Transcriptional and neurochemical profiles of cerebral blood flow changes in psychosis and prodrome
Poster No:
494
Submission Type:
Abstract Submission
Authors:
Samuel Knight1, Leyla Abbasova1, Yashar Zeighami2, Daniel Martins1, Fernando Zelaya1, Ottavia Dipasquale3, Thomas Liu4, David Shin5, Matthijs Bossong6, Matilda Azis1, Mathilde Antoniades7, Alice Egerton1, Paul Allen1, Owen O'Daly1, Philip McGuire8, Gemma Modinos1
Institutions:
1King's College London, London, United Kingdom, 2Douglas Research Centre, Montreal, Quebec, 3Olea Medical, La Ciotat, France, 4UC San Diego, San Diego, CA, 5GEHeathCare, Menlo Park, CA, 6University Medical Center Utrecht, Utrecht, Netherlands, 7University of Pennsylvania, Philadelphia, PA, 8Oxford University, Oxford, United Kingdom
First Author:
Co-Author(s):
Introduction:
In vivo investigations have demonstrated resting regional cerebral blood flow (rCBF) alterations in patients with schizophrenia (SCZ) and individuals at clinical high-risk for psychosis (CHR)(du Sert et al. 2023). Understanding how these differences in rCBF are related to dysfunction at the genetic or molecular level has the potential to inform the discovery of new therapeutic strategies. Recently, several tools have become available to combine MRI and genetic data: the Allen Human Brain Atlas (AHBA) and neurochemical data derived from PET atlases. These approaches have yet to be applied to investigate the underlying molecular mechanisms of rCBF alterations in psychosis. The study aimed to determine the gene expression and neuroreceptor density profiles that correspond to rCBF phenotypes in CHR and SCZ, thereby gaining insight into the cellular and molecular mechanisms underlying these alterations.
Methods:
Participants included 122 patients with SCZ and 116 healthy controls (HC1), and 129 CHR individuals and 58 healthy controls (HC2). Due to differences in scanning modality, pre-processing, and sample characteristics, each clinical group was compared to its own matched group of healthy controls, and not between clinical groups. rCBF maps from all participants were estimated from arterial spin labelling (ASL) data and pre-processed using the CBFBIRN pipeline(Shin et al. 2014) ('SCZ and HC1') and the ASL Toolbox (all participants)(Mato et al. 2015). SPM12 was used to derive case-control t-stat maps of and 'SCZvsHC1' and 'CHRvsHC2' comparisons. Unthresholded case-control rCBF difference maps and receptor atlases were segmented into the same standard space (Schaefer+Xiao atlases, 100 cortical and 22 subcortical regions)(Schaefer et al. 2018; Xiao et al. 2019) (Figure 1a). We next looked for spatial associations between rCBF t-stat maps and gene expression data, accessed via the AHBA, as well as neuroreceptor binding distribution of 19 freely available PET atlases across 9 neurotransmitter systems (Hansen et al. 2022). Gene expression correlations with rCBF maps were assessed with Spearman's rank correlations, and a dominance regression analysis was used to determine the unique contribution of mean receptor binding values to the prediction of mean rCBF differences in each region. Significance of all analyses was assessed against FDR-corrected spatial autocorrelation-preserving null models.
Results:
The rCBF profile of both 'SCZvsHC1' and 'CHRvsHC2' alterations was found to track the distribution of astrocytes, oligodendrocyte precursor and vascular leptomeningeal cell gene markers (PFDR<0.05)(Figure 1b). Receptor distribution significantly predicted 'SCZvsHC1' and 'CHRvsHC2' difference patterns (R2adj=.58,PFDR<.05; R2adj=.6, PFDR<.05 respectively) (Figure 1c). Dopamine D1 and D2 and GABA-A receptors, and acetylcholine transporters contributed most to 'SCZvsHC1' rCBF differences, while 5-HT1a, muscarinic 1, norepinephrine, CB1, and NMDA receptors, and dopamine transporter contributed most to the prediction of 'CHRvsHC2' rCBF differences.
·Figure 1
Conclusions:
In summary, our findings implicate cell types involved in stress response and neuroinflammation, as well as dopamine, GABA-A, and NMDA receptor systems as distinct cellular and neurochemical signatures of SCZ- and CHR-associated rCBF profiles. Such hypothesis-generating approaches could be utilised in future to guide the non-invasive stratification of mechanisms of risk, which may be amenable to pharmacological intervention.
Disorders of the Nervous System:
Psychiatric (eg. Depression, Anxiety, Schizophrenia) 1
Genetics:
Transcriptomics
Modeling and Analysis Methods:
Multivariate Approaches
Physiology, Metabolism and Neurotransmission :
Cerebral Metabolism and Hemodynamics 2
Pharmacology and Neurotransmission
Keywords:
Cerebral Blood Flow
Computational Neuroscience
FUNCTIONAL MRI
Neurotransmitter
Open Data
Preprint
Psychiatric
Psychiatric Disorders
Schizophrenia
1|2Indicates the priority used for review
Provide references using author date format
Shin, David D, (2014), 'Robust and Fast Quantification of CBF measures for Multiphase PCASL using Bayesian Nonlinear Model Fitting'. Proceedings of the International Society for Magnetic Resonance in Medicine, 22
Mato Abad V, (2016), 'ASAP (Automatic Software for ASL Processing): A toolbox for processing Arterial Spin Labeling images'. Magnetic Resonance Imaging. Apr;34(3):334-44.
Schaefer, A. (2018), ‘Local-Global Parcellation of the Human Cerebral Cortex from Intrinsic Functional Connectivity MRI’. Cerebral Cortex 28, no. 9: 3095–3114.
Xiao, Y. (2019), ‘An Accurate Registration of the BigBrain Dataset with the MNI PD25 and ICBM152 Atlases’. Scientific Data 6, no. 1: 210.
Hansen, J.Y., (2022), 'Mapping neurotransmitter systems to the structural and functional organization of the human neocortex'. Nature Neuroscience 25, 1569–1581.