The ciliary protein intraflagellar transport 88 is required for the maturation, homeostasis and mechanoadaptation of articular cartilage

Purpose: The integration of external cues, such as mechanics, with intrinsic cell signalling programmes, such as hedgehog (Hh) signalling, is crucial for the development, maturation and homeostasis of articular cartilage. Activation of Hh signalling in adulthood and pathophysiological mechanics, have both been associated with the development of murine and human OA. But, how chondrocytes might transduce and integrate these cues remains unknown. A microtubule-based organelle, the primary cilium, most noted for its crucial role in Hh signalling, is assembled by chondrocytes and possesses a devoted trafficking machinery, IntraFlagellar Transport or IFT. In vitro studies indicate chondrocyte helps tune the anabolic matrix response to compression and the response to Hh ligand. In vivo, the primary cilium has been proposed to be a platform for the integration of mechanics and Hh signalling in musculoskeletal tissues. While constitutive and peri-natal disruption of ciliary proteins, Hh signalling and altered mechanics, all drastically alter joint development in vivo, the influence of IFT in adult cartilage homeostasis remains unknown.

Methods: IFT88 was targeted using a cartilage-specific, inducible mouse line (ACANCreERT2;Ift88fl/fl : cKO hereafter). Cre activity was validated by qPCR, RNA scope and a ROSA26tdtomato reporter line. Ift88fl/fl mice, also receiving I.P injections of tamoxifen, were used as controls. Tibial articular cartilage was assessed 2, 14 or 26 weeks-post tamoxifen, at 8, 10, 22 and 34 weeks of age respectively, using histomorphometric analyses, including measurements of articular cartilage thickness, relative calcification, subchondral bone thickness, and OARSI score by means of immunohistochemistry (IHC). The surgical DMM model, which destabilises the joint, was performed at 10 weeks of age. To explore the role of physiological mechanics, mice were allowed two weeks of voluntary wheel exercise immediately following tamoxifen administration at 8 weeks of age. qPCR was performed on micro dissected articular cartilage at 10 weeks of age in control and cKO. RNAscope was performed on cryosections of articular cartilage from 10 week old mice control and cKO. Means ± S.D are quoted throughout, Mann-Whitney U-test or Fisher’s test were used for statistical comparisons.

Results: In our previous OARSI abstract of 2020 we described the phenotype arising in IFT88 cKO mice. Here we outline this in further detail and with an exploration of underlying mechanisms. Tamoxifen treatment of cKO mice resulted in a 50% reduction of Ift88 mRNA in articular cartilage (p=0.02, n=6 control, 14 cKO). Ift88 (cKO) mice had thinner medial articular cartilage (MAC), compared with controls, at all 5 time-points (Fig 1.A) In control mice, MAC thickness increased from 102.57μm (95% CI [94.30, 119.80]) at 8 weeks of age to 108.68 +/- (95% CI [101.32, 116.42]) at 10 weeks of age. Tamoxifen treatment, at 8 weeks of age, inhibited this increase in cKO mice (MC thickness at 10 weeks was 96.20 μm (95% CI [90.04, 102.36]), p=0.02, compared with 10 week ctrl, n=7). By 22 weeks of age mean MAC thickness in cKO was 90.16μm (95% CI [87.11, 93.22]) compared with 111.60μm (95% CI [104.34, 118.79]) in control animals (p=0.0002, n= 7 and 10 respectively). By 34 weeks MAC had continued to thin to 84.55μm (95% CI [75.43, 93.67]) in cKO, but this was now associated with surface damage and osteophyte formation. In the most extreme case, MAC was completely lost (Fig.1B). In contrast, lateral plateau thickness and OARSI score were unaffected. Calcified cartilage (below the tidemark) progressively increases on both plateaus, between 6 and 22 weeks of age and at all time-points thinning was attributable to the relative loss of calcified cartilage implying a failure of calcification. IHC analyses revealed no striking differences in collagen X expression, NITEGE neoepitope. There were no measurable increases in subchondral bone thickness or changes in osteoclastic activity in cKO mice. 12 weeks post DMM, OARSI scores were statistically significantly higher in cKO (29.83 +/- 7.69) than control (22.08 +/- 9.30, p< 0.05, n= 15 both groups). Two weeks of voluntary wheel exercise rescued cartilage atrophy in cKO mice (p< 0.0001), whilst no change was observed in controls. RNA isolated from microdissected articular cartilage of 10 week old control and cKO mice, two weeks post tamoxifen, revealed a statistically significant correlation between Ift88 and Tcf7l2 expression after Bonferroni correction (p=0.026). Ctgf, Gli2 and Enpp1 were also positively correlated with Ift88 expression before correction (p=0.002, p=0.0037, and p=0.009 respectively). RNA scope analysis of AC found a statistically significant (p< 0.0001, n=4 in both groups) decrease in Ift88 positive cells in cKO (27.78%) compared with controls (45.18%), whilst also showing an increase in Gli1 positive cells in cKO (50.42%) compared with controls (23.63%) (p< 0.0001, n=4 in both groups).

Conclusions: Progressive thickening and calcification in the mouse medial compartment illustrates the continued mechanoadaptation of adolescent and adult articular cartilage. Depletion of the ciliary gene Ift88 inhibits medial articular cartilage thickening, leading to atrophy, which then predisposes the joint to spontaneous OA. The lateral compartment is relatively unaffected. We propose this may be due, in part, to disruption of mechanotransduction and downstream anabolic remodelling in medial cartilage. Deletion of Ift88 impairs the progressive calcification of articular cartilage, in both compartments, which may be due to disruption of Hh signalling, which is also mechanosensitive. Ift88 expression was correlated with Tcf7l2, previously shown to interact and influence Hh signalling pathways in cartilage. On-going experiments are aiming to dissect the relative roles of IFT, mechanics and Hh in the context of adult cartilage. We conclude that Ift88 is crucial to post-natal articular cartilage homeostasis and chondroprotective against OA.