Skip to main content
Clinical and Experimental Immunology logoLink to Clinical and Experimental Immunology
. 2011 Sep;165(3):318–328. doi: 10.1111/j.1365-2249.2011.04426.x

Small molecule CXCR3 antagonist NIBR2130 has only a limited impact on type 1 diabetes in a virus-induced mouse model

S Christen *, M Holdener *, C Beerli , G Thoma , M Bayer *, J M Pfeilschifter *, E Hintermann *, H-G Zerwes , U Christen *
PMCID: PMC3170981  PMID: 21649647

Abstract

CXCL10 is one of the key chemokines involved in trafficking of autoaggressive T cells to the islets of Langerhans during the autoimmune destruction of beta cells in type 1 diabetes (T1D). Blockade of CXCL10 or genetic deletion of its receptor CXCR3 results in a reduction of T1D in animal models. As an alternative to the use of neutralizing monoclonal antibodies to CXCL10 or CXCR3 we evaluated the small molecule CXCR3 antagonist NIBR2130 in a virus-induced mouse model for T1D. We found that the overall frequency of T1D was not reduced in mice administered with NIBR2130. An initial slight delay of diabetes onset was not stable over time, because the mice turned diabetic upon removal of the antagonist. Accordingly, no significant differences were found in the islet infiltration rate and the frequency and activity of islet antigen-specific T cells between protected mice administered with NIBR2130 and control mice. Our data indicate that in contrast to direct inhibition of CXCL10, blockade of CXCR3 with the small molecule antagonist NIBR2130 has no impact on trafficking and/or activation of autoaggressive T cells and is not sufficient to prevent T1D.

Keywords: chemokines, inflammation, insulitis, RIP–LCMV model, virus infection

Introduction

Trafficking of autoaggressive T cells to the target tissue in autoimmune diseases such as type 1 diabetes (T1D) is critically dependent upon chemokine-mediated cell attraction. It has been shown that the chemokine CXCL10 [interferon-inducible protein 10 kDa (IP-10)] is a key inflammatory mediator orchestrating the migration of islet-specific T cells into the pancreas [1]. Blockade of CXCL10 by neutralizing anti-CXCL10 antibodies abrogates T1D in rat insulin promoter–lymphocytic choriomeningitis virus (RIP–LCMV) mice [2]. In the RIP–LCMV model, transgenic RIP–LCMV mice express the glycoprotein (GP) or nucleoprotein (NP) of the LCMV specifically under control of the RIP in the β cells of the islets of Langerhans [3]. Such RIP–LCMV–GP or RIP–LCMV–NP mice are healthy and develop T1D only upon infection with LCMV [3]. We could demonstrate previously that LCMV infection results in an early and massive up-regulation of CXCL10 expression [2]. Thus, CXCL10 neutralization seems to block the initial steps of virus-induced inflammation of the pancreas and subsequently prevents the autoimmune destruction of the β cells. Similarly, T1D was abrogated when RIP–LCMV mice have been crossed to CXCR3-deficient mice [4]. In contrast, overexpression of CXCL10 in the β cells leads to an accelerated form of T1D in the RIP–LCMV model [5]. Further, CXCR3 and its ligands, CXCL10, CXCL9 [monokine induced by gamma interferon (Mig)] and CXCL11 [interferon-inducible T cell chemoattractant (I-TAC)] have been suggested to mediate allograft rejection in human individuals as well as a variety of animal models [611], although recently the originally proposed pivotal role has been challenged [12,13].

Neutralizing antibodies and other biologicals, such as soluble receptors or fusion proteins, have proved successful in blocking many diseases, ranging from autoimmune diseases to cancer therapy and have found wide use as therapies targeting molecules not amenable by small molecule drugs [14,15]. However, their efficacy is often challenged by their pharmacokinetic and pharmacodynamic properties, including limited penetration and accelerated elimination after repetitive administration [14,15]. Chemokine receptors are G protein-coupled receptors and have been targeted successfully by small molecule antagonists [16], such as the CXCR4 antagonist plerixafor [17] or the CCR5 antagonist maraviroc [18]. We therefore evaluated the efficacy of a small molecule CXCR3 antagonist (NIBR2130) in the abrogation of T1D in the RIP–LCMV model. NIBR2130 has been demonstrated previously to exhibit the highest CXCR3 binding affinity among structurally related compounds [19] and to fully block CXCR3 in vivo[13]. Here, we report that administration of NIBR2130 has no impact on the overall frequency of T1D in the RIP–LCMV model. However, a significant delay of the onset of T1D was detected as long as the antagonist was present in the slow-onset RIP–LCMV–NP model. Thus, in contrast to earlier findings with a neutralizing anti-CXCL10 antibody [2], CXCR3 blockade with the small molecule antagonist NIBR2130 has only a minor effect on the pathogenesis of T1D in the RIP–LCMV mouse model.

Material and methods

Mice and virus

Generation and screening by polymerase chain reaction (PCR) of H-2b RIP–LCMV–GP and H-2b RIP–LCMV–NP transgenic mice were performed as described previously [3,20]. These mice have been back-crossed to a C57BL/6J background for more than 20 years. LCMV Armstrong clone 53b (LCMV-Arm) were plaque-purified three times on Vero cells and stocks were prepared by a single passage on BHK-21 cells [20]. All animal experiments have been approved by the local Ethics Animal Review Board, Darmstadt, Germany (reference number: V54-19c20/15-F143/02).

Administration of NIBR2130

Mini-osmotic pumps model 2002 (Alzet, Cupertino, CA, USA) were filled with 250 µl NIBR2130 buffer [16·7 mg/ml in 45 mM N-methyl-2-pyrrolidinon, 4·5 mM HCl, 59% polyethylene glycol (PEG) 200, 50 mM NaCl] and implanted under the skin of 6-week-old mice according to the manufacturer's guidelines. The pumps release NIBR2130 at 0·5 µl/h, which corresponds to a daily dose of 200 µg. After implantation, 1 ml 0·9% NaCl solution prewarmed to 37°C was injected next to the pump to facilitate movement of the pump. This procedure was repeated every second day. Due to their limited capacity, the pumps were replaced after 2 weeks of implantation. Total administration of NIBR2130 was for 4 weeks.

Determination of the NIBR2130 blood concentration

The amount of NIBR2130 in mouse whole blood was determined by micro-liquid chromatography and mass spectrum analysis (µLC-MS/MS). Before analysis whole blood samples were prepared by a protein precipitation step with 0·04 mol/l ZnSO4 in 90% methanol. Ten µl of each sample were injected on a Kromasil 100-C18 0·5 × 50 mm column with 3·5 µm particles (G&T Septech AS, Kolbotn, Norway). For separation, a linear gradient of solvent A (0·2% formic acid) and solvent B (0·2% formic acid in acetonitrile) was applied. The analytical system used was an Agilent1100 CapLC system (Agilent GmBH, Waldbronn, Germany). For detection, the column effluent was injected directly into the atmospheric pressure electrospray source (AP-ESI) of the LTQ linear trap MS (ThermoFisher Scientific, Reinach, Switzerland). For calculation, the compound/internal standard ratio of the corresponding extracted ion chromatogram areas was used. Concentration was determined with LCQuan version 2·5 software.

Transwell migration assay

Transwell migration assays were performed with a murine T helper type 1 (Th1) cell clone [21]. Cells in migration medium [RPMI-1640, 1% bovine serum albumin (BSA), 10 mM (4-(2-hydroxyethyl)-1-piperazine ethanesulphonic acid) (HEPES)] were seeded into the upper chamber of a Transwell insert (96-well format, 3 µM pore size; Corning Costar, Chorges, France) at 200 000 cells/well. The lower chamber was filled with migration medium containing 20 ng/ml recombinant mouse CXCL10 (PeproTech, London, UK). A stock solution of NIBR2130 at a concentration of 4 mM in NIBR2130 buffer (45 mM N-methyl-2-pyrrolidinon, 4·5 mM HCl, 59% PEG 200, 50 mM NaCl) was prepared. For blockade of Th1 cell migration a dilution series of NIBR2130 in migration medium (10 µM, 1 µM and 100 nM) was added to the lower chamber. As control, a dilution series of the NIBR2130 buffer in migration medium was also tested. For comparison, 50 µg/ml hamster anti-murine CXCL10 monoclonal antibody (mAb) (1F11 [22]) was used. Assays were incubated at 37°C for 2 h. In the meantime, a suspension of 2 × 106/ml carboxyfluorescein succinimidyl ester (CSFE)-labelled cells in phosphate-buffered saline (PBS)/1% fetal calf serum (FCS)/1% paraformaldehyde (PFA) was prepared. To stop migration, cells which had migrated through the pores to the lower chamber were collected by centrifugation. Cell pellets were resuspended in 175 µl of CSFE-labelled cell suspension. Assays were performed in triplicate. Results are mean ± standard deviation (s.d.) (n = 3) per condition.

Blood glucose values

Blood samples were obtained from the tail vein. Blood glucose was monitored with a OneTouch Ultra (LifeScan, Nackargemünd, Germany) at weekly intervals. Blood glucose values more than 300 mg/dl were considered diabetic [23].

Glucose tolerance assay

Mice were fasted for 12–16 h and then received a single intraperitoneal injection of 1·5 mg/g body weight glucose (Sigma, St Louis, MO, USA). Blood glucose was measured immediately before injection and then at 10, 20, 30, 40, 60, 120 and 240 min post-injection.

Immunohistochemistry

Tissues were immersed in Tissue-Tek octreotide (OCT) (Bayer AG, Leverkusen, Germany) and quick-frozen on dry ice. Using cryomicrotome and sialin-coated Superfrost Plus slides (Fisher Scientific, Schwerte, Germany), 6 µm tissue sections were cut. Sections were then fixed with 90% ethanol at −20°C and, after washing in PBS, an avidin/biotin-blocking step was included (Vector Laboratories). Primary and biotinylated secondary antibodies (Vector Laboratories/Biozol, Eching, Germany) were reacted with the sections for 60 min each, and colour reaction was obtained by sequential incubation with avidin–peroxidase conjugate (Vector Laboratories) and diaminobenzidine–hydrogen peroxide. Primary antibodies used were: rat anti-mouse CD8a (Ly2) (BD Biosciences, Heidelberg, Germany), rat anti-mouse forkhead box P3 (Foxp3; eBioscience, Frankfurt, Germany), goat anti-mouse NKp46/NCR1 (R&D Systems, Wiesbaden-Nordenstadt, Germany) and rabbit anti-mouse Ki67 (Abcam, Cambridge, UK). The rabbit anti-CXCR3 polyclonal antibody was generated as described elsewhere [13].

Flow cytometry

For intracellular stains, single-cell suspensions were restimulated overnight with major histocompatibility complex (MHC) class I- or class II-restricted viral peptides (2 µg/ml LCMV–GP33 or LCMV–NP396) in the presence of Brefeldin A. Cells were stained for surface expression of CD8 and fixed, permeabilized and stained for intracellular interferon (IFN)-γ as described previously [23]. All antibodies were obtained from BD Biosciences. Samples were acquired using a fluorescence activated cell sorter (FACS)Calibur or a FACSCanto II flow cytometer (BD Biosciences).

Determination of viral titres by plaque assay

Viral titres of organ homogenates were determined by infection of Vero cells, as described previously [20,24]. Tissues (spleen and pancreas) were obtained from NIBR2130-treated and control mice (three animals per group) at days 3 and 7 after infection with 105 LCMV [intraperitoneally (i.p.)]. Homogenates and sera were diluted serially, and viral titres were calculated from the number of counted plaques.

Statistical evaluations

Diabetes incidence curves (‘survival curves’) were analysed using the log-rank (Mantel–Cox) test (GraphPad Prism version 5·02 software). Contingency of islet infiltration/ insulitis score was analysed by the χ2 test and T cell frequencies by the unpaired t-test.

Results

NIBR2130 treatment of RIP–LCMV–NP mice results in a slight delay in onset but not a reduced frequency of T1D

RIP–LCMV–GP and RIP–LCMV–NP mice were infected with 5 × 104 plaque-forming units (pfu) LCMV. One day before LCMV infection, osmotic mini-pumps loaded with either active small molecule CXCR3 antagonist NIBR2130 or with vehicle alone were implanted under the skin. At several days after implantation blood was taken to determine the NIBR2130 blood concentration. During the time of administration the concentration of NIBR2130 in the blood was found to range between 500 and 1200 nM (Fig. 1a,b). It has been reported previously that the half maximal inhibitory concentration (IC50) of NIBR2130 for binding to mouse CXCR3 was approximately 200 nM [19]. Thus, the NIBR2130 blood concentration was 2·5 to six times higher than its IC50 throughout the treatment period of 4 weeks. However, the levels dropped rapidly as soon as the pumps had been removed and was close to background at day 35 after LCMV infection in both RIP–LCMV–GP and RIP–LCMV–NP mice (Fig. 1a,b). The serum concentration found in NIBR2130-treated mice was sufficient to block CXCL10-mediated migration of a Th1 cell clone [21] in an in vitro migration assay (Fig. 1c). Similar to the anti-CXCL10 antibody 1F11 used in our previous studies [2,25], NIBR2130 blocked the cell migration of a Th1 cells towards CXCL10 very efficiently (Fig. 1c). Significant blockade of cell migration has been detected at a concentration of 100 nM, which is 10 times lower than the concentration found in the serum of NIBR2130-treated mice (Fig. 1a–c).

Fig. 1.

Fig. 1

Administration of NIBR2130 results in a stable serum concentration sufficient to block CXCR3. Osmotic pumps releasing a daily dose of 200 µg NIBR2130 or vehicle only were implanted under the skin of rat insulin promoter–lymphocytic choriomeningitis virus–glycoprotein (RIP–LCMV–GP) (a) and RIP–LCMV–nucleoprotein (NP) mice (b) for 4 weeks. The delivery of the antagonist was verified by the measurement of the serum concentration of NIBR2130 at different times as indicated. Dotted line displays the half maximal inhibitory concentration (IC50) of NIBR2130 (a and b). (c) Transwell migration assay of T helper type 1 (Th1) cells towards CXCL10 in the presence or absence of NIBR2130. The neutralizing anti-CXCL10 antibody 1F11 was used as a positive control. A dilution series of NIBR2130 in migration medium was used as indicated. As control a dilution series of the NIBR2130 buffer in migration medium was tested. Results are mean ± standard deviation (n = 3) per condition. Note that NIBR2130 efficiently blocks CXCL10-mediated migration at concentrations found in the serum of NIBR2130-treated mice (i.e. at 1 µM).

NIBR2130-treated RIP–LCMV–NP had a significant delay in diabetes onset compared to untreated mice (Fig. 2d). After 4 weeks of treatment the frequency of diabetic mice was reduced by 85% (Fig. 2d), and the average blood glucose levels were reduced by almost 50% (Fig. 2c). However, the diabetes incidence increased rapidly as soon as the pumps were removed (week 4 post-infection) and no NIBR2130 was detectable in the blood. Thus, despite a profound delay in diabetes onset the long-term diabetes incidence was similar in NIBR2130-treated RIP–LCMV–NP mice compared to untreated mice (Fig. 2d). The incidence data obtained from RIP–LCMV–NP mice (percentage of mice with a blood glucose level of > 300 mg/dl) are reflected by the mean blood glucose values. Because this analysis includes both diabetic and non-diabetic mice, the standard errors are considerable (Fig. 2a,c). No significant reduction in blood glucose levels and diabetes incidence or onset was detected in RIP–LCMV–GP mice (fast-onset diabetes) treated with NIBR2130 compared to control mice (Fig. 2a,b).

Fig. 2.

Fig. 2

Blockade of CXCR3 by NIBR2130 delays the onset of type 1 diabetes (T1D) in rat insulin promoter–lymphocytic choriomeningitis virus–nucleoprotein (RIP–LCMV–NP) mice: NIBR2130-treated and control RIP–LCMV–glycoprotein (GP) (a,b) and RIP–LCMV–NP (c,d) mice were infected with LCMV at day 0, and blood glucose was measured at days indicated (a,c). Diabetes incidence was determined by considering blood glucose values > 300 mg/dl as diabetic (b,d). Data are the mean ± standard error of the mean of 10–11 and 14–18 RIP–LCMV–GP and RIP–LCMV–NP mice per group, respectively. Note that the frequency of T1D is not reduced by the NIBR2130 treatment, despite a significant delay in diabetes onset in NIBR2130-treated RIP–LCMV–NP mice (d). Survival curves have been compared using the log-rank (Mantel–Cox) test (GraphPad Prism version 5·02 software). P-values before and after NIBR2130 withdrawal are indicated (d).

Blockade of CXCR3 by NIBR2130 has no effect on β cell function

As we did not detect a difference in the frequency of T1D in both LCMV-infected RIP–LCMV–GP and RIP–LCMV–NP mice, but found a delayed diabetes onset in the slow-onset RIP–LCMV–NP line, we focused our subsequent studies on the RIP–LCMV–NP line. First, in order to investigate the impact of CXCR3 blockade on β cell function, NIBR2130-treated and control RIP–LCMV–NP mice were infected with LCMV and then subjected to glucose tolerance experiments. These experiments were performed at day 28 post-infection, when most of the NIBR2130-treated RIP–LCMV–NP mice were still non-diabetic but already a large fraction of control mice had turned diabetic, and at day 49, when most of the RIP–LCMV–NP mice of both experimental groups were diabetic. At day 28 after LCMV infection RIP–LCMV–NP mice that had not turned diabetic by that time were challenged with a single injection of 1·5 mg/g body mass glucose and the efficiency of regulating the blood glucose back to normoglycaemia was assessed over time. Both NIBR2130-treated and control RIP–LCMV–NP mice showed increased blood glucose values of ∼300 mg/dl at 10 min after glucose injection. The blood glucose levels declined similarly in NIBR2130-treated and control RIP–LCMV–NP mice, with an approximate half-life of 20 min (Fig. 3a). When the same assay was performed at day 49 post-LCMV infection, NIBR2130-treated RIP–LCMV–NP mice showed a slight but not significantly accelerated down-regulation of blood glucose compared to untreated control RIP–LCMV–NP mice (Fig. 3b).

Fig. 3.

Fig. 3

Inhibition of CXCR3 has no effect on β cell function in a glucose tolerance assay. NIBR2130-treated and untreated rat insulin promoter–lymphocytic choriomeningitis virus–nucleoprotein (RIP–LCMV–NP) mice were challenged with a single injection of 1·5 mg/g body weight glucose (intraperitoneally) at days 28 (a) and 49 (b) after LCMV infection. Blood glucose was measured at different times as indicated. Data are the mean ± standard error of the mean of three to four non-diabetic mice per group.

Administration of NIBR2130 has no significant influence on islet infiltration

In order to assess further the impact of the CXCR3 blockade by NIBR2130, we analysed the cellular infiltration by different leucocyte populations into the islets of Langerhans by immunohistochemistry and applied an infiltration score for statistical analysis. Therefore, we compared protected, non-diabetic RIP–LCMV–NP mice treated with NIBR2130 with diabetic control RIP–LCMV–NP mice (three mice per group) at day 28 after LCMV infection. Consecutive pancreas sections were stained for insulin production as well as cellular infiltration by CD8 T cells, natural killer (NK) cells and CXCR3-expressing cells (Fig. 4a). All RIP–LCMV–NP mice showed a similar reduction of insulin production as well as similar infiltration patterns of the different immune cells analysed. No difference between NIBR2130-treated and control RIP–LCMV–NP mice was observed. Interestingly, in NIBR2130-treated RIP–LCMV–NP mice the amount of cells expressing CXCR3 within the infiltrations of islets was the same as in control mice (Fig. 4a). Independent of the status of overt disease, islets from both experimental groups showed similar rates of infiltration as analysed by applying a infiltration score (Fig. 4b). Some remaining insulin-expressing β cells were found in both groups of RIP–LCMV–NP mice. Nevertheless, insulitis was widespread, and more than 30% of all islets analysed had heavy intra-insular infiltration and/or remained as ‘islet-scars’ in both experimental groups. Moreover, islet infiltrating cells were predominantly of the CD8 phenotype. CXCR3 is also expressed prominently on dendritic cells and on NK cells. However, we did not find differential recruitment of either NK cells or DCs in the pancreas of RIP–LCMV–NP mice that were treated with the CXCR3 antagonist (Fig. 4a). Thus, even though we detected a slight but significant delay of T1D onset in NIBR2130-treated RIP–LCMV–NP mice, we did not find a diminished rate of islet infiltration by aggressive lymphoctyes. An alternative explanation for the delayed disease onset could be an increased recruitment of regulatory T cells (Tregs) to the islet of Langerhans, where they might suppress the β cell destruction by aggressive CD8 T cells. However, staining of the pancreas sections for FoxP3 expression revealed no increase in the amount of Tregs within the islets of the NIBR2130-treated RIP–LCMV–NP mice compared with control RIP–LCMV–NP mice (Fig. 4a). Approximately 60 Tregs per mm2 area of infiltration were found in both groups of RIP–LCMV–NP mice analysed (Fig. 4c). Further, differences in the proliferation of either infiltrating lymphocytes (activation and clonal expansion) or β cells (regeneration) might account for the observed delay in T1D onset in NIBR2130-treated RIP–LCMV–NP mice. However, we did not find significantly increased frequencies of β cells or of infiltrating cells expressing the proliferation marker Ki67 (Fig. 4a). These observations indicate that the observed delay in the onset of T1D is not due to a decreased islet infiltration rate, differences in cellular proliferation or a redistribution of specific T cell subsets such as Tregs within the islets of Langerhans.

Fig. 4.

Fig. 4

Fig. 4

Inhibition of CXCR3 has no impact on different cell types infiltrating islets. (a) Pancreas sections of control (left panels) and NIBR2130-treated (right panels) rat insulin promoter–lymphocytic choriomeningitis virus–nucleoprotein (RIP–LCMV–NP) mice at day 28 after LCMV infection were probed for insulin production and cellular infiltrations of CD8 T cells, CXCR3-expressing cells, natural killer (NK) cells, CD11c and forkhead box P3 (FoxP3)+ regulatory T cells (Tregs). In addition, the cells were stained for presence of the proliferation marker Ki67. (b) An insulitis score was applied to sections of three mice per group (n indicates the number of islets analysed). Scoring system: 0, no infiltration; 1, some peri-insular infiltration; 2, heavy peri-insular infiltration with some intra-insular infiltrates; 3, heavy intra-insular infiltration and/or islet scars. Statistical analysis (χ2 test) revealed no significant differences between control and NIBR2130-treated mice. (c) Evaluation of the amount of Tregs per mm2 area of islet infiltration. Data are the mean ± standard error of the mean of three mice per group. Statistical analysis (unpaired t-test) revealed no significant differences between control and NIBR2130-treated mice.

Neutralization of CXCR3 has no significant influence on the functional activity of LCMV-specific lymphocytes

Further, we assessed the functional activity of antigen-specific T cells after CXCR3 neutralization by intracellular staining for IFN-γ after ex vivo stimulation with the immunodominant MHC class I H-2Db-restricted LCMV peptides NP396. As in previous experiments, RIP–LCMV–NP mice were treated with NIBR2130 by subcutaneously implanted osmotic pumps 1 day before LCMV infection. At days 7 and 28 post-infection, splenocytes as well as lymphocytes from pancreatic draining lymph nodes (PDLN) were isolated and different T cell subsets were analysed for intracellular IFN-γ production after overnight stimulation with the immunodominant LCMV peptide NP396 or medium only. We found that at day 7 post-infection the frequency of functional active LCMV–NP396-specific CD8 T cells in the spleen (∼3%) and in the PDLN (∼1–1·5%) was similar in NIBR2130-treated and control RIP–LCMV–NP mice (Fig. 5b). Similarly, no significant differences were found when splenocytes and lymphocytes from the PDLN were isolated at day 28 post-infection from either protected (non-diabetic) NIBR2130-treated or control RIP–LCMV–NP mice (Fig. 5c). These data indicate that the detected delay in diabetes onset of T1D in NIBR2130-treated RIP–LCMV–NP mice was not the result of a significant difference in the frequency of functional active islet antigen-specific CD8 T cells.

Fig. 5.

Fig. 5

Inhibition of CXCR3 has no significant influence on the functional activity of lymphocytic choriomeningitis virus (LCMV)-specific T cells. The frequency of interferon (IFN)-γ-producing, LCMV-specific CD8 T cells was determined in the spleen and the pancreatic draining lymph node (PDLN) of control and NIBR2130-treated rat insulin promoter–LCMV–nucleoprotein (RIP–LCMV–NP) mice at days 7 and 28 after LCMV infection by flow cytometry. (a) Representative flow cytometry dot-plots of splenocytes of a representative control and NIBR2130-treated mouse at 7 days post-infection stained for surface CD8 and intracellular IFN-γ after stimulation with the immunodominant LCMV peptide NP396. (b,c) Frequencies of NP396-specific CD8 T cells at days 7 (b) and 28 (c) after LCMV infection. Horizontal bars indicate the mean of four to six individual mice per group. Statistical analysis (unpaired t-test) revealed no significant differences between control and NIBR2130-treated mice.

Discussion

Previous studies have reported the importance of the CXCR3 receptor in various animal models of chronic autoimmune diseases, such as adjuvant arthritis in rats, murine lupus nephritis and experimental autoimmune encephalomyelitis (EAE) [2628]. It has been demonstrated previously that virus-induced T1D is delayed in CXCR3-deficient RIP–LCMV–GP mice [4]. Further, in earlier studies we have shown that blockade of CXCL10, a ligand of CXCR3, effectively abrogates T1D in the RIP–LCMV–GP mouse model [2]. In contrast, overexpression of CXCL10 in the islets of Langerhans resulted in a substantial acceleration of the disease process and subsequently caused an exacerbation of T1D in RIP–LCMV–NP mice [5]. Although it is known that CXCR3 is expressed predominantly on Th1-polarized activated/memory T cells [29], it has been shown that 95% of LCMV-specific CD8 T cells express CXCR3 after LCMV infection [5]. These data underline the importance of the CXCL10/CXCR3-mediated migration of aggressive T cells during acute inflammation and ongoing autoimmune tissue destruction (see [30] for review).

Here we used the RIP–LCMV mouse model to evaluate the efficacy of the novel small molecule CXCR3 antagonist NIBR2130. Therapeutic application of low-molecular-weight antagonists is of high pharmacological interest because of their bioavailability, stability and potential safety. Further, the possibility for oral administration brings additional advantage over neutralizing antibodies that are administered predominantly by intravenous injection. The compound NIBR2130 is a highly specific CXCR3 antagonist that does not induce functional activity. NIBR2130 inhibits binding to the human and mouse receptor with IC50 values of 54 nM and 200 nM, respectively [19]. Moreover, NIBR2130 does not significantly inhibit other chemokine receptors such as CCR5, CXCR4 and CXCR2 and only inhibits serotonin 5HT2A receptor with an IC50 value below 1 µM, whereas the structurally related lysergic acid diethylamide (LSD) efficiently binds to various serotonin, dopamine and adrenergic receptors [19]. By performing a CXCL11-induced CXCR3 receptor occupancy assay, Zerwes et al. illustrated in rat whole blood that NIBR2130 efficiently inhibited CXCR3 and that the receptor was not internalized after binding to NIBR2130 [13]. In the present study, we blocked CXCR3 with NIBR2130 by subcutaneously implanted osmotic pumps during the development of T1D in the fast-onset RIP–LCMV–NP as well as in the slow-onset RIP–LCMV–NP mouse model. The delivery of the antagonist was fully maintained and the blood concentration was always between three and six times higher than the reported IC50 value during the entire treatment period of 4 weeks, which includes the time of viral infection, virus elimination, islet infiltration by autoaggressive lymphocytes and ongoing autoimmune destruction of β cells. We found that continuous delivery of the CXCR3 antagonist did not reduce the frequency of virus-induced T1D in RIP–LCMV–GP and RIP–LCMV–NP mice. Although we detected a delay in the onset of T1D in RIP–LCMV–NP mice, but not in the fast-onset RIP–LCMV–GP strain, such mice rapidly developed T1D similar to the control group that did not receive the CXCR3 antagonist as soon as the pumps were removed. These data indicate that the bioactivity of NIBR2130 is not sufficient to block CXCR3 to a degree that persistently abrogates T1D. The fact that we detected a limited impact on the onset of disease only in the RIP–LCMV–NP, but not in the RIP–LCMV–GP mice, might be attributed to the higher aggressiveness of the RIP–LCMV–GP model, in which mice develop T1D in a very rapid fashion with the presence of high-affinity, autoaggressive CD8 T cells independently of CD4 T cell help [20]. Such a differential outcome of T1D in the two RIP–LCMV mouse strains was observed in other studies as well. For example, overexpression of CXCL10 in the pancreatic β cells massively accelerated T1D in LCMV infected RIP–LCMV–NP mice but not in RIP–LCMV–GP mice [5].

Unfortunately, the administration of the CXCR3-antagonist NIBR2130 was far less effective than treatment with a neutralizing antibody to CXCL10 [2] or genetic disruption of CXCR3 expression [4]. In the study by Frigerio et al. the authors used transgenic RIP–LCMV–GP mice that were developed initially in the laboratory of Hans Hengartner and Rolf Zinkernagel [31]. Although the Hengartner/Zinkernagel RIP–LCMV model uses a different LCMV strain for initiation of T1D, their RIP–LCMV–GP model is identical in several aspects to our fast-onset RIP–LCMV–GP model. In contrast to our CXCR3-antagonist studies, Figerio et al. observed a significant delay in the onset of LCMV-induced T1D in CXCR3-deficient RIP–LCMV–GP mice. Such mice were protected from T1D up to 30 days post-infection. Thereafter, T1D developed in at least 50% of the mice, indicating that the disease was not persistently abrogated [4]. It should be noted here that a somewhat high blood glucose level of 600 mg/dl was used as a diabetes threshold in the Frigerio study. Most studies with mouse models for T1D use thresholds ranging from 250 to 300 mg/dl. Such differences in the definition of diabetes might have an impact on the outcome of the study. Importantly, however, in contrast to our study, immunohistochemistry revealed that insulitis was indeed reduced clearly at days 6 and 8 post-LCMV infection in RIP–LCMV–GP × CXCR3–/– mice compared to regular RIP–LCMV–GP mice [4]. In addition, it should be stressed that we detected a delay in T1D in the slow-onset RIP–LCMV–NP, but not in the fast-onset RIP–LCMV–GP mouse line. In contrast to CXCR3-deficient mice, blockade of the CXCR3 ligand CXCL10 with a neutralizing antibody abrogated disease permanently in > 60% of mice [2]. Although the two studies used different strains of RIP–LCMV mice [3,31], it seems that the mechanistic consequences of CXCL10 neutralization are different from those of blocking CXCR3. There is ample evidence from studies in several different mouse model that investigate demyelinating diseases with similarities to human multiple sclerosis that the role of CXCR3 and its ligands is complex [32]. In addition, it had been shown in concanavalin A-induced hepatitis and fibrosis that different CXCR3 ligands, such as CXCL9 and CXCL10, might even have opposite effects on the progress of liver damage [33]. Similarly, blockade of CXCL9 with a neutralizing antibody had no influence on the incidence and onset of T1D in earlier experiments [2]. Thus, blocking the receptor rather than its individual chemokine ligands might have only a minor effect on the overall infiltration and the progress of tissue pathogenesis.

In accordance with our finding that administration of NIBR2130 had no effect on the frequency of T1D in the RIP–LCMV model, the pathogenic processes of T1D, as reflected by islet infiltration and the frequency of pathological T cells, did not differ in CXCR3 antagonist-treated compared to control RIP–LCMV–NP mice. During both the acute anti-viral (day 7) as well as the autoimmune (day 28) phase of the anti-LCMV immune response similar frequencies targeting antigen-specific T cells were found regardless of the inhibition of the CXCR3/CXCL10 interaction. Further, islet infiltration was similar and the functional impairment of β cells was comparable between NIBR2130-treated and control mice. Although the CXCR3 receptor was blocked in NIBR2130-treated mice, similar numbers of CXCR3 leucocytes infiltrating islets were found when compared to control animals. As an alternative to a reduction of the frequency of autoaggressive CD8 T cells, an increase in the frequency of Tregs might account for the observed delay in T1D. However, we did not find higher frequencies of FoxP3-positive cells among islet-infiltrating cells.

Overall, the fact that islet infiltration by aggressive CD8 T cells and was not blocked by neutralization of the interaction between CXCR3 and its ligands might indicate a certain redundancy of inflammatory factors driving the infiltration of the islets. An alternative and much simpler explanation would be that NIBR2130 is not as effective in blocking CXCR3 in vivo, as would have been expected from the in vitro neutralization assays [19]. In another study the CXCR3 antagonist NIBR2130 was used in a rat model of cardiac allograft rejection. It was found that a pharmacological blockade of CXCR3 did not result in prolonged cardiac allograft survival, although it could be demonstrated the antagonist was active in vivo and did indeed neutralize CXCR3 efficiently [13]. The lack of efficacy of the CXCR3 antagonist in the cardiac allograft model was corroborated by lack of prolonged allograft survival when CXCR3–/– mice were used as graft recipients [12,13], and indicate that CXCR3 does not play a pivotal role in allograft rejection. In another study antagonists to CXCR3 and CXCR4 were used in an experimental autoimmune encephalomyelitis (EAE) model. Interestingly, blockade of both receptors simultaneously significantly inhibited clinical EAE and was more efficient than blocking both receptors separately [28]. These data indicate that neutralization of more than one critical player involved in the pathogenic processes of T1D might be necessary to interfere successfully and permanently with the autodestructive process and to abrogate disease.

Acknowledgments

This research project was funded by the University Hospital Frankfurt, Frankfurt am Main, Germany and a grant of the German Research Foundation (DFG) to U. C. The study was also funded by Novartis Pharma AG, Basel Switzerland, which provided the CXCR3 antagonist NIBR2130 and the ALZET osmotic mini-pumps for its delivery.

Disclosure

The authors C.B., G.T. and H.G.Z. are employees of Novartis Pharma AG. The CXCR3 antagonist NIBR2130 was generated by Novartis Pharma AG. The authors S.C., M.H., M.B., J.M.P., E.H. and U.C. have nothing to disclose.

References

  • 1.Christen U, von Herrath MG. IP-10 and type 1 diabetes: a question of time and location. Autoimmunity. 2004;37:273–82. [PubMed] [Google Scholar]
  • 2.Christen U, McGavern DB, Luster AD, von Herrath MG, Oldstone MB. Among CXCR3 chemokines, IFN-gamma-inducible protein of 10 kDa (CXC chemokine ligand (CXCL) 10) but not monokine induced by IFN-gamma (CXCL9) imprints a pattern for the subsequent development of autoimmune disease. J Immunol. 2003;171:6838–45. doi: 10.4049/jimmunol.171.12.6838. [DOI] [PubMed] [Google Scholar]
  • 3.Oldstone MBA, Nerenberg M, Southern P, Price J, Lewicki H. Virus infection triggers insulin-dependent diabetes mellitus in a transgenic model: role of anti-self (virus) immune response. Cell. 1991;65:319–31. doi: 10.1016/0092-8674(91)90165-u. [DOI] [PubMed] [Google Scholar]
  • 4.Frigerio S, Junt T, Lu B, et al. Beta cells are responsible for CXCR3-mediated T-cell infiltration in insulitis. Nat Med. 2002;8:1414–20. doi: 10.1038/nm1202-792. [DOI] [PubMed] [Google Scholar]
  • 5.Rhode A, Pauza ME, Barral AM, et al. Islet-specific expression of CXCL10 causes spontaneous islet infiltration and accelerates diabetes development. J Immunol. 2005;175:3516–24. doi: 10.4049/jimmunol.175.6.3516. [DOI] [PubMed] [Google Scholar]
  • 6.Hancock WW, Gao W, Csizmadia V, Faia KL, Shemmeri N, Luster AD. Donor-derived IP-10 initiates development of acute allograft rejection. J Exp Med. 2001;193:975–80. doi: 10.1084/jem.193.8.975. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Hancock WW, Lu B, Gao W, et al. Requirement of the chemokine receptor CXCR3 for acute allograft rejection. J Exp Med. 2000;192:1515–20. doi: 10.1084/jem.192.10.1515. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Meyer M, Hensbergen PJ, van der Raaij-Helmer EM, et al. Cross reactivity of three T cell attracting murine chemokines stimulating the CXC chemokine receptor CXCR3 and their induction in cultured cells and during allograft rejection. Eur J Immunol. 2001;31:2521–7. doi: 10.1002/1521-4141(200108)31:8<2521::aid-immu2521>3.0.co;2-q. [DOI] [PubMed] [Google Scholar]
  • 9.Yun JJ, Fischbein MP, Whiting D, et al. The role of MIG/CXCL9 in cardiac allograft vasculopathy. Am J Pathol. 2002;161:1307–13. doi: 10.1016/S0002-9440(10)64407-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Goddard S, Williams A, Morland C, et al. Differential expression of chemokines and chemokine receptors shapes the inflammatory response in rejecting human liver transplants. Transplantation. 2001;72:1957–67. doi: 10.1097/00007890-200112270-00016. [DOI] [PubMed] [Google Scholar]
  • 11.Segerer S, Cui Y, Eitner F, et al. Expression of chemokines and chemokine receptors during human renal transplant rejection. Am J Kidney Dis. 2001;37:518–31. [PubMed] [Google Scholar]
  • 12.Kwun J, Hazinedaroglu SM, Schadde E, et al. Unaltered graft survival and intragraft lymphocytes infiltration in the cardiac allograft of Cxcr3–/– mouse recipients. Am J Transplant. 2008;8:1593–603. doi: 10.1111/j.1600-6143.2008.02250.x. [DOI] [PubMed] [Google Scholar]
  • 13.Zerwes HG, Li J, Kovarik J, et al. The chemokine receptor Cxcr3 is not essential for acute cardiac allograft rejection in mice and rats. Am J Transplant. 2008;8:1604–13. doi: 10.1111/j.1600-6143.2008.02309.x. [DOI] [PubMed] [Google Scholar]
  • 14.Beck A, Wurch T, Bailly C, Corvaia N. Strategies and challenges for the next generation of therapeutic antibodies. Nat Rev Immunol. 2010;10:345–52. doi: 10.1038/nri2747. [DOI] [PubMed] [Google Scholar]
  • 15.Chan AC, Carter PJ. Therapeutic antibodies for autoimmunity and inflammation. Nat Rev Immunol. 2010;10:301–16. doi: 10.1038/nri2761. [DOI] [PubMed] [Google Scholar]
  • 16.Proudfoot AE, Power CA, Schwarz MK. Anti-chemokine small molecule drugs: a promising future? Exp Opin Investig Drugs. 2010;19:345–55. doi: 10.1517/13543780903535867. [DOI] [PubMed] [Google Scholar]
  • 17.De Clercq E. The bicyclam AMD3100 story. Nat Rev Drug Discov. 2003;2:581–7. doi: 10.1038/nrd1134. [DOI] [PubMed] [Google Scholar]
  • 18.Lieberman-Blum SS, Fung HB, Bandres JC. Maraviroc: a CCR5-receptor antagonist for the treatment of HIV-1 infection. Clin Ther. 2008;30:1228–50. doi: 10.1016/s0149-2918(08)80048-3. [DOI] [PubMed] [Google Scholar]
  • 19.Thoma G, Baenteli R, Lewis I, et al. Special ergolines are highly selective, potent antagonists of the chemokine receptor CXCR3: discovery, characterization and preliminary SAR of a promising lead. Bioorg Med Chem Lett. 2009;19:6185–8. doi: 10.1016/j.bmcl.2009.09.002. [DOI] [PubMed] [Google Scholar]
  • 20.von Herrath MG, Dockter J, Oldstone MBA. How virus induces a rapid or slow onset insulin-dependent diabetes mellitus in a transgenic model. Immunity. 1994;1:231–42. doi: 10.1016/1074-7613(94)90101-5. [DOI] [PubMed] [Google Scholar]
  • 21.Giegold O, Ludwig RJ, Hardt K, et al. Computer-aided analysis of cell interactions under dynamic flow conditions. Exp Dermatol. 2009;18:238–45. doi: 10.1111/j.1600-0625.2008.00793.x. [DOI] [PubMed] [Google Scholar]
  • 22.Khan IA, MacLean JA, Lee FS, et al. IP-10 is critical for effector T cell trafficking and host survival in Toxoplasma gondii infection. Immunity. 2000;12:483–94. doi: 10.1016/s1074-7613(00)80200-9. [DOI] [PubMed] [Google Scholar]
  • 23.Christen U, Wolfe T, Mohrle U, et al. A dual role for TNF-alpha in type 1 diabetes: islet-specific expression abrogates the ongoing autoimmune process when induced late but not early during pathogenesis. J Immunol. 2001;166:7023–32. doi: 10.4049/jimmunol.166.12.7023. [DOI] [PubMed] [Google Scholar]
  • 24.Christen U, Benke D, Wolfe T, et al. Cure of prediabetic mice by viral infections involves lymphocyte recruitment along an IP-10 gradient. J Clin Invest. 2004;113:74–84. doi: 10.1172/JCI200417005. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 25.Hintermann E, Bayer M, Pfeilschifter JM, Luster AD, Christen U. CXCL10 promotes liver fibrosis by prevention of NK cell mediated hepatic stellate cell inactivation. J Autoimmun. 2010;35:424–35. doi: 10.1016/j.jaut.2010.09.003. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 26.Mohan K, Issekutz TB. Blockade of chemokine receptor CXCR3 inhibits T cell recruitment to inflamed joints and decreases the severity of adjuvant arthritis. J Immunol. 2007;179:8463–9. doi: 10.4049/jimmunol.179.12.8463. [DOI] [PubMed] [Google Scholar]
  • 27.Steinmetz OM, Turner JE, Paust HJ, et al. CXCR3 mediates renal Th1 and Th17 immune response in murine lupus nephritis. J Immunol. 2009;183:4693–704. doi: 10.4049/jimmunol.0802626. [DOI] [PubMed] [Google Scholar]
  • 28.Kohler RE, Comerford I, Townley S, Haylock-Jacobs S, Clark-Lewis I, McColl SR. Antagonism of the chemokine receptors CXCR3 and CXCR4 reduces the pathology of experimental autoimmune encephalomyelitis. Brain Pathol. 2008;18:504–16. doi: 10.1111/j.1750-3639.2008.00154.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 29.Sallusto F, Lenig D, Mackay CR, Lanzavecchia A. Flexible programs of chemokine receptor expression on human polarized T helper 1 and 2 lymphocytes. J Exp Med. 1998;187:875–83. doi: 10.1084/jem.187.6.875. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 30.Christen U. Chemokines as drug targets in type 1 diabetes. Endocr Metab Immune Disord Drug Targets. 2007;7:7–12. doi: 10.2174/187153007780059405. [DOI] [PubMed] [Google Scholar]
  • 31.Ohashi P, Oehen S, Buerki K, et al. Ablation of tolerance and induction of diabetes by virus infection in viral antigen transgenic mice. Cell. 1991;65:305–17. doi: 10.1016/0092-8674(91)90164-t. [DOI] [PubMed] [Google Scholar]
  • 32.Liu L, Callahan MK, Huang D, Ransohoff RM. Chemokine receptor CXCR3: an unexpected enigma. Curr Top Dev Biol. 2005;68:149–81. doi: 10.1016/S0070-2153(05)68006-4. [DOI] [PubMed] [Google Scholar]
  • 33.Wald O, Weiss ID, Wald H, et al. IFN-gamma acts on T cells to induce NK cell mobilization and accumulation in target organs. J Immunol. 2006;176:4716–29. doi: 10.4049/jimmunol.176.8.4716. [DOI] [PubMed] [Google Scholar]

Articles from Clinical and Experimental Immunology are provided here courtesy of British Society for Immunology

RESOURCES