VEGF-A regulates angiogenesis during osseointegration of Ti implants via paracrine/autocrine regulation of osteoblast response to hierarchical microstructure of the surface

Abstract

Establishment of a patent vasculature at the bone-implant interface plays a significant role in determining overall success of orthopedic and dental implants. Osteoblasts produce vascular endothelial growth factor-A (VEGF-A), an important regulator of angiogenesis during bone formation and healing, and the amount secreted is sensitive to titanium (Ti) surface microtopography and surface energy. The purpose of this study was to determine if surface properties modulate cellular response to VEGF-A. MG63 osteoblast-like cells were transfected with shRNA targeting VEGF-A at >80% knockdown. Cells stably silenced for VEGF-A secreted reduced levels of osteocalcin, osteoprotegerin, FGF-2, and angiopoietin-1 when cultured on grit-blasted/acid-etched (SLA) and hydrophilic SLA (modSLA) Ti surfaces and conditioned media from these cultures caused reduced angiogenesis in an endothelial tubule formation assay. Treatment of MG63 cells with 20 ng/mL rhVEGF-A₁₆₅ rescued production in silenced cells and increased production of osteocalcin, osteoprotegerin, FGF-2, and angiopoietin-1, with greatest effects on control cells cultured on modSLA. Addition of a neutralization antibody against VEGF receptor 2 (VEGFR2; Flk-1) resulted in a significant increase in VEGF-A production. Overall, this study indicates that VEGF-A has two roles in osseointegration: enhanced angiogenesis and an autocrine/paracrine role in maturation of osteoblast-like cells in response to Ti surface properties. © 2018 Wiley Periodicals, Inc. J Biomed Mater Res Part A: 107A: 423-433, 2019.

Keywords: VEGF; angiogenesis; osseointegration; osteoblast; titanium.

Figure 1.

Verification of VEGF-A silencing in an MG63 cell line using lentiviral-mediated transduction of shRNA specific for human VEGF-A. (A) VEGF-A protein levels and (B) VEGF-A mRNA levels were determined in 5 separate MG63 cell lines established using shRNA sequences specific for VEGF-A as well as a scrambled shRNA sequence (NT) and empty vector control (PLK0.1). *p<0.05 vs. MG63. (C) Total endothelial tube length in the presence of conditioned media from MG63 cells and VEGF-A silenced clone H4 cells. *p<0.05 vs. 4h; #p<0.05 vs. MG63.

Figure 2.

Paracrine effect of angiogenic growth factor production by MG63 and VEGF-A silenced MG63 cells. (A) VEGF-A, (B) FGF-2, and (C) Angiopoietin-1 levels in the conditioned media of MG63 and VEGF-A silenced MG63 cells were determined. Cells were cultured on control (TCPS), PT, SLA, and modSLA Ti surfaces. The ability of angiogenic growth factors to promote endothelial cell tubule formation was assessed using a fibrin gel assay. Total endothelial tube length was measured at (D) 12, (E) 24, and (F) 36 hours after the addition of conditioned media. Representative images using conditioned media from (G) siVEGF-A and (H) MG63 cell cultures. Values presented are mean ± SEM of six independent cultures. The data presented are from one of two separate experiments with comparable results. Data were analyzed using ANOVA, and statistical significance between groups was determined using Bonferroni’s modification of Student’s t-test. *p< 0.05 vs. TCPS; #p<0.05 vs. MG63.

Figure 3.

MG63 and VEGF-A silenced cell response to treatment with exogenous rhVEGF-A₁₆₅. (A) Cell number, (B) alkaline phosphatase specific activity in the cell lysate, (C) osteocalcin, (D) osteoprotegerin, (E) FGF-2, and (F) Angiopoietin-1 levels were determined for both MG63, and VEGF-A silenced MG63 cells as well as MG63 and VEGF-A silenced MG63 cells treated with 20 ng/mL of rhVEGF-A₁₆₅. Values presented are mean ± SEM of six independent cultures. Data were analyzed using ANOVA, and statistical significance between groups was determined using Bonferroni’s modification of Student’s t-test. *p<0.05 vs. TCPS; ^p<0.05 vs. MG63; #p<0.05 vs. no rhVEGF-A treatment.

Figure 4.

The response of MG63 cells cultured on TCPS and Ti surfaces in the presence of an Flk-1 neutralizing antibody. (A) Cell Number, (B) Alkaline phosphatase specific activity in the cell lysate, (C) osteocalcin, (D) osteoprotegerin, (E) VEGF-A, (F) FGF-2, and (G) Angiopoietin-1 levels in the conditioned media of MG63 cells cultured in the presence of 100 ng/mL of an Flk-1 or control IgG antibody were determined. Cells were cultured on control (TCPS), PT, SLA, and modSLA Ti surfaces. Values presented are mean ± SEM of six independent cultures. The data presented are from one of two separate experiments with comparable results. Data were analyzed using ANOVA, and statistical significance between groups was determined using Bonferroni’s modification of Student’s t-test. *p< 0.05 vs. TCPS.

Figure 5.

MG63 and VEGF-A silenced cell response to treatment with exogenous rhFGF-2. (A) Cell number, (B) alkaline phosphatase specific activity in the cell lysate, (C) osteocalcin, (D) osteoprotegerin, (E) VEGF-A, and (F) Angiopoietin-1 levels were determined for both MG63, and VEGF-A silenced MG63 cells as well as MG63 and VEGF-A silenced MG63 cells treated with 10 ng/mL of rhFGF-2. Values presented are mean ± SEM of six independent cultures. Data were analyzed using ANOVA, and statistical significance between groups was determined using Bonferroni’s modification of Student’s t-test. *p<0.05 vs. TCPS; ^p<0.05 vs. MG63; #p<0.05 vs. no rhFGF-2 treatment.