Inhibition of PHD3 by salidroside promotes neovascularization through cell–cell communications mediated by muscle-secreted angiogenic factors

Therapeutic angiogenesis has been considered as a potential strategy for treating peripheral artery diseases including hind-limb ischemia (HLI); however, no effective drug-based treatment is currently available. Here we showed that intramuscular administration of salidroside, an active compound of Chinese herb Rhodiola, could robustly enhance blood perfusion recovery by promoting neovascularization in HLI mice. We revealed that salidroside promoted skeletal muscle cell migration and paracrine function through inhibiting the transcriptional level of prolyl-hydroxylase domain 3 (PHD3) without affecting PHD1 and PHD2. Paracrine signals from salidroside-treated skeletal muscle cells enhanced endothelial and smooth muscle cells migration, while inhibition of FGF2/FGF2R and PDGF-BB/PDGFR-β pathways abolished this effect, as well as neovascularization in HLI mice. Furthermore, we elucidated that salidroside inhibition on PHD3 might occur through estrogen receptor alpha (ERα). Together, our findings highlights the potential application of salidroside as a novel pharmalogical inhibitor of ERα/PHD3 axis for therapeutic angiogenesis in HLI diseases.


Supplementary Figure 1. Schematic diagram for the administration schedule of salidroside.
(a) Salidroside was administered intramuscularly into the gastrocnemius muscle of the ischemic hind limb one day after femoral artery excision and every three days thereafter. (b) Salidroside was administered intramuscularly into the gastrocnemius muscle of the ischemic hind limb one day after femoral artery excision and every three days thereafter; while FGF2R inhibitor (PD173074) or PDGFR inhibitor (CP868596) was injected intramuscularly into the gastrocnemius muscle of the ischemic hind limb right after femoral artery excision and every two days thereafter. (c) Salidroside was administered intramuscularly into the gastrocnemius muscle of the ischemic hind limb one day after femoral artery excision and every three days thereafter; while PHD3 overexpression plasmid were injected intramuscularly into the gastrocnemius muscle of the ischemic hind limb once a week starting from the day after femoral artery excision.

Supplementary Figure 2. Prolonged salidroside administration does not result in obvious morphological changes.
Salidroside was administered intramuscularly into the gastrocnemius muscle of the ischemic hind limb of HLI mice every three days for two months starting from the day after femoral artery excision. Representative Hematoxylin & Eosin stainings of liver, kidney, spleen and heart were shown. Scale bars: 100 µm.

Supplementary Figure 3. Salidroside promotes skeletal muscle cells proliferation and migration. (a)
The proliferation of C2C12 cells treated with salidroside or PBS and cultured in hypoxia was analyzed using Ki67 staining. The nuclei were stained with DAPI. The representative microscopic images (Scale bars: 100 µm) were shown in left side; ratio of Ki67 positive cells to DAPI positive cells was shown in right side (**p < 0.01). (b) The percentage of apoptotic cells in C2C12 cells treated with salidroside or PBS and cultured in hypoxia were analyzed by using Annexin V-FITC/PI staining and FACS analysis. (c,d) the mobility of C2C12 cells treated with salidroside and cultured in hypoxia or treated with PBS and cultured in hypoxia or normoxia were analyzed using scratch assay: (c) representative images (scale bars: 100 µm) and (d) percentage of wound closure (**p < 0.01). N: normoxia, H: hypoxia, SA: salidroside. The quantitative results (a and d) are the means ± s.e.m of three independent experiments. Figure 4. Schematic diagram of the preparation and HPLC-MS analysis of conditioned medium collected from salidroside-treated C2C12 cells. (a) Salidroside (final concentration 100 µg/ml) or PBS was added into the culture medium of C2C12 cells, and twenty-four hours later, the medium was discarded, then the cells were washed twice with PBS. After the addition of fresh medium, cells were further cultured for 24 h in normoxia or hypoxia, and then the medium was collected and filtered using a 0.22-µm filter. (b-f) HPLC-MS analysis results: (b) DMEM supplemented with 10% FBS; (c) standard reference of salidroside; (d) culture medium collected from C2C12 cells cultured with DMEM supplemented with 10% FBS and salidroside for 24 h in normoxia; (e) culture medium collected from C2C12 cells cultured with DMEM supplemented with 10% FBS and salidroside for 24 h in normoxia, washed with PBS twice, added with fresh DMEM supplemented with 10% FBS and cultured in hypoxia for 0 h; (f) culture medium collected from C2C12 cells cultured with DMEM supplemented with 10% FBS and salidroside for 24 h in normoxia, washed with PBS twice, added with fresh DMEM supplemented with 10% FBS and cultured in hypoxia for 24 h (i.e., CM-SA). Salidroside's peak, as indicated by red arrow, could be detected at 323.1 U. shFGF2-2, shPDGFB-1 or shPDGFB-2 and cultured in hypoxic condition were analyzed by using quantitative RT-PCR (**p < 0.01, data were shown as relative to the expression level in the cells transfected with shCon). (b, c) The mobility of MOVAS cells cultured with CM-shFGF2/SA, CM-shPDGFB/SA or CM-shFGF2-shPDGFB/SA were analyzed by using transwell chamber assay: (b) representative images and (c) quantification of migrated cells (NS: not significant, **p < 0.01). (d,e) The expression levels of FGF2R and PDGFR-β in HUVECs (d) and MOVAS (e) cells transfected with shCon, shFGF2R-1, shFGF2R-2, shPDGFR-β-1 or shPDGFR-β-2 and cultured in hypoxic condition were analyzed by using quantitative RT-PCR (**p < 0.01, data were shown as relative to the expression level in the cells transfected with shCon). (f, g) The mobility of FGF2R-or PDGFR-β-silenced MOVAS cells cultured with CM-SA were analyzed by using transwell chamber assay: (f) representative images and ( and protein (c) expression levels of PHD1 and PHD2 in C2C12 cells transfected with shCon, shPHD3-1 or shPHD3-2 and cultured in hypoxic condition were analyzed by using quantitative RT-PCR (NS: not significant, data were shown as relative to the expression in the cells transfected with shCon) and western blotting: representative images (left) and the quantitative analysis (right) (NS: not significant), respectively. (d,e) The mRNA (d) and protein (e) expression levels of angiogenic factors in PHD3-silenced C2C12 cells were analyzed by using quantitative RT-PCR (*p < 0.05, **p < 0.01, data were shown as relative to cells transfected with shCon) and western blotting: representative images (left) and quantification (right) (*p < 0.05, **p < 0.01), respectively. (f,g) The mobility of C2C12 cells transfected with shCon, shPHD3-1 or shPHD3-2 in hypoxic condition were analyzed by using scratch assay: (f) representative images (left) and percentage of wound closure (right) (**p < 0.01); and (g) transwell chamber assay: representative images (left) and quantifiation of migrated cells (right) (**p < 0.01). (h) Morphological changes of F-actin were examined by using phalloidin staining: (left) representative images; (right) fractal dimension analysis (*p < 0.05). Scale bars: 100 µm (scratch and transwell chamber assays) or 25 µm (phalloidin staining). N: normoxia, H: hypoxia, SA: salidroside. The quantitative results (a, b, c, d, e, f, g and h) are the means ± s.e.m of three independent experiments.

RNA extraction and quantitative RT-PCR analysis
Total RNA from gastrocnemius muscles or cells was extracted with Trizol (Invitrogen) according to the manufacturer's instruction. Total RNA (1 μg) was reverse-transcribed into cDNA using the PrimeScript RT Reagent Kit with gDNA Eraser (Takara Bio), and quantitative RT-PCR was performed to assess the mRNA expression levels with SYBR Premix Ex Taq (Takara Bio). The sequences of the primers used for quantitative RT-PCR were shown in Supplementary Table 1. β-actin was used to normalize sample amplifications. The results were shown as relative to the expression level in the corresponding controls, which are assumed as 1.  Table 2, and immunoblotting with anti-β-actin antibody was conducted to ensure equal protein loading. The signal was measured by the SuperSignal West Femto Maximum Sensitivity Substrate detection system (Thermo Scientific, Waltham, MA). The quantitative analysis was performed by using Quantity One (Thermo Scientific), and β-actin was used as an internal control. The results were shown as relative to the expression level in the corresponding controls, which are assumed as 1.

Immunohistochemical analysis
Frozen gastrocnemius muscles were sectioned at 10 μm thickness using a cryostat and subjected to immunohistochemistry. Briefly, for immunohistochemistry against PECAM-1 and α-SMA, the frozen muscle tissue sections were incubated with a primary antibody against murine PECAM-1 for 1 h. The specimens were then incubated with a monoclonal antibody against murine α-SMA conjugated with Cy3

Scratch assay
Cells were seeded in 6-well plates at a density of 3 × 10 5 cells per well, treated with salidroside (final concentration: 100 µg/ml) or PBS for 24 h under normoxic condition, and washed with PBS. After the medium was changed to culture medium without salidroside, cells were treated with cyclohexamine (purity ≥ 95%, final concentration: 10 mg/ml, Cayman Chemicals, USA), then a scratch wound was created with a micropipette tip, and the cells were incubated under normoxia or hypoxia for the indicated times. For experiments with conditioned media, cells were cultured with conditioned medium under hypoxia after the scratch wound was created for the indicated times. For experiments using FGF2R and PDGFR inhibitors, cells were cultured with conditioned media and PD173074, CP868596 or both of them under hypoxia after the scratch wound was created for the indicated times.

Transwell chamber assay
Cells were treated with PBS or salidroside (final concentration: 100 µg/ml) and incubated under normoxic condition for 24 h. The cells were reseeded (5 × 10 3 cells per chamber) in the upper chambers of a transwell plate (Corning, NY, USA) and cultured under hypoxia for 24 h. Normal culture medium was placed in the lower chambers. Cells migrated to lower chamber were stained with crystal violet (Beyotime) for imaging or DAPI (Beyotime) for quantification by averaging the number of the cells in the lower chamber. Images were taken with Olympus IX7I (Japan). For experiments with conditioned media, cells were seeded (5 × 10 3 cells per chamber) in the upper chambers of a transwell plate (Corning) and cultured under hypoxia for 24 h. Conditioned media were placed in the lower chambers. For experiments using FGF2R and PDGFR inhibitors, conditioned media and PD173074, CP868596 or both of them were placed in the lower chambers.

Enzyme-linked immunosorbent assay (ELISA)
The amounts of FGF2 and PDGF-BB in the culture medium were measured using a commercially available FGF2 and PDGF-BB sandwich ELISA kit (CK-E91925M and CK-E9420M, respectively, Yuanye, Shanghai, China) according to the manufacturer's protocol.

Hematoxylin and Eosin staining
The gastrocnemius muscles were fixed with 4% paraformaldehyde for overnight prior to being embedded in paraffin and sectioned at 5 µm thickness using a cryostat. Sections were dewaxed using xylene and rehydrated before being stained with Hematoxylin and Eosin (Beyotime).