Effects of Etanercept on Experimental Osteoarthritis in Rats: Role of Histone Deacetylases

Abstract

Objective

Mounting evidence suggests that histone deacetylases (HDAC) inhibitors reduce cartilage destruction in animal models of osteoarthritis (OA). Tumor necrosis factor (TNF)-α-blocking treatment for OA may provide effective joint protection by slowing joint damage. To investigate the effects of intraperitoneal administration of etanercept (a TNF-α inhibitor) on OA development in rats and changes in the nociceptive behavior of rats and expression of HDACs, RUNX2, and MMP13 in cartilage.

Methods

Induction of OA in Wistar rats was accomplished through anterior cruciate ligament transection (ACLT). One or five milligrams (mg) of etanercept was administered intraperitoneally for 5 consecutive weeks after ACLT to the ACLT + etanercept (1 and 5 mg/kg) groups. Nociceptive behavior and changes in knee joint width were analyzed. Cartilage was evaluated histologically and immunohistochemically.

Results

ACLT + etanercept significantly improved mechanical allodynia and weight-bearing distribution compared to ACLT alone. In OA rats treated with etanercept, cartilage degeneration and synovitis were significantly less pronounced than those in ACLT rats. OA-affected cartilage also showed reduced expression of HDAC 6, 7, RUNX-2, and MMP-13 in response to etanercept but increased expression of HDAC4.

Conclusion

Our study demonstrated that etanercept therapy (1) attenuated the development of OA and synovitis in rats, (2) reduced nociception, and (3) regulated chondrocyte metabolism, possibly by inhibiting cell HDAC6 and HDAC7, RUNX2, and MMP13 and increasing HDAC4 expression. Based on new evidence, etanercept may have therapeutic potential in OA.

Introduction

Approximately 18% of women and 10% of men older than 60 years experience knee osteoarthritis (OA), which has a significant socioeconomic impact. ¹ Cartilage breakdown, bone thickening beneath the articular surface, synovial inflammation, and osteophyte formation are hallmarks of OA. ² Interleukin-1 (IL-1) and tumor necrosis factor-α (TNF-α) are inflammatory cytokines and key components of OA pathophysiology. ³ These cytokines lead to the overproduction of matrix metalloproteinases (MMPs) in the intra-articular area. Because cartilage is composed of an extracellular matrix, chondrocytes are the major site of inflammatory intermediate production in OA. ⁴ TNF-α is a cytokine that induces inflammation in cells, particularly macrophages. ⁵ TNF-α-blocking treatment for OA may provide effective joint protection by slowing joint damage. ⁶ Despite its lower levels in the synovial fluid of patients with OA than in that of patients with rheumatoid arthritis (RA), TNF-α participates in the pathogenesis of the disease.

Etanercept (Enbrel; Amgen, Thousand Oaks, CA), a TNF-α inhibitor, is a recombinant soluble p75 receptor linked to the Fc domain of human immunoglobulin G. ⁷ Etanercept is currently used to treat arthritis symptoms in humans, including RA, juvenile, and psoriatic arthritis. ⁸ In a variety of rat strains and inflammation models, etanercept reduces the reaction of C-fibers to mechanical stimulation. ⁷ Animal OA models treated with etanercept exhibit lubricin deposition on articular cartilage and decreased sulfated glycosaminoglycan release from the cartilage. ⁹

In addition to the acetylation of histone and non-histone proteins, histone deacetylases (HDAC) influence many physiological and pathological processes. Over the past 10–15 years, HDAC inhibitors have drawn significant interest in drug discovery owing to their influence on the cell cycle and apoptosis. ¹⁰ HDAC inhibitors reduce synovial inflammation by upregulating synovial fibroblast cell cycle regulators by modifying histone acetylation.^11,12 In vivo models of inflammatory arthritis and OA have demonstrated that HDAC inhibitors reduce joint damage, substantiating their chondroprotective properties.^13,14 Mounting evidence suggests that HDAC inhibitors reduce cartilage destruction in animal models of OA, suggesting a protective role of HDAC.^15,16 The function of HDAC inhibitors in OA has yet to be fully understood. Here, we explored the potential mechanisms underlying the effects of etanercept on OA by examining the effects of the TNF-α inhibitor-etanercept. OA was modeled using an anterior cruciate ligament transection (ACLT) in rats.^17-19 Changes in pain behavior (nociception) were also assessed, including secondary mechanical allodynia, changes in the weight-bearing distributions of both hind feet, and changes in knee joint width.^20,21 Etanercept modulated HDAC4, HDAC6, and HDAC7 expression, improving the cartilage microstructure and inhibiting cartilage degeneration in ACLT-induced OA animal model as we have previously hypothesized. Herein, we provided experimental evidence of the potential effects of etanercept on OA.

Materials and Methods

Animal Model of OA

Animal experiments have been conducted in accordance with the Guiding Principles in Animal Care and Use as approved by the Council of the American Physiology Society and by the National Sun Yat-Sen University Animal Care and Use Committee (approval no. 10538). OA was induced in in the right knee of male Wistar rats (body weight, 285–310 g, 10 weeks old) using ACLT. To anesthetize the rats, oxygen and room air (1:1) were used in combination with 3% isoflurane. Adequate anesthesia was identified as no flexor withdrawal after a noxious foot pinch. A medial parapatellar incision on the skin was necessary to perform arthrotomy under anesthesia. A #15 blade was used to cut through the mid-substance of the ACL under direct vision and expose the ACL. An anterior drawing sign was used to measure the sufficiency of the sections. Saline was used to irrigate the joint, vicryl 4-0 was used to stitch the capsule, and 4-0 nylon mattress sutures were used to close the skin. An earlier protocol provides the basis for this procedure. ²² Rats were housed individually and exposed to a daily 12-hour light/dark cycle with ad libitum access to food and water. Animals with pre-existing anatomical or gait abnormalities were excluded. All efforts were made to minimize the number of animals used and their suffering. As for antibiotic prophylaxis, cefazolin was intraperitoneally administered (20 mg/kg) in each animal. Immobilization after surgery was not performed, and cage activity was allowed daily without restrictions. We observed the rats daily during recovery to ensure that the wounds healed normally.

Study Design and Etanercept Administration

Animals were randomly and blindly allocated into the following experimental groups: Group I received intraperitoneal injections of etanercept (1 mg/kg) (Enbrel; Amgen) once a week for 5 weeks starting 12 weeks after the ACLT (n = 10); Group II received intraperitoneal injections of etanercept (5 mg/kg) once a week for 5 weeks starting 12 weeks after the ACLT (n = 10); Group III (ACLT group) received intraperitoneal injections of 0.1 mL of normal physiological saline once per week for 5 weeks starting 12 weeks after the ACLT (n = 6); Group IV (sham) received arthrotomy without any treatment (n = 6); and Group V received intraperitoneal injections of etanercept 5 mg/kg only (n = 6) with no other intervention. At 24 weeks after ACLT, rats were sacrificed, and gross morphological and histopathological examinations were performed on the cartilage and synovial membrane of the knee joint. Etanercept was assessed in the articular chondrocytes by immunohistochemical analysis to examine its effect on HDAC4, HDAC6, HDAC7 expessions, runt-related transcription factor-2 (RUNX2), and MMP13.

Nociception Evaluation

Two nociceptive behavioral tests were performed during the light phase, and the animals were habituated to the procedure room for at least 60 minutes. The observers were blinded to the surgical and pharmacological treatment of the animals.

Secondary Mechanical Allodynia Test

The rats were stimulated by von Frey filament (North Coast Medical, Morgan Hill, CA) for 5 seconds. We used Chaplan’s “up-down” method using alternate large and small fibers to determine the 50% withdrawal threshold. ²³ The rats were given a weaker filament when they lifted their paws in response to pressure. A spreadsheet was used to calculate the 50% response threshold based on the sequence of responses. Each paw was exposed to a von Frey filament for 5 trials spaced approximately 3-minute apart.

Weight-Bearing Distribution Test

A weight-bearing distribution averager (Singa Technology Corporation, New Taipei City, Taiwan) was used to analyze the proportional weight distribution between the joints in the right (OA) and left (contralateral) knees when the joints were damaged. As described previously, ²⁴ altered weight-bearing distribution of the hind paws can be an indicator of joint discomfort. The difference in weight-bearing distribution between the right (OA) and left (control) hind paws was used as a measure of joint pain following knee OA induction surgery. In an angled plexiglass chamber, each hind paw rested on its force plate. Our experiment measured the force (in grams) of each hind leg over a 5-second period. Each data point was the average of 3 measurements taken over the 5 seconds. Hind paw weight distribution was assessed using the difference between the contralateral and ipsilateral hind legs.

Gross Morphology and Histopathological Examination

Knee joint width was measured once weekly from 12 to 24 weeks after ACLT. Sodium pentobarbital (50 mg/kg) was administered to induce deep anesthesia at 24 weeks after ACLT. For 2 weeks, the joint sections were buffered with 12.5% ethylenediaminetetraacetic acid and formalin solution. Serial cuts were made in the articular cartilage using a microtome (HM340E; Microm, Walldorf, Germany). Hematoxylin/eosin and Safranin O/Fast Green were used to stain matrix proteoglycans and cartilage morphology. Immediately after sacrifice, each knee was examined for gross morphological changes in the cartilage lesions as previously described. ²⁵ The articular cartilage was examined under a microscope to determine its grade according to the Osteoarthritis Research Society International (OARSI) grading system. ²⁶ Histological grades and stages were divided into six categories. The grade and stage were combined to determine the OARSI score (score = grade × stage), ranges from 1 point (control articular cartilage) to 24 points (no repair). Synovial membrane specimens were carefully dissected from the suprapatellar pouch and tibiofemoral compartments and embedded in paraffin, and 10 slides prepared from each knee were stained with hematoxylin & eosin (H/E) for synovitis score assessment by histology. ²⁷

Immunohistochemistry

The immunohistochemical analysis of cartilage specimens has been previously described.^13,14 Briefly, paraffin-embedded specimens were cut into 2-µm sections and deparaffinized using xylene, followed by ethanol dehydration. To retrieve the antigen, 30 minutes of proteinase K (20 mM; Sigma, St Louis, MO, USA) digestion in phosphate-buffered saline (PBS) was used. To block non-specific binding, we gently washed the sections twice in PBS for 20 minutes and then incubated them for 1 hour in 5% normal goat serum in PBS. During incubation, the slides were treated with anti-HDAC4 (1:100; Abcam, Cambridge, MA; polyclonal antibody ab12172), anti-HDAC6 (1:100; Bioss, Woburn, MA; polyclonal antibody bs-2811R), anti-HDAC7 (1:100; Abcam; polyclonal antibody ab12174), and anti-Runx2 (1:150; Abcam; monoclonal ab76956), and MMP13 (1:150; Abcam, polyclonal ab39012). Next, the sections were treated with an avidin-biotin complex kit (Vectastain ABC kit; Vector Laboratories, Burlingame, CA). The negative control group was subjected to the same procedures but with goat anti-rabbit IgG (Cat: 111-035-144; Jackson ImmunoResearch Laboratories Inc., West Grove, PA, USA) or goat anti-mouse IgG (Cat: 115-035-166; Jackson ImmunoResearch Laboratories Inc., West Grove, PA) instead of the primary antibody. We used a Leica DM-6000 microscope (Leica, Hamburg, Germany) and a SPOT CCD ideal integrating camera (Diagnostic Instruments, Sterling Heights, MI) to observe and capture the images. Chondrocytes that stained positive throughout the cartilage thickness were counted to quantify and estimate the antigens present in each cartilage specimen. ²⁸ We averaged the results from six microscopic fields (three in each of the superficial and deep zones). Intact cartilage surfaces of each OA specimen were checked before the morphometric analysis for validation. In the final analysis, for each cartilage specimen, the total number of stained chondrocytes was expressed as a percentage (cell score). Each slide was reviewed by two independent readers who were blinded to the treatment groups.

Data Analysis and Statistics

The standard error of mean was calculated for all data. In the experimental groups and for scores with significant differences, a one-way analysis of variance (ANOVA) was conducted. To compare the mean differences between the treatment and sham groups, the Dunnett’s test post hoc test was applied. Using repeated-measures ANOVA, we examined the trends in nociceptive behavior and knee joint width over time. We considered P < 0.05 as statistically significant.

Results

Etanercept’s Effects on OA Nociception and Inflammation Induced by ACLT

We administered low and high doses (1 and 5 mg/kg, respectively) of etanercept intra-articularly to the rats after ACLT to investigate its inhibitory effects on nociceptive sensitization and pathological progression in OA. A diagram of hind paw weight distribution is shown in Figure 1A , and a diagram of secondary mechanical allodynia is shown in Figure 1B . There was a significant increase in hind paw weight distribution changes compared to the sham group (66.11 ± 3.32 g vs. 4.96 ± 0.83 g; P < 0.01; Fig. 1A ). At 12 weeks after ACLT surgery (before drug administration), the withdrawal threshold of the paws had decreased significantly (1.20 ± 0.09 g vs. 8.00 ± 0.73 g; P < 0.01; Fig. 1B ). Asymmetry in weight-bearing and mechanical allodynia was accompanied by ALCT-induced OA progression. As determined by the weight-bearing distribution test, etanercept exhibited dose-dependent inhibitory effects in the ACLT + etanercept (1 and 5 mg/kg) groups. Over the first 5 weeks of etanercept administration, a significant difference in weight-bearing distribution in the ACLT group was noted with 5 mg/kg etanercept versus the ACLT group without etanercept (62.73 ± 2.82 g vs. 64.95 ± 2.62 g; 42.53 ± 5.89 g vs. 66.35 ± 3.85 g; 36.65 ± 5.55 g vs. 63.51 ± 7.02 g; 49.25 ± 6.36 g vs. 66.75 ± 5.72 g; 47.82 ± 9.92 g vs. 70.98 ± 3.14 g, respectively; P < 0.05; Fig. 1A ) after drug injection. In addition, the ACLT + etanercept 1 mg/kg and ACLT groups showed the same trend in the first and fifth weeks (65.09 ± 4.85 g vs. 64.95 ± 2.62 g; 57.34 ± 5.57 g vs. 66.35 ± 3.85 g; 55.14 ± 3.00 g vs. 63.51 ± 7.02 g; 52.55 ± 4.72 g vs. 66.75 ± 5.72 g; 56.47 ± 6.00 g vs. 70.98 ± 3.14 g, respectively; P < 0.05; Fig. 1A ) after drug injection. In addition, the inhibitory effects persisted for up to 6 weeks after the end of etanercept administration (1 and 5 mg/kg; 14.20 ± 5.61 g vs. 66.11 ± 3.32 g; 23.26 ± 3.44 g vs. 66.11 ± 3.32 g; P < 0.05; Fig. 1A ). Etanercept administration at weeks 1–5 significantly increased paw withdrawal thresholds versus ACLT administration ( Fig. 1B ).

Figure 1.

Etanercept injections administered intraperitoneally attenuated ACLT-induced nociception and knee joint swelling. (A) Post-ACLT weight-bearing deficits in the hind limbs, (B) secondary mechanical allodynia, and (C) knee swelling (change in knee joint width) were studied over time after etanercept administration. Rats in the ACLT + etanercept groups were intra-articularly injected with etanercept (1 or 5 mg/kg weekly) in weeks 12–16 after ACLT. Data are expressed as mean ± standard error of the mean for each group (*P < 0.05 vs. sham group; #P < 0.05 vs. ACLT group). ACLT = anterior cruciate ligament transection.

Before etanercept treatment, the ACLT group experienced significantly increased knee joint swelling. Throughout the etanercept administration period, the width of the hind limb knee joint significantly decreased in the ACLT + etanercept 5 mg/kg group versus the ACLT group (0.82 ± 0.04 mm vs. 0.90 ± 0.04 mm; 0.66 ± 0.03 mm vs. 0.94 ± 0.07 mm; 0.53 ± 0.04 mm vs. 0.96 ± 0.04 mm; 0.56 ± 0.07 mm vs. 0.97 ± 0.07 mm; 0.45 ± 0.05 mm vs. 0.91 ± 0.06 mm, respectively; P < 0.05; Fig. 1C ). In the ACLT + etanercept 1 mg/kg versus ACLT groups, a similar trend was observed (0.83 ± 0.09 mm vs. 0.90 ± 0.04 mm; 0.76 ± 0.08 mm vs. 0.94 ± 0.07 mm; 0.68 ± 0.07 mm vs. 0.96 ± 0.04 mm; 0.72 ± 0.08 mm vs. 0.97 ± 0.07 mm; 0.61 ± 0.09 mm vs. 0.91 ± 0.06 mm, respectively; P < 0.05; Fig. 1C ). Based on these findings, etanercept treatment reduced ACLT-induced OA pain and inflammation in rats.

Attenuation of Gross Morphology, Cartilage Degradation, and Synovitis in ACLT-Induced OA by Etanercept Treatment

In the ACLT group, there were gross signs of cartilage degeneration, including fibrillation, erosions, and ulcers, as well as osteophytes. Significant differences in gross cartilage damage area (%) were found between the ACLT, ACLT + etanercept 1 mg/kg, and ACLT + etanercept 5 mg/kg groups and the sham group (Table 1). According to quantitative analysis, cartilage damage was significantly reduced in the ACLT+ etanercept 1 and 5 mg/kg group compared to ACLT alone (Table 1). Etanercept-treated rats had a smaller area of eroded cartilage surface in their joints (Table 1). Safranin O/Fast Green staining was used to determine whether etanercept treatment prevented cartilage degradation in the sham, ACLT, ACLT + etanercept (1 and 5 mg/kg), and etanercept 5 mg/kg groups. The articular cartilage in the sham group appeared normal. Cartilage erosion and superficial destruction were observed in the ACLT versus sham group. In contrast, intraperitoneal injections of etanercept (1 and 5 mg/kg) prevented cartilage and bone erosion ( Fig. 2A ). Using Safranin O/Fast Green staining, the histopathology of the cartilage was further assessed using the OARSI histopathological scoring system. The mean OARSI score of the ACLT group (13.00 ± 2.48; P < 0.05) was significantly higher than that of the sham group (1.50 ± 0.77; P < 0.05). The group that received etanercept (1 and 5 mg/kg) had significantly lower OARSI scores than the ACLT group (2.67 ± 1.89 vs. 13.00 ± 2.48; 5.42 ± 1.88 vs. 13.00 ± 2.48, respectively; P < 0.05; Fig. 2B ). As a result of ACLT + etanercept (5 mg/kg) and ACLT + etanercept (1 mg/kg), no significant differences in OARSI scores were found. Compared with the sham group, etanercept (5 mg/kg) alone had no significant effect on OARSI scores. Etanercept administration inhibited cartilage degradation in rats with ACLT-induced OA. Twenty-four weeks after surgery, none of the experimental groups showed any changes in the gross or histological appearance of their hips or ankles. Synovial membranes from the ACLT group were hypertrophic and appeared discolored reddish yellow, and those from the ACLT + etanercept 1- and 5-mg/kg groups were thinner and displayed less-intense discoloration ( Table 1 and Fig. 3 ). No evidence of synovitis or hyperemia was present in the synovial membranes of the sham and etanercept 5-mg-alone groups ( Fig. 3 ).

Table 1.

Macroscopic (Gross Cartilage Damage Area [%]) and Histological Evaluation of the Synovial Membrane.

Group	Score
	Macroscopic scores	Synovitis scores
Sham	0.5 ± 0.22	1.2 ± 0.82
ACLT	3.0 ± 0.36*	9.2 ± 1.03*, #
ACLT + etanercept 1 mg	2.2 ± 0.18*, #	6.1 ± 0.56*, #
ACLT + etanercept 5 mg	1.7 ± 0.21*, #	4.8 ± 0.49*, #
Etanercept 5 mg	0.2 ± 0.18	1.8 ± 0.31

For the macroscopic score and synovitis score, refer to Materials and Methods section. Data are expressed as mean ± SEM.

ACLT = anterior cruciate ligament transection.

*

P < 0.05 versus the sham group.

#

P < 0.05 versus the ACLT group.

Figure 2.

Evaluation of histopathological changes in knee joints of ACLT rats after etanercept treatment. Safranin O/Fast Green stain was applied to knee joints from the sham, ACLT, ACLT + etanercept (1 or 5 mg/kg), and etanercept (5 mg/kg) groups. OARSI scores were determined to measure histopathological quantitative changes in the knee joints. As shown in the histogram, the OARSI scores for sham, ACLT, ACLT, and etanercept (5 mg) were compared. Data are expressed as mean ± standard error of the mean for each group. The scale bar represents 200 μm (*P < 0.05 vs. the sham group; ^#P < 0.05 vs. the ACLT group). ACLT = anterior cruciate ligament transection.

Figure 3.

Histopathological evaluation of the synovial membrane in knee joints of ACLT rats after etanercept treatment. In both the sham and etanercept 5-mg groups, the synovial membrane showed no synovial lining cell hyperplasia (within normal limits). The specimens from the ACLT group were thick, with focal villi, hyperplasia of the lining cells, and moderate infiltration of mononuclear inflammatory cells. Both the specimens from the ACLT + etanercept (1 or 5 mg/kg) groups show less severe synovitis and tissue repair than both the ACLT groups. (H/E stain, original magnification, ×100). Scale bar = 100 μm.

Etanercept Affects HDAC4, HDAC6, and HDAC7 Expression in the ACLT-Induced OA Model

Etanercept treatment ameliorated ACLT-induced cartilage damage progression as shown by HDAC4, HDAC6, and HDAC7 immunostaining. Cells expressing HDAC4, HDAC6, and HDAC7 were noted in the cartilage tissues of knee joints treated with sham, ACLT, ACLT + etanercept (1 and 5 mg/kg), and etanercept (5 mg/kg) ( Fig. 4A ). Immunohistochemistry analysis showed that the number of HDAC4-positive cells significantly decreased in the ACLT group (14.95 ± 4.88%) compared to that of the sham group (49.96 ± 8.92%; P < 0.05; Fig. 4B ). Etanercept (1 mg/kg) and 5 mg/kg reversed the ACLT-induced reduction in the number of HDAC4-positive cells in a dose-dependent manner (36.96 ± 8.28% and 37.48 ± 7.38%, respectively; Fig. 3B ). HDAC-6- and HDAC-7-positive cells were significantly more abundant in the ACLT group versus the sham group (44.79 ± 7.89% vs. 5.86 ± 0.82%; 48.73 ± 5.52% vs. 11.51 ± 5.15%, respectively; P < 0.05; Fig. 4C and D). Attenuation of the ACLT’s effects on HDAC6- (13.60 ± 5.61% vs. 44.79 ± 7.89%; 10.03 ± 3.43% vs. 44.79 ± 7.89%, respectively; P < 0.05; Fig. 4C ) and HDAC7-positive (29.25 ± 7.79% vs. 48.73 ± 5.52%; 18.83 ± 5.52% vs. 48.73 ± 5.52%, respectively; P < 0.05; Fig. 4D ) cell numbers in chondrocytes was seen in the ACLT + etanercept groups. Accordingly, etanercept administration significantly increased HDAC4 activity and decreased HDAC6 and HDAC7 activities in the cartilage tissues after ACLT. Compared with the sham group, etanercept (5 mg/kg) alone did not affect the expression of HDAC4, HDAC6, or HDAC7.

Figure 4.

Etanercept affects HDAC4, HDAC6, and HDAC7 expression in cartilage tissues after ACLT. In the sham, ACLT, ACLT + etanercept (1 or 5 mg/kg), and etanercept 5 mg/kg groups, HDAC4, HDAC6, and HDAC7 immunohistochemical analyses were performed of joint sections from the knee joints. Negative control were incubated rabbit IgG without primary antibody showing no specific staining. The ratios of HDAC4-positive (B), HDAC6-positive (C), and HDAC7-positive (D) cells in the joint sections are presented quantitatively. Data are presented as mean ± standard error of the mean for each group. Scale bar represents 100 μm (*P < 0.05 vs. the sham group; ^#P < 0.05 vs. the ACLT group). ACLT = anterior cruciate ligament transection; HDAC, histone deacetylase.

Etanercept affects RUNX2 and MMP13 expression in the ACLT-induced OA model

In the present study, Figure 5A illustrates the distribution of RUNX2- and MMP13-positive cells in the cartilage tissues of knee joints after sham, ACLT, ACLT + etanercept (1 and 5 mg/kg), and etanercept (5 mg/kg) treatments. Quantitative analysis of immunohistochemical staining showed that compared with the sham group, RUNX2- and MMP13-positive cells were significantly higher in number after ACLT surgery (57.61 ± 5.13% vs. 9.25 ± 0.99%; 74.24 ± 4.65% vs. 11.48 ± 0.95%, respectively; P < 0.05; Fig. 5B and C). However, the low- and high-dose etanercept treatments markedly decreased the number of RUNX2- (43.18 ± 2.34% vs. 57.61 ± 5.13%; 36.74 ± 3.59% vs. 57.61 ± 5.13%, respectively; P < 0.05; Fig. 5B ) and MMP13-positive cells compared with ACLT-only treatment (58.82 ± 5.04% vs. 74.24 ± 4.65%; 52.94 ± 4.21% vs. 74.24 ± 4.65%, respectively; P < 0.05; Fig. 5C ). No significant differences were noted between the low- and high-dose etanercept treatment groups (Fig. 5B and C). Based on these findings, etanercept administration significantly decreased the expression of RUNX2 and MMP13 in cartilage tissues after ACLT, whereas RUNX2 and MMP13 expression levels in the sham group were unaffected by etanercept (5 mg/kg) treatment.

Figure 5.

Etanercept affects RUNX2 and MMP13 expression in cartilage tissues following ACLT. (A) Immunohistochemistry of RUNX2 and MMP13 in knee joint sections from sham, ACLT, ACLT + etanercept (1 or 5 mg/kg), and etanercept 5 mg/kg groups. Negative control were incubated mouse IgG without primary antibody showing no specific staining. We then quantitatively analyzed the ratios of (B) RUNX2-positive cells versus (C) MMP13-positive cells in the joint sections. In each group, the data are expressed as mean ± standard deviation. Scale bar represents 100 μm (*P < 0.05 vs. the sham group; ^#P < 0.05 vs. the ACLT group). ACLT = anterior cruciate ligament transection; MMP13 = matrix metallopeptidase 13; RUNX2 = runt-related transcription factor-2.

Discussion

To the best of our knowledge, this is the first report showing that the intraperitoneal etanercept administration reduces knee inflammation and prevents OA development as well as secondary mechanical allodynia and weight-bearing distribution in experimental OA rats. Interestingly, etanercept enhanced the expression of HDAC4, decreased the expression of HDAC6 and HDAC7, and attenuated the expression of RUNX2 in chondrocytes along with MMP13, its downstream effector.

RA, juvenile arthritis, and psoriatic arthritis are currently being treated with etanercept, a recombinant-soluble p75 receptor linked to the Fc portion of human immunoglobulin G. ²⁹ The increased production of TNF-α by activated synoviocytes and articular chondrocytes and increased expression of the p55 TNF-α receptor in chondrocytes suggests that TNF-α-mediated matrix degradation is a key factor in the pathogenesis of OA. ³⁰ Clinical scores for knee pain are correlated with the number and expression of activated macrophages in inflamed synovial tissues. ³¹ Etanercept injections effectively treat moderate and severe OA knee pain. ³² Furthermore, this finding suggests that TNF-α causes OA pain. Etanercept treatment significantly decreases nuclear factor-kappa B, Jun N-terminal kinase (JNK), and extracellular-signal-regulated kinase (ERK) levels but not p38 activation. ³³ With a half-life of 70–100 hours, etanercept reaches a peak serum concentration 48–60 hours after subcutaneous injection. ³⁴ At a systemic dose, etanercept does not cross the blood-brain barrier. ³⁵ Its systemic administration significantly reduces thermal hyperalgesia and tactile allodynia in chronic constriction and nerve injury models. ³⁶ In an ACLT rat model, Castro et al. ³⁷ created a quantitative model that correlates joint pain with weight-bearing. The present study examined secondary mechanical allodynia, performed a weight-bearing distribution test, and detected changes in knee joint width after the intraperitoneal injection of etanercept in OA knees. The results of the ACLT + etanercept group were significantly better than those of the ACLT group with regard to secondary mechanical allodynia, weight-bearing distribution, and knee joint width changes ( Fig. 1 ). In this study, etanercept treatment decreased nociception and inflammation. However, further study is required to determine exactly how etanercept exerts its antinociceptive and anti-inflammatory effects.

Destabilization caused by ACLT has been used as a model to study joint changes that closely match those observed in human OA. ²² Fibrillation, collagen and proteoglycan network disorganization, joint capsule thickening, and osteophyte formation can occur in the joint. ²⁴ Compared to animals that received only ACLT surgery, animals treated with etanercept scored significantly lower on the gross morphology and OARSI questionnaire (Table 1 and Fig. 2 ). Cartilage degradation severity after ACLT knee surgery significantly decreased with intraperitoneal etanercept administration ( Fig. 2 ). There is also evidence that synovial membrane inflammation plays a significant role in the development of OA. ¹⁵ Manifestations of inflammation associated with the ACLT model include joint effusions and synovial membrane hyperplasia,^14,22 both of which are important factors in the development of joint pain in patients with OA. ³⁴ In the present study, moderate synovitis and OA development were noted in the ACLT group, and treatment with both 1 and 5 mg/kg of etanercept reduced the severity of synovitis (Table 1 and Fig. 3 ). This is a novel finding that etanercept could reduce synovitis and OA in an experimental OA rat model.

HDACs are expressed in both healthy and diseased cartilage in patients with OA.^11,15,38 In terms of structural, functional, localization, and expression patterns, there are 18 known HDACs in the human and mouse genomes. HDACs are divided into four families (classes I, II, III, and IV). Class I HDACs (HDAC1, HDAC2, HDAC3, and HDAC8) use Zn²⁺ to perform their enzymatic activity. In contrast, class III HDACs, commonly referred to as sirtuins (Sirt1–7), rely on nicotinamide adenine dinucleotides (NAD⁺) for catalysis.^15,38 HDAC activity, expression, and distribution may be altered in OA, and HDAC inhibitors may prevent cartilage degradation and chondrocyte damage.^14,15,38 As skeletogenesis proceeds, HDAC4 inhibits RUNX2 activity, which is expressed in pre-hypertrophic chondrocytes. ³⁹ In OA cartilage degeneration, Cao et al. ⁴⁰ reported that HDAC4 negatively regulates RUNX2, MMP13, and COL10. The transcription factor RUNX2 induces COL10 and MMP13. ⁴⁰ MMP13 is more abundant in late-stage OA cartilage than in early OA cartilage or normal knee cartilage. ⁴³ The nuclear factor-kappa B pathway may also be involved in regulating MMP expression. ⁴² In addition to destroying cartilage, MMPs contribute to the breakdown of extracellular matrix components.^41,42 In addition to irreversible joint damage in OA, Ma et al. ⁴³ observed that MMP13, the main collagenase-degrading type II collagen in cartilage, contributes to OA onset and initiation. In addition, cartilage from OA patients showed lower HDAC4 mRNA levels than that from normal donors. ⁴⁰ Runx2, which regulates transcription, is connected to the upstream p38-caspase 3 signaling pathway by HDAC4. ⁴⁴ RUNX2 is inhibited by the caspase 3-HDAC4-p38 signaling cascade, which results in hypertrophic chondrocytes and bone formation. This study contributes to our understanding of bone growth by explaining chondrocyte hypertrophy in relation to p38 and HDAC4. In contrast, other studies reported that OA chondrocytes express high HDAC4 levels. ⁴⁰ In the ACLT group, HDAC4-positive cells were significantly diminished compared to HDAC4-positive cells in the control group, which supports our present observations. The inconsistent expression patterns of HDAC4 in OA chondrocytes could result from the different stages of the OA tissue samples. ⁴⁵ ACLT + etanercept treatment resulted in significantly more HDAC4-positive cells than ACLT alone as measured in the present study. In addition, we observed a significant decrease in the number of RUNX2- and MMP13-positive cells in the ACLT + etanercept group ( Fig. 5 ). HDAC4 plays an important role in OA pathogenesis; however, further research is required.

Murine models of RA and systemic lupus erythematosus have shown that HDAC6 overexpression promotes inflammatory responses, whereas its inhibition reduces disease activity. ⁴⁶ In the presence of high HDAC6 expression levels, mitochondrial dysfunction and reactive oxygen species production lead to increased extracellular matrix degradation. ⁴⁷ Using ricolinostat (ACY-1215), a selective inhibitor of HDAC6, in mice with OA reportedly reduced cartilage degradation. ⁴⁸ Therefore, OA can be treated by targeting HDAC6. In patients with OA, MMP13 overexpression is induced by HDAC7 overexpression in the knee cartilage. The development of cartilage and OA is critically dependent on HDAC7. ³⁸ The in vitro inhibition of HDAC7 inhibited inflammatory factor–induced MMP13 gene expression, consistent with the inhibition of HDAC7 in vivo. MMP13 is downregulated by miR-193b-5p in response to HDAC7 upregulation in human OA, which reduces cartilage degradation by inhibiting MMP13 expression. ⁴⁹ According to our results, the ACLT group had significantly more HDAC6- and HDAC7-positive cells than the control group ( Fig. 4 ). Etanercept treatment significantly reduced the number of HDAC6- and HDAC7-positive cells ( Fig. 4 ).

Etanercept is a Food and Drug Administration (FDA)-approved TNF-α inhibitor used to treat various inflammatory conditions such as psoriasis, RA, and ankylosing spondylitis. A prior study revealed that subcutaneous administration of etanercept at 0.5 mg/kg, either 2 or 7 times, resulted in chondroprotective effects, thereby reducing cartilage degradation in the knee joints of rats with ACLT-induced OA. ⁵⁰ In the present study, subcutaneous administration of etanercept 5 times at 1 mg/kg demonstrated inhibition of OA pain behaviors and cartilage degradation in ALCT rats. These findings suggest the potential application of etanercept in translational medicine. However, systemic administration of etanercept has been associated with common adverse reactions such as headache, pneumonia, and pruritis.^51,52 Injecting drugs directly into affected joints may enhance local drug availability while minimizing systemic toxicity risks. ⁵³ Another important consideration is identifying OA animal models that accurately mimic human OA.^54,55 Successful translation of therapies from the lab to clinical use hinges on using appropriate preclinical disease models. Therefore, conducting experiments in various animal models, such as rabbits or pigs, is imperative to facilitate future progress.

In summary ( Fig. 6 ), ACLT dysregulates the expression of HDAC3, HDAC4, and HDAC6 proteins in the chondrocytes of cartilage tissue. The resulting upregulation of RUNX2 and MMP-13 leads to the exacerbated degradation of the cartilage matrix, nociception, weight-bearing defects, and mechanical allodynia. Etanercept has a chondroprotective effect against articular cartilage degradation and inhibits nociception. The protective mechanisms of etanercept may be related to the attenuation of ACLT-induced OA-associated dysregulation of HDAC4/6/7 isoforms, which repress RUNX2 and its downstream effector MMP13.

Figure 6.

Schematic diagram of the possible protective mechanism of etanercept in ACLT-induced OA. ACLT = anterior cruciate ligament transection; OA = osteoarthritis; HDAC = histone deacetylase; RUNX2 = runt-related transcription factor 2; MMP = matrix metalloproteinase.

Conclusions

Our findings indicate that the intraperitoneal administration of etanercept aid in reducing the progression of OA and associated nociceptive behaviors. Upregulation of HDAC4 and downregulation of HDAC6 and HDAC7 in OA cartilage tissue may have chondroprotective effects in patients with early stage OA. The study findings could be beneficial in developing promising therapeutic strategies for OA and other forms of arthritis.