Innate IFN-γ ameliorates experimental autoimmune encephalomyelitis and promotes myeloid expansion and PDL-1 expression

The innate immune system plays a central role in the immune-mediated pathology of multiple sclerosis, and is a therapeutic target for progressive disease. Recently, it has been demonstrated that MIS416, a novel immunomodulatory microparticle that activates NOD-2 and TLR-9-signaling, has disease-modifying activity in multiple sclerosis models. This activity is dependent on innate IFN-γ; however, the precise immune regulatory mechanisms amplified by MIS416 have not previously been determined. Using the experimental autoimmune encephalomyelitis model, MIS416 treatment was associated with IFN-γ–dependant expansion of Treg number and increased suppressive function; however, these cells did not account for disease reduction. Additionally, MIS416 treatment stimulated increased nitric oxide production that was IFN-γ–dependant but dispensable for protection. Finally, MIS416-mediated protection was shown to correlate with IFN-γ–dependant expansion of PDL-1-expressing peripheral myeloid cells, a subset of which was found to be selectively recruited to the brain. This central nervous system trafficking was independent of neuro-inflammatory signals as it occurred in MIS416-treated healthy mice. Together, these findings provide insight into regulatory myeloid cell activities amplified by MIS416-mediated NOD-2 and TLR-9 signalling and highlight the potential importance of these cells in accessing the brain where they may act locally and contribute to the control of neuroinflammation.

Multiple sclerosis (MS) is a neuroinflammatory disease with an autoimmune component that is characterised by activation of self-reactive lymphocytes, which enter the central nervous system (CNS) and cause destruction of myelin producing cells and neurons leading to the formation of inflammatory lesions. A number of immune-modifying therapies are now available to treat the relapsing-remitting form of MS, most of which target the peripheral immune system. Unfortunately, such treatments are largely ineffective in patients with the secondary progressive form of MS (SPMS), supporting the hypothesis that CNS-compartmentalised, innate inflammation is the key driver of SPMS pathogenesis, which appears to be independent of peripheral adaptive immunity. Therefore, to treat progressive MS, new therapeutic strategies are required that have direct anti-inflammatory activity within the CNS and restore CNS homeostasis.
Modification of the innate immune system with MIS416, a TLR9 and NOD2 agonist derived from Propionibacterium acnes 1 , is currently under investigation as a treatment for SPMS. Previous studies have shown that MIS416 has a good safety profile in SPMS patients and works effectively as an immunomodulator 2,3 . Using the experimental autoimmune encephalomyelitis (EAE) model of MS, it has been shown that treatment with MIS416 significantly reduced disease burden in EAE mice when treated prophylactically or therapeutically 4 . MIS416 treatment induced rapid production of high levels of IFN-γ in the serum of treated mice and SPMS patients 4 , and IFN-γ was found to be essential for disease protection in EAE 4 . Whilst IFN-γ is typically considered to be a pro-inflammatory product of effector T cells, it has been also been demonstrated that IFN-γ −/− mice develop more severe EAE than wild type (WT) controls 5 , highlighting alternate, IFN-γ-dependant negative Isolation and in vitro culture of cells. A single cell suspension was prepared from spleen and lymph nodes as previously described 10 , and the number of viable cells were counted based on trypan blue dye exclusion. CD4 T cells were isolated using Dynabeads Untouched mouse CD4 Kit (Life Technologies, USA) and Treg were isolated using MACs CD4+ CD25+ mouse regulatory T cell isolation kit (Miltenyi Biotec, Cologne, Germany) as per manufacturer's instructions.
Cytokine analysis. Murine IFN-γ and IL-10 in culture supernatants were analyzed using a sandwich ELISA.
All reagents were purchased from Becton Dickinson Biosciences and used as per the manufacturer's instructions. Nitric oxide (NO) was measured in culture supernatants using the Griess reaction as previously described 13 and NO levels in the serum were detected using a Nitric Oxide Detection kit (Enzo Life Sciences, Farmingdale, New York) according to the manufacturer's instructions.
In vivo proliferation. Single cell suspensions were prepared from the spleen and lymph nodes of 2D2 mice and CFSE-labelled as indicated previously 7 . CFSE-labelled 2D2 cells (10 × 10 6 ) were injected i.v. into B6.SJL-ptprca mice, and the following day mice were immunized for EAE and treated with MIS416 as previously described 4 . Mice were culled 5 days later and the spleen, blood and draining lymph nodes were processed and analyzed by flow cytometry to assess 2D2 CD4 T cell proliferation, using rat anti-CD45.2 and rat anti-CD4 antibodies to clearly identify donor CD4 T cells.
Statistical analyses. All data were graphed and analysed using GraphPad Prism version 7 (La Jolla, CA, USA). In general, for two group comparisons, unpaired or paired Student's t test was used for parametric data, and Mann-Whitney for non-parametric data. For >2 groups, one-way or two-way ANOVA was used with the

Results
MIS416 reduced disease severity of EAE model and led to an increase in splenic T cell populations. As shown in previous studies 4 , weekly treatment with 100 μg MIS416 i.v. starting on the day of immunization effectively reduced disease severity in the EAE model (Fig. 1a). Analysis of T cell subsets in secondary lymphoid tissue (spleen) on day 22 after disease induction showed a significant increase in the total number of splenocytes in MIS416-treated mice (Fig. 1b). A similar increased was found in healthy MI416-treated mice 15 days after treatment initiation (Fig. 1b). This increase in splenocyte numbers in both EAE and healthy MIStreated mice was in part due to increased numbers of CD4 and CD8 T cells, as well as Tregs, whilst numbers of NK cells were not significantly altered ( Fig. 1c and Suppl Fig. 1).

CD4 T cell proliferation was reduced by MIS416 and was dependant on accessory cells.
Although MIS416 has been shown to reduce antigen-specific cytokine production during EAE 4 , it is not known how MIS416 affects CD4 T cell proliferation in vitro or in vivo. To determine the in vitro effect of MIS416 on CD4 T cell proliferation, MIS416 was added to MOG-stimulated splenocyte cultures from EAE immunized mice, and was shown to suppress MOG antigen-induced CD4 T cell proliferation (Fig. 2a). MIS416 added to cultures in the absence of antigen showed that MIS416 on its own did not support CD4 T cell proliferation (Fig. 2a). Furthermore, splenocyte cultures from healthy mice treated with MIS416 demonstrated reduced T cell proliferation as well as a decreased number of IFN-γ + CD4 T cells in response to antigen-independent stimulation by mitogen (ConA), when compared to untreated, healthy mice (Fig. 2b,c). Paradoxically, the number of IFN-γ + CD4 T cells was significantly enhanced in unstimulated cultures (Fig. 2c) suggesting an in vivo enhancement.
To determine if MIS416 reduced the proliferative capacity of CD4 T cells in vivo, CFSE-labelled CD4 T cells from 2D2 mice, the majority of which express a MOG-specific T cell receptor 15 were injected into congenic (CD45.1) mice prior to EAE immunization. Five days post-immunization, during the period of peak lymphocyte proliferation, we found that MIS416 treatment significantly reduced the number of proliferative events in the  Fig. 2). In contrast, purified CD4 T cells from MIS416-treated and untreated mice showed similar proliferative responses to anti-CD3/CD28 stimulator beads ( Fig. 2f,g). A decreased proliferative index was observed at bead ratios of 1:2 and 1:4 but this decrease was not reflected in the division index suggesting that MIS416 treatment was not directly inducing T cell anergy in the absence of other immune cells. Together, these results indicate that MIS416-mediated suppression of CD4 T cell proliferation is dependent on MIS416-associated alterations in the immune environment of the lymphoid tissue.

MIS416 administration enhanced Treg function in an IFN-γ-independent manner.
Given the significant increase in Treg number by MIS416 treatment (Fig. 1c) and the known capacity of Treg to suppress T cell proliferation, we investigated whether MIS416 treatment also enhanced Treg function. We found that purified splenic Treg from MIS416-treated healthy mice were more effective at suppressing CD4 T cell proliferation in vitro than those purified from untreated mice (Fig. 3a) indicating that MIS416 enhanced Treg function. Since IFN-γ is essential for MIS416-mediated suppression of EAE 4 , we assessed whether the enhancement in Treg number occurred in MIS416-treated IFN-γ −/− mice and found that MIS416 was still able to enhance Treg numbers in the absence IFN-γ (Fig. 3b). Although the expansion was not as dramatic as that seen in WT mice, we believe this reduced expansion may, in part, be due to increased numbers of Treg in the spleens of untreated IFN-γ −/− mice. Furthermore, in vitro CD4 T cell proliferation was not suppressed in splenocyte cultures from MIS416-treated . Splenocytes (1 × 10 6 /ml) were isolated from healthy mice treated with MIS416 15 days previously, labelled with CSFE, and stimulated in vitro with Con A (3 μg/ml), MIS416 (20 μg/ml), or both for 48 hours. The % of CD4 + cells that proliferated (b) and were IFN-γ + (c) was quantified by flow cytometry. Shown are the means and SEM of individual mice (n = 8/group; 4 for Con A + MIS) from 2 independent experiments. ***p < 0.001 and **p < 0.01 by two-way ANOVA with Sidak's multiple comparison test. (d,e) CFSE-labelled 2D2 cells were transferred to congenic mice (CD45.1) one day before MIS416 treatment and EAE immunization. After 5 days, proliferation of CD45.2 2D2 CD4 + T cells in the draining lymph node and spleen was determined by flow cytometry (gating strategy shown in Suppl Fig. 2). Shown are the means and SEM of values from individual mice (n = 9-10/group) from two independent experiments. ****p < 0.0001 by one-way ANOVA with Tukey's multiple comparison test. (f,g). Purified CD4 T cells were isolated from healthy, untreated and MIS416-treated mice at 24 hour post dose 3 (day 15), labelled with CFSE, and stimulated in vitro with anti-CD3/CD28 expander beads at increasing cell:bead ratios for 48 hours. The proliferation (f) and division (g) indexes were measured by flow cytometry. Shown are the means and SEM of duplicate wells from one experiment. ***p < 0.001 and **p < 0.01 by two-way ANOVA with Sidak's multiple comparison test. (Fig. 3c) suggesting that despite the increased presence of Treg, these cells are not the major contributors to the effects of MIS416. Together this data suggests MIS416 suppression of CD4 T cell proliferation is dependent on IFN-γ but Treg are not major contributors.

MIS416-stimulated NO was responsible for the suppression of CD4 T cell proliferation but not disease protection.
In light of MIS416-mediated expansion of splenic macrophages, we investigated their functional status by determining responsiveness to ex vivo innate stimulation with both MIS416 as well as the TLR4 ligand, LPS. Stimulation of splenocytes from MIS416-treated mice with MIS416 or LPS resulted in significantly higher levels of NO, IFN-γ and IL-10 in culture supernatants compared to those from healthy animals (Fig. 5a-c). Consistent with this, serum levels of NO were significantly elevated after MIS416 treatment in EAE as well as healthy mice (Fig. 5d). Additionally, the MIS416-associated increased NO production in vitro and in vivo was abolished in the absence of IFN-γ (Fig. 5e,f) whilst in vitro IL-10 production was not affected (Fig. 5g).
NO plays a role in both EAE disease development and disease pathogenesis; the timing of NO induction being a determinant for whether this molecule acts to enhance or suppress EAE disease [16][17][18][19] . Since NO can act to suppress T cell proliferation 18 , we determined whether MIS416-induced NO was involved in the suppression of EAE. Inhibition of iNOS, the enzyme responsible for NO synthesis, by addition of aminoguanidine to splenocyte cultures from MIS416-treated mice reduced the suppression of CD4 T cell proliferation but did not fully rescue T cell proliferation indicating that MIS416 may exert additional effects on T cells (Fig. 6a). To understand the effect of MIS416-induced NO in vivo, aminoguanidine was administered in the drinking water from the onset of EAE (day 12), but was not found to alter the disease-protective effects of MIS416 (Fig. 6b,c) despite reducing serum NO levels (Fig. 6d). Interestingly, inhibition of NO in vivo did not affect the expansion of red pulp macrophages nor the upregulation of PDL-1 on splenic myeloid cells in MIS416-treated immunized (Fig. 6d-f) or healthy mice (Fig. 6g-i). Taken together, these results indicate that while NO may contribute to the observed suppression of T

MIS416 reduced CNS invasion by immune cells during EAE while promoting the homeostatic trafficking of PDL-1-expressing myeloid populations in an IFN-γ-dependent manner. Recent
work has shown that IFN-γ promotes trafficking of regulatory immune cells 20 into the CNS through the choroid plexus 21 , a point of entry that is predominantly associated with innate immune surveillance and the maintenance of CNS homeostasis. Thus, we assessed the extent to which MIS416 altered CNS leukocyte trafficking during EAE-associated inflammation and also determined whether MIS416 promoted IFN-γ-dependant homeostatic trafficking to the CNS of healthy mice. As expected, MIS416-mediated attenuation of EAE was associated with a significant reduction in the number of CD45 high cells in the spinal cords during EAE (Fig. 7a and Suppl Figs 5 and 6), including a significant reduction in CD4 + and CD8 + T cell populations (Suppl Fig. 5b). In contrast, MIS416 treatment of healthy mice led to an increase in these cells in the spinal cord ( Fig. 7a; see inset) and brain (Fig. 7bf). The increase in CD45 high cells in the brains of MIS416-treated healthy animals consisted primarily of CD4 T cells, macrophages and a Ly6C + CD11b low monocytic myeloid population (Fig. 7b,d,e and Suppl Figs 5 and 7). Significantly, this recruitment was IFN-γ-dependent (Fig. 7b,d, and e), and leukocyte subset specific as MIS416 treatment did not lead to an increase in Treg, neutrophils, or microglia (Fig. 7b,c,f). Lastly, the expression of PDL-1 that was shown to be significantly upregulated on peripheral myeloid cells was also highly expressed on the peripheral myeloid subset (CD11b + CD45 high ) that were recruited to the brain by MIS416 treatment. Analysis of markers that discriminate non-inflammatory/non-classical blood monocytes from their pro-inflammatory/classical counterparts demonstrated these cells had the restricted phenotype of non-inflammatory monocytes cells with upregulation of markers associated with myeloid activation/macrophage maturation (CX3CR1 + , CD11c + , and IA-IE + ; Fig. 7g-k) similar to their CD11b − CD45 high peripheral blood counterpart (Fig. 7l,m). Overall, these results suggest that MIS416 may provide protection from EAE by driving innate IFN-γ production, which enhances homeostatic recruitment of PDL-1-expressing myeloid cells to the CNS, where they may establish a more immune regulatory environment that restricts neuroinflammation.

Discussion
The innate immune system is thought to play a pivotal role in MS pathogenesis, particularly in progressive disease; therefore, targeting innate cell receptors has recently been recognised as a potential therapeutic strategy 22 . Treatment with a NOD-2 and TLR9 agonist, MIS416, has shown success for reducing disease burden in the EAE model when given prophylactically and therapeutically. This protection depended on IFN-γ, a cytokine that was also significantly increased in MIS416-treated secondary progressive MS patients 4 . The dose of MIS416 that effectively reduced EAE disease also induced a number of immunological changes within the innate and adaptive immune compartments which may contribute to its disease-modifying activity. The current study aimed to determine which of these immune alterations were orchestrated by IFN-γ and evaluate their contribution to the mechanism of action of MIS416 in EAE.
It was determined that MIS416 treatment significantly reduced antigen-specific and non-specific T cell responses in vitro and in vivo, however in the absence of IFN-γ this capability of MIS416 to suppress T cells was lost. Additionally, the absence of MIS416-induced IFN-γ abrogated myeloid cell expansion and expression of NO and PDL-1, both of which are likely to contribute to suppressed T cell responses and ultimately, reduced EAE disease 18,23 . While NO was found to have an important function in suppressing T cell proliferation in vitro, depleting NO in vivo did not affect MIS416 disease protection. This result indicates that other factors such as myeloid expression of PDL-1 are likely to have a more central role in the mechanism underpinning suppressed T cell responses in vivo.
Central to immune homeostasis, the ligand PDL-1 is expressed by many different cell types, including both hematopoietic and non-hematopoietic cells, whereas expression of PDL-2 is limited to activated antigen presenting cells such as macrophages and dendritic cells. The engagement of PDL-1/PDL-2 with its receptor, PD-1, on T cells maintains homeostasis by limiting T cell responses during infection and preventing autoimmunity 24 . The PD-1/PDL-1 pathway has been extensively studied recently in the context of cancer immunology as the expression of PDL-1 in the tumour microenvironment is thought to be one of the mechanisms by which cancer cells suppress anti-tumour T cell responses and evade recognition by the host immune system. More recently, clinical trials have shown that treatment with antibodies that block T cell PD-1 signalling, such as Nivolumab, successfully initiate anti-cancer immune responses and increase patient survival 25 . Increased PDL-1 expression has been shown to be important in naturally limiting EAE disease, as work by Latchman et al. showed that PD-L1 −/− mice have enhanced T cell responses in vitro and develop more severe EAE than wild type mice 26 . In addition, PDL-1 expression was increased during the acute phase of disease indicating its role in limiting disease severity 27 . Other studies investigating EAE in PD-1 −/− and PDL-1 −/− mice have identified that the absence of PD-1 signalling Th1 and Th17-type cytokines were increased 23 . Such studies provide strong evidence that the PDL-1/PD1 axis is a pivotal check point that determines T cell fate and support the hypothesis that MIS416-mediated upregulation of the PDL-1 pathway is a key mechanism by which MIS416 reduces the severity of EAE. Notably, enhanced myeloid cell expression of PDL-1 is also associated with recombinant IFN-β treatment, an approved therapy for treating relapse remitting MS 28,29 . The activation of the Type I/II interferon axis by both recombinant IFN-β and MIS416, further highlights the significance of the PDL-1-Type I/II interferon axis to the proposed mechanism of action of MIS416.
In addition to its effects on T cell anergy and apoptosis, PDL-1 expression may be protective in EAE through the induction of Treg. Several studies have shown that the expression of PDL-1 on myeloid cells induces the formation of Treg as well as maintains the suppressive nature of the Treg population 30,31 . In agreement with these findings, we found MIS416-treated mice had increased Treg number and suppressive function in conjunction with increased PDL-1 expression. However, in the absence of IFN-γ while there was very little myeloid PDL-1 expression, Treg expansion still occurred, albeit to a lesser extent, suggesting that Treg expansion alone was not responsible for disease protection.
In addition to suppressed peripheral T cell responses and reduced inflammatory cell trafficking to the CNS, MIS416 was demonstrated to induce IFN-γ-dependent recruitment of regulatory PDL-1 + myeloid cells into the CNS in the absence of neuroinflammation -another mechanism by which MIS416 may exert its effects in vivo. The concept that IFN-γ can enhance CNS-trafficking 20 of non-inflammatory myeloid cells through the choroid plexus is well supported 21,32,33 . Such PDL-1 expressing myeloid cells may be able to suppress T cell activity within the CNS, directly reducing CNS inflammation and recruitment of pathogenic leukocytes from the periphery.
In summary, this work has uncovered a unique mechanism by which an innate-myeloid targeted therapeutic can selectively modulate CNS-trafficking and reduce neuroinflammation. We have found that this mechanism is dependent upon IFN-γ, which appears to mediate its protective actions through myeloid cell expansion Figure 6. Inhibition of NO restored CD4 T cell proliferation but did not alter the ability of MIS416 to inhibit EAE or modulate spleen myeloid subsets. (a) Splenocytes (1 × 10 6 /ml) were isolated from mice 15 days post MIS416 treatment initiation, labelled with CSFE, and stimulated in vitro with Con A (3 μg/ml) +/− MIS416 (20 μg/ml) in the presence of absence of aminoguanidine (AG; 200 mM) for 48 hours. The % of CD4 + cells that divided was quantified by flow cytometry. Shown are the means and SEM of individual mice (n = 8-10/group) from 2 independent experiments. ***p < 0.001 and **p < 0.01 by two-way ANOVA with Sidak's multiple comparison test. (b,c) Mice were treated with MIS416 weekly (i.v.; 100 μg/mouse) starting on day 0, immunized to induce EAE, and scored daily for EAE (b) with the cumulative disease shown as area under the curve (AUC; c). Aminoguanidine (AG; 100 mM) was administered in the drinking water starting on day 12 (dotted line). Shown are the means and SEM of individual mice (n = 9-10/group) from 2 independent experiments. ***p < 0.001 and *p < 0.05 by two-way ANOVA with Sidak's multiple comparisons test (b) and ***p < 0.001 and *p < 0.05 by one-way ANOVA with Tukey's multiple comparisons test. (d-f) Splenocytes and serum were isolated from EAE mice treated with MIS416 (weekly; starting day 0) and/or AG (100 mM; starting day 12) 22 days post immunization. Serum NO (d) was assessed directly, and the total number of red pulp macrophages (RPMϕ; e) and expression of PDL-1 (geoMFI; f) was determined by flow cytometry. Shown are the means and SEM from individual mice (n = 4-5/group) from one experiment. (g-i) Splenocytes and serum were isolated from healthy mice treated with MIS416 and/or AG (100 mM; starting day 0) 15 days post MIS416 treatment initiation. Serum NO (g) was measured directly, and the total number of red pulp macrophages (RPMϕ; h) and expression of PDL-1 (geoMFI; i) were determined by flow cytometry. Shown are the means and SEM from individual mice from two independent experiments (h; n = 7-8/group), and one experiment (i; n = 5/group). (d-i) ****p < 0.0001, ***p < 0.001, **p < 0.01 and *p < 0.05 by one-way ANOVA with Tukey's multiple comparisons test.  and PDL-1 upregulation. This finding is particularly significant as MIS416 has been shown to induce elevated serum IFN-γ in secondary progressive MS patients treated with MIS416 4 . Having shown clinical promise during preliminary trials 2,3 , MIS416 is currently completing a phase 2b exploratory trial in secondary progressive MS. This work provides critical insight into how this novel MS therapeutic may reduce compartmentalized neuroinflammation that occurs in progressive MS. MIS416 has the capacity to favour CNS recruitment of myeloid cells associated with anti-inflammatory/tissue repair activity in the presence of an intact blood barrier and represents the first example of an MS therapeutic that can amplify peripheral, regulatory myeloid cells and licence their trafficking to the CNS 34,35 .