The establishment of a rheumatoid arthritis primate model in Macaca fascicularis

Background Rheumatoid arthritis (RA) is a long-term autoimmune disorder that mostly affects the joints and leads to the destruction of cartilage. An RA model in non-human primates is especially useful because of their close phylogenetic relationship to humans in terms of cross-reactivity to compounds developed using modern drug technologies. Methods We used a collagen-induced arthritis (CIA) model in Macaca fascicularis. CIA was induced by the immunization of chicken type II collagen. Swelling was measured as the longitudinal and transverse axes of 16 proximal interphalangeal joints. Results A new system for visual evaluation was created, with a perfect score of 16. Individual behavioral analysis was also conducted. Serum was collected once a week after the first immunization. Blood chemistry and inflammatory cytokine parameters were higher in the CIA group than in the wild type group. Conclusion In conclusion, we established CIA in M. fascicularis, and the results can be used for drug evaluation models.


Background
Rheumatoid arthritis (RA) is an autoimmune-mediated inflammatory disease. Some pathogenic mechanisms of RA have been revealed, and several novel treatment modalities, based on blocking pro-inflammatory cytokines or modulating intra-cellular signals such as Janus kinase (JAK)/signal transducer and activator of transcription proteins (STATs), are used nowadays [1]. Despite advances in our understanding of RA pathogenesis and improvements in RA treatment, some RA patients are refractory to conventional disease-modifying anti-rheumatic drugs (DMARDs), biological DMARDs, and even to the most recently developed small-molecule treatments, such as JAK inhibitors. To find solutions for these refractory RA patients, research into novel pathogenic mechanisms of RA and treatment strategies for RA are actively on-going.
To develop and prove the efficacy of new medication for RA patients, several stepwise pre-clinical and clinical studies are required. Most of the pre-clinical studies are performed on rodent models of RA, such as collagen-induced arthritis (CIA) or antibody-induced arthritis. Rodents are relatively easy to manage and are more economical than primates; however, rodent models have

Open Access
Journal of Translational Medicine *Correspondence: iammila@catholic.ac.kr; rapark@catholic.ac.kr † Hyun Sik Na, Seon-yeong Lee, Hong Ki Min, Mi-La Cho and Sung-Hwan Park contributed equally to this work 5  Full list of author information is available at the end of the article a critical failing due to the differences between humans and rodents in terms of morphology, physiology, and phylogenetic closeness [2,3]. Therefore, any beneficial effects proven in rodent models do not guarantee the same or similar therapeutic effects in RA patients. There is an unmet need to establish a better animal model for RA to enhance our knowledge of RA pathogenesis and to conduct pre-clinical experiments using an RA animal model that is similar to humans.
Non-human primates are an ideal animal model to investigate RA pathogenesis and treatment responses because they resemble humans more closely than rodents in terms of their physiology, morphology, and genetic background. Several non-human primate background RA models have been introduced, however, there are some limitations. The rhesus monkey, Macaca mulatta, was first used as an animal RA model; however, some individuals had specific major histocompatibility complex (MHC) class I alleles that were resistant to the induction of arthritis via type II collagen injections [4,5]. These CIA models, based on M. mulatta, had only a 70% induction rate for arthritis [4]. The cynomolgus monkey, M. fascicularis, has emerged as an alternative primate animal model for RA [6][7][8][9], however, a standardized scoring system has not yet been established for this model. Furthermore, osteoporosis is one of the major comorbidities of RA [10], and an evaluation of the degree of osteoporosis, such as bone density in primates models of RA, has not been conducted previously.
In this study, we induced CIA in M. fascicularis and established a scoring system of various clinical markers. Furthermore, we also investigated the MHC class I allele responsible for arthritis resistance in the M. mulatta. Finally, we performed micro-computed tomography (micro-CT) of the paw to evaluate the induction of osteoporosis in a primate model of RA.

Animals
Macaca fascicularis monkeys weighing 3-4 kg (age, 3-5 years) at the start of the experiment were purchased from GENIA (Seongnam, South Korea). Animals were housed in single cages in a controlled temperature room (21-29 °C) and light (12-h light-dark cycle) conditions. The monkeys had free access to a gamma-raysterilized diet (5048, LabDiet, St. Louis, MO, USA) and autoclaved R/O water. All animal research procedures were conducted in accordance with the Laboratory Animals Welfare Act, the Guide for the Care and Use of Laboratory Animals, and the Guidelines and Policies for Non-human Primate (NHP) Experiments provided by the Institutional Animal Care and Use Committee (ORIENT-IACUC-16255).

fascicularis B*01 gene analysis
Total mRNA was isolated from B cells of each M. fascicularis monkey and converted to cDNA. The cDNA was subjected to polymerase chain reaction (PCR) to identify the CIA-related B*01 gene. The following B-specific primers were used: MBS (sense) 5′-AAT TCA TGG CGC CCC GAA CCC TCC TCC TGC -3′ and (anti-sense) 5′-CTA GAC CAC ACA AGA CAG TTG TCT CAG-3′.

Induction of arthritis
CIA was induced in M. fascicularis. Type II collagen (CII) was dissolved overnight in 0.1 N acetic acid (4 mg/mL) with gentle rotation at 4 °C. Female M. fascicularis monkeys were immunized intradermally at the base of the tail with 100 µg of chicken CII (Chondrex Inc., Redmond, WA, USA) in complete Freund's adjuvant (Chondrex Inc.). M. fascicularis were boosted with 400 µg of CII emulsified with incomplete Freund's adjuvant (Chondrex Inc.) and injected intradermally into 10 sites on the back, 21 days after the primary immunization.

Arthritis scoring
The arthritis score was created by grading for disease severity. We used four scoring methods. First, clinical scoring was conducted as follows: 0, no disease symptoms; 0.5, fever; 1, apathy, lessened mobility and loss of appetite; 2, weight loss, warm extremities, and treatable pain without sinus tarsi syndrome (STS); 3, redness of joints, normal flexibility of extremities; 4, severe STS of joints, joint stiffness; and 5, untreatable pain, immobility of joints, weight loss > 25%. Second, phenotype scoring was as follows: 0, no evidence of swelling; 1, over 0.5 cm swelling of the proximal interphalangeal (PIP) joint compared with week 0 (< 0.5 cm 2 ); 2, score 1 evidence in two PIP joints; 3, score 1 evidence in three PIP joints; and 4, score 1 evidence in four PIP joints. Third, behavior scoring was as follows: 0, no evidence of behavior; 1, drag or limp in a hind leg; 2, slipping over in cage; and 3, one leg is not working. Fourth, pain scoring was as follows: 0, no evidence of pain; 1, tremor in paws; 2, tremor in paws and losing grip; and 3, severe tremor in paws and a crouching position.

Histopathological analysis
The PIP joints were collected from each group at 8 weeks after the first immunization. The tissues were fixed in 10% neutral buffered formalin solution, decalcified using Decalcifying Solution-Lite (Sigma, St. Louis, MO, USA), and embedded in paraffin. Sections of 4-to 5-μm thickness were cut, dewaxed using xylene, dehydrated in an alcohol gradient, and stained with hematoxylin and eosin (H&E) and safranin O.

Immunohistochemistry
Paraffin-embedded sections were incubated at 4 °C with the following primary monoclonal antibodies: anti-IL-1β, anti-IL-6 (Novus Biologicals, Littleton, CO, USA), anti-IL-17, and anti-TNF-α (Abcam, Cambridge, UK). The samples were then incubated with horseradish peroxidase-conjugated secondary antibody for 30 min. The reaction product was developed using 3,3-diaminobenzidine chromogen (Dako, Carpinteria, CA, USA). Histological assessments were conducted by three independent blinded observers. Images were captured using a DP71 digital camera (Olympus, Shinjuku, Tokyo, Japan) attached to a photomicroscope at a magnification of 200 × . Positively stained cells were counted using Adobe Photoshop software (Adobe, San Jose, CA, USA), and the mean values were calculated.

In vivo micro-CT imaging and analysis
Micro-CT imaging and analysis were performed using an animal scanner (SkyScan 1176, Billerica, MA, USA). The animals were imaged at settings of 60 kVp and 417 μA using an aluminum 0.5-mm thick filter. The pixel size was 35.76 μm, and the rotation step was 0.4°. Cross-sectional images were reconstructed using a filtered back-projection algorithm (NRecon Software, Bruker Micro CT, Kontich, Belgium). For each scan, a stack of 286 cross-sections was reconstructed at 4000 × 2670 pixels. Bone volume and bone surfaces were analyzed at the proximal interphalangeal and wrist joints.

Enzyme-linked immunosorbent assay (ELISA)
The concentrations of IL-6 in the blood serum of M. fascicularis were measured using a sandwich ELISA (Duo-Set; R&D Systems, Lille, France).

Hematology and serum biochemistry
Hematology values were determined using a Cell-Dyn ® 3700 hematologic analyzer (Abbott Diagnostics, Wiesbaden, Germany), and serum biochemistry was evaluated using a Hitachi clinical analyzer 7180 (Hitachi Ltd., Tokyo, Japan). The hematology analysis included hemoglobin (Hb; g/dL), and biochemistry analysis included C-reactive protein (CRP; mg/dL).

Statistical analysis
The results are expressed as mean ± standard deviation and were obtained from three separate experiments.

Resistance to CIA by B*01 genotype
Macaca mulatta monkeys are resistant to CIA when they have the B*01 genotype. However, we confirmed that M. fascicularis has no resistance to CIA [4,11]. PCR products of 1.1 kb were detected, and sequenced data of each M. fascicularis included the Gogo-B*01 gene. The M. fascicularis had the B*01 gene among their MHC class I genes, but all were susceptible to CIA induction (Fig. 1).

Clinical scoring of CIA in M. fascicularis
CIA monkeys were intradermally administered a CII/ CFA emulsion. The second booster immunization was performed after 3 weeks. Clinical changes in the M. fascicularis CIA model were assessed with a clinical  (Table 1). Clinical symptoms of CIA were monitored weekly for 8 weeks after the first immunization. All clinical signs were monitored weekly and included body weight, body temperature, number of joints affected, soft tissue swelling, warmth, and redness. These CIA monkeys showed more severe CIA symptoms and a high clinical score compared to wildtype M. fascicularis (Fig. 2).

Phenotype scoring of CIA in M. fascicularis
Phenotype scoring of the PIP joints and wrists was introduced to fully quantify the assessment of arthritis. The phenotype scoring system included the PIP joints of all limbs except the thumbs ( Table 2). In addition, pain and behavior scoring was analyzed from video data captured at 1-min intervals (Tables 3 and 4). Severe edema was observed in the CIA group (Fig. 3a). Also, the pain and behavior scores were higher in the CIA group (Fig. 3c). Phenotype scores of 7-8 were recorded at 8 weeks after CIA induction. There was also a markedly higher incidence of arthritis in the CIA groups (Fig. 3b). Safranin O and H&E staining showed increased infiltration of immune cells in the synovium and damage to the cartilage (Fig. 4a). Additionally, the CIA group had a higher score for inflammation, cartilage damage, and bone erosion than the WT group (Fig. 4b). Immunohistochemical staining against pro-inflammatory cytokine-(IL-1β, IL-6, IL-17, and TNF-α) presenting cells were significantly increased in the CIA group compared to the WT group (Fig. 5a). Serum levels of IL-6 and CRP were also significantly higher, and the hemoglobin level was lower in the CIA group (Fig. 5b).

Inflammatory T cell expansion in PBMCs from M. fascicularis with arthritis
Next, we analyzed helper T cells and regulatory T cells 6 weeks after induction of arthritis using PBMCs extracted from M. fascicularis. The isolated PBMCs were stained with T cell-and cytokine-specific antibodies. The results showed that inflammatory T cells such as IL-17 + CD4 T cells (Th17) was increased in arthritisinduced M. fascicularis. On the other hand, IFNγ + CD4 T cells (Th1) decreased (Fig. 6). Both Th1 and Th17 cells contribute to RA pathogenesis, but the role of Th1 cells is controversial [13][14][15]. Foxp3 + Treg cells decreased slightly in CIA, and Treg cells are known to show correlation with Th17 cells in terms of counter-regulation of Th17 cells [15]. These results suggest that our M. fascicularis CIA model is useful for pre-clinical studies of RA.

Micro-CT analysis of CIA in M. fascicularis
Quantitative micro-CT analysis showed the progression of bony damage of the PIP joints and wrist joint in CIA monkeys (Fig. 7a). Quantitative analysis of micro-CT demonstrated that the bone surface and the bone volume/tissue volume ratio were significantly decreased in the CIA group compared to the WT group (Fig. 7b).

Discussion
In this study, we demonstrated a CIA model in M. fascicularis and evaluated this animal RA model according to various clinical, serologic, and histologic parameters. In particular, the combined osteoporotic changes were noted and evaluated in the CIA model in M. fascicularis.
Although there exist several RA animal models based on primates, these RA animal models were not standardized to evaluate the degree of arthritis. Macaca mulatta-based RA animal models were the first primate models used. However, arthritis induction was successful in only 70% of M. mulatta, and arthritisresistant rhesus monkeys possess a specific MHC allele, Mamu-B*01 [4,11], which is identical to Gogo-B*01 in M. fascicularis. The present study showed that Gogo-B*01 allele-positive M. fascicularis were susceptible to CIA induction. This suggests that the arthritis-resistant MHC allele may differ between species, even when they belong to the same genus. Previous studies have demonstrated CIA induction in M. fascicularis [6][7][8][9], however, there has been no evaluation of the genetic background of the  Table 2 (*p < 0.05, **p < 0.01, and ***p < 0.001). c The pain and behavior scoring of hyperalgesia were analyzed, as seen in Tables 3 and 4 (*p < 0.05 and **p < 0.01) arthritis-resistant MHC allele. The present study showed that the specific MHC alleles known to play crucial roles in the induction of CIA in M. mulatta were not active in M. fascicularis. This is the first report to demonstrate the influence of the genetic background in arthritis induction in M. fascicularis.
Previous studies have demonstrated CIA induction in M. fascicularis, and have shown the therapeutic effects of IL-6 neutralizing agents in CIA [6,7,[16][17][18]. However, these studies did not use a standardized scoring system or conduct behavior and pain analysis, and did not evaluate bone mineral density. In the present study, we focused on evaluating clinical, phenotypical, behavior, and pain parameters in a M. fascicularis-based CIA model. Osteoporosis, a well-known co-morbidity of RA, was also quantitatively assessed by micro-CT scanning, which is used in osteoporosis evaluation [19]. This information could expand the use of M. fascicularis-based RA animal models to evaluate and demonstrate the therapeutic effects of novel agents in aspects of RA-related osteoporosis. The present study has several limitations. First, although we set the clinical/phenotypical/behavior/pain score in a CIA model using M. fascicularis, further studies are needed to clarify the usefulness of the scoring system. Several pain and behavior assessment tools are available for rodent or mice models such as the Von Frey or Randall Selitto test [20]; these methods can be modified for use in monkey models in further studies. Second, although the positive Gogo-B*01 allele in M. fascicularis did not show resistance to CIA induction, the induction rate for arthritis was approximately 80%. This suggests another MHC allele may be responsible for arthritis resistance in M. fascicularis, and the specific arthritisresistant MHC allele was not demonstrated in the present study.

Conclusion
The present study aimed to establish a novel scoring system for a M. fascicularis-based CIA and RA animal model that can be used to score clinical data, but also to score for behavior and pain. In addition, osteoporotic changes were clearly observed in this model, and therefore, a M. fascicularis-based CIA model could assess