KCa1.1 and Kv1.3 channels regulate the interactions between fibroblast-like synoviocytes and T lymphocytes during rheumatoid arthritis
Fibroblast-like synoviocytes (FLS) and CCR7− effector memory T (TEM) cells are two of the major cell types implicated in the progression of rheumatoid arthritis (RA). In particular, FLS become highly invasive, whereas TEM cells proliferate and secrete proinflammatory cytokines, during RA. FLS and T cells may also interact and influence each other’s phenotypes. Inhibition of the pathogenic phenotypes of both FLS and TEM cells can be accomplished by selectively blocking the predominant potassium channels that they upregulate during RA: KCa1.1 (BK, Slo1, MaxiK, KCNMA1) upregulated by FLS and Kv1.3 (KCNA3) upregulated by activated TEM cells. In this study, we investigated the roles of KCa1.1 and Kv1.3 in regulating the interactions between FLS and TEM cells and determined if combination therapies of KCa1.1- and Kv1.3-selective blockers are more efficacious than monotherapies in ameliorating disease in rat models of RA.
We used in vitro functional assays to assess the effects of selective KCa1.1 and Kv1.3 channel inhibitors on the interactions of FLS isolated from rats with collagen-induced arthritis (CIA) with syngeneic TEM cells. We also used flow cytometric analyses to determine the effects of KCa1.1 blockers on the expression of proteins used for antigen presentation on CIA-FLS. Finally, we used the CIA and pristane-induced arthritis models to determine the efficacy of combinatorial therapies of KCa1.1 and Kv1.3 blockers in reducing disease severity compared with monotherapies.
We show that the interactions of FLS from rats with CIA and of rat TEM cells are regulated by KCa1.1 and Kv1.3. Inhibiting KCa1.1 on FLS reduces the ability of FLS to stimulate TEM cell proliferation and migration, and inhibiting Kv1.3 on TEM cells reduces TEM cells’ ability to enhance FLS expression of KCa1.1 and major histocompatibility complex class II protein, as well as stimulates their invasion. Furthermore, we show that combination therapies of selective KCa1.1 and Kv1.3 blockers are more efficacious than monotherapies at reducing signs of disease in two rat models of RA.
Our results demonstrate the importance of KCa1.1 and Kv1.3 in regulating FLS and TEM cells during RA, as well as the value of combined therapies targeting both of these cell types to treat RA.
KeywordsSynovial fibroblast Immunomodulation Cell interactions Dual therapy Autoimmunity
Carboxyfluorescein succinimidyl ester
Intercellular adhesion molecule 1
Large-conductance calcium-activated potassium channel 1.1
Voltage-gated potassium channel 1.3
Major histocompatibility complex
Receptor activator of nuclear factor κΒ ligand
- TEM cell
Effector memory T cell
Tumor necrosis factor
Rheumatoid arthritis (RA) is a chronic autoimmune disease featuring inflammation centralized within the synovial joints [1, 2]. Despite major advances in treatment strategies for RA, remission remains uncommon and is achieved in only a subset of patients . Current treatments also render patients immunosuppressed and at increased risk for infections . Therefore, there is an unmet need for novel and innovative strategies to treat this disease without further immunosuppressing patients. T cells have a role in disease pathogenesis; in particular, CD4+CD45RA−CCR7− effector memory T lymphocytes (TEM cells) are a primary effector T cell population responsible for the inflammatory aspect of this disease, with characteristic phenotypes of being highly proliferative and the secretion of proinflammatory cytokines within the synovium, leading to joint inflammation [5, 6, 7]. Following activation, TEM cells upregulate the potassium channel Kv1.3 at their plasma membrane, as opposed to naïve and central memory T cells, which primarily express the KCa3.1 potassium channel [6, 7, 8]. Selective blockade of Kv1.3 reduces TEM cell proliferation and cytokine secretion while leaving naïve and central memory T cells able to become activated. Kv1.3-selective blockers are effective at reducing disease severity in multiple animal models of autoimmunity, including in the pristane-induced arthritis (PIA) model of RA, without affecting the clearance of acute infections [6, 7, 9, 10]. As such, Kv1.3 blockers have emerged as promising therapeutics for the treatment of RA [6, 7, 9, 11]. Indeed, ShK-186 (Dalazatide; Kv1.3 Therapeutics, Inc., Seattle, WA, USA), an analog of a venom peptide from the sea anemone Stichodactyla helianthus, is a highly potent and selective Kv1.3 blocker that has shown promise in early clinical trials for the treatment of TEM cell-mediated autoimmune disease [12, 13]. As promising Kv1.3 blockade is as an RA therapy, in experimental models of autoimmunity, ShK-186-treated animals still exhibit signs of joint damage, albeit at significantly lower levels than in control vehicle-treated animals [6, 9]. It is likely that other cell types involved in the pathogenesis of RA and its animal models that do not express Kv1.3 are still active following Kv1.3 block and continue to cause disease progression.
Fibroblast-like synoviocytes (FLS) are resident synovial joint cells that develop a highly invasive phenotype during RA, contributing to joint damage, and secrete a variety of proinflammatory cytokines and chemokines [14, 15]. We have previously demonstrated that FLS from patients with RA and from animal models of RA upregulate the potassium channel KCa1.1 at their plasma membrane [16, 17, 18, 19, 20]. KCa1.1 blockade with selective inhibitors, such as the fungal alkaloid paxilline and the Buthus tamulus scorpion venom toxin iberiotoxin (IbTX), reduces FLS invasion and cytokine and chemokine secretion ex vivo [17, 19, 21]. Furthermore, KCa1.1 blockers reduce disease severity in animal models of RA [17, 21]. However, similar to what is observed in animal models of autoimmunity treated with Kv1.3 blockers, rats with a model of RA treated with a KCa1.1 blocker still exhibit signs of disease, but at lower levels than vehicle-treated animals [17, 21].
FLS and T cells interact within ex vivo settings in which, when stimulated with interferon (IFN)-γ, FLS express major histocompatibility complex (MHC) class II molecules along with the costimulatory molecule B7-H3 (CD276), intercellular adhesion molecule (ICAM)-1 (CD54), and CD40, allowing FLS to serve as antigen-presenting cells to CD4+ T cells [22, 23, 24, 25, 26, 27]. TEM cells also secrete a variety of cytokines, including tumor necrosis factor (TNF)-α, interleukin (IL)-17, and IFN-γ, that are known to induce or enhance the highly invasive, pathogenic phenotype of FLS [28, 29, 30]. Therefore, it is likely that FLS and TEM cells interact during RA to increase each other’s pathogenic features. It may be possible to reduce these interactions through modulating the predominant potassium channels each cell expresses. Importantly, FLS do not express Kv1.3, and the Kv1.3 blocker ShK-186 does not inhibit the RA-FLS pathogenic phenotype, because ShK-186 does not block KCa1.1 channels [19, 31, 32]. Likewise, no T cell populations are known to express KCa1.1, and the KCa1.1 blockers paxilline and IbTX do not block Kv1.3, the potassium channel predominantly expressed by TEM cells [7, 33, 34].
In this study, we show that KCa1.1 is a regulator of MHC class II molecule expression in FLS from the collagen-induced arthritis (CIA) model of RA. KCa1.1 block reduces the CIA-FLS ability to stimulate the proliferation and migration of TEM cells. We further show that blocking Kv1.3 reduces TEM cells’ ability to induce the invasion of CIA-FLS and induce an increase in expression of KCa1.1 and MHC class II molecules on CIA-FLS. Finally, we show that a combined therapy of potassium channel blockers targeting both KCa1.1 and Kv1.3 is more effective than monotherapies at reducing disease severity in two rat models of RA. Our studies highlight the importance of KCa1.1 on FLS and Kv1.3 on TEM cells as moderators of disease severity in RA, and they further validate the use of selective, potent potassium channel blockers as novel therapies for RA.
All experiments involving rats were approved by the Institutional Animal Care and Use Committee at Baylor College of Medicine. Female Lewis rats (8–11 weeks old; Charles River Laboratories, Wilmington, MA, USA) and female Dark Agouti rats (8–11 weeks old; Envigo, Indianapolis, IN, USA) were housed in autoclaved setups in an Association for Assessment and Accreditation of Laboratory Animal Care International-accredited facility in which they were provided food and water ad libitum.
Isolation and culture of FLS
FLS from patients with RA, as defined by criteria of the American College of Rheumatology , were isolated as described previously . FLS from rats with CIA, induced with disease as described below, were isolated 14 days after the rats developed signs of disease, as described previously  by isolating the synovial paw joints, incubating them with Gibco type IV collagenase (Life Technologies, Carlsbad, CA, USA) for 1 h at 37 °C, and culturing adherent cells in DMEM supplemented with 2 mg/ml L-glutamine, 0.1 μg/ml streptomycin, 10 U/ml penicillin, and 10% FBS. CIA-FLS and RA-FLS were considered pure after the third passage of the adherent cells and were used between passages 3 and 10.
KCa1.1 and Kv1.3 channel blockers
The KCa1.1 blocker paxilline was purchased from Fermentek (Jerusalem, Israel), and the Kv1.3 blocker ShK-186/Dalazatide, synthesized under good manufacturing practice conditions by CSBio (Menlo Park, CA, USA), was a kind gift from Kineta, Inc. (Seattle, WA, USA). The KCa1.1 blocker IbTX was synthesized as described previously . Each batch of blockers was tested for channel block by patch-clamping on HEK 293 cells stably expressing KCa1.1 and on L929 cells stably expressing Kv1.3  using a Port-a-Patch automated patch-clamp system (Nanion, Munich, Germany) as described elsewhere [11, 21]. For all in vitro and in vivo studies, potassium channel blockers were used at concentrations known to significantly inhibit the pathogenic phenotypes of FLS and TEM cells and were chosen on the basis of pharmacokinetic and dose-dependence studies [6, 17, 19].
Measuring MHC class II molecule, B7-H3, ICAM-1, and CD40 expression levels in CIA-FLS
CIA-FLS were treated with 100 ng/ml recombinant IFN-γ (MilliporeSigma, Burlington, MA, USA) for 72 h in the presence or absence of 20 μM paxilline. To measure levels of MHC class II molecules, cells were scraped from culture dishes and left either intact or permeabilized with 0.5% saponin, followed by staining with an anti-MHC class II molecule antibody (clone HIS19; LSBio, Seattle, WA, USA), recognizing the RT1L haplotype expressed by Lewis rats , followed by a secondary antibody labeled with the Alexa Fluor 488 fluorophore (Life Technologies). For measurement of B7-H3, ICAM-1, and CD40 expression levels, cells were scraped from their culture flasks and stained with antibodies against B7-H3 (clone MIH42; BioLegend, San Diego, CA, USA) or CD40 (clone 5c3; BioLegend), followed by Alexa Fluor 488-conjugated secondary antibodies, or ICAM-1 (clone HA58; BD Biosciences, San Jose, CA, USA) conjugated to allophycocyanin (APC). Fluorescence was measured using a FACSCanto II flow cytometer (BD Biosciences) and analyzed using FlowJo software (FlowJo, Ashland, OR, USA).
Preparation of TEM cell conditioned medium
Primary Lewis rat ovalbumin-specific CD4+ Th1 TEM cells were provided by Alexander Flügel (University Medical Center, Göttingen, Germany) and maintained in culture as described previously . TEM cell conditioned medium was prepared by stimulating 3 × 106 TEM cells with 150 × 106 irradiated (30 Gy) Lewis rat thymus-derived antigen-presenting cells, which were obtained by making single-cell suspensions of freshly harvested thymuses of healthy Lewis rats  that were loaded with 10 μg/ml ovalbumin (MilliporeSigma) in the presence or absence of 100 nM ShK-186. Supernatants were collected after 48 h of stimulation and stored at −80 °C for later use.
Measuring ex vivo antigen presentation
CIA-FLS antigen presentation was measured as described elsewhere . CIA-FLS (10,000 cells/well) were plated in 24-well plates and treated for 48 h with 100 ng/ml recombinant IFN-γ with or without 20 μM paxilline. Medium was then changed to one containing IFN-γ, paxilline, and either 10 μg/ml ovalbumin or myelin basic protein as relevant or irrelevant antigens, respectively, and cells were cultured for an additional 72 h. CIA-FLS were then washed; 10,000 ovalbumin-specific TEM cells were added to each well; and cocultures were grown for 72 h in medium containing 10 μg/ml ovalbumin or myelin basic protein, but without IFN-γ or paxilline. Proliferation of the TEM cells was measured by [3H]thymidine incorporation as described elsewhere [10, 11, 42].
Imaging conjugates between CIA-FLS and TEM cells
CIA-FLS were stimulated for 72 h with 100 ng/ml IFN-γ and loaded with 10 μg/ml ovalbumin for 2–3 h in the presence of IFN-γ. The CIA-FLS were then labeled with CellTrace Violet dye (Thermo Fisher Scientific, Waltham, MA, USA), and ovalbumin-specific TEM cells were labeled with carboxyfluorescein succinimidyl ester (CFSE) (Thermo Fisher Scientific). The CIA-FLS were then removed from their culture flasks and placed in 1.5-ml Eppendorf tubes with the ovalbumin-specific TEM cells at a 4:1 ratio of TEM cells to CIA-FLS. The cocultures were briefly centrifuged, incubated at 37 °C for 30 min in the presence or absence of either 20 μM paxilline or 100 nM ShK-186, and fixed with 4% paraformaldehyde. Cells were then stained with an anti-CD3 antibody conjugated to APC (antibody clone 1F4; BD Biosciences) and analyzed with an Amnis Imaging Flow Cytometer (MilliporeSigma) to image cell conjugates and the formation of an immune synapse or with a FACSCanto II flow cytometer to quantify cell conjugates, with gating completed to exclude single cells.
Measuring ex vivo T cell migration
CIA-FLS (5000 cells/well) were plated in 24-well plates and stimulated for 24 h with 100 ng/ml IFN-γ with or without 20 μM paxilline. Medium was then changed, and Transwell inserts with 5-μm pores (Corning, Corning, NY, USA) were placed on top of the CIA-FLS-containing wells. Fifty thousand Lewis rat TEM cells were placed in the Transwell, and T cells were allowed to migrate toward the FLS. After 6 h, the number of T cells that migrated through the Transwell and into the well containing the CIA-FLS was counted with a hemocytometer, as described previously .
Measuring ex vivo FLS invasion toward T cell supernatants
CIA-FLS invasion assays were completed as previously described [17, 19, 36, 37]. Briefly, CIA-FLS were placed in the top well of a Matrigel-coated Transwell insert (Corning) in serum-free culture medium. Culture medium containing 20% T cell conditioned medium or 10% FBS was placed in the well beneath the Transwell, and CIA-FLS were allowed to migrate for 24 h. The Matrigel and noninvading cells were then removed, and cells that invaded through the Transwell were counted.
Measuring the effect of T cells or recombinant cytokines on FLS MHC class II and KCa1.1 expression
CIA-FLS (10,000 cells/well) were plated in 96-well plates and incubated for 72 h in TEM cell-conditioned medium supplemented with 10% FBS. For RA-FLS, cells were stimulated for 24 h in the presence or absence of 100 ng/ml recombinant TNF-α (Bachem, Bubendorf, Switzerland), IL-1β (R&D Systems, Minneapolis, MN, USA), receptor activator of nuclear factor κΒ ligand (RANKL) (MilliporeSigma), or IFN-γ (MilliporeSigma). The FLS were then analyzed by flow cytometry for the plasma membrane expression of MHC class II molecules, as described above. For detection of KCa1.1 levels, cells were permeabilized with 0.5% saponin and stained with an anti-KCa1.1α subunit antibody (clone L6/60; NeuroMab/UC Davis, Davis, CA, USA) followed by an Alexa Fluor 488-conjugated secondary antibody, with analysis by flow cytometry.
Inducing, monitoring, and treating rat models of RA
CIA was induced in Lewis rats by a subcutaneous injection at the base of the tail with 200 μl of a 1:1 emulsion of 2 mg/ml porcine type II collagen (Chondrex, Redmond, WA, USA) with incomplete Freund’s adjuvant. Seven days after the first injection, rats were given a booster of 100 μl of the collagen and adjuvant emulsion [21, 44]. PIA was induced in Dark Agouti rats by a subcutaneous injection of 150 μl of pristane (MP Biomedicals, Santa Ana, CA, USA) at the base of the tail [6, 17, 21]. Disease onset was defined as development of at least one swollen or red paw joint. Clinical scores were determined daily as described previously [6, 9, 17] by assigning 1 point for each swollen or red toe joint and 5 points for each swollen wrist or ankle, giving each rat a maximum possible score of 60. Upon disease onset, rats were treated every other day with vehicle (P6N buffer; 10 mM sodium phosphate, 0.8% NaCl, 0.05% polysorbate 20, pH 6.0) , 20 mg/kg paxilline through intraperitoneal injection, 0.5 mg/kg IbTX subcutaneously, 0.1 mg/kg ShK-186 subcutaneously, or a combination of 0.1 mg/kg ShK-186 and either 20 mg/kg paxilline or 0.5 mg/kg IbTX, doses known to reduce disease severity in rat models of RA and chosen on the basis of pharmacokinetic studies and dose-dependence studies [6, 17, 21]. Randomization of rats to treatment groups was completed in which every fourth rat that developed signs of disease on a given day was placed in the same treatment group, thereby disregarding differences in basal disease severity on the day each rat developed signs of disease and accounting for differences in the time between immunization and when a rat developed signs of disease.
Measuring KCa1.1 expression by FLS from rats with CIA
FLS were isolated from rats with CIA in each treatment group 14 days after disease onset. They were permeabilized with 0.5% saponin and stained with an anti-KCa1.1α antibody (NeuroMab), followed by an Alexa Fluor 488-conjugated secondary antibody, and cells were analyzed by flow cytometry as described above. Linear regression analysis between KCa1.1α staining intensity of FLS from individual rats and each rat’s clinical score on the day of cell collection was completed with Prism software (GraphPad Software, La Jolla, CA, USA).
T cell phenotyping of lymph node cells of rats with CIA
The draining inguinal lymph nodes were collected from rats with CIA in each treatment group 14 days after disease onset. Single-cell suspensions were made of each lymph node and stained for expression of CD3 conjugated to APC (clone 1F4; BD Biosciences), CD4 conjugated to phycoerythrin (PE)-cyanine 7 (clone W3/25; BioLegend), CD8 conjugated to PE (clone OX-8; BD Biosciences), CD25 conjugated to fluorescein isothiocyanate (clone OX-39; BD Biosciences), and CD45RC conjugated to biotin (clone OX-22; Thermo Fisher Scientific) and further stained with streptavidin conjugated to Alexa Fluor 405. Cells were analyzed by flow cytometry as described above.
X-rays and histology
Rats with either PIA or CIA were killed after either 21 or 14 days of treatment, and their paws were collected for histology and x-ray analysis. X-rays were completed using an In-Vivo Xtreme Imaging System (Bruker BioSpin, Billerica, MA, USA) on the hind paws and were used by an investigator blinded to treatment groups to assess the presence or absence of abnormal bone structures and erosions. Following collection, hind paws were fixed, decalcified, embedded in paraffin, sectioned, and stained with either H&E or Safranin O/Fast Green. Images of synovial paw joints were taken with an Olympus BX41 microscope equipped with an Olympus Q Color 5 camera at 10× magnification (Olympus, Center Valley, PA, USA). Scoring of disease parameters from histology was completed by an investigator blinded to treatment groups, as described elsewhere , in which immune infiltrates, pannus extensions, hyperplasia, and cartilage erosions were quantified by giving a score of 0 when absent, a score of 1 when mild, a score of 2 when moderate, and a score of 3 when a severe amount of these parameters was present.
Serum chemistry and cytokine analyses
At the end of the CIA trial, blood was collected by terminal cardiac puncture from each rat, and serum was stored at −80 °C. Blood chemistry panels were completed on five randomly selected rats per group through the Center for Comparative Medicine at Baylor College of Medicine by investigators blinded to each rat’s treatment. Serum cytokine analyses were completed on five or six randomly selected rats per group by using a rat cytokine array (Eve Technologies, Calgary, AB, Canada) in a blinded manner.
All statistical analyses were completed by using the Mann-Whitney U test or Wilcoxon matched-pairs signed-rank test, except for the CIA and PIA clinical scores analyses, which were completed using repeated measures one-way analysis of variance with Bonferroni post hoc test. All data are shown as mean ± SEM, and p values less than 0.05 were considered significant. All statistical analyses were completed using Prism software.
KCa1.1 regulates the ability of CIA-FLS to activate and attract TEM cells
Besides attracting T cells toward them, FLS can stimulate the proliferation of CD4+ T cells by acting as antigen-presenting cells . To confirm the antigen-presenting capability of FLS, we first verified that CIA-FLS physically bind to and form cell-cell conjugates with CD4+ TEM cells. CIA-FLS were stimulated with IFN-γ for 72 h, loaded with ovalbumin, and labeled with the CellTrace Violet fluorescent dye. They were allowed to interact with ovalbumin-specific Lewis rat CD4+ TEM cells labeled with the CFSE fluorescent dye. We observed physical interactions between CIA-FLS and TEM cells, with CD3 accumulation observed at the contact site between the CIA-FLS and TEM cells (Fig. 1b), suggesting formation of an immune synapse between the two cell types [50, 51]. To determine if either KCa1.1 or Kv1.3 block prevents the formation of cell conjugates, either paxilline or ShK-186 was added to the mixtures of IFN-γ-treated CIA-FLS and TEM cells while the cells were allowed to interact. Neither paxilline nor ShK-186, alone or in combination, acutely altered the proportion of cell conjugates (Fig. 1c and d).
To confirm that FLS can stimulate the proliferation of CD4+ T cells and determine if KCa1.1 block reduces the FLS ability to stimulate T cell proliferation, CIA-FLS were stimulated with IFN-γ in the presence or absence of paxilline, loaded with ovalbumin, and cocultured with syngeneic ovalbumin-specific CD4+ TEM cells. The cocultures of ovalbumin-specific TEM cells with CIA-FLS stimulated with IFN-γ and loaded with ovalbumin exhibited significantly greater proliferation than the cocultures containing T cells with CIA-FLS stimulated with IFN-γ in the presence of paxilline and loaded with ovalbumin. Loading IFN-γ-stimulated CIA-FLS with an irrelevant antigen (myelin basic protein) failed to induce the proliferation of the ovalbumin-specific TEM cells (Fig. 1e).
KCa1.1 regulates surface expression of MHC class II by FLS
RA-FLS also express the costimulatory molecules B7-H3, ICAM-1, and CD40 [24, 25, 27]. We examined their expression by CIA-FLS following stimulation with IFN-γ in the presence or absence of paxilline. All of these proteins were present in CIA-FLS, and neither IFN-γ nor paxilline had an effect on the proportion of cells expressing them (Fig. 2e). This result was confirmed in RA-FLS (data not shown). We also observed no differences in the expression levels of these proteins between treatment groups, as determined by mean fluorescence intensity, in both RA-FLS and CIA-FLS (data not shown).
TEM cells and Kv1.3 regulate ex vivo phenotype of CIA-FLS
Invasiveness is a hallmark of aggressive FLS during RA and its animal models and is enhanced by proinflammatory cytokines such as IFN-γ, IL-17, and TNF-α [28, 29, 30]. We therefore examined the influence of the TEM cell conditioned medium on CIA-FLS invasion through Matrigel-coated Transwell inserts. CIA-FLS exposed to conditioned medium from antigen-stimulated TEM cells were significantly more invasive than those exposed to medium from unstimulated TEM cells, indicating that activated TEM cells can enhance CIA-FLS invasion. However, CIA-FLS exposed to conditioned medium of TEM cells stimulated in the presence of ShK-186 were significantly less invasive (Fig. 3d), indicating that Kv1.3 activity regulates the TEM cell ability to induce FLS invasiveness.
A combined therapy of Kv1.3 and KCa1.1 blockers is more effective than monotherapies in reducing disease severity in CIA
We previously found that monotherapies with either Kv1.3 or KCa1.1 blockers reduce disease severity in rat models of RA, but that each monotherapy does not completely stop disease [6, 7, 11, 17, 21]. This is presumably due to each monotherapy only directly inhibiting either TEM cells or FLS while leaving other pathogenic cell types intact. We therefore sought to determine if a combined therapy of Kv1.3 and KCa1.1 blockers could work in synergy to further ameliorate disease severity in animal models of RA.
X-rays of the hind paws of rats from each treatment group indicated that vehicle-treated rats developed significant bone erosions around the synovial joints, which were reduced by both the monotherapies and the combined therapy (Fig. 4b). Safranin O/Fast Green staining of tissue sections from paw joints in each treatment group demonstrated fewer cartilage erosions in the synovial joints of monotherapy-treated rats and in the combined therapy animals. H&E staining showed that immune infiltrates were reduced in both the monotherapy-treated and combined therapy-treated rats (Fig. 4b and c). We also collected serum from CIA rats in each treatment group 14 days after disease onset and from healthy rats for serum chemistry analyses, which are summarized in Additional file 1: Table S1.
FLS from potassium channel blocker-treated CIA rats have decreased pathogenic phenotypes
Potassium channel blockers alter T cell populations in CIA rats
A combined therapy of Kv1.3 and KCa1.1 blockers is more efficacious than monotherapies in treating PIA
Twenty-one days after disease onset, rats with PIA were killed, and FLS were isolated from ankle and toe joints of each rat. Invasion through Matrigel indicated that FLS from the potassium channel blocker-treated PIA rats were significantly less invasive than FLS isolated from vehicle-treated PIA rats (Fig. 8b). Paws were collected, and bone damage was assessed on x-rays. Similar to our results with CIA, rats with PIA treated with vehicle exhibited a significant amount of bone erosion around the synovial joints, which was reduced in the potassium channel blocker-treated rats (Fig. 8c). Safranin O/Fast Green staining and H&E staining of paw joints indicated that cartilage damage and immune infiltrates were reduced in the monotherapy-treated PIA rats, and were even further reduced in the combined therapy-treated PIA rats (Fig. 8c).
We used FLS isolated from Lewis rats with CIA to examine their in vitro interactions with Lewis rat TEM cells in this study. Others have used human FLS isolated from patients with RA and examined their interactions with T cells isolated from peripheral blood of healthy volunteers with superantigens used to stimulate T cells [23, 24, 25, 26], or they have used antigen-specific T cell hybridomas developed from transgenic mice expressing a particular human MHC class II allele . Although using human cells is ideal for the study of disease processes in RA, the mixing of FLS and T cells from different individuals or species could lead to the observation of phenotypes that would not be observed naturally in patients with RA. For this reason, and because of the difficulty in obtaining both FLS and sufficient numbers of TEM cells from the same individuals, we used FLS and TEM cells from inbred Lewis rats. The choice of model species is carefully made because ion channel phenotype, function, and response to pharmacological agents vary between species . We chose rats for this work because the phenotype and function of KCa1.1 in FLS are conserved between humans and rats, as are the phenotype and function of Kv1.3 in T cell subsets in these two species. In contrast, mouse and human T cells diverge in potassium channel expression , precluding the use of mice in the present study.
The conditioned medium from activated TEM cells induced an increase in MHC class II and KCa1.1α protein expression in CIA-FLS. In addition, CIA-FLS invaded through Matrigel toward the conditioned medium, indicating that activated TEM cells secrete cytokines which increase the pathogenic features of FLS and chemokines which direct FLS motility. Because the TEM cells used in these experiments are Th1 cells , the cytokines responsible for these changes in FLS phenotype were likely IFN-γ or TNF-α, or a combination of both. IFN-γ is the only known cytokine to induce MHC class II expression in FLS , whereas both IFN-γ and TNF-α enhance FLS invasion [28, 29, 30]. It is likely that one or both of these cytokines induced FLS invasion. These results may also indicate that blocking Kv1.3 can be a novel way to locally interfere with TNF-α production, a key disease mediator, without systemically suppressing a patient’s immune system.
The mechanism by which KCa1.1α expression is regulated in FLS, and how it becomes upregulated during disease, is unknown. However, it is likely through a cytokine-mediated pathway, because our data show that the medium of stimulated TEM cells can induce its upregulation in CIA-FLS and that recombinant proinflammatory cytokines (IFN-γ, TNF-α, and IL-1β) induce KCa1.1α upregulation by RA-FLS. Interestingly, the osteoclastogenic cytokine RANKL, which is secreted by RA-FLS , also induces KCa1.1α upregulation in RA-FLS, leading to a possible autocrine or paracrine regulation of KCa1.1. Furthermore, Kv1.3 blockade on TEM cells prevents secretion of some proinflammatory cytokines by RA synovial fluid TEM cells . The conditioned medium of TEM cells stimulated in the presence of the Kv1.3 blocker ShK-186 did not alter FLS invasion, KCa1.1α expression, or MHC class II expression compared with conditioned medium from unstimulated TEM cells. This further verifies that the cytokines secreted by TEM cells alter FLS phenotypes and that Kv1.3 can regulate these phenotypes.
IFN-γ-stimulated CIA-FLS induced the migration of TEM cells toward them, and blocking KCa1.1 on CIA-FLS reduced this effect. This indicates that upon IFN-γ stimulation, CIA-FLS secrete chemokines that influence T cell migration. CIA and RA-FLS are known to produce a variety of T cell chemoattractants upon cytokine stimulation, including CXCL10 [46, 49], fractalkine , IL-6, IL-8, CCL2, CCL5, and CXCL12 . KCa1.1 blockade prevents FLS from secreting the chemokines IL-8 and VEGF, but not IL-6 . It is therefore likely that KCa1.1 block also prevents FLS secretion of one or more T cell chemoattractants. In order to determine which particular chemokines are responsible for the FLS-induced T cell migration, it is necessary to first identify the complete milieu of chemokines secreted by IFN-γ-stimulated FLS and determine which of these chemokines are regulated by KCa1.1.
We found that IFN-γ induced an increase in MHC class II protein expression on CIA-FLS, in concordance with previous studies in RA-FLS [23, 26, 53]. Interestingly, cotreatment with the KCa1.1 blocker paxilline reduced the plasma membrane expression of MHC class II molecules, but not the total amount of MHC class II protein present in FLS. These data show that KCa1.1 activity is not necessary for IFN-γ to induce MHC class II protein production, but it is required for its localization to the plasma membrane. The mechanism underlying this observation has yet to be determined, but it may involve KCa1.1 serving as a regulator of protein and vesicle trafficking, cytoskeletal rearrangements, or membrane fluidity and turnover within FLS, which could all result in this channel regulating protein expression at the cell’s surface.
Though stimulating FLS with IFN-γ has been a widely studied means to induce the antigen-presenting phenotype of FLS [22, 23, 24, 26, 53], the relevance of FLS antigen presentation and IFN-γ to disease severity and the pathophysiology of RA is subject to debate. Indeed, TNF-α, and not IFN-γ, is generally seen as the prototypical driver of inflammation in RA, and these two cytokines exhibit a mutual antagonism on their effects on FLS phenotypes, including MHC class II protein expression . IFN-γ also reduces IL-1β-induced matrix metalloprotease secretion in FLS, limiting their ability to degrade cartilage . IFN-γ receptor-knockout mice, or mice treated with neutralizing antibodies against IFN-γ, have accelerated onset of CIA, higher disease severity during CIA, and increased Th17 cells [56, 57]. Furthermore, recombinant IFN-γ progressed through several clinical trials as a potential RA therapy [58, 59, 60]. However, IFN-γ and its receptor expression are upregulated in the synovium of patients with RA compared with patients with osteoarthritis , and IFN-γ promotes FLS motility and invasion [29, 30]. Therefore, it remains to be determined how relevant these observations are within the actual disease state. Furthermore, cytokine receptors can have converging signaling pathways and other proinflammatory cytokines that activate the same signaling cascades as those downstream of the IFN-γ receptor and may induce the same phenotypes observed in FLS following IFN-γ stimulation .
Determining if FLS can present antigen on MHC class II and stimulate CD4+ T cells in vivo and thereby serve as nontraditional professional antigen-presenting cells may reveal important information regarding inflammation in RA, because it would reveal FLS as an in situ activator of T cells at a site of inflammation. This could have major implications regarding the importance of FLS and IFN-γ in initiating inflammation in RA, because it is possible that localized increases in IFN-γ levels, perhaps from inflammation due to an infection or physical insult to the joint, could activate FLS antigen presentation and therefore initiate T cell-driven inflammation.
We found that several cytokines were elevated in the serum of vehicle-treated rats with CIA compared with healthy rats and that ShK-186 and IbTX normalized their levels. These included IL-4, IL-12, IL-17A, MCP-1, and TNF-α. Given that blocking Kv1.3 on TEM cells reduces their secretion of cytokines, it was expected that ShK-186 treatments would reduce circulating cytokine levels. The mechanisms by which IbTX reduced these cytokines is less clear, though it is possible that inhibiting FLS indirectly causes a reduction in T cell cytokine secretion. Several cytokines were also found in decreased concentrations in the serum of rats with CIA, including fractalkine, leptin, and RANTES, among others. Fractalkine and RANTES are increased in the serum of patients with RA, and their inhibition decreases disease severity in mouse CIA and rat adjuvant-induced arthritis, respectively [63, 64, 65, 66]. The mechanisms by which these cytokines decrease in rat CIA remains unknown. Leptin is also associated with RA and is increased in the serum of patients with RA. However, its role in rodent models of RA is less clear and may be decreased in the serum during mouse models of RA . Our data suggest that a similar trend exists in rat models of RA, and overall our data indicate potential divergences in the pathogenesis between RA and its animal models.
We found that rats with either PIA or CIA treated with a Kv1.3 blocker or a KCa1.1 blocker had reduced disease severity, in agreement with previous studies [6, 17]. However, tandem therapies of Kv1.3 and KCa1.1 blockers were even more beneficial than the monotherapies, indicating the value of directly targeting both TEM cells and FLS as a novel and potent therapeutic approach to treating RA. Furthermore, because the combined therapy of Kv1.3 and KCa1.1 blockers was more beneficial than the monotherapies, it can be inferred that the monotherapies do not have large downstream effects on the reciprocal cell type through limiting either the TEM cell-induced increase in FLS pathogenicity when blocking Kv1.3 or limiting the FLS-induced increase in TEM cell pathogenicity when blocking KCa1.1. This also implies that both FLS and TEM cells drive disease progression, as opposed to one cell type inducing the pathogenic phenotype of the other.
However, through examining the ex vivo invasiveness of FLS isolated from rats of each treatment group, we found that FLS from arthritic rats treated with ShK-186 had reduced invasion compared with FLS from vehicle-treated animals. FLS do not express Kv1.3, and ShK-186 does not block KCa1.1 and does not have an effect on FLS phenotypes [19, 31, 32]. This indicates that our observations in the present study were not a result of ShK-186 directly affecting FLS. Therefore, our data show that even though the FLS were not directly inhibited by ShK-186, the decrease in TEM cell pathogenicity did have at least some downstream effects on FLS in vivo. Similarly, TEM cells do not express KCa1.1, and paxilline and IbTX do not block Kv1.3 [7, 33, 34]. Therefore, the effects of KCa1.1 blockers in ameliorating disease in PIA and CIA were not due to directly inhibiting TEM cells.
The populations of T cells within the draining inguinal lymph nodes of rats with CIA treated with potassium channel blockers were altered compared with those of vehicle-treated rats. For example, those treated with ShK-186, IbTX, or both ShK-16 and IbTX had an increase in the proportion of CD8+ T cells compared with those from vehicle-treated rats with CIA. CD8+ T cells may have a protective role in RA , and our data suggest that CD8+ T cells are correlated with a decreased disease burden. We also found large differences in the proportion of CD4−CD8− T cells in the draining inguinal lymph nodes of rats with CIA, in which those treated with IbTX in the presence or absence of ShK-186 had dramatically reduced populations of these cells compared with vehicle-treated animals. Interestingly, although the proportion of these cells were decreased in these rats, they expressed more of the activation marker CD25 than vehicle-treated rats with CIA. The role of CD4−CD8− T cells in RA and the roles of Kv1.3 and KCa1.1 in them is not well understood. Overall, our studies suggest that both Kv1.3 and KCa1.1 blockers affect T cell populations in CIA, though the mechanisms by which they do so, along with the roles of some of these cells, are yet to be investigated.
The ex vivo invasion of FLS is directly correlated with disease severity in patients with RA [37, 69, 70], and we found that FLS isolated from rats with PIA or CIA that were treated with Kv1.3 or KCa1.1 blockers, alone or in combination, exhibited reduced ex vivo invasion compared with vehicle-treated rats. These results agree with previous findings regarding FLS invasion as a measure of disease severity and joint destruction . However, although the rats with PIA or CIA that were treated with a combined therapy of KCa1.1 and Kv1.3 blockers had reduced disease severity compared with monotherapy-treated rats, the ex vivo invasiveness was approximately the same. This may be due to limitations of the technique used to measure invasion, because the method we used involved measuring the number of cells that invaded through Matrigel at a single time point and did not account for differences in their rates of invasion.
Overall, these studies provide further insights into the role of FLS and T cell interactions during RA and the importance of the potassium channels these cells express as mediators of these interactions. We also validated a novel therapeutic approach to treating RA by simultaneously inhibiting FLS and TEM cells through targeting the predominant potassium channels by which these cells are regulated. In doing so, we further validated the central role of FLS and TEM cells in the pathogenesis of RA and the importance of KCa1.1 and Kv1.3 in driving disease progression.
This work was supported by National Institutes of Health award NS073712 (to CB and MWP) and Arthritis Foundation award 6483 (to CB and MWP). MRT was supported by T32 awards GM088129, AI053831, and HL007676 and F31 award AR069960 from the National Institutes of Health. The Cytometry and Cell Sorting, Mouse Phenotyping, and Pathology & Histology cores at Baylor College of Medicine are supported in part by funding from the National Institutes of Health (grants HG006348, RR024574, and CA125123) and the Dan L Duncan Comprehensive Cancer Center at Baylor College of Medicine.
Availability of data and materials
The data supporting the authors’ conclusions are included in the article.
MRT and CB designed the study. MWP, SSC, TL, and PSG provided critical reagents. MRT, MWP, and CB performed experiments. MRT and CB wrote the manuscript with input from all authors. All authors read and approved the final manuscript.
The experiments involving the use of rats were approved by the Institutional Animal Care and Use Committee at Baylor College of Medicine. De-identified human FLS were isolated by PSG’s group after patients provided written consent for their tissues to be studied, as approved by the Institutional Review Board at the Feinstein Institute for Medical Research. Baylor College of Medicine’s Institutional Review Board determined that the study of these cells did not constitute human research, because the samples were de-identified.
Consent for publication
CB and MWP are inventors on the patent for ShK-186/dalazatide. CB and MWP are cofounders of Airmid, Inc., and sit on its board of directors. CB and MWP are investors in Kineta, Inc. The other authors declare that they have no competing interests.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
- 9.Tarcha E, Chi V, Munoz-Elias E, Bailey D, Londono L, Upadhyay S, Norton K, Banks A, Tjong I, Nguyen H, et al. Durable pharmacological responses from the peptide ShK-186, a specific Kv1.3 channel inhibitor that suppresses T cell mediators of autoimmune disease. J Pharmacol Exp Ther. 2012;342:642–53.CrossRefGoogle Scholar
- 13.Munoz-Elias E, Peckham D, Norton K, Duculan J, Cueto I, Li X, Qin J, Lustig K, Tarcha E, Odegard J, et al. Dalazatide (ShK-186), a first-in-class blocker of Kv1.3 potassium channel on effector memory T cells: safety, tolerability and proof of concept of immunomodulation in patients with active plaque psoriasis [abstract]. Arthritis Rheumatol. 2015;67(suppl 10).Google Scholar
- 35.Aletaha D, Neogi T, Silman A, Funovits J, Felson D, Bingham C, Birnbaum N, Burmester G, Bykerk V, Cohen M, et al. Rheumatoid arthritis classification criteria: an American College of Rheumatology/European League Against Rheumatism collaborative initiative. Ann Rheum Dis. 2010;2010(69):1580–8.CrossRefGoogle Scholar
- 38.Grissmer S, Nguyen A, Aiyar J, Hanson D, Mather R, Gutman G, Karmilowicz M, Auperin D, Chandy K. Pharmacological characterization of five cloned voltage-gated K+ channels, types Kv1.1, 1.2, 1.3, 1.5, and 3.1, stably expressed in mammalian cell lines. Mol Pharmacol. 1994;45(6):1227–34.PubMedGoogle Scholar
- 41.Beeton C, Chandy K. Induction and monitoring of adoptive delayed-type hypersensitivity in rats. J Vis Exp. 2007;8:325.Google Scholar
- 53.Alvaro-Gracia J, Zvaifler N, Firestein G. Cytokines in chronic inflammatory arthritis: V. Mutual antagonism between interferon-γ and tumor necrosis factor-α on HLA-DR expression, proliferation, collagenase production, and granulocyte macrophage colony-stimulating factor production by rheumatoid arthritis synoviocytes. J Clin Invest. 1990;86:1790–8.CrossRefGoogle Scholar
- 55.Page C, Smale S, Carty S, Amos N, Lauder S, Goodfellow R, Richards P, Jones S, Topley N, Williams A. Interferon-γ inhibits interleukin-1β-induced matrix metalloproteinase production by synovial fibroblasts and protects articular cartilage in early arthritis. Arthritis Res Ther. 2010;12:R49.CrossRefGoogle Scholar
- 56.Vermeire K, Heremans H, Vandeputte M, Huang S, Billiau A, Matthys P. Accelerated collagen-induced arthritis in IFN-γ receptor-deficient mice. J Immunol. 1997;158:5507–13.Google Scholar
- 61.Dolhain R, ter Haar N, Hoefakker S, Tak P, de Ley M, Claassen E, Breedveld F, Miltenburg A. Increased expression of interferon (IFN)-γ together with IFN-γ receptor in the rheumatoid synovial membrane compared with synovium of patients with osteoarthritis. Br J Rheumatol. 1996;35:24–32.CrossRefGoogle Scholar
- 62.Yuan J, Wegenka U, Lutticken C, Buschmann J, Decker T, Schindler C, Heinrich P, Horn F. The signalling pathways of interleukin-6 and γ interferon converge by the activation of different transcription factors which bind to common responsive DNA elements. Mol Cell Biol. 1994;14:1657–68.CrossRefGoogle Scholar
- 69.Tolboom T, van der Helm-Van Mil A, Nelissen R, Breedveld F, Toes R, Huizinga T. Invasiveness of fibroblast-like synoviocytes is an individual patient characteristic associated with the rate of joint destruction in patients with rheumatoid arthritis. Arthritis Rheum. 2005;52:1999–2002.CrossRefGoogle Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.