Cell Type-Specific Profiles and Developmental Trajectories of Transcriptomes in Primate Prefrontal Layer 3 Pyramidal Neurons: Implications for Schizophrenia

Abstract

Objective: In schizophrenia, impaired working memory is associated with transcriptome alterations in layer 3 pyramidal neurons (L3PNs) in the dorsolateral prefrontal cortex (DLPFC). Distinct subtypes of L3PNs that send axonal projections to the DLPFC in the opposite hemisphere (callosal projection [CP] neurons) or the parietal cortex in the same hemisphere (ipsilateral projection [IP] neurons) play critical roles in working memory. However, how the transcriptomes of these L3PN subtypes might shift during late postnatal development when working memory impairments emerge in individuals later diagnosed with schizophrenia is not known. The aim of this study was to characterize and compare the transcriptome profiles of CP and IP L3PNs across developmental transitions from prepuberty to adulthood in macaque monkeys.

Methods: The authors used retrograde labeling to identify CP and IP L3PNs in the DLPFC of prepubertal, postpubertal, and adult macaque monkeys, and used laser microdissection to capture these neurons for RNA sequencing.

Results: At all three ages, CP and IP L3PNs had distinct transcriptomes, with the number of genes differentially expressed between neuronal subtypes increasing with age. For IP L3PNs, age-related shifts in gene expression were most prominent between prepubertal and postpubertal animals, whereas for CP L3PNs such shifts were most prominent between postpubertal and adult animals.

Conclusions: These findings demonstrate the presence of cell type-specific profiles and developmental trajectories of the transcriptomes of PPC-projecting IP and DLPFC-projecting CP L3PNs in monkey DLPFC. The evidence that IP L3PNs reach a mature transcriptome earlier than CP L3PNs suggests that these two subtypes differentially contribute to the maturation of working memory performance across late postnatal development and that they may be differentially vulnerable to the disease process of schizophrenia at specific stages of postnatal development.

Keywords: Laser Microdissection; Pyramidal Neuron; RNAseq; Retrograde Labeling; Schizophrenia; Working Memory.

Figure 1.. Transcriptome differences between CP and IP L3PNs over development.

In each heatmap, rows indicate individual genes and columns represent individual monkeys within each age group. A. Bold outline indicates a heat map of genes qualifying as DEG_FD between CP and IP L3PNs in prepubertal animals. The adjacent heat maps show the relative expression levels of these same genes in postpubertal and adult animals. B. Bold outline indicates a heat map of genes qualifying as DEG_FD between CP and IP L3PNs in postpubertal animals. The adjacent heat maps show the relative expression levels of those same genes in prepubertal and adult animals. C. Bold outline indicates a heat map of genes qualifying as DEG_FD between CP and IP PNs in adult animals. The adjacent heat maps show the relative expression levels of those same genes in prepubertal and postpubertal animals. In all 3 panels, the DEG_FD in the highlighted age group show a similar pattern of differential gene expression in the other two age groups. Abbreviations: CP, callosal projection; IP, ipsilateral projection; pre, prepubertal; post, postpubertal.

Figure 2.. Consistent differences in gene expression between CP and IP L3PNs over development.

Plots of log2 expression values for two voltage-gated sodium channels (SCN1A and SCN3A) and two synaptotagmins (SYT2 and SYT10) in CP (green) and IP (yellow) L3PNs from each age group. Hash bars indicate group means and filled circles indicate values for individual monkeys; lines connect the CP and IP samples from each monkey. Log2 expression values can be compared between L3PN subtypes within a given age group, but not across age groups due to the different sequencing batches. Abbreviations: CP, callosal projection; IP, ipsilateral projection; pre, prepubertal; post, postpubertal.

Figure 3.. Examples of genes enriched in IP or CP L3PNs at all three developmental time points.

Plots of log2 expression values for the structural gene NEFM and the neurite outgrowth gene HGF in CP (green) and IP (yellow) L3PNs from each age group. Hash bars indicate group means and filled circles indicate values for individual monkeys; lines connect the CP and IP samples from each monkey. Log2 expression values can be compared between L3PN subtypes within a given age group, but not across age groups due to the different sequencing batches. Abbreviations: CP, callosal projection; IP, ipsilateral projection; pre, prepubertal; post, postpubertal.

Figure 4.. Differences in gene expression within CP or IP L3PNs across development.

A. Bold outline indicates a heat map of genes whose expression levels in CP L3PNs differ across the 3 age groups studied. The heat map to the right shows the expression levels of those same genes in IP L3PNs. B. Plots of DEGs in CP L3PNs with progressively increasing (top) or declining (bottom) expression levels over late postnatal development. C. Bold outline indicates heat map of genes whose expression levels in IP L3PNs differ across the 3 age groups studied. The heat map to the left shows the expression levels of those same genes in CP L3PNs. D. Plots of DEGs in IP L3PNs with progressively increasing (left) or declining (right) expression levels over late postnatal development. E. Plots of representative genes whose expression levels change in CP and IP L3PNs over different time periods. In each plot, hash bars indicate group means and filled circles indicate values for individual monkeys; lines connecting the means are included to highlight the different developmental trajectories. Abbreviations: CP, callosal projection; IP, ipsilateral projection; pre, prepubertal; post, postpubertal.

Figure 5.. RRHO2 analyses of gene expression shifts over development.

A, B. RRHO2 plots of gene expression differences between prepubertal and postpubertal monkeys for CP and IP L3PNs (panel A) and between postpubertal and adult monkeys for CP and IP L3PNs (panel B). Number of concordant genes that increase (lower left quadrant) or decrease (upper right quadrant) in expression level with age are shown. C, D. Heat maps of the discordant genes from the prepubertal-to-postpubertal transition (panel C) and from the postpubertal-to-adult transition (panel D). Each column of the heat maps represents an individual animal for the indicated cell type and each row an individual gene. The numbers below each heat map are the average normalized expression level for all the discordant genes in that cell type and age group. Note that for the prepubertal-to-postpubertal transition the shift in gene expression is larger for IP than for CP L3PNs, whereas for the postpubertal-to-adult transition the shift in gene expression is larger for CP than for IP L3PNs. E. RRHO2 plots showing gene expression differences between the prepubertal and adult age groups compared to the prepubertal-to-postpubertal transition (left side) and the postpubertal-to-adult transition (right side) for CP L3PNs (top) and IP L3PNs (bottom). The earlier transition accounts for most of the overall difference with age for IP L3PNs, whereas the later transition accounts for more of the overall difference with age for CP L3PNs.