NEW YORK — Different types of neurons in a key brain region undergo changes in schizophrenia, a new single-nucleus RNA sequencing study has found, suggesting there is general network impairment in patients with the condition.
Animal models have identified a number of brain cell types that are perturbed in schizophrenia, including different types of cortical γ-aminobutyric acid-expressing (GABAergic) interneurons. But as researchers led by the University of Copenhagen's Konstantin Khodosevich noted, these studies have largely focused on a few neuronal subtypes.
In their new study, appearing on Wednesday in Science Advances, Khodosevich and colleagues examined more than 220,000 neurons of all types, isolated from the dorsolateral prefrontal cortex of people with and without schizophrenia. Their single-nucleus RNA-sequencing study uncovered shifts in the number of GABAergic and principal neurons present among individuals with schizophrenia as well as changes in gene expression in those cells.
"Our work also shows that schizophrenia-associated changes do not affect only a selected family of neurons but involve multiple neuronal subtypes, indicating a general network impairment," the researchers wrote in their paper.
Khodosevich and colleagues analyzed post-mortem tissue samples from the Brodmann area 9 region of the dorsolateral prefrontal cortex of nine individuals who had schizophrenia and 14 control individuals who had no history of mental illness. This brain region had previously been tied to schizophrenia through functional and anatomical studies.
For each sample, the researchers isolated neuronal nuclei based on the expression of the neuronal marker NeuN for snRNA-seq. In all, they sequenced 225,012 nuclei, 209,053 of which passed quality control analyses.
Overall, the neuronal composition of the cortex differed among individuals with schizophrenia. There were fewer GABAergic interneurons, especially those belonging to the PVALB, SST, and VIP subtypes, but an increase in the portion of principal neurons belonging to the L2_3_CUX2 family and the L4_5_FEZF2_LRRK1 subtype. Additional compositional analysis underscored these changes in schizophrenia, while transcriptional analyses noted large differences in gene expression between the L2_3_CUX2_LAMP5 subfamily of principal neurons and the PVALB subfamily of GABAergic interneurons in schizophrenia versus controls.
The researchers noted, though, that schizophrenia is a highly variable condition with high inter-individual variability. When they compared the magnitude of the transcriptional differences between the affected and control individuals, the researchers found that the control samples clustered closely together, while the schizophrenia samples were less concentrated and more scattered.
Still, the researchers homed in on differentially expressed genes within each neuronal subtype among individuals with schizophrenia versus controls. These genes tended to affect certain pathways: Genes associated with energy metabolism and protein biogenesis pathways were downregulated, while upregulated genes were enriched within neurotransmission, plasticity, and developmental processes.
Further, a higher fraction of these differentially expressed genes was dysregulated in the upper cortical layer neuron subtypes, suggesting that this layer is a hotspot of schizophrenia-related changes. Upper layers, the researchers noted, are the most expanded cortical layers in the human cortex, as compared to rodent as well as other primate brains.
"[O]ur data support the theory that schizophrenia-associated developmental events occur in the [dorsolateral prefrontal cortex] during late gestation due to the disturbed arrival and integration of GABAergic interneurons into the cortical circuitry and differentiation of principal neurons," they added in their paper.
But single-nucleus studies like this one draw on a small number of patient samples. The researchers noted that additional studies examining larger numbers of samples with richer patient history data would provide additional insight into how potential covariates affect cell type abundance and gene expression.