Reduction in GABAB on glia induce Alzheimer's disease related changes

Abstract

Alzheimer's Disease (AD) is a neurodegenerative disorder characterized by beta-amyloid plaques (Aβ), neurofibrillary tangles (NFT), and neuroinflammation. Data have demonstrated that neuroinflammation contributes to Aβ and NFT onset and progression, indicating inflammation and glial signaling is vital to understanding AD. A previous investigation demonstrated a significant decrease of the GABA_B receptor (GABA_BR) in APP/PS1 mice (Salazar et al., 2021). To determine if changes in GABA_BR restricted to glia serve a role in AD, we developed a mouse model with a reduction of GABA_BR restricted to macrophages, GAB/CX3ert. This model exhibits changes in gene expression and electrophysiological alterations similar to amyloid mouse models of AD. Crossing the GAB/CX3ert mouse with APP/PS1 resulted in significant increases in Aβ pathology. Our data demonstrates that decreased GABA_BR on macrophages leads to several changes observed in AD mouse models, as well as exacerbation of AD pathology when crossed with existing models. These data suggest a novel mechanism in AD pathogenesis.

Keywords: Alzheimer’s Disease; Amyloid; Electrophysiology; Flow cytometry; GABABR; Glia; Mouse models; NanoString.

Figure 1.. Flow cytometric verification of GABA_B1 receptor subunit knockdown and expression of YFP in males.

(A) Representation of gating strategy used for data analysis. (B) Percent of parent population (CD11b⁺/CD45⁺) that was positive for GABA_B1, GABA_B2, and total CD11b⁺/CD45⁺ cells. Significant differences in GABA_B1 and GABA_B2 receptor subunits between controls and GAB/CX3ert mice indicate recombination resulted in knockdown of the GABA_B receptor. (C) YFP is significantly expressed in the GAB/CX3ert mice indicating genetic recombination. Data analysis through IBM SPSS Statistics v24. All data are shown as mean (±SEM); *p<.05.

Figure 2.. Flow cytometric evaluation of GABA_B1 receptor subunit knockdown and expression of YFP in females.

(A) Representation of gating strategy used for data analysis. (B) Percent of parent population (CD11b⁺/CD45⁺) that was positive for GABA_B1, GABA_B2, and total CD11b⁺/CD45⁺ cells. No significant differences were observed in GABA_B1 and GABA_B2 receptor subunits between controls and GAB/CX3ert mice. (C) YFP is significantly expressed in the GAB/CX3ert mice indicating some genetic recombination in samples. Data analysis through IBM SPSS Statistics v24. All data are shown as mean (±SEM); *p<.05.

Figure 3.. GABA/CX3ert and APP/PS1 mice have fewer sharp-wave ripples and hippocampal-prefrontal hypersynchrony.

(A) Example raw and filtered (150 – 250 (Hz)) traces and spectrograms from C57Bl/6J (top), GABACX3rt (middle), and APP/PS1 (bottom) from raw and filtered hippocampal wires. APP/PS1 and GABA/CX3rt exhibit fewer sharp-wave ripples (SWRs) than C57Bl/6Jmice. (B - G) Comparison of SWR activity of C57Bl/6J, GABACX3rt, and APP/PS1 mice. (B) Frequency of SWR events; (C) Max frequency of hippocampal activity during SWR; (D) max power spectral density in the hippocampus during SWR; (E) duration of SWRs; (F) the maximum amplitude of hippocampal wires during SWR; (G) max power spectral density in the prefrontal cortex during SWR. Average values are shown for each of the five mice group mean + S.E.M; * p < 0.05. (H - L) Hippocampal - prefrontal Coherence. (H) Average hippocampal - prefrontal coherogram across groups. Note the color bar represents the minimum and maximum coherence values. (I) Delta coherence between the mPFC and HC is higher in APP/PS1 mice than C57Bl/6J and GABA/CX3rt mice. (J) Theta coherence between the mPFC and HC is not significantly different across groups. (K) Slow gamma coherence between the mPFC and HC is significantly higher in APP/PS1 and GABA/CX3rt mice compared to C57Bl/6J. (L) Fast gamma coherence between the mPFC and HC is not significantly different across groups. Average values are shown for each of the five mice group mean + S.E.M; * p < 0.05.

Figure 4:. Comparison of gene expression changes of the 770 transcripts on the NanoString panel between pairs of mouse models.

(A) GAB/CX3ert mice showed strong positive correlation (Pearson correlation coefficient = 0.64, p < 2.2 × 10⁻¹⁶) with APP/PS1 mice. (B) GAB/CX3ert x APP/PS1 mice showed strong positive correlation (Pearson correlation coefficient = 0.61, p < 2.2 × 10⁻¹⁶) with APP/PS1 mice. (C) GAB/CX3ert mice showed strong positive correlation (Pearson correlation coefficient = 0.66, p < 2.2 × 10⁻¹⁶) with GAB/CX3ert x APP/PS1 mice. In all panels the total App gene expression is labeled, which reports the sum of the mouse gene and human transgene expression in APP/PS1 carriers, to illustrate the effect of the transgenic construct.

Figure 5:. Correlation analysis between mouse models and 30 human co-expression modules using the NanoString Mouse AD panel.

(A) Our mouse models showed a significant positive correlation (p < 0.05) with immune related modules STGblue and IFGturquoise in Consensus Cluster B. Circles within a square correspond to significant (p < 0.05) positive (blue) and negative (red) Pearson correlation coefficients. Color intensity and size of the circles are proportional to the correlation. (B) Common genes exhibiting directional coherence for gene expression changes between immune related STGblue module and all mouse models. (C) Common genes exhibiting directional coherence for gene expression changes between immune related IFGturquoise module and all mouse models.

Figure 6.. Quantification of Aβ-40 and Aβ-42 in GAB/CX3ert x APP/PS1 mice compared to APP/PS1 mice.

(A) Significant increase in Aβ-40 and Aβ-42 was observed in male GAB/CX3ert mice compared to APP/PS1 mice. (B) No significant differences were observed in female GAB/CX3ert x APP/PS1 mice compared to female APP/PS1 mice. Data analysis through IBM SPSS Statistics v24. All data are shown as mean (±SEM); *p<.05.