Depletion of primary cilia in articular chondrocytes results in reduced Gli3 repressor to activator ratio, increased Hedgehog signaling, and symptoms of early osteoarthritis

Abstract

Objective: Primary cilia are present in almost every cell type including chondrocytes. Studies have shown that defects in primary cilia result in skeletal dysplasia. The purpose of this study was to understand how loss of primary cilia affects articular cartilage.

Design: Ift88 encodes a protein that is required for intraflagellar transport and formation of primary cilia. In this study, we used Col2aCre;Ift88(fl/fl) transgenic mice in which primary cilia were deleted in chondrocytes. Col2aCre;Ift88(fl/fl) articular cartilage was characterized by histological staining, real time RT-PCR, and microindentation. Hedgehog (Hh) signaling was measured by expression of Ptch1 and Gli1 mRNA. The levels of Gli3 proteins were determined by western blot.

Results: Col2aCre;Ift88(fl/fl) articular cartilage was thicker and had increased cell density, likely due to decreased apoptosis during cartilage remodeling. Mutant articular cartilage also showed increased expression of osteoarthritis (OA) markers including Mmp13, Adamts5, ColX, and Runx2. OA was also evident by reduced stiffness in mutant cartilage as measured by microindentation. Up-regulation of Hh signaling, which has been associated with OA, was present in mutant articular cartilage as measured by expression of Ptch1 and Gli1. Col2aCre;Ift88(fl/fl) cartilage also demonstrated reduced Gli3 repressor to activator ratio.

Conclusion: Our results indicate that primary cilia are required for normal development and maintenance of articular cartilage. It was shown that primary cilia are required for processing full length Gli3 to the truncated repressor form. We propose that OA symptoms in Col2aCre;Ift88(fl/fl) cartilage are due to reduced Hh signal repression by Gli3.

Figure 1. Primary cilium was deleted in Col2aCre;Ift88^fl/fl articular cartilage

Sections from 2 month old control or Col2aCre;Ift88^fl/fl mutant mice were used for immunostaining of primary cilia (A–D). Primary cilia were stained with anti-Arl13b antibody (green), and the basal bodies were stained with anti-γ-tubulin antibody (red). The nuclei were stained with DAPI (blue). Primary cilia were seen as either green dots (arrowheads) or short green lines (arrows) next to the red basal bodies in control articular cartilage. In mutant (B), primary cilia were deleted; however, the basal bodies were still intact (arrows). (C) Close view of primary cilium, which is absent in mutant (D). (Scale bar in A–D: 5μm). (E) Percentage of ciliated cells in control and mutant articular cartilage. Five microscopic fields were randomly chosen along tibial cartilage, and there were least 40 cells in each field. In control, 24.1 ± 7.5% cells were ciliated, whereas in mutant, there were only 2.1± 1.1% ciliated cells, indicating that most primary cilia were deleted in mutant cartilage (t-test, p=0.0002).

Figure 2. Deletion of primary cilia in chondrocytes results in abnormal articular cartilage

Hematoxylin and eosin staining of sections from 2 month old control (A, C) and mutant (B, D) articular cartilage. The proximal tibia of mutant mice showed abnormal shape (B) and thicker cartilage (D, double arrow), compared to control mice (A, C). (E, F) Cell number and cell size were increased in mutant articular cartilage. The cell density (E) of articular cartilage from proximal tibia was compared in 3 controls and 3 mutants. The cell number was counted in three microscopic fields (one field = 1*10⁴ μm²), which spanned the articular surface to the deep zone. Data is presented as individual points from each field from each mouse. An increase in cell density was observed in mutant articular cartilage (**t-test, p=0.005). (F) Cell size was measured using Olympus imaging software (cellSens^® standard). Each data point represents the average cell size in each field counted in (E). The result indicated that cell size was increased (***t-test, p=0.0001) in mutant articular cartilage. Apoptotic cells were measured using the TUNEL assay on sections from control (G, I) and Col2aCre;Ift88^fl/fl (H, J) knee joints. I and J are high magnification images from the boxed area in G and H, respectively. Apoptosis was seen in hypertrophic cells within the calcified zone in control articular cartilage; however, no apoptosis was detected in the corresponding area in mutants. (Scale bar in A and B: 200 μm; in C, D, and G–J: 20 μm).

Figure 3. Markers of osteoarthritis are increased in Col2aCre;Ift88^fl/fl articular cartilage

(A) RNA was isolated from articular cartilage of 2 month old mice and markers of osteoarthritis Mmp13, Adamts5, ColX and Runx2 were determined by real time RT-PCR. The markers were all increased in mutant articular cartilage (n=3 separate mice). The expression of Nckx3 (sodium/potassium/calcium exchanger 3) was used as a control gene and was not altered in mutants. Data are presented by whisker-box plots. Top and bottom whiskers represent the maximum and minimum observations, respectively, and the box represents the middle 50% of observations. Median is shown by a dotted line. Runx2 (B, C) and type X Collagen (D, E) expression were visualized by immunostaining. In control articular cartilage, Runx2 was expressed in chondrocytes in the calcified region (B, bottom right: magnified box area), and type X Collagen was expressed in hypertrophic area (D). In mutant articular cartilage, Runx2 expression was seen in cells in the midzone (C, bottom right: magnified box area), and type X Collagen was seen throughout the cartilage (E). (Scale bar in B–E: 10 μm). (F, G) Mechanical properties were altered in Col2aCre;Ift88^fl/fl articular cartilage compared to controls. Mechanical properties were tested on articular cartilage of tibiae from 2 month old mice. Individual data points from 4 controls and 5 mutants are presented. Instantaneous and equilibrium stiffness were significantly lower in mutant cartilage (F) (t-test, *p=0.032 and **p=0.004, respectively). The thickness was also measured (G), showing an increase in mutants (*t-test, p=0.028).

Figure 4. Expressions of matrix proteins are increased in Col2aCre;Ift88^fl/fl articular cartilage

The expression of matrix proteins was examined by histochemistry and immunostaining. Proteoglycans were examined by Safranine O (A, B) and Toluidine Blue staining (C, D). The staining showed an increase in pericellular and interterritorial proteoglycan content, which may due to increased expression of Aggrecan protein, as shown in (E, F). The pericellular staining of Aggrecan was also increased in mutant, compared to control. (G, H) Collagens were stained with Sirius Red. The staining showed more pericellular staining in the mutant articular cartilage, compared to control. (I, J) Immunostaining of Type II Collagen indicates that the increase in pericelluar collagen content may due to the increased expression of Type II Collagen. Scale bars in A, B equal 200 μm. Scale bars in C–J equal 20 μm.

Figure 5. Hedgehog signaling is up-regulated Col2aCre;Ift88^fl/fl articular cartilage

(A) The expression of downstream targets of Hedgehog signaling, Ptch1 and Gli1, were determined in RNA samples from articular cartilage. The result indicated that hedgehog signaling was up-regulated in mutant articular cartilage relative to controls. Data are presented by whisker-box plots. Top and bottom whiskers represent the maximum and minimum observations, respectively, and the box represents the middle 50% of observations. Median is shown by a dotted line. (B) Gli3 full length and processed repressor forms were determined in postnatal 7 days epiphyseal cartilage from 3 control and 2 mutant mice using western blot. The data showed an accumulation of the full length Gli3 form, resulting in alterations in the ratio of the full length to repressor forms in the absence of primary cilia. Western blot also confirmed reduced expression of Ift88 in the mutant cells. GAPDH was used as a loading control.