PPP1R3A inhibits osteogenesis and negatively regulates intracellular calcium levels in calcific tendinopathy

Abstract

Calcific tendinopathy (CT) is defined by the progressive accumulation of calcium crystals in tendonic regions that results in severe pain in patients. The etiology of CT is not fully elucidated. In this study, we elucidate the role of PPP1R3A in CT. A significant decrease in PPP1R3A expression was observed in CT patient tissues, which was further confirmed in tissues from a CT-induced rat model. Overexpression of PPP1R3A ex vivo reduced the expression of osteo/chondrogenic markers OCN and Sox9, improved tendon tissue architecture, and reduced intracellular Ca2+ levels. Overexpression of SERCA2 and knockdown of Piezo1 decreased expression of osteo/chondrogenic markers and intracellular calcium in PPP1R3A-knockdown tendon cells. Lastly, PPP1R3A expression was regulated at the posttranscriptional level by binding of HuR. Collectively, the present study indicates that PPP1R3A plays an important role in regulating calcium homeostasis in tendon cells via Piezo1/SERCA2, rendering it a promising target for therapeutic interventions of CT.

Keywords: Cell biology; Molecular biology; Orthopedics.

Graphical abstract

Figure 1

Expression of PPP1R3A in calcific tendinopathy (A) Heatmap of PPP1R3A expression in tissues from healthy controls (n = 23) and tendinopathy patients (n = 23) from microarray dataset (GSE26051). Scale bar indicates high expression in red and low expression in green. (B) Individual values from data shown in A plotted as log₂ values. p value calculated by Mann-Whitney U test. (C) Representative immunohistochemistry images of PPP1R3A expression and H&E staining in human normal and calcific tendons. Scale bars, 50 μm. (D) Quantification of images shown in C for total area positive for PPP1R3A expression in calcific and normal tendons (n = 10). Values are means ± SD. ∗p < 0.05, versus normal tendons, calculated by two-tailed paired t test. (E) PPP1R3A mRNA levels were analyzed by qRT-PCR with GAPDH as control, n = 10. p value calculated by Mann-Whitney U test. (F) Western blotting analysis was performed to detect the levels of PPP1R3A protein in human calcific (C) and normal tendons (N). The band intensity was normalized to GAPDH. The intensity of the blots was represented as mean ± standard deviation (SD) of ten pair tissues samples. Results are shown as the fold change of the normal tendon group on the bar graph. ∗∗p < 0.01, versus normal tendons, calculated by two-tailed paired t test.

Figure 2

PPP1R3A overexpression inhibits ectopic calcification in a collagenase-induced rat tendon calcification model (A–D) Analysis of PPP1R3A expression in rat Achilles tendon tissues from control (Sham) or tendon calcification model (Model) treated with PPP1R3A overexpressing lentivirus or empty control vector. (A) qRT-PCR analysis of PPP1R3A mRNA expression in rat Achilles tendons. (B) Western blot analysis indicating PPP1R3A protein expression in rat Achilles tendons from control (Sham) or pathological mice group (Model). The band intensity was normalized to GAPDH. (C) Representative immunohistochemistry images of PPP1R3A expression in rat Achilles tendons. Scale bars, 50 μm. Arrows indicate the PPP1R3A + positive cells. (D) Quantification of images shown in C for total area positive for PPP1R3A expression. Values are means ± SD (n = 6). p value calculated was determined by two-tailed unpaired Student’s t test. ∗∗, ##p < 0.01. ∗ vs. the Sham-Empty vector group, # vs. the Model-Empty vector group. (E) Representative images from Safranin O staining of Achilles tendon sections. Scale bar: 50 μm. (F) Representative micrographs of OCN and SOX9 expression in Achilles tendon tissues visualized by immunofluorescence. Scale bar: 50 μm. (G and H) The fluorescence intensity of OCN (G) and SOX9 (H) staining in Achilles tendon tissues. Results are shown as mean ± SD (n = 6). p value calculated was determined by two-tailed unpaired Student’s t test. ∗, #p < 0.05, ∗∗, ##p < 0.01. ∗ vs. the Sham-Empty vector group, # vs. the Model-Empty vector group.

Figure 3

PPP1R3A overexpression promotes tendon regeneration in a collagenase-induced rat tendon calcification model (A) Representative images from H&E staining in tissues from control (Sham) and calcification-induced (Model) rats. (B) Histological score of the repaired tendon (n = 6). p value calculated was determined by two-tailed unpaired Student’s t test. ∗, #p < 0.05, ∗∗p < 0.01. ∗ vs. the Sham-Empty vector group, # vs. the Model-Empty vector group. (C) Masson staining of the Achilles tendon in four groups. (D) Representative images of immunohistochemistry staining of type Ⅰ collagen (COLⅠ) and TNMD in the Achilles tendon tissue in each group. Scale bar: 50 μm. Arrows indicate the COLⅠ+ or TNMD + positive cells. (E and F) Quantification of images shown in D for total area positive for COLⅠ (E) and TNMD (F) expression. Results are shown as mean ± SD (n = 6). p value calculated was determined by two-tailed unpaired Student’s t test. ∗, #p < 0.05. ∗ vs. the Sham-Empty vector group, # vs. the Model-Empty vector group.

Figure 4

PPP1R3A regulates osteogenesis and intracellular Ca²⁺ level in tendon cells (A) qRT-PCR analysis of PPP1R3A in tendon cells treated with PPP1R3A overexpression lentivirus or siRNA. (B) Representative immunoblots and quantification of PPP1R3A protein expression in tendon cells. (C) Cell proliferation assessed by CCK-8 assay for the different treatment groups. (D) Representative images from alkaline phosphatase staining of tendon cells and ALP quantification on Day 7. (E) Representative images from Alcain blue staining of tenocytes and ARS quantification on Day 14. (F) Representative immunoblots and quantification of OCN and Sox9 in tendon cells cultured 7 days in the osteogenic medium. The densitometric values of the proteins were normalized to that of GAPDH. (G) Visualization of the distribution of intracellular Ca²⁺ in tendon cells by Fluo-4AM Ca²⁺ fluorescence probe. Blue: cell nucleus stain, Green: intracellular Ca²⁺. (H) Quantification of intracellular Ca²⁺ levels by flow cytometry. Shown are representative histograms and quantification under different conditions. Results are shown as mean ± SD of three independent repeats. ∗p < 0.05, ∗∗p < 0.01.p value calculated was determined by two-tailed unpaired Student’s t test.

Figure 5

SERCA2/Piezo1 is involved in the regulation of osteogenesis and Ca²⁺ level in a PPP1R3A-dependent manner in tendon cells (A) Representative immunoblots and densitometric quantification of Piezo1 and SERCA2 expression in rat Achilles tendons. The band intensity was normalized to GAPDH. Values are means ± SD (n = 6). ∗, #p < 0.05, ∗∗p < 0.01. ∗ vs. the Sham-Empty vector group, # vs. the Model-Empty vector group. p value calculated was determined by two-tailed unpaired Student’s t test. (B) Representative micrographs of SERCA2 and Piezo1 expression in rat Achilles tendons visualized by immunofluorescence (×400; n = 6). Scale bars, 50 μm. (C) Representative immunoblots and densitometric quantification of Piezo1 and SERCA2 expression in tendon cells treated with PPP1R3A overexpression lentivirus or siRNA. n = 3 replicates for each group. p value calculated was determined by two-tailed unpaired Student’s t test. ∗, #p < 0.05, ∗∗, ##p < 0.01. ∗ vs. Empty vector, # vs. siControl. (D and E) Representative immunoblots from co-immunoprecipitation experiments with antibodies against endogenous Piezo1, SERCA2 or control IgG in tendon cells treated with PPP1R3A overexpression lentivirus or siRNA (E) Representative images from alkaline phosphatase staining of tendon cells and ALP quantification on Day 7. n = 3 replicates for each group. p value calculated was determined by two-tailed unpaired Student’s t test. ∗∗p < 0.01, vs. Empty vector, #p < 0.05, vs. siControl. (F) Representative images from alcain blue staining of tendon cells and ARS quantification on Day 14. n = 3 replicates for each group. p value calculated was determined by two-tailed unpaired Student’s t test. ∗∗p < 0.01, vs. Empty vector, #p < 0.05, vs. siControl. (G) Representative immunoblots and corresponding densitometric quantification of OCN and Sox9 expression in tendon cells cultured for 3 days in the osteogenic medium. Protein expression was normalized to GAPDH. n = 3 replicates for each group. p value calculated was determined by two-tailed unpaired Student’s t test. ∗∗, ##p < 0.01, ∗ vs. Empty vector, # vs. siControl. (H) Quantification of intracellular Ca²⁺ levels by flow cytometry. Shown are representative histograms and quantification under different conditions. Results are shown as mean ± SD of at least three independent repeats. n = 3 replicates for each group. p value calculated was determined by two-tailed unpaired Student’s t test. ∗∗, ##p < 0.01. ∗ vs. siControl, # vs. siPPP1R3A.

Figure 6

HuR interacts with PPP1R3A mRNA and increases its stability and translation in tendon cells (A) Representative immunoblots and densitometric quantification of HuR and PPP1R3A expression in tendon cells treated with control or HuR siRNA. n = 3 replicates for each group. p value calculated was determined by two-tailed unpaired Student’s t test. ∗∗p < 0.01. (B) mRNA levels of HuR and PPP1R3A in tendon cells measured by qRT-PCR. n = 3 replicates for each group. p value calculated was determined by two-tailed unpaired Student’s t test. ∗∗p < 0.01. (C) RIP analysis to evaluate the association of endogenous HuR with endogenous PPP1R3A mRNA. Quantification of PPP1R3A by RT-qPCR analysis in immunoprecipitated fractions using either anti-HuR antibody or control IgG in tendon cells. Values are the means ± SD from triplicate samples. n = 3 replicates for each group. p value calculated was determined by two-tailed unpaired Student’s t test. ∗∗p < 0.01 compared with IgG IP. (D and E) Schematic representation of the different regions in PPP1R3A mRNAs used for RNA pulldown assays (E) Immunoblots probed for HuR and control TUBA from different biotinylated pulldown fractions: PPP1R3A 5′-UTR, CR (coding region), and 3′-UTR. p27 5′UTR and CR served as positive (P) and negative (N) controls, respectively. α-Tubulin (TUBA) as the internal control. (F and G) Schematic representation of the different pGL3-derived reporter vectors generated bearing the PPP1R3A mRNA fragments for luciferase assay (G) Luciferase activity in tendon cells. Cells were transfected with siControl or siHuR for 24 h. Cells were subsequently transfected with each of the pGL3-derived reporters together with a pRL-CMV vector and cultured for an additional 48 h. Luciferase activity was measured by a luminometer. n = 3 replicates for each group. p value calculated was determined by two-tailed unpaired Student’s t test. ∗∗p < 0.01. (H) Half-lives of PPP1R3A and GAPDH mRNAs were examined at different time points post actinomycin D treatment in siControl or siHuR transfected tendon cells. p value calculated was determined by two-tailed unpaired Student’s t test. ∗p < 0.05 compared with siControl. Results are shown as mean ± SD of three independent repeats.