IL-3 Is a Marker of Encephalitogenic T Cells, but Not Essential for CNS Autoimmunity

Identifying molecules that are differentially expressed in encephalitogenic T cells is critical to the development of novel and specific therapies for multiple sclerosis (MS). In this study, IL-3 was identified as a molecule highly expressed in encephalitogenic Th1 and Th17 cells, but not in myelin-specific non-encephalitogenic Th1 and Th17 cells. However, B10.PL IL-3-deficient mice remained susceptible to experimental autoimmune encephalomyelitis (EAE), a mouse model of MS. Furthermore, B10.PL myelin-specific T cell receptor transgenic IL-3−/− Th1 and Th17 cells were capable of transferring EAE to wild-type mice. Antibody neutralization of IL-3 produced by encephalitogenic Th1 and Th17 cells failed to alter their ability to transfer EAE. Thus, IL-3 is highly expressed in myelin-specific T cells capable of inducing EAE compared to activated, non-encephalitogenic myelin-specific T cells. However, loss of IL-3 in encephalitogenic T cells does not reduce their pathogenicity, indicating that IL-3 is a marker of encephalitogenic T cells, but not a critical element in their pathogenic capacity.

Multiple sclerosis (MS) is a CNS demyelinating disease that is postulated to be mediated by encepha litogenic CD4 + T cells reactive to myelin (1). Most individuals harbor myelinreactive T cells, illustrating that the mere presence of myelinspecific T cells is insufficient to cause MS (2)(3)(4). It is still unclear what characteristics of a myelinspecific T cell make them encephalitogenic and capable of mediating CNS pathology. Identifying molecules that are differentially expressed in encepha litogenic T cells is critical to understanding how they mediate pathology, as well as identifying new and specific therapeutic targets for MS. IL3 is a molecule that is primarily produced by activated Tcells (5), but plays a critical role in the activation and survival of a diverse group of cells, includ ing mast cells and basophils, monocytes, B cells, T cells, and endothelial cells (6)(7)(8)(9)(10)(11)(12)(13). Studies have shown that IL3 contributes to inflammation and autoimmunity (14)(15)(16). The Il3 gene is located on chromosome 11 in mice and chromosome 5 in humans, within the same gene cluster as Csf2, the gene coding GMCSF. There is a potential binding site for the transcriptional factor Tbet close to this gene cluster. Tbet and GMCSF have been shown to be associated with EAE development (17)(18)(19)(20)(21)(22)(23)(24), and so it is possible that the activation of this gene cluster is a contributing factor to T cell encephalitogenicity. A previous study found that increased IL3 expression by CD4 + T cells is associ ated with relapses in MS patients, encephalitogenic CD4 + T cells are the primary source of IL3 in EAE, and the administration of IL3 to mice with EAE exacerbates disease (16,25). IL3 has also been identified as a signature gene of pathogenic Th17 cells in EAE (26), and IL3 was shown to IL-3 in Encephalitogenic T Cells Frontiers in Immunology | www.frontiersin.org June 2018 | Volume 9 | Article 1255 be higher in myelin tetramerpositive memory T cells from MS patients (27). In this study, we analyze the role of IL3 produced by T cells in EAE using B10.PL mice and myelinspecific T cell receptor (TCR) transgenic cells, which allows us to efficiently generate encephalitogenic and nonencephalitogenic T cells with a single in vitro stimulation so that we can compare the role of IL3 in T cell encephalitogenicity (28).

resULts
The encephalitogenicity of myelinspecific T cells varies using different in vitro activation protocols. In Figure 1A, naïve myelinspecific T cells from a MBP Ac111specific TCR trans genic mouse were differentiated with antiCD3/CD28 + IL12 or APCs + MBP Ac111 peptide + IL12 to generate myelin specific Th1 cells. Th1 cells differentiated with APC/Ag + IL12 were highly encephalitogenic, while the myelinspecific Th1 cells differentiated with antiCD3/CD28 + IL12 were significantly less encephalitogenic. Analysis of IL3 by these Th1 cells found that IL3 expression was significantly elevated in these APC/Ag driven Th1 cells ( Figure 1B). Similarly, myelinspecific Th17 cells differentiated with APC/Ag + IL6 + TGFβ were not encepha litogenic, while differentiation of MBPspecific TCR transgenic T cells into Th17 with APC/Ag + IL6 + antiIL4/IL12/IFNγ were highly encephalitogenic (Figures 1C,D), consistent with previous studies (22,28,(31)(32)(33)(34). Similar to the encephalitogenic Th1 cells, IL3 was highly expressed in the encephalitogenic Th17 cells ( Figure 1E). The percentage of IL3 + cells in the activated CD4 + T cell population was enhanced in the encephalitogenic Th1 and Th17 cells (Figure 1F). Since the Il3 allele is located in the same gene cluster with Csf2 (encodes GMCSF) on chromosome 11 in mouse and chromosome 5 in human ( Figure 1G) and GMCSF was found to be critical for encephalitogenic T cells (17,21,23,24), we analyzed GMCSF + activated CD4 T cells via flow cytometry. As expected, GMCSF + cells were increased in encephalitogenic Th1 and Th17 cells (Figure 1H). Given that Il3 and Csf2 are in the same gene cluster, it is unclear whether increased IL3 is due to a specific upregulation of the Il3 gene or an indirect effect due to upregulation of the Il3/Csf2 gene cluster.
To determine whether IL3 was critical in the encephalito genic capacity of myelinspecific T cells, IL3deficient mice were backcrossed with the B10.PL MBPspecific TCR transgenic mice. Since the Il3 and Csf2 alleles are in close proximity, we needed to  Th1 and Th17 conditions ( Figure 2B). For Th1 cells, IFNγ levels were not significantly changed, but there was a reduction in GMCSF levels. There was no difference between Th1 cells cultured with IL3 +/+ or IL3 −/− APCs. There was a reduction in IL17 and GMCSF levels in the Th17 cells, and surprisingly IL17 expression was higher with IL3 −/− APCs. The reduction in GMCSF in both the Th1 and Th17 cells was suggestive of reduced encephalitogenicity since GMCSF was found to be essential in encephalitogenic T cells.
To determine if IL3 deficiency protects mice from EAE development, B10.PL IL3 +/+ , IL3 +/− , and IL3 −/− mice were immunized with MBP Ac111/CFA. The incidence, day of onset, disease course, and maximum clinical course were not signifi cantly different between the groups (Figure 2C). To specifically address the role of IL3 in CD4 T cells, myelinspecific IL3 +/+ and IL3 −/− T cells were differentiated into encephalitogenic Th1 and Th17 cells and transferred into B10.PL mice. The incidence of disease, day of onset, and maximum clinical score were not significantly different between the mice that received the IL3 +/+ and IL3 −/− T cells (Figures 2D,E). There was also no difference in disease course between IL3 +/+ and IL3 −/− Th1 cells (Figure 2D). There was a slight decrease in disease course in the IL3 −/− Th17 cells (Figure 2E), but this was not reproducible.
To validate these data, myelinspecific CD4 T cells were cultured with a neutralizing IL3 antibody during the in vitro dif ferentiation into encephalitogenic Th1 and Th17 cells. Although there was a complete abrogation of IL3 with the antibody (data not shown), there was no change in IFNγ, IL17, or GMCSF expression ( Figure 2F). There was also no reduction in the encephalitogenicity of myelinspecific Th1 or Th17 cells with IL3 neutralization (Figures 2G,H), indicating that IL3 was not essential for the generation of encephalitogenic T cells. These data suggest that IL3 levels are associated with the encephalitogenic capacity of myelinspecific Th1 and Th17 cells, but IL3 plays no major role in the encephalitogenic capacity of the cells.

DiscUssioN
As the role of IL3 in MS is not welldescribed, the goal of this study was to determine if myelinspecific IL3 T cell produc tion was necessary for encephalitogenicity. IL3 expression was significantly higher on encephalitogenic T cells, but loss of IL3 in T cells had no significant effect on the development or severity of EAE. Previous studies have demonstrated that IL3 is highly expressed in encephalitogenic cells (16,25,26), consistent with the data from this study. However, the high expression of IL3 in combination with the high expression of GMCSF, a cytokine located in the same gene cluster, indicates that the upregulation of these two genes may be a product of increased upstream promotor function and not as a causative factor in encephalito genicity. Renner et al. (16) have shown results contradictory to the findings in this study, which may be due to differences in experimental design. Their study found a modest reduction in B6/MOG EAE severity at days 18-20 using a parametric ttest, instead of analyzing EAE using a nonparametric test since the EAE scoring system is not linear. Their study also found that EAE was more severe in IL3 −/− mice than IL3 +/− mice, which is inconsistent with their conclusion. However, they did observe that administration of IL3 in vivo enhanced EAE severity and administration of antiIL3 modestly lessened disease severity, indicating that IL3 may contribute to pathology.
IL3 has been implicated as both a proinflammatory and antiinflammatory cytokine. IL3 has been shown to not only contribute to microglia activation (35) but also protect neurons from inflammation following mechanical strain (36). IL3 transcripts have been shown to be upregulated in CNS lesions of MS patients (37), and chronic expression of IL3 by astrocytes results in an MSlike disease in mice (38). In the present study, we differentiated the effects of systemic loss of IL3 compared to T cellspecific loss of IL3 and found that IL3 was not a major contributor to the incidence or severity of EAE. The observations that IL3 is expressed in encephalitogenic T cells and myelin specific memory T cells from MS patients (27) indicate that IL3 may be a marker of encephalitogenic T cells, possibly due to transcriptional upregulation of the Il3/Csf2 gene cluster, but an unlikely therapeutic target.

ethics stateMeNt
The protocols used for these experiments received prior approval by the OSU Institutional Animal Care and Use Committee and were conducted in accordance with the United States Public Health Service's Policy on Humane Care and Use of Laboratory Animals.
aUthor coNtriBUtioNs PL and MX designed and performed experiments, analyzed data, and assisted with writing the manuscript. WP assisted with the in vivo experiments. YY and ALR designed experiments, ana lyzed data, and prepared the manuscript.

acKNoWLeDGMeNts
The IL3deficient mice were developed by Glenn Dranoff (Dana Farber Cancer Institute) and provided to us by Booki Min (Cleveland Clinic). This study was funded by grants to ALR from the NIH (1 R01 NS067441) and National Multiple Sclerosis Society (RG 3812). PL was supported by award TL1TR000091 from Clinical Translational Science Award, funded by NIH.