Smooth muscle cells of human veins show an increased response to injury at valve sites

Abstract

Objective: Venous valves are essential but are prone to injury, thrombosis, and fibrosis. We compared the behavior and gene expression of smooth muscle cells (SMCs) in the valve sinus vs nonvalve sites to elucidate biologic differences associated with vein valves.

Methods: Tissue explants of fresh human saphenous veins were prepared, and the migration of SMCs from explants of valve sinus vs nonvalve sinus areas was measured. Proliferation and death of SMCs were determined by staining for Ki67 and terminal deoxynucleotidyl transferase dUTP nick end labeling. Proliferation and migration of passaged valve vs nonvalve SMCs were determined by cell counts and using microchemotaxis chambers. Global gene expression in valve vs nonvalve intima-media was determined by RNA sequencing.

Results: Valve SMCs demonstrated greater proliferation in tissue explants compared with nonvalve SMCs (19.3% ± 5.4% vs 6.8% ± 2.0% Ki67-positive nuclei at 4 days, respectively; mean ± standard error of the mean, five veins; P < .05). This was also true for migration (18.2 ± 2.7 vs 7.5 ± 3.0 migrated SMCs/explant at 6 days, respectively; 24 veins, 15 explants/vein; P < .0001). Cell death was not different (39.6% ± 16.1% vs 41.5% ± 16.0% terminal deoxynucleotidyl transferase dUTP nick end labeling-positive cells, respectively, at 4 days, five veins). Cultured valve SMCs also proliferated faster than nonvalve SMCs in response to platelet-derived growth factor subunit BB (2.9 ± 0.2-fold vs 2.1 ± 0.2-fold of control, respectively; P < .001; n = 5 pairs of cells). This was also true for migration (6.5 ± 1.2-fold vs 4.4 ± 0.8-fold of control, respectively; P < .001; n = 7 pairs of cells). Blockade of fibroblast growth factor 2 (FGF2) inhibited the increased responses of valve SMCs but had no effect on nonvalve SMCs. Exogenous FGF2 increased migration of valve but not of nonvalve SMCs. Unlike in the isolated, cultured cells, blockade of FGF2 in the tissue explants did not block migration of valve or nonvalve SMCs from the explants. Thirty-seven genes were differentially expressed by valve compared with nonvalve intimal-medial tissue (11 veins). Peptide-mediated inhibition of SEMA3A, one of the differentially expressed genes, increased the number of migrated SMCs of valve but not of nonvalve explants.

Conclusions: Valve compared with nonvalve SMCs have greater rates of migration and proliferation, which may in part explain the propensity for pathologic lesion formation in valves. Whereas FGF2 mediates these effects in cultured SMCs, the mediators of these stimulatory effects in the valve wall tissue remain unclear but may be among the differentially expressed genes discovered in this study. One of these genes, SEMA3A, mediates a valve-specific inhibitory effect on the injury response of valve SMCs.

Figure 1

Valve SMC from intimal/medial explants migrate more than non-valve SMCs. Cell migration from the valve sinus (hereafter just called valve) compared to the non-valve intimal/medial explants (A) or adventitial explants (B). Migration is presented in the left hand panels as the percent of migration-positive explants (≥ 1 cell/explant, which measures only migration) and in the right hand panels as the number of cells per explant (which measures a combination of migration and post-migration proliferation). * P<.001 paired valve vs non-valve; N=24 and 16 different veins for A and B, respectively.

Figure 2

Valve SMCs proliferate more than non-valve SMCs in response to injury. Cell density (A; cells/mm²), death (B; % TUNEL positive cells), and proliferation (C; % Ki67 positive nuclei) in intima/media (left hand panels) vs adventitia (right hand panels) of valve and non-valve wall in response to tissue dissection and culture over 4 days. *P<.05 paired valve vs non-valve; N=4 veins.

Figure 3

Valve and non-valve SMCs have equivalent expression of SMC markers. Representative immunofluorescent staining for (A) smooth muscle myosin heavy chain (SM-Myosin) and (B) smooth muscle α-actin (SM- α-actin) in valve and non-valve SMCs was quantified for 5 pairs of valve and non-valve SMCs (C). (D) Levels of SMC marker mRNA in valve and non-valve SMCs. Smooth muscle α-actin (ACTA2), smooth muscle myosin heavy chain (MYH11), caldesmon-1 (CALD1), and NG2 (CSPG4). N=4 pairs of valve and non-valve SMCs.

Figure 4

Valve SMCs show increased chemotactic migration to PDGF-BB and increased chemokinetic migration to serum. Valve and non-valve SMC migration were measured using a microchemotaxis chamber assay. (A) Migration in response to PDGF-BB or 10% serum. *P<.01 valve vs non-valve; results are from paired cells from 7 veins. (B and C) Valve and non-valve SMC migration were determined in microchemotaxis chambers with either 10 ng/ml PDGF-BB (B) or 10% serum (C) in only the bottom chamber or in both top and bottom chambers. * P<.01 top only vs top & bottom; N=5 experiments. (D) Attachment of SMCs to polymeric collagen measured using the crystal violet assay. N=4 experiments.

Figure 5

Valve SMC proliferation is stimulated differentially by PDGF-BB. Proliferation of valve vs non-valve SMCs in response to PDGF-BB + 2% serum (A), 1–10% serum (B) or a commercial SMC growth medium (C). Results are presented as the mean ratio (± SEM) of day 4 to day 1 cell counts or as fold of control. *P<.05 valve vs non-valve; N=5, 6, or 9 pairs of cells for A, B, and C, respectively.

Figure 6

FGF2 antibody blocks increased PDGF-BB induced valve SMC migration and proliferation. The effect of a blocking antibody to FGF2 on valve vs non-valve SMC migration (microchemotaxis chamber) (A) and proliferation (B) mediated by 10 ng/ml PDGF-BB. Results are presented as fold of control in A and as the ratio (± SEM) of day 4 to day 1 cell counts in B. *P<.05 antiFGF2 vs IgG control; N= 5–7 pairs of cells. (C) The effect of exogenous FGF2 (10 ng/ml) on migration of valve and non-valve SMCs in a chemotaxis chamber assay. * P=.003 FGF2 vs unstimulated control; N= 5 different pairs of valve and non-valve SMCs.

Figure 7

Migration from explants is not affected by anti-FGF2, and PDGFR blockade is more effective in non-valve SMC. The effect of the blocking antibody to FGF2 (or goat IgG; both at 60 ug/ml) on SMC migration from valve (A) and non-valve (B) intimal/medial tissue (N= 5 veins). The effect of a combination of antibodies to PDGFRα and β (5 ug/ml each or 10 ug/ml mouse IgG) on SMC migration from valve (C) and non-valve (D) intimal/medial tissue (*P<.01 vs IgG; N=4 veins). Migration is presented as mean migration positive explants (≥1 cell migrating from explant).

Figure 8

SEMApeptide selectively increases valve SMC migration from explants. (A) Schematic of NRP-1 structure with the blocking peptides used in this experiment. (B) The effect of the SEMApeptide, MAMpeptide, or scrambled peptide (SCRpeptide) each at 20 uM on SMC migration/proliferation (cell number/explant) from valve (left panel) and non-valve (right panel) intimal/medial tissue (*P<.05 vs SCRpeptide; N= 5 veins). (C) The effect of the SEMApeptide, MAMpeptide, or scrambled peptide (SCRpeptide) on the % migration positive explants (≥ 1 cell migrating from explant) from valve (left panel) and non-valve (right panel) intimal/medial (N= 5 veins).