Alterations in retrotransposition, synaptic connectivity, and myelination implicated by transcriptomic changes following maternal immune activation in non-human primates

Nicholas F. Page, Michael J. Gandal, Myka L. Estes, Scott Cameron, Jessie E. Buth, Sepideh Parhami, Gokul Ramaswami, Karl D. Murray, David G. Amaral, Judy Van de Water, Cynthia M. Schumann, Cameron S. Carter, Melissa D. Bauman, A. Kimberley McAllister, Daniel H. GeschwindbioRxiv 2020 · DOI

TL;DR

This study uses a non-human primate model to investigate how maternal immune activation (MIA) during pregnancy affects brain development. They found that MIA leads to widespread changes in gene expression, particularly impacting transposable elements, synaptic connectivity, and myelination, with the hippocampus showing particular vulnerability. These findings offer crucial insights into the molecular underpinnings of neurodevelopmental and psychiatric disorders.

What Problem Does This Paper Address?

Maternal immune activation (MIA) during pregnancy is a known risk factor for neurodevelopmental and psychiatric disorders like schizophrenia and autism. However, the precise molecular and neurobiological mechanisms by which MIA confers this risk remain poorly understood. This paper aims to bridge this knowledge gap by using a non-human primate (NHP) model to explore the brain-wide transcriptomic consequences of MIA, providing a more direct link to human brain development than rodent models.

Methods

The researchers employed RNA sequencing (RNA-seq) to analyze gene expression in four key brain regions (prefrontal cortex, anterior cingulate, hippocampus, and primary visual cortex) of adolescent male macaques. These macaques were exposed to a viral mimic (poly:ICLC) during gestation to simulate MIA, or received a saline control. They also utilized bioinformatic tools for differential gene expression analysis, co-expression network analysis (WGCNA), and assessment of transposable element (TE) activity.

Datasets

The study utilized RNA-seq data generated from the brains of adolescent male rhesus macaques (3.5-4 years old) exposed to either maternal immune activation (poly:ICLC) or a control (saline) during gestation. No external datasets were used for training or evaluation in this specific study.

Key Insights and Ideas

  • Hippocampal Vulnerability: The hippocampus appears to be particularly susceptible to MIA-induced transcriptomic changes, showing the largest number of differentially expressed genes.
  • Transposable Element Dysregulation: MIA significantly impacts the expression of transposable elements (TEs), which are mobile DNA sequences that can influence gene regulation and genome stability. This suggests a novel mechanism by which MIA might contribute to neurodevelopmental disorders.
  • Synaptic and Myelination Alterations: The study identified widespread changes in genes related to synaptic connectivity and myelination, processes critical for proper brain function and development.
  • PIWIL2 as a Key Regulator: The gene PIWIL2, a known regulator of DNA methylation and transposon silencing, was consistently downregulated across multiple brain regions and at the protein level, suggesting its central role in mediating MIA's effects.
  • Timing of MIA Matters, But Concordance Exists: While there were some differences in gene expression based on whether MIA occurred in the first or second trimester, overall transcriptomic changes were largely concordant, indicating a robust effect of MIA regardless of the specific timing.

Key Results

  • 266 Differentially Expressed Genes: A total of 266 unique genes were found to be differentially expressed in at least one brain region following MIA.
  • PIWIL2 and MGARP as Top DE Genes: PIWIL2 and MGARP were identified as the top downregulated genes, with PIWIL2 being a master regulator of piRNA-mediated DNA methylation and an inhibitor of retrotransposition.
  • Region-Specific Signatures: While some genes were affected across multiple regions, distinct regional patterns of gene expression changes were observed, with the hippocampus showing the most significant alterations.
  • TEs and Myelination Pathways Affected: Significant dysregulation was observed in pathways related to transposable elements, synaptic signaling, and myelination.

Limitations and Future Work

  • Age of Subjects: The study examined adolescent macaques (3.5-4 years old), which is younger than human subjects typically included in studies of schizophrenia and autism. This means the findings may represent prodromal changes rather than the full spectrum of adult disease. Future work should investigate older NHP offspring to see if these changes persist or evolve.
  • Transcriptomic vs. Protein Level: While RNA-seq provides valuable gene expression data, it doesn't directly reflect protein levels or functional outcomes. The authors acknowledge that transcriptomic approaches cannot uncover functional protein level changes.
  • Limited Sample Size: Working with non-human primates inherently involves limitations in sample size, which can affect statistical power.
  • Future Directions: The authors suggest further investigation into the role of PIWIL2 methylation and transposon regulation, as well as the progression of MIA-induced changes in myelination and synaptic connectivity throughout development. They also propose examining striatal and limbic regions, given the observed vulnerability of the hippocampus and its connection to repetitive behaviors.

Summary updated

Abstract

Abstract Background Maternal immune activation (MIA) is a proposed risk factor for multiple neurodevelopmental and psychiatric disorders, including schizophrenia. However, the molecular and neurobiological mechanisms through which MIA imparts risk for these disorders remain poorly understood. A recently developed nonhuman primate model of exposure to the viral mimic poly:ICLC during pregnancy shows abnormal social and repetitive behaviors and elevated striatal dopamine, a molecular hallmark of human psychosis, providing an unprecedented opportunity for mechanistic dissection. Methods We performed RNA-sequencing across four psychiatrically-relevant brain regions (prefrontal cortex, anterior cingulate, hippocampus, and primary visual cortex) from 3.5-4-year old male MIA-exposed and control offspring—an age comparable to mid adolescence in humans. Results We identify 266 unique genes differentially expressed (DE) in at least one brain region with the greatest number observed in hippocampus. Co-expression networks identified region-specific alterations in synaptic signaling and oligodendrocytes. Across regions, we observed temporal and regional differences, but transcriptomic changes were largely similar across 1 st or 2 nd trimester MIA exposures, including for the top DE genes—PIWIL2 and MGARP. In addition to PIWIL2, several other known regulators of retrotransposition, as well as endogenous transposable elements were dysregulated in MIA offspring. Conclusions Together, these results begin to elucidate the brain-level molecular mechanisms through which MIA may impart risk for psychiatric disease.
BibTeX
@misc{page2020alterations,
  author = {Nicholas F. Page and Michael J. Gandal and Myka L. Estes and Scott Cameron and Jessie E. Buth and Sepideh Parhami and Gokul Ramaswami and Karl D. Murray and David G. Amaral and Judy Van de Water and Cynthia M. Schumann and Cameron S. Carter and Melissa D. Bauman and A. Kimberley McAllister and Daniel H. Geschwind},
  title = {{Alterations in retrotransposition, synaptic connectivity, and myelination implicated by transcriptomic changes following maternal immune activation in non-human primates}},
  year = {2020},
  doi = {10.1101/2020.03.31.019190},
  howpublished = {bioRxiv},
}
Figure