Prenatal allergic inflammation in rats programs the developmental trajectory of dendritic spine patterning in brain regions associated with cognitive and social behavior

Abstract

Allergic inflammation during pregnancy increases risk for a diagnosis of neurodevelopmental disorders such as Attention Deficit/Hyperactivity Disorder (ADHD) and Autism Spectrum Disorder (ASD) in the offspring. Previously, we found a model of such inflammation, allergy-induced maternal immune activation (MIA), produced symptoms analogous to those associated with neurodevelopmental disorders in rats, including reduced juvenile play behavior, hyperactivity, and cognitive inflexibility. These behaviors were preceded by perinatal changes in microglia colonization and phenotype in multiple relevant brain regions. Given the role that microglia play in synaptic patterning as well as evidence for altered synaptic architecture in neurodevelopmental disorders, we investigated whether allergic MIA altered the dynamics of dendritic spine patterning throughout key regions of the rat forebrain across neurodevelopment. Adult virgin female rats were sensitized to the allergen, ovalbumin, with alum adjuvant, bred, and allergically challenged on gestational day 15. Brain tissue was collected from male and female offspring on postnatal days (P) 5, 15, 30, and 100-120 and processed for Golgi-Cox staining. Mean dendritic spine density was calculated for neurons in brain regions associated with cognition and social behavior, including the medial prefrontal cortex (mPFC), basal ganglia, septum, nucleus accumbens (NAc), and amygdala. Allergic MIA reduced dendritic spine density in the neonatal (P5) and juvenile (P15) mPFC, but these mPFC spine deficits were normalized by P30. Allergic inflammation reduced spine density in the septum of juvenile (P30) rats, with an interaction suggesting increased density in males and reduced density in females. MIA-induced reductions in spine density were also found in the female basal ganglia at P15, as well as in the NAc at P30. Conversely, MIA-induced increases were found in the NAc in adulthood. While amygdala dendritic spine density was generally unaffected throughout development, MIA reduced density in both medial and basolateral subregions in adult offspring. Correlational analyses revealed disruption to amygdala-related networks in the neonatal animals and cortico-striatal related networks in juvenile and adult animals in a sex-specific manner. Collectively, these data suggest that communication within and between these cognitive and social brain regions may be altered dynamically throughout development after prenatal exposure to allergic inflammation. They also provide a basis for future intervention studies targeted at rescuing spine and behavior changes via immunomodulatory treatments.

Keywords: Allergy; Dendritic Spines; Development; Inflammation; Maternal Immune Activation; Microglia; Sex Differences.

Figure 1:

Graphical depiction of the experimental timeline. Offspring were euthanized on P5, P15, P30 and P100-120 and brain tissue was processed for Golgi-Cox histology.

Figure 2:

mPFC spine densities throughout development. (A) Allergic MIA reduces developmental dendritic spine density in the mPFC that are normalized by P30. Data are normalized to the female vehicle group. (B) Representative images of dendritic spines in mPFC of P15 animals, taken at 100x. Scale bar = 10 μm. M = male, F = female.

Figure 3:

Basal ganglia spine densities throughout development. (A) Males have higher spine density compared to females in the basal ganglia at P5 and P15, and allergic MIA reduces basal ganglia spine density at P15, particularly in females. Data are normalized to the female vehicle group. (B) Representative images of dendritic spines in basal ganglia at P15, taken at 100x. Scale bar = 10 μm. M = male, F = female.

Figure 4:

Septum spine densities throughout development. (A) Allergic MIA reduced septal spine density at P30. Significant interactions were found at both P15 and P30. Though post-hoc tests were not significant, it appears allergic MIA may reduce spine density in males and increase it in females at P15, while the opposite may be true for P30. Males also had higher dendritic spine density than females at P30. Data are normalized to the female vehicle group.

Figure 5:

Nucleus accumbens spine densities throughout development. (A) Allergic MIA did not produce any effects on NAc spine density at early timepoints, however, (B & C) MIA reduced spine density at P30 and increased it at P120 in both core and shell subregions. Note that core versus shell subregions were not easily resolved at P5 or P15, so subregion-based analyses were limited to P30 and P120 when these regions were more easily distinguished. Data are normalized to the female vehicle group. (D & E) Representative images of dendritic spines at P30 and 120, respectively. Images taken at 100x. Scale bar = 10 μm. M = male, F = female.

Figure 6:

Amygdala spine densities throughout development. (A) Allergic MIA did not affect spine density in the amygdala at P5 or P15. (B & C) Both MeA and BLA showed MIA-induced spine deficits at P120. Note that subregions were not easily resolved at P5 or P15, so subregion-based analyses were limited to P30 and P120 when these regions were more easily distinguished. Data are normalized to the male vehicle group. (D) Representative images of dendritic spines in the amygdala at 120, taken at 100x. Scale bar = 10 μm. M = male, F = female.

Figure 7:

Spine density correlations at P5. (A & C) At baseline, all relationships were strongly positively correlated. (A & B) In males, allergic MIA strengthens the significance of the correlations amongst most region pairs. (C & D) In females, allergic MIA reverses the positive correlations of the amygdala with the PFC, basal ganglia, septum, and nucleus accumbens. * = p ≤ 0.005 (statistically significant following Bonferroni adjustment). mPFC = medial prefrontal cortex, BG = basal ganglia, NAc = nucleus accumbens, AMY = amygdala.

Figure 8:

Spine density correlations at P15. (A & C) At baseline, spine density correlated amongst all regions examined in vehicle females, while many striatal relationships exhibited negative correlations in males. (B & D) Allergic MIA may reverse or reduce striatal relationships in P15 males. 5. * = p ≤ 0.005 (statistically significant following Bonferroni adjustment). mPFC = medial prefrontal cortex, BG = basal ganglia, NAc = nucleus accumbens, AMY = amygdala.

Figure 9:

Spine density correlations at P30. (A & C) At baseline, males and females exhibited different patterns in spine density correlations, particularly regarding comparisons with the mPFC. (B & D) Allergic MIA may reverse corticostriatal correlational relationships found in both males and females. * = p ≤ 0.002 (statistically significant following Bonferroni adjustment). PFC = prefrontal cortex, BG = basal ganglia, NAcc = nucleus accumbens core, NAcs = nucleus accumbens shell, MeA = medial amygdala, BLA = basolateral amygdala.

Figure 10:

Spine density correlations at P120. (A) At baseline, spine density positively correlated for two comparisons in vehicle males prior to adjustment the septum and accumbens shell and the mPFC and basolateral amygdala. (C). Conversely, strong negative correlations were found in females between for the PFC and basal ganglia and the septum and medial amygdala pairs prior to adjustment. (B & D) MIA eliminated the trends seen in vehicle animals and produced a strong significant negative correlation between mPFC and septum in males, and a strong positive correlation between the basal ganglia and medial amygdala in females. * = p ≤ 0.002 (statistically significant following Bonferroni adjustment). PFC = prefrontal cortex, BG = basal ganglia, NAcc = nucleus accumbens core, NAcs = nucleus accumbens shell, MeA = medial amygdala, BLA = basolateral amygdala.