Article Text


The novel small molecule drug Rabeximod is effective in reducing disease severity of mouse models of autoimmune disorders
  1. M Hultqvist1,2,3,
  2. K S Nandakumar1,2,3,
  3. U Björklund,
  4. R Holmdahl1,2,3
  1. 1
    Section for Medical Inflammation Research, I11 BMC, Lund University, Lund, Sweden
  2. 2
    Section for Medical Inflammation Research, MBB, Karoliska Institutet, Stockholm, Sweden
  3. 3
    OxyPharma AB, Stockholm, Sweden
  1. R Holmdahl, Medical Inflammation Research, I11 BMC, Lund University, S-221 84 Lund, Sweden; rikard.holmdahl{at}


Objectives: Autoimmune diseases such as rheumatoid arthritis (RA) and multiple sclerosis (MS) affect a relatively large portion of the population, leading to severe disability if left untreated. Even though pharmaceutics targeting the immune system have revolutionised the therapy of these diseases, there is still a need for novel, more effective therapeutic substances. One such substance is the new chemical entity 9-chloro-2,3 dimethyl-6-(N,N-dimthylamino-2-oxoethyl)-6H-indolo [2,3-b] quionoxaline, Rabeximod, currently being investigated for efficiency in treatment of human RA. In this study we aimed to evaluate Rabeximod as a treatment for autoimmune diseases, using animal models.

Methods: In the present investigation we have evaluated Rabeximod as a treatment for autoimmune diseases using mouse models of RA and MS, ie, collagen-induced arthritis, collagen antibody induced arthritis and experimental autoimmune encephalomyelitis.

Results: Rabeximod efficiently prevented arthritis and encephalomyelitis in mice. In addition, this effect correlated to the timepoint when cells migrate into the joints.

Conclusions: We conclude that Rabeximod reduces disease severity in animal models of autoimmunity and should be considered as a new therapeutic substance for MS and RA.

Statistics from

The pathogenesis of autoimmune diseases is difficult to dissect, as these diseases are dependent on genes and environment. This complicates the issue of identifying efficient therapies, since the pathway to target is mainly unknown. Two such complex diseases are rheumatoid arthritis (RA) and multiple sclerosis (MS). RA and MS are chronic inflammatory diseases affecting joints and the central nervous system (CNS), respectively. Therapy of these disorders to a large extent relies on the use of disease-modifying anti-inflammatory drugs (DMARDs). Lately, the use of biological agents, targeting cytokines involved in the inflammatory reaction, has revolutionised the therapy. Interfering with tumour necrosis factor (TNF)α has proved to be successful in therapy of RA,1 2 but such treatment seems to worsen the disease in MS.3 In MS treatments targeting interferon (IFN)β are effective and currently used. However, there is a risk of severe infection with these drugs since cytokines play an important role in immune defence.4 Although these biological drugs have proven to be beneficial and acceptable, development of small molecular therapeutic drugs for these diseases would be beneficial. One such compound is the new small molecular drug Rabeximod (9-chloro-2,3 dimethyl-6-(N,N-dimthylamino-2-oxoethyl)-6H-indolo[2, 3-b] quionoxaline), formerly known as Rob 803, that is currently being trialled in patients with RA with moderate to severe disease. In this study, we investigated the efficacy of Rabeximod in animal models of RA and MS.



B10.Q mice with a mutation in Ncf1 (B10.QNcf1*/*) have been described previously.5 The mice were sex-matched and age-matched (7–12 weeks) at the start of the experiment. QB mice (10–20 weeks) generated from a (BALB/cxB10.Q)F1 cross were used to identify the critical timepoint of action. SJL mice (bred and kept at Harlan Biotech, Rehovot, Israel) were used for the experimental autoimmune encephalomyelitis (EAE) experiments. B10.QNcf1*/* and QB mice were kept and bred in a climate-controlled environment with a 12-h light/dark cycle and fed standard rodent chow and water ad libitum in the animal facility of Medical Inflammation Research, Lund University, Lund, Sweden ( BALB/c mice used for collagen antibody-induced arthritis (CAIA) experiments were bred and kept at Harlan Biotech, Israel. All experiments were approved by the local (Malmö, Lund, Sweden) ethical committee (license M70/04) or performed following review by the Committee for Ethical Conduct in the Care and Use of Laboratory Animals of the Hebrew University, Jerusalem, the Institutional Animal Care and Use Committee (IACUC) responsible for approving Harlan Biotech (Israel) animal usage application in compliance with its respective registration under National Institues of Health (NIH) accreditation no. OPR-A01-5011 HU.

Induction of disease

Collagen-induced arthritis (CIA)

A total of 100 μg of rat collagen type II (rCII) emulsified in complete Freund adjuvant (CFA; Difco, BD Diagnostics Systems, Sparks, Maryland, USA) in a total volume of 50 μl was injected subcutaneously at the base of the tail day 0. rCII was purified from the Swarm rat chondrosarcoma as previously described.6 On day 35 mice were given a subcutaneous booster injection of 50 μg rCII in incomplete Freund adjuvant (IFA; Difco) in a total volume of 50 μl.


Arthritis was induced day 0 by injection of 4 mg of a mixture of four purified anti-CII monoclonal antibodies (M2139, CIIC1, CIIC2 and UL-1; produced at Medical Inflammation Research) or with cocktail of monoclonal antibodies (D1, F10, A2 and D8; MD Biosciences, a division of Morwell Diagnostics GmBH, Zürich, Switzerland) to CII at a dose level of 100 mg/kg (2 mg/mouse). At day 5 or 3 respectively, lipopolysaccharide (LPS) was injected intraperitonally to enhance the incidence and severity of the disease (50 μg/mouse).


EAE was induced in SJL mice by subcutaneous injection of proteolipid protein (PLP; MD Biosciences) emulsified in CFA (MD Biosciences) on day 0 (125 μg PLP/300 μg CFA per animal). Supplemental immunostimulation with 20 μg/kg pertussis toxin (Sigma Aldrich, St Louis, Missouri, USA) was performed on day 0 and after 48 h.

Disease evaluation


Mice were scored blindly 1–3 times a week. The scoring protocol was described in detail previously.7 Following this protocol each mouse can score a maximum of 15 points/paw and a maximum of 60 points/mouse. Alternatively, arthritis severity was evaluated as hind paw thickness.


Mice were examined for signs of any neurological responses and symptoms on a daily basis. EAE was scored according to a classical 0–5 scale as follows: 0 = normal, absence of neurological signs; 1 = tail weakness, tail limp and droops; 2 = hind leg weakness and paresis, wobbly walk and/or hind legs unsteady; 3 = hind leg paralysis, when moving animal dragging its hind legs; 4 = quadriplegia, full paralysis, animal unable to move, thin and emaciated; 5 = morbid and/or death.

Drug preparation

Carboxylmethyl cellulose (CMC) vehicle

Water (100 ml) was heated to 50°C and 0.22 g of carboxylmethyl cellulose sodium salt (NaCMC; medium viscosity; Sigma Aldrich) was added under stirring for 70 min. Then, 0.9 ml polysorbate (Tween 80; Sigma Aldrich) was added and stirred for 30 min. The final volume was made up with 100 ml purified water and stirred for 5 min.

Oral administration in Ncf1 mutated mice

A total of 4 ml of CMC vehicle was added to a volumetric flask containing 100 mg of Rabeximod, ultrasonicated (10 min) before dilution in vehicle (final concentration 5 mg/ml). The solution was used for two administrations and stored under dark and cold conditions.

Corn oil

Rabeximod was dissolved in corn oil (Sigma Aldrich) and ultrasonicated in a heated water bath (68°C) until a clear solution was obtained (2×30 min with pipetting inbetween).

Treatment protocol


40 mg/kg of Rabeximod was administered orally starting from day 1 and continued every second day for 20 days.


Mice were treated on 6 successive days starting from day 0 with different concentrations of Rabeximod or using a 40 mg/kg dose (subcutaneously or orally). Alternatively, mice were divided into groups where treatment was initiated at different timepoints. Treatment started either on day 1, 3, 5, 7 or 9 after antibody transfer and continued for 12 days with subcutaneous administrations every second day.


In the EAE experiments, Rabeximod was injected at a dose of 7 mg/kg or 20 mg/kg subcutaneously or 20 mg/kg orally every day for 10 successive treatment days starting day 0. Corn oil (5 ml/kg; Sigma Aldrich) was used as control.

Determination of reactive oxygen species (ROS) production

B10.Q mice were injected subcutaneously with 40 mg/kg Rabeximod in corn oil on day 0. The level of intracellular oxidative burst ex vivo was measured in blood taken on day 1 and 5 after injection. Blood collected in tubes containing heparin was haemolysed with ammonium chloride (0.84%, pH 7.4). Granulocytes were stained with allophycocyanin conjugated anti-Gr-1 antibody (RB6-8C5, BD Biosciences Pharmingen, San Jose, California, USA) (30 min, 4°C). To determine the ROS production a modified version of the oxidative burst flow cytometry assay as previously described was used.8 Briefly, cells were resuspended in Dulbecco complete medium and incubated with 3 μM dihydrorhodamine-123 (10 min at 37°C) (Molecular Probes, Invitrogen, Eugene, Oregon, USA), which after oxidation to rhodamine-123 emits a fluoroscent signal upon excitation. Phorbol 12-myrestate 13-acetate (PMA; 200 ng/ml; Sigma Aldrich) was used to stimulate the cells (20 min, 37°C). After washing, cells were acquired on a FACSort (BD Biosciences, Franklin Lakes, New Jersey, USA) and gated on cell-type and R-123 fluorescence intensity. Results are expressed in relative fluorescence units.

Cytokine production after LPS stimulation of peripheral blood mononuclear cells (PBMCs)

Human PBMCs were isolated from whole blood using a Ficoll gradient. Cells (2×106/ml) were then plated in a volume of 0.5 ml (0.5 ml of vehicle control (1% methanol in RPMI 1640 media) or Rabeximod resuspended in methanol and diluted to a concentration of 2.5 or 25 μg/ml media was added). After incubation (1 h 37°C) with LPS (0.1 μg/ml), cells were further incubated for 18–20 h before supernatant was removed for analysis of cytokine production using cytokine multiplex assays (Biosource, Nivelles, Belgium), according to the manufacturer’s instructions. Data were collected using a Luminex 100 (Luminex Corporation, Austin, Texas, USA). All samples were analysed in duplicates and a mean value for each sample was calculated.


Quantitative data was expressed as mean (SEM). All results were compared to the control group if not otherwise stated. Statistics were calculated with the Mann–Whitney U test. p Values <0.05 were considered statistically significant (*, p⩽0.05, **, p⩽0.01, ***, p⩽0.001 throughout the present work).


Rabeximod reduces arthritis severity

To study the effect of the novel compound Rabeximod on arthritis in mice, we first used a model of RA where arthritis is induced with a cocktail of CII-specific monoclonal arthritogenic antibodies911 and subsequently enhanced by injection of LPS, stimulating Toll-like receptor 4 (TLR4) expressing immune cells.12 In this model of CAIA, the effect on the inflammatory phase can be studied directly without involving the priming phase of the immune response, since the antibodies penetrate and bind to cartilage followed by activation of complement and Fc receptor bearing inflammatory cells.13 Subsequently, granulocytes are attracted and macrophages are activated in a fashion that is independent of the adaptive immune system.11

First, we investigated dose and route dependency. BALB/c mice were injected with titrated doses of Rabeximod for 6 successive days, starting on day 0. We found that a dose of 40 mg/kg resulted in the best preventive effect with an arthritis-reducing capacity similar to the TNFα blocking agent etanercept, used as a golden standard for treatment of RA (fig 1A). In addition, different routes of administration were investigated in a similar experimental setup, suggesting subcutaneous injections to be superior for maximal disease suppression (fig 1B).

Figure 1 Rabeximod effectively reduces arthritis induced with arthritogenic antibodies. Arthritis was induced in BALB/c mice with an intravenous injection of a cocktail containing four monoclonal antibodies on day 0 followed by a single intraperitoneal injection of lipopolysaccharide (LPS) on day 3. Hind paw thickness was measured as an evaluation of arthritis severity. A. Starting from the day of immunisation the mice received different doses (subcutaneous) of Rabeximod in corn oil (mg/kg). Treatment was repeated for 6 successive days and compared to treatment with etanercept (6,25 mg/kg by repeated intraperitoneal injections during 11 dosing sessions) and control mice were injected with corn oil only (n = 9). The data series for etanercept-treated mice are masked under the data series for treatment with 40 mg/kg Rabeximod. B. Starting day 0, Rabeximod in corn oil was administered either as a daily oral gavage or subcutaneously at a dose of 40 mg/kg during 6 successive days. Control mice were given corn oil orally (n = 8–9). * = p<0.05, ** = p<0.01, p<0.001*** = p<0.001.

Rabeximod decreases EAE severity

Similar to RA, MS is a T cell-dependent disease and both diseases are presumed to involve related pathogenic pathways. It has been shown that experimental models for MS and RA have overlapping quantitative trait loci (QTL) in rat14 and mouse15 crosses, suggesting regulating genes shared between the diseases. To investigate Rabeximod in a T cell-dependent autoimmune model, we used myelin peptide induced experimental autoimmune encephalomyelitis (EAE), a model for MS. In contrast to the CAIA model, peptide-induced EAE is not dependent on pathogenic antibodies or granulocyte infiltration. SJL mice were treated with Rabeximod 10 successive days starting day 0 (ie, days 0–9). A clear reduction of disease severity could be seen after the treatment period (fig 2). Thus, we conclude that Rabeximod is effective also in EAE, a model operating without major involvement of neutrophils.

Figure 2 Rabeximod effectively reduced experimental autoimmune encephalomyelitis (EAE) severity. EAE was induced in SLJ mice on day 0 with a subcutaneous injection of proteolipid protein/complete Freund adjuvant (PLP/CFA) followed by an intraperitoneal injection of pertussis toxin on days 0 and 2. Mice were treated with Rabeximod in different doses (mg/kg) and through different routes of administration starting from day 0 and continuing for 10 successive days. Control mice were given corn oil subcutaneously. The line below the curves represents the difference between control mice and mice given Rabeximod orally and the line above curves represents difference between control mice and mice treated with 20 mg/kg Rabeximod subcutaneously. *, p Value of at least 0.05 (n = 10).

Rabeximod ameliorates CIA independently of NADPH oxidase ROS production

It has been found previously that B-220, a Rabeximod analogue, blocks production of ROS from neutrophils.16 This mechanism presumably operates downstream of the NADPH oxidase complex producing ROS in granulocytes,17 since no effect on assembly or function of the complex could be seen after B-220 administration in vitro.16 We have previously shown that ROS are highly involved in the regulation of arthritis as mice lacking ROS production are more susceptible to disease.5 18 These mice, mutated in Ncf1, were used to investigate the role of ROS in the effectiveness of Rabeximod. Rabeximod was administered orally on 10 occasions, starting on day 1 and continued on every other day until day 19. This treatment led to a partial reduction of disease severity that was most pronounced during two periods of disease development (ie, during the initiation of the first acute disease and the relapsing disease course) (fig 3A). This argues in favour of the therapeutic effect operating before the priming phase. As these mice lacked oxidative burst capacity an apparent conclusion is also that Rabeximod does not mediate its effect through modifying the oxidative burst response.16 In addition, when the effect of Rabeximod injection on intracellular ROS levels in granulocytes was investigated on day 1 (fig 3B) or day 5 (fig 3C), no significant differences could be seen compared to control mice injected with corn oil only. Taken together these data suggest that the main action of Rabeximod is not mediated by interfering with ROS levels.

Figure 3 The arthritis ameliorating effect of Rabeximod is not mediated via the NADPH oxidase complex. A. B10.Q mice with a mutation in Ncf1, resulting in nearly absent reactive oxygen species (ROS) production, were immunised on day 0 with 100 μg CII in complete Freund adjuvant (CFA) in a total volume of 50 μl/mouse. At day 35 the mice were boosted with 50 μg CII in incomplete Freund adjuvant (IFA). Rabeximod (40 mg/kg in vehicle) was administered orally on 10 occasions starting from day 1 and then on every second day (n = 17). Control mice (n = 25) were administered vehicle only. To investigate the effect of Rabeximod on ROS levels B10.Q mice (n = 5) were injected subcutaneously with 40 mg/kg Rabeximod in corn oil on day 0. Blood granulocytes were analysed for relative intracellular ROS levels on (B) day 1 and (C) day 5 after Rabeximod injection. Control mice were injected with corn oil only. *, p<0.05. PMA, phorbd 12-myrestate 13-acetate.

Disease amelioration is strictly time dependent

Apparently, Rabeximod influences disease activity if administered after priming, but before disease onset. To exactly investigate the therapeutic time window we used the CAIA model, which is very predictable in its timecourse. We divided mice into different groups where treatment was initiated at different days after disease induction. Rabeximod prevented development of severe arthritis when treatment started day 3 or day 5 after antibody injection (fig 4). However, treatment started before day 3 or later than day 5 had no therapeutic effect. Importantly, mice treated with Rabeximod from day 1 that had a treatment period covering the sensitive period from days 3–5 did not have the course of the disease affected. Thus, there is a precise time when Rabeximod mediated its effect, coinciding with the LPS-induced activation but preceding the clinical onset of arthritis.

Figure 4 The effect of Rabeximod is time dependent. QB mice were injected on day 0 with a mixture of four monoclonal antibodies (4 mg/mouse) followed by a lipopolysaccharide (LPS) injection on day 5. Treatment with Rabeximod (40 mg/mouse, subcutaneously) was started on different days after disease induction and repeated six times with 1 day between each injection (n = 8 for control group and 6 for treated groups). *, p<0.05.

Rabeximod suppresses release of proinflammatory cytokines in vitro

Since Rabeximod operates in a small time window coinciding with the timepoint when macrophages are activated, we wanted to investigate if the effect was due to alteration of cytokine production. Thus, we stimulated PBMCs with LPS in vitro in the presence of Rabeximod. Addition of Rabeximod to the cultures reduced the release of IL1β, IL6 and TNFα (fig 5). In the absence of LPS however, Rabeximod did not affect the cytokine production (data not shown). Since Rabeximod binds plasma proteins the effective concentrations in this experiment are 100 times lower than is indicated in fig 5 (12.5 ng/ml and 125 ng/ml respectively).

Figure 5 Prevention of cytokine release by Rabeximod. Supernatant from human peripheral blood mononuclear cells (PBMCs) stimulated with lipopolysaccharide (LPS) in the presence of Rabeximod at different concentrations were assayed for (A) interleukin (IL)1β, (B) IL6 and (C) tumour necrosis factor (TNF)α production. All samples were assayed in duplicates (n = 7). **, p<0.01.


In the present study we have evaluated the therapeutic effect of the small molecular compound Rabeximod on autoimmune diseases using animal models of RA and MS. We found that Rabeximod was efficient in preventing disease onset and in reducing disease severity. It operated in a specific time window coinciding with macrophage activation in the target tissue and before clinical onset of the disease.

Rabeximod is a first in the class of molecules that are under development as an improved first-line treatment for moderate to severe RA, and has demonstrated potential in its antiarthritic effects in disease models, including downregulation of clinical symptoms. To date, one phase I clinical trial has been completed. This first study assessed safety, tolerability, pharmacokinetics and pharmacodynamics of single and multiple oral rising doses of Rabeximod in healthy male volunteers. A study in patients with moderate to severe RA has been initiated. The mode of action of Rabeximod is, however, still unclear in spite of testing a number of hypotheses. We therefore decided to confirm the effect of Rabeximod in well defined animal models and to better define at which phase of the disease it mediated its effect.

From experiments using the CIA, CAIA and EAE models of autoimmunity, it is clear that the most pronounced effect of Rabeximod is in the emerging inflammatory phase. This was demonstrated with the CAIA model, which operates independent of T and B cells.19 The effect had a surprisingly narrow window of 3–5 days after the injection of antibodies, coinciding with the activation of macrophages with LPS. This is compatible with an effect on infiltration or activation of inflammatory cells such as macrophages or neutrophils, known to be of importance in the CAIA model at this stage.11 A therapeutic effect of Rabeximod observed when injected on day 0 in mice that had already received LPS on day 3 (fig 1) suggested a dependency of Rabeximod treatment on LPS activation. EAE, by contrast, is a T cell-dependent autoimmune model and the PLP peptide-induced EAE used here is independent of B cells and most likely also independent of neutrophils. Regardless, Rabeximod treatment had a suppressive effect when administered after priming but before the disease onset, similar to the CAIA model. Thus, an interesting possibility arises where Rabeximod modifies the activation of macrophages in the local tissue during the early stages of inflammation. A number of in vitro experimental results are also compatible with this observation. Rabeximod seems to suppress TNFα production and the oxidative burst response. Obviously, reduction of TNFα production could be one of many factors contributing to the ameliorative effect. The reduction of ROS response is, however, not obviously a protective effect as ROS production could be protective in the development of CIA and EAE.5 18 20 To directly address this possibility we used mice with deficient ROS production. Even though the results did not conclusively rule out an effect on ROS production, they clearly showed that Rabeximod could not solely operate through interacting with ROS production. Interestingly, Rabeximod treatment had protective effects at the initiation and relapsing phases of arthritis. We conclude that Rabeximod has ameliorative effects in autoimmune animal models. The disease-preventing mechanism is presumably operating through regulation of effector cell activation in the inflammatory site preceding the clinical onset of an acute inflammatory phase or relapse.


We wish to thank Rebecka Ljungqvist for taking excellent care of the mice.


View Abstract


  • Competing interests: UB is employed by OxyPharma AB, one of the spsonsors of this study.

  • Funding: Grants were received from OxyPharma AB, the Strategic Research Foundation and the EU projects AUTOCURE (LSHB-CT-2006-018661) and NEUROPROMISE (LSHM-CT-2005-018637).

  • Ethics approval: All experiments were approved by the local (Malmö, Lund, Sweden) ethical committee or performed following review by the Committee for Ethical Conduct in the Care and Use of Laboratory Animals of the Hebrew University, Jerusalem.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.