Cutting Edge: Mast Cells Regulate Disease Severity in a Relapsing–Remitting Model of Multiple Sclerosis Free

Multiple sclerosis (MS) is associated with visual, sensory, motor, and cognitive dysfunction and is the most common cause of chronic neurologic disability affecting 2 million individuals worldwide (reviewed in Refs. 1, 2). It is a heterogeneous disease, and clinical presentation ranges from a benign-type MS with little or no progression following initial onset to a primary progressive MS characterized by a relatively acute and severe clinical decline. The most common form, relapsing–remitting MS, accounts for 85% of cases and manifests as a series of discrete episodes separated by complete or partial remission (1, 2). Although the etiology of MS is not well understood, it is widely agreed that myelin-specific autoreactive T cells initiate CNS inflammation and neuronal tissue injury.

Experimental autoimmune encephalomyelitis (EAE) is the prototypic animal model of MS (reviewed in Ref. 3). EAE shares many characteristics with MS, including an early breach of the blood–brain barrier (BBB), focal perivascular inflammatory cell infiltration, and myelin sheath damage leading to progressive ascending paralysis. Analogous to MS, there are multiple disease phenotypes, and these are dependent upon the rodent strain and inducing myelin peptide. For example, primary progressive EAE (PP-EAE) is elicited in C57BL/6 (B6) mice by immunization with myelin oligodendrocyte glycoprotein (MOG_35–55) peptide in CFA and injection of Bordetella pertussis toxin. In contrast, immunization of SJL mice with proteolipid protein_139–151 in CFA results in a relapsing–remitting EAE (RR-EAE) disease course and does not require pertussis toxin. Myelin-specific CD4⁺ Th cells producing IL-17 (Th17) and IFN-γ (Th1) are the major mediators of the pathogenic autoimmune response in both EAE types (reviewed in Refs. 4, 5). Many other cell types, including CD8⁺ T cells, B cells, dendritic cells, monocytes, and mast cells (MCs), have also been implicated in the human and rodent disease (2, 5).

MCs, innate immune cells present in most tissues, are well-established contributors to allergic disease, and there is much correlative evidence that supports MC involvement in amplifying the severity of both MS and EAE (reviewed in Ref. 6). MCs are present in and around MS CNS lesions, and there is clear evidence of their disease-associated activation. For example, transcripts of MC-associated genes, including those encoding tryptase and FcεRI, are present in chronic MS plaques, and MC products including tryptase, histamine, and platelet activating factor are detected in the cerebrospinal fluid of MS patients and increase during relapses (7, 8). Genetically determined susceptibility to EAE induction correlates with increased MC numbers, and MC-stabilizing drugs can decrease disease in rodents (9–13). Direct evidence that MCs contribute to EAE comes from studies showing that MC-deficient WBB6F₁-Kit^W/W-v develop less severe disease than their wild-type (WT) littermates after MOG_35–55 immunization (14). Reconstitution of WBB6F₁-Kit^W/W-v mice with B6 bone marrow-derived MCs (BMMCs) restores MCs to most normal MC-bearing tissues and is sufficient to restore WT levels of disease (14, 15). In this model of PP disease, meningeal MCs, through expression of TNF, promote cellular influx into the CNS and alter BBB permeability (16).

Despite a pathologic role in the MOG-induced model of primary progressive MS, there are several reports of strain-specific variability in murine MC numbers and responses to distinct activation signals (11, 17–19). These data suggest the possibility that genetically disparate MCs can exert variable effects on EAE. In this study, we show that B6 and SJL BMMCs have striking qualitative and quantitative differences in their responses to FcεRI cross-linking as well as to TLR agonists. To test the contribution of MCs in RR disease, MC-deficient SJL-Kit^W/W-v mice were generated. SJL-Kit^W/W-v mice exhibit significantly reduced disease severity and incidence compared with WT mice, but still retain the characteristic RR course. These data confirm that MCs also promote disease in a pertussis toxin-independent and more clinically relevant model of EAE.

WB-Kit^W/+ and B6-Kit^W-v/+ mice were individually backcrossed for at least 10 generations onto the SJL background (The Jackson Laboratory, Bar Harbor, ME), and then SJL-Kit^W/+ and SJL-Kit^W-v/+ progeny were intercrossed to generate SJL-Kit^W/W-v animals.

Mice were immunized as previously described (20) with 100 μg proteolipid protein_139–151 peptide (Genemed Synthesis) emulsified in 5 mg/ml CFA (VWR International, West Chester, PA). Clinical disease was scored every other day as described (20). Cumulative and mean high scores were calculated as described (14). Disease incidence is defined as a score of ≥1 for 2 consecutive days or a score of ≥2 for at least 1 d. A blinded scoring system was used to eliminate subjective bias.

Bone marrow harvest, MC differentiation, and i.v. reconstitution were performed as previously described (14). At >8 wk posttransfer, reconstitution was confirmed by histology.

A standard protocol measuring β-hexosaminidase release was used (21).

BMMCs were stimulated with either anti-DNP IgE (0.5 μg/ml) plus DNP (10 ng/ml), Mycobacterium tuberculosis (100 μg/ml), LPS (5 ng/ml), or peptidoglycan (100 μg/ml) for 2 h. Quantitative real-time PCR was performed according to the manufacturer’s protocol (Quanta BioSciences, Gaithersburg, MD).

Tail DNA was used as a template for PCR-based genotyping of mice. The PCR products were digested by either HphI or NsiI and resolved by gel electrophoresis (Fig. 2A). W mutation primers were: forward: 5′-TTCCTTGCAGAGCAAATCCAG-3′ and reverse: 5′-ATACATGGGTTTCTGGAGGAG-3′; and W^v mutation primers were: forward: 5′-CCTGTACCCACCACCAGTTT-3′ and reverse: 5′-TGATCCACTGAGGGCAGAAT-3′.

The sectioning and staining of most tissues was performed by Histo-Scientific Research Laboratories (Mt. Jackson, VA) after fixation in 10% buffered formalin for 24 h. Peritoneal MCs were analyzed by flow cytometry. Dura maters were prepared as previously described (16).

Hematocrit values were determined as previously described (14).

All statistics were performed using Prism 4 software (GraphPad). For comparison of two groups, an unpaired Student two-tailed t test was used.

SJL- and B6-derived BMMCs were compared for their response to a variety of MC agonists. As shown in Fig. 1A, SJL BMMCs exhibit significantly more degranulation at high Ag concentrations after FcεRI cross-linking as measured by β-hexosaminidase release. There are also notable differences in cytokine production and expression of MC-specific mediators. Although mRNA levels of matrix metalloproteinase 9, IFN-γ, IL-17 (data not shown), and IL-6 are comparable (Fig. 1B), the high LPS-induced expression of IL-1β and TNF in SJL but not B6 cells and the virtual lack of tryptase expression by SJL cells under every activation condition is striking (Fig. 1B). The lower inducible expression of histidine decarboxylase mRNA in B6 MCs correlates with our previous observation that MC production of this molecule, which is required for histamine production, does not contribute to disease in the B6 model of EAE (16). Increased cell-surface expression of programmed death ligand-1 and ICAM-1 was observed in SJL BMMCs and was not dependent on the type or duration (0, 6, or 24 h) of stimulation (Fig. 1C). Disparities in CD40L, VLA-2, VLA-4, OX40 ligand, or glucocorticoid-induced TNFR-related protein expression (GITR) between the two strains was not observed (data not shown).

SJL-Kit^W/+ × SJL-Kit^W-v/+ offspring, unlike WBB6F₁-Kit^W/W-v mice and their WT counterparts, are indistinguishable by coat color and were distinguished using a PCR-based genotyping assay (Fig. 2A). The frequency of mice born with Kit^+/+, Kit^W/+, Kit^W-v/+, Kit^W/W-v genotypes did not meet the expected Mendelian ratios, and only 15% of progeny are SJL-Kit^W/W-v. The low numbers of SJL-Kit^W/W-v mice likely reflect in utero or neonatal death of the double mutant mice and suggest that the c-kit^W/Wv mutations predispose the mice to early developmental abnormalities. Although some SJL-Kit^W/W-v animals appear smaller than their WT littermates at weaning, they are phenotypically identical by 4–6 wk of age.

SJL-Kit^W/W-v animals lack MCs in all tissues examined including the peritoneum (Fig. 2B), stomach, spleen, lung, trachea, liver, brain, lymph nodes (LN), skin, ear, colon, heart, dura mater, and pancreas (Fig. 2C and data not shown). Following BMMC reconstitution of SJL-Kit^W/W-v with WT BMMCs, only some of the tissues where MCs normally reside were repopulated, including the skin, spleen, LN, and dura mater (Fig. 2C). Notably, MC repopulation of the subcranial dura mater was only variably detected. SJL-Kit^W/W-v mice are also neutropenic (Fig. 2D) and anemic (Fig. 2E). T cell compartments in the thymus, LN, and spleen of naive SJL-Kit^W/W-v mice are relatively normal (data not shown).

SJL-Kit^W/W-v mice exhibit significantly decreased disease severity (Fig. 3A), decreased disease incidence (Fig. 3B), and decreased cumulative disease scores (Fig. 3C), all of which are restored to WT levels by BMMC reconstitution (Fig. 3A–C, Supplemental Table I). Thus, as in PP-EAE, MCs play a pathogenic role.

Separate analyses of the 60% of SJL-Kit^W/W-v mice that developed measurable, albeit mild, clinical disease reveal that these mice exhibit an RR course (Fig. 3D). Remission and relapse rates, changes in diseases scores (indices) (Supplemental Table II), and remission kinetics (Fig. 3E) were similar. It has been proposed that disease relapses depend on the release of new myelin epitopes generated by ongoing CNS injury, which in turn activates new waves of encephalitogenic T cells, a phenomenon termed epitope spreading (22). If so, our data indicate that epitope spreading still occurs in the absence of significant MC-dependent CNS damage.

It is possible that the difference in the phenotype of activated B6 versus SJL BMMCs in vitro reflects quantitative and/or qualitative differences in MC influence on disease. Although the proinflammatory contributions of MCs appear to dominate in both models, it is still unclear whether SJL strain-associated differences in histidine decarboxylase, tryptase, or TNF, for example, are relevant. However, the conservation of pathologic MC effects in these distinct models indicates that targeting MCs, in concert with T cells, may lead to more effective therapies for the heterogeneous MS population.

Our original claim of MC contributions to EAE in WBB6F₁-Kit^W/W-v as well as subsequent confirming studies in Kit^W-sh/W-sh mice by others (14, 23) have been refuted by at least one laboratory (24). However, BBB permeability and CNS infiltration, two parameters that can be objectively evaluated, are MC-dependent in the PP-EAE model (16). Furthermore, WT and MC-deficient SJL mice cannot be distinguished by coat color, which provided us a system for unbiased clinical scoring. Taken together, these data firmly establish that MCs regulate EAE.

The establishment of this SJL model has other important implications for future investigations as there is now a system to examine MC contributions in other disease models that are well established in the SJL strain such as autoimmune neuritis, the mouse model of Guillian-Barré syndrome (25, 26). MC-dependent strain-specific differences in susceptibility to Theiler’s virus-induced disease, mouse hepatitis virus, and measles virus as well as in bacterial infections can now also be explored in the SJL versus B6 mice (27–29). These studies will be particularly relevant to human diseases, including MS, in which the genetically determined heterogeneity of cellular responses likely impacts individual susceptibility and disease outcomes.

The authors have no financial conflicts of interest.