Cartilage-specific Sirt6 deficiency represses IGF-1 and enhances osteoarthritis severity in mice

Abstract

Objectives: Prior studies noted that chondrocyte SIRT6 activity is repressed in older chondrocytes rendering cells susceptible to catabolic signalling events implicated in osteoarthritis (OA). This study aimed to define the effect of Sirt6 deficiency on the development of post-traumatic and age-associated OA in mice.

Methods: Male cartilage-specific Sirt6-deficient mice and Sirt6 intact controls underwent destabilisation of the medial meniscus (DMM) or sham surgery at 16 weeks of age and OA severity was analysed at 6 and 10 weeks postsurgery. Age-associated OA was assessed in mice aged 12 and 18 months of age. OA severity was analysed by micro-CT, histomorphometry and scoring of articular cartilage structure, toluidine blue staining and osteophyte formation. SIRT6-regulated pathways were analysed in human chondrocytes by RNA-sequencing, qRT-PCR and immunoblotting.

Results: Sirt6-deficient mice displayed enhanced DMM-induced OA severity and accelerated age-associated OA when compared with controls, characterised by increased cartilage damage, osteophyte formation and subchondral bone sclerosis. In chondrocytes, RNA-sequencing revealed that SIRT6 depletion significantly repressed cartilage extracellular matrix (eg, COL2A1) and anabolic growth factor (eg, insulin-like growth factor-1 (IGF-1)) gene expression. Gain-of-function and loss-of-function studies in chondrocytes demonstrated that SIRT6 depletion attenuated, whereas adenoviral overexpression or MDL-800-induced SIRT6 activation promoted IGF-1 signalling by increasing Akt^ser473 phosphorylation.

Conclusions: SIRT6 deficiency increases post-traumatic and age-associated OA severity in vivo. SIRT6 profoundly regulated the pro-anabolic and pro-survival IGF-1/Akt signalling pathway and suggests that preserving the SIRT6/IGF-1/Akt axis may be necessary to protect cartilage from injury-associated or age-associated OA. Targeted therapies aimed at increasing SIRT6 function could represent a novel strategy to slow or stop OA.

Keywords: arthritis, experimental; biological therapy; chondrocytes; osteoarthritis; osteoarthritis, knee.

Figure 1

The effect of Sirt6 deficiency on OA severity after DMM surgery. Sirt6 intact and Sirt6 deficient (Sirt6 cKO) mice received sham or DMM surgery at 16 weeks of age and OA severity was analysed at 6 and 10 weeks post-DMM surgery by histology and micro-CT. (A,B) Representative images of H&E stained midcoronal sections showing the MTP and MFC from Sirt6 intact and Sirt6 cKO mice at 6 and 10 weeks post-DMM surgery. (C,D) Summed (MTP, MFC, LTP, LFC) ACS and toluidine blue scores, summed (MTP, LTP) osteophyte scores and synovial hyperplasia scores (medial side) at 6 and 10 weeks post-DMM surgery. Black arrows indicate areas of complete articular cartilage loss. (E–F) Representative three-dimensional micro-CT reconstructions of knee joints from representative sham, Sirt6 intact and Sirt6 cKO mice at 6 and 10 weeks post-DMM surgery. Upper panels show transverse images of the tibial plateau and lower panels show images of the whole joint. (G–H) Micro-CT analysis of subchondral bone changes (BV/TV, Tb.Th, Tb.Sp, SCBP.Th) on the medial tibial plateaus of Sirt6 intact and Sirt6 cKO mice (DMM and sham groups) at 6 and 10 weeks post-DMM surgery. (I) Micro-CT analysis of max osteophyte area and osteophyte volume on the MTP of Sirt6 intact and Sirt6 cKO mice (DMM groups only) at 6 weeks and (J) 10 weeks post-DMM surgery. Individual data points are presented with mean±SD. Significant differences between groups were detected by Mann-Whitney U test comparing sham and DMM groups for each genotype (A–H) or t-test (I,J). Exact p values are presented. ACS, articular cartilage structure; cKO, Sirt6-deficient; DMM, destabilisation of the medial meniscus; LFC, lateral femoral condyle; LTP, lateral tibial plateau; MFC, medial femoral condyle; MTP, medial tibial plateau; OA, osteoarthritis; SCBP, subchondral bone plate.

Figure 2

The effect of Sirt6 deficiency on OA severity during ageing. Sirt6 intact and Sirt6 deficient (cKO) mice were aged 12 and 18 months and spontaneous OA was analysed by histology and micro-CT. (A,B) Representative images of H&E stained midcoronal sections showing the medial tibial plateau and medial femoral condyle from Sirt6 intact and Sirt6 cKO mice at 12 and 18 months of age. (C,D) Summed (MTP, MFC, LTP, LFC) ACS and toluidine blue scores, summed (MTP, LTP) osteophyte scores and synovial hyperplasia scores (medial side) at 12 and 18 months of age. Black arrows indicate areas of complete articular cartilage loss. (E) Micro-CT analysis of subchondral bone changes (BV/TV, Tb.Th, Tb.Sp, SCBP.Th) on the medial tibial plateaus of Sirt6 intact and Sirt6 cKO mice at 12 months and (F) 18 months of age. Individual data points are presented with mean±SD. Significant differences between groups were detected by unpaired t-test. Exact p values are presented. ACS, articular cartilage structure; cKO, Sirt6-deficient; DMM, destabilisation of the medial meniscus; LFC, lateral femoral condyle; LTP, lateral tibial plateau; MFC, medial femoral condyle; MTP, medial tibial plateau; OA, osteoarthritis; SCBP, subchondral bone plate.

Figure 3

RNA-sequencing analysis of Sirt6 depleted human chondrocytes. To assess SIRT6-mediated transcriptional regulation, RNA-sequencing was conducted on primary human chondrocytes nucleofected (72 hours) with small interfering RNA (siRNA) to Sirt6 (siSirt6, Sirt6 knockdown) and compared with cells nucleofected with a scrambled siRNA as a control (siCtrl). (A) Heatmap showing significant differentially expressed genes identified in our dataset. Selected upregulated and downregulated genes are highlighted when comparing Sirt6 depleted cells with controls. (B) Gene Ontology (GO) enrichment showing downregulated processes in Sirt6 depleted cells, when compared with control, are presented. (C) RT-PCR was conducted on human chondrocytes to validate expression of selected genes found in the RNA-sequencing dataset. Significant differences between groups were detected by paired t-test. Exact p values are presented. (D) Upstream regulator analysis using the ingenuity pathway analysis tool showing the effect of downregulated insulin-like growth factor-1 (IGF-1) expression on its targets.

Figure 4

SIRT6 regulates insulin-like growth factor-1 (IGF-1) signalling in chondrocytes. (A) Femoral caps from Sirt6 intact and Sirt6-deficient mice were treated with 4-hydroxytamoxifen to activate Cre-mediated recombination ex vivo. Protein levels of IGFBP2, phospho-Akt (Ser473) and SIRT6 were assessed by immunoblotting (n=3). (B) Immunohistochemistry to detect IGF-1 levels was performed on Sirt6 intact and Sirt6-deficient mouse joint tissue sections derived from sham control mice from our destabilisation of the medial meniscus (DMM) study (n=5), and percentage IGF-1 positively stained cells were quantified. (C) Primary non-osteoarthritis (non-OA) older human chondrocytes were transduced with an adenoviral vector to overexpress SIRT6 or an empty vector control for 24 hours prior to immunoblotting for phospho-Akt (Ser473, Thr308), phospho-PRAS40 and SIRT6 (n=5). (D) Primary non-OA human chondrocytes were treated with MDL-800 (12.5 µM) for 0–24 hours prior to immunoblotting for phospho-Akt (Ser473), phospho-PRAS40 and SIRT6 (n=4). Presented immunoblots are representative and protein bands were normalised to total protein or housekeeping proteins as indicated. Individual data points are presented with mean±SD. Significant differences were detected by t-test (A–C) or two-way analysis of variance (D). *P<0.05.