Methylprednisolone alleviates multiple sclerosis by expanding myeloid‐derived suppressor cells via glucocorticoid receptor β and S100A8/9 up‐regulation

Abstract Methylprednisolone is an effective drug in the treatment of autoimmune disease, such as multiple sclerosis (MS), due to long‐acting anti‐inflammatory, antiallergic and immunosuppressant. Previous studies have noted the importance of myeloid‐derived suppressor cells (MDSC) in MS progression. However, it is still not known whether methylprednisolone could influence the ratio and function of MDSC during MS treatment. In the current study, we found an increased ratio of MDSC at the onset of EAE in mice model; but methylprednisolone pulse therapy (MPPT) did not alter the percentage and suppressive function of MDSC during disease attenuation. However, the percentage of G‐MDSC in PBMC significantly increased in patients with MS. Surprisingly, relapsing MS patients showed a significant increase in both M‐MDSC and G‐MDSC after MPPT. The disease remission positively correlated expansion of MDSC and expression of arginase‐1. Additionally, MPPT reduced the expression of inhibitory glucocorticoid (GCs) receptor β subunit on MDSC while elevating serum levels of immune regulatory S100A8/A9 heterodimer. Thus, MDSC dynamics and function in mouse EAE differ from those in human MS during MPPT. Our study suggested that GCs treatment may help relieve the acute phase of MS by expanding MDSC through up‐regulating of GR signalling and S100A8/A9 heterodimers.


| INTRODUC TI ON
Methylprednisolone is a potent glucocorticosteroid with extensive therapeutic effect in inflammatory and autoimmune diseases treatment. It can effectively relieve ulcerative colitis, allergies, arthritis, lupus and multiple sclerosis (MS) due to immunosuppressive effects. 1 A considerable amount of literature has been demonstrated the low incidence of side effects of methylprednisolone. 1,2 In MS patients, intravenous methylprednisolone treatment modulates gene expression profiles of CD4 + T lymphocytes, it can also induce Treg cells expansion and reduce proinflammatory cytokine secretion, and thus alleviates MS. 3 However, data about the molecular mechanisms of methylprednisolone in MS treatment still need further research.
Multiple sclerosis is a complex inflammatory autoimmune disease with unknown aetiology. It is established that the pathogenesis of this disease is mainly attributed to abnormally activated T cells triggering an immune response against myelin or myelin-forming cells (ie oligodendrocytes) in the central nervous system (CNS).
Inflammatory cytokines produced by autoreactive T cells that penetrate the CNS after crossing the blood-brain barrier cause damage to the myelin and surrounding tissues. 4 Given the inflammatory nature associated with acute MS lesions, glucocorticoids (GCs) have been used for clinical management of MS. Methylprednisolone pulse therapy (MPPT) are considered in case of moderate functional severity or relapse. 5,6 However, these treatments are often associated with significant side effects. As no drugs are currently available for spe-  7 With their inherent T-cell suppressive activity, MDSC were reported to alleviate the pathology of EAE, 8 We also define the molecular mechanisms underlying MDSC activity in MS patients, which may help guide the development of new therapeutic targets for treatment of MS.

| EAE induction and treatment
C57BL/6 mice were immunized with 50 μg myelin oligodendrocyte glycoprotein peptide 35-55 (MOG35-55) in CFA and treated twice with 400 ng pertussis toxin in total as described. 11 Animals were weighed daily and scored for clinical signs of the disease on a scale from 0 to 5 depending on severity. Treatment with methylprednisolone (Pfizer Manufacturing Belgium NV) was performed by intravenous (iv) injection on three consecutive days at a dose of 100 mg/kg, starting once the mice had reached an average clinical score of 1 score. 13 The dose was halved every 3 days gradually (namely 100 mg/kg over 3 days, 50 mg/kg over 3 days,  The assessment of responses to GC therapy was made by physicians not involved in this study.

| Cell preparations
PBMC were prepared by density gradient centrifugation using Histopaque ® -1083(Sigma, 10831) and Histopaque-1077 (Sigma-Aldrich). Naive CD4 + T cells were isolated from PBMC using the naive CD4 + T cell isolation kit II (Miltenyi Biotec) according to the manufacturer's instructions, and the purity of cells after separation was >98%. MDSC were isolated from PBMC by cell sorting using a cell sorter (Influx; Becton Dickinson). Subsequently, T-cell proliferation was determined by measuring BrdU incorporation (BrdU was added 6 hours prior to cell harvest) (Roche).

| Quantification of serum arginase activity, iNOS and S100A8/A9
Arginase activity was quantified by the QuantiChrom arginase assay kit (Bioassay Systems) according to the manufacturer's instructions and data expressed as enzyme activity (units per litre). Serum levels of human iNOS were measured using an ELISA kit according to the manufacturer's instructions. Serum levels of S100A8/A9 were measured using an ELISA kit according to the manufacturer's instructions (439707; BioLegend).

| Statistical analysis
Statistical analyses were performed on GraphPad Prism 5.0 software. Data are expressed as means ± SD. Between-group comparisons were performed using two-tailed t test, whereas multiple-group comparisons were performed using one-way analysis of variance (ANOVA) followed by the Newman-Keuls test. The Spearman rank test was used for analysis of correlation. A P value of <0.05 was considered significant.

| The expansion of MDSC was independent of GC treatment in EAE
Our previous work showed that MDSC in PBMC and spleen significantly increased at the early stage of EAE and contributed to the pathogenesis of EAE 11 ( Figure 1B-C). To clarify the effect of MPPT on MDSC in EAE, we injected Methylprednisolone intravenously to EAE mice when the score reached average 1( Figure 1A).
Then, we found that GC did not significantly change the levels of  Figure 1F). Additionally, GC treatment did not alter T-cell suppressive activity of either total MDSC or the MDSC subsets ( Figure 1H). Consistent with our previous work, M-MDSC appeared to be more immunosuppressive than G-MDSC from mice with EAE ( Figure S1).

| Increased G-MDSC in RRMS patients
We next examined the frequency of human MDSC and their subsets in PBMC from patients with RRMS. MDSC were defined as CD11b + CD33 + HLA-DR − , which were further divided into SSC low CD14 + CD66b − M-MDSC and SSC high CD14 − CD66b + G-MDSC subsets (Figure 2A). The percentages of MDSC or F I G U R E 1 The expansion of MDSC is independent of GC treatment in EAE. A, EAE models were successfully established. Methylprednisolone (100 mg/kg) was administered after EAE induction on the 13th day (blue arrow, average clinical score = 1). The dose was halved every 3 d gradually (100 mg/kg for 3 d, 50 mg/kg for 3 d, 25 mg/kg for 3 d). The injection was stopped on the 10th day after the initial injection. EAE development in mice (each group n = 10) was followed, and clinical scores were recorded (P < 0.001, MPPT vs EAE). B, Representative staining profiles of MDSC from PBMC from the healthy control (HC) and EAE groups. The cells were collected on the 13th day. C, Percentages of MDSC from PBMC (P < 0.0001) and whole splenocyte population (P < 0.0001) of the EAE and MPPT groups.  (Figure 2A,B). However, the PBMC from RRMS patients showed a significant increase in G-MDSC (Figure 2A,B). Interestingly, while the percentage of G-MDSC in total MDSC increased that of M-MDSC in total MDSC decreased ( Figure 2C).

| Alleviation of MS by GC correlated with expansion of G-MDSC
To determine an impact of GC on MDSC in MS patients, we assessed the changes of MDSC before and after MPPT. We showed that percentages of MDSC in PBMC elevated significantly after GC treatment ( Figure 3A

| Increased serum Arg-1 activity and Arg-1 production by G-MDSC
To better understand the molecular pathways by which MDSC, which expanded in MS patients, exerts their suppressive function, we examined several factors including Arg-1 that is known to mediate the immunosuppressive activity of MDSC. 7 Intracellular staining for Arg-1 showed that G-MDSC express much more Arg-1 compared with M-MDSC ( Figure 4A,B) and production of Arg-1 in G-MDSC was further enhanced by GC treatment ( Figure 4B). We next addressed the question whether GC treatment altered production of Arg-1 by measuring serum Arg-1 activity in MS patients before and after MPPT.
There was no obvious difference in Arg-1 activity between untreated MS patients and HCs ( Figure 4C). However, a significant increase in serum Arg-1 activity was detected in all MS patients after treatment ( Figure 4C). Additionally, our results showed a marked up-regulation of arg-1 mRNA after MPPT in PBMC ( Figure 4D).  Figure 5A). qPCR analysis of PBMC also did not detect a significant change in iNOS mRNA after MPPT ( Figure 5B). As IDO is also produced by MDSC, 15

| Negative correlation between expansion of MDSC and hGRβ expression
To explore how GC treatment expanded MDSC in MS patients, we assessed the expression of human glucocorticoid receptor (hGR), which has two isoforms, hGRα and hGRβ. 16 The expression of total hGR or hGRα on MDSC subsets did not change before or after GC treatment ( Figure 6A,B). However, the levels of hGRβ expression on these cells significantly decreased after GC treatment ( Figure 6A,B), implicating a potential role of hGRβ for modulating MDSC amplification.

| Association of S100A8/A9 with MDSC accumulation in MS patients
Activating transcription factor 3 (ATF3)/S100A9 signalling plays an important role in GC-mediated G-MDSC accumulation in the fatty liver disease. 17 GC can directly induce A8/A9 mRNA in human monocytes or A8/A9-positive cells in the rheumatoid synovial membrane. 18 To examine a potential role that S100A8/A9 protein may play in GC-induced MDSC expansion, serum samples from MS patients were assayed for the levels of S100A8/A9 heterodimer. We showed that methylprednisolone significantly increased S100A8/A9 protein levels ( Figure 7A), which was associated with up-regulation of their mRNA levels relative to untreated patients ( Figure 7B-C).
We further showed that there was a positive correlation between the serum Arg-1 activity and serum S100A8/A9 heterodimer levels in MS patients ( Figure 7D). However, the mRNA expression of cyclic AMP-dependent transcription factor 3 (ATF3), a negative regulator of S100A8/A9, remained unchanged ( Figure 7E). Considering that interferon regulatory factor-7 (IRF7) regulates the development of G-MDSC through S100A9 gene transcription in cancer 19 and modulates progression of EAE in mice, 20 we also compared relative expression of IRF7 mRNA in PBMC of MS patients before and after MPPT. However, there was no change in IRF7 mRNA levels ( Figure 7F), indicating IRF7 may not a crucial factor in regulating G-MDSC development after MPPT.

| D ISCUSS I ON
Glucocorticoids are the most widely used drugs to inhibit immune activation during transplantation and autoimmune diseases. It has been well established in EAE model that CD11b + Gr-1 − MDSC increase at the onset of disease, continue to accumulate prior to disease remission, and contracted upon disease resolution. 9,11,[22][23][24][25][26][27][28] However, we show that CD11b + CD33 + HLA-DR − MDSC in MS patients do not significantly change in the relapsing phase. Indeed, there is a positive correlation between the levels of circulating G-MDSC and disease relapse in MS patients, which is consistent with a previous report. 23 Administration of a synthetic GC dexamethasone (Dex) was previously shown to cause expansion of distinctive proinflammatory monocytes. 33 Accumulating studies suggest that the increased GR expression is involved in this process. [34][35][36][37] While there is only one receptor for GC in mice, human GRs include three isoforms, that is GRα, GRβ and GRγ. GRα is believed to direct the appropriate GC signalling transduction, whereas GRβ and GRγ antagonize GC activity.
Our data show that MPPT significantly reduce expression of GRβ, not GRα on MDSC, suggesting that methylprednisolone induces the expansion of MDSC or G-MDSC by promoting GC-GRα signalling in the acute relapse phase of MS. S100A8 (A8, MRP8, calgranulin A) is generally co-expressed with S100A9 (A9, MRP14, calgranulin B). Recently, it was reported that MDSC and S100A8/A9, co-expressed members of the S100 family of calcium-binding proteins that have an immunoregulatory role, 18,38 can operate through a positive feedback loop to promote tumour development and metastasis. 39 For the first time, we show that methylprednisolone significantly increases S100A8/A9 protein expression that is associated with the serum Arg-1 activity, suggesting that the immunoregulatory S100A8/A9 heterodimer may mediate GC-caused MDSC expansion in MS patients. The transcription factor interferon Regulatory Factor-7 is constitutively expressed at low levels in the cytoplasm 40 and becomes activated by innate receptor signalling, resulting in translocation to the nucleus and induction of type I IFN. 41 Mice lacking either IFN-β or IFNAR develop more severe EAE, with increased CNS infiltration. 42,43 Our results showed that there was no significant change in IRF7 mRNA levels in PBMC of MS patients before and after MPPT, indicating IRF7 may not a crucial factor in regulating G-MDSC development after MPPT. F I G U R E 6 Glucocorticoid receptor expression in two MDSC subsets of relapsing MS patients before and after treatment. A, B, Glucocorticoid receptors (glucocorticoid total receptor, glucocorticoid receptor alpha and glucocorticoid receptor beta) expression in M-MDSC (n = 4 per group; glucocorticoid total receptor P > 0.05; glucocorticoid receptor alpha P > 0.05; glucocorticoid receptor beta P = 0.0483) and G-MDSC (n = 4 per group; glucocorticoid total receptor P > 0.05; glucocorticoid receptor alpha P > 0.05; glucocorticoid receptor beta P = 0.0476) from MS patients before MPPT and MS patients after treatment. Fig A shows staining profiles of cells from a representative MS patient before treatment (blue filled) and a representative MS patient after treatment (red line); grey filled: isotype control. Fig  B shows representative results of four independent experiments In summary, there is a distinction in MDSC from EAE and from MS in that these cells are not primarily involved in MPPT-reduced disease severity in EAE model; however, they play an important role in the acute GC therapy of MS patients. Our findings indicate the pathogenic complexity of MS as well as functional diversity of MDSC during treatment. Additionally, this is the first study that reveals a novel mechanism involving expansion of immunosuppressive MDSC, particularly G-MDSC, which is responsible for GC treatment in the acute relapse phase of MS. Furthermore, our molecular studies uncover that amplification of MDSC by GC may be attributed to GR signalling and S100A8/A9 elevation. Therefore, targeting inherently immunosuppressive MDSC may be exploited for efficacious treatment of T-cell driven autoimmune disorders such as MS.

CO N FLI C T O F I NTE R E S T
The authors declare no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

DATA AVA I L A B I L I T Y S TAT E M E N T
The data used to support the findings of this study are available from the corresponding author upon request.

S U PP O RTI N G I N FO R M ATI O N
Additional supporting information may be found online in the Supporting Information section.