CXCR2 signaling protects oligodendrocytes and restricts demyelination in a mouse model of viral-induced demyelination

Abstract

Background: The functional role of ELR-positive CXC chemokines during viral-induced demyelination was assessed. Inoculation of the neuroattenuated JHM strain of mouse hepatitis virus (JHMV) into the CNS of susceptible mice results in an acute encephalomyelitis that evolves into a chronic demyelinating disease, modeling white matter pathology observed in the human demyelinating disease Multiple Sclerosis.

Methodology/principal findings: JHMV infection induced the rapid and sustained expression of transcripts specific for the ELR+ chemokine ligands CXCL1 and CXCL2, as well as their binding receptor CXCR2, which was enriched within the spinal cord during chronic infection. Inhibiting CXCR2 signaling with neutralizing antiserum significantly (p<0.03) delayed clinical recovery. Moreover, CXCR2 neutralization was associated with an increase in the severity of demyelination that was independent of viral recrudescence or modulation of neuroinflammation. Rather, blocking CXCR2 was associated with increased numbers of apoptotic cells primarily within white matter tracts, suggesting that oligodendrocytes were affected. JHMV infection of enriched oligodendrocyte progenitor cell (OPC) cultures revealed that apoptosis was associated with elevated expression of cleaved caspase 3 and muted Bcl-2 expression. Inclusion of CXCL1 within JHMV infected cultures restricted caspase 3 cleavage and increased Bcl-2 expression that was associated with a significant (p<0.001) decrease in apoptosis. CXCR2 deficient oligodendrocytes were refractory to CXCL1 mediated protection from JHMV-induced apoptosis, readily activating caspase 3 and down regulating Bcl-2.

Conclusion/significance: These findings highlight a previously unappreciated role for CXCR2 signaling in protecting oligodendrocyte lineage cells from apoptosis during inflammatory demyelination initiated by viral infection of the CNS.

Figure 1. CXCR2 and ELR+ chemokines are upregulated within the spinal cord during chronic JHMV infection.

C57BL/6 mice were i.c. infected with 500 pfu JHMV and spinal cords removed at defined times p.i. to assess viral burden (A) and transcripts specific for CXCR2, CXCL1, CXCL2, and CXCL5 using semi-quantitative real time PCR (B). For PCR results, data represents the fold induction relative to sham infected mice (n = 3–5, *p<0.05, **p<0.01, ***p<0.001). Representative immunofluorescence of the spinal cord white matter at 15 days p.i reveals CXCL1 expression within reactive hypertrophic GFAP+ astrocytes (C).

Figure 2. CXCR2 neutralization mutes spontaneous recovery from JHMV induced demyelination.

C57BL/6 mice were i.c. infected with 500 pfu JHMV and treated with neutralizing CXCR2 antiserum or control normal rabbit serum (NRS) i.p. from day 12 to 20, indicated by the shaded box. Treated mice were allowed to recover for a further two weeks and clinical severity was assessed (A). In addition, brains and spinal cords from treated mice were removed at the beginning, middle, and end of treatment to assess JHMV viral burden (B). For panel A data is a summation of four independent experiments (anti-CXCR2 n = 32; NRS n = 31). For panel B data is a summation of two independent experiments.

Figure 3. CXCR2 neutralization does not alter neuroinflammation during chronic JHMV infection.

C57BL/6 mice were i.c. infected with 500 pfu JHMV and treated with neutralizing CXCR2 antiserum or control normal rabbit serum (NRS) i.p. from day 12 to 20. Brains (A–D) and/or spinal cords (E) were removed from perfused mice at defined times p.i. and were assessed for CD4+ (A), CD8+ T lymphocyte (B), macrophage (C), and neutrophil (D & E) accumulation within the CNS. Data in panels A–C are representative of four independent experiments, and data in panels D & E are representative of two independent experiments.

Figure 4. CXCR2 neutralization does not diminish virus – specific T cell infiltration during chronic JHMV infection.

C57BL/6 mice were i.c. infected with 500 pfu JHMV and treated with neutralizing CXCR2 antiserum or control normal rabbit serum (NRS) i.p. from day 12 to 18. Brains (A & B) and spinal cords (C & D) were removed at defined times p.i. and isolated cells were stimulated ex vivo for six hours with 5 µM of the immunodominant CD4+ epitope M_133–147 (A & C) or 5 µM of the immunodominant CD8+ epitope S_510–518 (B & D) before intracellular IFNγ staining was assessed. Data is representative of two independent experiments.

Figure 5. CXCR2 neutralization increases white matter demyelination.

C57BL/6 mice were i.c. infected with 500 pfu JHMV and treated with neutralizing CXCR2 antiserum or control normal rabbit serum (NRS) i.p. from day 12 to 20. Spinal cords were removed at day 35 p.i. and coronal sections along the length of the spinal cord were stained with luxol fast blue to assess myelin integrity (A). Quantification of the total area of demyelination along the entire length of the spinal cord (B) revealed a significant increase in the percent demyelinated area of mice treated with CXCR2 antiserum (solid outline indicates total white matter, dashed outline indicates area of demyelinated white matter). Data is a summation of four independent experiments (anti-CXCR2 n = 19, NRS n = 23).

Figure 6. CXCR2 neutralization increases oligodendrocyte lineage cell apoptosis in the spinal cord.

C57BL/6 mice were i.c. infected with 500 PFU JMHV and were treated with CXCR2 neutralizing antiserum or control normal rabbit serum (NRS) every other day beginning at day 12 p.i. Mice were sacrificed at day 15 and 17 p.i. and spinal cords were removed and processed for TUNEL labeling. (A) Quantification of the number of white matter TUNEL-positive nuclei revealed an approximate 3 to 4 fold increase in the number of apoptotic cells in anti-CXCR2 treated mice (*** p<0.001) compared to NRS-treated mice at days 15 and 17 p.i. (B) Co-immuno fluorescence of TUNEL with cleaved caspase 3 confirmed that TUNEL reactive cells were indeed apoptotic. Immunophenotyping the TUNEL-positive cells revealed apoptotic (C) CC1⁺ oligodendrocytes and (D) Olig2⁺ oligodendrocyte precursor cells. Data presented is representative of two independent experiments. Bars: (B) 20 µm; (C & D) 50 µm; (C & D) inset 5 µm.

Figure 7. The CXCR2 ligand CXCL1 protects oligodendrocytes from JHMV – mediated apoptosis in vitro.

Differentiated oligodendrocyte enriched cultures were infected with JHMV (moi 1.0), treated with varying concentrations of recombinant mouse CXCL1, and assessed for apoptosis via TUNEL 24 hours p.i. (A) CXCL1 protected oligodendrocytes from JHMV – mediated apoptosis in a dose dependent manner (*** p<0.001, relative to JHMV only; * p<0.05 relative to Media only). (B) A representative image of an apoptotic oligodendrocyte following JHMV infection. (C) Immuno-staining for JHMV nucleocapsid (green) in addition to TUNEL (red) revealed both apoptotic JHMV infected and uninfected oligodendrocytes; note the absence of apoptosis within JHMV infected oligodendrocytes treated with CXCL1 at 10 ng/ml. (D) Quantification of the frequency of JHMV infected cells amongst TUNEL+ cells revealed that slightly greater than half of the apoptotic cells (p≤0.03) were reactive for JHMV nucleocapsid. Data in panels A & D is a summation of three independent experiments. Data in panels B & C is representative of three independent experiments. Bars: (B) 60 µm; (C) 30 µm.

Figure 8. CXCR2 deficient oligodendrocytes are insensitive to CXCL1 – mediated protection from apoptotis.

Differentiated CXCR2+/+ and CXCR2−/− oligodendrocyte enriched cultures were infected with JHMV (moi 1.0) and treated with recombinant mouse CXCL1 at 10 ng/ml. (A) While CXCL1 protected CXCR2+/+ oligodendrocytes from JHMV-mediated apoptosis, CXCR2−/− oligodendrocytes were not protected (*** p≤0.001, relative to Media). Protein lysates were collected from JHMV infected CXCR2 +/+ and CXCR2−/− cultures and cleaved caspase 3, the caspase target PARP, and Bcl-2 expression were assessed (B). Densitometric quantification of the cleaved caspase 3 and Bcl-2 blots is presented panels C and D. Data in panels A–D is representative of three independent experiments.