Advertisement
Archival Report|Articles in Press

Pathway-based polygenic risk scores for schizophrenia and associations with reported psychotic-like experiences and neuroimaging phenotypes in UK Biobank

Open AccessPublished:March 24, 2023DOI:https://doi.org/10.1016/j.bpsgos.2023.03.004
Pathway-based polygenic risk scores for schizophrenia and associations with reported psychotic-like experiences and neuroimaging phenotypes in UK Biobank
Previous ArticleRepeated Social Defeat Stress induces an Inflammatory Gut Milieu by altering the Mucosal Barrier Integrity and Gut Microbiota Homeostasis
Next ArticleGenetic decomposition of the heritable component of reported childhood maltreatment

      Abstract

      Background

      Schizophrenia is a heritable psychiatric disorder with a polygenic architecture. Genome-wide association studies (GWAS) report an increasing number of risk-associated variants and polygenic risk scores (PRS) explain 17% of the variance in the disorder. There exists substantial heterogeneity in the effect of these variants and aggregating them based on biologically-relevant functions may provide mechanistic insight into the disorder.

      Methods

      Using the largest schizophrenia GWAS we associated PRS based on 5 gene-sets previously found to contribute to schizophrenia pathophysiology: postsynaptic density of excitatory synapses, postsynaptic membrane, dendritic spine, axon, and histone H3-K4 methylation, along with respective whole-genome PRS, with neuroimaging (N>29,000) and reported psychotic-like experiences (PLEs) (N>119,000) variables in healthy UK Biobank subjects.

      Results

      Several variables were significantly associated with the axon gene-set (psychotic-like communications, parahippocampal gyrus volume, FA thalamic radiations and FA posterior thalamic radiations (β range: -0.016-0.0916, pFDR=<0.05), postsynaptic density gene-set (PLE-distress, global surface area and cingulate lobe surface area (β range: -0.014-0.0588, pFDR=<0.05) and histone gene-set (entorhinal surface area: β=-0.016, pFDR=0.035). From these, whole-genome PRS were significantly associated with psychotic-like communications (β=0.2218, pFDR=1.34x10-7), distress (β=0.1943, pFDR=7.28x10-16), and FA thalamic radiations (β=-0.0143, pFDR=0.036). Permutation analysis revealed these associations were not due to chance.

      Conclusions

      Our results indicate that genetic variation in 3 gene-sets relevant to schizophrenia may confer risk for the disorder through effects on previously implicated neuroimaging variables. As associations were overall stronger for whole-genome PRS, findings here highlight that selection of biologically relevant variants is not yet sufficient to address the heterogeneity of the disorder.

      Keywords

      Introduction

      Schizophrenia is a psychiatric disorder characterised by positive and negative symptoms, as well as marked cognitive impairment (
      • Owen MJ
      • Sawa A
      • Mortensen PB
      Schizophrenia. Lancet. 2016 Jul 2; 388: 86-97
      ). Schizophrenia is thought to result from a complex combination of genetic and environmental factors, and its heritability is estimated at 80% (
      • Hilker R
      • Helenius D
      • Fagerlund B
      • Skytthe A
      • Christensen K
      • Werge TM
      • et al.
      Heritability of Schizophrenia and Schizophrenia Spectrum Based on the Nationwide Danish Twin Register.
      Biol Psychiatry. 2018 Mar 15; 83: 492-498
      ).
      Genome-wide association studies (GWAS) have reported increasing numbers of genomic loci associated with schizophrenia, lending support to the contribution of common genetic variants to the pathophysiology of schizophrenia (

      Lam M, Chen CY, Li Z, Martin AR, Bryois J, Ma X, et al. Comparative genetic architectures of schizophrenia in East Asian and European populations. Nat Genet [Internet]. 2019 Dec 1 [cited 2022 Apr 22];51(12):1670–8. Available from: https://pubmed.ncbi.nlm.nih.gov/31740837/

      ,

      Trubetskoy V, Pardiñas AF, Qi T, Panagiotaropoulou G, Awasthi S, Bigdeli TB, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nat 2022 6047906 [Internet]. 2022 Apr 8 [cited 2022 Apr 22];604(7906):502–8. Available from: https://www.nature.com/articles/s41586-022-04434-5

      ). Genome-wide polygenic risk scores (PRS) calculated from GWAS explain ∼17% (Nagelkerke’s R2) of the variance in schizophrenia (

      Trubetskoy V, Pardiñas AF, Qi T, Panagiotaropoulou G, Awasthi S, Bigdeli TB, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nat 2022 6047906 [Internet]. 2022 Apr 8 [cited 2022 Apr 22];604(7906):502–8. Available from: https://www.nature.com/articles/s41586-022-04434-5

      ). However, there exists substantial heterogeneity in the effect of risk variants, and genome-wide approaches may not be sufficient for patient stratification in downstream analyses. Mechanistic insight may be derived from GWASs, including the identification of biologically relevant gene-sets in which risk variants are aggregated. For instance, The Psychiatric Genomics Consortium (PGC) identified a number of pathways specific to schizophrenia, major depressive disorder (MDD), and bipolar disorder, but also common across the three, including histone methylation, immune, and neuronal signalling pathways (

      O’dushlaine C, Rossin L, Lee PH, Duncan L, Parikshak NN, Newhouse S, et al. Psychiatric genome-wide association study analyses implicate neuronal, immune and histone pathways. Nat Neurosci [Internet]. 2015 Feb 17 [cited 2022 Apr 22];18(2):199. Available from: /pmc/articles/PMC4378867/

      ).
      Combining the predictive power of PRS with findings from pathway analysis by examining genetic variation within biologically relevant gene-sets may address the inherent heterogeneity in schizophrenia and provide additional mechanistic insight by detecting associations with biologically informative traits. This methodology has been applied to a number of schizophrenia-relevant phenotypes. Rampino et al. (2017) found that schizophrenia PRS calculated based on single nucleotide polymorphisms (SNPs) implicated in glutamatergic signalling were associated with attention, a cognitive process known to be impaired in schizophrenia (
      • Rampino A
      • Taurisano P
      • Fanelli G
      • Attrotto M
      • Torretta S
      • Antonucci LA
      • et al.
      A Polygenic Risk Score of glutamatergic SNPs associated with schizophrenia predicts attentional behavior and related brain activity in healthy humans.
      Eur Neuropsychopharmacol. 2017 Sep 1; 27: 928-939
      ). Yao et al. (2021) found that PRS calculated based on neural microRNA-137 (MIR137) explained a disproportionately larger schizophrenia risk variance than genomic control PRS, when accounting for gene-set size (∼2% MIR137 N=∼1,000 genes; ∼10%, whole-genome N=∼20,000 genes) (
      • Yao Y
      • Guo W
      • Zhang S
      • Yu H
      • Yan H
      • Zhang H
      • et al.
      Cell type-specific and cross-population polygenic risk score analyses of MIR137 gene pathway in schizophrenia.
      iScience. 2021 Jul 23; 24102785
      ). It is therefore possible to interrogate genetic risk aggregated to biologically relevant gene-sets to gain insight into the association between aggregated genetic risk and disorder-specific traits of interest.
      Previous evidence has also shown schizophrenia PRS associations with disruptions in white matter microstructure and global and regional brain volumes, although results are inconsistent (

      Terwisscha Van Scheltinga AF, Bakker SC, Van Haren NEM, Derks EM, Buizer-Voskamp JE, Boos HBM, et al. Genetic schizophrenia risk variants jointly modulate total brain and white matter volume. Biol Psychiatry [Internet]. 2013 Mar 15 [cited 2022 Apr 26];73(6):525. Available from: /pmc/articles/PMC3573254/

      ,

      Alnæs D, Kaufmann T, Van Der Meer D, Córdova-Palomera A, Rokicki J, Moberget T, et al. Brain Heterogeneity in Schizophrenia and Its Association With Polygenic Risk. JAMA Psychiatry [Internet]. 2019 Jul 1 [cited 2022 Apr 26];76(7):739–48. Available from: https://jamanetwork.com/journals/jamapsychiatry/fullarticle/2730004

      ). As such, it was suggested that biologically relevant gene-sets may reveal stronger associations with neuroimaging phenotypes, as demonstrated recently (
      • Grama S
      • Willcocks I
      • Hubert JJ
      • Pardiñas AF
      • Legge SE
      • Bracher-Smith M
      • et al.
      Polygenic risk for schizophrenia and subcortical brain anatomy in the UK Biobank cohort.
      Transl Psychiatry. 2020 101; ([Internet]. 2020 Sep 9 [cited 2022 Apr 22];10(1):1–10. Available from:)
      https://www.nature.com/articles/s41398-020-00940-0
      ). Grama et al. (2020) investigated whether behaviour- and neuronal-related gene-sets, previously implicated in schizophrenia, are associated with subcortical volumes. They found that one gene-set, “abnormal behaviour”, was associated with right thalamic volume, and this association was robust across different p-value thresholds (
      • Grama S
      • Willcocks I
      • Hubert JJ
      • Pardiñas AF
      • Legge SE
      • Bracher-Smith M
      • et al.
      Polygenic risk for schizophrenia and subcortical brain anatomy in the UK Biobank cohort.
      Transl Psychiatry. 2020 101; ([Internet]. 2020 Sep 9 [cited 2022 Apr 22];10(1):1–10. Available from:)
      https://www.nature.com/articles/s41398-020-00940-0
      ). This methodology can be applied to other psychiatric disorders. For instance, PRS calculated for a previously established biological pathway (NETRIN1) in relation to MDD were associated with relevant neuroimaging phenotypes, shedding light on links between biology and neuroimaging (
      • Barbu MC
      • Zeng Y
      • Shen X
      • Cox SR
      • Clarke TK
      • Gibson J
      • et al.
      Association of Whole-Genome and NETRIN1 Signaling Pathway–Derived Polygenic Risk Scores for Major Depressive Disorder and White Matter Microstructure in the UK Biobank.
      Biol Psychiatry Cogn Neurosci Neuroimaging. 2019 Jan 1; 4: 91-100
      ). This indicates that meaningful associations with traits of interest may be revealed when applying genomic methods addressing relevant parts of the genome.
      Based on this evidence, we hypothesised that schizophrenia PRS aggregated in biologically relevant pathways previously shown to play a role in schizophrenia, would be associated with structural neuroimaging and reported-PLE phenotypes in a general population adult sample. Identifying associations with specific neuroimaging phenotypes may provide an opportunity to disentangle the heterogeneity of the disorder, both in terms of genetic risk and inconsistent previous neuroimaging findings. We therefore selected 5 gene-sets: postsynaptic density (PSD), postsynaptic membrane (PSM), dendritic spine, axon, and histone H3-K4 methylation, previously identified by the PGC (

      O’dushlaine C, Rossin L, Lee PH, Duncan L, Parikshak NN, Newhouse S, et al. Psychiatric genome-wide association study analyses implicate neuronal, immune and histone pathways. Nat Neurosci [Internet]. 2015 Feb 17 [cited 2022 Apr 22];18(2):199. Available from: /pmc/articles/PMC4378867/

      ). These 5 cellular components and biological processes have been associated with schizophrenia in a number of studies investigating human and animal models (
      • Moyer CE
      • Shelton MA
      • Sweet RA
      Dendritic spine alterations in schizophrenia.
      Neurosci Lett. 2015 Aug 5; 601: 46-53
      ,
      • Föcking M
      • Lopez LM
      • English JA
      • Dicker P
      • Wolff A
      • Brindley E
      • et al.
      Proteomic and genomic evidence implicates the postsynaptic density in schizophrenia.
      Mol Psychiatry. 2015 204; ([Internet]. 2014 Jul 22 [cited 2022 May 5];20(4):424–32. Available from:)
      https://www.nature.com/articles/mp201463
      ,

      Shen E, Shulha H, Weng Z, Akbarian S. Regulation of histone H3K4 methylation in brain development and disease. Philos Trans R Soc B Biol Sci [Internet]. 2014 [cited 2022 May 5];369(1652). Available from: https://royalsocietypublishing.org/doi/full/10.1098/rstb.2013.0514

      ). We calculated PRSs for each gene-set-specific set of SNPs and SNPs excluded from the gene-sets for paired comparisons (gene-set SNPs versus whole-genome (WG) minus gene-set SNPs PRS). We then tested their association with brain structural phenotypes (cortical volume, thickness, and surface area; white matter microstructure indexed by fractional anisotropy (FA) and mean diffusivity (MD); and subcortical volumes) and reported-psychotic-like experiences (PLEs, Mental Health Questionnaire (MHQ)) in UK Biobank (UKB), utilising the most up-to-date genetic, mental health, and imaging data. We hypothesized distinctive roles in the pathophysiology of schizophrenia for the different biologically relevant pathways tested, after adjustment for whole-genome PRSs (excluding SNPs in each gene-set), highlighting important mechanistic processes underlying the different phenotypes associated with schizophrenia. We expect this methodology to partly address the heterogeneity in schizophrenia through the identification of biologically relevant mechanisms.

      Methods and Materials

      Study population

      UKB comprises of 502,411 community-dwelling individuals recruited between 2006-2010 in the United Kingdom (https://biobank.ctsu.ox.ac.uk/crystal/field.cgi?id=200) (

      Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, et al. UK Biobank: An Open Access Resource for Identifying the Causes of a Wide Range of Complex Diseases of Middle and Old Age. PLOS Med [Internet]. 2015 Mar 1 [cited 2022 Apr 25];12(3):e1001779. Available from: https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1001779

      ). UKB received ethical approval from the research ethics committee (REC reference 11/NW/0382). This study was approved by the UKB Access Committee (Project No. 4844 and 16124). Written informed consent was obtained from all participants. This study was conducted using the latest release of UKB neuroimaging data (Ncortical=29,791, Nsubcortical=29,536, Nwhite matter=27,917) and N=119,947 with MHQ data. We excluded individuals with a diagnosis of schizophrenia as indicated by ICD-10 (F20, https://biobank.ndph.ox.ac.uk/showcase/field.cgi?id=41270) due to the small proportion of diagnosed individuals (N=989) and due to the fact that our analyses focussed on associations in the general population.

      Gene-set selection

      The top 5 gene-sets associated with schizophrenia were selected from The Network and Pathway Analysis Subgroup of the PGC (

      O’dushlaine C, Rossin L, Lee PH, Duncan L, Parikshak NN, Newhouse S, et al. Psychiatric genome-wide association study analyses implicate neuronal, immune and histone pathways. Nat Neurosci [Internet]. 2015 Feb 17 [cited 2022 Apr 22];18(2):199. Available from: /pmc/articles/PMC4378867/

      ) and identified on Gene Ontology (GO) by searching for each pathway’s GO identifier in the discovery manuscript above (

      O’dushlaine C, Rossin L, Lee PH, Duncan L, Parikshak NN, Newhouse S, et al. Psychiatric genome-wide association study analyses implicate neuronal, immune and histone pathways. Nat Neurosci [Internet]. 2015 Feb 17 [cited 2022 Apr 22];18(2):199. Available from: /pmc/articles/PMC4378867/

      ,

      Gene Ontology Resource [Internet]. [cited 2022 Apr 26]. Available from: http://geneontology.org/

      ): postsynaptic density (GO:0014069), postsynaptic membrane (GO:0045211), dendritic spine (GO:0043197), axon (GO:0030424), and histone H3-K4 methylation (GO:0051568). The gene-sets were selected based on their robust association with schizophrenia in this, and previous studies (

      O’dushlaine C, Rossin L, Lee PH, Duncan L, Parikshak NN, Newhouse S, et al. Psychiatric genome-wide association study analyses implicate neuronal, immune and histone pathways. Nat Neurosci [Internet]. 2015 Feb 17 [cited 2022 Apr 22];18(2):199. Available from: /pmc/articles/PMC4378867/

      ,
      • Föcking M
      • Lopez LM
      • English JA
      • Dicker P
      • Wolff A
      • Brindley E
      • et al.
      Proteomic and genomic evidence implicates the postsynaptic density in schizophrenia.
      Mol Psychiatry. 2015 204; ([Internet]. 2014 Jul 22 [cited 2022 May 5];20(4):424–32. Available from:)
      https://www.nature.com/articles/mp201463
      ,

      Gupta S, Kim SY, Artis S, Molfese DL, Schumacher A, Sweatt JD, et al. Histone Methylation Regulates Memory Formation. J Neurosci [Internet]. 2010 Mar 10 [cited 2022 May 16];30(10):3589. Available from: /pmc/articles/PMC2859898/

      ,

      Zhu Y, Wang S, Gong X, Edmiston EK, Zhong S, Li C, et al. Associations between hemispheric asymmetry and schizophrenia-related risk genes in people with schizophrenia and people at a genetic high risk of schizophrenia. Br J Psychiatry [Internet]. 2021 Jul 1 [cited 2022 May 16];219(1):392–400. Available from: https://www.cambridge.org/core/journals/the-british-journal-of-psychiatry/article/associations-between-hemispheric-asymmetry-and-schizophreniarelated-risk-genes-in-people-with-schizophrenia-and-people-at-a-genetic-high-risk-of-schizophrenia/8506769E0926B8AF52DB15E6D3FBA614

      ). Further details on gene-sets are included in Supplementary Materials and Supplementary Excel File 1.

      Genotyping, SNP annotation and PRS profiling

      Genotyping of 488,000 blood samples from UKB participants was carried out using the UK BiLEVE (https://biobank.ctsu.ox.ac.uk/crystal/refer.cgi%3fid%3d149600) or UKB Axiom (https://biobank.ctsu.ox.ac.uk/crystal/refer.cgi%3fid%3d149601) arrays (

      Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp K, et al. The UK Biobank resource with deep phenotyping and genomic data. Nat 2018 5627726 [Internet]. 2018 Oct 10 [cited 2022 Apr 25];562(7726):203–9. Available from: https://www.nature.com/articles/s41586-018-0579-z

      ). Further information on genotyping procedures and QC are provided in https://biobank.ctsu.ox.ac.uk/crystal/crystal/docs/genotyping_qc.pdf and in Bycroft et al. (

      Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp K, et al. The UK Biobank resource with deep phenotyping and genomic data. Nat 2018 5627726 [Internet]. 2018 Oct 10 [cited 2022 Apr 25];562(7726):203–9. Available from: https://www.nature.com/articles/s41586-018-0579-z

      ). Genetic data from both the base and target datasets were annotated in reference to human genome build 19.
      We used SNPs from the largest available GWAS of schizophrenia (N=320,404, N=76,755 cases), carried out by Trubetskoy et al. (

      Trubetskoy V, Pardiñas AF, Qi T, Panagiotaropoulou G, Awasthi S, Bigdeli TB, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nat 2022 6047906 [Internet]. 2022 Apr 8 [cited 2022 Apr 22];604(7906):502–8. Available from: https://www.nature.com/articles/s41586-022-04434-5

      ). Here, they identified schizophrenia associations with common variants at 287 distinct loci, with PRSs explaining ∼17% of the variance in a European ancestry target sample (

      Trubetskoy V, Pardiñas AF, Qi T, Panagiotaropoulou G, Awasthi S, Bigdeli TB, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nat 2022 6047906 [Internet]. 2022 Apr 8 [cited 2022 Apr 22];604(7906):502–8. Available from: https://www.nature.com/articles/s41586-022-04434-5

      ). The GWAS sample utilised here did not include any individuals in UKB. Full SNP QC is detailed in Supplementary Materials and was based on Choi et al. ((
      • Choi SW
      • Mak TSH
      • O’Reilly PF
      Tutorial: a guide to performing polygenic risk score analyses.
      Nat Protoc. 2020 Sep 1; 15: 2759-2772
      ), https://choishingwan.github.io/PRS-Tutorial/base/). Additionally, we ensured that our sample consisted of unrelated, white British participants with no overlap with PGC (

      Smith BH, Campbell A, Linksted P, Fitzpatrick B, Jackson C, Kerr SM, et al. Cohort Profile: Generation Scotland: Scottish Family Health Study (GS:SFHS). The study, its participants and their potential for genetic research on health and illness. Int J Epidemiol [Internet]. 2013 Jun 1 [cited 2022 Mar 23];42(3):689–700. Available from: https://academic.oup.com/ije/article/42/3/689/909916

      ). The final genetic sample consisted of N=365,125 participants and N=5,974,990 SNPs, further reduced when combining with reported-PLE and imaging samples (see below).
      Following functional annotation (

      Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res [Internet]. 2010 Sep 1 [cited 2022 Apr 26];38(16):e164–e164. Available from: https://academic.oup.com/nar/article/38/16/e164/1749458

      ) (See Supplementary Materials), SNPs located within each gene-set were extracted. See Supplementary Table 1 and Supplementary Excel File 1 for the number of genes and SNPs within each gene-set. From these lists, gene-set PRSs for each individual in UKB were computed (See Supplementary Materials) using PRSice (

      Euesden J, Lewis CM, O’Reilly PF. PRSice: Polygenic Risk Score software. Bioinformatics [Internet]. 2015 May 1 [cited 2022 Apr 26];31(9):1466–8. Available from: https://pubmed.ncbi.nlm.nih.gov/25550326/

      ) at 5 p-value thresholds (0.01, 0.05, 0.1, 0.5, 1). Each gene-set PRS had its respective whole-genome (WG PRS) (i.e. each gene-set SNP list was input as an exclusion flag to create whole-genome PRS that did not include the specific gene-sets). The primary analysis in this manuscript comprises SNPs that met a significance level of 0.1 (

      Trubetskoy V, Pardiñas AF, Qi T, Panagiotaropoulou G, Awasthi S, Bigdeli TB, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nat 2022 6047906 [Internet]. 2022 Apr 8 [cited 2022 Apr 22];604(7906):502–8. Available from: https://www.nature.com/articles/s41586-022-04434-5

      ). Analyses at other thresholds are included in Supplementary Excel File 2 (Tables 5-11).

      Phenotypes

      Psychotic-like experiences

      A Mental Health Questionnaire (MHQ) was sent to participants who provided an e-mail address between July 2016-July 2017 (N=∼157,000). This included four unusual and psychotic experience items, the dichotomised responses to which were used to create the four lifetime PLEs utilised in this study which will be referred to as “conspiracies”, “communications”, “voices”, and “visions”, based on their clinical significance (See Supplementary Materials for the list of questions). We also examined distress by selecting individuals who reported PLEs as distressing (as opposed to neutral or positive) and those who reported no PLE to determine associations with experiencing stress (
      • Schoorl J
      • Barbu MC
      • Shen X
      • Harris MR
      • Adams MJ
      • Whalley HC
      • et al.
      Grey and white matter associations of psychotic-like experiences in a general population sample (UK Biobank).
      Transl Psychiatry. 2021; 111 ([Internet]. 2021 Jan 7 [cited 2022 Apr 26];11(1):1–11. Available from:)
      https://www.nature.com/articles/s41398-020-01131-7
      ,
      • Alloza C
      • Blesa-Cábez M
      • Bastin ME
      • Madole JW
      • Buchanan CR
      • Janssen J
      • et al.
      Psychotic-like experiences, polygenic risk scores for schizophrenia, and structural properties of the salience, default mode, and central-executive networks in healthy participants from UK Biobank.
      Transl Psychiatry. 2020 101; ([Internet]. 2020 Apr 27 [cited 2022 Apr 26];10(1):1–13. Available from:)
      https://www.nature.com/articles/s41398-020-0794-x
      ,

      Bosma MJ, Cox SR, Ziermans T, Buchanan CR, Shen X, Tucker-Drob EM, et al. White matter, cognition and psychotic-like experiences in UK Biobank. Psychol Med [Internet]. 2021 [cited 2022 Apr 26];1–10. Available from: https://www.cambridge.org/core/journals/psychological-medicine/article/white-matter-cognition-and-psychoticlike-experiences-in-uk-biobank/5A6A3F2BE82DF66341271996187D2200

      ) which will be referred to as “distress”. Frequencies for all items are noted in Table 1.
      Table 1Demographic characteristics for the reported-PLEs samples.
      MHQ Sample
      Age mean (SD)56.09 (7.69)
      Sex, F (%)67,432 (56%)
      MHQ – heard unreal voice (Yes, %)1,918 (1.6%)
      MHQ – seen unreal vision (Yes, %)3,696 (3%)
      MHQ – believed in unreal conspiracy (Yes, %)724 (0.60%)
      MHQ – believed in unreal communications (Yes, %)652 (0.54%)
      Distressing PLEs (Yes, %)2,046 (1.76%)

      Neuroimaging phenotypes

      A brain MRI scan was conducted for a subset of participants (

      Alfaro-Almagro F, Jenkinson M, Bangerter NK, Andersson JLR, Griffanti L, Douaud G, et al. Image processing and Quality Control for the first 10,000 brain imaging datasets from UK Biobank. Neuroimage [Internet]. 2018 Feb 1 [cited 2022 Apr 25];166:400–24. Available from: https://pubmed.ncbi.nlm.nih.gov/29079522/

      ,

      Smith SM, Alfaro-Almagro F, Miller KL. UK Biobank Brain Imaging Documentation UK Biobank Brain Imaging Documentation Contributors to UK Biobank Brain Imaging. [cited 2022 Apr 25]; Available from: http://www.ukbiobank.ac.uk

      ), and imaging-derived phenotypes (IDPs) of T1 weighted and diffusion (DTI) magnetic resonance imaging (MRI) images were used in this study. MRI acquisition, pre-processing, and QC protocols can be found in Supplementary Materials. Individuals with global values >3 SDs from the sample mean, as ascertained by conducting PCA on each modality’s imaging variables to derive a global value, were excluded (range N excluded based on imaging modality=105-232 participants, distribution plots created using “ggplot” in R are shown in Supplementary Figure 1) (
      • Shen X
      • Adams MJ
      • Ritakari TE
      • Cox SR
      • McIntosh AM
      • Whalley HC
      White Matter Microstructure and Its Relation to Longitudinal Measures of Depressive Symptoms in Mid- and Late Life.
      Biol Psychiatry. 2019 Nov 15; 86: 759-768
      ,
      • Shen X
      • Reus LM
      • Cox SR
      • Adams MJ
      • Liewald DC
      • Bastin ME
      • et al.
      Subcortical volume and white matter integrity abnormalities in major depressive disorder: findings from UK Biobank imaging data.
      Sci Reports. 2017; 71 ([Internet]. 2017 Jul 17 [cited 2022 May 19];7(1):1–10. Available from:)
      https://www.nature.com/articles/s41598-017-05507-6
      ). A list of neuroimaging phenotypes investigated is included in Supplementary Table 2.
      15 white matter tracts (3 unilateral) from two DTI scalars, FA and MD, were utilised (https://biobank.ndph.ox.ac.uk/showcase/label.cgi?id=135). Cortical regions of interest (ROIs) were identified using Desikan-Killiany-Tourville parcellation in FreeSurfer, resulting in 31 cortical structures per hemisphere for cortical thickness (CT), surface area (SA), and volume (CV) (
      • Alexander B
      • Loh WY
      • Matthews LG
      • Murray AL
      • Adamson C
      • Beare R
      • et al.
      Desikan-Killiany-Tourville Atlas compatible version of m-CRIB neonatal parcellated whole brain atlas: The m-Crib 2.0.
      Front Neurosci. 2019; 13: 34
      ,

      Elliott LT, Sharp K, Alfaro-Almagro F, Shi S, Miller KL, Douaud G, et al. Genome-wide association studies of brain imaging phenotypes in UK Biobank. Nat 2018 5627726 [Internet]. 2018 Oct 10 [cited 2022 May 19];562(7726):210–6. Available from: https://www.nature.com/articles/s41586-018-0571-7

      ) (https://biobank.ndph.ox.ac.uk/showcase/label.cgi?id=196). 8 subcortical grey matter ROIs per hemisphere were also identified (

      Fischl B, Salat DH, Busa E, Albert M, Dieterich M, Haselgrove C, et al. Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron [Internet]. 2002 Jan 31 [cited 2022 Apr 25];33(3):341–55. Available from: https://pubmed.ncbi.nlm.nih.gov/11832223/

      ).
      For white matter microstructure, we derived global and regional (association and projection fibres, thalamic radiations) measures of FA and MD by conducting PCA on sets of tracts and extracting scores of the first unrotated principal component. For cortical ROIs, we derived global and lobar measures by summing up all (global) or lobar-specific ROIs, as in previous studies (

      O’Connell KS, Sønderby IE, Frei O, Van Der Meer D, Athanasiu L, Smeland OB, et al. Association between complement component 4A expression, cognitive performance and brain imaging measures in UK Biobank. Psychol Med [Internet]. 2021 [cited 2022 Apr 25];1–11. Available from: https://www.cambridge.org/core/journals/psychological-medicine/article/association-between-complement-component-4a-expression-cognitive-performance-and-brain-imaging-measures-in-uk-biobank/D26127B6301C8C58148F3F4233403022

      ,

      Green C, Stolicyn A, Harris MA, Shen X, Romaniuk L, Barbu MC, et al. Hair glucocorticoids are associated with childhood adversity, depressive symptoms and reduced global and lobar grey matter in Generation Scotland. Transl Psychiatry 2021 111 [Internet]. 2021 Oct 12 [cited 2022 Apr 25];11(1):1–9. Available from: https://www.nature.com/articles/s41398-021-01644-9

      ). Sample size and descriptive statistics for neuroimaging phenotypes are in Table 2.
      Table 2Demographic characteristics for the neuroimaging samples. N is different for each variable as outlier exclusion (3 SDs from mean) was applied individually to each phenotype.
      Neuroimaging Sample
      Age mean (SD)63.71 (7.48)
      Sex, F (%)
      Cortical regions15,666 (53%)
      Subcortical volumes15,557 (53%)
      White matter microstructure14,745 (53%)
      Scan site
      Cheadle24,910
      Reading5,064
      Newcastle9,968

      Statistical analysis

      All analyses were conducted using R (version 4.1.0) in a Linux environment. We used linear mixed-effects models (function “lme” in package “nlme”) for bilateral neuroimaging phenotypes, with age, age2, sex, 15 genetic principal components (PC), scan site, three MRI head position coordinates (lateral brain position X, https://biobank.ctsu.ox.ac.uk/crystal/field.cgi?id=25756, transverse brain position Y, https://biobank.ctsu.ox.ac.uk/crystal/field.cgi?id=25757, longitudinal brain position Z https://biobank.ctsu.ox.ac.uk/crystal/field.cgi?id=25757), and genotype array set as covariates. Hemisphere was included as a within-subject covariate in all mixed-effects models. Intracranial volume (ICV) was included as a covariate for grey matter phenotypes. We used general linear models (function “lm”) for unilateral, regional, and global neuroimaging phenotypes, using the same covariates as above without hemisphere included. Finally, we used logistic regression for dichotomised PLEs, with age, sex, 15 genetic PCs, and genotype array as covariates. All models included gene-set PRS and each gene-set’s respective whole-genome (excluding SNPs in each gene-set) PRS as predictor variables. False Discovery Rate (FDR) was used to correct for multiple testing and was applied separately for each neuroimaging and reported-PLE phenotype and across all p-value thresholds (
      • Benjamini Y
      • Hochberg Y
      Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing.
      Source J R Stat Soc Ser B. 1995; 57: 289-300
      ) (See Supplementary Table 3 for detailed protocol). Effect sizes from linear models were standardized throughout.
      In order to establish that the effect of gene-set PRS on neuroimaging and reported-PLE phenotypes was not due to chance (as the SNP set sizes in all gene-sets were much smaller compared to the whole-genome), permutation analysis was carried out using a method developed by Cabrera et al. (

      Cabrera CP, Navarro P, Huffman JE, Wright AF, Hayward C, Campbell H, et al. Uncovering networks from genome-wide association studies via circular genomic permutation. G3 (Bethesda) [Internet]. 2012 Sep [cited 2022 Apr 26];2(9):1067–75. Available from: https://pubmed.ncbi.nlm.nih.gov/22973544/

      ) for all significant associations (See Supplementary Materials).

      Results

      All results presented below concern PRS at p-value threshold <0.1. Descriptive statistics for reported-PLE and neuroimaging phenotypes are noted in Tables 1 and 2, respectively. Analyses involving other thresholds are included in Supplementary Excel File 2 (Tables 5-11).

      Demographic characteristics

      Associations with psychotic-like experiences

      Gene-set-specific significant findings and corresponding WG (excluding SNPs in each gene-set) results are noted in Figure 1. Axon gene-set PRS was significantly associated with communications MHQ item (gene-set PRS β=0.0916, pFDR=0.04; whole-genome PRS (excluding SNPs in each gene-set) β=0.2218, pFDR=1.34x10-7), while the PSD gene-set PRS was significantly associated with distress associated with reported-PLEs (gene-set PRS β=0.0588, pFDR=0.02; whole-genome PRS (excluding SNPs in each gene-set) β=0.1943, pFDR=7.28x10-16). Whole-genome PRS (excluding SNPs in each gene-set) were associated with all four MHQ items, as well as distress, and effect sizes were of greater magnitude than associations with gene-set PRS (see Supplementary Excel File 2, Table 5). Effect sizes for gene-set-specific PRS ranged from 1x10-4 to 0.092 and for whole-genome PRS (excluding SNPs in each gene-set) from 0.087 to 0.300. Results concerning MHQ items at other p-values thresholds are in Supplementary Excel File 2 (Table 5).
      Figure thumbnail gr1
      Figure 1Significant associations between the axon and postsynaptic density gene-set PRSs and reported-PLEs phenotypes. X-axis indicates standardized effect sizes; *=significant associations after correction for multiple comparisons; WG = whole-genome (excluding SNPs in each gene-set).

      Associations with neuroimaging phenotypes

      Gene-set PRS associations

      Axon gene-set PRS were associated with parahippocampal gyrus volume (gene-set PRS β=0.0156, pFDR=0.03; whole-genome (excluding SNPs in each gene-set) PRS β=-0.003, pFDR=0.833), FA thalamic radiations tract (gene-set PRS (excluding SNPs in each gene-set) β=-0.014, pFDR=0.036; whole-genome PRS (excluding SNPs in each gene-set) β=-0.0143, pFDR=0.036), and FA posterior thalamic radiations (gene-set PRS β=-0.016, pFDR=0.048; whole-genome PRS (excluding SNPs in each gene-set) β=-0.011, pFDR=0.126).
      PSD gene-set PRS were associated with global surface area (gene-set PRS β=-0.012, pFDR=0.034; whole-genome PRS (excluding SNPs in each gene-set) β=-0.003, pFDR=0.517) and cingulate lobe surface area (gene-set PRS β=-0.014, pFDR=0.04; whole-genome PRS (excluding SNPs in each gene-set) β=1x10-4, pFDR=0.977).
      Finally, the histone H3-K4 methylation gene-set PRS were associated with entorhinal surface area (gene-set PRS β=-0.016, pFDR=0.035; whole-genome PRS (excluding SNPs in each gene-set) β=0.01, pFDR=0.164).
      Neuroimaging phenotypes associated with gene-set PRSs are indicated in Figures 2 and 3.
      Figure thumbnail gr2
      Figure 2Significant associations between the axon, postsynaptic density, and histone H3-K4 methylation gene-set PRSs and neuroimaging phenotypes. X-axis indicates standardized effect sizes; *=significant associations after correction for multiple comparisons; WG = whole-genome (excluding SNPs in each gene-set).
      Figure thumbnail gr3
      Figure 3Cortical and white matter microstructure phenotypes that were associated with the axon, PSD, and histone gene-set PRSs. The thalamic radiations tract category, comprised of the anterior, superior, and posterior thalamic radiations, was associated with the axon gene-set PRS; the cingulate lobe, comprised of caudal anterior, rostral anterior, posterior and isthmus cingulate, was associated with PSD gene-set PRSs. Global surface area (i.e. the entire brain’s surface area) was also associated with PSD but is not indicated in the graph above.

      Whole-genome PRS associations

      Whole-genome PRS, irrespective of which SNPs were removed prior to calculating PRSs, were associated with a number of neuroimaging phenotypes, indicated below, in Supplementary Excel File 2 (Tables 6-11) and in Supplementary Table 4. All standardized beta values are absolute, and ranges are not reported if they are rounded to the same number. Results presented below are consistent across whole-genome PRS that had different gene-set SNPs excluded.
      FA: global FA (β range=0.016-0.019, pFDR range=0.004-0.03), association fibres (β range=0.018 -0.021, pFDR range=0.003-0.019), thalamic radiations (β range=0.014-0.018, pFDR range=0.006-0.03), cingulate gyrus (β range=0.014-0.015, pFDR range=0.014-0.034), anterior (β range=0.016-0.017, pFDR range=0.019-0.038) and posterior thalamic radiations (β range=0.015-0.016, pFDR range=0.025-0.03), inferior longitudinal (β range=0.015-0.017, pFDR range=0.015-0.043) and inferior fronto-occipital fasciculi (β range=0.015-0.016, pFDR range=0.024-0.035).
      MD: global MD (β range=0.015-0.017, pFDR range=0.011-0.04), association fibres (β range=0.013-0.014, pFDR range=0.034-0.042), thalamic radiations (β range=0.014-0.016, pFDR range=0.011-0.04), projection fibres (β range=0.014, pFDR range=0.04), corticospinal tract (β range=0.02-0.022, pFDR range=0.005-0.023), anterior (β=0.015, pFDR=0.037), superior (β=0.014, pFDR=0.043), and posterior thalamic radiations (β=0.013, pFDR=0.04), inferior longitudinal fasciculus (β=0.015, pFDR=0.04), and cingulate gyrus (β=0.015, pFDR=0.04).
      Thalamus (β range=0.019-0.023, pFDR range=3.6x10-5-0.0009) and accumbens (β range=0.013-0.015, pFDR range=0.005-0.032) were also significantly associated with whole-genome PRS (excluding SNPs in each gene-set) after FDR correction. Finally, significant cortical regions included volume of the medial orbitofrontal cortex (β range=0.016-0.019, pFDR range=0.003-0.027) and superior temporal gyrus (β range=0.018-0.02, pFDR range=8.3x10-4-0.009), and surface area of superior temporal gyrus (β range=0.015-0.016, pFDR range=0.022-0.04).

      Permutation analysis

      For gene-set PRSs that were significantly associated with reported-PLEs, or neuroimaging measures, we performed circular genomic permutation analysis. We found that the significant associations with the gene-set PRSs in these phenotypes were not due to chance, based on a permutation p-value computed by comparing t-values from the real associations with t-values from the permuted associations (Table 3).
      Table 3Permutation results for gene-set PRS where significant associations were identified (PRS p-value threshold 0.1).
      Gene-set PRSPhenotypeGene-set PRS βGene-set PRS t-valueGene-set PRS calculated p-value
      AxonCommunications0.09162.3060.026
      Parahippocampal gyrus volume0.01563.1690.003
      Thalamic radiations (FA)-0.014-2.440.01
      Posterior thalamic radiations (FA)-0.016-3.0550.004
      Postsynaptic densityDistress0.05882.6180.01
      Cingulate lobe surface area-0.014-2.8120.012
      Global surface area-0.012-2.4980.024
      Histone H3-K4 methylationEntorhinal cortex-0.016-3.4540.002

      Discussion

      In this study, we investigated associations between biologically relevant pathway PRSs and N=119,947reported-PLE and N=∼29,000neuroimaging phenotypes, after adjustment for respective whole-genome PRSs (excluding SNPs in each gene-set). We calculated PRSs using the largest, most recent schizophrenia GWAS to date (

      Trubetskoy V, Pardiñas AF, Qi T, Panagiotaropoulou G, Awasthi S, Bigdeli TB, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nat 2022 6047906 [Internet]. 2022 Apr 8 [cited 2022 Apr 22];604(7906):502–8. Available from: https://www.nature.com/articles/s41586-022-04434-5

      ). We found significant associations for axon, PSD, and histone H3-K4 methylation gene-sets with a number of cortical regions and white matter tracts, and with reported-PLEs, specifically psychotic-like communications and distress associated with a reported-PLE. Associations with reported-PLEs were stronger for whole-genome PRS (excluding SNPs in each gene-set), while associations with neuroimaging variables were stronger for gene-set PRS.
      The three gene-sets with identified significant associations here have previously been involved in schizophrenia. The PGC (

      O’dushlaine C, Rossin L, Lee PH, Duncan L, Parikshak NN, Newhouse S, et al. Psychiatric genome-wide association study analyses implicate neuronal, immune and histone pathways. Nat Neurosci [Internet]. 2015 Feb 17 [cited 2022 Apr 22];18(2):199. Available from: /pmc/articles/PMC4378867/

      ) identified pathways’ implication in schizophrenia through robust computational analyses, by aggregating gene-sets from a large number of databases. The most significant genetic contribution to schizophrenia was found in genes encoding proteins located in excitatory synapses, in particular the proteome of the PSD, followed by histone methylation-related processes. The PSD of excitatory synapses comprises protein complexes that assemble glutamate-sensitive neurotransmitter receptors to intracellular proteins (

      Bayés Á, Van De Lagemaat LN, Collins MO, Croning MDR, Whittle IR, Choudhary JS, et al. Characterization of the proteome, diseases and evolution of the human postsynaptic density. Nat Neurosci [Internet]. 2011 Jan [cited 2022 Jul 18];14(1):19–21. Available from: https://pubmed.ncbi.nlm.nih.gov/21170055/

      ,

      Sorokina O, Mclean C, Croning MDR, Heil KF, Wysocka E, He X, et al. A unified resource and configurable model of the synapse proteome and its role in disease. Sci Reports 2021 111 [Internet]. 2021 May 11 [cited 2022 Jul 18];11(1):1–9. Available from: https://www.nature.com/articles/s41598-021-88945-7

      ). Postsynaptic networks have long been implicated in schizophrenia (

      Fromer M, Pocklington AJ, Kavanagh DH, Williams HJ, Dwyer S, Gormley P, et al. De novo mutations in schizophrenia implicate synaptic networks. Nature [Internet]. 2014 [cited 2022 Jul 18];506(7487):179–84. Available from: https://pubmed.ncbi.nlm.nih.gov/24463507/

      ,

      Purcell SM, Moran JL, Fromer M, Ruderfer D, Solovieff N, Roussos P, et al. A polygenic burden of rare disruptive mutations in schizophrenia. Nature [Internet]. 2014 [cited 2022 Jul 18];506(7487):185–90. Available from: https://pubmed.ncbi.nlm.nih.gov/24463508/

      ,
      • Skene NG
      • Roy M
      • Grant SG
      A genomic lifespan program that reorganises the young adult brain is targeted in schizophrenia.
      Elife. 2017 Sep 12; : 6
      ) as they play a role in cognition and synaptic plasticity, features known to be disrupted in schizophrenia (
      • Föcking M
      • Lopez LM
      • English JA
      • Dicker P
      • Wolff A
      • Brindley E
      • et al.
      Proteomic and genomic evidence implicates the postsynaptic density in schizophrenia.
      Mol Psychiatry. 2015 204; ([Internet]. 2014 Jul 22 [cited 2022 May 5];20(4):424–32. Available from:)
      https://www.nature.com/articles/mp201463
      ,

      Nithianantharajah J, Komiyama NH, McKechanie A, Johnstone M, Blackwood DH, Clair DS, et al. Synaptic scaffold evolution generated components of vertebrate cognitive complexity. Nat Neurosci 2012 161 [Internet]. 2012 Dec 2 [cited 2022 Jul 18];16(1):16–24. Available from: https://www.nature.com/articles/nn.3276

      ). In addition, both post-synaptic and pre-synaptic networks were uncovered through gene prioritisation in the latest schizophrenia GWAS, highlighting the consistency of associations with this pathway and further rationale to investigate the pathway here (

      Trubetskoy V, Pardiñas AF, Qi T, Panagiotaropoulou G, Awasthi S, Bigdeli TB, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nat 2022 6047906 [Internet]. 2022 Apr 8 [cited 2022 Apr 22];604(7906):502–8. Available from: https://www.nature.com/articles/s41586-022-04434-5

      ).
      We identified associations between the PSD gene-set and decreased global and cingulate lobe surface area, as well as higher PLE-associated distress. The cingulate lobe is part of the limbic system and comprises four cortical regions, rostral, caudal, posterior, and isthmus cingulate. The region is involved in behaviour, emotion regulation and cognitive processes including memory, attention, and motivation, all disrupted in schizophrenia (
      • Bersani FS
      • Minichino A
      • Fojanesi M
      • Gallo M
      • Maglio G
      • Valeriani G
      • Biondi M FP
      Cingulate Cortex in Schizophrenia: its relation with negative symptoms and psychotic onset. A review study.
      Eur Rev Med Pharmacol Sci. 2014; 18: 3354-3367
      ). Volumes in this region were reduced in previous studies investigating their association with schizophrenia, and a systematic review concluded that hypoactivity of the cingulate cortex underlies the manifestation of negative symptoms in many patients, although the studies analysed provided inconsistent results (
      • Bersani FS
      • Minichino A
      • Fojanesi M
      • Gallo M
      • Maglio G
      • Valeriani G
      • Biondi M FP
      Cingulate Cortex in Schizophrenia: its relation with negative symptoms and psychotic onset. A review study.
      Eur Rev Med Pharmacol Sci. 2014; 18: 3354-3367
      ,
      • Wang L
      • Hosakere M
      • Trein JCL
      • Miller A
      • Ratnanather JT
      • Barch DM
      • et al.
      Abnormalities of cingulate gyrus neuroanatomy in schizophrenia.
      Schizophr Res. 2007 Jul 1; 93: 66-78
      ). Interestingly, whole-genome PRS did not show associations with cingulate or global surface area in previous studies (
      • Neilson E
      • Shen X
      • Cox SR
      • Clarke TK
      • Wigmore EM
      • Gibson J
      • et al.
      Impact of Polygenic Risk for Schizophrenia on Cortical Structure in UK Biobank.
      Biol Psychiatry. 2019 Oct 1; 86: 536-544
      ,

      Zhu X, Ward J, Cullen B, Lyall DM, Strawbridge RJ, Smith DJ, et al. Polygenic Risk for Schizophrenia, Brain Structure, and Environmental Risk in UK Biobank. Schizophr Bull Open [Internet]. 2021 Jan 1 [cited 2022 May 12];2(1). Available from: https://academic.oup.com/schizbullopen/article/2/1/sgab042/6375001

      ). Associations identified here indicate that narrowing the genome to a biologically relevant gene-set may reveal associations not observed genome-wide.
      The axon gene-set, a cellular component that conducts electrical signals to presynaptic boutons that store and release neurotransmitters, has also been found to play an important role in schizophrenia. Specifically, disruption in axon guidance and axonal growth have been associated with increased schizophrenia risk in both human and mouse models (
      • Wang Z
      • Li P
      • Wu T
      • Zhu S
      • Deng L
      • Cui G
      Axon guidance pathway genes are associated with schizophrenia risk.
      Exp Ther Med. 2018 Dec 1; ([Internet]) ([cited 2022 May 13];16(6):4519–26. Available from:)
      http://www.spandidos-publications.com/10.3892/etm.2018.6781/abstract
      ,
      • Mukai J
      • Tamura M
      • Fénelon K
      • Rosen AM
      • Spellman TJ
      • Kang R
      • et al.
      Molecular Substrates of Altered Axonal Growth and Brain Connectivity in a Mouse Model of Schizophrenia.
      Neuron. 2015 May 6; 86: 680-695
      ). A recent study showed that individuals at high genetic risk for schizophrenia had hemispherical asymmetry in whole-brain structural networks, indicating that genetic susceptibility to schizophrenia modulated white matter network abnormalities. Additionally, gene-set enrichment analysis found that genes participating in the PRS threshold used were involved in multiple relevant pathways, including axonal growth and axon guidance (

      Zhu Y, Wang S, Gong X, Edmiston EK, Zhong S, Li C, et al. Associations between hemispheric asymmetry and schizophrenia-related risk genes in people with schizophrenia and people at a genetic high risk of schizophrenia. Br J Psychiatry [Internet]. 2021 Jul 1 [cited 2022 May 16];219(1):392–400. Available from: https://www.cambridge.org/core/journals/the-british-journal-of-psychiatry/article/associations-between-hemispheric-asymmetry-and-schizophreniarelated-risk-genes-in-people-with-schizophrenia-and-people-at-a-genetic-high-risk-of-schizophrenia/8506769E0926B8AF52DB15E6D3FBA614

      ). The axon gene-set is therefore a highly relevant route of investigation in the context of schizophrenia. Here, the axon gene-set was associated with thalamic radiations, white matter microstructural tracts that link the thalamus to the rest of the cerebral cortex (
      • Lebel C
      • Deoni S
      The development of brain white matter microstructure.
      Neuroimage. 2018 Nov 15; 182: 207-218
      ), and volume of the parahippocampal gyrus, a cortical region that plays a role in memory processes such as encoding and retrieval (

      Van Hoesen GW, Augustinack JC, Dierking J, Redman SJ, Thangavel R. The Parahippocampal Gyrus in Alzheimer’s Disease: Clinical and Preclinical Neuroanatomical Correlates. Ann N Y Acad Sci [Internet]. 2000 Jun 1 [cited 2022 May 13];911(1):254–74. Available from: https://onlinelibrary.wiley.com/doi/full/10.1111/j.1749-6632.2000.tb06731.x

      ). Both neuroimaging phenotypes have been implicated in schizophrenia (
      • McIntosh AM
      • Maniega SM
      • Lymer GKS
      • McKirdy J
      • Hall J
      • Sussmann JED
      • et al.
      White Matter Tractography in Bipolar Disorder and Schizophrenia.
      Biol Psychiatry. 2008 Dec 15; 64: 1088-1092
      ) and have recently been associated with state anhedonia and PRS for anhedonia, a core negative symptom of schizophrenia (
      • Zhu X
      • Ward J
      • Cullen B
      • Lyall DM
      • Strawbridge RJ
      • Lyall LM
      • et al.
      Phenotypic and genetic associations between anhedonia and brain structure in UK Biobank.
      Transl Psychiatry. 2021; 111 ([Internet]. 2021 Jul 16 [cited 2022 May 13];11(1):1–8. Available from:)
      https://www.nature.com/articles/s41398-021-01522-4
      ). An opposite directionality of effect for whole-genome PRS and gene-set PRS on parahippocampal gyrus volume could denote differential pathway-specific action on this region. The findings here indicate that genes conferring risk for schizophrenia aggregated in the axon gene-set are strengthening evidence for brain structural regions already implicated in schizophrenia psychopathology.
      Finally, here the histone H3-K4 methylation gene-set was associated with entorhinal surface area. Histone H3-K4 methylation involves the modification of histone H3 by the addition of one or more methyl groups to lysine at position 4 of the histone. Histones, and specifically H3, are used to package DNA, and modifications lead to changes in gene expression (
      • Greer EL
      • Shi Y
      Histone methylation: a dynamic mark in health, disease and inheritance.
      Nat Rev Genet. 2012 135; ([Internet]. 2012 Apr 3 [cited 2022 May 13];13(5):343–57. Available from:)
      https://www.nature.com/articles/nrg3173
      ). Epigenetic processes, including histone and DNA methylation, have been associated with schizophrenia through candidate gene regulation (HDAC1, GAD67) and epigenome-wide studies, lending support to the investigation of epigenetic modifications in schizophrenia (
      • Gavin DP
      • Sharma RP
      Histone modifications, DNA methylation, and Schizophrenia.
      Neurosci Biobehav Rev. 2010 May 1; 34: 882-888
      ,

      Montano C, Taub MA, Jaffe A, Briem E, Feinberg JI, Trygvadottir R, et al. Association of DNA Methylation Differences With Schizophrenia in an Epigenome-Wide Association Study. JAMA psychiatry [Internet]. 2016 May 1 [cited 2022 May 16];73(5):506–14. Available from: https://pubmed.ncbi.nlm.nih.gov/27074206/

      ,

      Huang HS, Matevossian A, Whittle C, Se YK, Schumacher A, Baker SP, et al. Prefrontal Dysfunction in Schizophrenia Involves Mixed-Lineage Leukemia 1-Regulated Histone Methylation at GABAergic Gene Promoters. J Neurosci [Internet]. 2007 Oct 17 [cited 2022 May 13];27(42):11254. Available from: /pmc/articles/PMC6673022/

      ). A key feature of schizophrenia is that it has its age of onset in young adults and the transcriptional regulation of schizophrenia risk genes encoding synaptic proteins occurs at the age of onset, potentially through mechanisms involving H3-K4 methylation (
      • Skene NG
      • Roy M
      • Grant SG
      A genomic lifespan program that reorganises the young adult brain is targeted in schizophrenia.
      Elife. 2017 Sep 12; : 6
      ).
      The entorhinal cortex plays a role in the integration of multisensory information between cortical and subcortical structures. It is strongly connected to the hippocampus and is involved in memory processes such as formation and consolidation (

      Schultz CC, Koch K, Wagner G, Roebel M, Schachtzabel C, Nenadic I, et al. Psychopathological correlates of the entorhinal cortical shape in schizophrenia. Eur Arch Psychiatry Clin Neurosci [Internet]. 2010 Jun 7 [cited 2022 May 13];260(4):351–8. Available from: https://link.springer.com/article/10.1007/s00406-009-0083-4

      ). In studies investigating a neurodevelopmental animal model of schizophrenia, lesions in the entorhinal cortex, created early in the developmental period, were associated with enhanced mesolimbic dopamine release at a later timepoint, expressed through increased locomotor activity. This indicates that structural disruptions in this area are relevant for schizophrenia-like presentations (
      • Sumiyoshi T
      • Tsunoda M
      • Uehara T
      • Tanaka K
      • Itoh H
      • Sumiyoshi C
      • et al.
      Enhanced locomotor activity in rats with excitotoxic lesions of the entorhinal cortex, a neurodevelopmental animal model of schizophrenia: Behavioral and in vivo microdialysis studies.
      Neurosci Lett. 2004 Jul 1; 364: 124-129
      ). Interestingly, in an animal model, H3-K4 methylation was found to be upregulated in the hippocampus 1 hour after inducing contextual fear conditioning, a memory-related task which aims to rapidly create context-related fear memories (

      Gupta S, Kim SY, Artis S, Molfese DL, Schumacher A, Sweatt JD, et al. Histone Methylation Regulates Memory Formation. J Neurosci [Internet]. 2010 Mar 10 [cited 2022 May 16];30(10):3589. Available from: /pmc/articles/PMC2859898/

      ). Although not directly related to the entorhinal cortex, this finding indicates that H3-K4 methylation may be a useful candidate in the investigation of some schizophrenia-related symptoms in a subcortical structure closely linked to the entorhinal region. Lastly, in a recent human study investigating cortical thickness, surface area, and folding index of the entorhinal cortex, Schultz et al. (2009) uncovered a link between left surface area and folding index and increased psychotic symptom severity, further supporting the role of the region in schizophrenia (

      Schultz CC, Koch K, Wagner G, Roebel M, Schachtzabel C, Nenadic I, et al. Psychopathological correlates of the entorhinal cortical shape in schizophrenia. Eur Arch Psychiatry Clin Neurosci [Internet]. 2010 Jun 7 [cited 2022 May 13];260(4):351–8. Available from: https://link.springer.com/article/10.1007/s00406-009-0083-4

      ). Notably, we observed an opposite directionality of effect for H3-HK methylation whole-genome PRS and gene-set PRS on entorhinal surface area, which we hypothesise could be due to differential action of the pathway SNPs versus the whole-genome SNPs.
      Interestingly, the dendritic spine and PSM gene-sets were not associated with any phenotypes investigated. Both cellular components have been linked to schizophrenia previously and are closely related to the other biologically relevant pathways investigated here (i.e. PSD, axon gene-sets) (
      • Glausier JR
      • Lewis DA
      Dendritic spine pathology in schizophrenia.
      Neuroscience. 2013 Oct 22; 251: 90-107
      ,

      De Bartolomeis A, Latte G, Tomasetti C, Iasevoli F. Glutamatergic postsynaptic density protein dysfunctions in synaptic plasticity and dendritic spines morphology: Relevance to schizophrenia and other behavioral disorders pathophysiology, and implications for novel therapeutic approaches. Mol Neurobiol [Internet]. 2014 Sep 3 [cited 2022 May 16];49(1):484–511. Available from: https://link.springer.com/article/10.1007/s12035-013-8534-3

      ). This may indicate that, while these gene-sets are relevant in schizophrenia, we may need additional information (e.g. interaction with other relevant biological pathways or investigation of copy number variants in relation to these cellular components (
      • Kirov G
      • Pocklington AJ
      • Holmans P
      • Ivanov D
      • Ikeda M
      • Ruderfer D
      • et al.
      De novo CNV analysis implicates specific abnormalities of postsynaptic signalling complexes in the pathogenesis of schizophrenia.
      Mol Psychiatry. 2012; 17: 142-153
      )) to identify associations with these reported-PLE and structural brain features. Despite correlation analyses of PRSs across participants revealing a strong positive correlation between PSD and PSM pathway PRSs (r=0.63, p=≤0.01), this did not translate to similarly strong associations within the main analyses.
      All PLEs, as well as distress, were significantly associated with whole-genome PRS (excluding SNPs in each gene-set), as expected. This indicates that the PRS generated from the schizophrenia GWAS are able to predict schizophrenia-related traits and are therefore a valuable tool. In addition, a number of neuroimaging phenotypes were associated with whole-genome PRS (irrespective of which gene-set-specific SNPs were removed prior to calculation), including white matter microstructure, cortical, and subcortical volume regions (See Supplementary Table 4 for list of regions). These areas have been associated with whole-genome PRS in previous studies, calculated both with the GWAS summary statistics utilised here, and with older GWAS (

      van der Merwe C, Passchier R, Mufford M, Ramesar R, Dalvie S, Stein DJ. Polygenic risk for schizophrenia and associated brain structural changes: A systematic review. Compr Psychiatry [Internet]. 2019 Jan 1 [cited 2022 Apr 26];88:77–82. Available from: https://pubmed.ncbi.nlm.nih.gov/30529765/

      ,
      • Stauffer EM
      • Bethlehem RAI
      • Warrier V
      • Murray GK
      • Romero-Garcia R
      • Seidlitz J
      • et al.
      Grey and white matter microstructure is associated with polygenic risk for schizophrenia.
      Mol Psychiatry. 2021 2612; ([Internet]. 2021 Aug 30 [cited 2022 May 16];26(12):7709–18. Available from:)
      https://www.nature.com/articles/s41380-021-01260-5
      ). These findings confirm previous results and provide additional evidence of an association between increased genetic risk for schizophrenia and disruptions in grey and white matter. However, that whole-genome PRS (excluding SNPs in each gene-set) more accurately predict schizophrenia-relevant phenotypes over pathway PRS for these measures at this stage caveats our aim to address the disorder’s heterogeneity by selecting biologically relevant pathways. Further studies could incorporate additional factors such as expression quantitative trait loci (eQTL) to PRS to determine if pathway-specific gene expression on brain tissue can better differentiate variation in specific phenotypes over non-pathway expression trends.
      There are a number of strengths in the current study. We utilised the largest, most recent GWAS of schizophrenia within a large sample comprising reported-PLEs and neuroimaging datasets. Our findings were further tested utilising permutation analysis, indicating that results were not due to chance. Finally, we were able to identify associations that could in future be utilised to identify stratified patient populations, potentially leading to earlier diagnosis and applied interventions.
      Limitations include the investigation of these associations in a general population sample, where participants are generally healthier and wealthier than the general population, as shown in Fry et al. (2017) (
      • Fry A
      • Littlejohns TJ
      • Sudlow C
      • Doherty N
      • Adamska L
      • Sprosen T
      • et al.
      Comparison of Sociodemographic and Health-Related Characteristics of UK Biobank Participants with Those of the General Population.
      Am J Epidemiol. 2017; 186: 1026-1034
      ). Due to the low number of ICD-10 diagnosed schizophrenia participants with reported-PLE (N=123) and imaging (N=49) measures, we were unable to attempt a replication and highly encourage this in larger cohorts. This may partially explain low effect sizes observed throughout, and it is expected that effect sizes will increase as clinical samples with available genetic and neuroimaging data increase. We are unable to generalise these findings to populations of non-European ancestry, as both GWAS and target samples were selected based on individuals of European descent. However, efforts are continually being made to collect data from other ancestries, and the associations identified here could be explored in these cohorts.
      A further limitation includes the use of the MHQ, which reports lifetime occurrence of PLEs, and not at a specific timepoint. Therefore, our results may differ in other age populations. Further, responses are self-reported, and it is unclear whether reported-PLEs arose from normal experiences or were linked to a psychiatric disorder. The strong associations with schizophrenia for whole-genome PRS (excluding SNPs in each gene-set) indicate that the items pick up on some psychosis-related features. However, although PLEs are a hallmark of schizophrenia, they may be shared by a number of mental health disorders and non-clinical phenotypes (

      Kelleher I, Cannon M. Psychotic-like experiences in the general population: characterizing a high-risk group for psychosis. Psychol Med. 2011 Jan;41(1):1-6. doi: 10.1017/S0033291710001005. Epub 2010 May 19. PMID: 20624328.

      ,

      Legge SE, Jones HJ, Kendall KM, Pardiñas AF, Menzies G, Bracher-Smith M, et al. Association of Genetic Liability to Psychotic Experiences With Neuropsychotic Disorders and Traits. JAMA Psychiatry. 2019 Dec 1;76(12):1256-1265. doi: 10.1001/jamapsychiatry.2019.2508. PMID: 31553412; PMCID: PMC6764002.

      ). As such, schizophrenia-implicated pathways may not fully capture the genetic basis for PLEs. A multi-disorder study focused on the severity of PLEs encompassing pathways associated with other mental health disorders could disentangle potentially shared genetic pathways associated with PLEs.
      The differential directionality of effect for a number of pathway PRS and whole-genome (excluding SNPs in each gene-set) PRS may signal the presence of other disorder-contributing factors which may have differential pathway and whole-genome effects, not controlled for here. Further studies could include a wider range of covariates, encompassing both environmental and genomic factors, such as smoking status and methylation.
      Lastly, a number of methodological limitations should be noted. Reference based approaches, such as the pathway of interest selection employed here, may limit the strength of their findings (
      • Pergola G
      • Penzel N
      • Sportelli L
      • Bertolino A
      Lessons learned from parsing genetic risk for schizophrenia into biological pathways.
      Biological Psychiatry. 2022 Oct 28;
      ) by lowering the threshold required for a pathway to be significantly associated with a phenotype. An omnibus study utilising all available pathways could be useful in contextualising the strength of these findings in relation to the larger pool of gene ontologies not previously associated with schizophrenia. Although Freesurfer is a standard methodological approach to analysing sMRI data across psychiatry, it may have limitations compared to manual tracing approaches, especially concerning segmentation variability (
      • Morey RA
      • Petty CM
      • Xu Y
      • Hayes JP
      • Wagner II, HR
      • Lewis DV
      • LaBar KS
      • Styner M
      • McCarthy G
      A comparison of automated segmentation and manual tracing for quantifying hippocampal and amygdala volumes.
      Neuroimage. 2009 Apr 15; 45: 855-866
      ). Further studies could utilise a combination of image processing software, such as Freesurfer and voxel-based morphometry (VBM) to validate their findings. Finally, window-based approaches for SNP-gene mapping are limited by their chosen arbitrary distances and may be less informative than methodologies using gene expression data. Ultimately, this study is intrinsically limited by its inclusion of only 5 biologically relevant pathways and a limited number and type of reported-PLE and neuroimaging measures, and the methods used to obtain these. Further studies could complement this investigation by diversifying these parameters, especially in the case of utilising functional imaging measures, as well as replicating our findings in other large cohorts.
      The pathways investigated here were previously found to be implicated in schizophrenia in both animal and human studies (

      Trubetskoy V, Pardiñas AF, Qi T, Panagiotaropoulou G, Awasthi S, Bigdeli TB, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nat 2022 6047906 [Internet]. 2022 Apr 8 [cited 2022 Apr 22];604(7906):502–8. Available from: https://www.nature.com/articles/s41586-022-04434-5

      ,

      Gupta S, Kim SY, Artis S, Molfese DL, Schumacher A, Sweatt JD, et al. Histone Methylation Regulates Memory Formation. J Neurosci [Internet]. 2010 Mar 10 [cited 2022 May 16];30(10):3589. Available from: /pmc/articles/PMC2859898/

      ,
      • Zhu X
      • Ward J
      • Cullen B
      • Lyall DM
      • Strawbridge RJ
      • Lyall LM
      • et al.
      Phenotypic and genetic associations between anhedonia and brain structure in UK Biobank.
      Transl Psychiatry. 2021; 111 ([Internet]. 2021 Jul 16 [cited 2022 May 13];11(1):1–8. Available from:)
      https://www.nature.com/articles/s41398-021-01522-4
      ). In this study, we identified structural neuroimaging and reported-PLE phenotypes associated with biologically informed PRSs, and these were shown to be more strongly associated with neuroimaging phenotypes than whole-genome PRS. However, since associations were stronger overall for whole-genome PRS, our findings also indicate that genetic risk aggregated to biologically relevant pathways is not yet more informative than genome-wide risk but it may be still of relevance to future studies attempting to address heterogeneity and stratify individuals by genetic risk.

      Acknowledgements

      The authors wish to thank all the participants who are part of the UK Biobank cohort, and all of the UK Biobank staff who make this work possible. We would also like to thank our funders, Wellcome Trust, the Royal College of Physicians of Edinburgh, the Medical Research Council, the Chief Scientist Office for Scotland, NHS Education for Scotland, the Guarantors of Brain, and the British Medical Association. The article has been uploaded to the preprint server medRxiv.

      Supplementary Material

      References

        • Owen MJ
        • Sawa A
        • Mortensen PB
        Schizophrenia. Lancet. 2016 Jul 2; 388: 86-97
        View in Article
        • Hilker R
        • Helenius D
        • Fagerlund B
        • Skytthe A
        • Christensen K
        • Werge TM
        • et al.
        Heritability of Schizophrenia and Schizophrenia Spectrum Based on the Nationwide Danish Twin Register.
        Biol Psychiatry. 2018 Mar 15; 83: 492-498
        View in Article

      3. Lam M, Chen CY, Li Z, Martin AR, Bryois J, Ma X, et al. Comparative genetic architectures of schizophrenia in East Asian and European populations. Nat Genet [Internet]. 2019 Dec 1 [cited 2022 Apr 22];51(12):1670–8. Available from: https://pubmed.ncbi.nlm.nih.gov/31740837/

        View in Article

      4. Trubetskoy V, Pardiñas AF, Qi T, Panagiotaropoulou G, Awasthi S, Bigdeli TB, et al. Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nat 2022 6047906 [Internet]. 2022 Apr 8 [cited 2022 Apr 22];604(7906):502–8. Available from: https://www.nature.com/articles/s41586-022-04434-5

        View in Article

      5. O’dushlaine C, Rossin L, Lee PH, Duncan L, Parikshak NN, Newhouse S, et al. Psychiatric genome-wide association study analyses implicate neuronal, immune and histone pathways. Nat Neurosci [Internet]. 2015 Feb 17 [cited 2022 Apr 22];18(2):199. Available from: /pmc/articles/PMC4378867/

        View in Article
        • Rampino A
        • Taurisano P
        • Fanelli G
        • Attrotto M
        • Torretta S
        • Antonucci LA
        • et al.
        A Polygenic Risk Score of glutamatergic SNPs associated with schizophrenia predicts attentional behavior and related brain activity in healthy humans.
        Eur Neuropsychopharmacol. 2017 Sep 1; 27: 928-939
        View in Article
        • Yao Y
        • Guo W
        • Zhang S
        • Yu H
        • Yan H
        • Zhang H
        • et al.
        Cell type-specific and cross-population polygenic risk score analyses of MIR137 gene pathway in schizophrenia.
        iScience. 2021 Jul 23; 24102785
        View in Article

      8. Terwisscha Van Scheltinga AF, Bakker SC, Van Haren NEM, Derks EM, Buizer-Voskamp JE, Boos HBM, et al. Genetic schizophrenia risk variants jointly modulate total brain and white matter volume. Biol Psychiatry [Internet]. 2013 Mar 15 [cited 2022 Apr 26];73(6):525. Available from: /pmc/articles/PMC3573254/

        View in Article

      9. Alnæs D, Kaufmann T, Van Der Meer D, Córdova-Palomera A, Rokicki J, Moberget T, et al. Brain Heterogeneity in Schizophrenia and Its Association With Polygenic Risk. JAMA Psychiatry [Internet]. 2019 Jul 1 [cited 2022 Apr 26];76(7):739–48. Available from: https://jamanetwork.com/journals/jamapsychiatry/fullarticle/2730004

        View in Article
        • Grama S
        • Willcocks I
        • Hubert JJ
        • Pardiñas AF
        • Legge SE
        • Bracher-Smith M
        • et al.
        Polygenic risk for schizophrenia and subcortical brain anatomy in the UK Biobank cohort.
        Transl Psychiatry. 2020 101; ([Internet]. 2020 Sep 9 [cited 2022 Apr 22];10(1):1–10. Available from:)
        https://www.nature.com/articles/s41398-020-00940-0
        View in Article
        • Barbu MC
        • Zeng Y
        • Shen X
        • Cox SR
        • Clarke TK
        • Gibson J
        • et al.
        Association of Whole-Genome and NETRIN1 Signaling Pathway–Derived Polygenic Risk Scores for Major Depressive Disorder and White Matter Microstructure in the UK Biobank.
        Biol Psychiatry Cogn Neurosci Neuroimaging. 2019 Jan 1; 4: 91-100
        View in Article
        • Moyer CE
        • Shelton MA
        • Sweet RA
        Dendritic spine alterations in schizophrenia.
        Neurosci Lett. 2015 Aug 5; 601: 46-53
        View in Article
        • Föcking M
        • Lopez LM
        • English JA
        • Dicker P
        • Wolff A
        • Brindley E
        • et al.
        Proteomic and genomic evidence implicates the postsynaptic density in schizophrenia.
        Mol Psychiatry. 2015 204; ([Internet]. 2014 Jul 22 [cited 2022 May 5];20(4):424–32. Available from:)
        https://www.nature.com/articles/mp201463
        View in Article

      14. Shen E, Shulha H, Weng Z, Akbarian S. Regulation of histone H3K4 methylation in brain development and disease. Philos Trans R Soc B Biol Sci [Internet]. 2014 [cited 2022 May 5];369(1652). Available from: https://royalsocietypublishing.org/doi/full/10.1098/rstb.2013.0514

        View in Article

      15. Sudlow C, Gallacher J, Allen N, Beral V, Burton P, Danesh J, et al. UK Biobank: An Open Access Resource for Identifying the Causes of a Wide Range of Complex Diseases of Middle and Old Age. PLOS Med [Internet]. 2015 Mar 1 [cited 2022 Apr 25];12(3):e1001779. Available from: https://journals.plos.org/plosmedicine/article?id=10.1371/journal.pmed.1001779

        View in Article

      16. Gene Ontology Resource [Internet]. [cited 2022 Apr 26]. Available from: http://geneontology.org/

        View in Article

      17. Gupta S, Kim SY, Artis S, Molfese DL, Schumacher A, Sweatt JD, et al. Histone Methylation Regulates Memory Formation. J Neurosci [Internet]. 2010 Mar 10 [cited 2022 May 16];30(10):3589. Available from: /pmc/articles/PMC2859898/

        View in Article

      18. Zhu Y, Wang S, Gong X, Edmiston EK, Zhong S, Li C, et al. Associations between hemispheric asymmetry and schizophrenia-related risk genes in people with schizophrenia and people at a genetic high risk of schizophrenia. Br J Psychiatry [Internet]. 2021 Jul 1 [cited 2022 May 16];219(1):392–400. Available from: https://www.cambridge.org/core/journals/the-british-journal-of-psychiatry/article/associations-between-hemispheric-asymmetry-and-schizophreniarelated-risk-genes-in-people-with-schizophrenia-and-people-at-a-genetic-high-risk-of-schizophrenia/8506769E0926B8AF52DB15E6D3FBA614

        View in Article

      19. Bycroft C, Freeman C, Petkova D, Band G, Elliott LT, Sharp K, et al. The UK Biobank resource with deep phenotyping and genomic data. Nat 2018 5627726 [Internet]. 2018 Oct 10 [cited 2022 Apr 25];562(7726):203–9. Available from: https://www.nature.com/articles/s41586-018-0579-z

        View in Article
        • Choi SW
        • Mak TSH
        • O’Reilly PF
        Tutorial: a guide to performing polygenic risk score analyses.
        Nat Protoc. 2020 Sep 1; 15: 2759-2772
        View in Article

      21. Smith BH, Campbell A, Linksted P, Fitzpatrick B, Jackson C, Kerr SM, et al. Cohort Profile: Generation Scotland: Scottish Family Health Study (GS:SFHS). The study, its participants and their potential for genetic research on health and illness. Int J Epidemiol [Internet]. 2013 Jun 1 [cited 2022 Mar 23];42(3):689–700. Available from: https://academic.oup.com/ije/article/42/3/689/909916

        View in Article

      22. Wang K, Li M, Hakonarson H. ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res [Internet]. 2010 Sep 1 [cited 2022 Apr 26];38(16):e164–e164. Available from: https://academic.oup.com/nar/article/38/16/e164/1749458

        View in Article

      23. Euesden J, Lewis CM, O’Reilly PF. PRSice: Polygenic Risk Score software. Bioinformatics [Internet]. 2015 May 1 [cited 2022 Apr 26];31(9):1466–8. Available from: https://pubmed.ncbi.nlm.nih.gov/25550326/

        View in Article
        • Schoorl J
        • Barbu MC
        • Shen X
        • Harris MR
        • Adams MJ
        • Whalley HC
        • et al.
        Grey and white matter associations of psychotic-like experiences in a general population sample (UK Biobank).
        Transl Psychiatry. 2021; 111 ([Internet]. 2021 Jan 7 [cited 2022 Apr 26];11(1):1–11. Available from:)
        https://www.nature.com/articles/s41398-020-01131-7
        View in Article
        • Alloza C
        • Blesa-Cábez M
        • Bastin ME
        • Madole JW
        • Buchanan CR
        • Janssen J
        • et al.
        Psychotic-like experiences, polygenic risk scores for schizophrenia, and structural properties of the salience, default mode, and central-executive networks in healthy participants from UK Biobank.
        Transl Psychiatry. 2020 101; ([Internet]. 2020 Apr 27 [cited 2022 Apr 26];10(1):1–13. Available from:)
        https://www.nature.com/articles/s41398-020-0794-x
        View in Article

      26. Bosma MJ, Cox SR, Ziermans T, Buchanan CR, Shen X, Tucker-Drob EM, et al. White matter, cognition and psychotic-like experiences in UK Biobank. Psychol Med [Internet]. 2021 [cited 2022 Apr 26];1–10. Available from: https://www.cambridge.org/core/journals/psychological-medicine/article/white-matter-cognition-and-psychoticlike-experiences-in-uk-biobank/5A6A3F2BE82DF66341271996187D2200

        View in Article

      27. Alfaro-Almagro F, Jenkinson M, Bangerter NK, Andersson JLR, Griffanti L, Douaud G, et al. Image processing and Quality Control for the first 10,000 brain imaging datasets from UK Biobank. Neuroimage [Internet]. 2018 Feb 1 [cited 2022 Apr 25];166:400–24. Available from: https://pubmed.ncbi.nlm.nih.gov/29079522/

        View in Article

      28. Smith SM, Alfaro-Almagro F, Miller KL. UK Biobank Brain Imaging Documentation UK Biobank Brain Imaging Documentation Contributors to UK Biobank Brain Imaging. [cited 2022 Apr 25]; Available from: http://www.ukbiobank.ac.uk

        View in Article
        • Shen X
        • Adams MJ
        • Ritakari TE
        • Cox SR
        • McIntosh AM
        • Whalley HC
        White Matter Microstructure and Its Relation to Longitudinal Measures of Depressive Symptoms in Mid- and Late Life.
        Biol Psychiatry. 2019 Nov 15; 86: 759-768
        View in Article
        • Shen X
        • Reus LM
        • Cox SR
        • Adams MJ
        • Liewald DC
        • Bastin ME
        • et al.
        Subcortical volume and white matter integrity abnormalities in major depressive disorder: findings from UK Biobank imaging data.
        Sci Reports. 2017; 71 ([Internet]. 2017 Jul 17 [cited 2022 May 19];7(1):1–10. Available from:)
        https://www.nature.com/articles/s41598-017-05507-6
        View in Article
        • Alexander B
        • Loh WY
        • Matthews LG
        • Murray AL
        • Adamson C
        • Beare R
        • et al.
        Desikan-Killiany-Tourville Atlas compatible version of m-CRIB neonatal parcellated whole brain atlas: The m-Crib 2.0.
        Front Neurosci. 2019; 13: 34
        View in Article

      32. Elliott LT, Sharp K, Alfaro-Almagro F, Shi S, Miller KL, Douaud G, et al. Genome-wide association studies of brain imaging phenotypes in UK Biobank. Nat 2018 5627726 [Internet]. 2018 Oct 10 [cited 2022 May 19];562(7726):210–6. Available from: https://www.nature.com/articles/s41586-018-0571-7

        View in Article

      33. Fischl B, Salat DH, Busa E, Albert M, Dieterich M, Haselgrove C, et al. Whole brain segmentation: automated labeling of neuroanatomical structures in the human brain. Neuron [Internet]. 2002 Jan 31 [cited 2022 Apr 25];33(3):341–55. Available from: https://pubmed.ncbi.nlm.nih.gov/11832223/

        View in Article

      34. O’Connell KS, Sønderby IE, Frei O, Van Der Meer D, Athanasiu L, Smeland OB, et al. Association between complement component 4A expression, cognitive performance and brain imaging measures in UK Biobank. Psychol Med [Internet]. 2021 [cited 2022 Apr 25];1–11. Available from: https://www.cambridge.org/core/journals/psychological-medicine/article/association-between-complement-component-4a-expression-cognitive-performance-and-brain-imaging-measures-in-uk-biobank/D26127B6301C8C58148F3F4233403022

        View in Article

      35. Green C, Stolicyn A, Harris MA, Shen X, Romaniuk L, Barbu MC, et al. Hair glucocorticoids are associated with childhood adversity, depressive symptoms and reduced global and lobar grey matter in Generation Scotland. Transl Psychiatry 2021 111 [Internet]. 2021 Oct 12 [cited 2022 Apr 25];11(1):1–9. Available from: https://www.nature.com/articles/s41398-021-01644-9

        View in Article
        • Benjamini Y
        • Hochberg Y
        Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing.
        Source J R Stat Soc Ser B. 1995; 57: 289-300
        View in Article

      37. Cabrera CP, Navarro P, Huffman JE, Wright AF, Hayward C, Campbell H, et al. Uncovering networks from genome-wide association studies via circular genomic permutation. G3 (Bethesda) [Internet]. 2012 Sep [cited 2022 Apr 26];2(9):1067–75. Available from: https://pubmed.ncbi.nlm.nih.gov/22973544/

        View in Article

      38. Bayés Á, Van De Lagemaat LN, Collins MO, Croning MDR, Whittle IR, Choudhary JS, et al. Characterization of the proteome, diseases and evolution of the human postsynaptic density. Nat Neurosci [Internet]. 2011 Jan [cited 2022 Jul 18];14(1):19–21. Available from: https://pubmed.ncbi.nlm.nih.gov/21170055/

        View in Article

      39. Sorokina O, Mclean C, Croning MDR, Heil KF, Wysocka E, He X, et al. A unified resource and configurable model of the synapse proteome and its role in disease. Sci Reports 2021 111 [Internet]. 2021 May 11 [cited 2022 Jul 18];11(1):1–9. Available from: https://www.nature.com/articles/s41598-021-88945-7

        View in Article

      40. Fromer M, Pocklington AJ, Kavanagh DH, Williams HJ, Dwyer S, Gormley P, et al. De novo mutations in schizophrenia implicate synaptic networks. Nature [Internet]. 2014 [cited 2022 Jul 18];506(7487):179–84. Available from: https://pubmed.ncbi.nlm.nih.gov/24463507/

        View in Article

      41. Purcell SM, Moran JL, Fromer M, Ruderfer D, Solovieff N, Roussos P, et al. A polygenic burden of rare disruptive mutations in schizophrenia. Nature [Internet]. 2014 [cited 2022 Jul 18];506(7487):185–90. Available from: https://pubmed.ncbi.nlm.nih.gov/24463508/

        View in Article
        • Skene NG
        • Roy M
        • Grant SG
        A genomic lifespan program that reorganises the young adult brain is targeted in schizophrenia.
        Elife. 2017 Sep 12; : 6
        View in Article

      43. Nithianantharajah J, Komiyama NH, McKechanie A, Johnstone M, Blackwood DH, Clair DS, et al. Synaptic scaffold evolution generated components of vertebrate cognitive complexity. Nat Neurosci 2012 161 [Internet]. 2012 Dec 2 [cited 2022 Jul 18];16(1):16–24. Available from: https://www.nature.com/articles/nn.3276

        View in Article
        • Bersani FS
        • Minichino A
        • Fojanesi M
        • Gallo M
        • Maglio G
        • Valeriani G
        • Biondi M FP
        Cingulate Cortex in Schizophrenia: its relation with negative symptoms and psychotic onset. A review study.
        Eur Rev Med Pharmacol Sci. 2014; 18: 3354-3367
        View in Article
        • Wang L
        • Hosakere M
        • Trein JCL
        • Miller A
        • Ratnanather JT
        • Barch DM
        • et al.
        Abnormalities of cingulate gyrus neuroanatomy in schizophrenia.
        Schizophr Res. 2007 Jul 1; 93: 66-78
        View in Article
        • Neilson E
        • Shen X
        • Cox SR
        • Clarke TK
        • Wigmore EM
        • Gibson J
        • et al.
        Impact of Polygenic Risk for Schizophrenia on Cortical Structure in UK Biobank.
        Biol Psychiatry. 2019 Oct 1; 86: 536-544
        View in Article

      47. Zhu X, Ward J, Cullen B, Lyall DM, Strawbridge RJ, Smith DJ, et al. Polygenic Risk for Schizophrenia, Brain Structure, and Environmental Risk in UK Biobank. Schizophr Bull Open [Internet]. 2021 Jan 1 [cited 2022 May 12];2(1). Available from: https://academic.oup.com/schizbullopen/article/2/1/sgab042/6375001

        View in Article
        • Wang Z
        • Li P
        • Wu T
        • Zhu S
        • Deng L
        • Cui G
        Axon guidance pathway genes are associated with schizophrenia risk.
        Exp Ther Med. 2018 Dec 1; ([Internet]) ([cited 2022 May 13];16(6):4519–26. Available from:)
        http://www.spandidos-publications.com/10.3892/etm.2018.6781/abstract
        View in Article
        • Mukai J
        • Tamura M
        • Fénelon K
        • Rosen AM
        • Spellman TJ
        • Kang R
        • et al.
        Molecular Substrates of Altered Axonal Growth and Brain Connectivity in a Mouse Model of Schizophrenia.
        Neuron. 2015 May 6; 86: 680-695
        View in Article
        • Lebel C
        • Deoni S
        The development of brain white matter microstructure.
        Neuroimage. 2018 Nov 15; 182: 207-218
        View in Article

      51. Van Hoesen GW, Augustinack JC, Dierking J, Redman SJ, Thangavel R. The Parahippocampal Gyrus in Alzheimer’s Disease: Clinical and Preclinical Neuroanatomical Correlates. Ann N Y Acad Sci [Internet]. 2000 Jun 1 [cited 2022 May 13];911(1):254–74. Available from: https://onlinelibrary.wiley.com/doi/full/10.1111/j.1749-6632.2000.tb06731.x

        View in Article
        • McIntosh AM
        • Maniega SM
        • Lymer GKS
        • McKirdy J
        • Hall J
        • Sussmann JED
        • et al.
        White Matter Tractography in Bipolar Disorder and Schizophrenia.
        Biol Psychiatry. 2008 Dec 15; 64: 1088-1092
        View in Article
        • Zhu X
        • Ward J
        • Cullen B
        • Lyall DM
        • Strawbridge RJ
        • Lyall LM
        • et al.
        Phenotypic and genetic associations between anhedonia and brain structure in UK Biobank.
        Transl Psychiatry. 2021; 111 ([Internet]. 2021 Jul 16 [cited 2022 May 13];11(1):1–8. Available from:)
        https://www.nature.com/articles/s41398-021-01522-4
        View in Article
        • Greer EL
        • Shi Y
        Histone methylation: a dynamic mark in health, disease and inheritance.
        Nat Rev Genet. 2012 135; ([Internet]. 2012 Apr 3 [cited 2022 May 13];13(5):343–57. Available from:)
        https://www.nature.com/articles/nrg3173
        View in Article
        • Gavin DP
        • Sharma RP
        Histone modifications, DNA methylation, and Schizophrenia.
        Neurosci Biobehav Rev. 2010 May 1; 34: 882-888
        View in Article

      56. Montano C, Taub MA, Jaffe A, Briem E, Feinberg JI, Trygvadottir R, et al. Association of DNA Methylation Differences With Schizophrenia in an Epigenome-Wide Association Study. JAMA psychiatry [Internet]. 2016 May 1 [cited 2022 May 16];73(5):506–14. Available from: https://pubmed.ncbi.nlm.nih.gov/27074206/

        View in Article

      57. Huang HS, Matevossian A, Whittle C, Se YK, Schumacher A, Baker SP, et al. Prefrontal Dysfunction in Schizophrenia Involves Mixed-Lineage Leukemia 1-Regulated Histone Methylation at GABAergic Gene Promoters. J Neurosci [Internet]. 2007 Oct 17 [cited 2022 May 13];27(42):11254. Available from: /pmc/articles/PMC6673022/

        View in Article

      58. Schultz CC, Koch K, Wagner G, Roebel M, Schachtzabel C, Nenadic I, et al. Psychopathological correlates of the entorhinal cortical shape in schizophrenia. Eur Arch Psychiatry Clin Neurosci [Internet]. 2010 Jun 7 [cited 2022 May 13];260(4):351–8. Available from: https://link.springer.com/article/10.1007/s00406-009-0083-4

        View in Article
        • Sumiyoshi T
        • Tsunoda M
        • Uehara T
        • Tanaka K
        • Itoh H
        • Sumiyoshi C
        • et al.
        Enhanced locomotor activity in rats with excitotoxic lesions of the entorhinal cortex, a neurodevelopmental animal model of schizophrenia: Behavioral and in vivo microdialysis studies.
        Neurosci Lett. 2004 Jul 1; 364: 124-129
        View in Article
        • Glausier JR
        • Lewis DA
        Dendritic spine pathology in schizophrenia.
        Neuroscience. 2013 Oct 22; 251: 90-107
        View in Article

      61. De Bartolomeis A, Latte G, Tomasetti C, Iasevoli F. Glutamatergic postsynaptic density protein dysfunctions in synaptic plasticity and dendritic spines morphology: Relevance to schizophrenia and other behavioral disorders pathophysiology, and implications for novel therapeutic approaches. Mol Neurobiol [Internet]. 2014 Sep 3 [cited 2022 May 16];49(1):484–511. Available from: https://link.springer.com/article/10.1007/s12035-013-8534-3

        View in Article
        • Kirov G
        • Pocklington AJ
        • Holmans P
        • Ivanov D
        • Ikeda M
        • Ruderfer D
        • et al.
        De novo CNV analysis implicates specific abnormalities of postsynaptic signalling complexes in the pathogenesis of schizophrenia.
        Mol Psychiatry. 2012; 17: 142-153
        View in Article

      63. van der Merwe C, Passchier R, Mufford M, Ramesar R, Dalvie S, Stein DJ. Polygenic risk for schizophrenia and associated brain structural changes: A systematic review. Compr Psychiatry [Internet]. 2019 Jan 1 [cited 2022 Apr 26];88:77–82. Available from: https://pubmed.ncbi.nlm.nih.gov/30529765/

        View in Article
        • Stauffer EM
        • Bethlehem RAI
        • Warrier V
        • Murray GK
        • Romero-Garcia R
        • Seidlitz J
        • et al.
        Grey and white matter microstructure is associated with polygenic risk for schizophrenia.
        Mol Psychiatry. 2021 2612; ([Internet]. 2021 Aug 30 [cited 2022 May 16];26(12):7709–18. Available from:)
        https://www.nature.com/articles/s41380-021-01260-5
        View in Article
        • Fry A
        • Littlejohns TJ
        • Sudlow C
        • Doherty N
        • Adamska L
        • Sprosen T
        • et al.
        Comparison of Sociodemographic and Health-Related Characteristics of UK Biobank Participants with Those of the General Population.
        Am J Epidemiol. 2017; 186: 1026-1034
        View in Article

      66. Kelleher I, Cannon M. Psychotic-like experiences in the general population: characterizing a high-risk group for psychosis. Psychol Med. 2011 Jan;41(1):1-6. doi: 10.1017/S0033291710001005. Epub 2010 May 19. PMID: 20624328.

        View in Article

      67. Legge SE, Jones HJ, Kendall KM, Pardiñas AF, Menzies G, Bracher-Smith M, et al. Association of Genetic Liability to Psychotic Experiences With Neuropsychotic Disorders and Traits. JAMA Psychiatry. 2019 Dec 1;76(12):1256-1265. doi: 10.1001/jamapsychiatry.2019.2508. PMID: 31553412; PMCID: PMC6764002.

        View in Article
        • Pergola G
        • Penzel N
        • Sportelli L
        • Bertolino A
        Lessons learned from parsing genetic risk for schizophrenia into biological pathways.
        Biological Psychiatry. 2022 Oct 28;
        View in Article
        • Morey RA
        • Petty CM
        • Xu Y
        • Hayes JP
        • Wagner II, HR
        • Lewis DV
        • LaBar KS
        • Styner M
        • McCarthy G
        A comparison of automated segmentation and manual tracing for quantifying hippocampal and amygdala volumes.
        Neuroimage. 2009 Apr 15; 45: 855-866
        View in Article

      Article info

      Publication history

      Accepted: March 5, 2023
      Received in revised form: March 2, 2023
      Received: October 17, 2022

      Publication stage

      In Press Accepted Manuscript

      Footnotes

      Funding

      This research was funded in whole, or in part, by the Wellcome Trust [216767/Z/19/Z]. This analysis uses linked data from UKB applications 4844 and 16124. For the purpose of open access, the author has applied a CC BY public copyright licence to any Author Accepted Manuscript version arising from this submission. MCB is supported by a Guarantors of Brain Non-clinical Post-Doctoral Fellowship and British Medical Association Margaret Temple Grant. HCW is supported by the Royal College of Physicians of Edinburgh (Sim Fellowship). MJA is supported by MRC Mental Health Data Pathfinder Award (MC_PC_17209). KM is supported by the Chief Scientist Office for Scotland and NHS Education for Scotland (PCL/17/01). This study is supported by a Wellcome Trust Strategic Award “Stratifying Resilience and Depression Longitudinally” (STRADL) (Reference 104036/Z/14/Z).

      Financial disclosures

      AMM has previously received funding from commercial sources (Pfizer, Roche, Abbvie, Sunovion, Jannssen and Lilly), but none of these funds or funders were used or involved in the current study. All other authors report no biomedical financial interests or potential conflicts of interest.

      Data sharing statement

      Data used in the preparation of this article were obtained from the UK Biobank cohort (https://www.ukbiobank.ac.uk/).

      Scripts for the analyses in this project can be accessed at the following GitHub repository: https://github.com/mirunabarbu8/Fellowship-project.

      Identification

      DOI: https://doi.org/10.1016/j.bpsgos.2023.03.004

      Copyright

      © 2023 Published by Elsevier Inc. on behalf of Society of Biological Psychiatry.

      User license

      Creative Commons Attribution (CC BY 4.0) |
      How you can reuse Information Icon
      Creative Commons Attribution (CC BY 4.0)

      Permitted

      • Read, print & download
      • Redistribute or republish the final article
      • Text & data mine
      • Translate the article
      • Reuse portions or extracts from the article in other works
      • Sell or re-use for commercial purposes

      Elsevier's open access license policy

      ScienceDirect

      Access this article on ScienceDirect

      The content on this site is intended for healthcare professionals.


      We use cookies to help provide and enhance our service and tailor content. To update your cookie settings, please visit the Cookie Preference Center for this site.
      Copyright © 2023 Elsevier Inc. except certain content provided by third parties.


      RELX