Cytokines and schizophrenia revisited: a two-sample multi-marker Mendelian randomization approach
Ren, H.; Meng, Y.; Zhang, Y.; Wang, Q.; Deng, W.; Ma, X.; Zhao, L.; Li, X.; Wang, Y.; Sham, P. C.; Li, T.
Show abstract
BackgroundSchizophrenia is a complex mental disorder with recent evidence suggesting a critical immune component underpinning its pathophysiology. Two-sample Mendelian randomization (MR) provided an opportunity to probe the immune changes in schizophrenia by harnessing the increasing availability of summary-level data from large GWAS consortia. ObjectiveTo map the extensive immune response of schizophrenia in terms of cytokines/chemokines and to explore the effect of cytokines induced by schizophrenia (SCZ-induced cytokines) on the brain structure and function Sources and methodsUsing the summary-level data generated from GWAS of schizophrenia, cytokines in the peripheral blood and imaging-derived phenotypes (IDPs), we performed two rounds of two-sample MR analysis; the identified cytokines from first round of analysis (schizophrenia => cytokines) were modeled for its underlying structure and subsequent clustering analysis further grouped SCZ-induced cytokines based on their genetic similarities. The multi-phenotype summary statistics of each cytokine module were then used as instrumental variables (IVs) for the second round of MR analysis to detect their effect on brain structure and function. ResultsThe first round of MR analysis identified nine cytokines, the highlight of which includes IL18 (OR = 1.292, P = 8.37 x 10-42) and TNFa (OR = 0.721, P = 7.33 x 10-6), to be causally associated with schizophrenia. These SCZ-induced cytokines could be clustered into three modules. The second round of MR analysis (cytokine module => IDPs) indicated that module B (SCGFb-IP10-CTACK-IL6) significantly increased the level of IDPs including IDP_T1_SIENAX_peripheral_grey_normalised_volume ({beta} = 0.0453, P = 4.40x1010), IDP_dMRI_TBSS_MD_Posterior_corona_radiata_R ({beta}= 0.0584, P = 8.89x 10-16) and IDP_dMRI_TBSS_MD_Cingulum_hippocampus_R ({beta} = 0.0563, P = 9.88x 10-15), with module C (IL18-GROa-TNFa) increasing the level of IDP_dMRI_TBSS_L2_Posterior_thalamic_radiation_R ({beta}= 0.0341, P = 2.67x 10-6). ConclusionOur study, for the first time, mapped the causal link from schizophrenia to the comprehensive immune responses, and the findings suggest immune networks play a role in pathophysiology of schizophrenia by mediating the deviations of total gray matter volume and white matter fibers possibly in the mesolimbic system.
Matching journals
The top 8 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
- Computational modelling of EEG and fMRI paradigms reveals a consistent loss of pyramidal cell synaptic gain in schizophrenia 94%
- Connectivity patterns of task-specific brain networks allow individual prediction of cognitive symptom dimension of schizophrenia and link to molecular architecture 94%
- Pathogenic mis-splicing of CPEB4 in schizophrenia 94%
Similar papers in this journal
- Schizophrenia Risk Alleles Often Affect The Expression of Many Genes and Each Gene May Have a Different Effect On The Risk; A Mediation Analysis. 94%
- Polygenic risk scores for psychiatric, inflammatory, and cardio-metabolic traits and diseases highlight possible genetic overlaps with suicide attempt and treatment-emergent suicidal ideation 93%
- SLC39A8.p.(Ala391Thr) is associated with poorer cognitive ability: a cross-sectional study of schizophrenia and the general UK population 92%
Similar papers in this journal
- Neuronal-specific methylome and hydroxymethylome analysis reveal replicated and novel loci associated with alcohol use disorder 91%
- Polygenic heterogeneity across obsessive-compulsive disorder subgroups defined by a comorbid diagnosis 91%
- Personality and MB-COMT gene: molecular-genetic and epigenetic associations with NEO-PI-R personality domains and facets in monozygotic twins 88%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.