Chronic intermittent hypoxia exposure induces a unique microglial transcriptome in 5XFAD mice

Abstract

Clinical observations suggest that obstructive sleep apnea (OSA) and Alzheimer's disease (AD) pathology may be linked; however, causal mechanisms and relationships are unclear. To investigate the potential interaction between amyloidosis and intermittent hypoxia (IH), a hallmark of OSA, starting at 4-months of age 5XFAD mice were exposed to chronic IH (CIH) consisting of 20 episodes per hour of hypoxia for 12 hours/day, daily for 4- (males) or 6-months (females). CIH did not induce significant changes in amyloid burden or the number of astrocytes in males or females, but there was a slight decrease in the number of microglia observed in the cortex of 5XFAD mice of both sexes. To further explore this effect, we performed bulk RNA sequencing on isolated microglia. In WT mice, the most robust gene changes induced by CIH were identified in male microglia, many of which were pro-inflammatory. In microglia from 5XFAD mice, compared to NX, CIH exposure induced comparatively more DEGs in males. Further, in genes that were upregulated by CIH in WT vs 5XFAD mice of both sexes, there was an enrichment of pathways associated with oxidative phosphorylation, aerobic and cellular respiration, and ATP synthesis. These changes indicate that CIH has a more robust effect on the microglial transcriptome in 5XFAD mice than in WT mice, suggesting that the synergy between AD and OSA pathologies may be driven by metabolic changes in the microglial transcriptome. These observations are particularly interesting given the known sex differences in OSA and AD pathology in human disease.

Keywords: amyloid; intermittent hypoxia; microglia; neuroinflammation.

Figure 1

CIH exposure alters weight of WT and 5XFAD mice. Weight (in grams) of female, 10-month old (A) and male, 8-month old (B) mice at harvest. Two-way ANOVA with Tukey multiple comparisons * p<0.05, **p<0.01, ***p<0.001, ****p<0.0001

Figure 2

Plaque load and plaque morphology are unchanged with CIH exposure. Imaris surfaces generated from methoxy-X04 stained, confocal z-stack images of a high-power field in the cortex of NX or CIH treated 5XFAD female or male mice (A, F). Imaris quantification of the number of plaques per field and average plaque volume per image, each data point represents 6 images (B,C,G,H). Tiled images of normoxia or CIH treated 5XFAD female or male mice (D, I). ImageJ quantification of the percent area of the cortex that contains plaques (E,J; each point represents 3 sections per animal). There were no differences detected by unpaired t-test, p>0.05. Scale bar is 70 μm(A, F) and 1000 μm(D, I)

Figure 3

The astrocyte response to amyloid deposition is unchanged by CIH exposure. Sections were stained with anti-GFAP antibodies (astrocytes), TO-PRO-3 (nuclei) and methoxy-X04 (plaques) and then imaged by confocal microscopy (A, C). Images were analyzed in Imaris (Bitplane) by colocalization. Each point represents 3 images per animal. There was no significant effect of CIH exposure by two-way ANOVA with Tukey multiple comparisons; p>0.05. Scale bar is 70 μm

Figure 4

The microglial response is slightly depressed with CIH exposure. Representative images of cortical sections from NX and CIH WT and 5XFAD mice stained with anti-iba-1 (microglia), TO-PRO-3 (nuclei) and methoxy-X04 (plaques) and then imaged by confocal microscopy (A (female), B (male)). Images were analyzed in Imaris (Bitplane) by colocalization (C, E). Microglia quantity and percentage of microglia 30 μm in female (C, D) and male (E, F) WT and 5XFAD mice exposed to NX or CIH. Each data point represents 3 images per animal. No significant effect of CIH exposure by two-way ANOVA with Tukey multiple comparisons, p>0.05. Scale bar is 70 μm

Figure 5

CIH exposure drives few changes in gene expression by RNA sequencing in WT females but induces a unique gene program in 5XFAD females. RNA was isolated from CD11b+ cells immunomagnetically-isolated from the cortex of 10-month-old WT and 5XFAD female mice. Volcano plots with significant DEGs compared by genotype (A,D) and treatment (B,C) for all genes with FDR<0.05 and LFC>1. Venn diagram showing the number of overlapping and unique genes between the female WT and 5XFAD comparisons in NX and CIH treatment (E). Biological processes gene ontology (GO) analyses of those genes specific to the female WT NX vs 5XFAD NX comparison that are down- (F) and up-regulated (G). Biological processes GO analyses of those genes specific to the female WT CIH vs 5XFAD CIH comparison that are up-regulated (H).

Figure 6

CIH exposure in males drives significant differential expression by RNA sequencing. RNA was isolated from CD11b+ cells immunomagnetically-isolated from the cortex of 8-month-old WT and 5XFAD male mice. Volcano plots with significant DEGs compared by genotype (A,D) and treatment (B,C) for all genes FDR<0.05, LFC>1. Venn diagram showing the number of overlapping and unique genes between the male WT and 5XFAD comparisons in NX and CIH treatment (E). Biological processes gene ontology (GO) analyses of those genes specific to the male WT NX vs 5XFAD NX comparison that are down- (F) and up-regulated (G). Biological processes GO analyses of those genes specific to the male WT CIH vs 5XFAD CIH comparison that are up-regulated (H).