Microglial TREM2 promotes phagocytic clearance of damaged neurons after status epilepticus

Abstract

In the central nervous system, triggering receptor expressed on myeloid cells 2 (TREM2) is exclusively expressed by microglia and is critical for microglial proliferation, migration, and phagocytosis. Microglial TREM2 plays an important role in neurodegenerative diseases, such as Alzheimer's disease and amyotrophic lateral sclerosis. However, little is known about how TREM2 affects microglial function within epileptogenesis. To investigate this, we utilized male TREM2 knockout (KO) mice within the intra-amygdala kainic acid seizure model. Electroencephalographic analysis, immunocytochemistry, and RNA sequencing revealed that TREM2 deficiency significantly promoted seizure-induced pathology. We found that TREM2 KO increased both the severity of acute status epilepticus and the number of spontaneous recurrent seizures characteristic of chronic focal epilepsy. Phagocytic clearance of damaged neurons by microglia was also impaired by TREM2 KO and reduced phagocytic activity correlated with increased spontaneous seizures. Analysis of human tissue from patients who underwent surgical resection for drug resistant temporal lobe epilepsy also showed a negative correlation between expression of the microglial phagocytic marker CD68 and focal to bilateral tonic-clonic generalized seizure history. These results indicate that microglial TREM2 and phagocytic activity are important to epileptogenic pathology.

Keywords: CD68; Epilepsy; Microglia; Phagocytosis; Seizures; TREM2.

Fig. 1:. TREM2 alters EEG profiles of mice undergoing status epilepticus.

(A) No difference was observed between control (CX3CR1^GFP/+) and TREM2 KO (TREM2^−/−:CX3CR1^GFP/+) in regards to when focal to bilateral tonic clonic (FBTC) events (Racine Score 4+) first occur. (B) TREM2 KO mice displayed a higher incidence of 5+ T-C events than controls. (C) The average number of T-C events was higher in TREM2 KO than controls. (D) TREM2 KO and control displayed similar post-KA survival. (E) Representative EEG profiles during status epilepticus. Green area = pre-ictal period, Red area = ictal period, Red arrow indicates beginning of ictal period, Green arrows indicate FBTC events. (F) Average FBTC event duration was similar between TREM2 KO and controls. EEG profiles for each animal were then investigated for differences in individual frequency band power spectral density (PSD). Delta: 1–3 Hz; Theta: 3–9 Hz; Alpha: 8–12 Hz; Sigma: 12–15 Hz; Beta: 15–30 Hz; Low Gamma: 30–55 Hz; High Gamma: 65–110 Hz. (G) PSD evaluation for the pre-ictal period. (H) PSD evaluation for the interictal periods. (I) PSD evaluation for the aggregated FBTC events. Sample sizes and analysis: (A-D) All control and TREM2 KO animals treated with KA during the course of this study underwent initial seizure characterization, N = 234 and 170 animals respectively. Each point represents an individual animal. (F-I) N = 18 and 21 animals for control and TREM2 KO respectively. (F) Each point represents an individual animal. Unpaired T-test with Welch correction was used for all comparisons. All data presented as Mean ± S.E.M. * P < 0.05, ** P < 0.01, *** P < 0.001.

Fig. 2:. TREM2 affects microglial proliferation and morphology following IA-KA.

(A) Representative images of the general CA3 region of control (CX3CR1^GFP/+) and TREM2 KO (TREM2^−/−:CX3CR1^GFP/+) animals at baseline and 3, 7, 14 days post-KA administration. Scale bar = 400 μM. (B) Quantification of microglia density within the general CA3 region (Within the yellow dashed line). (C) Experimental timeline of BrdU administration and sample collection. Animals were administered BrdU twice daily on days 0–2, or days 4–6 post IA-KA. (D) Representative images of BrdU staining. Scale bar = 200 μm. (E) Quantification of BrdU⁺:IBA1⁺ cells divided by total IBA1⁺ cells. (F) Representative images of the CA3 region at baseline and 3, 7, 14 days post-KA administration. Scale = 200 bar μm. (G) Quantification of CA3 region microglial soma size within the image field. (H) Representative images from tissues collected 7 days post-KA. Scale bar = 100 μm. (I) Quantification of the number of microglia displaying either ramified/bushy (small soma, long branched processes) or amoeboid (large soma, short processes) morphology in both control and TREM2 KO tissues. Sample sizes and analysis: N = 5–10 animals per group at each time point. Each point represents an individual animal. The quantification of 2–3 non-consecutive tissue sections were averaged for each N. Unpaired T-test with Welch correction was used for all comparisons. All data presented as Mean ± S.E.M. * P > 0.05, *** P > 0.001, **** P > 0.0001.

Fig. 3:. TREM2 KO affects neuronal atrophy following IA-KA administration.

(A) Representative CA3 region images of NeuN staining and GFP:NeuN co-localization within control (CX3CR1^GFP/+) and TREM2 KO (TREM2^−/−:CX3CR1^GFP/+) animals. Scale bar = 200 μm. (B) Positively stained NeuN area divided by total area (between yellow dashed lines). (C) GFP:NeuN co-localization divided by microglia number within the image field. (D) Representative images from 7 days post-KA. Scale bar = 100 μm. There are indications of microglia engulfing neurons following IA-KA in controls, but not TREM2 KO. Magnification, White arrows. (E) Representative images of CD68 staining within the CA3 region. Scale bar = 200 μm. (F) GFP and CD68 staining signal co-localization. Sample sizes and analysis: N = 5–10 animals per group at each time point. Each point represents an individual animal. The quantification of 2–3 non-consecutive tissue sections was averaged for each N. Unpaired T-test with Welch correction was used for each comparison. All data presented as Mean ± S.E.M. * P < 0.05, ** P < 0.01, **** P < 0.0001.

Fig. 4:. TREM2 KO affects apoptotic neuron phagocytosis in vitro.

(A) Representative images of primary microglia incubated with apoptotic N2a cells over time. Scale bar = 200 μm. (B) pHrodo red positive microglia relative to total microglia number. (C) Average pHrodo red object integrated intensity normalized to time 0 h. Sample sizes and analysis: N = 11–12 replicates from 2 independent experiments. The quantification of 4 image fields were averaged for each N. 2-way ANOVA followed by post-hoc multiple comparisons was used for each comparison. All data presented as Mean ± S.E.M. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001.

Fig. 5:. TREM2 KO affects microglial response to apoptotic neurons.

(A) Representative images of PSVue staining and GFP:PSVue co-localization within the CA3 region of control (CX3CR1^GFP/+) and TREM2 KO (TREM2^−/−:CX3CR1^GFP/+) animals. Scale bar = 200 μm. (B) PSVue⁺ cells divided by total area (between yellow dashed lines). (C) GFP:PSVue co-localization area divided by microglia number. (D) Representative Z direction images of microglia interacting with PSVue⁺ cells. Scale bar = 6 μm. (E) Representative images and intensity plots for various levels of microglial interaction along the outer edge of PSVue⁺ cells. (F) Binned 3 days post-KA cellular contact measurements. (G) Average contact percentage at 3 days post-KA. (H) Binned 7 days post-KA cellular contact measurements. (I) Average contact percentage at 7 days post-KA. Sample sizes and analysis: (B, C) N = 4–5 animals per group at each time point. Each point represents an individual animal. The quantification of 2–3 non-consecutive tissue sections was averaged for each N. (G, I) N = 4–5 animals for each group. Each solid point represents an individual animal. Contact measurements from 10–50 cells per animal were averaged for each N. Unpaired T-test with Welch correction was used for all comparisons. All data presented as Mean ± S.E.M. * P < 0.05, ** P < 0.01.

Fig. 6:. Comparison of TREM2 KO and control gene expression profiles.

(A) Volcano plot illustrating differentially expressed genes (DEGs) with a log₂ ≥ 1. (B) Network analysis of identified DEGs with expression log₂ ≥ 1 higher in TREM2 KO (TREM2^−/−:CX3CR1^GFP/+) or control (CX3CR1^GFP/+). (C) Biological process gene ontology (GO) of identified DEGs. (D) Heat map of identified cell cycle related DEGs at both baseline and 7 days post-KA. (E) Heat map of homeostatic microglial genes. (F) Heat map of disease associated microglial genes. Comparisons in gene expression for (G) Trem2, (H) Tyrobp (DAP12) and (I) Cd68. FPKM = Fragments per kilobase of transcript per million fragments mapped. Sample sizes and analysis: N = 4–5 animals per group at each time point. (G-I) Each point represents an individual animal. Analysis and comparison between gene sets was carried out via DESeq2. (E, F) Asterisks indicate statistically significate differences in expression between TREM2 KO and control at 7 days post-KA. For all comparisons significance is expressed by Q value. * Q < 0.05, ** Q < 0.01, *** Q < 0.001, **** Q < 0.0001.

Fig. 7:. TREM2 KO increases spontaneous recurrent seizure frequency.

(A) Representative EEG traces captured during spontaneous recurrent seizure (SRS) events from control (CX3CR1^GFP/+) and TREM2 KO (TREM2^−/−:CX3CR1^GFP/+) animals. (B) Fraction of animals that presented SRS versus the total number of animals enrolled. (C) Percentage of animals presenting SRS activity displaying +2 SRS events during the recording period (Day 7–11 post-KA). (D) SRS frequency. (E) Average SRS event duration. (F) No difference was observed as to when SRS events occurred during the housing day/night cycle. (G) PSD evaluations of 3 consecutive hours of non-FBTC event containing EEG recordings collected 7 days post-KA. Sample sizes and analysis: (B, C) N = 15 and 22 animals for control and TREM2 KO respectively. (D-G) N = 7 and 16 animals for control and TREM2 KO respectively. Each point represents an individual animal. Unpaired T-test with Welch correction was used for all comparisons. All data presented as Mean ± S.E.M. * P < 0.05, ** P < 0.01.

Fig. 8:. CD68 expression within human epilepsy derived tissues sections.

(A) Representative images of hippocampal CA1, CA4, dentate gyrus (DG) and proximal white matter (WM) and temporal lobe grey matter (GM) and WM. Scale bar = 200 μm. (B) Analysis of Iba1⁺ cell number in the hippocampus. (C) CD68 staining area per Iba1⁺ cell in the hippocampus. (D) Iba1⁺ cell number in the temporal lobe. (E) CD68 staining area per Iba1⁺ cell in the temporal lobe. Sorting results by generalized seizure history, no difference was found in Iba1⁺ cell number in temporal lobe (F) GM and (G) WM. (H) Temporal lobe WM CD68 expression was found to be higher in patients with no known history of generalized seizures. (I) Plotting the sorted data in H against severity score (0 = 0, 1 = ≤1, 2 = 2–4, 3 = 5–7, 4 = 8–12, 5 = 13≤ monthly focal seizures) revealed that patients without generalized seizures had higher CD68 expression at every severity level. Sample sizes and analysis: N ≤ 10 patients per group. 31 patients in total were assessed. Each point represents an individual patient. The quantification of 2–5 image fields per region was averaged for each N. (B, C) 1-way ANOVA was used for each comparison. (F-H) Unpaired T-test with Welch correction was used for each comparison. (I) Difference is based on linear regression elevation. All data presented as Mean ± S.E.M, except I. * P < 0.05, ** P < 0.01, *** P < 0.001, **** P < 0.0001.