1. Lenti-siRIP1 successfully inhibits compressive mechanical force-induced RIP1 expression in vivo
To explore the role of RIP1 in mechanical force-induced TMJ OA, we used our original TMJ OA rat model6–10, 13,15,16 with the intra-articular injection of Lenti-siRIP1 to mediate RIP1 knockdown (Fig. 1). After loading compressive mechanical force for 4 or 7 days, RIP1 mRNA was significantly increased and the increase was more obvious in the 4 d group (3.69-fold in the 4 d group and 2.2-fold in the 7 d group, P༜0.01). With the injection of Lenti-siRIP1, the mRNA level of RIP1 was significantly decreased compared to the F groups (50.9% in the 4 d group and 39.8% in the 7 d group, P༜0.01). Also, the protein level of RIP1 declined (37.1% in the 4 d group and 34.9% in the 7 d group by Immunohistochemistry, P༜0.01). In both 4 d and 7 d, the decrease was more obvious in proliferative and hypertrophic layers. Considering the low expression of RIP1 in mandibular cartilage, the siRIP1-only group did not reveal a statistical difference compared with Con groups. Thus, the rise of RIP1 in mandibular cartilage under mechanical force could be inhibited by the injection of Lenti-siRIP1.
2. RIP1 inhibition protects mandibular cartilage thinning under compressive mechanical force stimulation
During the progress of compressive mechanical stress-induced TMJ OA, destruction of mandibular cartilage is an obvious pathological change, including the change of cell morphology, cell number, and cartilage thickness. Therefore, we performed H&E staining to detect the change of cartilage to test whether Lenti-siRIP1 had a protective effect on mandibular cartilage.
In the F group, cartilage damage increased time-dependently, showing apparent thinning and chondrocyte loss (Fig. 2). Chondrocytes under mechanical force shrunk in size, arranged irregularly and flattened. The number of chondrocytes also decreased (20.2% in the 4 d group and 29.0% in the 7 d group, P༜0.01). Consistent with the cell number, the thickness of mandibular cartilage became thinning (25.7% in 4 d group and 39.8% in 7 d group, P༜0.01). Interestingly, by measuring the thickness of different layers, we found that in the 4 d group, the thickness of the hypertrophic layer decreased more obviously (32.4% in the 4 d group’s hypertrophic layer, P༜0.01). While in the 7 d group, the proliferative layer showed more thinning (46.8% in the 7 d group’s proliferative layer, P༜0.01).
In the F + siRIP1 group, the thickness of mandibular cartilage recovered (31.8% in the 4 d group and 35.0% in the 7 d group, P༜0.01) and the number of chondrocytes raised (22.8% in the 4 d group and 23.4% in the 7 d group, P༜0.01) compared with the F group. In the 4 d group, hypertrophic layers’ thickness recovered most obviously (56.5% in the 4 d group and 43.3% in the 7 d group, P༜0.01). This demonstrated that Lenti-siRIP1 significantly relieves mandibular cartilage thinning under compressive mechanical force.
3. RIP1 inhibition decreases the mortality of chondrocytes under compressive mechanical force
To further investigate the protective efficiency of Lenti-siRIP1 among chondrocytes, we used TUNEL staining to observe dead cells. Without mechanical force, there was nearly no dead cell in mandibular cartilage. After loading force, dead chondrocytes occurred and showed a time-dependent increasing trend (Fig. 3). The distribution of dead cells also has certain characteristics. In the 4 d group, dead cells were mainly distributed in the transition layer and the hypertrophic layer, while in the 7 d group, dead chondrocytes were scattered throughout the whole layer of mandibular cartilage.
RIP1 inhibition rescued the death of chondrocytes under compressive mechanical force in both 4 d (30.4%, P༜0.01) and 7 d (36.0%, P༜0.01) groups, strongly demonstrating the therapeutic effectiveness. Chondrocytes’ death in proliferative, transition and hypertrophic layers were all alleviated by Lenti-siRIP1, showing that Lenti-virus transfection had effects on the whole layer of mandibular cartilage. Among this, the transition layer recovered most in the 4 d group (37.3%, P༜0.01) while the proliferative layer recovered most in the 7 d group (50.7%, P༜0.01).
4. RIP1 inhibition in mandibular cartilage could not protect subchondral bone under compressive mechanical force
Another pathological change during stress-induced TMJ OA is the destruction and remodeling of subchondral bone. Consistent with our previous study (Fig. 4), at the early stage (4 d), subchondral bone showed decreased percent bone volume (BV/TV), indicating the loss of bone mass. Bone trabecula which supports subchondral bone plate also showed damage, including decreased trabecular number (Tb.N), trabecular thickness (Tb.Th) and increased trabecular separation (Tb.Sp). After loading mechanical force for 7 days, the destruction remained but recovered a little, indicating the initiation of bone remodeling under compressive mechanical stress.
Nonetheless, there was no significant difference between the F + siRIP1 group and the F group on both 4 d and 7 d. The loss of bone mass remained and bone trabecula still showed diffuse microdamage. Meanwhile, in the 7 d group, there was no signal of earlier bone remodeling due to Lenti-siRIP1. Thus, RIP1 inhibition in mandibular cartilage by injecting Lenti-siRIP1 into TMJ could not protect subchondral bone under mechanical force.
5. RIP1 inhibition reduces the mechanical stress-induced expression of inflammatory factors in the cartilage
After confirming the protective role of Lenti-siRIP1 in mandibular cartilage under mechanical force, we further determined the underlying mechanism. As for inflammatory markers, TNF-α and IL-1β both increased in the F group compared with the Con group, indicating the activation of inflammation in the mandibular cartilage during the whole process of force application (Fig. 5). The mRNA level of TNF-α rose rapidly in the 4 d group (5.8-fold, P༜0.01) while the protein level also showed increase in the F group (2.1-fold in the 4 d group and 3.9-fold in the 7 d group, P༜0.01). IL-1β increased obviously on 4 d (1.9-fold in mRNA and 3.4-fold in protein, P༜0.01). While in the 7 d group, the mRNA level of IL-1β had no significant difference compared with the Con group (1.2-fold, P༞0.05) but the protein level was higher (2.5-fold, P༜0.01). TNF-α was mainly located in the proliferation and transition layers while IL-1β was more expressed in the transition and hypertrophic layers.
Lenti-siRIP1 prevented the increase of TNF-α and IL-1β on both 4 d and 7 d, suggesting the inhibition of stress-induced inflammation in mandibular cartilage. In the F + siRIP1 group, TNF-α decreased both at the mRNA (51.7% in the 4 d group and 32.0% in the 7 d group, P༜0.01) and protein (28.0% in the 4 d group and 43.5% in the 7 d group, P༜0.01) levels. The expression of IL-1β also declined both at the mRNA (58.8% in the 4 d group and 65.8% in the 7 d group, P༜0.01) and protein (55.3% in the 4 d group and 44.6% in the 7 d group, P༜0.01) levels. The decrease of inflammatory factors appeared in all layers, suggesting the function of Lenti-siRIP1 could penetrate mandibular cartilage and alleviate inflammation induced by compressive mechanical force.
6. RIP1 inhibition reduces the mechanical stress-induced apoptosis and late-stage necroptosis of chondrocytes
TUNEL staining (Fig. 3) revealed that a large number of chondrocytes died after loading mechanical force. In order to clarify the form of cell death, we further detected the expression of Caspase-8 and RIP3, which represent apoptosis and necroptosis respectively (Fig. 6).
After loading mechanical force for 4 d, the expression level of Caspase-8 in mandibular cartilage showed an obvious upward trend compared with the Conrol group (2.1-fold in mRNA and 6.4-fold in protein, P༜0.01), indicating the activation of apoptosis at the early stage. In the 7 d F group, the level of Caspase-8 was still higher than the Con group (1.4-fold in mRNA and 3.53-fold in protein, P༜0.01), but lower than the 4 d F group, suggesting that apoptosis was reduced but still remained under 7 days’ mechanical force. Different from the early high expression of Caspase-8, RIP3 did not show any increase in the 4 d F group, suggesting that in the early stage of stress-induced TMJ OA, necroptosis did not appear in mandibular cartilage. On 7 d, the expression of RIP3 increased (1.5-fold in mRNA and 2.0-fold in protein, P༜0.01). Combined with the decrease of Caspase-8 on 7 d, it represented that necroptosis was activated in chondrocytes at the later stage of stress-induced TMJ OA. Under mechanical stress, Caspase-8 could be found in all layers, especially in the hypertrophic layer, whereas RIP3 was mainly found in the transition layer.
Compared with the F group, Caspase-8 in the F + siRIP1 group decreased both at the mRNA (49.3% on 4d and 36.2% on 7d, P༜0.01) and protein (62.8% on 4d and 25.4% on 7d, P༜0.01) levels, indicating the inhibition of apoptosis during the whole progress. In the 7 d group, RIP3 decreased obviously (65.5% in mRNA and 25.9% in protein, P༜0.01) in the F + siRIP1 group, demonstrating the effectiveness of Lenti-siRIP1 in preventing the necroptosis. Therefore, as an important upstream regulator, RIP1 inhibition could protect mandibular cartilage in stress-induced TMJ OA by alleviating persisted apoptosis and late-stage necroptosis.