Repository corticotropin injection attenuates collagen-induced arthritic joint structural damage and has enhanced effects in combination with etanercept

The rat CIA model was used to characterize the effects of RCI therapy on established arthritis [25‐29]. Female dark Agouti rats (weight: 120–170 g) were obtained from Harlan Laboratories, Inc. (Indianapolis, IN) and acclimated for approximately 8 days before initiation of CIA. All animal use was in accordance with the guidelines cited in the Guide for the Care and Use of Laboratory Animals [30] (Institutional Animal Care Use Committee Protocol BBP12–002). Rats were anesthetized with isoflurane, which has been shown to have a minimal effect on inflammation, and injected with 400 μL of Freund’s incomplete adjuvant (Difco, Detroit, MI) emulsified with 2 mg/mL porcine type II collagen (Chondrex, Inc., Redmond, WA) at 2 sites on the back (200 μL per site) on day 0, and then 100 μL at 1 site on day 7. Arthritis disease was represented by an increase in ankle diameter observed at day 12 and reaching a plateau at days 18 through 20.

Disease induction was performed on study days 0 and 7 between the time of 8:00 am and 12:00 pm, and animals were dosed in the same time frame. On study day 13 after the onset of disease, rats were assigned to treatment groups (see Additional file 1, Fig. 1 and Table 1) according to body weight and mean ankle caliper measurements. Dose selection was determined on the basis of results from an unpublished study that used a prophylactic design and showed a dose-responsive benefit. Treatments were initiated on day 13 and continued through day 19. On study day 20, all animals were anesthetized with isoflurane and were sacrificed by bleeding the descending aorta to exsanguinate (for serum collection) followed by cervical dislocation before tissue collection. Spleens were weighed, and hind paws were transected at the level of the medial and lateral malleolus for determination of paw weight as another measure of inflammation, then fixed for histopathologic evaluation and microfocal computed tomography (micro-CT).

Animals studies were conducted at the laboratory of Bolder BioPATH (Boulder, CO). Rat hind paws were scanned using micro-CT to visualize CIA-induced damage to joints and quantify various bone parameters at the laboratory of Numira, (Salt Lake City, UT). All animals were maintained at 5 rats/cage in plexiglass shoebox cages containing aspen wood bedding. Animals were provided Lab Diet 5001 and ad libitum tap water. Environmental controls were set to maintain temperatures of 20–26 °C with relative humidity between 30 and 70% and a 12-h light/dark cycle.

All animals were required to weigh ≥120 g prior to use in the study. Animals that showed signs of morbidity, including the loss of more than 20% body weight within 1 week, were removed from the study and humanly sacrificed via CO₂ inhalation according to the Bolder BioPATH Program of Veterinary Care. The study protocol was reviewed and approved by the Institutional Animal Care and Use Committee of Mallinckrodt Pharmaceuticals.

Clinical assessment of CIA during the in-life phase was performed as previously described on days 9 through 20 [31]. Body weight and caliper measurements of right and left ankle diameters were taken on study days −1, 1, 6, and 9 through 20. Ankle caliper measurements were made with a Digitrix II micrometer (Fowler & NSK, Newton, MA). Baseline measurements were taken using 1 ankle with values rounded to 0.00254 cm. Measurements were confirmed as clinically normal (0.66–0.67 cm) by comparison with historical values for rats with a range of body weights. Experiments were performed twice with 6 animals in the naive group and 10 animals in all other groups. Animals were placed into groups such that each group consisted of similar mean ankle caliper measurements. Some animals may have been enrolled with no disease, with the assumption that disease onset would occur during the study.

Ankle joints were collected and kept in 10% neutral buffered formalin for at least 24 h before placement in 5% formic acid for approximately 1 week for decalcification. When decalcification was complete, ankle joints were cut in half. Joints were processed through graded alcohols and a clearing agent, paraffin embedded, sectioned, and stained with toluidine blue to visualize cartilage proteoglycan content for general and specific evaluation of cartilage changes. Tissues from all animals were examined microscopically by a board-certified veterinary pathologist. Ankles were given scores of 0 to 5 for all histopathologic criteria. Bone resorption, ankle inflammation, and ankle cartilage damage were scored as previously described [27, 31] with the addition of 0.5, reflecting very minimal resorption affecting only marginal zones and only a few joints, minimal focal inflammation, and very minimal decrease in toluidine blue staining affecting only marginal zones, respectively. Ankle pannus was scored as previously described [32], and ankle periosteal new bone formation (measured at 16× magnification) was scored according to the following criteria: 0 = normal (no periosteal proliferation); 0.5 = minimal focal or multifocal proliferation (width measures < 127 μm [1–2 U] at any location); 1 = minimal multifocal proliferation (width at any location measures 127–252 μm [3–4 U]; 2 = mild multifocal proliferation on tarsals, diffuse in some locations, width at any location 253–441 μm [5–7 U]; 3 = moderate multifocal proliferation on tarsals, diffuse in most other locations, width at any location measures 442–630 μm [8–10 U]; 4 = marked multifocal proliferation on tarsals, diffuse at most other locations, width at any location measures 631–819 μm [11–13 U]; and 5 = severe, multifocal proliferation on tarsals, diffuse at most other locations, width at any location measures > 819 μm [> 13 U] [25, 33].

Hind paws were scanned with a high-resolution, volumetric micro-CT scanner (μCT40; ScanCo Medical, Zürich, Switzerland). The image data were acquired with the following parameters: 18-μm isotropic voxel resolution at 300-ms exposure time, 2000 views, and 2 frames per view scanned from the distal tibia to the mid metatarsal bones. Voxel counts were calculated using VHLab software (Numira, Salt Lake City, UT). Bone density (a measure of the amount of magnesium hydroxyapatite [mgHA] per unit volume) was calculated as the fraction of bone mineral content over total bone volume. The cortical shell is represented in the CT image by voxels of varying densities. The density of mature, solid bone voxels is > 600 mgHA/cm³, whereas that of young bone is < 600 mgHA/cm³. The numbers and percentages of voxels > 600 mgHA/cm³ (representing old bone, Ct_BV_high, Ct_BV_high_pc) and < 600 mgHA/cm³ (representing new bone, Ct_BV_low, Ct_BV_low_pc) were calculated to reveal the prevalence of new bone growth. Bone volume difference of the cortical shell (cortical volume difference), a determination of old versus new bone formation, was calculated as Ct_BV_high − Ct_BV_low.

Immunohistochemistry (IHC) was performed on the Ventana Discovery-Ultra IHC/in situ hybridization automated staining platform (Ventana Medical Systems Inc., Tucson, AZ). Paraffin sections (4 μM) were used for staining with CD68 (ab125212, Abcam, Cambridge, MA) or cathepsin K (ab19027; Abcam, Cambridge, MA) antibodies. Samples were deparaffinized and treated with Ventana Cell Conditioning Solution for 32 min at 95 °C before staining. Primary antibodies were detected with anti-rabbit horse radish peroxidase and 3,3-diaminobenzidine (DAB) substrate (Ventana Roche OmniMap catalog #760–4311). Whole slide tissue sections were scanned with the NanoZoomer S210 digital slide scanner (Hamamatsu Photonics K. K., Hamamatsu, Shizuoka, Japan). Visiopharm quantitative digital pathology software (Visiopharm, Hoersholm, Denmark) was used to quantify CD68-positive or cathepsin K−positive cells using the cell classification application. The software was trained to identify DAB-positive staining, and all slides were batched and analyzed using the same algorithm. The calculated end point was the number of CD68-positive or cathepsin K−positive cells per area.

Serum was collected at necropsy and frozen at −80^o C until analysis. Rat anti-collagen type II (CII) immunoglobulin G (IgG) was quantified using an enzyme-linked immunosorbent assay (ELISA) kit (Chondrex, Inc., Redmond, WA) following the manufacturer’s instructions.

Clinical data for ankle diameter and histopathology scores were analyzed by 2-way analysis of variance (ANOVA) followed by a Newman-Keuls, Dunn’s, or Dunnet’s multiple comparison test. Bone density and cortical shell volume differences by micro-CT analysis were analyzed for differences using 1-way ANOVA followed by a Holm-Sidak multiple comparisons test using GraphPad Prism 6.0 (GraphPad Software, San Diego, CA).

Rats were assigned to treatment groups according to body weight (weight: 120–170 g) and mean ankle caliper measurements (see Additional file 1, Table 1). Rats immunized with collagen developed arthritis, as assessed by paw edema by measuring ankle diameter. Treatment with ETN caused significant decreases in ankle diameter versus the disease control group (Fig. 1). Treatment with RCI also caused a significant dose-dependent reduction in ankle swelling compared with the disease control group, and the effects were additive when combined with ETN (Fig. 1). Neither ETN nor RCI treatment caused a reduction in body weight (see Additional file 1, Fig. 2). Treatment with ETN alone did not reduce paw or spleen weight (see Additional file 1, Fig. 2). At 400 U/kg, RCI significantly reduced paw and spleen weight, and paw weight was further reduced with combined RCI and ETN treatment (see Additional file 1, Fig. 2).

Microscopic examination of diseased paws revealed a destructive polyarthritis in 100% of the disease control rats. This was characterized by synovial and periarticular edema, mixed inflammatory cell infiltration (monocyte/macrophages, neutrophils, and lymphocytes) within the joint space and synovial/periarticular tissue, marked cartilage damage, and mild to moderate bone resorption (see Additional file 1, Fig. 3; Fig. 2a and b). At 400 U/kg, RCI significantly reduced histologic evidence of injury as reflected by mean summed histopathology score and individual histologic parameters, including inflammation, pannus, cartilage damage, and bone resorption, versus the disease control group (Fig. 2a and b). Combined RCI and ETN treatment showed additional benefits on these parameters. Histologic evidence of ankle inflammation was significantly reduced in rats receiving 160 U/kg RCI compared with the disease control group, although summed histopathology scores and other parameters were not significantly attenuated at this lower dose. At 40 U/kg, RCI had no significant effects on histologic parameters (see Additional file 1, Fig. 4). Treatment with ETN alone did inhibit disease, as assessed by the summed ankle histopathology score (Fig. 2b); however, a significant effect of ETN was seen only on histologic evidence of cartilage damage compared with disease control rats (Fig. 2a).

Table 1 shows the percent inhibition of various disease parameters after ETN and/or RCI treatment compared with the disease control group. Monotherapy with ETN showed a significant reduction in cartilage damage. However, monotherapy with RCI 400 U/kg showed significant decreases in inflammation, pannus formation, cartilage damage, bone resorption, and paw edema. These disease parameters were further reduced with combined treatment with RCI (160 or 400 U/kg) and ETN.

To better understand the cellular response in rat CIA, we used IHC analysis (see Additional file 1, Fig. 5) to semi-quantitatively determine the monocyte and osteoclast numbers as well as serum anti-CII antibody levels to evaluate the B-cell response (Fig. 3). Monocyte/macrophage infiltration and activation have been correlated with joint pain and the general inflammatory status of patients with RA [34, 35]. Ankles from rats treated with RCI or the RCI/ETN combination displayed a substantial reduction in CD68-positive macrophages versus the disease control group (Fig. 3a). Furthermore, rats treated with ETN and RCI in combination had significantly fewer cathepsin K−positive osteoclasts than did disease control rats (Fig. 3b).

B cells are believed to play a critical role in arthritis disease induction, autoantibody production, and disease progression [36]. Treatment with RCI in combination with ETN inhibited anti-CII antibody levels by > 50%. Neither ETN nor RCI alone had a significant effect on serum anti-CII antibody levels (Table 1).

To further evaluate the effects of RCI alone or in combination with ETN on disease-mediated bone destruction, micro-CT of hind limb ankle joints was performed at the termination of the study. Rendering of micro-CT images demonstrated the smooth bone surface and distinctive bone architecture of the ankle and foot for naive (non-diseased) control rats. In contrast, diseased ankle joints from rats with CIA displayed extensive bone destruction, highlighted by rough bone surface and severe dimpling, periosteal proliferation, loss of bone mass, and fused bone joints. Treatment with RCI (400 U/kg) substantially improved the damaged appearance of the bone, and treatment with RCI in combination with ETN displayed even more improvement in bone appearance compared with disease control rats (Fig. 4).

To quantify changes in bone architecture, average bone density and cortical shell volume differences were calculated. As shown in Table 2, CIA disease control rats had a lower average bone density and cortical shell volume difference than naive rats. Treatment with RCI 400 U/kg inhibited bone density loss and cortical shell volume difference, reflecting a positive effect on structural damage. The protective effects of RCI on average bone density were further potentiated in combination with ETN; the CIA-induced decrease in average bone density was significantly improved by combined treatment with 400 U/kg RCI and ETN (Table 2). Furthermore, rats that received combination treatment (RCI 160 or 400 U/kg + ETN) had significantly higher cortical shell volume bone measurements than both the disease control and ETN monotherapy groups. In contrast, treatment with ETN alone had no significant effect on these bone parameters.