Interleukin 10 Modulation of Pathogenic Th17 Cells During Fatal Alphavirus Encephalomyelitis

Interleukin 10 modulation of pathogenic Th17 cells during fatal alphavirus encephalomyelitis

Kirsten A. Kulcsar, Victoria K. Baxter, Ivorlyne P. Greene1, and Diane E. Griffin2

Program in Cellular and Molecular Medicine, Johns Hopkins University School of Medicine, Baltimore, MD 21205; and W. Harry Feinstone Department of Molecular Microbiology and Immunology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD 21205

Contributed by Diane E. Griffin, October 6, 2014 (sent for review July 25, 2014; reviewed by Lawrence Steinman) Mosquito-borne alphaviruses are important causes of epidemic Results encephalomyelitis. Neuronal cell death during fatal alphavirus IL-10 Deficiency Accelerates the Course of Fatal Encephalomyelitis. encephalomyelitis is immune-mediated; however, the types of Interleukin (IL) 10 is a critical regulatory cytokine produced by cells involved and their regulation have not been determined. We many types of cells that determines the balance between in- show that the virus-induced inflammatory response was accom- flammation and immune regulation by influencing antigen pre- panied by production of the regulatory cytokine IL-10, and in the sentation, T-cell differentiation, cytokine production, and intensity absence of IL-10, paralytic disease occurred earlier and mice died of inflammation (12, 13). During fatal CNS infection induced by faster. To determine the reason for accelerated disease in the intranasal inoculation of 4–6-wk-old C57BL/6J mice with 105 pfu −/− absence of IL-10, immune responses in the CNS of IL-10 and NSV, Il10 mRNA was increased in both the brain and spinal cord wild-type (WT) mice were compared. There were no differences (P < 0.0001; Fig. 1A). To determine the importance of IL-10 in the amounts of brain inflammation or peak virus replication; production in the CNS during the induction of fatal encephalo- −/− however, IL-10 animals had accelerated and increased infiltra- myelitis, we compared the disease course of NSV-infected wild- + + + + γ+ tion of CD4 IL-17A and CD4 IL-17A IFN cells compared with type (WT) C57BL/6J mice to that of IL-10–deficient C57BL/6J − − WT animals. Th17 cells infiltrating the brain demonstrated a path- mice (B6.129P2-Il10tm1Cgn/J; IL-10 / ). Disease was accelerated in − − ogenic phenotype with the expression of the transcription factor, IL-10 / mice with more rapid onset of paralysis (P < 0.0001; Fig. Tbet, and the production of granzyme B, IL-22, and GM-CSF, with C −/− 1 ) and earlier death (mean day of death was 8 d compared with greater production of GM-CSF in IL-10 mice. Therefore, in fatal 10 d, P < 0.0001; Fig. 1D). Accelerated disease progression oc- alphavirus encephalomyelitis, pathogenic Th17 cells enter the CNS curred during the time of Il10 mRNA increase in WT mice, 5–7d at the onset of neurologic disease and, in the absence of IL-10, after infection (Fig. 1 A and C). IL-10 deficiency did not affect appear earlier, develop into Th1/Th17 cells more often, and have virus replication or peak virus titers, although titers at day 7 sug- greater production of GM-CSF. This study demonstrates a role for gest that virus clearance was delayed (Fig. 1B). pathogenic Th17 cells in fatal viral encephalitis. To identify the reason for accelerated disease in the absence of IL-10, we first assessed the magnitude of the CNS inflammatory GM-CSF | Sindbis virus | immunopathology | viral encephalitis |

response by isolating and counting cells infiltrating the brain (Fig. INFLAMMATION

interleukin-10 2A) and by scoring hematoxylin and eosin (H&E)-stained coronal IMMUNOLOGY AND brain sections for inflammation (Fig. 2 B and C). The numbers of ncephalitic arthropod-borne viruses are important causes of infiltrating cells and amount of inflammation increased through 7 d Emorbidity and mortality worldwide (1, 2). In the Americas, after infection, but these parameters were not different between − − the mosquito-borne alphaviruses infect neurons and cause out- WT and IL-10 / mice. To determine whether the relative breaks of encephalomyelitis with high mortality in horses and proportions of myeloid and lymphoid cells contributing to the humans. In addition, Chikungunya virus, a newly emerging and rapidly spreading old-world alphavirus, can also cause neurologic Significance disease (3–5). The immune-mediated inflammatory response in the central nervous system (CNS) is necessary for virus clear- Mosquito-borne alphaviruses are important causes of epidemic ance, but can also cause neuronal damage. Several lines of evi- encephalomyelitis. The immune response plays an important dence suggest that fatal alphavirus encephalomyelitis is mediated role in disease; however, immune-mediated mechanisms of by the immune response to virus infection rather than virus in- pathogenesis and regulation are not understood. In this study, fection per se (6, 7), but the components of the immune response we determined that a pathogenic Th17 response occurs during involved in causing neuronal death and mechanisms for regu- fatal alphavirus encephalitis. Furthermore, the regulatory cy- lating this immunopathology have not been identified. tokine, interleukin 10, plays an important role in modulating We have used a well-characterized model of fatal alphavirus the pathogenic Th17 response. In the absence of interleukin 10, encephalomyelitis to identify immune contributors to fatal dis- the Th17 response is increased in magnitude and displays ease. Neuroadapted Sindbis virus (NSV) infects neurons in the a more pathogenic phenotype, resulting in accelerated disease brain and spinal cord with a particular tropism for hippocampal progression. These findings are important for understanding and motor neurons and causes fatal paralytic disease in adult the pathogenesis of virus infections in the central nervous C57BL/6 mice (8–10). Infected neurons die during this disease system and the identification of therapeutic interventions that focus on immune modulation in the central nervous system. process, but neuronal death is caused by the immune response to

infection rather than by damage from virus replication (7, 10). Author contributions: K.A.K. and D.E.G. designed research; K.A.K., V.K.B., and I.P.G. per- Mice can be protected from fatal disease if the CNS in- formed research; K.A.K. and V.K.B. analyzed data; and K.A.K. and D.E.G. wrote the paper. flammatory response is inhibited, despite the fact that this results Reviewers included: L.S., Beckman Center for Molecular Medicine. in a failure of virus clearance (7, 11). Studies of immune- The authors declare no conflict of interest. deficient mice have implicated T cells in this immunopathologic 1Present address: Quidel, Inc., San Diego, CA 92130. process (6, 9), but the types of T cells involved, their pathologic 2To whom correspondence should be addressed. Email: [email protected]. function, and the regulation of this response during fatal disease This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10. are not known. 1073/pnas.1418966111/-/DCSupplemental.

www.pnas.org/cgi/doi/10.1073/pnas.1418966111 PNAS | November 11, 2014 | vol. 111 | no. 45 | 16053–16058 Downloaded by guest on September 29, 2021 − − producing granzyme B in IL-10 / mice would be predicted to decrease, rather than increase, cytotoxicity, but could slow virus clearance (Fig. 1B) (15, 16). At 7 d after infection, no differences − − were observed between WT and IL-10 / mice in the frequency + or number of CD8 T cells expressing IFNγ, TNFα, or granzyme B (Fig. 3 H–N). Analysis of fluorescence intensity indicated that + amounts of granzyme B per CD8 T cell were similar in WT and − − IL-10 / mice (Fig. 3 A and H). Therefore, differences in num- + bers or function of CD8 T cells did not explain accelerated − − disease in IL-10 / mice.

IL-10 Deficiency Increases the Numbers of Th17 Cells in the CNS. + CD4 T cells can differentiate into several functional groups defined by transcription factor expression and cytokine production. Because both Th1 and Th17 subsets have been impli- Fig. 1. IL-10 is important in regulating disease progression independent of cated in autoimmune disease in the CNS (17), we used flow − − virus replication. WT and IL-10 / mice were infected intranasally with 105 cytometry and intracellular cytokine staining to analyze brain + pfu NSV. (A) Il10 mRNA expression measured by quantitative real-time PCR CD4 cells for production of signature cytokines IFNγ (Th1) and in the brains (filled square) and spinal cords (open square) of NSV-infected IL-17A (Th17) (Fig. 4). At 5 d after infection, when signs of − − WT mice. Ct values were normalized to GAPDH. Ct values and fold change neurologic disease appear in IL-10 / mice (Fig. 1C), the IFNγ- + were calculated relative to uninfected controls (ΔΔCt). Data are pooled from producing CD4 population was similar (Fig. 4 A–C), but there ± – two independent experiments and represent the mean SEM of 4 6 mice at was a fivefold increase in the frequency (2.97% vs. 14.62%, P = each time point; *P < 0.05, ***P < 0.001, ****P < 0.0001. (B) Virus titers were 0.0063; Fig. 4D) and in the number (6.4 × 102 vs. 5.2 × 103 cells determined using brain homogenates prepared from WT (filled square, solid + − − P = E – line) and IL-10 / (open circle, dashed line) mice. Data are pooled from three per brain, 0.0153; Fig. 4 ) of IL-17 producing CD4 T cells −/− independent experiments and presented as the mean ± SEM of 6–9 mice at in IL-10 mice compared with WT mice. At 7 d after infection, each time point; **P < 0.01. (C) Signs of disease in WT and IL-10−/− mice when neurologic signs of disease become prominent in WT mice, infected with NSV were monitored daily. The clinical score scale was as fol- the numbers of IL-17–producing as well as IFNγ-producing + lows: 0, no symptoms; 1, abnormal hind-limb and tail posture, ruffled fur, CD4 T cells increased in WT mice, but the percentage (0.41% and/or hunched back; 2, unilateral hind-limb paralysis; 3, bilateral hind-limb vs. 1.94%, P = 0.0120; Fig. S2 A and D) and numbers (1.2 × 103 paralysis and/or moribund; 4, dead. Data are pooled from three independent × 3 P = E −/− vs. 5.4 10 cells per brain, 0.0508; Fig. S2 ) of Th17 cells experiments and presented as the mean ± SEM; WT, n = 27 and IL-10 , n = −/− < – remained higher in IL-10 mice. In the spinal cord, Th17, but 36; ****P 0.0001. (D) Survival was assessed using a Kaplan Meier curve and −/− −/− not Th1, cells were also more prevalent in IL-10 mice than log-rank test. The mean day of death was 10 for WT mice and 8 for IL-10 P = F–J mice. Data are pooled from three independent experiments; WT, n = 35 and WT mice ( 0.0420; Fig. S2 ). − − IL-10 / , n = 37; ****P < 0.0001. To assess the effect of IL-10 deficiency on Th17-related gene expression in the CNS, we compared the levels of mRNAs for Il17a and for the Th17-specific chemokine Ccl20 (18) in the − − inflammatory process differed, we used flow cytometry for brains and spinal cords of WT and IL-10 / mice (Fig. 4 K–N). − − identification of cells isolated from the brain (Fig. 2 D–O). There Il17a mRNA levels were higher in IL-10 / compared with WT were no differences in numbers or percentages of microglia (Fig. mice in the brain (P = 0.0144; Fig. 4K) and spinal cord (P = + − − 2 D and E), natural killer (NK) cells (Fig. 2 J and K), CD4 T 0.0095; Fig. 4L). Ccl20 mRNA levels were higher in IL-10 / + cells (Fig. 2 L and M), or CD8 T cells (Fig. 2 N and O). WT mice in the spinal cord (P = 0.0126; Fig. 4N), but not the brain mice had more monocyte/macrophages (P < 0.05; Fig. 2F), and (Fig. 4M). − − IL-10 / mice had more neutrophils (P < 0.01; Fig. 2 H and I) Therefore, both Th1 and Th17 cells began infiltrating the CNS at 7 d after infection. of WT mice coincident with the onset of paralysis and were To determine whether increased neutrophil infiltration into present during fatal NSV infection. In the absence of IL-10, − − the CNS was responsible for accelerated disease in IL-10 / accelerated disease correlated with an increase in the Th17, but mice, neutrophils were depleted using antibody to Ly6G or not the Th1, response. control antibody given at the time of infection and 4 d after infection (14). Depletion of circulating neutrophils was confirmed Th17 Cells in the CNS of Mice with Fatal Viral Encephalomyelitis Have to be greater than 90% by blood smears (Fig. S1A) and greater a Pathogenic Phenotype That Is Enhanced in the Absence of IL-10. than 65% in the brain (Fig. S1 B and C). The presence or ab- Th17 cells are polyfunctional, have multiple phenotypes, and − − sence of neutrophils did not alter the disease course in IL-10 / can play both detrimental and beneficial roles in disease path- mice (Fig. S1 D and E). Therefore, neither differences in num- ogenesis (19–21). To determine the phenotype of the Th17 cells + bers nor overall types of CNS inflammatory cells explained ac- entering the brain at the onset of neurological disease, CD4 IL- + celerated disease in the absence of IL-10. 17A cells from the brain at 5 d after NSV infection were ex- + + As both CD8 and CD4 T cells have been implicated in amined for production of granzyme B, IL-22, and GM-CSF, immunopathologic processes during NSV infection (6), we next cytokines associated with a pathogenic phenotype in experi- assessed the possibility that the absence of IL-10 led to func- mental autoimmune encephalomyelitis (EAE) (19–21) (Fig. 5). − − tional differences in these cell populations even though overall Th17 cells from both WT and IL-10 / mice expressed all three numbers were similar. Flow cytometry and intracellular cytokine cytokines. The proportions of Th17 cells producing granzyme B staining were used to measure production of the effector mole- (23.0% vs. 10.8%, P = 0.028; Fig. 5 A and E) and IL-22 (41.5% + cules TNFα, IFNγ, and granzyme B by CD8 T cells isolated vs. 22.6%, P = 0.0104; Fig. 5 A and G) were greater in WT than − − + from the brain at 5 d (Fig. 3 A–G) and 7 d (Fig. 3 H–N) after IL-10 / mice, but there were more granzyme B (9.5 × 102 vs. + + infection. At 5 d after infection, the percentage of CD8 T cells 9.0 × 101, P = 0.0328; Fig. 5F) and IL-22 (6.5 × 102 vs. 7.9 × 101, − − − − expressing granzyme B was lower in IL-10 / mice than WT mice P = 0.0369; Fig. 5H) Th17 cells in the brains of IL-10 / com- (68.8% vs. 86.4%, P = 0.0247; Fig. 3B), whereas the number and pared with WT mice. Analysis of fluorescence intensity indicated + percentage of CD8 T cells producing IFNγ (Fig. 3 D and E) and that the amount of granzyme B expressed by each positive Th17 + TNFα (Fig. 3 F and G) were similar. A decrease in CD8 T cells cell was similar (Fig. 5A).

16054 | www.pnas.org/cgi/doi/10.1073/pnas.1418966111 Kulcsar et al. Downloaded by guest on September 29, 2021 Greater differences were observed for Th17 cell production of − − GM-CSF. In IL-10 / mice, both the proportion (23.8% vs. 15.7%, P = 0.0173; Fig. 5 A and B) and numbers (1.1 × 103 vs. 4.1 × 101, P = 0.0268; Fig. 5C) of Th17 cells producing GM-CSF were higher than in WT mice. Additionally, the Th17 cells in the brains − − of IL-10 / mice had higher mean fluorescence intensities for GM- CSF than those in the brains of WT mice, indicating a higher level of GM-CSF production per cell (P = 0.0179; Fig. 5 A and D). Together, these data show that the Th17 cells in the CNS during fatal NSV infection have a pathogenic phenotype and that, in the absence of IL-10, the pathogenic Th17 response is amplified and GM-CSF production increased. A definitive role for a particular Th17 product could not be identified because neither treatment with neutralizing antibody to IL-17 nor GM-CSF improved the − − outcome in IL-10 / or WT mice compared with control antibody (Fig. S3 A–H). Th17 cells develop independently of Tbet and IFNγ, but in tissue can acquire expression of these signature Th1 factors along with an increase in pathogenicity (22). An important pathogenic role is suggested by the target organ presence of IL- + + 17A IFNγ Th1/Th17 cells during EAE and autoimmune colitis (21) and by the fact that mice deficient in the transcription factor Tbet are protected from development of EAE (23, 24). Because pathogenic Th17 cells can develop into multifunctional patho- + genic Th1/17 cells, we examined brain CD4 T cells for expression of Tbet and the Th17 transcription factor, RORγt, in conjunction with the production of IL-17A and IFNγ (Fig. 6). + + WT mice had mostly Th1 cells (IFNγ Tbet ) and Th17 cells (IL- + + + + 17A RORγt ) with very few Th1/Th17 cells (IFNγ IL-17A ) (Fig. 6 A and B). In the absence of IL-10, however, in addition to an increase in the proportion of Th17 cells (8.98% vs. 2.53%, P = 0.0003), there was also an increase in the Th1/Th17 population (4.28% vs. 0.46%, P = 0.0262; Fig. 6 A and B). + − − The IFNγ cells in the brains of WT and IL-10 / mice expressed Tbet, with little expression of RORγt, consistent with + INFLAMMATION classic Th1 cells (Fig. 6 C and D, Top). The IL-17A cells IMMUNOLOGY AND expressed RORγt, the canonical Th17 lineage transcription factor, as well as Tbet, a marker of pathogenicity (Fig. 6 C and D, + + − − Middle). The IFNγ IL-17A cells in the brains of IL-10 / mice also expressed RORγt and Tbet (Fig. 6 C and D, Bottom). These data show that populations of pathogenic Th17 cells are present in the CNS at the onset of neurological disease in WT and − − IL-10 / mice. The increased presence of Th1/Th17 cells in the absence of IL-10 suggests that these cells are regulated by IL-10 and may contribute to the accelerated disease progression. Discussion T-cell–mediated damage drives pathogenesis of fatal alphavirus encephalomyelitis. We identified Th1 and pathogenic Th17 cells in the CNS at the onset of neurological disease. IL-10 expression increased in the CNS after infection and was identified as a critical regulator of the Th17 and Th1/Th17 response. In the Fig. 2. The inflammatory response to NSV infection is similar in WT and IL- absence of IL-10, paralytic disease was accelerated concomitant −/− −/− 10 mice. WT (filled square or bar, solid line) and IL-10 (open circle or with increased infiltration of pathogenic Th17 and Th1/Th17 bar, dashed line) mice were infected with NSV and inflammation assessed. cells into the CNS. Pathogenic Th17 cells expressed effector (A) The average number of live cells isolated from the brains (n = 6–10, ± cytokines IL-17A, IL-22, granzyme B, and GM-CSF and tran- pooled) of mice after infection. Data represent the mean SEM from four γ separate experiments. (B and C) Brains were collected from uninfected and scription factors ROR t and Tbet, with a marked increase in NSV-infected mice at 3, 5, and 7 d after infection. (B) Extent of inflammation production of GM-CSF in the absence of IL-10. in coded H&E-stained coronal sections scored on a scale from 0 to 3, with an The role of IL-10 regulation of the immune response during extra point given to sections that had excessive parenchymal cellularity. Data viral encephalitis has received limited analysis. IL-10 has a pro- − − are presented as the mean ± SEM of three WT (black bars) and three IL-10 / tective role during acute coronavirus and flavivirus encephalitis (open bars) mice. (C) Representative H&E-stained coronal sections of un- in mice, but the mechanisms by which outcome is improved were × – infected mice and infected mice 7 d after infection (20 ). (D O) Flow cyto- not identified (25–27). Although IL-10 deficiency can promote metric analysis of isolated cells pooled from the brains (n = 6–10) of WT or IL- − − 10 / mice without infection and 3, 5, and 7 d after infection. The absolute number of cells per brain and frequency (percentage of live cells) of microglia (D and E), macrophages/monocytes (F and G), neutrophils (H and I), determined. The data represent the mean ± SEM from two to four in- + + NK cells (J and K), CD4 T cells (L and M), and CD8 T cells (N and O) were dependent experiments; *P < 0.05, **P < 0.01.

Kulcsar et al. PNAS | November 11, 2014 | vol. 111 | no. 45 | 16055 Downloaded by guest on September 29, 2021 The effector function(s) of Th17 cells that cause neurologic disease are not clear, but multiple factors may contribute. IL-17 production is reported to be both protective and damaging (17, 37–40). IL-17 can induce neuronal cell death in vitro (41), and EAE is attenuated in mice lacking IL-17 (42). Neutralization of IL-17 did not improve outcome after NSV infection, but a role for IL-17 in neuronal damage cannot be excluded because of unknown CNS penetration of the antibody. Granzyme B induces neurotoxicity by activation of protease-activated G protein-cou- pled receptors on neurons (43), and both granzyme B and IL-22 contribute to blood–brain barrier disruption (44). GM-CSF is a potent mediator of Th17 cell damage during EAE (45). GM-CSF activates microglia, enhances myeloid cell recruitment, and promotes Th17 cell differentiation (46–48). The greatest influence of IL-10 deficiency on Th17 function was increased production of GM-CSF on both a population and single-cell basis. Thus, although a neutralizing antibody did not improve the outcome, GM-CSF cannot be excluded as a con- tributor to immune-mediated damage in viral encephalomyelitis. Together, these studies have characterized the functional features of Th17 cells that develop in the absence of IL-10 and have identified pathogenic Th17 and Th1/17 cells as participants in the immunopathologic process leading to fatal viral encephalomyelitis. This understanding of the pathogenesis of acute virus-induced neurologic disease will aid in identifying therapeutic interventions that focus on immune modulation as well as antiviral drugs.

+ Fig. 3. CD8 T-cell effector molecule expression is similar in the brains of WT and IL-10−/− mice during NSV infection. Isolated cells were pooled (n = 6– 10) from the brains of WT and IL-10−/− mice at 5 d (A–G) and 7 d (H–N) after infection; stimulated with PMA and ionomycin in the presence of brefeldin A to assess IFNγ, TNFα, and granzyme B production; and analyzed by flow + cytometry. Contour plots show the gating of CD8 T cells that produce IFNγ, TNFα, and granzyme B at 5 d (A) and 7 d (H) after infection and are representative of three independent experiments. Histograms show the fluorescence intensity of granzyme B expression in CD8+ T cells isolated from the brains of WT (black) and IL-10−/− (orange) mice relative to the isotype control + (gray filled). The frequency of CD8 T cells that produce IFNγ (B and I), TNFα (D and K), and granzyme B (F and M) calculated as the mean ± SEM from + three independent experiments. The number of CD8 T cells that produce IFNγ (C and J), TNFα (E and L), and granzyme B (G and N) was calculated as the mean ± SEM from three independent experiments; *P < 0.05.

an overall increase in inflammation (28–30), this did not occur in alphavirus-induced encephalomyelitis. Rather, IL-10 deficiency + facilitated a selective increase in the proportion of CD4 T cells −/− in the inflammatory response that were Th17 cells. Fig. 4. The Th17 response is increased in the brains of IL-10 mice compared with WT mice at the onset of clinical disease. Mononuclear cells were − − IL-10 directly inhibits both differentiation and proliferation of isolated from the brains (n = 6–10) of WT and IL-10 / mice at 5 d after NSV Th17 cells (31–33), and IL-10 deficiency leads to an increase in infection and stimulated with PMA and ionomycin in the presence of bre- + Th17 cells in mice infected with leishmania (34), influenza virus feldin A to assess the production of IFNγ (Th1) and IL-17A (Th17) in CD4 T (35), and respiratory syncytial virus (RSV) (36). During pulmo- cells. (A) Contour plots show the gating for IFNγ+ and IL-17A+ cells within the brain CD4+ T-cell population. Plots are representative of three independent nary infection with influenza virus, an increase in Th17 cells is + experiments. (B–E) Quantification of the frequency of CD4 T cells that ex- associated with a better outcome, whereas for infection with + + press IFNγ (B) and IL-17A (D), as well as the number of CD4 IFNγ (C)and + + − − RSV and leishmania, the disease was more severe. In none of CD4 IL-17A (E) T cells in the brains of WT (black) and IL-10 / (white) mice. these studies was the function or cytokine profile of the Th17 Data are shown as the mean ± SEM from three independent experiments; cells characterized to provide insight into how IL-10 modulated *P < 0.05, **P < 0.01. (F–I) Analysis of Il17a (F and G) and Ccl20 (H and I) Th17-associated disease. During NSV infection of the CNS, IL- mRNAs in the brains (F and H) and spinal cords (G and I) of WT (filled square, −/− 10 deficiency also led to more severe disease and an increase solid line) and IL-10 (open circle, dashed line) mice during NSV infection. Gene Ct values were normalized to Gapdh, and fold change was calculated in Th17 cells. Functional characterization of the Th17 cells relative to uninfected controls (ΔΔCt). Data are pooled from two independent revealed a pathogenic phenotype with production of granzyme B, experiments and presented as the mean ± SEM from four to six mice at each IL-22, and GM-CSF. time point; *P < 0.05, **P < 0.01, ***P < 0.001.

16056 | www.pnas.org/cgi/doi/10.1073/pnas.1418966111 Kulcsar et al. Downloaded by guest on September 29, 2021 incubated again, and filtered (70 μm). Myelin debris and red blood cells were removed by centrifugation on a 30/70% percoll gradient for 30 min at 4 °C. Mononuclear cells at the interface were collected, resuspended in PBS + 2 mM EDTA, and live cells counted using trypan blue exclusion.

Flow Cytometry. Approximately 1–2 × 106 cells were stained with the violet Live/Dead Fixable Cell Stain Kit (Invitrogen) in PBS + 2 mM EDTA, blocked with rat anti-mouse CD16/CD32 (BD Pharmingen), diluted in PBS + 2mM EDTA + 0.5% BSA, surface-stained for 25 min on ice, fixed, and resuspended in 200 μLPBS+ 2 mM EDTA + 0.5% BSA. All antibodies were from BD Pharmingen or eBioscience: CD45 (clone 30-F11), CD11b (clone M1-70), Ly6G (clone 1A8), Ly6C (clone HK1.4), CD3 (clone 17A2), CD4 (clone RM4-5), CD8 (clone 53–6.7), and NK1.1 (clone PK136). Cell types were defined as follows: + − − + microglia (CD45loCD11b Ly6G Ly6C ), macrophages/monocytes (CD45hiCD11b − + + + + + − Ly6G Ly6C ), neutrophils (CD45 CD11b Ly6G Ly6Cint), NK cells (CD45 CD3 + + + + + + NK1.1 ), T cells (CD3 ), CD4 T cells (CD3 CD4 ), and CD8 T cells (CD3 CD8 ). − − – × 6 + Fig. 5. Th17 cells found in the brains of NSV-infected WT and IL-10 / mice For intracellular cytokine staining, 2 3 10 cells were stimulated in RPMI have a pathogenic phenotype. (A) Mononuclear cells were isolated and 1% FBS containing 50 ng/mL of phorbol-12-myristate 13-acetate (PMA) and − − μ pooled from brains (n = 6–10) of WT and IL-10 / mice 5 d after infection and 1 g/mL ionomycin in the presence of GolgiPlug-brefeldin A (BD Pharmingen) for stimulated with PMA and ionomycin in the presence of brefeldin A to assess 4 h. After surface staining, cells were fixed and permeabilized using either the + + the production of GM-CSF, granzyme B, and IL-22 by Th17 cells (CD4 IL-17A ). CytoFix/CytoPerm kit (BD Pharmingen) or the Foxp3 Transcription Factor Fix/ (A) Contour plots show the gating for GM-CSF, granzyme B, and IL-22 pro- Perm kit (eBioscience). Intracellular markers were stained for 25 min on ice and duction within the CD4+ T-cell population. Histograms show the fluorescence resuspended in 200 μLofPBS+ 2mMEDTA+ 0.5% BSA. All antibodies were intensity of GM-CSF and granzyme B expression in Th17 cells isolated from from BD Pharmingen or eBioscience: IFNγ (clone XMG1.2), IL-17a (clone − − the brains of WT (black) and IL-10 / (orange) mice relative to the isotype eBio17B7), TNFα (clone MP6-XT22), granzyme B (clone NGZB), Tbet (clone control (gray, filled). Contour plots and histograms are representative of ebio4B10), RORγt (clone B2D, eBioscience), GM-CSF (clone MP1-22E9), granzyme + three independent experiments. (B–H) The frequencies and numbers of CD4 B (clone NGZB), and IL-22 (clone 1H8PWSR). Data were acquired with a BD FACS + IL-17A T cells producing GM-CSF (B and C), granzyme B (E and F), and IL-22 Canto II using FACS Diva software (version 6.0) and analyzed using FlowJo 8.8.7 (G and H) in NSV-infected WT (black) and IL-10−/− (white) mice. (D) The mean (TreeStar Inc.). fluorescence intensity of CD4+IL-17A+GM-CSF+ T cells in the brains of WT − − (black) and IL-10 / (white) mice. Data represent the mean ± SEM from three Gene Expression Analysis Using Real-Time PCR. RNA was isolated from frozen independent experiments; *P < 0.05. tissue using the RNeasy Lipid Mini RNA Isolation Kit (Qiagen). RNA was quantified using a nanodrop spectrophotometer, and cDNA was prepared with the High Capacity cDNA Reverse Transcription Kit (Life Technologies) Methods and Materials using 0.5–2.5 μg RNA. Quantitative real-time PCR was performed using 2.5 μL Animals and Infection. C57BL/6J WT and B6.129P2-Il10tm1Cgn/J (C57BL/6J IL- cDNA, TaqMan gene expression arrays (Il10, Il17a, and Ccl20), and 2× Uni- − − 10 / ) (49) mice were purchased from Jackson Laboratories and bred in versal PCR Mastermix (Applied Biosystems). Gapdh mRNA levels were de- house. The NSV strain of Sindbis virus (50) was grown and assayed by plaque termined using the rodent primer and probe set (Applied Biosystems). All – reactions were run on the Applied Biosystems 7500 real-time PCR machine INFLAMMATION

formation in BHK cells. For infection, sex-matched 4 6-wk-old mice were IMMUNOLOGY AND inoculated intranasally with 105 pfu NSV in 20 μL HBSS. For assessment of with the following conditions: 50 °C for 2 min, 95 °C for 10 min, 95 °C for morbidity and mortality, mice were monitored daily. The scoring system 15 s, and 60 °C for 1 min for 50 cycles. Transcript levels were determined by used was as follows: 0, no signs of disease; 1, abnormal hind-limb and tail normalizing the target gene Ct value to the Ct value of Gapdh. This posture, ruffled fur, and/or hunched back; 2, unilateral hind-limb paralysis; 3, bilateral hind-limb paralysis or full-body paralysis; 4, dead. For tissue collection, mice were anesthetized with isoflurane and bled by cardiac puncture. Mice were perfused with ice-cold PBS, and brains and spinal cords were collected and used fresh or snap frozen and stored at −80 °C. All experiments were performed according to guidelines approved by the Johns Hopkins University Institutional Animal Care and Use Committee.

Virus Assays. Tissue was homogenized in cold PBS to make 10% (wt/vol) brain and spinal cord homogenates and was clarified by centrifugation. Infectious virus was assayed by plaque formation on BHK-21 cells. Data are plotted as

the mean of the log10 value of plaque forming units ± SEM. For statistical purposes, samples in which no virus was detected at a 1:10 dilution were assigned a value of 0.85, halfway between the limit of detection and 0.

Histology. Uninfected and NSV-infected mice at 3, 5, and 7 d after infection were perfused with ice-cold PBS followed by 4% (wt/vol) paraformaldehyde. − − Brains were removed and cut into 2-mm coronal slices with an adult Mouse Fig. 6. Th17 cells in the brains of NSV-infected WT and IL-10 / mice express Brain Slicer (Zivic Instruments) that were postfixed in 4% (wt/vol) para- Tbet and RORγt and develop into Th1/Th17 cells in the absence of IL-10. formaldehyde for 24 h at 4 °C. Slices were washed in PBS and embedded in Mononuclear cells isolated and pooled from the brains (n = 6–10) of WT and − − paraffin. Coronal brain sections (10 μm) were stained with H&E, coded, and IL-10 / mice 5 d after infection were stimulated with PMA and ionomycin in scored blindly as previously described (51) using a 0–3 scale: 0, no detectable the presence of brefeldin A to assess the production of IFNγ and IL-17A by + inflammation; 1, one to two small inflammatory foci per section; 2, mod- CD4 T cells. (A) Contour plots showing IFNγ and IL-17A production within + erate inflammatory foci in up to 50% of 10× fields; 3, moderate to large the CD4 T-cell population. Plots are representative of three independent inflammatory foci in greater than 50% of 10× fields. An additional point was experiments. (B) Quantification of the IFNγ+ (Th1), IL-17A+ (Th17), and IFNγ+ added for excessive parenchymal cellularity, allowing for a maximal score IL-17A+ (Th1/Th17) CD4+ T cells in the brains of WT (black) and IL-10−/− of 4. (white) mice presented as the mean ± SEM from three independent experiments; *P < 0.05, ***P < 0.001. (C) Tbet and RORγt transcription factor + + + + + Mononuclear Cell Isolation. Single-cell suspensions were made from brain and expression in IFNγ (Th1), IL-17A (Th17), and IFNγ IL-17A (Th1/Th17) CD4 spinal cord tissue homogenized in RPMI + 1% FBS, 1 mg/mL collagenase T cells. (D) Histograms show Tbet and RORγt expression in WT (black) and IL- (Roche), and 0.1 mg/mL DNase (Roche) using the GentleMACS system (Mil- 10 KO (orange) relative to an isotype control (gray, filled). Contour plots and tenyi). The homogenate was incubated (37 °C, 15 min), homogenized again, histograms are representative of three independent experiments.

Kulcsar et al. PNAS | November 11, 2014 | vol. 111 | no. 45 | 16057 Downloaded by guest on September 29, 2021 normalized value was used to calculate the fold change relative to the av- 3A2, BioXcell) antibody intraperitoneally in a volume of 200 μL (dlluted in erage of the uninfected control (ΔΔCt method). PBS) at 2, 4, 6, and 8 d after infection.

Neutrophil Depletion. Mice were administered 0.5 mg rat anti-Ly6G antibody Statistical Analysis. Data from two to four independent experiments or at (clone 1A8, BioXcell) or the rat IgG2a isotype control (clone 3A2, BioXcell) least three mice per group were used. Survival was compared using Kaplan– intraperitoneally in 200 μL PBS at the time of infection and 4 d after in- Meier survival curves (log rank test). Differences during the course of infection. To document neutrophil depletion, blood smears were examined for fection in a single group were determined using one-way ANOVA and Dunn the percentage of white blood cells that were neutrophils. To document posttests. Differences between groups during the course of infection were depletion in the CNS, brains collected 7 d after infection were analyzed by determined using two-way ANOVA and Bonferroni posttests. Differences flow cytometry. between groups at a single time point were determined using an unpaired, two-tailed Student t test with a 95% confidence interval. All statistical analyses used Prism 5 (GraphPad). Cytokine Neutralization. To neutralize IL-17a, mice were treated with 250 μg of mouse anti–IL-17a (clone 17F3, BioXcell) or the mouse IgG1 isotype con- ACKNOWLEDGMENTS. We thank Tricia Niles for help with flow cytometry, trol (clone MOPC-21, BioXcell) antibody intraperitoneally in a volume of Damien Chopy for help with early experiments, and Drs. Alan Scott and Jay μ 200 L (diluted in PBS) at the time of infection and again at 3 and 6 d after Bream for helpful advice. This work was supported by Grants F31 NS076223 infection. To neutralize GM-CSF, mice were treated with 250 μg of rat anti– (to K.A.K.), T32 OD011089 (to V.K.B.), T32 AI007247 (to I.P.G.), and R01 GM-CSF (clone MP1-22E9, BioXcell) or the rat IgG2a isotype control (clone NS087539 (to D.E.G.) from the National Institutes of Health.

1. Hollidge BS, González-Scarano F, Soldan SS (2010) Arboviral encephalitides: Trans- 27. Tun MMN, et al. (2014) Protective role of TNF-α, IL-10 and IL-2 in mice infected with mission, emergence, and pathogenesis. J Neuroimmune Pharmacol 5(3):428–442. the Oshima strain of Tick-borne encephalitis virus. Sci Rep 4:5344. 2. Go YY, Balasuriya UBR, Lee C-K (2014) Zoonotic encephalitides caused by arboviruses: 28. Lin MT, Hinton DR, Parra B, Stohlman SA, van der Veen RC (1998) The role of IL-10 in Transmission and epidemiology of alphaviruses and flaviviruses. Clin Exp Vaccine Res mouse hepatitis virus-induced demyelinating encephalomyelitis. Virology 245(2): 3(1):58–77. 270–280. 3. Chusri S, Siripaitoon P, Hirunpat S, Silpapojakul K (2011) Case reports of neuro-Chi- 29. Cheeran MC-J, Mutnal MB, Hu S, Armien A, Lokensgard JR (2009) Reduced lympho- kungunya in southern Thailand. Am J Trop Med Hyg 85(2):386–389. cyte infiltration during cytomegalovirus brain infection of interleukin-10-deficient 4. Staples JE, Breiman RF, Powers AM (2009) Chikungunya fever: An epidemiological mice. J Neurovirol 15(4):334–342. review of a re-emerging infectious disease. Clin Infect Dis 49(6):942–948. 30. Marques CP, et al. (2004) Interleukin-10 attenuates production of HSV-induced in- 5. Nelson J, Waggoner JJ, Sahoo MK, Grant PM, Pinsky BA (2014) Encephalitis caused by flammatory mediators by human microglia. Glia 47(4):358–366. Chikungunya virus in a traveler from the Kingdom of Tonga. J Clin Microbiol 52(9): 31. Huber S, et al. (2011) Th17 cells express interleukin-10 receptor and are controlled by 3459–3461. Foxp3⁻ and Foxp3+ regulatory CD4+ T cells in an interleukin-10-dependent manner. 6. Rowell JF, Griffin DE (2002) Contribution of T cells to mortality in neurovirulent Immunity 34(4):554–565. Sindbis virus encephalomyelitis. J Neuroimmunol 127(1-2):106–114. 32. Chaudhry A, et al. (2011) Interleukin-10 signaling in regulatory T cells is required for 7. Greene IP, Lee E-Y, Prow N, Ngwang B, Griffin DE (2008) Protection from fatal viral suppression of Th17 cell-mediated inflammation. Immunity 34(4):566–578. encephalomyelitis: AMPA receptor antagonists have a direct effect on the in- 33. McGeachy MJ, et al. (2007) TGF-beta and IL-6 drive the production of IL-17 and IL-10 by flammatory response to infection. Proc Natl Acad Sci USA 105(9):3575–3580. T cells and restrain T(H)-17 cell-mediated pathology. Nat Immunol 8(12):1390–1397. 8. Jackson AC, Moench TR, Trapp BD, Griffin DE (1988) Basis of neurovirulence in Sindbis 34. Gonzalez-Lombana C, et al. (2013) IL-17 mediates immunopathology in the absence virus encephalomyelitis of mice. Lab Invest 58(5):503–509. of IL-10 following Leishmania major infection. PLoS Pathog 9(3):e1003243. 9. Kimura T, Griffin DE (2003) Extensive immune-mediated hippocampal damage in mice 35. McKinstry KK, et al. (2009) IL-10 deficiency unleashes an influenza-specific Th17 re- surviving infection with neuroadapted Sindbis virus. Virology 311(1):28–39. sponse and enhances survival against high-dose challenge. J Immunol 182(12): 10. Nargi-Aizenman JL, Griffin DE (2001) Sindbis virus-induced neuronal death is both 7353–7363. necrotic and apoptotic and is ameliorated by N-methyl-D-aspartate receptor antag- 36. Weiss KA, Christiaansen AF, Fulton RB, Meyerholz DK, Varga SM (2011) Multiple CD4+ onists. J Virol 75(15):7114–7121. T cell subsets produce immunomodulatory IL-10 during respiratory syncytial virus 11. Nargi-Aizenman JL, et al. (2004) Glutamate receptor antagonists protect from virus- infection. J Immunol 187(6):3145–3154. induced neural degeneration. Ann Neurol 55(4):541–549. 37. Crowe CR, et al. (2009) Critical role of IL-17RA in immunopathology of influenza in- 12. Moore KW, de Waal Malefyt R, Coffman RL, O’Garra A (2001) Interleukin-10 and the fection. J Immunol 183(8):5301–5310. interleukin-10 receptor. Annu Rev Immunol 19:683–765. 38. Mukherjee S, et al. (2011) IL-17-induced pulmonary pathogenesis during respiratory 13. Couper KN, Blount DG, Riley EM (2008) IL-10: The master regulator of immunity to viral infection and exacerbation of allergic disease. Am J Pathol 179(1):248–258. infection. J Immunol 180(9):5771–5777. 39. Suryawanshi A, et al. (2011) Role of IL-17 and Th17 cells in herpes simplex virus- 14. Garraud K, et al. (2012) Differential role of the interleukin-17 axis and neutrophils in induced corneal immunopathology. J Immunol 187(4):1919–1930. resolution of inhalational anthrax. Infect Immun 80(1):131–142. 40. Savarin C, et al. (2012) IFN-γ protects from lethal IL-17 mediated viral encephalomy- 15. Haile Y, et al. (2011) Granule-derived granzyme B mediates the vulnerability of hu- elitis independent of neutrophils. J Neuroinflammation 9:104. man neurons to T cell-induced neurotoxicity. J Immunol 187(9):4861–4872. 41. Wang D-D, et al. (2009) IL-17 potentiates neuronal injury induced by oxygen-glucose 16. Wang T, et al. (2006) Granzyme B mediates neurotoxicity through a G-protein-cou- deprivation and affects neuronal IL-17 receptor expression. J Neuroimmunol 212(1-2): pled receptor. FASEB J 20(8):1209–1211. 17–25. 17. Petermann F, Korn T (2011) Cytokines and effector T cell subsets causing autoimmune 42. Komiyama Y, et al. (2006) IL-17 plays an important role in the development of ex- CNS disease. FEBS Lett 585(23):3747–3757. perimental autoimmune encephalomyelitis. J Immunol 177(1):566–573. 18. Yamazaki T, et al. (2008) CCR6 regulates the migration of inflammatory and regu- 43. Wang T, et al. (2012) Granzyme B-induced neurotoxicity is mediated via activation of latory T cells. J Immunol 181(12):8391–8401. PAR-1 receptor and Kv1.3 channel. PLoS ONE 7(8):e43950. 19. Lee Y, et al. (2012) Induction and molecular signature of pathogenic TH17 cells. Nat 44. Kebir H, et al. (2007) Human TH17 lymphocytes promote blood-brain barrier dis- Immunol 13(10):991–999. ruption and central nervous system inflammation. Nat Med 13(10):1173–1175. 20. Korn T, Bettelli E, Oukka M, Kuchroo VK (2009) IL-17 and Th17 cells. Annu Rev Im- 45. McQualter JL, et al. (2001) Granulocyte macrophage colony-stimulating factor: A new munol 27:485–517. putative therapeutic target in multiple sclerosis. J Exp Med 194(7):873–882. 21. Peters A, Lee Y, Kuchroo VK (2011) The many faces of Th17 cells. Curr Opin Immunol 46. Codarri L, Greter M, Becher B (2013) Communication between pathogenic T cells and 23(6):702–706. myeloid cells in neuroinflammatory disease. Trends Immunol 34(3):114–119. 22. Hirota K, et al. (2011) Fate mapping of IL-17-producing T cells in inflammatory re- 47. El-Behi M, et al. (2011) The encephalitogenicity of T(H)17 cells is dependent on IL-1- sponses. Nat Immunol 12(3):255–263. and IL-23-induced production of the cytokine GM-CSF. Nat Immunol 12(6):568–575. 23. Bettelli E, et al. (2004) Loss of T-bet, but not STAT1, prevents the development of 48. van Nieuwenhuijze A, et al. (2013) GM-CSF as a therapeutic target in inflammatory experimental autoimmune encephalomyelitis. J Exp Med 200(1):79–87. diseases. Mol Immunol 56(4):675–682. 24. Yang Y, et al. (2009) T-bet is essential for encephalitogenicity of both Th1 and Th17 49. Kühn R, Löhler J, Rennick D, Rajewsky K, Müller W (1993) Interleukin-10-deficient cells. J Exp Med 206(7):1549–1564. mice develop chronic enterocolitis. Cell 75(2):263–274. 25. Trandem K, et al. (2011) Virally expressed interleukin-10 ameliorates acute enceph- 50. Griffin DE, Johnson RT (1977) Role of the immune response in recovery from Sindbis alomyelitis and chronic demyelination in coronavirus-infected mice. J Virol 85(14): virus encephalitis in mice. J Immunol 118(3):1070–1075. 6822–6831. 51. Rowell JF, Griffin DE (1999) The inflammatory response to nonfatal Sindbis virus in- 26. Hayasaka D, et al. (2013) TNF-α acts as an immunoregulator in the mouse brain by fection of the nervous system is more severe in SJL than in BALB/c mice and is asso- reducing the incidence of severe disease following Japanese encephalitis virus in- ciated with low levels of IL-4 mRNA and high levels of IL-10-producing CD4+ T cells. fection. PLoS ONE 8(8):e71643. J Immunol 162(3):1624–1632.

16058 | www.pnas.org/cgi/doi/10.1073/pnas.1418966111 Kulcsar et al. Downloaded by guest on September 29, 2021