Validity evaluation of a rat model of monoiodoacetate-induced osteoarthritis with clinically effective drugs

Abstract

Background

Knee osteoarthritis (KOA) is the most common type of joint disease in elderly people and is characterized by pain and dysfunction. Although the monoiodoacetate (MIA)-induced model is widely used as a rodent KOA model, it is important to acknowledge the inherent limitations of this model, as the MIA model develops complex pathological phases on a daily basis. An accurate understanding of this model and the selection of an appropriate time point according to the target for drug candidates can lead to the development of clinically effective drugs.

Methods

Changes in the pathological state of the MIA model were assessed via histopathological evaluation. Clodronate, a bisphosphonate, and diclofenac, a nonsteroidal anti-inflammatory drug (NSAID), were selected as models of clinically effective drugs due to their different mechanisms of action. The analgesic effects of both drugs on the MIA model were evaluated. The long-term effect of clodronate on subchondral bone osteoclasts was also evaluated.

Results

Histopathological evaluation revealed that MIA-induced symptomatic behavior occurred in the early and late phases and was accompanied by synovial inflammation and osteoclast-related joint degeneration, respectively. Although clodronate inhibited symptomatic behavior and prevented cartilage degeneration from the early to late phases, diclofenac inhibited symptomatic behavior only in the early phase. Clodronate acted locally and inhibited the activation of subchondral osteoclasts.

Conclusions

Pathological changes, such as synovial changes in the early phase and knee joint degeneration in the late phase, in the MIA model are similar to those in human KOA. Our results indicate that the early phase in the MIA model is appropriate for evaluating the effects of anti-inflammatory agents such as NSAIDs and corticosteroids. The late phase in the MIA model is appropriate for evaluating the effects of drugs that act on cartilage and subchondral bone.

Introduction

Knee osteoarthritis (KOA) is the most common type of joint disease among individuals aged > 60 years. There are approximately 365 million KOA patients worldwide, and this number is expected to increase with the aging population [1]. This condition, characterized by chronic knee pain, not only significantly diminishes patients' quality of life but also imposes a considerable burden in terms of healthcare needs and expenses over prolonged periods [2–4].

Presently, treatment options for KOA primarily include exercise therapy, pharmacotherapy, and surgical interventions [5]. Total knee replacement (TKR) surgery has emerged as a principal surgical solution; however, its application is often discouraged in young patients because of the need for revision TKR [6]. Given the absence of disease-modifying osteoarthritis drugs, analgesics that relieve knee pain are of paramount importance in the management of this condition. Despite the approval and use of certain analgesics, such as nonsteroidal anti-inflammatory drugs (NSAIDs) and corticosteroids, for the treatment of KOA, their clinical efficacy remains limited. A significant proportion of patients find NSAIDs inadequate for pain control [7], while intra-articular corticosteroid injections often provide only short-term relief after injection, necessitating repeated administration [8]. Consequently, a pressing need exists for the development of more effective analgesic agents.

Identifying potential drug candidates in appropriate animal models reflecting the pathology of human KOA is essential for developing clinically effective drugs. The monoiodoacetate (MIA)-induced model is widely used as a rodent osteoarthritis model [9, 10]. Several reports have shown that the behavioral and symptomatic changes in the MIA model are similar to those in human KOA [9–11], which suggests the effectiveness of this model as a preclinical animal model. On the other hand, it is necessary to consider the limitation of this model, which is that the MIA model developed a complex pathological phase depending on the days after MIA injection. Therefore, an accurate understanding of this model by comparing it to the pathology of human OA and selecting the most appropriate time points for drug candidates according to the target may lead to the development of clinically effective drugs.

In this study, we first investigated pathological changes in the knee joint following MIA injection. The effects of diclofenac and clodronate on MIA-induced osteoarthritis-related knee pain were subsequently evaluated to determine the relationship between the pathological phase and the efficacy of the drugs. We chose diclofenac and clodronate because both of these drugs have analgesic effects in clinical studies [12–14], and the pathology of the MIA model is associated with changes in the synovium and subchondral erosion, which are targeted by diclofenac and clodronate, respectively [9, 10]. Oral and topical diclofenac, an NSAID, is an agent approved for treating KOA in humans (Zipsor, Voltaren Gel). Diclofenac is thought to exert an anti-inflammatory effect by suppressing the activation of inflammatory cells [15]. The other drug, clodronate, a first-generation bisphosphonate, is approved as an antiresorptive agent and has an analgesic effect on KOA in humans after intra-articular injection [13, 14]. Although clodronate is thought to have chondroprotective and anti-inflammatory effects [16, 17], its underlying mechanisms have not been elucidated.

Our study revealed that both diclofenac and clodronate exhibited analgesic effects on rats with MIA-induced OA. Although both drugs ameliorated symptomatic behavior during the early inflammatory phase, only clodronate improved symptomatic behavior during the late stages of the disease and effectively suppressed cartilage degeneration. Notably, clodronate treatment significantly reduced the number of subchondral osteoclasts.

In conclusion, our findings clarify the utility of the MIA model as a relevant osteoarthritis model that reflects the acute and chronic pathology of human OA and the validity of the model for the evaluation of potential drug candidates, provided that the pathological phase aligns with the drug's mode of action. This approach holds promise for advancing the development of more effective therapies for KOA.

Materials and methods

Animals

Male Sprague‒Dawley rats (6 weeks old) were obtained from The Jackson Laboratory Japan, Inc. (Kanagawa, Japan). Rats were housed in plastic cages under specific pathogen-free conditions at a room temperature of 23 ± 3°C and an air humidity of 50 ± 20% on a 12-hour light/dark cycle and had free access to food and water. The Institutional Animal Care and Use Committee of Seikagaku Corporation reviewed and approved all the animal studies.

Reagents

MIA was obtained from Sigma‒Aldrich Japan (Tokyo, Japan). Clodronate was obtained from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan). Diclofenac and 0.5 w/v% methyl cellulose 400 solution were obtained from FUJIFILM Wako Pure Chemical Corporation (Osaka, Japan). Saline was obtained from Otsuka Pharmaceutical Factory, Inc. (Tokushima, Japan).

Rat model of MIA-induced OA

Rats were anesthetized with isoflurane, and the left hindlimb was shaved. Subsequently, 1 mg of MIA (50 μl/joint) was intra-articularly injected using a 29G needle. Sham rats received an intra-articular injection of saline (50 μl/joint). The day of MIA administration was set to day 0.

Weight bearing test

Hindlimb weight distributionwas measured weekly using an incapacitance tester (Linton Instrumentation, Norfolk, UK) as previously described [18]. Briefly, rats were placed on an incapacitance tester, and hindlimb weight distribution was measured five times under the following conditions: (1) each hindlimb on each pad and each forelimb on the slope board; (2) facing the front; (3) not leaning on the wall; and (4) with the foot position unchanged during measurement. The average of five measurements was converted to weight distribution (%) using the following formula:

$$\mathrm{Weight}\mathrm{distribution}\left(\%\right)=\left(\frac{\mathrm{weight}\mathrm{on}\mathrm{left}\mathrm{hindlimb}}{\left[\mathrm{weight},,\mathrm{on},,\mathrm{left},,\mathrm{hindlimb},+,\mathrm{weight},,\mathrm{on},,\mathrm{right},,\mathrm{hindlimb}\right\}}\right)\times 100(\%)$$

Before drug treatment, animals with weight distributions exceeding 45% at 6 or 7 days after MIA injection were excluded. The rats were then divided into groups based on body weight and weight distribution.

Drug treatment

The dose of clodronate was determined by reference to the human treatment dose; the maximum dose was 2 mg/joint [19]. Clodronate was dissolved in saline at a concentration of 2 mg/mL and filtered through a 0.22 μm filter. Clodronate solutions at concentrations of 1, 0.2 and 0.02 mg/mL were prepared by further dilution. For intra-articular administration, 50 µl of clodronate at each concentration was administered twice a week: 0.02 mg/mL clodronate (1 µg/joint, 0.58 µg/kg as the human equivalent), 0.2 mg/mL clodronate (10 µg/joint, 5.8 µg/kg as the human equivalent), and 2 mg/mL clodronate (100 µg/joint, 58 µg/kg as the human equivalent). Additionally, sham rats received intra-articular injections of 50 μl of saline. For subcutaneous administration, 1 mg/mL clodronate was administered in a volume of 1 mL/kg twice a week.

Diclofenac, dissolved in 0.5 w/v% methyl cellulose 400 solution at a concentration of 0.6 mg/mL, was administered orally to the rats daily at 5 mL/kg/day according to previous methods [20].

Gross evaluation of cartilage degeneration

The rats were sacrificed by transecting the abdominal aorta under 2% isoflurane anesthesia 27 days after MIA injection. Knee joint samples were collected and stained with India ink. The area of cartilage degeneration (the India ink-stained area) was measured using ImageJ software (National Institutes of Health, MD, USA).

Histology

The rats were sacrificed by transecting the abdominal aorta under 2% isoflurane anesthesia 2, 7, 14, 21, and 28 days after MIA injection. Knee joint samples were collected, fixed in neutral buffered formalin, and embedded in paraffin. Sections were cut and stained with hematoxylin-eosin (HE) or toluidine blue (TB), a method for staining cartilage tissue. Tartrate-resistant acid phosphate (TRAP)-stained sections, a method of staining osteoclasts, were generated using a TRAP Staining Kit (Cosmo bio, Tokyo, Japan). The Mankin score, osteophyte score, and number of TRAP-positive cells in the subchondral bone were evaluated as previously reported [21–23].

Statistics

Statistical analysis was performed using a parametric Dunnett’s test (Holm’s method), Wilcoxon rank sum test, or Tukey’s test. The results are presented as the mean ± SEM. p values of 0.05 or less were considered significant. Statistical analyses were performed using the Statistical Analysis System (SAS Institute, Inc., NC, USA).

Results

Pathological changes in the MIA-injected knee joint

We first assessed the pathological changes in the knee joint, especially the tibia and synovial membrane, following MIA injection to confirm the similarity between human OA and the rat MIA model. In the synovial membrane, severe necrosis and moderate inflammatory cell infiltration were observed as degenerative changes on day 2, which gradually improved but subsequently worsened on day 21 compared to day 14 (Fig. 1 and Table 1). Changes in the synovial membrane on day 21 appeared to correlate with the progression of articular cartilage and bone destruction, as cartilage fragments and bone fragments were observed in the synovial membrane on day 21. Multilayered synovial cells were observed as a regenerative change after day 7. The degree of this change was mild.

Fig. 1.

Pathological changes in the knee joints of MIA-injected rats. Representative histopathology of HE-stained sections of the synovial membrane, cartilage, and subchondral bone at 2, 7, 14, and 21 days after MIA injection. The black triangles indicate multilayered synovial cells. White triangles indicate bone or cartilage fragments. Asterisks indicate replacement by mesenchymal tissue. Scale bars are 100 µm

Table 1.

Pathological changes in the synovial membrane in the knee joints of MIA-injected rats

Tissue	Findings	Degree	Number of cases (out of 3)
			Day 2	Day 7	Day 14	Day 21
Synovial membrane	Necrosis	Mild	0	1	2	2a
		Moderate	0	0	0	1
		Severe	3	0	0	0
	Inflammatory cell infiltration	Mild	0	0	0	2c
		Moderate	3b	0	0	0
	Multilayered synovial cells	Mild	0	3	2	3
		Moderate	0	0	0	0

^abone fragment and/or cartilage fragment, ^bmacrophages-dominated including neutrophils and lymphocytes and edema, ^clymphocyte-dominated and edema

In the loading area of the articular cartilage, degenerative changes were observed beginning on day 2, with erosions becoming apparent by day 14 and increasing in severity over time (Fig. 1 and Table 2). Increased chondrocytes around the loading area were observed beginning on day 7, likely as a compensatory response against the applied physical load (Table 2). These findings were accompanied by endochondral ossification and myelogenesis over time, eventually leading to osteophyte formation.

Table 2.

Pathological changes in the articular cartilage in the knee joints of MIA-injected rats

Tissue	Findings	Degree	Number of cases (out of 3)
			Day 2	Day 7	Day 14	Day 21
Articular cartilage	Degeneration (Loading area)	Mild	0	0	0	0
		Moderate	3	1	0	0
		Severe	0	2	3a	3a
	Erosion (Loading area)	Mild	0	0	1	0
		Severe	0	0	1	3
	Increased chondrocyte (Surrounding loading area)	Mild	0	1	0	0
		Moderate	0	2	3b	3bc

^acrack, ^bbone marrow formation, ^cendochondral ossification

In the subchondral bone, degenerative changes, including thinning, were observed beginning on day 2 and worsened over time (Fig. 1 and Table 3). In the thinning areas of subchondral bone, replacement by mesenchymal tissue occurred and progressed until day 21. On day 21, this change was accompanied by cartilage formation. Thickening of the trabecular bone was observed to compensate for subchondral bone thinning after day 14.

Table 3.

Pathological changes in subchondral bone and bone marrow in the knee joints of MIA-injected rats

Tissue	Findings	Degree	Number of cases (out of 3)
			Day 2	Day 7	Day 14	Day 21
Subchondral bone and bone marrow	Thinning of subchondral bone	Mild	2	1	0	0
		Moderate	0	2	1	0
		Severe	0	0	2	3
	Thickness of subchondral cancellous bone	Mild	0	0	3	2
		Moderate	0	0	0	1
	Replacement by mesenchymal tissue	Mild	2	2	0	0
		Moderate	0	1	2	0
		Severe	0	0	1	3a

^acartilage formation (aggregation or cluster)

These results indicate that the degree of synovial membrane degeneration (necrosis and inflammatory cell infiltration) peaked on day 2, while the changes in articular cartilage (erosion and degeneration) and subchondral bone lesions (thinning) worsened from day 2 to day 21. Thus, the pathology in the knee joint of the MIA model varied greatly depending on the time point of evaluation: synovial inflammation and necrosis in the early phase and cartilage degeneration in the late phase.

Effects of clodronate and diclofenac on MIA-induced knee pain and cartilage degeneration

Having characterized the MIA model, we clarified the effects of clinically effective drugs in this model. Clodronate, a bisphosphonate, and diclofenac, an NSAID, were selected for their clinical efficacy and different mechanisms of action [12–14]. To investigate whether clodronate and diclofenac prevent MIA-induced knee pain, both substances were administered to MIA-injected rats beginning on day 6 or 7 after MIA injection. Consistent with the findings of previous reports [24], compared with sham rats, MIA-injected rats exhibited significant reductions in weight-bearing ratios from day 6 to day 27 after MIA injection (Fig. 2A). Due to the lack of data on the appropriate dose of clodronate, three different doses were administered intra-articularly twice a week. To further confirm the effects of different routes of administration, we also evaluated the efficacy of the drug when administered systemically. Clodronate significantly reversed the decrease in weight-bearing ratios in a dose-dependent manner from day 13 to day 27 after MIA injection (Fig. 2A). Subcutaneously injected clodronate also significantly reversed the decrease in weight-bearing ratios on day 13.

Fig. 2.

Effects of clodronate and diclofenac on MIA-induced knee pain. Hindlimb weight distribution was measured weekly using an incapacitance tester. A MIA-injected rats were injected with clodronate twice a week starting 6 days after MIA injection. B MIA-injected rats were injected with clodronate (100 μg/joint) twice a week or diclofenac (5 mg/kg) daily starting 7 days after MIA injection. The results are shown as the mean ± SEM. Dunnett’s test (Holm’s method) was performed for comparison with the MIA/Vehicle group (^*, p < 0.05; ^**, p < 0.01). CLO-L, clodronate (1 μg/joint); CLO-M, clodronate (10 μg/joint); CLO-H, clodronate (100 μg/joint); CLO-sc, clodronate (1 mg/kg)

Subsequently, diclofenac was administered orally to the rats. Diclofenac reversed the reduction in the weight-bearing ratio from day 10 to day 14; however, it was ineffective from day 21 onward (Fig. 2B). These findings align with previous reports indicating that the late phase of the MIA-induced OA model is insensitive to NSAIDs, resulting in chronic pain [25]. In the present study, clodronate reversed the reduction in the weight-bearing ratio from day 10 to 28.

Site for clodronate treatment

A previous report suggested that intraperitoneally injected clodronate liposomes inhibit the early phase of synovial inflammation by depleting synovial macrophages [16], suggesting a systemic anti-inflammatory effect of clodronate. Indeed, subcutaneously injected clodronate reversed the reduction in the weight-bearing ratio in our study (Fig. 2A). This raises the possibility that some clodronate injected intra-articularly may be transferred systemically and exert analgesic effects. To elucidate the site of action of clodronate, we evaluated the effect of clodronate injection into the contralateral knee joint. Contralaterally injected clodronate did not improve the decreased weight-bearing ratio, whereas ipsilaterally injected clodronate did improve this parameter (Fig. 3). Therefore, these results suggest that the site of action for clodronate is within the knee joint rather than within systemic tissues.

Fig. 3.

The site for clodronate action. Hindlimb weight distribution was measured using an incapacitance tester. MIA-injected rats were injected contralaterally or ipsilaterally with clodronate (100 μg/joint) twice a week starting 7 days after MIA injection. The results are shown as the mean ± SE. Dunnett’s test (Holm’s method) was performed for comparison with the vehicle group (^**, p < 0.01)

Clodronate protects cartilage from degeneration and reduces the number of subchondral osteoclasts

The presence of cartilage and subchondral bone lesions in the late phase of the MIA model suggested that mechanisms other than inflammatory suppression may contribute to the analgesic effects of clodronate. Thus, we evaluated the effect of clodronate on cartilage degeneration after the last measurement of weight distribution, as shown in Fig. 2A. As a result, MIA-induced knee pain was associated with severe cartilage degeneration, and clodronate inhibited cartilage degeneration in a dose-dependent manner (Fig. 4A and B).

Fig. 4.

Chondroprotective effects of clodronate in the rat MIA model. A, B MIA-injected rats were injected with clodronate twice a week starting 6 days after MIA injection. The tibia bones were harvested 27 days after MIA injection. Gross appearance of the representative cartilage surface after India ink staining (A) and the area of the India ink-stained cartilage surface (B). Dunnett’s tests were performed for comparison with the MIA/Vehicle group (^*, p < 0.05). C, D, E MIA-injected rats were injected with clodronate (100 μg/joint) biweekly starting 7 days after MIA injection. The tibia bones were harvested 28 days after MIA injection. A representative TB-stained tibia bone section is shown (C). The results of the Mankin score measurements (D). The results of the osteophyte score measurements (E). The results are shown as the mean ± SEM. The Mankin score and osteophyte score in the MIA/Vehicle and MIA/CLO groups were compared by the Wilcoxon rank sum test (^*, p < 0.05; ^**, p < 0.01). The yellow line indicates the thickness of the osteophyte. Scale bars are 500 µm. CLO-L, clodronate (1 μg/joint); CLO-M, clodronate (10 μg/joint); CLO-H, clodronate (100 μg/joint); CLO-sc, clodronate (1 mg/kg)

We also histologically evaluated the effect of clodronate on cartilage degeneration after the last measurement of weight distribution, as shown in Fig. 2B. The Mankin score based on TB staining of the tibia, a method of staining cartilage tissue, was significantly lower in the clodronate-treated group than in the sham group (Fig. 4C and D), indicating that clodronate prevents MIA-induced cartilage degeneration. Similarly, the osteophyte score was significantly lower in the clodronate-treated group (Fig. 4E). These results suggest that clodronate mitigates MIA-induced knee pain by preventing cartilage degeneration.

Given the antiresorptive effect of clodronate, we hypothesized that clodronate improves the subchondral bone environment by inhibiting subchondral osteoclasts, thereby protecting cartilage, as demonstrated elsewhere [24, 26, 27]. As shown in Figure 5A and B, the number of TRAP-positive multinucleated giant cells was significantly greater in the MIA-treated joints than in the sham joints. Conversely, clodronate significantly reduced the number of TRAP-positive cells in subchondral bone. These results suggest that clodronate has inhibitory effects on subchondral osteoclasts, thereby suppressing cartilage degeneration.

Fig. 5.

Inhibitory effects of clodronate on subchondral osteoclasts. MIA-injected rats were injected with clodronate (100 μg/joint) twice a week starting 7 days after MIA injection. The tibia bones were harvested 28 days after MIA injection. A Representative histopathology of TRAP-stained tibia bone sections. The black triangles indicate TRAP-positive multinucleated giant cells. B The number of TRAP-positive cells in the subchondral bone was measured in histological sections. The results are shown as the mean ± SEM. The number of TRAP-positive cells was compared by Tukey’s test (^**, p < 0.01). Scale bars are 200 µm. CLO, clodronate

Discussion

The pathological changes, such as synovial membrane and knee joint degeneration, in the MIA model are similar to those in human KOA. In the synovial membrane, inflammatory changes such as synovial membrane hyperplasia and inflammatory cell infiltration were observed in both human OA [28] and rat MIA models (Fig. 1). In the cartilage and subchondral bone, cartilage erosion, bone marrow replacement and subchondral bone thickening were observed in both the human OA [29] and rat MIA models (Figure 1). In addition, the degree of synovitis is related to cartilage damage in both human OA and rat MIA models [28]. However, there were some differences between them. Although synovitis in human OA has a patchy distribution [30], synovitis in the MIA model has a uniform distribution. In addition, the synovial necrosis which was observed in the MIA model has not been reported in human OA. The most notable difference is that disease progression in the MIA model is much faster than that in human pathology. Pathology of the knee joint revealed dramatic changes within a few weeks after MIA injection. Thus, it is necessary to understand the pathological changes in the MIA model and evaluate the effect of drug candidates according to their mode of action. Our results suggest that MIA-induced symptomatic behavior in the early and late phases can be attributed to synovial inflammation and joint degeneration, respectively. We then evaluated two different types of compounds to determine their analgesic effects in a rat MIA model.

Since clodronate has a high affinity for macrophages, it has been shown to deplete macrophages via intraperitoneal injection [31]. Although recent reports have shown the clinical effectiveness of clodronate for treating human osteoarthritic knee pain in a randomized clinical trial [13, 14], the underlying mechanisms have yet to be elucidated. This study demonstrated that clodronate attenuated MIA-induced osteoarthritic knee pain in the early and late phases. Additionally, clodronate prevents cartilage degeneration and reduces the number of subchondral osteoclasts. A previous report showed that synovial macrophages induce synovium formation and that synovial macrophage-derived cytokines such as IL-1β and TNF-α stimulate the autocrine production of ADAMTS-4, a cartilage degradation factor, in an MIA model [16]. Furthermore, synovial macrophages enhance the production of NGF, one of the main pain inducers in patients with knee osteoarthritis [16]. Considering that the ability of macrophages to enhance inflammation and synovial inflammation is related to the degree of osteoarthritic knee pain, depleting macrophages improved the loss of weight distribution, which is reasonable. Thus, clodronate may exert analgesic effects by preventing the activation of synovial macrophages in the early phase of MIA.

In the late phase of MIA, clodronate inhibited osteoclasts by inhibiting ATP metabolism [32]. It has been reported that multinucleated giant cells in the synovium and subchondral bone are elevated in OA patients [33]. Another group reported that the number of TRAP-positive cells was significantly greater in inflammatory patients than in noninflammatory OA patients [34]. Subchondral bone remodeling disrupts the weight-load balance to the tibial cartilage and leads to sustained mechanical stress in a specific region, inducing cartilage degeneration [29, 35–37]. Osteoclasts are also involved in tidemark erosion by inducing bone remodeling and angiogenesis [27, 37]. Although there are no nerves in normal articular cartilage, vascular invasion of articular cartilage is accompanied by sensory nerves extending over the calcified cartilage [37, 38], which might contribute to the cause of osteoarthritic knee pain. Angiogenesis exacerbates osteoarthritic knee pain and enhances cartilage degeneration by enhancing the crosstalk of humoral factors between chondrocytes and subchondral bone. Several reports have shown that vascular factors induce chondrocyte hypertrophy, enhancing the expression of type X collagen and MMP-13 [37, 39]. Thus, it is plausible that the effects of clodronate on the late phase of the MIA model are attributed to preventing subchondral osteoclasts. In addition to suppressing osteoclast functions, clodronate has been reported to exert an anabolic effect on chondrocytes in vitro [17]. Because we could not reveal the involvement of the anabolic effect of clodronate on chondrocytes in this study, further study is needed to determine the specific chondroprotective mechanisms of clodronate in the MIA-induced model.

In the present study, diclofenac inhibited symptomatic behavior only in the early phase of the MIA model, which is consistent with previous reports [25]. These findings suggested that diclofenac inhibits only synovial inflammation and does not prevent cartilage or subchondral bone damage. Although one study examined the effect of diclofenac on osteoarthritic chondrocytes [40], the mechanism by which diclofenac protects cartilage was not shown. Therefore, diclofenac is thought to suppress pain via its anti-inflammatory effects. In conclusion, the early phase of the MIA model seems suitable for the evaluation of anti-inflammatory agents such as NSAIDs and corticosteroids.

One limitation of this study is that disease progression in the MIA model was much faster than that in the human pathology model. Therefore, the course of events leading to the onset of the disease may influence the efficacy of the drugs. In addition, it is unclear how close the degree of knee pain in the MIA model is to that in human KOA. Therefore, it is also useful to compare the MIA model with other OA models, among which the anterior cruciate ligament transection model, the medial transection model, the medial meniscectomy model, and destabilization of the medial meniscus are well-established models of surgically induced rat OA [22, 41]. In these studies, severe cartilage degeneration, osteophyte formation, subchondral sclerosis, and bone marrow fibrosis were observed when using surgical OA models. Contrastingly, other studies have reported limited synovial inflammation in such models [42, 43]. The findings of these studies tend to indicate that surgical OA models have a stronger effect on cartilage and bone tissue than the MIA model and a weaker effect on the synovial membrane. Accordingly, the advantage of using surgical OA models lies in their suitability for evaluating chondroprotective drugs, as the pathology and causes in these models are similar to those of human OA. However, it should be noted that surgical OA models may not be suitable for OA pain research due to the limited pain behavior observed in these models [44]. In addition to surgical models, the genetic OA model has also been utilized [22], which, although reproducible, is characterized by a mild pathology and is limited to OA-like lesions. This makes it less suitable for studying advanced stages of OA. Overall, it is thus plausible to select an appropriate OA model according to the aims of a given study. For instance, while surgical models are advantageous for studying cartilage degeneration and chondroprotective therapies, the MIA model may be more appropriate for studying the pain of OA caused by the synovial inflammation and/or the cartilage degeneration.

Although the MIA-induced osteoarthritis model is widely used as a preclinical animal model of OA, its effectiveness has been debated since some drugs did not show any effect on osteoarthritic pain in clinical studies [45, 46], despite strongly inhibiting MIA-induced osteoarthritic knee pain [47, 48]. On the other hand, synovial inflammation and cartilage/subchondral bone changes are commonly reported in humans and in MIA-induced osteoarthritis models. These findings suggest that the MIA-induced osteoarthritis model is suitable for predicting the analgesic effect of drugs that target synovial inflammation and cartilage/subchondral bone by choosing appropriate evaluation points. Additionally, the fact that clodronate and diclofenac, which are clinically applied KOA drugs, had analgesic effects in this model indicates the effectiveness of this model to some extent. Therefore, by understanding pathological characteristics and the mode of action of drug candidates at each time point, our results indicate that the MIA model is a useful preclinical model of osteoarthritis.

Conclusions

Our results showed that the early phase in the MIA model is suitable for evaluating the effects of anti-inflammatory agents such as NSAIDs and corticosteroids. The late phase in the MIA model is suitable for acting on cartilage and subchondral bone.

Abbreviations

KOA

Knee osteoarthritis

MIA

Monoiodoacetate

NSAID

Nonsteroidal anti-inflammatory drug

TKR

Total knee replacement

HE

Hematoxylin-eosin

TB

Toluidine blue

TRAP

Tartrate-resistant acid phosphate

Authors’ contributions

KY: Supervised this research. YS: Designed and performed the animal experiments. KK and YS: Analyze histology. YS and KY: Wrote and revised the manuscript. All the authors have read and approved the manuscript.

Funding

Not applicable.

Data availability

The dataset supporting the conclusions of this article is stored in Seikagaku Corporation, Tokyo, Japan. Further inquiries on the data may be submitted to Yamato Sasaki (yamato.sasaki@seikagaku.co.jp).

Declarations

Ethics approval and consent to participate

All animal studies were reviewed and approved by the Institutional Animal Care and Use Committee of Seikagaku Corporation and were performed under an animal husbandry/management system in an appropriate environment with animal protection and welfare (Approval number: 70–101, 72–15, 73–187, and 75–119). Clinical trial number: not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.