Conditional Deletion of EphA4 on Cx3cr1-Expressing Microglia Fails to Influence Histopathological Outcome and Blood Brain Barrier Disruption Following Brain Injury

Abstract

Erythropoietin-producing human hepatocellular receptors play a major role in central nervous system injury. Preclinical and clinical studies revealed the upregulation of erythropoietin-producing human hepatocellular A4 (EphA4) receptors in the brain after acute traumatic brain injury. We have previously reported that Cx3cr1-expressing cells in the peri-lesion show high levels of EphA4 after the induction of controlled cortical impact (CCI) injury in mice. Cx3cr1 is a fractalkine receptor expressed on both resident microglia and peripheral-derived macrophages. The current study aimed to determine the role of microglial-specific EphA4 in CCI-induced damage. We used Cx3cr1 ^CreER/+ knock-in/knock-out mice, which express EYFP in Cx3cr1-positive cells to establish microglia, EphA4-deficient mice following 1-month tamoxifen injection. Consistent with our previous findings, induction of CCI in wild-type (WT) Cx3cr1 ^CreER/+ EphA4 ^+/+ mice increased EphA4 expression on EYFP-positive cells in the peri-lesion. To distinguish between peripheral-derived macrophages and resident microglia, we exploited GFP bone marrow-chimeric mice and found that CCI injury increased EphA4 expression in microglia (TMEM119+GFP-) using immunohistochemistry. Using Cx3cr1 ^CreER/+ EphA4 ^f/f (KO) mice, we observed that the EphA4 mRNA transcript was undetected in microglia but remained present in whole blood when compared to WT. Finally, we found no difference in lesion volume or blood-brain barrier (BBB) disruption between WT and KO mice at 3 dpi. Our data demonstrate a nonessential role of microglial EphA4 in the acute histopathological outcome in response to CCI.

Keywords: Eph signaling; TMEM119; innate immune; neuroinflammation; peripheral monocytes; traumatic brain injury.

Figure 1

EphA4 is upregulated on Cx3cr1-expressing cells in the peri-lesion 3 days post CCI injury. (A–D) Representative confocal images at max z-projection for immunohistochemical analysis for EphA4 (red), EYFP in Cx3cr1-expressing cells (green), and DAPI (blue) at 3 dpi in the ipsilateral cortex of Cx3cr1^CreER/+mice. CCI injury increased EphA4 expression in cx3cr1⁺ cells (arrows) in the peri-lesion areas. (E1–E3) Co-localization of EphA4 and EYFP in amoeboid-shaped microglia and/or PDM. Scale bar: (A) 100 μm; (B–D) 50 μm and (E1–E3) 20 μm.

Figure 2

Microglial EphA4 is upregulated in the damaged cortex after CCI injury. (A) Timeline for generation of GFP bone marrow chimeric mice and CCI injury. Mice were euthanized 3 dpi, and brain sections were used for IHC analysis of TMEM119 and EphA4. (B) Representative confocal images for peripheral-derived immune cells (GFP), infiltrating the ipsilateral injured cortex. (C–G) Representative images of TMEM119 (purple) and EphA4 (red). EphA4 is expressed on amoeboid-resident microglia (white arrow, EphA4⁺/GFP⁻/TMEM119⁺) as well as peripheral-derived GFP immune cells, which also express TMEM119 (yellow arrowhead, EphA4⁺/GFP⁺/TMEM119⁺). (C1–C4) Representative images showing the colocalization of TMEM119 (white) and EphA4 (red) in GFP⁻ microglia. Scale bar in (B) 1,000 μm; (C1–C4) 10 μm.

Figure 3

Generation of microglia-specific EphA4-deficient mice. (A) Relative EphA4 expression in cortical microglia of naive Cx3cr1^CreER/+EphA4 ^+/+ and Cx3cr1^CreER/+EphA4 ^f/f was measured at 2 weeks and 1-month post tamoxifen (TAM) injection using qRT-PCR. (B) PCR amplification of EphA4⁻WT allele (286 bp), loxP-flanked EphA4 allele (390 bp), EphA4-excised allele (250 bp), and Cre (100 bp) in the DNA extracted from Cx3cr1^CreER/+EphA4^+/+ and Cx3cr1^CreER/+EphA4^f/f microglia at 2 weeks and at 1-month post-tamoxifen injection. Positive (+) control showing loxP-flanked EphA4 allele (390 bp), EphA4-excised allele (250 bp), and Cre (100 bp). M = 100 bp low-scale DNA ladder. (C-G)qRT-PCR analysis for microglial genes (Cx3cr1, C and TMEM119, D), neuronal gene (NeuN, E), astrocyte gene (GFAP, F), and vascular endothelial gene (VE-cadherin, G) in isolated microglia, non-adherent remaining cortical cells, as well as whole blood collected from naïve Cx3cr1^CreER/+EphA4^+/+ and Cx3cr1^CreER/+EphA4^f/f mice. N = 3–6, ^*p < 0.05, significant difference between the designated groups; nd, not detected.

Figure 4

EphA4 expression in microglia at 2 h post-CCI injury. (A) Cx3cr1^CreER/creEREphA4 ^+/+, Cx3cr1^CreER/+EphA4 ^+/+, and Cx3cr1^CreER/+EphA4 ^f/f mice were intraperitoneally injected with five daily doses of tamoxifen (100 mg/kg) and then subjected to CCI injury 30 days after the last injection. Microglia were isolated from ipsilateral and contralateral cortices for qRT-PCR analysis at 2 h. (B) Relative EphA4 mRNA expression was measured in isolated microglia and non-adherent cells from ipsilateral (ipsi) and contralateral (contra) cortices (Ctx) at 2 h. N = 5, ^*p < 0.05, significant difference between the designated groups; nd, not detected. (C–E) Representative confocal images for IHC analysis of TMEM119 (white) and EphA4 (red) in GFP bone marrow chimeric WT mice, showing increased EphA4 expression in resident microglia (white arrowhead, EphA4⁺GFP⁻TMEM119⁺). Scale bar = 100 μm.

Figure 5

Conditional deletion of microglial EphA4 does not affect acute tissue damage or BBB disruption, following CCI injury. (A) Lesion volume (mm³) 3 dpi was estimated in the ipsilateral cortex of five Nisslstained serial coronal sections of each brain using Cavalieri Estimator from StereoInvestigator. (B–E) Representative mosaic images for the ipsilateral cortices of Nissl-stained sections taken at 4× magnification. (F) Immunoglobulin G (IgG) deposition volume was measured at 3 dpi in the ipsilateral cortex of five serial coronal sections of each brain using Cavalieri Estimator from StereoInvestigator. (G–J) Representative confocal images for the ipsilateral cortices of anti-mouse IgG-stained sections (red) taken at 4× magnification. N = 5–15, ^*p < 0.05, significant difference between the designated groups. ns, nonsignificant difference was observed.

Figure 6

Conditional deletion of microglial EphA4 does not influence microglia activation or ramification, following CCI injury. (A, B) Representative confocal images at max z-projection for immunostaining of CCR2 (red) and CD68 (white) in the cortical lesion and peri-lesion of Cx3cr1^CreER/+EphA4 ^+/+ and Cx3cr1^CreER/+EphA4 ^f/f mice at 3 dpi. (C, F) Co-localization of CCR2 (red) with Cx3cr1-EYFP (green), showing EYFP⁺ CCR2⁻ resident microglia in the peri-lesion. (D, G) Microglia ramification in the peri-lesion. EYFP⁺ CCR2⁻ microglia are either ramified (orange arrowhead), hypertrophic (yellow arrowhead), or amoeboid (cyan arrowhead). (E, H) Representative confocal images, showing CD68 expression (purple) in activated microglia (CD68⁺/Cx3cr1⁺/CCR2⁻, cyan arrow). (I) Quantification of the percentage of resting ramified, activated hypertrophic and activated amoeboid microglia (Cx3cr1⁺/CCR2⁻) in the peri-lesion. (J) Quantification of the percentage of CD68^hi microglia (CD68^hi/Cx3cr1⁺/CCR2⁻) in the peri-lesion. N = 4–5; ns, nonsignificant difference was observed between different groups.