Contribution of pannexin1 to experimental autoimmune encephalomyelitis

Abstract

Pannexin1 (Panx1) is a plasma membrane channel permeable to relatively large molecules, such as ATP. In the central nervous system (CNS) Panx1 is found in neurons and glia and in the immune system in macrophages and T-cells. We tested the hypothesis that Panx1-mediated ATP release contributes to expression of Experimental Autoimmune Encephalomyelitis (EAE), an animal model for multiple sclerosis, using wild-type (WT) and Panx1 knockout (KO) mice. Panx1 KO mice displayed a delayed onset of clinical signs of EAE and decreased mortality compared to WT mice, but developed as severe symptoms as the surviving WT mice. Spinal cord inflammatory lesions were also reduced in Panx1 KO EAE mice during acute disease. Additionally, pharmacologic inhibition of Panx1 channels with mefloquine (MFQ) reduced severity of acute and chronic EAE when administered before or after onset of clinical signs. ATP release and YoPro uptake were significantly increased in WT mice with EAE as compared to WT non-EAE and reduced in tissues of EAE Panx1 KO mice. Interestingly, we found that the P2X7 receptor was upregulated in the chronic phase of EAE in both WT and Panx1 KO spinal cords. Such increase in receptor expression is likely to counterbalance the decrease in ATP release recorded from Panx1 KO mice and thus contribute to the development of EAE symptoms in these mice. The present study shows that a Panx1 dependent mechanism (ATP release and/or inflammasome activation) contributes to disease progression, and that inhibition of Panx1 using pharmacology or gene disruption delays and attenuates clinical signs of EAE.

Figure 1. Blockade of Panx1 channels ameliorates signs of EAE.

Time course of the mean ± s.e.m. values of neurological scores recorded from rats (A, B) and mice (C, D) with EAE. (A) Daily intraperitoneal injections of 5 mg/kg mefloquine (MFQ) but not 1 mg/kg MFQ improved EAE outcome in rats. (B–D) Daily injections of MFQ (5 mg/kg) administered to (B) rats and (C) mice at 7 days post immunization (dpi) till, respectively 14 and 20 dpi, and to (D) mice starting at 14 dpi till 38 dpi is shown to ameliorate the EAE symptoms. P values were calculated from all data points using Mann-Whitney test. (E) Bar histograms showing the mean ± s.e.m. values of conduction latency in the corticospinal pathway of mice with EAE untreated (black bar) and treated (gray bar) with MFQ. In parentheses are the numbers of animals used. P values were obtained from unpaired t-test.

Figure 2. Delayed onset of EAE in Panx1 KO mice.

(A - right) Time course of clinical signs recorded from Panx1 wild type (WT) and Panx1 knockout (KO) mice immunized for MOG. (A-left) The graph depicts the initial acute phase of the disease (blue rectangle in A) showing that up to day 12 post-immunization, a significant difference in clinical scores was detected between the two genotypes. Symbols represent mean ± s.e.m. Twenty two Panx1 WT and 22 Panx1 KO female mice were immunized. Thirteen days post-immunization (dpi), 5 animals of each genotype were used for histopathology; 6 Panx1 WT and 2 Panx1 KO died or were euthanized due to the severity of EAE symptoms. *P<0.01, unpaired t-test. (B) Hematoxylin & eosin stained sections of spinal cord from Panx1 WT and Panx1 KO mice obtained at the acute phase of EAE (12–13 dpi). Ventral funiculus is outlined in white box and inflammatory lesions are outlined in cyan. Bar histograms on the right represent the mean ± s.e.m values of the percent lesion area in each genotype. Note that Panx1 KO EAE spinal cords exhibit significantly less infiltrating cells than spinal cords from WT mice. Five Panx1 WT and 5 Panx1 KO mice were used for histology; three spinal cord sections from sacral to thoracic areas were analyzed from each mouse. (C) Hematoxylin & eosin stained sections of spinal cord from Panx1 WT and Panx1 KO mice obtained at the chronic phase of EAE (35 dpi). Bar histograms on the right represent the mean ± s.e.m values of the percent lesion area in each genotype. At this stage of disease, Panx1 WT and Panx1 KO spinal cords exhibit similar extent of lesion areas. Five Panx1 WT and 5 Panx1 KO mice were used for histology. Three sections from sacral to thoracic spinal cords were analyzed from each mouse. Quantification of lesions is presented as percent area of ventral funiculus white matter occupied by inflammatory cells. P values were obtained using unpaired t-test.

Figure 3. Deficient IL-1β release from activated Panx1 KO splenic macrophages.

Bar histograms showing the mean ± s.e.m. values of interleukin-1β (IL-1β) measured from media bathing Panx1 wild type (WT) and Panx1 knockout (KO) macrophage cultures that were untreated, treated with lipopolysaccharide (LPS) and treated with LPS and stimulated with 5 mM ATP. Samples were obtained from 9 mice pooled into three groups and ELISA run in triplicates. ***P<0.001 (ANOVA followed by Newman-Keuls multiple comparison test).

Figure 4. Increased Panx1 channel activity in EAE.

(A) Bar histograms of the mean ± s.e.m. values of the relative intensity of YoPro uptake (test/control) recorded from Panx1 wild type (WT) and Panx1 knockout (KO) spinal cord slices of control and EAE mice. **P<0.005, ***P<0.001 (ANOVA followed by Newman-Keuls multiple comparison test). Nine Panx1 WT non-EAE, 9 Panx1 WT EAE, 8 Panx1 KO non-EAE, and 13 Panx1 KO EAE mice were used, with a minimum of 167 slices examined per group. Samples from EAE mice were then separated into mildly affected (flaccid tail, hind-limb weakness, clinical score range 0–3) and severely affected (hind-limb paralysis with possible trunk and forelimb involvement, clinical score range 4–7). (B) Bar histograms of the mean ± s.e.m. values of the relative intensity of propidium iodide (P.I.) uptake (test/control) recorded from Panx1 wild type (WT) and Panx1 knockout (KO) spinal cord slices of control and EAE mice. ***P<0.001, (ANOVA followed by Newman-Keuls post hoc comparison). Seven Panx1 WT non-EAE, 5 Panx1 WT EAE, 5 Panx1 KO non-EAE, and 7 Panx1 KO EAE mice were used, with a minimum of 71 slices per group examined. (C) Bar histograms of the mean ± s.e.m. values of ATP (pmol/mg protein) recorded from ACSF bathing Panx1 wild-type (WT) and Panx1 knockout (KO) spinal cord slices of control and EAE mice. In parentheses are the numbers of animals used. (D) Bar histograms of the mean ± s.e.m. values of Panx1 transcript (normalized to ribosomal 18S) measured from RNA samples extracted from cerebellum of naïve and EAE Panx1 WT mice. Samples are from 3 naïve and 3 EAE mice. ***P<0.001 and *P<0.05 (ANOVA followed by Newman-Keuls multiple comparison test); P value in part D was obtained using unpaired t-test analysis.

Figure 5. Increased P2X₇ receptor expression levels in spinal cords of mice with EAE.

(Top) Western blot showing bands corresponding to the P2X₇ receptor (69 and 75 kDa) and to GAPDH (37 kDa). Band at 50 kDa likely corresponds to nonspecific staining due to the simultaneous use of two antibodies. (Bottom) Bar histograms of the mean ± s.e.m values of P2X₇R/GAPDH obtained from western blots showing the increased P2X₇R expression in Panx1 WT and Panx1 KO mice with EAE compared to naïve animals. (***; &&&) P<0.01; t-test.