The Influence of Platelet-Derived Growth Factor and Bone Morphogenetic Protein Presentation on Tubule Organization by Human Umbilical Vascular Endothelial Cells and Human Mesenchymal Stem Cells in Coculture

Abstract

A three-dimensional in vitro Matrigel plug was used as a model to explore delivery patterns of platelet-derived growth factor (PDGF) and bone morphogenetic protein-2 (BMP-2) to a coculture of human mesenchymal and endothelial cells. While BMP-2 is well recognized for its role in promoting fracture healing through proliferation and differentiation of osteoclast precursors, it is not a growth factor known to promote the process of angiogenesis, which is also critical for complete bone tissue repair. PDGF, in contrast, is a known regulator of angiogenesis, and also a powerful chemoattractant for osteoblast precursor cells. It has been suggested that presentation of PDGF followed by BMP may better promote vascularized bone tissue formation. Yet, it is unclear as to how cells would respond to various durations of delivery of each growth factor as well as to various amounts of overlap in presentation in terms of angiogenesis. Using a three-dimensional in vitro Matrigel plug model, we observed how various presentation schedules of PDGF and BMP-2 influenced tubule formation by human mesenchymal stem cells and human umbilical vascular endothelial cells. We observed that sequential presentation of PDGF to BMP-2 led to increased tubule formation over simultaneous delivery of these growth factors. Importantly, a 2-4 day overlap in the sequential presentation of PDGF and BMP-2 increased tubule formation as compared with groups with zero or complete growth factor overlap, suggesting that a moderate amount of angiogenic and osteogenic growth factor overlap may be beneficial for processes associated with angiogenesis.

Keywords: 3D cell culture; angiogenesis; endothelial cells; growth factors; mesenchymal stem cells.

FIG. 1.

Sequential growth factor delivery regimens showing type of growth factor delivery per day over 10 days. Growth factors were delivered by pipetting 20 μL of medium with concentrations of 10 ng/mL PDGF and/or 100 ng/mL BMP. BMP, bone morphogenetic protein; PDGF, platelet-derived growth factor.

FIG. 2.

Sequential delivery of PDGF to BMP and delivery of PDGF alone results in organized tubule formation in vitro. (A) CD31 staining of growth factor delivery groups, and (B) software analysis overlay of the same images, showing quantification methods for tubule area, tubule length, branching points, and loops. All scale bars = 50 μm.

FIG. 3.

Quantification of tubule area, tubule length, branches, and loops for sequential growth factor delivery groups showing that PDGF delivery alone (positive control) results in significantly greater tubule formation when compared with all other treatment groups at each parameter. Sequential PDGF to BMP delivery results in significantly greater tubule area coverage, tubule length, and branching points when compared with several other treatment groups. ^#Indicates significant difference from PDGF to BMP treatment group, *indicates significant difference from PDGF treatment group. ANOVA followed by post hoc multiple comparisons testing using Tukey's test. α = 0.05. ANOVA, analysis of variance.

FIG. 4.

Immunofluorescent staining of CD31 (red), α-SMA (green), and nuclei (blue) of Matrigel cross sections suggests that longer PDGF delivery before switch to BMP (7 days, 9 days) results in greater amounts of tubule formation and presence of pericytes * indicates significantly more area coverage versus days 1, 3, and 5. ^#Indicates significantly more area coverage versus days 1, 3 when compared using ANOVA and Tukey's test for post hoc comparisons, α = 0.05. All scale bars = 50 μm.

FIG. 5.

Overlapping growth factor delivery regimens showing type of growth factor delivery per day over 10 days. Growth factors were delivered by pipetting 20 μL of medium with concentrations of 10 ng/mL PDGF and/or 100 ng/mL BMP.

FIG. 6.

(A) Moderate overlap (2–4 days) in PDGF and BMP delivery results in greater tubule formation (CD31 staining). All scale bars = 50 μm. (B) Tubule quantification (area, length, branches, loops) of groups represented in Figure 6A showing that 4 days of overlap in growth factor delivery results in significantly greater tubule formation when compared with all other groups for every measured parameter. Additionally, 2 days of overlap in growth factor delivery results in significantly greater tubule area coverage and length versus 8 days and 10 days of overlap. *Indicates significant difference from 4 days overlap, ^#indicates significant difference from 2 days overlap when compared using ANOVA followed by post hoc multiple comparisons with Tukey's test. α = 0.05.