Aberrant Calreticulin Expression in Articular Cartilage of Dio2 Deficient Mice

Abstract

Objective: To identify intrinsic differences in cartilage gene expression profiles between wild-type- and Dio2-/--mice, as a mechanism to investigate factors that contribute to prolonged healthy tissue homeostasis.

Methods: Previously generated microarray-data (Illumina MouseWG-6 v2) of knee cartilage of wild-type and Dio2 -/- -mice were re-analyzed to identify differential expressed genes independent of mechanical loading conditions by forced treadmill-running. RT-qPCR and western blot analyses of overexpression and knockdown of Calr in mouse chondro-progenitor cells (ATDC5) were applied to assess the direct effect of differential Calr expression on cartilage deposition.

Results: Differential expression analyses of articular cartilage of Dio2-/- (N = 9) and wild-type-mice (N = 11) while applying a cutoff threshold (P < 0.05 (FDR) and FC > |1,5|) resulted in 1 probe located in Calreticulin (Calr) that was found significantly downregulated in Dio2-/- mice (FC = -1.731; P = 0.044). Furthermore, overexpression of Calr during early chondrogenesis in ATDC5 cells leads to decreased proteoglycan deposition and corresponding lower Aggrecan expression, whereas knocking down Calr expression does not lead to histological differences of matrix composition.

Conclusion: We here demonstrate that the beneficial homeostatic state of articular cartilage in Dio2-/- mice is accompanied with significant lower expression of Calr. Functional analyses further showed that upregulation of Calr expression could act as an initiator of cartilage destruction. The consistent association between Calr and Dio2 expression suggests that enhanced expression of these genes facilitate detrimental effects on cartilage integrity.

Fig 1. Calreticulin expression stratified for treadmill-running and knockout.

Representation of the Calr expression values of all individual samples plotted by genotype-group (Wild-type and Dio2^-/-) and treadmill-running-group (No Run and Run). Showing the significant reduction of Calr expression upon forced treadmill-running in wild-type-mice (P = 0,005) and the absence of change in the knockout-mice. P-values depicted are derived from Student T-Test.

Fig 2. Overexpressing Calr resulted in significantly lower Alcian Blue (AB) staining intensities.

Staining intensities of Alcian Blue staining measured with a photospectrometer (620 nm). Values are displayed as the average±SEM, relative to the control sample. Differences were analyzed with Student T-Test ((*) P < 0.05).

Fig 3. Expression analyses in Calr overexpressing ATDC5 cells.

RT-qPCR expression of Calr, Dio2 and Acan in ATDC5 cells transfected with either a Calr-containing vector for overexpression (Calr+) or 3 unique 27mer siRNA duplexes, specific for Calr, for knockdown (Calr-). Values of the RT-qPCR are displayed as the average±SEM, normalized for GAPDH expression and relative to the control samples (dashed line). Differences were analyzed with Student T-Test ((*) P < 0.05).

Fig 4. Calreticulin expression in DIO2 overexpressing hBMSCs.

qPCR expression of Calr in hBMSCs transduced with either a control virus-vector (eGFP) or a DIO2-eGFP vector. Values of the RT-qPCR are displayed as the average±SEM, normalized for GAPDH expression and relative to the control sample at every timepoint. Differences were analyzed with Student T-Test ((*) P < 0.05).