Cocaine perturbs neurodevelopment and increases neuroinflammation in a prenatal cerebral organoid model

Abstract

Prenatal exposure to cocaine causes abnormalities in foetal brain development, which are linked to later development of anxiety, depression and cognitive dysfunction. Previous studies in rodent models have indicated that prenatal cocaine exposure affects proliferation, differentiation and connectivity of neural cell types. Here, using cerebral organoids derived from the human iPSC cell line HPSI1213i-babk_2, we investigated cocaine-induced changes of the gene expression regulatory landscape at an early developmental time point, leveraging recent advances in single cell RNA-seq and single cell ATAC-seq. iPSC-cerebral organoids replicated well-established cocaine responses observed in vivo and provided additional information about the cell-type specific regulation of gene expression following cocaine exposure. Cocaine altered gene expression patterns, in part through epigenetic landscape remodelling, and revealed disordered neural plasticity mechanisms in the cerebral organoids. Perturbed neurodevelopmental cellular signalling and an inflammatory-like activation of astrocyte populations were also evident following cocaine exposure. The combination of altered neuroplasticity, neurodevelopment and neuroinflammatory signalling suggests cocaine exposure can mediate substantial disruption of normal development and maturation of the brain. These findings offer new insights into the cellular mechanism underlying the adverse effects of cocaine exposure on neurodevelopment and point to the possible pathomechanisms of later neuropsychiatric disturbances.

Fig. 1. scRNAseq characterisation of the effect of cocaine on the 36-day old cerebral organoid.

a UMAP projection of 4437 single cells revealing 10 distinct clusters within control and cocaine-treated organoids. b Differentially expressed genes used for annotation within each cluster. c UMAP projection of astrocyte, neurogenic niche and choroid/roof plate lineage cell types in cocaine-treated organoids coloured by pseudotime. d Overlayed UMAP of 1681 cells for control and 2756 cells for cocaine. e Proportion of cells within each cluster in control and cocaine-treated organoids. f Differentially expressed genes within each cluster in response to cocaine. Genes were selected based on a |log₂FC| > 0.4, pct > 0.25, and p < 0.05. g Top dysregulated pathways in cocaine-treated organoids identified by IPA analysis. h A selection of previously identified cocaine-associated genes are increased in cocaine-treated organoids.

Fig. 2. Cocaine alters neural plasticity, developmental signalling and neuroinflammatory tone.

a Immunostaining revealed increased FOSB expression in organoids treated with cocaine. Magnification 40X. Scale: 50μm. b Increased expression of FOS, JUNB, and JUND and c Immediate early genes (IEGs) were evident across the organoids. d Increased expression of synaptic plasticity-associated genes, the glutamate ionotropic receptor GluD2 (GRID2), and Na⁺-dependent excitatory amino acid transporters, EAAT1 (SLC1A3), and SNAT1 (SLC38A1) were observed in the neurogenic niche clusters. e Increased expression of the vesicle glutamate transporter VGLUT3 (SLC17A8) and NRN1 was observed in the choroid/plate niche. f Increased expression of oxidative stress and neuroinflammation associated genes were observed in the astrocyte-like clusters.

Fig. 3. Cocaine-mediated remodelling of the epigenetic landscape.

a Increased expression of chromatin modifiers was observed across clusters in response to cocaine. b UMAP plot of annotated scATACseq dataset with overlayed UMAP of 3647 cells for control and 3328 cells for cocaine. c UMAP projection of chromatin accessibility in astrocytes, neurogenic niche, and choroid/roof plate cell types in cocaine-treated organoids coloured by pseudotime. d Differentially accessible regions (DARs) within each cluster in response to cocaine. DARs were selected based on a log₂FC ≤ −0.5 or ≥ 0.5, pct ≥ 0.25, and p < 0.05. e Global changes in chromatin accessibility at DNase I hypersensitivity sites, transcription start sites, promoters and enhancers. ****P < 0.0001, ***P < 0.001, **P < 0.01, *P < 0.05, ns ≥ 0.05; pairwise t-test with Bonferroni correction. f Heatmap of chromVAR computed deviations z-score for the topmost variable TF motifs enriched in cocaine-treated organoids. g UMAPs of scRNAseq gene expression (left panel) and scATACseq based ChromVAR motif deviation scores (right panel) for FOS, JUND, TCF7L2, ATF4, GATA6, and NFIB. Prediction of cis co-accessibility networks between enhancer (blue) and promoter (red) elements at sample Immediate Early Gene loci, h FOSL1 and i MAPK in response to cocaine. Higher co-accessibility score indicates higher co-accessibility between promoter and enhancer elements.

Fig. 4. Identifying the transcription factor networks underpinning cocaine-mediated actions in the cerebral organoid.

a Heatmap of top differentially regulated transcription factor networks (“regulons”) across all clusters in response to cocaine. b UMAP visualisation of TF regulons for JUND, FOS, and ATF4. c Motif-centric footprinting estimating transcription factor binding for JUND, FOS, ATF4 in response to cocaine and d across clusters.

Fig. 5. Cocaine-mediated alteration in cellular communication in the cerebral organoid.

a Increased number of inferred interactions between most clusters were apparent in cocaine-treated organoids relative to control. b Circle plots depicting interaction numbers and strength between all cell clusters in control and cocaine-treated organoids. Line thickness represents increased interaction between clusters. c Signalling in PTN, NCAM, MK, CD99, JAM, ncWNT, Laminin, Notch, and Collagen pathways were increased in cocaine-treated organoids. d Heatmap of notch signalling pathway network contributing to mostly outgoing or incoming signalling between cell groups. e Circle plot depicting the notch signalling pathway network. f Overall contribution of each ligand-receptor pair to the notch signalling pathway amongst cell types. g Heatmap of non-canonical Wnt signalling pathway network contributing to mostly outgoing or incoming signalling between cell groups. h Chord diagram showing the non-canonical Wnt (ncWnt) signalling pathway network. i Overall contribution of each ligand-receptor pair to the ncWnt signalling pathway amongst cell types. j Notch and ncWnt ligand receptor interactions between AS2 and all other cell subtypes.

Fig. 6. Altered midkine and pleiotrophin signalling in cocaine-treated organoids.

a Roof plate and astrocyte-like cells contributed to the majority of midkine (MDK) signalling in cocaine-treated organoids. b Circle plot depicting the MDK signalling pathway network. c Increased expression of midkine and pleiotrophin (PTN)-associated genes were evident in cocaine-treated organoids. d Overall contribution of each ligand-receptor pair to the MDK pathway amongst cell types. e Cell-cell communication network for the MDK-Nucleolin (NCL) ligand-receptor pair. f Proliferating radial glia, roof plate, and radial glial cells contribute to the majority of pleiotrophin signalling. g Circle plot depicting the PTN signalling pathway network. h Overall contribution of each ligand-receptor pair to the PTN pathway amongst cell types. i Cell-cell communication network for the PTN-NCL ligand-receptor pair.