Kang, X and Woods, DL (2012) Automated measurement of the human corpus callosum using MRI. Frontiers in Neuroinformatics 6, 25. DOI: 10.3389/fninf.2012.00025.CrossRefGoogle ScholarPubMed
Hofer, S and Frahm, J (2006) Topography of the human corpus callosum revisited—comprehensive fiber tractography using diffusion tensor magnetic resonance imaging. NeuroImage 32(3), 989–994. DOI: 10.1016/j.neuroimage.2006.05.044.CrossRefGoogle ScholarPubMed
Hopkins, WD, Westerhausen, Ré, Schapiro, S and Sherwood, CC (2022) Heritability in corpus callosum morphology and its association with tool use skill in chimpanzees (Pan troglodytes): reproducibility in two genetically isolated populations. Genes Brain and Behavior 21(2), e12784. DOI: 10.1111/gbb.12784.CrossRefGoogle ScholarPubMed
Jansen, IE, Savage, JE, Watanabe, K, et al. (2019) Genome-wide meta-analysis identifies new loci and functional pathways influencing Alzheimers disease risk. Nature Genetics 51(3), 404–413. DOI: 10.1038/s41588-018-0311-9.CrossRefGoogle ScholarPubMed
Kranzler, HR, Zhou, H, Kember, RL, et al. (2019) Genome-wide association study of alcohol consumption and use disorder in 274,424 individuals from multiple populations. Nature Communications 10(1), 1499. DOI: 10.1038/s41467-019-09480-8.CrossRefGoogle ScholarPubMed
Liu, J, Li, S, Li, X, et al. (2021) Genome-wide association study followed by trans-ancestry meta-analysis identify 17 new risk loci for schizophrenia. BMC Medicine 19(1), 177. DOI: 10.1186/s12916-021-02039-9.CrossRefGoogle ScholarPubMed
Liu, M, Wu, Y, Wu, X, et al. (2020) White matter microstructure changes and cognitive impairment in the progression of chronic kidney disease. Frontiers in Neuroscience 14. Availabl at https://www.frontiersin.org/articles/10.3389/fnins.2020.559117 (accessed 2 April 2023).CrossRefGoogle ScholarPubMed
Liu, M, Jiang, Y, Wedow, R, et al. (2019) Association studies of up to 1.2 million individuals yield new insights into the genetic etiology of tobacco and alcohol use. Nature Genetics 51(2), 237–244. DOI: 10.1038/s41588-018-0307-5.CrossRefGoogle ScholarPubMed
Loughnan, RJ, Shadrin, AA, Frei, O, et al. (2022) Generalization of cortical MOSTest genome-wide associations within and across samples. NeuroImage 263, 119632. DOI: 10.1016/j.neuroimage.2022.119632.CrossRefGoogle ScholarPubMed
Mountjoy, E, Schmidt, EM, Carmona, M, et al. (2021) An open approach to systematically prioritize causal variants and genes at all published human GWAS trait-associated loci. Nature Genetics 53(11), 1527–1533. DOI: 10.1038/s41588-021-00945-5.CrossRefGoogle ScholarPubMed
Mufford, MS, van der Meer, D, Kaufmann, T, et al. (2021) The genetic architecture of amygdala nuclei. DOI: 10.1101/2021.06.30.21258615.CrossRefGoogle Scholar
Mullins, N, Forstner, AJ, O’Connell, KS, et al. (2021) Genome-wide association study of over 40,000 bipolar disorder cases provides new insights into the underlying biology. medRxiv. DOI: 10.1101/2020.09.17.20187054.Google Scholar
Nagel, M, Jansen, PR, Stringer, S, et al. (2018) Meta-analysis of genome-wide association studies for neuroticism in 449,484 individuals identifies novel genetic loci and pathways. Nature Genetics 50(7), 920–927. DOI: 10.1038/s41588-018-0151-7.CrossRefGoogle Scholar
Newbury, A and Rosen, G (2012) Genetic, morphometric, and behavioral factors linked to the midsagittal area of the corpus callosum. Frontiers in Genetics 3, 91. DOI: 10.3389/fgene.2012.00091.CrossRefGoogle Scholar
Ou, Y-N, Wu, B-S, Ge, Y-J, et al. (2023) The genetic architecture of human amygdala volumes and their overlap with common brain disorders. Translational Psychiatry 13(1), 90. DOI: 10.1038/s41398-023-02387-5.CrossRefGoogle ScholarPubMed
Pasman, JA, Verweij, KJH, Gerring, Z, et al. (2018) GWAS of lifetime cannabis use reveals new risk loci, genetic overlap with psychiatric traits, and a causal effect of schizophrenia liability. Nature Neuroscience 21(9), 1161–1170.CrossRefGoogle Scholar
PGC ASD (2017) Meta-analysis of GWAS of over 16,000 individuals with autism spectrum disorder highlights a novel locus at 10q24.32 and a significant overlap with schizophrenia. Molecular Autism 8(1), 21. DOI: 10.1186/s13229-017-0137-9.CrossRefGoogle Scholar
Pouget, JG, Wu, Y, Mignot, E, et al. (2019) Cross-disorder analysis of schizophrenia and 19 immune-mediated diseases identifies shared genetic risk. Human Molecular Genetics 28(20), 3498–3513. DOI: 10.1093/hmg/ddz145.CrossRefGoogle Scholar
Prunas, C, Delvecchio, G, Perlini, C, et al. (2018) Diffusion imaging study of the Corpus Callosum in bipolar disorder. Psychiatry Research: Neuroimaging 271, 75–81. DOI: 10.1016/j.pscychresns.2017.11.001.CrossRefGoogle ScholarPubMed
Rosenbloom, M, Sullivan, EV and Pfefferbaum, A (2003) Using magnetic resonance imaging and diffusion tensor imaging to assess brain damage in alcoholics. Alcohol Research & Health 27, 146–152.Google ScholarPubMed
Satizabal, CL, Adams, HHH, Hibar, DP, et al. (2019) Genetic architecture of subcortical brain structures in 38,851 individuals. Nature Genetics 51(11), 1624–1636. DOI: 10.1038/s41588-019-0511-y.CrossRefGoogle ScholarPubMed
Scamvougeras, A, Kigar, DL, Jones, D, Weinberger, DR and Witelson, SF (2003) Size of the human corpus callosum is genetically determined: an MRI study in mono and dizygotic twins. Neuroscience Letters 338, 91–94. DOI: 10.1016/S0304-3940(02)01333-2.CrossRefGoogle ScholarPubMed
Stahl, EA, Breen, G, Forstner, AJ, et al. (2019) Genome-wide association study identifies 30 loci associated with bipolar disorder. Nature Genetics 51(5), 793–803.CrossRefGoogle ScholarPubMed
Stergiakouli, E, Hamshere, M, Holmans, P, et al. (2012) Investigating the contribution of common genetic variants to the risk and pathogenesis of ADHD. The American Journal of Psychiatry 169(2), 186–194. DOI: 10.1176/appi.ajp.2011.11040551.CrossRefGoogle Scholar
Thompson, PM, Narr, KL, Blanton, RE and Toga, AW (2003) Mapping structural alterations of the corpus callosum during brain development and degeneration.Google Scholar
Turley, P, Walters, RK, Maghzian, O, et al. (2018) Multi-trait analysis of genome-wide association summary statistics using MTAG. Nature Genetics 50(2), 229–237. DOI: 10.1038/s41588-017-0009-4.CrossRefGoogle ScholarPubMed
Valenti, M, Pino, MC, Mazza, M, Panzarino, G, Di Paolantonio, C and Verrotti, A (2020) Abnormal structural and functional connectivity of the corpus callosum in autism spectrum disorders: a review. Review Journal of Autism and Developmental Disorders 7(1), 46–62. DOI: 10.1007/s40489-019-00176-9.CrossRefGoogle Scholar
van der Meer, D, Frei, O, Kaufmann, T, et al. (2020) Understanding the genetic determinants of the brain with MOSTest. Nature Communications 11(1), 3512. DOI: 10.1038/s41467-020-17368-1.CrossRefGoogle ScholarPubMed
van der Meer, D, Rokicki, J, Kaufmann, T, et al. (2020) Brain scans from 21,297 individuals reveal the genetic architecture of hippocampal subfield volumes. Molecular Psychiatry 25(11), 3053–3065. DOI: 10.1038/s41380-018-0262-7.CrossRefGoogle Scholar
Walterfang, M, Malhi, GS, Wood, AG, et al. (2009) Corpus callosum size and shape in established bipolar affective disorder. Australian & New Zealand Journal of Psychiatry 43(9), 838–845. DOI: 10.1080/00048670903107534.CrossRefGoogle ScholarPubMed
Wang, P, Meng, C, Yuan, R, et al. (2020) The organization of the human corpus callosum estimated by intrinsic functional connectivity with white-matter functional networks. Cerebral Cortex 30(5), 3313–3324. DOI: 10.1093/cercor/bhz311.CrossRefGoogle ScholarPubMed
Watanabe, K, Taskesen, E, van Bochoven, A and Posthuma, D (2017) Functional mapping and annotation of genetic associations with FUMA. Nature Communications 8(1), 1–11.CrossRefGoogle ScholarPubMed
Witelson, SF (1989) Hand and sex differences in the isthmus and genu of the human corpus callosum: a postmortem morphological study. Brain 112, 799–835. DOI: 10.1093/brain/112.3.799.CrossRefGoogle ScholarPubMed
Woldehawariat, G, Martinez, PE, Hauser, P, et al. (2014) Corpus callosum size is highly heritable in humans, and may reflect distinct genetic influences on ventral and rostral regions. PLoS ONE 9(6), e99980. DOI: 10.1371/journal.pone.0099980.CrossRefGoogle ScholarPubMed
Wray, NR, Ripke, S, Mattheisen, M, et al. (2018) Genome-wide association analyses identify 44 risk variants and refine the genetic architecture of major depression. Nature Genetics 50(5), 668–681. DOI: 10.1038/s41588-018-0090-3.CrossRefGoogle ScholarPubMed
Zhao, B, Li, T, Yang, Y, et al. (2020) Common genetic variation influencing human white matter microstructure. preprint. Genetics. DOI: 10.1101/2020.05.23.112409.Google Scholar
Zhao, G, Lau, WK W, Wang, C, et al. (2022) A comparative multimodal meta-analysis of anisotropy and volume abnormalities in white matter in people suffering from bipolar disorder or schizophrenia. Schizophrenia Bulletin 48(1), 69–79. DOI: 10.1093/schbul/sbab093.CrossRefGoogle ScholarPubMed
Zhou, L, Wang, L, Wang, M, et al. (2022) Alterations in white matter microarchitecture in adolescents and young adults with major depressive disorder: a voxel-based meta-analysis of diffusion tensor imaging. Psychiatry Research: Neuroimaging 323, 111482. DOI: 10.1016/j.pscychresns.2022.111482.CrossRefGoogle Scholar
Figure 0

Figure 1. Miami plot of the −log10(p) statistics from the multivariate (top) and univariate (bottom) GWAS of total CC volume. (A) Top: The multivariate analysis was conducted using MOSTest and illustrate the distributive genetic architecture of the CC. Bottom: Individual univariate GWAS were conducted for total CC volume and each subregion, indicating both a shared and unique genetic aetiology of the CC subregions. (B) Top SNPs from the multivariate GWAS show subregion-specific effects across the CC. The top 10 SNPs from the multivariate GWAS were mapped to the CC subregions using the Z-scores from the respective univariate subregion GWAS. The results indicate a gradient of effect across the posterior and anterior subregions.

Figure 1

Figure 2. (A) Manhattan plots from univariate GWAS of each CC subregion. (B) Quantile-Quantile plot (QQ plot) of observed versus expected −log10(p-values) the GWAS of CC subregions. The shaded regions represent the expected null distribution of −log10(p-values) under the assumption of no association. (C) Volumetric correlations of CC subregions, each square represents the r-squared values.

Figure 2

Figure 3. (A) SNP-based heritability estimates of total CC volume and volume of the subregions. LDSC was used to estimate the SNP-based heritability of total CC volume and the five subregions. A schematic of the CC and the five subregions, extracted using FreeSurfer, based on the Witelson segmentation. The CC and subregions have been overlaid onto an anatomical reference image. (B) Heat map of genetic correlations between CC subregions and relevant psychiatric and substance use traits each square represents the r-squared values, and significant correlations are denoted by *

Supplementary material: File

Campbell et al. supplementary material

Campbell et al. supplementary material

Download Campbell et al. supplementary material(File)
File 2.2 MB