Cyclic tensile culture promotes fibroblastic differentiation of marrow stromal cells encapsulated in poly(ethylene glycol)-based hydrogels

Abstract

To inform future efforts in tendon/ligament tissue engineering, our laboratory has developed a well-controlled model system with the ability to alter both external tensile loading parameters and local biochemical cues to better understand marrow stromal cell differentiation in response to both stimuli concurrently. In particular, the synthetic, poly(ethylene glycol)-based hydrogel material oligo(poly(ethylene glycol) fumarate) (OPF) has been explored as a cell carrier for this system. This biomaterial can be tailored to present covalently incorporated bioactive moieties and can be loaded in our custom cyclic tensile bioreactor for up to 28 days with no loss of material integrity. Human marrow stromal cells encapsulated in these OPF hydrogels were cultured (21 days) under cyclic tensile strain (10%, 1 Hz, 3 h of strain followed by 3 h without) or at 0% strain. No difference was observed in cell number due to mechanical stimulation or across time (n = 4), with cells remaining viable (n = 4) through 21 days. Cyclic strain significantly upregulated all tendon/ligament fibroblastic genes examined (collagen I, collagen III, and tenascin-C) by day 21 (n ≥ 6), whereas genes for other pathways (osteogenic, chondrogenic, and adipogenic) did not increase. After 21 days, the presence of collagen I and tenascin-C was observed via immunostaining (n = 2). This study demonstrates the utility of this hydrogel/bioreactor system as a versatile, yet well-controlled, model environment to study marrow stromal cell differentiation toward the tendon/ligament phenotype under a variety of conditions.

FIG. 1.

(A) Custom tensile culture system. Up to 24 biomaterial constructs can be cultured in tensile wells (dotted-line arrows) and strained by the tensile rakes (double-head arrow shows direction of tensile strain) at the same time. The rake is moved by a linear motor with positional accuracy monitored by an optical encoder. Inset: individual construct with hydrogel (bracket) flanked by end blocks (arrowheads). (B) Laminated hydrogel constructs in culture wells. Polyethylene end blocks (arrowheads) are interfaced with the tensile rake to allow mechanical stimulation to be transduced to the hydrogel section (bracket) and create a uniform strain field in the construct during culture. Double-head arrow indicates direction of strain. (C) Diagram demonstrating possibilities for future use of this culture system with laminated hydrogels. For example, this permits coculture of two different cell types (top) or two biomaterial environments with different bioactive factors (bottom). Double-head arrow represents direction of strain. (D) Overall experimental design for human marrow stromal cell loading studies. Arrows represent direction of loading on dynamic samples. Color images available online at www.liebertonline.com/ten.

FIG. 2.

OPF-3K, OPF-10K, and laminated hydrogel constructs all maintained structural integrity over 28 days of cyclic tensile culture with minimal degradation (n ≥ 3 ± standard deviation). Fold swelling was normalized to day 1 for each construct type. ^#Significance difference for laminated constructs versus day 1 (p < 0.05). OPF-3K, oligo(poly(ethylene glycol) fumarate) with a 3 kDa poly(ethylene glycol) chain; OPF-10K, oligo(poly(ethylene glycol) fumarate) with a 10 kDa poly(ethylene glycol) chain.

FIG. 3.

A majority of live (green) cells were observed in both cyclically (A) and statically (B) cultured OPF hydrogels over 21 days (n = 4). No significant difference in cell number (C) was found between static conditions and cyclic strain at any time point (n = 4 ± standard deviation). *Significance (p ≤ 0.05). Scale bar = 100 μm. Color images available online at www.liebertonline.com/ten.

FIG. 4.

Collagen I (A), collagen III (B), and tenascin-C (C), markers for tendon/ligament fibroblastic differentiation, were upregulated in human marrow stromal cells encapsulated in constructs under cyclic tensile strain compared to static constructs by day 21 (n ≥ 6 ± standard deviation). Expression normalized to glyceraldehyde phosphate dehydrogenase was then normalized to day 1 static results for all genes (note: y-axis is different for each graph). *Significance over static constructs at same time point (p ≤ 0.05). ⁺Significance over day 1 for the same sample type.

FIG. 4.

Collagen I (A), collagen III (B), and tenascin-C (C), markers for tendon/ligament fibroblastic differentiation, were upregulated in human marrow stromal cells encapsulated in constructs under cyclic tensile strain compared to static constructs by day 21 (n ≥ 6 ± standard deviation). Expression normalized to glyceraldehyde phosphate dehydrogenase was then normalized to day 1 static results for all genes (note: y-axis is different for each graph). *Significance over static constructs at same time point (p ≤ 0.05). ⁺Significance over day 1 for the same sample type.

FIG. 5.

Immunohistochemistry (n = 2) showing the presence of collagen I (A, D) and tenascin-C (B, E) after 21 days of culture in both static and cyclic constructs. Collagen I and tenascin-C were detected mainly pericellularly (arrows). No staining was evident in control sections with the primary antibody omitted (C). Scale bar = 100 μm. Color images available online at www.liebertonline.com/ten.