27689
Sex-Specific Fetal Programming Impact of Testosterone on Later Developing Intrinsic Functional Connectivity between Social Brain Default Mode Subsystems
Objectives: As a model for sex-specific mechanisms involved in autism, we investigate how fetal testosterone (FT) biases neural circuitry development in a sex-specific manner. We also examined the question of how androgens directly affect gene expression in these circuits during early prenatal periods of brain development.
Methods: Age-matched neurotypical male and female adolescents (n=32 per group, 13.22-17.18 years) were scanned with resting state fMRI (rsfMRI) on a multi-echo EPI sequence at 3T. FT was measured with radioimmunoassays on amniotic fluid samples collected during midgestation. rsfMRI data were preprocessed using multi-echo independent components analysis (ME-ICA). Functional connectivity between neural circuits was measured utilizing group-ICA and dual regression as implemented in FSL. Between-component connectivity was estimated with partial correlations (ridge regression) using FSLNets. To gain insight into how androgens affect gene expression in a model of human prenatal neurodevelopment, we further examined how gene expression changes in human neural stem cells (hNSC) after treatment with a potent androgen, dihydrotestosterone (DHT) (a re-analysis of data from Quartier et al., in revision, Biol Psychiatry). We then tested for enrichment between DHT-dysregulated genes and genes that show spatial patterns of gene expression that are similar to the default mode network. Spatial gene expression similarity is measured using the Neurosynth Gene Expression Decoding feature on the Allen Institute Human Brain Gene Expression Atlas data.
Results: Variation in fetal testosterone (FT) exerts sex-specific effects on later adolescent functional connectivity between social brain default mode network (DMN) subsystems. Increasing FT is associated with later dampening of functional connectivity between DMN subsystems in males (r = -0.69), but has no effect in females (r = 0.02). This difference between sexes was highly significant (z = 3.35, p = 7.88e-4). No other circuits were identified at FDR q<0.05 as affected by FT in a sex-specific manner. Within hNSCs, DHT dysregulates genes involved in early neurodevelopmental biological processes such as neurogenesis, cell differentiation, regionalization, and pattern specification. Gene sets with a high degree of spatial expression similarity in the adult brain to cortical midline DMN circuitry are also highly enriched in DHT-dysregulated genes (OR = 1.88, p = 0.000002). These DMN-related and DHT-dysregulated genes are highly enriched (fold enrichment > 5) in synaptic biological processes.
Conclusions: This work highlights sex-specific prenatal androgen influence on human social brain DMN circuitry and is of high-relevance to male-biased conditions affecting social brain development such as autism. Prenatal androgens may act as male-specific influences on DMN cortical circuit formation and maintenance by affecting genes involved in both early neurodevelopmental processes (e.g., neurogenesis, cell differentiation) and later synaptic processes.
See more of: Brain Function (fMRI, fcMRI, MRS, EEG, ERP, MEG)