Promotion of Joint Degeneration and Chondrocyte Metabolic Dysfunction by Excessive Growth Hormone in Mice

Abstract

Objective: Many patients with acromegaly, a hormonal disorder with excessive growth hormone (GH) production, report pain in joints. We undertook this study to characterize the joint pathology of mice with overexpression of bovine GH (bGH) or a GH receptor antagonist (GHa) and to investigate the effect of GH on regulation of chondrocyte cellular metabolism.

Methods: Knee joints from mice overexpressing bGH or GHa and wild-type (WT) control mice were examined using histology and micro-computed tomography for osteoarthritic (OA) pathologies. Additionally, cartilage from bGH mice was used for metabolomics analysis. Mouse primary chondrocytes from bGH and WT mice, with or without pegvisomant treatment, were used for quantitative polymerase chain reaction and Seahorse respirometry analyses.

Results: Both male and female bGH mice at ~13 months of age had increased knee joint degeneration, which was characterized by loss of cartilage structure, expansion of hypertrophic chondrocytes, synovitis, and subchondral plate thinning. The joint pathologies were also demonstrated by significantly higher Osteoarthritis Research Society International and Mankin scores in bGH mice compared to WT control mice. Metabolomics analysis revealed changes in a wide range of metabolic pathways in bGH mice, including beta-alanine metabolism, tryptophan metabolism, lysine degradation, and ascorbate and aldarate metabolism. Also, bGH chondrocytes up-regulated fatty acid oxidation and increased expression of Col10a. Joints of GHa mice were remarkably protected from developing age-associated joint degeneration, with smooth articular joint surface.

Conclusion: This study showed that an excessive amount of GH promotes joint degeneration in mice, which was associated with chondrocyte metabolic dysfunction and hypertrophic changes, whereas antagonizing GH action through a GHa protects mice from OA development.

Figure 1.. bGH mice develop increased cartilage degeneration, synovitis, and subchondral plate thinning at 13-months-old.

(A) Safranin O staining of knee joints of WT (upper panel) and bGH (bottom panel) mice. Low-magnification images (4x) of whole joints are on the left, high-magnification (20x) images of each joint compartments (lateral femur, medial femur, lateral tibia, medial tibia) are on the right. White arrowhead shows cartilage defects or loss, black arrowhead shows cartilage expansion and hypertrophy, * shows chondrocyte hypertrophic changes. Scale bar, 500 μm for 4x images, 100 μm for 20x images. (B-C) Quantification of OA severity in male (B) and female (C) WT and bGH mice using average (AVG) and maximal (Max) OARSI Scores, as well as average (AVG) Mankin Score that includes osteophyte, cartilage damage, tidemark duplication, proteoglycan Saf O staining, and hypertrophic chondrocytes. (D) Representative images of synovium from WT and bGH mice. Dashed black lines indicate areas of synovium hyperplasia. (E) Semi-quantitative synovitis scores in male and female WT and bGH mice. (F) Representative histological images of WT and bGH mice displaying thinning of the subchondral plate. Dashed white lines indicate areas of subchondral plate. (G) Representative images of μCT of tibial subchondral bone of WT and bGH mice. (H) Subchondral plate thickness (left graph) and percentage of bone volume to total volume (BV/TV) (right graph) of tibial subchondral bone of WT and bGH mice. *p<0.05, **p<0.01, ns=not significant. n=5 for male WT, n=4 for male bGH, n=5 for female WT, n=5 for female bGH mice.

Figure 2.. Chondrocytes in the articular cartilage tissue express functional GH receptor (GHR).

(A) Immunohistochemical staining for GHR in lateral tibial cartilage of WT and GHR knockout (GHR^−/−) mice. Scale bar, 100 μm. n=3. (B) Immunohistochemical staining for GHR in human OA and normal articular cartilage. Scale bar, 100 μm. n=3 for Normal, n=4 for OA. (C) Ratio of GHR positively stained (GHR+) chondrocytes to total chondrocytes in Normal or OA human cartilage tissue. (D) Mouse primary chondrocytes were treated with 0, or 0.5 or 2.5 nM bGH with or without the GH receptor antagonist, Peg, and expression of phos-STAT5 (~95 kDa), total STAT5 (Tot-STAT5) (~95 kDa), and beta-actin (ACTB) (45 kDa) was determined, n=5. (E) Densitometry analysis of ratio of phos-STAT5/total-STAT5 using blotting results from 5 independent experiments. (F) Immunofluorescent staining for phos-STAT5 in the cartilage of WT and bGH mice. (G) Ratio of phos-STAT5 positively stained (p-STAT5+) chondrocytes to total chondrocytes in WT and bGH mice cartilage. n=4, *p<0.05, scale bar, 100 μm.

Figure 3.. Chondrocytes from bGH mice express higher levels of chondrocyte hypertrophy associated genes.

Mouse primary chondrocytes isolated from the articular cartilage of WT and bGH mice treated with or without Pegvisomant (200 nM) for two days were then used to detect gene expression of Col2a1 (A), aggrecan (Acan) (B), comp (C), Col10a1 (D), Mmp13 (E), and Adamts5 (F). *p<0.05, **p<0.01, ***p<0.001. n=3. (G) Safranin O staining and immunohistochemical staining for collagen type X (COLX) in the articular cartilage of WT and bGH mice. (H & I) Quantification of chondrocyte sizes in both non-calcified (n=75 for male WT, n=60 for male bGH, n=60 for female WT, n=75 for female bGH) and calcified (n=57 for male WT, n=48 for male bGH, n=40 for female WT, n=58 for female bGH) articular cartilage tissue in male (H) and female (I) mice. *p<0.05, ****p<0.0001, ns=not significant.

Figure 4:. Metabolite analysis of bGH versus WT cartilage.

Panel (A): PCA and PLS-DA plots showing extent of sample clustering per condition. Panel (B): Heatmap for metabolites with significantly different (t-test FDR < 0.1) concentrations in bGH versus WT samples. Also depicted is each feature’s log2 fold-change (LFC; with red and blue shading analogous with upregulation and downregulation in bGH samples, respectively) and PLS-DA derived variable in importance projection (VIP) score. Panel (C): Pathway analysis of important metabolites (t-test FDR < 0.1 and VIP score > 1). Circle size is proportional to number of important metabolite ‘hits’ in given pathway, while strength of color shading is representative of −Log10(P) for pathway enrichment (with darker shading analogous having stronger significance). Significantly enriched pathways are consequently labelled using in-plot. Panel (D): Interaction network(s) for important metabolites (t-test FDR < 0.1 and VIP score > 1), as defined using the Search Tool for Interactions of Chemicals (STITCH) database. Color shading is representative of LFC, red being upregulated and blue downregulated, in bGH samples (versus WT controls, respectively, with connecting edge thickness corresponding to strength of interaction. Panel (E): Box plots depicting normalized concentration for the two most connected metabolites in panel (D).

Figure 5.. Chondrocytes from bGH mice are more efficient in oxidizing long-chain fatty acid palmitate using a Seahorse Respirometry assay.

(A) Chondrocytes isolated from WT and bGH mice were provided BSA-conjugated palmitate as substrate. The cells were then subject to ‘Mito Stress Assay (acute injection)’ using Seahorse XFe24 analyzer. The first injection from Port A is either medium or the fatty acid oxidation inhibitor, Etomoxir (4 μM), while the following three injections from Port B, C, D are: 2 μM Oligomycin, 1 μM FCCP, and 0.5 μM Rotenone/Antimycin-A. Oxygen consumption rate (OCR, pmol/min/million cells) were measured during the assay. Data were then normalized to cell number. (B-I) Different aspects of mitochondrial functions including non-mitochondrial OCR (B), Basal Respiration_before extomoxir (C), Basal Respiration_after extomoxir (D), Maximal Respiration (E), Proton Leak (F), ATP Production linked Respiration (G), Spare Respiratory Capacity (H), and Coupling Efficiency (I) are calculated. *p<0.05, ****p<0.0001, n=4.

Figure 6.. Mice expressing a GH receptor antagonist (GHa) are protected from aging associated OA development.

(A) Representative images of joints from ~23 months-old male WT and GHa mice. Upper panel, low-magnification images (4x). Lower panel, high-magnification images (20x) of the lateral tibia. Scale bar, 500 μm for 4x images, 100 μm for 20x images. Average (AVG) OARSI score (B), maximal (Max) OARSI score (C), average (AVG) Mankin score (D), cartilage structure (E), tidemark (F), Safranin O staining (G), oteophyte (H), hypertrophic chondrocyte (I) scores of WT and GHa mice joints. White arrowhead shows cartilage defects or loss, # shows loss of proteoglycan staining, *p<0.05, ns=not significant. n=4 for WT, n=5 for GHa.