Valvular Interstitial Cells Suppress Calcification of Valvular Endothelial Cells
Background
Calcific aortic valve disease (CAVD) is the most common heart valve disease in the Western world. We previously proposed that valvular endothelial cells (VECs) replenish injured adult valve leaflets via endothelial-to-mesenchymal transformation (EndMT); however, whether EndMT contributes to valvular calcification is unknown. We hypothesized that aortic VECs undergo osteogenic differentiation via an EndMT process that can be inhibited by valvular interstitial cells (VICs).
Approach and Results
VEC clones underwent TGF-β1-mediated EndMT, shown by significantly increased mRNA expression of the EndMT markers α-SMA (5.3±1.2), MMP-2 (13.5±0.6) and Slug (12±2.1) (p<0.05), (compared to unstimulated controls). To study the effects of VIC on VEC EndMT, clonal populations of VICs were derived from the same valve leaflets, placed in co-culture with VECs, and grown in control/TGF-β1 supplemented media. In the presence of VICs, EndMT was inhibited, shown by decreased mRNA expression of α-SMA (0.1±0.5), MMP-2 (0.1±0.1), and Slug (0.2±0.2) (p<0.05). When cultured in osteogenic media, VECs demonstrated osteogenic changes confirmed by increase in mRNA expression of osteocalcin (8.6±1.3), osteopontin (3.7±0.3), and Run×2 (5.5±1.5). The VIC presence inhibited VEC osteogenesis, demonstrated by decreased expression of osteocalcin (0.4±0.1) and osteopontin (0.2±0.1) (p<0.05). Time course analysis suggested that EndMT precedes osteogenesis, shown by an initial increase of α-SMA and MMP-2 (day 7), followed by an increase of osteopontin and osteocalcin (day 14).
Conclusions
The data indicate that EndMT may precede VEC osteogenesis. This study shows that VICs inhibit VEC EndMT and osteogenesis, indicating the importance of VEC–VIC interactions in valve homeostasis.
Introduction
Calcific aortic valve disease (CAVD) and subsequent aortic valve stenosis is the most common heart valve disease in the Western world.1, 2 CAVD is currently considered an actively regulated and progressive disease, characterized by a cascade of cellular changes that initially cause fibrotic thickening, followed by extensive calcification of the aortic valve leaflets. This in turn leads to significant aortic valve stenosis and eventual left ventricular outflow obstruction,3, 4 for which surgical replacement remains the only viable treatment option.
Heart valves contain a heterogeneous population of valvular endothelial cells (VECs) and valvular interstitial cells (VICs), which maintain valve homeostasis and structural leaflet integrity. VICs, the most abundant cell type in the heart valve, play a key role in CAVD progression. Various VIC phenotypes have been identified in diseased human heart valves,5 including quiescent fibroblast-like VICs (qVICs), which upon pathological cues can differentiate into activated myofibroblast-like VICs (aVICs); and osteoblast-like VICs (oVICs), which are responsible for the active deposition of calcium in CAVD.6-8 Additionally, numerous studies have demonstrated the ability of VICs to undergo osteogenic differentiation.9-11 Relatively little is known about the role of VECs in CAVD. VECs cover the surface of the heart valve to form an endothelial monolayer, and are unique in that they can undergo endothelial-to-mesenchymal transformation (EndMT) — a critical process in developmental valvulogenesis.12-15 During development, EndMT occurs in the endocardial cushions, where a subset of endothelial cells detach from the endothelium, transiently enhance the contractile protein α-SMA, and migrate into the interstitium of the embryonic valve to become VICs.13,16,17 EndMT also occurs in adult valves, where cells co-expressing endothelial markers and α-SMA have been detected along the valve endothelium and in subendothelial locations.18 These observations prompted the hypothesis that low or basal levels of EndMT contribute to the replenishment of VICs as part of physiologic valve remodeling that is required throughout postnatal life19. We recently demonstrated that EndMT plays a role in the adaptive pathologic remodeling of mitral valve leaflets in an ovine model of functional mitral regurgitation.20,21 In addition, EndMT has shown to be potentiated by exposure to cyclic mechanical strain.22 We furthermore showed that mitral VECs are able to differentiate in vitro into mesenchymal lineages, including osteogenic cells.14
VECs have been indicated as key regulators in early CAVD via recruitment of immune cells,23 dysregulation of protective nitric oxide (NO) signaling,24,25 and phenotypic plasticity through the expression of procalcific proteins.14 Collectively, these studies indicate a potential role for dysregulated VECs in valvular disease, but the role of EndMT in aortic valve leaflet calcification in CAVD is yet unknown. Furthermore, factors that regulate EndMT-associated VEC differentiation into osteogenic cells remain unclear. Based on our previous investigations, we hypothesized that VEC–VIC interaction serves as a native barrier to prevent excessive VEC EndMT and subsequent osteogenic differentiation.
Methods
Aortic valve cell isolation
Ovine tissues from animals 8-10 months of age, weighing 20–25 kg, were obtained under approved NIH guidelines for animal experimentation as performed at Children’s Hospital Boston. Valve leaflets were incubated in endothelial basal medium (EBM-2) (CC- 3156, Cambrex Bio Science, Walkersville, MD) with 5% fetal bovine serum (FBS) (Hyclone, Logan, UT), 1% GPS (Invitrogen, Carlsbad, California), 2 mmol/L l-glutamine, and 100 μg/ml gentamicin sulfate for 1–4 hours. They were then minced into 2-mm pieces and incubated with 0.2% collagenase A (Roche Diagnostics, Indianapolis, IN) in EBM-2 for 5 minutes at 37°C, and diluted wit h Hanks’ balanced salt solution containing 5% FBS, 1.26 mmol/L CaCl2, 0.8 mmol/L MgSO4, and 1% GPS (wash buffer). The supernatant was sedimented at 200×g, resuspended in VEC medium (EBM-2 medium, 10% heat-inactivated FBS, 1% GPS, and 2 ng/mL basic fibroblast growth factor [bFGF]; Roche Diagnostics, Indianapolis, IN) and plated. The following day, primary cultures were washed to remove unattached cells. Primary cultures were trypsinized, resuspended in growth medium at 3.3 cells/ml, and 100 μl were plated in each well of a 96-well plate, at a concentration of approximately one cell to every third well; visual inspection was performed to assess that single colonies appeared in a subset of wells. When the colonies covered two-thirds of the well, cells were split into 24-well dishes. Based on morphology, these clones were initially visually identified as either endothelial or interstitial. These observations were confirmed with phenotypic characterization and designated as either endothelial (VEC)26, 27 or interstitial (VIC-K3, VIC-K5, VIC-K6), and expanded on 1% gelatin-coated dishes in VEC media. Cells were passaged 1:3 or 1:4 every 6 to 14 days, and experiments were performed using VEC and VIC clones between passages 8 and 14. VECs were grown in VEC media, and VICs were cultured in VIC media (DMEM, 10% FBS, 1% GPS).
Human calcified aortic valves
Aortic valve leaflets (n=5) were harvested from patients undergoing aortic valve replacement for aortic valve stenosis and a healthy valve was obtained from autopsy, performed using criteria established by the declaration of Helsinki. Tissue samples were frozen in optimal cutting temperature compound (OCT, Sakura Finetech, Torrance, CA) and 7 μm serial sections were cut and stained. Tissue was collected according to Brigham and Women’s Hospital IRB Protocols.
Mouse model of CAVD
Male apolipoprotein E-deficient mice (apoE mice; 10 weeks old) were purchased from Jackson Laboratory (Bar Harbor, ME). High-fat diet (21% fat and 0.21% cholesterol) was obtained from Research Diets (D12079B, New Brunswick, NJ). Mice were fed with an atherogenic diet for a total of 22 weeks. Mice were euthanized for tissue collection and histopathology at 32 weeks. The animal procedures were performed conform the NIH guidelines and approved by the Brigham and Women’s Hospital Animal Care and Use Committee. This model is characterized by the presence of functional and histopathological changes found in human CAVD.33
Endothelial-to-Mesenchymal Transformation (EndMT)
EndMT was induced as described previously.27 The assay was also performed in the presence of VICs in an indirect co-culture system, using a Transwell system (12-mm Transwells with 0.4-μm pore polycarbonate membrane inserts; Corning Life Sciences, Acton, MA) (Supplemental Figure I). Briefly, VECs were grown alone or in co-culture with VICs in VEC media, or VEC media supplemented with 2 ng/mL TGF-β1 (R&D Systems, Minneapolis, MN), for up to 14 days. Media was changed every 2–3 days.
VEC and VIC osteogenic potential
The osteogenic potential of VEC and VIC clones was tested both separately and in co-culture for up to 21 days via treatment with osteogenic media (OM) (DMEM with 10% FBS, 1% GPS with 10 nmol/L beta-glycerolphosphate, 50 μmol/L ascorbic acid, 10 μmol/L dexamethasone). DMEM with 10% FBS, 1% GPS was used as control medium (NM). Media was changed every 2–3 days.
VEC-VIC indirect co-culture
Co-culture experiments were performed using Transwell plates (Corning, Tewksbury, MA) (Supplemental Figure I). Cells were plated either in the Transwell inserts or on the bottom of 6-well plates at a density of 1 × 10 cells/cm and allowed to adhere overnight. After 24 hours, the two cell types were combined to begin the co-culture by placing the inserts into the corresponding well of the 6 well plate. For the conditioned media experiments, cells were treated with media that had been conditioned by the specific cell type for 24 hours. Conditioned media was sterile-filtered and added in a 1:1 ratio with fresh media to treated cells. Media was changed every 2–3 days.
Histological detection of VEC and VIC mineralization
Upon completion of experiments, cells were washed with PBS and fixed using 4% paraformaldehyde for 15 minutes. Cells were subsequently washed with PBS. Mineralization was analyzed by staining with 0.02 mg/L of Alizarin Red S (Sigma-Aldrich). The area of positive Alizarin Red S staining was normalized to cell number. To detect the expression of alkaline phosphatase (ALP), nitro-blue tetrazolium/indolylphosphate (NBT/BCIP) staining was performed. Before staining, the cells were washed with PBS, 0.5 ml of NBT/BCIP was added, and the samples were then incubated at 37°C in a humidified chamber cont aining 5% CO2 for 30 minutes. Samples were then washed with PBS and fixed with 4% PFA, after which they underwent a counter stain with 0.1% eosin for 5 minutes. Images were taken with an Eclipse 80i microscope (Nikon) and processed with Elements 3.20 software (Nikon).
Alkaline phosphatase activity and calcium measurements
A colorimetric kit was used to measure ALP activity (BioVision, Milpitas, CA) according to the manufacturer’s instructions. 16 μl of supernatant from 12-well plates was added to 64 μl of ALP assay buffer and 50 μl of PnPP solution was added and incubated for 1 hour at room temperature. The absorbance was read at a wavelength of 405 nm. Values were normalized to the standard curve. Calcium content was quantified using a colorimetric kit (BioVision) according to the manufacturer’s protocol. Briefly, 50 μL of sample was added to 60 μL of calcium assay buffer, after which 90 μL of chromogenic reagent was added and incubated for 10 minutes at room temperature in the dark. Absorbance was read at a wavelength of 575 nm with a plate reader. Values were normalized to the standard curve.
Immunofluorescence and western blotting
Immunofluorescence staining was performed on methanol-fixed cells and human and mouse aortic valve leaflets using anti-human VE-cadherin (Santa Cruz) or CD31 (Cell Signaling), anti-human α-SMA (clone 1A4, Sigma-Aldrich), anti-vimentin antibody (Abcam, Cambridge, MA) and anti-osteocalcin (Abcam). Secondary antibodies conjugated with AlexaFluor 488 (Invitrogen) were used. Images were taken with an Eclipse 80i microscope (Nikon) and processed with Elements 3.20 software (Nikon). For western blotting, cells were lysed as previously described14. Briefly, cells were lysed with 4 mol/L urea, 0.5% SDS, 0.5% NP-40, 100 mmol/L Tris, and 5 mmol/L EDTA, pH 7.4, containing 100 μmol/L leupeptin 10 mmol/L benazmidine, 1 mmol/L PMSF, and 12.5 μg/mL aprotinin. Lysates were subjected to 10% SDS-PAGE (13 μg of protein per lane) and transferred to Immobilon-P membrane (Millipore, Bedford, MA). Membranes were incubated with murine anti-human α-SMA, goat anti-human CD31, and goat anti-human VE-Cadherin diluted in 5% dry milk in 1× PBS-T (0.1% Tween-20, 25 μM Tris-HCL, 0.15 M NaCl in PBS), and then with secondary antibody (peroxidase-conjugated anti-mouse or anti-goat). Antigen-Ab complexes were visualized using chemiluminescent sensitive film. Equal protein amounts (12-15 μg) were loaded in each lane (determined by u-BCA assay, Pierce), and expression was quantified via densitometry analysis and normalized to that of β-actin (Sigma). All antibodies were shown to cross-react with their ovine homologs.
Real-time Polymerase Chain Reaction
Total RNA was isolated using RNeasy Mini Kit (Qiagen), supplemented with DNase I treatment (Qiagen). Reverse transcription was performed with Superscript II cDNA synthesis kit (Invitrogen/Life Technologies, Grand Island, NY) to obtain a target cDNA concentration of 0.335 μg/mL, followed by RT- PCR using SYBR Green (BioRad, Hercules, CA), and annealing temperatures of 95° C and 60° C for 35 cycles. Oligonucleotide primer sequences are presented in Table 1. All PCR products were sequenced using ABi DNA sequences (Children’s Hospital Boston core facility) to verify the sequence corresponding to the gene of interest (Suppl. Table 1).
Table 1
| Gene | Primer Sequence |
|---|---|
| GAPDH-F | 5’-ACCACAGTCCATGCCATCAC-‘3 |
| GAPDH-R | 5’-TTCACCACCCTGTTGCTGTA-‘3 |
| ShVE-Cadherin-F | 5’-ACATCCGTGGTTCTGGACTC-‘3 |
| ShVE-Cadherin-R | 5’-AGATGGGGAAGTTGTCGTTG-‘3 |
| ShSMA-F | 5’-TGCCATGTATGTGGCTATTCA-‘3 |
| ShSMA-R | 5’-ACCAGTTGTACGTCCAGAAGC-‘3 |
| ShSlug.2-F | 5’-GGACGCACACCTTACCTTGT-‘3 |
| ShSlug.2-R | 5’-CGAGAAGGTTTTGGAGCAAC-‘3 |
| MMP-2.3-F | 5’-GAGACTCCCACTTCGACGAC-‘3 |
| MMP-2.3-R | 5’-AACACCAGAGGAAACCATCG-‘3 |
| Osteocalcin-F | 5’-AGCTCATCACAGTCAGGGTTG-‘3 |
| Osteocalcin-R | 5’-AGCGAGGTGGTGAAGAGA-‘3 |
| Osteopontin-F | 5’-CTGATTTTCCCACTGACATT-‘3 |
| Osteopontin-R | 5’-CTATGGAATTCTTGGCTGAG-‘3 |
| Osteonectin-F | 5’-CGACTCTTCCTGCCACTTCT-‘3 |
| Osteonectin-R | 5’-TTGTGGCCCTTCTTGT-‘3 |
| Run×2-F | 5’-CGACAGTCCCAACTTCCTGT-‘3 |
| Run×2-R | 5’-CGGTAACCACAGTCCCATCT-‘3 |
Migration assay
100 μg/mL rat tail collagen type I (0.02 N acetic acid) was used to coat 6.5-mm Transwells with 8.0-μm pore polycarbonate membrane inserts for 1 hour at 37 °C, followed by one wash with PBS. Cells were treated as specified, trypsinized, and seeded in the upper chamber of the transwell plate at a density of 1×10 cells/well (100 μl volume). 300 μL of control media (EBM-2 / serum-free and growth factor-free) or VEC media supplemented with 20% FBS was added to the lower chamber. Cells were then allowed to migrate for 4 hours at 37° C. The cells in the upper chamber were gently removed using a cotton swab, and the lower surface was fixed with ice-cold methanol and mounted on glass slides in mounting media containing DAPI. Cells were counted using a fluorescent microscope.
Statistical analysis
Results are presented as mean +/− standard deviation (SD) unless indicated otherwise. Unpaired Student’s t-test was used for comparisons between two groups. One-way ANOVA was used to evaluate statistical significant differences in multiple groups. P < 0.05 was considered statistically significant.
Aortic valve cell isolation
Ovine tissues from animals 8-10 months of age, weighing 20–25 kg, were obtained under approved NIH guidelines for animal experimentation as performed at Children’s Hospital Boston. Valve leaflets were incubated in endothelial basal medium (EBM-2) (CC- 3156, Cambrex Bio Science, Walkersville, MD) with 5% fetal bovine serum (FBS) (Hyclone, Logan, UT), 1% GPS (Invitrogen, Carlsbad, California), 2 mmol/L l-glutamine, and 100 μg/ml gentamicin sulfate for 1–4 hours. They were then minced into 2-mm pieces and incubated with 0.2% collagenase A (Roche Diagnostics, Indianapolis, IN) in EBM-2 for 5 minutes at 37°C, and diluted wit h Hanks’ balanced salt solution containing 5% FBS, 1.26 mmol/L CaCl2, 0.8 mmol/L MgSO4, and 1% GPS (wash buffer). The supernatant was sedimented at 200×g, resuspended in VEC medium (EBM-2 medium, 10% heat-inactivated FBS, 1% GPS, and 2 ng/mL basic fibroblast growth factor [bFGF]; Roche Diagnostics, Indianapolis, IN) and plated. The following day, primary cultures were washed to remove unattached cells. Primary cultures were trypsinized, resuspended in growth medium at 3.3 cells/ml, and 100 μl were plated in each well of a 96-well plate, at a concentration of approximately one cell to every third well; visual inspection was performed to assess that single colonies appeared in a subset of wells. When the colonies covered two-thirds of the well, cells were split into 24-well dishes. Based on morphology, these clones were initially visually identified as either endothelial or interstitial. These observations were confirmed with phenotypic characterization and designated as either endothelial (VEC)26, 27 or interstitial (VIC-K3, VIC-K5, VIC-K6), and expanded on 1% gelatin-coated dishes in VEC media. Cells were passaged 1:3 or 1:4 every 6 to 14 days, and experiments were performed using VEC and VIC clones between passages 8 and 14. VECs were grown in VEC media, and VICs were cultured in VIC media (DMEM, 10% FBS, 1% GPS).
Human calcified aortic valves
Aortic valve leaflets (n=5) were harvested from patients undergoing aortic valve replacement for aortic valve stenosis and a healthy valve was obtained from autopsy, performed using criteria established by the declaration of Helsinki. Tissue samples were frozen in optimal cutting temperature compound (OCT, Sakura Finetech, Torrance, CA) and 7 μm serial sections were cut and stained. Tissue was collected according to Brigham and Women’s Hospital IRB Protocols.
Mouse model of CAVD
Male apolipoprotein E-deficient mice (apoE mice; 10 weeks old) were purchased from Jackson Laboratory (Bar Harbor, ME). High-fat diet (21% fat and 0.21% cholesterol) was obtained from Research Diets (D12079B, New Brunswick, NJ). Mice were fed with an atherogenic diet for a total of 22 weeks. Mice were euthanized for tissue collection and histopathology at 32 weeks. The animal procedures were performed conform the NIH guidelines and approved by the Brigham and Women’s Hospital Animal Care and Use Committee. This model is characterized by the presence of functional and histopathological changes found in human CAVD.33
Endothelial-to-Mesenchymal Transformation (EndMT)
EndMT was induced as described previously.27 The assay was also performed in the presence of VICs in an indirect co-culture system, using a Transwell system (12-mm Transwells with 0.4-μm pore polycarbonate membrane inserts; Corning Life Sciences, Acton, MA) (Supplemental Figure I). Briefly, VECs were grown alone or in co-culture with VICs in VEC media, or VEC media supplemented with 2 ng/mL TGF-β1 (R&D Systems, Minneapolis, MN), for up to 14 days. Media was changed every 2–3 days.
VEC and VIC osteogenic potential
The osteogenic potential of VEC and VIC clones was tested both separately and in co-culture for up to 21 days via treatment with osteogenic media (OM) (DMEM with 10% FBS, 1% GPS with 10 nmol/L beta-glycerolphosphate, 50 μmol/L ascorbic acid, 10 μmol/L dexamethasone). DMEM with 10% FBS, 1% GPS was used as control medium (NM). Media was changed every 2–3 days.
VEC-VIC indirect co-culture
Co-culture experiments were performed using Transwell plates (Corning, Tewksbury, MA) (Supplemental Figure I). Cells were plated either in the Transwell inserts or on the bottom of 6-well plates at a density of 1 × 10 cells/cm and allowed to adhere overnight. After 24 hours, the two cell types were combined to begin the co-culture by placing the inserts into the corresponding well of the 6 well plate. For the conditioned media experiments, cells were treated with media that had been conditioned by the specific cell type for 24 hours. Conditioned media was sterile-filtered and added in a 1:1 ratio with fresh media to treated cells. Media was changed every 2–3 days.
Histological detection of VEC and VIC mineralization
Upon completion of experiments, cells were washed with PBS and fixed using 4% paraformaldehyde for 15 minutes. Cells were subsequently washed with PBS. Mineralization was analyzed by staining with 0.02 mg/L of Alizarin Red S (Sigma-Aldrich). The area of positive Alizarin Red S staining was normalized to cell number. To detect the expression of alkaline phosphatase (ALP), nitro-blue tetrazolium/indolylphosphate (NBT/BCIP) staining was performed. Before staining, the cells were washed with PBS, 0.5 ml of NBT/BCIP was added, and the samples were then incubated at 37°C in a humidified chamber cont aining 5% CO2 for 30 minutes. Samples were then washed with PBS and fixed with 4% PFA, after which they underwent a counter stain with 0.1% eosin for 5 minutes. Images were taken with an Eclipse 80i microscope (Nikon) and processed with Elements 3.20 software (Nikon).
Alkaline phosphatase activity and calcium measurements
A colorimetric kit was used to measure ALP activity (BioVision, Milpitas, CA) according to the manufacturer’s instructions. 16 μl of supernatant from 12-well plates was added to 64 μl of ALP assay buffer and 50 μl of PnPP solution was added and incubated for 1 hour at room temperature. The absorbance was read at a wavelength of 405 nm. Values were normalized to the standard curve. Calcium content was quantified using a colorimetric kit (BioVision) according to the manufacturer’s protocol. Briefly, 50 μL of sample was added to 60 μL of calcium assay buffer, after which 90 μL of chromogenic reagent was added and incubated for 10 minutes at room temperature in the dark. Absorbance was read at a wavelength of 575 nm with a plate reader. Values were normalized to the standard curve.
Immunofluorescence and western blotting
Immunofluorescence staining was performed on methanol-fixed cells and human and mouse aortic valve leaflets using anti-human VE-cadherin (Santa Cruz) or CD31 (Cell Signaling), anti-human α-SMA (clone 1A4, Sigma-Aldrich), anti-vimentin antibody (Abcam, Cambridge, MA) and anti-osteocalcin (Abcam). Secondary antibodies conjugated with AlexaFluor 488 (Invitrogen) were used. Images were taken with an Eclipse 80i microscope (Nikon) and processed with Elements 3.20 software (Nikon). For western blotting, cells were lysed as previously described14. Briefly, cells were lysed with 4 mol/L urea, 0.5% SDS, 0.5% NP-40, 100 mmol/L Tris, and 5 mmol/L EDTA, pH 7.4, containing 100 μmol/L leupeptin 10 mmol/L benazmidine, 1 mmol/L PMSF, and 12.5 μg/mL aprotinin. Lysates were subjected to 10% SDS-PAGE (13 μg of protein per lane) and transferred to Immobilon-P membrane (Millipore, Bedford, MA). Membranes were incubated with murine anti-human α-SMA, goat anti-human CD31, and goat anti-human VE-Cadherin diluted in 5% dry milk in 1× PBS-T (0.1% Tween-20, 25 μM Tris-HCL, 0.15 M NaCl in PBS), and then with secondary antibody (peroxidase-conjugated anti-mouse or anti-goat). Antigen-Ab complexes were visualized using chemiluminescent sensitive film. Equal protein amounts (12-15 μg) were loaded in each lane (determined by u-BCA assay, Pierce), and expression was quantified via densitometry analysis and normalized to that of β-actin (Sigma). All antibodies were shown to cross-react with their ovine homologs.
Real-time Polymerase Chain Reaction
Total RNA was isolated using RNeasy Mini Kit (Qiagen), supplemented with DNase I treatment (Qiagen). Reverse transcription was performed with Superscript II cDNA synthesis kit (Invitrogen/Life Technologies, Grand Island, NY) to obtain a target cDNA concentration of 0.335 μg/mL, followed by RT- PCR using SYBR Green (BioRad, Hercules, CA), and annealing temperatures of 95° C and 60° C for 35 cycles. Oligonucleotide primer sequences are presented in Table 1. All PCR products were sequenced using ABi DNA sequences (Children’s Hospital Boston core facility) to verify the sequence corresponding to the gene of interest (Suppl. Table 1).
Table 1
| Gene | Primer Sequence |
|---|---|
| GAPDH-F | 5’-ACCACAGTCCATGCCATCAC-‘3 |
| GAPDH-R | 5’-TTCACCACCCTGTTGCTGTA-‘3 |
| ShVE-Cadherin-F | 5’-ACATCCGTGGTTCTGGACTC-‘3 |
| ShVE-Cadherin-R | 5’-AGATGGGGAAGTTGTCGTTG-‘3 |
| ShSMA-F | 5’-TGCCATGTATGTGGCTATTCA-‘3 |
| ShSMA-R | 5’-ACCAGTTGTACGTCCAGAAGC-‘3 |
| ShSlug.2-F | 5’-GGACGCACACCTTACCTTGT-‘3 |
| ShSlug.2-R | 5’-CGAGAAGGTTTTGGAGCAAC-‘3 |
| MMP-2.3-F | 5’-GAGACTCCCACTTCGACGAC-‘3 |
| MMP-2.3-R | 5’-AACACCAGAGGAAACCATCG-‘3 |
| Osteocalcin-F | 5’-AGCTCATCACAGTCAGGGTTG-‘3 |
| Osteocalcin-R | 5’-AGCGAGGTGGTGAAGAGA-‘3 |
| Osteopontin-F | 5’-CTGATTTTCCCACTGACATT-‘3 |
| Osteopontin-R | 5’-CTATGGAATTCTTGGCTGAG-‘3 |
| Osteonectin-F | 5’-CGACTCTTCCTGCCACTTCT-‘3 |
| Osteonectin-R | 5’-TTGTGGCCCTTCTTGT-‘3 |
| Run×2-F | 5’-CGACAGTCCCAACTTCCTGT-‘3 |
| Run×2-R | 5’-CGGTAACCACAGTCCCATCT-‘3 |
Migration assay
100 μg/mL rat tail collagen type I (0.02 N acetic acid) was used to coat 6.5-mm Transwells with 8.0-μm pore polycarbonate membrane inserts for 1 hour at 37 °C, followed by one wash with PBS. Cells were treated as specified, trypsinized, and seeded in the upper chamber of the transwell plate at a density of 1×10 cells/well (100 μl volume). 300 μL of control media (EBM-2 / serum-free and growth factor-free) or VEC media supplemented with 20% FBS was added to the lower chamber. Cells were then allowed to migrate for 4 hours at 37° C. The cells in the upper chamber were gently removed using a cotton swab, and the lower surface was fixed with ice-cold methanol and mounted on glass slides in mounting media containing DAPI. Cells were counted using a fluorescent microscope.
Statistical analysis
Results are presented as mean +/− standard deviation (SD) unless indicated otherwise. Unpaired Student’s t-test was used for comparisons between two groups. One-way ANOVA was used to evaluate statistical significant differences in multiple groups. P < 0.05 was considered statistically significant.
Results
TGF-β1 and osteogenic media induce VEC EndMT
Clonal populations from ovine aortic valve leaflets were isolated using a brief collagenase-A procedure, which has been described previously.12 The VEC clone demonstrated cobblestone endothelial morphology and stained positively for VE-Cadherin and negatively for α-SMA (Supplementary Figure II). VIC clonal populations from the same valve were isolated and characterized. Immunofluorescence staining of the aortic VIC clones (VIC-K3, VIC-K5, and VIC-K6) confirmed the characteristic myofibroblastic phenotype of VICs grown in culture, with positive staining for α-SMA and vimentin and negative staining for the endothelial cell marker VE-Cadherin (Supplementary Figure III).
In accordance with our earlier work,12 this VEC clone was able to undergo TGF-β1–induced EndMT, as visualized by anti-α-SMA immunofluorescence after 8 days of TGF-β1 stimulation (Supplementary Figure IVA). Western blot analysis confirmed the changes in protein expression (Supplemental Figure IVB). Significant increases in mRNA expression of the EndMT markers α-SMA (5.3±1.2), MMP-2 (13.5±0.6) and Slug (12±2.1), as well as a decrease of the endothelial marker VE-Cadherin (0.2±0.1) (p<0.05) confirmed TGF-β1–induced EndMT of the VECs (n=3, Supplementary Figures IVC-F).
We performed a time-course analysis of EndMT-associated proteins in VECs treated with either TGF-β1 or osteogenic media (OM) for up to 14 days. VECs were stained for VE-cadherin and α-SMA at days 1, 7, and 14 of culture (Figure 1A). Both TGF-β1 and OM induced a progressive loss of VE-cadherin compared to NM in the VECs. Quantification of α-SMA–positive cells confirmed the increase of myofibroblast-like differentiation of VECs after day 7 and day 14, as compared to day 1 (p<0.05). When VECs were cultured in OM, a higher percentage of α-SMA–positive cells were present on day 14, as compared to day 7 (Figure 1B). RT-PCR analysis of EndMT markers revealed a significant increase in mRNA expression of α-SMA at day 7 relative to day 1 when cultured with TGF-β1 (9.8±4.3) and OM (7.9±1.1), but α-SMA expression decreased at day 14 as compared to day 7 (Figure 1C). In cells stimulated with TGF-β1 or OM, MMP-2 expression increased significantly from day 7 (TGF-β1: 2.3±0.5; OM: 2.7±1.1) to day 14 (TGF-β1: 5.4±1.8; OM: 5.6±0.7). Slug expression follows a similar pattern to α-SMA, demonstrating a significant increase at day 7 (TGFβ1 5.8±1.9; OM: 6.4±2.2) compared to day 1 (p<0.05). While day 14 for Slug was not significantly different from day 7, it remained significantly increased as compared to day 1 (TGF-β1: 4.1±0.3; OM: 2.6±0.7) (Figure 1C). The increase in expression of these EndMT markers between days 1 and 7 suggests that a process of EndMT occurs in both the VECs treated with TGF-β1 and those treated with OM.
A-C. VECs (n=3) were treated with normal media (NM) or osteogenic media (OM) +/− TGF-β1 for 1, 7 or 14 days. A: Immunofluorescence staining of VE-cadherin and α-SMA (green), cell nuclei (DAPI/blue). (n=3). Bar=50 μm. B: Quantification of α-SMA staining. C: mRNA expression of α-SMA, MMP-2 and Slug. Data is depicted as mean +/− SD fold change, *p<0.05.
D. ALP staining of VECs in media + TGF-β1 or OM at day 14. (n=3). Bar=50 μm. E: ALP activity of VECs in media + TGF-β1 or OM. Data is depicted as mean +/− SD/normalized to NM conditions, *p<0.05. F: mRNA expression of osteopontin, osteocalcin and Run×2. Data is depicted as mean +/− SD fold change, *p<0.05.
Osteogenic differentiation follows EndMT in VECs
We next evaluated osteogenic differentiation over time. We detected a small number of ALP-positive cells following 14 days of stimulation with TGF-β1, but when VECs were cultured in OM, a more pronounced ALP staining was observed (Figure 1D). This observation was confirmed by quantification of the ALP activity (OM: 3.2±0,8 vs. TGF β1: 1.9 ±0.4 (Figure 1E). TGF-β1 significantly increased osteopontin (2.9±0.6) osteocalcin (4.8±0.5), and Run×2 (3.2±0.5) mRNA expression (p<0.05) at day 14 (Figure 1F). In addition, when we cultured VECs in OM, we found a significant increase in osteopontin (15.5±8.1), osteocalcin (11.5±4.3), and Run×2 (6.7±2.1) expression at day 14, as compared with earlier time points (p <0.05) (Figure 1F).
VICs suppress TGFβ1–induced EndMT of VECs
The presence of VICs (in co-culture) attenuated TGF-β1–induced EndMT. VICs suppressed TGF-β1–induced expression of α-SMA in VECs, as demonstrated by immunofluorescence staining (Figure 2A) and confirmed by Western blot. Three different VIC clones suppressed the EndMT marker α-SMA when the VECs were stimulated with TGF-β1 (Figure 2B, Supplementary Figure V). VICs significantly suppressed the expression of three EndMT markers in TGF-β1 treated VECs: α-SMA (0.1±0.5), MMP-2 (0.1±0.1) and Slug (0.2±0.2) (three different VIC clones, p<0.05; Figures 2C, 2D, 2E). A similar inhibition of EndMT markers was observed when using VIC conditioned media at a 1:1 ratio (Supplementary Figure VI). The presence of VICs also inhibited TGF-β1-induced migration potential of VECs that were first co-cultured with VICs and TGF-β1, as compared to VECs treated with TGF-β1 alone (Supplementary Figure VII).
VECs (n=3) were co-cultured with VICs (n=3) in a Transwell culture system and treated with TGF-β1 for 8 days. A: Immunofluorescence staining of α-SMA (green), cell nuclei (DAPI/blue). (n=3). Bar=50 μm. B: Western blot for endothelial markers (VE-cadherin, CD31) and myofibroblastic marker (α-SMA). C: mRNA expression of EndMT markers α-SMA, MMP-2 and Slug. Data is depicted as mean +/− SD fold change, *p<0.05.
VICs inhibit osteogenic differentiation of VECs
VECs were cultured in OM to evaluate their osteogenic differentiation capacity, using DMEM-based normal growth media (NM) as a control. VECs cultured in OM for 21 days demonstrated a loss of VE-Cadherin, compared with VECs cultured in NM (Figure 3A). Mineralized matrix, visualized using Alizarin Red S staining, was observed after culturing VECs for 21 days in OM, but was not detected in VECs cultured in NM (Figure 3A). The presence of VICs (in co-culture) prevented both the OM-mediated decrease in VE-Cadherin and the increase in mineralized matrix (Figure 3B). There was no difference in cell number between groups (Supplementary Figure VIII). Analyses of mRNA expression at day 21 confirmed the inhibitory effect of VICs on the osteogenic differentiation of VECs. VE-Cadherin expression decreased significantly in VECs in OM compared to NM (0.3±0.2 p<0.05). This decrease was inhibited by the presence of VICs in co-culture (0.8±0.5) (Figure 3C). Expression of α-SMA increased when VECs were cultured in OM (17.3±7.5, p<0.05), and this increase was mitigated by the presence of VICs in co-culture (1.3±0.7) (Figure 3D). VEC cultured in OM showed increased expression of osteocalcin (8.6±1.3, p<0.05), osteopontin (3.7±0.3, p<0.05) and Run×2 (5.5±1.5, p<0.05), compared with cells cultured in NM (Figures 3E, 3F, and 3G). The co-culture of VECs with VICs in OM abolished the induction of osteogenic differentiation markers. A functional consequence of osteogenic differentiation, calcium deposition, increased when VECs were cultured in OM alone, but was significantly impaired when VECs were co-cultured with VICs (OM: 4.2±1.7 μg/mL, OM+VICs: 1.9±0.8 μg/mL, n=3, p<0.05) (Figure 3H).
VECs were co-cultured with VICs in a Transwell culture system in osteogenic media (OM) for 21 days. A-B: Immunofluorescence staining of VE-cadherin (green), α-SMA (green), cell nuclei (DAPI/blue), and Alizarin Red S (ARS) (red/orange). (n=3). Bar=50 μm. C: mRNA expression of VE-cadherin, D: α-SMA, E: Osteocalcin, F: Osteopontin, G: Run×2. H: Calcium content. Data is depicted as mean +/− SD fold change, *p<0.05.
VECs do not suppress osteogenic differentiation of VICs
We evaluated whether VECs have a similar inhibitory effect on the osteogenic differentiation of VICs. VICs cultured in OM for 21 days demonstrated mineralized matrix by Alizarin Red S staining (Figure 4A). When VICs were co-cultured with VECs in NM or OM, VICs also stained positively for both α-SMA and calcium (Figure 4B). Expression of α-SMA increased in VICs cultured for 21 days in OM (1.4±0.3, p<0.05) (Figure 4C). VICs cultured in OM with VECs demonstrated a significant decrease in α-SMA expression (0.2±0.1 p<0.05). The mRNA expression of osteogenic differentiation markers osteocalcin, osteopontin, and Run×2 and the activity of ALP further revealed that VECs do not exhibit an inhibitory effect on VIC osteogenic differentiation (Figures 4D, 4E, and 4F, Supplementary Figure IX). Functionally, we observed a significant increase in VIC calcium deposition in the co-culture samples (Figure 4G).
VICs (n=3) were co-cultured with VECs in a Transwell culture system in osteogenic media (OM) for 21 days. A-B: Immunofluorescence staining of Alizarin Red S (red/orange), α-SMA (green), cell nuclei (DAPI/blue). (n=3). Bar=50 μm. C: mRNA expression of α-SMA, D: Osteocalcin, E: Osteopontin, F: Run×2. G: Calcium content. Data is depicted as mean +/− SD fold change, *p<0.05.
Human and mouse calcified aortic valves leaflets demonstrate EndMT
After observing EndMT in isolated aortic VECs, we evaluated the presence of EndMT in human calcified aortic valve leaflets. Using immunofluorescence we demonstrate co-expression of α-SMA and CD31 (Figure 5), confirming the presence of EndMT in calcific valves. In addition, both α-SMA and CD31 co-expressed with osteocalcin, indicating a potential role for EndMT in human calcific aortic valve disease. Further, α-SMA was not observed in the endothelium of a non-calcified human aortic valve leaflets. To further evaluate the in vivo relevance of EndMT in CAVD we assessed α-SMA expression in the aortic valve of wild type and Apoe −/− mice, a common mouse model of cardiovascular calcification.33 Increased expression of a-SMA was observed in the endothelium of the Apoe −/− mice.
Human non-calcified and calcified aortic valves (n=6) stained for CD31, α-SMA and osteocalcin. Aortic root sections from wild type (n=2) and apoE−/− mice (n=3) stained with CD31 and α-SMA. Representative images of the leaflets are shown. * Aortic side. Bar=20 μm.
TGF-β1 and osteogenic media induce VEC EndMT
Clonal populations from ovine aortic valve leaflets were isolated using a brief collagenase-A procedure, which has been described previously.12 The VEC clone demonstrated cobblestone endothelial morphology and stained positively for VE-Cadherin and negatively for α-SMA (Supplementary Figure II). VIC clonal populations from the same valve were isolated and characterized. Immunofluorescence staining of the aortic VIC clones (VIC-K3, VIC-K5, and VIC-K6) confirmed the characteristic myofibroblastic phenotype of VICs grown in culture, with positive staining for α-SMA and vimentin and negative staining for the endothelial cell marker VE-Cadherin (Supplementary Figure III).
In accordance with our earlier work,12 this VEC clone was able to undergo TGF-β1–induced EndMT, as visualized by anti-α-SMA immunofluorescence after 8 days of TGF-β1 stimulation (Supplementary Figure IVA). Western blot analysis confirmed the changes in protein expression (Supplemental Figure IVB). Significant increases in mRNA expression of the EndMT markers α-SMA (5.3±1.2), MMP-2 (13.5±0.6) and Slug (12±2.1), as well as a decrease of the endothelial marker VE-Cadherin (0.2±0.1) (p<0.05) confirmed TGF-β1–induced EndMT of the VECs (n=3, Supplementary Figures IVC-F).
We performed a time-course analysis of EndMT-associated proteins in VECs treated with either TGF-β1 or osteogenic media (OM) for up to 14 days. VECs were stained for VE-cadherin and α-SMA at days 1, 7, and 14 of culture (Figure 1A). Both TGF-β1 and OM induced a progressive loss of VE-cadherin compared to NM in the VECs. Quantification of α-SMA–positive cells confirmed the increase of myofibroblast-like differentiation of VECs after day 7 and day 14, as compared to day 1 (p<0.05). When VECs were cultured in OM, a higher percentage of α-SMA–positive cells were present on day 14, as compared to day 7 (Figure 1B). RT-PCR analysis of EndMT markers revealed a significant increase in mRNA expression of α-SMA at day 7 relative to day 1 when cultured with TGF-β1 (9.8±4.3) and OM (7.9±1.1), but α-SMA expression decreased at day 14 as compared to day 7 (Figure 1C). In cells stimulated with TGF-β1 or OM, MMP-2 expression increased significantly from day 7 (TGF-β1: 2.3±0.5; OM: 2.7±1.1) to day 14 (TGF-β1: 5.4±1.8; OM: 5.6±0.7). Slug expression follows a similar pattern to α-SMA, demonstrating a significant increase at day 7 (TGFβ1 5.8±1.9; OM: 6.4±2.2) compared to day 1 (p<0.05). While day 14 for Slug was not significantly different from day 7, it remained significantly increased as compared to day 1 (TGF-β1: 4.1±0.3; OM: 2.6±0.7) (Figure 1C). The increase in expression of these EndMT markers between days 1 and 7 suggests that a process of EndMT occurs in both the VECs treated with TGF-β1 and those treated with OM.
A-C. VECs (n=3) were treated with normal media (NM) or osteogenic media (OM) +/− TGF-β1 for 1, 7 or 14 days. A: Immunofluorescence staining of VE-cadherin and α-SMA (green), cell nuclei (DAPI/blue). (n=3). Bar=50 μm. B: Quantification of α-SMA staining. C: mRNA expression of α-SMA, MMP-2 and Slug. Data is depicted as mean +/− SD fold change, *p<0.05.
D. ALP staining of VECs in media + TGF-β1 or OM at day 14. (n=3). Bar=50 μm. E: ALP activity of VECs in media + TGF-β1 or OM. Data is depicted as mean +/− SD/normalized to NM conditions, *p<0.05. F: mRNA expression of osteopontin, osteocalcin and Run×2. Data is depicted as mean +/− SD fold change, *p<0.05.
Osteogenic differentiation follows EndMT in VECs
We next evaluated osteogenic differentiation over time. We detected a small number of ALP-positive cells following 14 days of stimulation with TGF-β1, but when VECs were cultured in OM, a more pronounced ALP staining was observed (Figure 1D). This observation was confirmed by quantification of the ALP activity (OM: 3.2±0,8 vs. TGF β1: 1.9 ±0.4 (Figure 1E). TGF-β1 significantly increased osteopontin (2.9±0.6) osteocalcin (4.8±0.5), and Run×2 (3.2±0.5) mRNA expression (p<0.05) at day 14 (Figure 1F). In addition, when we cultured VECs in OM, we found a significant increase in osteopontin (15.5±8.1), osteocalcin (11.5±4.3), and Run×2 (6.7±2.1) expression at day 14, as compared with earlier time points (p <0.05) (Figure 1F).
VICs suppress TGFβ1–induced EndMT of VECs
The presence of VICs (in co-culture) attenuated TGF-β1–induced EndMT. VICs suppressed TGF-β1–induced expression of α-SMA in VECs, as demonstrated by immunofluorescence staining (Figure 2A) and confirmed by Western blot. Three different VIC clones suppressed the EndMT marker α-SMA when the VECs were stimulated with TGF-β1 (Figure 2B, Supplementary Figure V). VICs significantly suppressed the expression of three EndMT markers in TGF-β1 treated VECs: α-SMA (0.1±0.5), MMP-2 (0.1±0.1) and Slug (0.2±0.2) (three different VIC clones, p<0.05; Figures 2C, 2D, 2E). A similar inhibition of EndMT markers was observed when using VIC conditioned media at a 1:1 ratio (Supplementary Figure VI). The presence of VICs also inhibited TGF-β1-induced migration potential of VECs that were first co-cultured with VICs and TGF-β1, as compared to VECs treated with TGF-β1 alone (Supplementary Figure VII).
VECs (n=3) were co-cultured with VICs (n=3) in a Transwell culture system and treated with TGF-β1 for 8 days. A: Immunofluorescence staining of α-SMA (green), cell nuclei (DAPI/blue). (n=3). Bar=50 μm. B: Western blot for endothelial markers (VE-cadherin, CD31) and myofibroblastic marker (α-SMA). C: mRNA expression of EndMT markers α-SMA, MMP-2 and Slug. Data is depicted as mean +/− SD fold change, *p<0.05.
VICs inhibit osteogenic differentiation of VECs
VECs were cultured in OM to evaluate their osteogenic differentiation capacity, using DMEM-based normal growth media (NM) as a control. VECs cultured in OM for 21 days demonstrated a loss of VE-Cadherin, compared with VECs cultured in NM (Figure 3A). Mineralized matrix, visualized using Alizarin Red S staining, was observed after culturing VECs for 21 days in OM, but was not detected in VECs cultured in NM (Figure 3A). The presence of VICs (in co-culture) prevented both the OM-mediated decrease in VE-Cadherin and the increase in mineralized matrix (Figure 3B). There was no difference in cell number between groups (Supplementary Figure VIII). Analyses of mRNA expression at day 21 confirmed the inhibitory effect of VICs on the osteogenic differentiation of VECs. VE-Cadherin expression decreased significantly in VECs in OM compared to NM (0.3±0.2 p<0.05). This decrease was inhibited by the presence of VICs in co-culture (0.8±0.5) (Figure 3C). Expression of α-SMA increased when VECs were cultured in OM (17.3±7.5, p<0.05), and this increase was mitigated by the presence of VICs in co-culture (1.3±0.7) (Figure 3D). VEC cultured in OM showed increased expression of osteocalcin (8.6±1.3, p<0.05), osteopontin (3.7±0.3, p<0.05) and Run×2 (5.5±1.5, p<0.05), compared with cells cultured in NM (Figures 3E, 3F, and 3G). The co-culture of VECs with VICs in OM abolished the induction of osteogenic differentiation markers. A functional consequence of osteogenic differentiation, calcium deposition, increased when VECs were cultured in OM alone, but was significantly impaired when VECs were co-cultured with VICs (OM: 4.2±1.7 μg/mL, OM+VICs: 1.9±0.8 μg/mL, n=3, p<0.05) (Figure 3H).
VECs were co-cultured with VICs in a Transwell culture system in osteogenic media (OM) for 21 days. A-B: Immunofluorescence staining of VE-cadherin (green), α-SMA (green), cell nuclei (DAPI/blue), and Alizarin Red S (ARS) (red/orange). (n=3). Bar=50 μm. C: mRNA expression of VE-cadherin, D: α-SMA, E: Osteocalcin, F: Osteopontin, G: Run×2. H: Calcium content. Data is depicted as mean +/− SD fold change, *p<0.05.
VECs do not suppress osteogenic differentiation of VICs
We evaluated whether VECs have a similar inhibitory effect on the osteogenic differentiation of VICs. VICs cultured in OM for 21 days demonstrated mineralized matrix by Alizarin Red S staining (Figure 4A). When VICs were co-cultured with VECs in NM or OM, VICs also stained positively for both α-SMA and calcium (Figure 4B). Expression of α-SMA increased in VICs cultured for 21 days in OM (1.4±0.3, p<0.05) (Figure 4C). VICs cultured in OM with VECs demonstrated a significant decrease in α-SMA expression (0.2±0.1 p<0.05). The mRNA expression of osteogenic differentiation markers osteocalcin, osteopontin, and Run×2 and the activity of ALP further revealed that VECs do not exhibit an inhibitory effect on VIC osteogenic differentiation (Figures 4D, 4E, and 4F, Supplementary Figure IX). Functionally, we observed a significant increase in VIC calcium deposition in the co-culture samples (Figure 4G).
VICs (n=3) were co-cultured with VECs in a Transwell culture system in osteogenic media (OM) for 21 days. A-B: Immunofluorescence staining of Alizarin Red S (red/orange), α-SMA (green), cell nuclei (DAPI/blue). (n=3). Bar=50 μm. C: mRNA expression of α-SMA, D: Osteocalcin, E: Osteopontin, F: Run×2. G: Calcium content. Data is depicted as mean +/− SD fold change, *p<0.05.
Human and mouse calcified aortic valves leaflets demonstrate EndMT
After observing EndMT in isolated aortic VECs, we evaluated the presence of EndMT in human calcified aortic valve leaflets. Using immunofluorescence we demonstrate co-expression of α-SMA and CD31 (Figure 5), confirming the presence of EndMT in calcific valves. In addition, both α-SMA and CD31 co-expressed with osteocalcin, indicating a potential role for EndMT in human calcific aortic valve disease. Further, α-SMA was not observed in the endothelium of a non-calcified human aortic valve leaflets. To further evaluate the in vivo relevance of EndMT in CAVD we assessed α-SMA expression in the aortic valve of wild type and Apoe −/− mice, a common mouse model of cardiovascular calcification.33 Increased expression of a-SMA was observed in the endothelium of the Apoe −/− mice.
Human non-calcified and calcified aortic valves (n=6) stained for CD31, α-SMA and osteocalcin. Aortic root sections from wild type (n=2) and apoE−/− mice (n=3) stained with CD31 and α-SMA. Representative images of the leaflets are shown. * Aortic side. Bar=20 μm.
Discussion
We report that VICs can inhibit EndMT of VECs even when stimulated with TGF-β1, a well-established inducer of EndMT.12,13,19,28 We have also demonstrated that VEC osteogenic differentiation is inhibited by VICs when cultured in an osteogenic environment. Conversely, in our study, VECs did not inhibit VIC mineralization. In addition, we have shown that EndMT may precede VEC osteogenesis. Finally, EndMT was observed to co-express with osteogenic markers in a mouse model of aortic valve calcification and human aortic valves obtained from patients with calcific aortic valve disease. We thus propose that VECs contain the capacity to differentiate into endothelial-derived VICs (eVICs) through an EndMT process. In certain disease conditions where communication between VICs and VECs is disrupted, EndMT may be promoted. EndMT-derived eVICs may populate the valve leaflet and differentiate into osteoblastic cells (oVICs),contributing to the pathological remodeling observed in CAVD (Figure 6).
(i) Quiescent VICs (qVICs) may differentiate into activated myofibroblast-like VICs (aVICs), responsible for functional remodeling of the heart valve ECM. The interplay between qVICs and aVICs is thought to be the cornerstone of valve homeostasis. (ii) Upon pathological stimulation, aVICs can further differentiate into osteoblastic VICs (oVICs), which may be responsible for calcium deposition in CAVD. (iii) VICs normally inhibit VEC EndMT. (iv) However, when VEC-VIC interactions are disrupted as CAVD progresses, VECs can differentiate into endothelial-derived VICs (eVICs) via EndMT. (v) In turn, eVICs may also differentiate into oVICs and contribute to calcification of the valve.
This study builds on our previous work, in which we demonstrated that VEC clones from ovine mitral valve leaflets might be a source for osteoblastic VICs.14 Endothelial osteoblastic differentiation potential has been proposed in VECs,14 prostate tumor EC,29 mutant ECs with constitutively active ALK2 30 and in arterial endothelial cells in matrix Gla protein-deficient mice.31 We previously showed that the in vitro osteogenic differentiation potential of mitral VEC corresponded with focal regions of osteogenic endothelium in tethered mitral valves in vivo.20 The role of the endothelium in valve calcification was also suggested by the finding that the endothelial activation marker VCAM-1 expression correlated with VIC osteoblastic differentiation in a model of aortic valve stenosis32. This work also underscores the unique plasticity within subsets of VECs, which is reflected not only in diseased states but also in normal valve physiology, as evident by the co-expression of CD31 and α-SMA in human fetal and postnatal semilunar valves.19
These findings have prompted our hypothesis that — when needed — a progenitor-like subset of VECs can replenish the VIC population via EndMT.14 In turn, this transdifferentiation could contribute to maintaining structural integrity and function of the heart valve (Figure 6). In normal valves, qVICs are activated by environmental cues and can differentiate into myofibroblast-like VICs (aVICs; α-SMA-positive), which maintain tissue integrity by adaptive remodeling of the valve ECM through the secretion of various cytokines,33, 34 matrix metalloproteinases,35, 36 and deposition of ECM proteins.34, 37 But when persistent activation of VICs occurs, an excessive, lasting remodeling of the valve ECM may also take place. Such a maladaptive process may lead to a pathological disruption of the valve’s normal connective tissue homeostasis, leading to fibrosis and eventual calcification by differentiation of oVICs.38 Although the role of VECs in CAVD remains to be elucidated, mounting evidence indicates that endothelial dysfunction correlates with such continuous maladaptive VIC activation.38 To our knowledge, the present work is the first to investigate VEC–VIC direct interaction in relation to osteogenesis in an in vitro culture model system. VICs in culture mostly demonstrate a myofibroblast-like phenotype attributed to the unnatural stiff substrate of the tissue culture plates.39 To what extent qVICs affect VEC phenotype in co-culture remains to be elucidated. In the current experimental setup, we cannot separate the culture conditions to modulate VIC and VEC phenotypes independently. Therefore, both VICs and VECs were cultured in an osteogenic environment. This may be similar to the tissue, wherein both cell types are likely exposed to pathologic stimuli simultaneously, but it is possible that certain cues lead to phenotypic changes in only one cell type (e.g., hemodynamic changes that affect VECs only). Future studies may try to build on the current work to isolate changes in each cell type.
It remains unclear how closely the EndMT we have observed in VECs in vitro reflects the EndMT that occurs in vivo, either during valve development or disease. As such, it is important to note that although we demonstrate a correlation of EndMT and osteogenesis, our in vivo results cannot offer a causal role for VEC EndMT in CAVD. Our in vivo knowledge of valve EndMT mostly stems from end-point analyses of human postnatal pulmonary valve specimens or studies of the murine endocardial cushion19, where hallmarks of EndMT consist of a loss of cell–cell contact in the EC monolayer; increased expression of α-SMA, MMP-2, and Slug; and increased cellular invasion. The present study confirms our earlier work that EndMT — as determined by these hallmarks — can be simulated in vitro.12 Therefore, our current results build upon previous work suggesting that EndMT plays an important role in the onset of CAVD. Aortic VECs represent a cell population with the intrinsic plasticity to differentiate into myofibroblast-like aVICs, and further into osteoblast-like oVICs. That both phenotypes have been shown to possess the potential to contribute to the development of CAVD underscores the importance of understanding the role of the valvular endothelium in the disease process. Future studies may use cell lineage tracing and cell fate models to determine the source of cells that populate the aortic valve leaflet during tissue remodeling and pathogenesis.40 By building a better understanding the cellular contributions valve remodeling, specific populations of cells may be targeted to control valvular homeostasis and develop therapeutics for CAVD.
Supplementary Material
1
2
3
1
2
3
Acknowledgments
The authors would like to thank Sara Karwacki for excellent editorial assistance.
Sources of Funding
This study was supported by grants from the Netherlands Heart Foundation (NHS-2011T024), Netherlands Scientific Council (NWO) to J.H., the Tommy Kaplan Discretionary Fund to J.E.M., and National Institutes of Health R01HL114805, to E.A.; and R01HL109506, to E.A. and J.B.
Abstract
Background
Calcific aortic valve disease (CAVD) is the most common heart valve disease in the Western world. We previously proposed that valvular endothelial cells (VECs) replenish injured adult valve leaflets via endothelial-to-mesenchymal transformation (EndMT); however, whether EndMT contributes to valvular calcification is unknown. We hypothesized that aortic VECs undergo osteogenic differentiation via an EndMT process that can be inhibited by valvular interstitial cells (VICs).
Approach and Results
VEC clones underwent TGF-β1-mediated EndMT, shown by significantly increased mRNA expression of the EndMT markers α-SMA (5.3±1.2), MMP-2 (13.5±0.6) and Slug (12±2.1) (p<0.05), (compared to unstimulated controls). To study the effects of VIC on VEC EndMT, clonal populations of VICs were derived from the same valve leaflets, placed in co-culture with VECs, and grown in control/TGF-β1 supplemented media. In the presence of VICs, EndMT was inhibited, shown by decreased mRNA expression of α-SMA (0.1±0.5), MMP-2 (0.1±0.1), and Slug (0.2±0.2) (p<0.05). When cultured in osteogenic media, VECs demonstrated osteogenic changes confirmed by increase in mRNA expression of osteocalcin (8.6±1.3), osteopontin (3.7±0.3), and Run×2 (5.5±1.5). The VIC presence inhibited VEC osteogenesis, demonstrated by decreased expression of osteocalcin (0.4±0.1) and osteopontin (0.2±0.1) (p<0.05). Time course analysis suggested that EndMT precedes osteogenesis, shown by an initial increase of α-SMA and MMP-2 (day 7), followed by an increase of osteopontin and osteocalcin (day 14).
Conclusions
The data indicate that EndMT may precede VEC osteogenesis. This study shows that VICs inhibit VEC EndMT and osteogenesis, indicating the importance of VEC–VIC interactions in valve homeostasis.
Abbreviations
| CAVD | Calcific aortic valve disease |
| VEC | valvular endothelial cell |
| EndMT | Endothelial-to-mesenchymal transformation |
| VIC | valvular interstitial cell |
| α-SMA | alpha-smooth muscle actin |
| MMP-2 | matrix metalloproteinase 2 |
| (TGF-β1) | transforming growth factor beta 1 |
| qVICs | quiescent fibroblast-like VIC |
| aVICs | activated myofibroblast-like VICs |
| oVICs | osteoblast-like VICs |
| EBM-2 | endothelial basal media |
| FBS | fetal bovine serum |
| GPS | glutamine-penicillin-streptomycin |
| bFGF | basic fibroblast growth factor |
| DMEM | Dulbecco’s modified eagles media |
| OM | osteogenic media |
| NM | control media |
| PBS | phosphate buffered saline |
| PFA | paraformaldehyde |
| NBT/BCIP | nitro-blue tetrazolium/indolylphosphate |
| ALP | alkaline phosphatase |
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Disclosures
None.
References
- 1. Stewart BF, Siscovick D, Lind BK, Gardin JM, Gottdiener JS, Smith VE, Kitzman DW, Otto CM. Clinical factors associated with calcific aortic valve disease. Cardiovascular health study. J Am Coll Cardiol. 1997;29:630–634.[PubMed]
- 2. Takkenberg JJ, Rajamannan NM, Rosenhek R, Kumar AS, Carapetis JR, Yacoub MH. The need for a global perspective on heart valve disease epidemiology. The shvd working group on epidemiology of heart valve disease founding statement. J Heart Valve Dis. 2008;17:135–139.[PubMed]
- 3. Rajamannan NM, Evans FJ, Aikawa E, Grande-Allen KJ, Demer LL, Heistad DD, Simmons CA, Masters KS, Mathieu P, O’Brien KD, Schoen FJ, Towler DA, Yoganathan AP, Otto CM. Calcific aortic valve disease: Not simply a degenerative process: A review and agenda for research from the national heart and lung and blood institute aortic stenosis working group. Executive summary: Calcific aortic valve disease-2011 update. Circulation. 2011;124:1783–1791.
- 4. Otto CMCalcific aortic stenosis--time to look more closely at the valve. N Engl J Med. 2008;359:1395–1398.[PubMed][Google Scholar]
- 5. Liu AC, Joag VR, Gotlieb AIThe emerging role of valve interstitial cell phenotypes in regulating heart valve pathobiology. The American journal of pathology. 2007;171:1407–1418.[Google Scholar]
- 6. Otto CM, Kuusisto J, Reichenbach DD, Gown AM, O’Brien KD. Characterization of the early lesion of ‘degenerative’ valvular aortic stenosis. Histological and immunohistochemical studies. Circulation. 1994;90:844–853.[PubMed]
- 7. Mohler ER, 3rd, Gannon F, Reynolds C, Zimmerman R, Keane MG, Kaplan FSBone formation and inflammation in cardiac valves. Circulation. 2001;103:1522–1528.[PubMed][Google Scholar]
- 8. Aikawa E, Whittaker P, Farber M, Mendelson K, Padera RF, Aikawa M, Schoen FJHuman semilunar cardiac valve remodeling by activated cells from fetus to adult: Implications for postnatal adaptation, pathology, and tissue engineering. Circulation. 2006;113:1344–1352.[PubMed][Google Scholar]
- 9. Peacock JD, Levay AK, Gillaspie DB, Tao G, Lincoln JReduced sox9 function promotes heart valve calcification phenotypes in vivo. Circ Res. 2010;106:712–719.[Google Scholar]
- 10. Osman L, Yacoub MH, Latif N, Amrani M, Chester AHRole of human valve interstitial cells in valve calcification and their response to atorvastatin. Circulation. 2006;114:I547–552.[PubMed][Google Scholar]
- 11. Rajamannan NM, Subramaniam M, Rickard D, Stock SR, Donovan J, Springett M, Orszulak T, Fullerton DA, Tajik AJ, Bonow RO, Spelsberg THuman aortic valve calcification is associated with an osteoblast phenotype. Circulation. 2003;107:2181–2184.[Google Scholar]
- 12. Paranya G, Vineberg S, Dvorin E, Kaushal S, Roth SJ, Rabkin E, Schoen FJ, Bischoff JAortic valve endothelial cells undergo transforming growth factor-beta-mediated and non-transforming growth factor-beta-mediated transdifferentiation in vitro. The American journal of pathology. 2001;159:1335–1343.[Google Scholar]
- 13. Armstrong EJ, Bischoff JHeart valve development: Endothelial cell signaling and differentiation. Circ Res. 2004;95:459–470.[Google Scholar]
- 14. Wylie-Sears J, Aikawa E, Levine RA, Yang JH, Bischoff JMitral valve endothelial cells with osteogenic differentiation potential. Arterioscler Thromb Vasc Biol. 2011;31:598–607.[Google Scholar]
- 15. Madri JA, Pratt BM, Tucker AMPhenotypic modulation of endothelial cells by transforming growth factor-beta depends upon the composition and organization of the extracellular matrix. The Journal of cell biology. 1988;106:1375–1384.[Google Scholar]
- 16. Combs MD, Yutzey KEHeart valve development: Regulatory networks in development and disease. Circ Res. 2009;105:408–421.[Google Scholar]
- 17. Person AD, Klewer SE, Runyan RBCell biology of cardiac cushion development. International review of cytology. 2005;243:287–335.[PubMed][Google Scholar]
- 18. Frid MG, Kale VA, Stenmark KRMature vascular endothelium can give rise to smooth muscle cells via endothelial-mesenchymal transdifferentiation: In vitro analysis. Circ Res. 2002;90:1189–1196.[PubMed][Google Scholar]
- 19. Paruchuri S, Yang JH, Aikawa E, Melero-Martin JM, Khan ZA, Loukogeorgakis S, Schoen FJ, Bischoff JHuman pulmonary valve progenitor cells exhibit endothelial/mesenchymal plasticity in response to vascular endothelial growth factor-a and transforming growth factor-beta2. Circ Res. 2006;99:861–869.[Google Scholar]
- 20. Dal-Bianco JP, Aikawa E, Bischoff J, Guerrero JL, Handschumacher MD, Sullivan S, Johnson B, Titus JS, Iwamoto Y, Wylie-Sears J, Levine RA, Carpentier AActive adaptation of the tethered mitral valve: Insights into a compensatory mechanism for functional mitral regurgitation. Circulation. 2009;120:334–342.[Google Scholar]
- 21. Chaput M, Handschumacher MD, Guerrero JL, Holmvang G, Dal-Bianco JP, Sullivan S, Vlahakes GJ, Hung J, Levine RA, Leducq Foundation MTN Mitral leaflet adaptation to ventricular remodeling: Prospective changes in a model of ischemic mitral regurgitation. Circulation. 2009;120:S99–103.
- 22. Balachandran K, Alford PW, Wylie-Sears J, Goss JA, Grosberg A, Bischoff J, Aikawa E, Levine RA, Parker KKCyclic strain induces dual-mode endothelial-mesenchymal transformation of the cardiac valve. Proc Natl Acad Sci U S A. 2011;108:19943–19948.[Google Scholar]
- 23. Skowasch D, Schrempf S, Wernert N, Steinmetz M, Jabs A, Tuleta I, Welsch U, Preusse CJ, Likungu JA, Welz A, Luderitz B, Bauriedel GCells of primarily extra-valvular origin in degenerative aortic valves and bioprostheses. European heart journal. 2005;26:2576–2580.[PubMed][Google Scholar]
- 24. Bosse K, Hans CP, Zhao N, Koenig SN, Huang N, Guggilam A, LaHaye S, Tao G, Lucchesi PA, Lincoln J, Lilly B, Garg VEndothelial nitric oxide signaling regulates notch1 in aortic valve disease. Journal of molecular and cellular cardiology. 2013;60:27–35.[Google Scholar]
- 25. Richards J, El-Hamamsy I, Chen S, Sarang Z, Sarathchandra P, Yacoub MH, Chester AH, Butcher JTSide-specific endothelial-dependent regulation of aortic valve calcification: Interplay of hemodynamics and nitric oxide signaling. The American journal of pathology. 2013;182:1922–1931.[Google Scholar]
- 26. Melero-Martin JM, De Obaldia ME, Kang SY, Khan ZA, Yuan L, Oettgen P, Bischoff JEngineering robust and functional vascular networks in vivo with human adult and cord blood-derived progenitor cells. Circ Res. 2008;103:194–202.[Google Scholar]
- 27. Yang JH, Wylie-Sears J, Bischoff JOpposing actions of notch1 and vegf in post-natal cardiac valve endothelial cells. Biochem Biophys Res Commun. 2008;374:512–516.[Google Scholar]
- 28. Dvorin EL, Jacobson J, Roth SJ, Bischoff JHuman pulmonary valve endothelial cells express functional adhesion molecules for leukocytes. J Heart Valve Dis. 2003;12:617–624.[Google Scholar]
- 29. Dudley AC, Khan ZA, Shih SC, Kang SY, Zwaans BM, Bischoff J, Klagsbrun MCalcification of multipotent prostate tumor endothelium. Cancer cell. 2008;14:201–211.[Google Scholar]
- 30. Medici D, Shore EM, Lounev VY, Kaplan FS, Kalluri R, Olsen BRConversion of vascular endothelial cells into multipotent stem-like cells. Nat Med. 2010;16:1400–1406.[Google Scholar]
- 31. Yao Y, Jumabay M, Ly A, Radparvar M, Cubberly MR, Bostrom KIA role for the endothelium in vascular calcification. Circ Res. 2013;113:495–504.[Google Scholar]
- 32. Aikawa E, Nahrendorf M, Sosnovik D, Lok VM, Jaffer FA, Aikawa M, Weissleder RMultimodality molecular imaging identifies proteolytic and osteogenic activities in early aortic valve disease. Circulation. 2007;115:377–386.[PubMed][Google Scholar]
- 33. Chester AH, Taylor PMMolecular and functional characteristics of heart-valve interstitial cells. Philosophical transactions of the Royal Society of London. Series B, Biological sciences. 2007;362:1437–1443.[Google Scholar]
- 34. Walker GA, Masters KS, Shah DN, Anseth KS, Leinwand LAValvular myofibroblast activation by transforming growth factor-beta: Implications for pathological extracellular matrix remodeling in heart valve disease. Circ Res. 2004;95:253–260.[PubMed][Google Scholar]
- 35. Rabkin E, Aikawa M, Stone JR, Fukumoto Y, Libby P, Schoen FJActivated interstitial myofibroblasts express catabolic enzymes and mediate matrix remodeling in myxomatous heart valves. Circulation. 2001;104:2525–2532.[PubMed][Google Scholar]
- 36. Rabkin-Aikawa E, Farber M, Aikawa M, Schoen FJDynamic and reversible changes of interstitial cell phenotype during remodeling of cardiac valves. J Heart Valve Dis. 2004;13:841–847.[PubMed][Google Scholar]
- 37. Cushing MC, Liao JT, Anseth KSActivation of valvular interstitial cells is mediated by transforming growth factor-beta1 interactions with matrix molecules. Matrix biology: journal of the International Society for Matrix Biology. 2005;24:428–437.[PubMed][Google Scholar]
- 38. Hinton RB, Jr., Lincoln J, Deutsch GH, Osinska H, Manning PB, Benson DW, Yutzey KEExtracellular matrix remodeling and organization in developing and diseased aortic valves. Circ Res. 2006;98:1431–1438.[PubMed][Google Scholar]
- 39. Wang H, Tibbitt MW, Langer SJ, Leinwand LA, Anseth KSHydrogels preserve native phenotypes of valvular fibroblasts through an elasticity-regulated pi3k/akt pathway. Proc Natl Acad Sci U S A. 2013;110:19336–19341.[Google Scholar]
- 40. Visconti RP, Ebihara Y, LaRue AC, Fleming PA, McQuinn TC, Masuya M, Minamiguchi H, Markwald RR, Ogawa M, Drake CJAn in vivo analysis of hematopoietic stem cell potential: Hematopoietic origin of cardiac valve interstitial cells. Circ Res. 2006;98:690–696.[PubMed][Google Scholar]