Synovial Membrane Is a Major Producer of Extracellular Inorganic Pyrophosphate in Response to Hypoxia

Abstract

Calcium pyrophosphate dehydrate (CPPD) crystals are found in the synovial fluid of patients with articular chondrocalcinosis or sometimes with osteoarthritis. In inflammatory conditions, the synovial membrane (SM) is subjected to transient hypoxia, especially during movement. CPPD formation is supported by an increase in extracellular inorganic pyrophosphate (ePPi) levels, which are mainly controlled by the transporter Ank and ectonucleotide pyrophosphatase/phosphodiesterase 1 (ENPP1). We demonstrated previously that transforming growth factor (TGF)-β1 increased ePPi production by inducing Ank and Enpp1 expression in chondrocytes. As the TGF-β1 level raises in synovial fluid under hypoxic conditions, we investigated whether hypoxia may transform SM as a major source of ePPi production. Synovial fibroblasts and SM explants were exposed to 10 ng/mL of TGF-β1 in normoxic or hypoxic (5% O₂) culture conditions. Ank and Enpp1 expression were assessed by quantitative PCR, Western blot and immunohistochemistry. ePPi was quantified in culture supernatants. RNA silencing was used to define the respective roles of Ank and Enpp1 in TGF-β1-induced ePPi generation. The molecular mechanisms involved in hypoxia were investigated using an Ank promoter reporter plasmid for transactivation studies, as well as gene overexpression and RNA silencing, the respective role of hypoxia-induced factor (HIF)-1 and HIF-2. Our results showed that TGF-β1 increased Ank, Enpp1, and therefore ePPi production in synovial fibroblasts and SM explants. Ank was the major contributor in ePPi production compared to ENPP1. Hypoxia increased ePPi levels on its own and enhanced the stimulating effect of TGF-β1. Hypoxic conditions enhanced Ank promoter transactivation in an HIF-1-dependent/HIF-2-independent fashion. We demonstrated that under hypoxia, SM is an important contributor to ePPi production in the joint through the induction of Enpp1 and Ank. These findings are of interest as a rationale for the beneficial effect of anti-inflammatory drugs on SM in crystal depositions.

Keywords: Ank; ectonucleotide pyrophosphatase/phosphodiesterase 1 (ENPP1); extracellular inorganic pyrophosphate (ePPi); hypoxia; synovial fibroblasts.

Figure 1

TGF-β1 stimulates the expression of Ank and Enpp1 to increase the production of ePPi and Enpp1 activity by synovial fibroblasts in vitro. (A–D) Synovial fibroblasts cultured in monolayer were exposed or not to 10 ng/mL of TGF-β1. (A) Kinetics from 1 to 48 h of Ank and Enpp1 expression (mRNA). Total RNA was extracted from synovial fibroblasts and subjected to RT-qPCR analysis. The abundance of Ank (left panel) and Enpp1 (right panel) mRNA was normalized to that of Rp29 mRNA. Results are expressed as means (±SD) over Rp29 values. (B) Kinetics from 6 to 48 h of ANK and ENPP1 expression (protein). Total proteins were extracted from synovial fibroblasts and subjected to Western blotting using polyclonal anti-Ank and anti-Enpp1 antibodies. Band intensities (left panel) were quantified by densitometry (right panel). The abundance of these proteins was normalized to that of β-actin protein and expressed as induction folds over control value. (C) ePPi level was assayed radiometrically using supernatant of synovial fibroblasts cultured from 6 to 48 h and normalized to the amount of total cell proteins. Data are expressed as mean (±SD) in picomoles per microgram of protein. (D) Enpp1 activity. Proteins were extracted from synovial fibroblasts cultured for 24 h. Enzyme activity was standardized to the total cell proteins. Results are stated as mean (±SD) in micromoles of paranitrophenol per minute per milligram of protein. Statistically significant differences from the control are indicated as * for p < 0.05. (Ank: inorganic pyrophosphate transport regulator; Cont: control condition meaning no TGF-β1 treatment; Enpp1: ectonucleotide pyrophosphatase/phosphodiesterase 1; ePPi: extracellular inorganic pyrophosphate; Rp29: ribosomal protein 29; RT-qPCR: reverse transcription–quantitative polymerase chain reaction; SD: standard deviation; TGF: transforming growth factor).

Figure 2

Hypoxia increases Ank and Enpp1 mRNA levels and the production of ePPi ex vivo. (A–C) SM explants were transfected or not with scramble siRNA or siRNAs directed against Ank or Enpp1 for 12 h, before being challenged with 10 ng/mL of TGF-β1 for 48 h either in normoxia (21% O₂) or hypoxia (5% O₂). Total RNA was extracted from explants and then subjected to RT-qPCR. The abundance of Ank (A) and Enpp1 (B) mRNA was normalized to that of Rp29 mRNA. Results are expressed as means (±SD) over Rp29 values. (C) ePPi levels were assessed in culture supernatants of SM explants. They were normalized to the amount of total cell proteins and are expressed as means (±SD) in picomoles per microgram of protein. Statistically significant differences are indicated as: * for p < 0.05 from the control in normoxia, # for p < 0.05 from control in hypoxia, £ for p < 0.05 from TGF-β1-stimulated cells in normoxia, and $ for p < 0.05 from TGF-β1-stimulated cells in hypoxia. (Ank: inorganic pyrophosphate transport regulator; Cont: control condition meaning no TGF-β1 treatment and no transfection; Enpp1: ectonucleotide pyrophosphatase/phosphodiesterase 1; ePPi: extracellular inorganic pyrophosphate; Rp29: ribosomal protein 29; RT-qPCR: reverse transcription–quantitative polymerase chain reaction; Scr: scramble RNA meaning non-silencing RNA; SD: standard deviation; Si: means siRNA or small interfering RNA; SM: synovial membrane; TGF: transforming growth factor).

Figure 3

Hypoxia enhances Ank and Enpp1 protein expression to increase ePPi production and Enpp1 activity ex vivo. (A–E) SM explants were cultured either in normoxia (21% O₂) or hypoxia (5% O₂) for 48 h. (A,B) Samples were subjected to immunohistochemistry using polyclonal anti-Ank (A) and anti-Enpp1 (B) antibodies. The exhibited pictures are representative of at least three independent experiments. The tissue presence of proteins is characterized by a brown staining and the counterstaining of cell nuclei appear as blue. Magnification is ×10. (C) The percentage of positively stained cells in SM was quantified by two independent observers and results are expressed as mean (±SD). (D) ePPi levels were assessed in culture supernatants of SM explants and normalized to the amount of total cell proteins. Results are expressed as means (±SD) in picomoles per microgram of protein. (E) Enpp1 activity. Proteins were extracted from SM. Enzyme activity was normalized to the amount of total cell proteins. Results are expressed as mean (±SD) in micromoles of paranitrophenol per minute per milligram of protein. Statistically significant differences are indicated as * for p < 0.05 from the control in normoxia. (Ank: inorganic pyrophosphate transport regulator; Enpp1: ectonucleotide pyrophosphatase/phosphodiesterase 1; ePPi: extracellular inorganic pyrophosphate; SD: standard deviation; SM: synovial membrane).

Figure 4

Hypoxia increases the transactivation of Ank promoter in vitro. (A,B) Synovial fibroblasts were transfected or not for 24 h then cultured either in normoxia (21% O₂) or hypoxia (various oxygen percentage). (A) Synovial fibroblasts were transfected or not with pGL3-Ank-Luc and pCMV-Renilla reporter. (B) Control of hypoxia efficiency using synovial fibroblasts transfected or not with pGL3-HRE-Luc and pCMV-Renilla reporter. (C,D) Synovial fibroblasts were transfected or not then cultured in hypoxia (5% O₂). (C) Synovial fibroblasts were transfected or not with pcDNA3.1, or pcDNA3.1 construct to overexpress HIF-1 or HIF-2, and with pGL3-Ank-Luc and pCMV-Renilla reporter. (D) Synovial fibroblasts were transfected or not with pcDNA3.1, or pcDNA3.1 construct to overexpress HIF-1 or HIF-2, and with pGL3-HRE-Luc and pCMV-Renilla reporter before being cultured in hypoxia. Results are presented as mean luciferase activity ratio of Firefly/Renilla (±S.D.). Statistically significant differences are indicated as * for p < 0.05 from the control in normoxia (A,B) and as * for p < 0.05 from the control in hypoxia (5% O₂) (C,D). (Ank: inorganic pyrophosphate transport regulator; Cont: control condition meaning transfection with pGL3-Ank-Luc (A,C) or pGL3-HRE-Luc (B,D) and pCMV-Renilla reporter; Hif: hypoxia-induced factor; HRE: Hif-response element; Luc: Firefly luciferase; SD: standard deviation).

Figure 5

HIF-1, but not HIF-2, increases Ank promoter transactivation in vitro. (A,D) Synovial fibroblasts were transfected or not with siRNA for 12 h, then with plasmids for 24 h, before being challenged with 10 ng/mL of TGF-β1 for 48 h either in normoxia (21% O₂) or hypoxia (5% O₂). (A,C,D) Synovial fibroblasts were transfected or not with scramble siRNA, or siRNAs directed against Hif-1 or Hif-2, then with pGL3-Ank-Luc and pCMV-Renilla reporter. (B) Control of hypoxia efficiency using synovial fibroblasts transfected or not with scramble siRNA, or siRNAs directed against Hif-1 or Hif-2, then with pGL3-HRE-Luc and pCMV-Renilla reporter. (A) Results for Ank promoter transactivation are presented in histograms as mean luciferase activity ratio of Firefly/Renilla (±S.D.). (B) Results for HRE transactivation are presented in histograms as mean luciferase activity ratio of Firefly/Renilla (±S.D.). Statistically significant differences are indicated as * for p < 0.05 from the control in normoxia and as # for p < 0.05 from control in hypoxia (5% O₂). (Ank: inorganic pyrophosphate transport regulator; Cont: control condition meaning no siRNA transfection; ePPi: extracellular inorganic pyrophosphate; Hif: hypoxia-induced factor; HRE: Hif-response element; Luc: Firefly luciferase; Rp29: ribosomal protein 29; RT-qPCR: reverse transcription–quantitative polymerase chain reaction; Scr: scramble RNA meaning non-silencing RNA; SD: standard deviation; Si: means siRNA or small interfering RNA).