Systemic Lack of Canonical Histamine Receptor Signaling Results in Increased Resistance to Autoimmune Encephalomyelitis

Histamine (HA) [2-(4-imidazole) ethylamine] is an important mediator involved in regulating various physiological processes, such as neurotransmission, secretion of pituitary hormones, and gastrointestinal and circulatory functions (1). In addition, HA is a potent mediator of inflammation and regulates innate and adaptive immune responses (2). Histidine decarboxylase (HDC) synthesizes HA through the decarboxylation of histidine, and mast cells and basophils provide the major source of stored HA in the body (3). However, other cellular sources of HA have recently been identified, including dendritic cells (DCs), T cells, neutrophils, and macrophages (4), and induced or nascent HA secretion occurs in conjunction with increased HDC activity in these cell types. HA mediates its effect through binding to four distinct histamine receptors (HRs), namely, H₁–H₄. All four HRs are 7-transmembrane G-protein–coupled receptors (GPCRs). H₁R and H₂R couple to the Gα_q/11 and G_s class of G proteins, respectively, whereas H₃R and H₄R are coupled to G_i/o (1).

HA plays an important role in the development of both allergic inflammation and autoimmune diseases such as multiple sclerosis (MS) and experimental allergic encephalomyelitis (EAE), the principal animal model of MS. HA and HA-releasing agents from mast cells have a dramatic effect on the permeability of the blood–brain barrier (BBB) (5, 6). The use of first-generation H₁R antihistamines, which readily cross the BBB, is associated with a decrease in MS risk (7). MS patients given an H₁R antagonist remained stable and improved neurologically (6). In addition, microarray analysis on the chronic plaques of MS patients revealed increased levels of H₁R transcripts (8). Similarly, in EAE, T cell clones activated against myelin peptides have increased levels of H₁R and H₂R transcripts (9). Mast cell granule stabilizers and H₁R-specific antagonists reduce EAE severity (10, 11), and mice treated with the H₂R agonist dimaprit showed reduced clinical severity and pathological changes (12). In contrast, the absence of HA leads to an elevation in levels of proinflammatory cytokines and increased susceptibility to EAE in HDCKO mice (13). In both MS and EAE, it is well accepted that MHC class II–restricted CD4⁺ T cells, which are capable of secreting either IFN-γ (Th1) or IL-17 (Th17), are necessary and sufficient to induce neuropathological conditions (14).

Using HRKO mice, we have extensively studied the role of HRs in the development of EAE (4, 15–18). H₁RKO mice show a significant delay in the development of EAE and have reduced clinical signs, compared with their wild-type (WT) counterparts (15). During myelin oligodendrocyte gp35–55 (MOG_35–55)–induced EAE, T cells from H₁RKO mice produce significantly less IFN-γ and increased Th2 cytokines (19). H₂RKO mice are also less susceptible to EAE, with a blunted Th1 cytokine response in in vitro recall assays (16). H₃R is an inhibitory auto/hetero-receptor expressed presynaptically on neurons. H₃RKO mice develop severe acute early phase EAE, which supports the existence of a novel H₃R-mediated CNS component in the neurogenic control of BBB permeability and peripheral T cell responses (17). H₄R is predominantly expressed on hematopoietic cells and exhibits diverse functions (20). H₄RKO mice showed increased susceptibility to MOG_35–55–induced EAE in association with decreased CNS regulatory T cell activity (18).

Although the majority of MS and EAE studies have focused on the role of HA signaling through the four known GPC-HRs, evidence does exist for HA signaling through non-GPCRs—for example, GABA_AR—which are ligand-gated ion channels named for their ability to bind the inhibitory neurotransmitter γ-aminobutyric acid (GABA) (21–23). Therefore, to test the hypothesis that HA signaling through noncanonical GPC-HR signaling pathways plays a role in allergic inflammation and the immune responses, we generated mice deficient for the four known HRs (H_1–4RKO) and studied them for susceptibility to EAE. In this article, we report that H_1–4RKO mice develop less severe EAE and neuropathological conditions than WT and HDCKO mice. Furthermore, splenocytes from immunized H_1–4RKO mice produce significantly less IFN-γ, and H_1–4RKO Th1 effector cells are less encephalitogenic under adoptive transfer conditions. Therefore, our data support the possible existence of an alternative HAergic pathway, which, in the absence of GPC-HRs, significantly reduces susceptibility to EAE.

C57BL/6J (B6) mice were purchased from Jackson Laboratory (Bar Harbor, ME). B6.129P-Hrh1^tm1Wat (H₁RKO) (24), B6.129P-Hrh2^tm1Wat (H₂RKO) (16), B6.129P2-Hrh3^tm1Twl (H₃RKO) (17), and B6.129P-Hrh4^tm1Thr (H₄RKO) mice (Lexicon Genetics, Woodlands Park, TX) (25) were maintained at the University of Vermont, Burlington, VT. All the above strains were backcrossed to B6 background for > 10 generations. H_1–4RKO mice were generated by intercrossing individual HR knockout mice (H₁RKO × H₂RKO × H₃RKO × H₄RKO). The genetic background of H_1–4RKO mice was assessed by the DartMouse Speed Congenic Core Facility at the Geisel School of Medicine at Dartmouth (Lebanon, NH). DartMouse uses the Illumina (San Diego, CA) GoldenGate Genotyping Assay to interrogate 1449 single nucleotide polymorphisms (SNPs) spread throughout the genome. The raw SNP data were analyzed using DartMouse’s SNaP-Map and Map-Synth software, allowing for the determination of the SNP allele at each location. The SNP analysis revealed that the H_1–4RKO mice are 97% B6, with minor carryover (3%) from 129 only at the retained knockout loci. HDC^Δ6–8/HDC^Δ6–8 (HDCKO) mice were obtained from Dr. Paul J Bryce, Northwestern University, Division of Allergy-Immunology, Feinberg School of Medicine, Chicago, IL (26). The experimental procedures used in this study were approved by the Animal Care and Use Committee of the University of Vermont.

Mice were immunized for the induction of EAE, using either a 1× or a 2× immunization protocol (27). For the 2× protocol, mice were injected s.c. in the posterior right and left flanks with an emulsion containing 100 μg MOG_35–55 and an equal volume of CFA (Sigma-Aldrich, St. Louis, MO) having 200 μg Mycobacterium tuberculosis H37RA (Difco Laboratories, Detroit, MI); 1 wk later all mice received an identical injection of MOG_35–55–CFA. For the 1× protocol, mice were immunized with an emulsion containing 200 μg MOG_35–55 and an equal volume of CFA containing 200 μg Mycobacterium tuberculosis H37RA. On the day of immunization each mouse received 200 ng pertussis toxin (PTX) i.v. (List Biological Laboratories, Campbell, CA).

For passively induced disease, donor mice were immunized using the 1× immunization protocol, and at day 12 post immunization single-cell suspensions of draining lymph nodes (DLNs) were prepared. Cells (10 × 10⁶ cells/ml in 75-ml tissue culture flasks) were stimulated with MOG_35–55 (10 μg/ml) and IL-12 (0.5 ng/ml) for 72 h. Before transferring the cells into recipient mice, cell viability was assessed by trypan blue exclusion; cells stained for intracellular cytokines and culture supernatant were also screened for IFN-γ, IL-17, and GM-CSF production by ELISA. After 72 h of culture, cells were washed twice at room temperature with PBS, and 1 × 10⁷ cells/200 μl PBS were injected i.v. into B6 recipient mice. The mice were monitored for the onset of EAE for 30 d.

Mice were ranked, scored daily for clinical quantitative trait variables beginning at day 5 after injection, as follows: 0, no clinical expression of disease; 1, flaccid tail without hind limb weakness; 2, hind limb weakness; 3, complete hind limb paralysis and floppy tail; 4, hind leg paralysis accompanied by a floppy tail and urinary or fecal incontinence; and 5, moribund. Assessments of clinical quantitative trait variables were performed as previously described (27).

For ex vivo cytokine assays, mice were immunized using the 2× immunization protocol, spleens and DLNs were harvested on day 10, and single-cell suspensions were prepared (1 × 10⁶ cells/ml) in RPMI 1640 (10% FBS) and restimulated with 50 μg/ml MOG_35–55. Cell culture supernatants were recovered after 72 h and assayed for IFN-γ, IL-4, and IL-17 by ELISA, using anti–IFN-γ, anti–IL-4, and anti–IL-17 mAbs and their respective biotinylated mAbs (BD Biosciences–Pharmingen, San Jose, CA). For proliferation assays, 5 × 10⁵ cells per well in RPMI 1640 were plated on standard 96-well U-bottom tissue culture plates and stimulated with 0, 1, 2, 10, and 50 μg MOG_35–55 for 72 h at 37°C. During the last 18 h of culture, 1 μCi [³H]thymidine (PerkinElmer) was added. Cells were harvested onto glass fiber filters, and thymidine uptake was determined with a liquid scintillation counter.

At day 15 post immunization, animals were perfused with saline, and brains and spinal cords were removed. A single-cell suspension was obtained and passed through a 70-μm strainer. Mononuclear cells were obtained by Percoll gradient (37%/70%) centrifugation and collected from the interphase. Cells were washed and stimulated for 4 h with PMA + ionomycin in the presence of Brefeldin A (GolgiPlug; BD Biosciences). Cells were labeled with LIVE/DEAD ultraviolet-blue dye (Invitrogen) followed by surface staining (CD45 from Invitrogen and CD4, CD8, TCR-γδ, CD11b, and TCR-β from BD Biosciences). Afterward, cells were fixed, permeabilized, and stained for intracellular IL-17A (BD Biosciences) and IFN-γ (Invitrogen).

Single-cell suspensions of thymocytes, lymph node cells, and splenocytes were prepared, and the RBCs were lysed with ammonium chloride. Total numbers of cells were counted using the Advia 120 hematology analyzer (Bayer/Siemens, Tarrytown, NY). For flow cytometric analysis, the cells were washed twice and incubated for 30 min on ice with the desired fluorochrome-conjugated mAbs or isotype control Ig at 0.5 μg/10⁶ cells. For the identification and phenotypic analysis of T_R cells (CD4⁺CD8⁻TCR-β⁺Foxp3⁺), the following surface anti-mouse mAbs were used: anti-CD4 (MCD0417; Caltag); anti-CD8 and anti-CD25 (53-6.7, PC61; BD Pharmingen); and anti-TCR-β and anti-Foxp3 staining set (H57-5987, FJK-16s; eBioscience), according to the manufacturer’s instructions. Viable cells were selected for flow cytometric analysis (LSR II; BD) based on forward and side scatter properties, and analysis was performed using FlowJo software (TreeStar Software).

HA concentrations were assessed using a Histamine EIA Kit according to the manufacturer’s instructions (Cayman Chemicals, Ann Arbor, MI). Briefly, 50 μl derivatization buffer was added to 200 μl undiluted supernatants, followed by the addition of 20 μl derivatization reagent. The samples, controls, and standards were added in duplicate to the plate; 100 μl histamine AChE Tracer was added to each well; and the plate was incubated at 4°C for 24 h. The wells were washed; 200 μl Ellman’s Reagent was added; and the plate was incubated, while shaking, for 30 min in the dark at room temperature. The plate was read at 405 nm, when the maximum binding control wells reached an absorbance of 0.2–0.8.

Blood was collected from H_1–4RKO and B6 mice at day 30 post immunization, and sera were stored at –20°C until analyzed. MOG_35–55–specific IgG Ab was measured by ELISA, as previously described (13). Briefly, 96-well microtiter plates were coated overnight at 4°C with 100 μl MOG_35–55 (0.010 mg/ml) diluted in coating buffer (0.1 M NaHCO₃, pH 9.5). The plates were blocked with PBS/1% BSA (blocking buffer) for 2 h. Dilutions of mouse sera from B6 and H_1–4RKO were incubated in MOG_35–55–coated wells. Ab binding was tested by the addition of peroxidase-conjugated monoclonal goat anti-mouse IgG (Southern Biotechnology Associates, Birmingham, AL), each at a 1:5000 dilution in blocking buffer. Enzyme substrate was added, and plates were read at 450 nm on a microplate reader.

From the lymph node and spleen, CD4⁺ T cells were isolated by negative selection (QIAGEN, Valencia, CA). In culture, purified CD4⁺ T cells (1 × 10⁶ cells per milliliter) were stimulated with anti-CD3 (5 μg/ml) and anti-CD28 (1 μg/ml) mAbs (BD Biosciences–Pharmingen). Supernatants were collected at different time points (24, 48, and 72 h) and analyzed for IFN-γ, IL-4, and IL-2 production by ELISA. CD4⁺ T cells (1 × 10⁶ cells/ml) were polarized toward Th1, Th2, and Th17 effector cells, as previously described (15), and analyzed for IFN-γ, IL-4, and IL-17 production by ELISA.

Statistical analyses as indicated in the figure legends were performed using GraphPad Prism 5 software (GraphPad software).

Previously, we demonstrated that mice lacking individual HRs display differential susceptibility to EAE elicited by immunization with a single injection of MOG_35–55 + CFA + PTX (1× protocol) or two injections of MOG_35–55 + CFA (2× protocol) (15–18). In addition, 1× immunized HDCKO mice, which lack HA, exhibit exacerbated EAE (13). In the current study, we assessed susceptibility to MOG_35–55–induced EAE in H_1–4RKO mice, lacking all four known GPC-HRs, relative to that of WT and HDCKO mice, using the 1× immunization protocol. The severity of the clinical disease courses differed significantly among the strains (F = 277.7; p < 0.0001). Surprisingly, we found that H_1–4RKO mice exhibited a significantly less severe clinical disease course than both WT (F = 307.7; p < 0.0001) and HDCKO (F = 485.4; p < 0.0001) mice. The severity of clinical disease in HDCKO mice was significantly greater than that in WT (F = 74.6; p < 0.0001) and H_1–4RKO (F = 485.4; p < 0.0001) mice (Fig. 1A).

Analysis of EAE-associated clinical quantitative trait variables revealed that the incidence, cumulative disease score, peak score, number of days affected, mean day of onset, overall severity index, and frequency of lethal disease were significantly lower in H_1–4RKO mice than in both WT and HDCKO mice (Table I). Analysis of EAE susceptibility in H_1–4RKO and WT mice, using the 2× immunization protocol, yielded similar results. H_1–4RKO mice, compared with WT mice, exhibit a less severe disease course (F = 82.7; p < 0.0001) (Fig. 1B) and lower incidence, cumulative disease score, peak score, and number of days affected (Table II). Therefore, in contrast to the increase in EAE susceptibility observed in HDCKO mice, H_1–4RKO mice exhibited a dramatic resistance to EAE.

To determine whether the absence of H_1–4R inherently influenced immune cell profiles and HA production, we determined the frequency of different cell types in the central and peripheral immune compartments of naive WT and H_1–4RKO mice. No significant difference was noted in the total number of cells in either the lymph node or the spleen, but H_1–4RKO mice had greater numbers of thymocytes than did WT mice (Supplemental Fig. 1A). Further analysis revealed a higher frequency of CD4 and CD8 double negative cells and a lower frequency of CD4 and CD8 double positive cells in the H_1–4RKO mice compared with WT mice (Supplemental Fig. 1B, 1C). We did not observe any difference between H_1–4RKO and WT thymocytes in the frequency of single positive CD4 and CD8 cells (Supplemental Fig. 1B, 1C). We also analyzed the frequency of different immune cell subtypes in the lymph node and spleen and found no significant differences (Supplemental Fig. 1D–F). The only exception was the frequency of splenic B cells, which was decreased in H_1–4RKO compared with WT mice (Supplemental Fig. 1E). Analysis of the WBC differentials in the peripheral blood of H_1–4RKO and WT mice revealed no significant differences between them (Supplemental Fig. 1G, 1H). Therefore, the absence of the four GPC-HRs inherently affects the frequency of neither T cell subsets nor inflammatory leukocytes in the peripheral immune system, but does exert an influence on total thymic cell numbers and the frequency of double negative and double positive cells. Finally, we measured the level of HA in the plasma and found that H_1–4RKO mice have significantly higher levels of HA than do WT mice (Supplemental Fig. 2).

Previous reports indicated that HRs have a role in T cell differentiation and cytokine production (2). To address whether the lack of HRs has an intrinsic effect on the differentiation and/or cytokine production by CD4⁺ T cells, we stimulated purified CD4⁺ T cells from the spleen and lymph nodes of naive WT and H_1–4RKO mice with plate-bound anti-CD3 and soluble anti-CD28 mAb for 24, 48, and 72 h and screened the culture supernatants for IL-17, IFN-γ, IL-4, and IL-2 production by ELISA. IL-17 was undetectable among the strains. Interestingly, CD4⁺ T cells from H_1–4RKO mice produced less IFN-γ than did WT CD4⁺ T cells at 48 and 72 h (Fig. 2A). However, IL-4 production from the stimulated cells was greater in H_1–4RKO than in WT CD4⁺ T cells (Fig. 2B). We observed no significant difference in the production of IL-2 by these mice (Fig. 2C). These results indicate that CD4⁺ T cells from H_1–4RKO mice upon polyclonal stimulation have an inherent bias toward the Th2 phenotype.

To address whether the lack of HRs can influence CD4⁺ T cell differentiation, we purified CD4⁺ T cells from naive WT and H_1–4RKO mice and in vitro differentiated them into different effector Th cell subsets. In vitro differentiated Th1 effector cells from H_1–4RKO mice produced less IFN-γ than did Th1 effectors from WT mice (Fig. 2D). We found no difference in the production of IL-4 (Fig. 2E) and IL-17 (Fig. 2F) from Th2 and Th17 effector cells, respectively. Thus, under in vitro nonpolarizing or polarizing conditions, CD4⁺ T cells from H_1–4RKO mice produce less IFN-γ, indicating a deficiency in the Th1 response, with a bias toward Th2 under nonpolarizing conditions.

EAE is primarily associated with pathogenic Th1 and Th17 cells (14). HA and HRs play a role in T cell polarization, proliferation, and cytokine production (2), as well as Ab production by B cells (28). Therefore, to elucidate the immune mechanisms associated with differential EAE susceptibility observed in WT and H_1–4RKO mice, we compared MOG_35–55–specific T cell responses on day 10 post immunization. In ex vivo proliferation assays, splenocytes and DLN cells from both strains responded equivalently in a dose-dependent fashion to MOG_35–55 (Fig. 3A). Splenic and DLN cells from H_1–4RKO mice restimulated with MOG_35–55 produced less IFN-γ (Fig. 3C) and trended toward reduced IL-17 (Fig. 3D), compared with restimulated cells from WT mice. We also analyzed IL-4 production by these cells, which is an indicator of an EAE protective Th2 response, but the level was below the limit of detection. Finally, as HA and HRs can influence Ab production and the frequency of B cells (24, 28), and H_1–4RKO mice have slightly fewer B cells (Supplemental Fig. 1E), we measured the MOG_35–55–specific IgG levels in the sera on day 30 after 1× immunization and found no difference in the anti-MOG Ab titers between H_1–4RKO and the WT mice (Fig. 3B).

We also evaluated immune responses in the target organ of 1× immunized WT and H_1–4RKO mice at day 15 post immunization. Mononuclear cells from the CNS were isolated, stimulated with PMA/ionomycin for 4 h in the presence of Brefeldin A, stained for cell surface molecules, and analyzed by flow cytometry for intracellular cytokines. The total number of infiltrating cells and the number and frequency of CD45⁺ CNS-infiltrating cells were comparable between WT and H_1–4RKO mice (Supplemental Fig. 3A, 3B). The frequency of TCRαβ⁺, TCRγδ⁺, CD4⁺, CD8⁺ T cells, and CD11b⁺ cells (Supplemental Fig. 3C–E) was also similar. In contrast, the frequency of IFN-γ–producing CD4⁺ T cells was significantly decreased in H_1–4RKO mice compared with WT mice, whereas that of IL-17–producing cells did not change (Fig. 4A, 4B). We also evaluated the frequency of IFN-γ– and IL-17–producing TCRγδ⁺ and CD8⁺ T cells and found no difference (Supplemental Fig. 3F, 3G).

In addition to the CNS, we also examined the peripheral T cell response on day 15 post immunization. No difference was detected in the frequency of TCRγδ⁺, TCRαβ⁺, CD4⁺, CD8⁺ T cells, Foxp3⁺ regulatory T cells, and CD11b⁺ cells in the DLNs of H_1–4RKO mice compared with WT mice (Supplemental Fig. 3H–K). In addition, no difference was noted in the frequency of IFN-γ– or IL-17–producing cells among CD8⁺ T cells (Supplemental Fig. 3L). As with the CNS, we found that the frequency of IFN-γ–producing CD4⁺ T cells was significantly decreased in H_1–4RKO mice compared with WT mice (Fig. 4A, 4C), with no effect noted on the frequency of IL-17–producing cells. Taken together, these results suggest that a reduced MOG-specific Th1 response underlies the increased EAE resistance observed in H_1–4RKO mice.

In the current study, we observed that upon polyclonal stimulation, during in vitro differentiation, and in ex vivo recall assays, Th1 effector cells from H_1–4RKO mice produced significantly less IFN-γ. Therefore, to address whether the lack of all four GPC-HRs had any impact on the encephalitogenic potential of T cells, we immunized donor WT and H_1–4RKO mice, using the 1× immunization protocol; and at day 12 post immunization, DLN cells were ex vivo restimulated with MOG_35–55 and IL-12 for 72 h. Before adoptive transfer, representative cells were stimulated with PMA/ionomycin + Brefeldin A for 4 h and stained for intracellular IFN-γ and IL-17, and supernatants were screened for IFN-γ, IL-17, and GM-CSF production. Compared with WT restimulated DLN cells, H_1–4RKO cells produced significantly less IFN-γ and IL-17, with no effect on GM-CSF production (Fig. 5A–E). At the end of ex vivo restimulation, an equal number of WT cells were transferred into naive WT, HDCKO, and H_1–4RKO mice. In addition, we transferred restimulated H_1–4RKO cells into WT recipients and monitored them for the development of EAE. Of interest, we found that restimulated cells from WT mice were fully capable of inducing EAE in both WT and H_1–4RKO recipients. In contrast, restimulated H_1–4RKO cells were unable to induce EAE in WT recipients. Furthermore, we observed that the restimulated WT cells transferred into HDCKO mice, compared with WT recipients, also induced EAE, but with an intermediary disease course (Fig. 5F). These results demonstrate that in the absence of H_1–4Rs, T cells are markedly less encephalitogenic.

HA mediates its effects by signaling through the four known GPC-HRs. However, it was shown that HA can signal through non-GPCRs, such as GABA_AR (21–23). Xenopus oocytes or HEK-293T cells transfected with homomultimeric subunits of GABA_AR can form functional HA-gated chloride channels (23). In addition, evidence exists for HA-mediated chloride conductance in the mammalian brain, suggesting the existence of HA signaling through non–GPC-HRs (22). We hypothesized that in the absence of HRs in H_1–4RKO mice HA may similarly signal through GABA_AR expressed by CD4⁺ T cells, thereby diminishing their encephalitogenic capacity through altered cytokine production. To address this possibility, we stimulated purified CD4⁺ T cells from the spleen and lymph nodes of naive WT and H_1–4RKO mice with plate-bound anti-CD3 and soluble anti-CD28 mAb for 72 h in the presence or absence of the GABA_AR antagonist picrotoxin, and screened the culture supernatants for IL-2, IFN-γ, and IL-4 production by ELISA. We observed no difference in the production of IL-2 (Fig. 6A) between WT and H_1–4RKO CD4⁺ T cells stimulated in the presence or absence of picrotoxin. In the absence of picrotoxin, stimulated CD4⁺ T cells from H_1–4RKO mice produced significantly less IFN-γ (Fig 2A) and more IL-4 (Fig. 2B) than did WT CD4⁺ T cells. Blocking GABA_AR in H_1–4RKO CD4⁺ T cells with picrotoxin significantly increased IFN-γ (Fig. 6B) and decreased IL-4 production (Fig. 6C), supporting the concept that HA can exert its effect through GABA_AR expressed in CD4⁺ T cells and alter cytokine production.

In this article, we have assessed the role of HA in the absence of all known GPC-HRs in EAE susceptibility to test the hypothesis that a noncanonical GPC-HR signaling pathway may influence allergic inflammation and immune responses. The results of our study demonstrate that compared with WT and HDCKO mice, H_1–4RKO mice are remarkably resistant to two different protocols of MOG_35–55–induced EAE. In contrast, HDCKO mice, compared with WT and H_1–4RKO mice, exhibit increased susceptibility to EAE. The absence of GPC-HRs results in an intrinsic inability of CD4⁺ T cells to produce IFN-γ and differentiate into Th1 effector T cells. Consequently, decreased susceptibility to EAE in H_1–4RKO mice is associated with decreased IFN-γ production by both MOG_35–55–specific peripheral and CNS-infiltrating CD4⁺ T cells. Furthermore, in passively induced disease, restimulated WT cells transferred into HDCKO mice also induced EAE, but with a reduced disease course compared with that in WT recipients, suggesting a role for HA in the effector phase of the disease. In addition, we show that compared with WT effector T cells, H_1–4RKO effector T cells are markedly less encephalitogenic and produce less IFN-γ and IL-17, with no difference in the production of GM-CSF. Our results are in contrast to the observation that GM-CSF is required for the full encephalitogenic potential of T cells (29–31). Our findings suggest that under these in vitro polarizing conditions, cells from H_1–4RKO mice are less encephalitogenic because of their diminished Th1 and Th17 cytokines. However, it is possible that in vivo within the CNS, infiltrating cells may produce less GM-CSF.

Our results confirm a previous observation showing increased EAE susceptibility in HDCKO mice (13). Because H_1–4RKO mice lack all known GPC-HRs and are therefore deficient in HA signaling, we anticipated that H_1–4RKO mice would be as equally susceptible to EAE as HDCKO mice. Rather, these two models of impaired HA signaling have opposing EAE phenotypes. It is known that the prolonged lack of HA affects the level of HR expression, mast cell granule content, and the development of DCs, all of which are important in EAE. Compared with WT mice, HDCKO MOG_35–55–specific T cells produce more IFN-γ, TNF-α, and IL-6 (13). In addition, DCs from HDCKO mice have greater Ag-presenting activity than their WT counterparts and tend to more readily polarize T cells toward a Th1 phenotype. Furthermore, HDCKO mice have reduced mast cell numbers with drastically decreased granularity (26).

The contrasting results between H_1–4RKO and HDCKO mice suggest a novel HAergic pathway capable of actively regulating resistance to EAE in H_1–4RKO mice or alternate ligand binding to canonical GPC-HRs in the absence of HA that leads to increased disease severity in HDCKO mice. Support of the latter hypothesis comes from a single unreplicated study suggesting that CCL16 may be a low-affinity functional ligand for the H₄R capable of eliciting chemotaxis in human and mouse eosinophils (32). Although CCL16 binding to H₄R in other immune cells expressing the gene has not been studied, it is theoretically possible that this mechanism leads to the increased lymphocyte chemotaxis, production of proinflammatory cytokines, and EAE susceptibility seen in HDCKO. However, we favor the existence of an alternative HAergic pathway capable of actively regulating increased resistance to EAE in H_1–4RKO mice.

We propose that this alternative HAergic pathway may reflect HA acting either through an additional or an as-yet-unidentified high-affinity GPC-HR or through less well characterized HA activation of known receptors and/or signaling pathways. Regarding the existence of an unidentified high-affinity GPC-HR, it is unlikely that such a receptor has gone undiscovered, given the extensive in vitro and in vivo based screens that led to the identification of H₃R and H₄R (33). Support for the latter hypothesis comes from the fact that biogenic amines, including HA, can activate ligand-gated ion channels (34). Ligand-gated ion channels can be either cation channels, which are activated by acetylcholine and serotonin, or anion channels, which are activated by GABA and glycine (35). Evidence exists for HA signaling through GABA_AR, in that homomultimeric subunits of GABA_AR expressed in Xenopus and HEK-293T cells form functional HA-gated chloride channels (23).

Additional evidence for HA activation and signaling through known receptors comes from studies in invertebrates (34, 36–38). Through immunohistochemistry, high levels of HA and HDC were seen in Drosophila photoreceptors (39, 40); however, the functional significance of HA localization in these receptors was not evident until the cloning of two novel HA-gated chloride channels, HisCl-α1 and HisCl-α2 (41). Importantly, the elements that are most important for HA binding and function of HisCl have the greatest mammalian sequence homology with particular subunits of GABA_AR and glycine-gated chloride channels (42). In addition, evidence can be found in mammalian brain for HA signaling through a noncanonical GPC-HR that is picrotoxin sensitive and mediated by chloride conductance (22). Taken together, these observations support the concept that HA can bind to and signal through GABA and/or glycine receptors. However, unlike the GABA_AR β subunits, the glycine receptor β subunits expressed in HEK293 cells do not elicit gating of the channel in response to HA (42).

GABA is a major neurotransmitter in the CNS but is also known to have immunomodulatory activity (43). GABA_AR subunits are expressed by immune cells, including CD4⁺ T cells (44), and the production of IL-6 and IL-12 by peritoneal macrophages can be inhibited by exogenous GABA treatment (45). GABA can also modulate CD8⁺ T cell cytotoxicity and decrease cutaneous delayed-type hypersensitivity reactions (46, 47). Importantly, GABA was shown to reduce myelin basic protein-specific T cell proliferative responses (48), and treatment of SJL/J mice with GABAergic agents delays the onset of EAE in association with decreased production of Th1 and Th17 cytokines (49), which is similar to the H_1–4RKO T cell effector response reported herein. Similarly, low-dose GABA treatment of type 1 diabetes–prone NOD mice inhibits the development of diabetogenic T cells and suppresses disease progression (44). Furthermore, in the current study, we show that purified CD4⁺ T cells from H_1–4RKO mice stimulated with anti-CD3 + anti-CD28 in the presence of the GABA_AR antagonist picrotoxin significantly decreased IFN-γ and increased IL-4 production, suggesting that HA can exert its effect through GABA_AR expressed in CD4⁺ T cells, leading to altered cytokine production. Therefore, we anticipate that HA binding to GABA receptors expressed by T cells has the potential to mediate the increased resistance to EAE in H_1–4RKO mice. Moreover, we observed that H_1–4RKO mice had significantly higher plasma levels of HA than did WT mice. HRs, especially H₂R and H₃R, are known to regulate the levels of HA by feedback mechanisms in various systems (50–52). It is likely that higher plasma levels of HA observed in the absence of canonical HRs in H_1–4RKO mice may bind to low-affinity receptors like GABA_AR. In addition, HA may also act as an endogenous ligand at an unknown allosteric site on the GABA_AR subunit, potentiating the action of GABA (23).

An additional mechanism whereby HA may mediate disease resistance in H_1–4RKO mice is through HA transport. HA transporters are organic cation transporters (OCT) with polyspecificity for transporting molecules across cell membranes. One of these, OCT3, is ubiquitously expressed, and HA uptake through OCT3 has been studied in basophils (53). Once inside the cell, HA is thought to play a role in cell signaling as a second messenger by binding to cytochrome P450 (P450). P450 is a member of a large and diverse group of heme-containing enzymes involved in the metabolism of xenobiotics, lipids, and hormones (54). In liver microsomes, polyamines, hormones, biogenic amines, and antihistamines can inhibit or displace HA binding to P450 (55). Furthermore, HA binding to P450 has been suggested to modulate its catalytic activity and influence cell growth (56). Although the functional significance of HA transport and binding to P450 in the pathogenesis of EAE is unclear, this pathway nevertheless provides for an additional mechanism whereby HA may influence T cell responses and actively suppress EAE.

In summary, we have studied the function of endogenous HA on EAE susceptibility in H_1–4RKO and HDCKO mice, both of which are deficient in HA signaling. Rather than being equivalent in EAE susceptibility and immune responses, as predicted, H_1–4RKO mice were found to be significantly less susceptible to MOG_35–55 induced EAE, whereas HDCKO mice were highly susceptible. Taken together, our findings strongly support the concept that HA acting through mechanisms independent of the four known GPC-HRs mediates increased resistance to EAE. Clearly, delineating the mechanism(s) whereby this alternative inhibitory pathway(s) leads to immune deviation and EAE resistance has the potential to lead to the development of new disease-modifying therapies for both allergic and autoimmune diseases.

We thank Dr. Dimitry N. Kremenstov and members of the Teuscher laboratory for helpful discussions and Drs. Robin L. Thurmond and Timothy W. Lovenberg (Johnson and Johnson Pharmaceutical Research & Development, San Diego, CA) for providing H₃RKO and H₄RKO mice.

The authors have no financial conflicts of interests.