Modulating In Vivo Degradation Rate of Injectable Extracellular Matrix Hydrogels

Abstract

Extracellular matrix (ECM) derived hydrogels are increasingly used as scaffolds to stimulate endogenous repair. However, few studies have examined how altering the degradation rates of these materials affect cellular interaction in vivo. This study sought to examine how crosslinking or matrix metalloproteinase (MMP) inhibition by doxycycline could be employed to modulate the degradation rate of an injectable hydrogel derived from decellularized porcine ventricular myocardium. While both approaches were effective in reducing degradation in vitro, only doxycycline significantly prolonged hydrogel degradation in vivo without affecting material biocompatibility. In addition, unlike crosslinking, incorporation of doxycycline into the hydrogel did not affect mechanical properties. Lastly, the results of this study highlighted the need for development of novel crosslinkers for in situ modification of injectable ECM-derived hydrogels, as none of the crosslinking agents investigated in this study were both biocompatible and effective.

Keywords: MMP inhibition; crosslinking; degradation; extracellular matrix; hydrogel.

Figure 1. Cytotoxicity of crosslinked and doxycycline loaded hydrogels

Effect of elution products collected from myocardial matrix hydrogels crosslinked with 0.05% glutaraldehyde (GA), 1 mM genipin (Gen), 120 mU/mL transglutaminase (TG), and 1 mg/mL doxycycline (Dox) on L929 fibroblasts cell viability (measured by MTT assay) and cell number (measured by PicoGreen). Unmodified hydrogel and 0.5% GA crosslinking served as negative and positive controls, respectively ** p < 0.01 compared to unmodified hydrogels.

Figure 2. In vitro bacterial collagenase degradation

Degradation of unmodified, crosslinked, and doxycycline loaded hydrogels by bacterial collagenase were measured by amount of amines released in the supernatant through ninhydrin reactivity. Degradation in PBS alone, without addition of collagenase, is shown as negative control. GA: glutaraldehyde, Gen: genipin, TG: transglutaminase, Dox: doxycycline. * p < 0.05, ** p < 0.01 compared to unmodified hydrogel controls.

Figure 3. Doxycycline sequestration with the myocardial matrix hydrogel

A) Release profile of doxycycline from myocardial matrix hydrogels follows the typical burst release profile, but approximately 30% of loaded doxycycline is sequestered by hydrogel. B) Degradation of the same hydrogels after burst release has occurred showed that the sequestered doxycycline is still sufficient to inhibit bacterial collagenase degradation. * p < 0.05.

Figure 4. Rheological properties of crosslinked and doxycycline loaded hydrogels

Storage modulus (A) and loss modulus (B) of hydrogels measured by a parallel plate rheometer from a frequency sweep of 0.5 to 10 rad/s. At 0.4 rad/s, both 0.05% glutaraldehyde (GA) and 1 mM genipin (Gen) crosslinking significantly increased storage modulus (C) and loss modulus (D), while transglutaminase (TG) and doxycycline (Dox) had no effect. ** p < 0.01 compared to unmodified hydrogel controls.

Figure 5. Cell migration through crosslinked and doxycycline loaded hydrogels

A) Fluorescently labeled L929 migration through unmodified, 0.05% glutaldehyde (GA), 1 mM genipin (Gen), 120 mU/mL transglutaminase (TG), and 1 mg/mL doxycycline (Dox) hydrogels within transwell migration inserts over 24 hours (* p < 0.05 for one-way ANOVA). B) Migration through transglutaminase crosslinked hydrogels was significantly faster at earlier time points, hour 2 and hour 6, compared to unmodified matrices (* p < 0.05).

Figure 6. Biocompatibility of crosslinked and doxycycline loaded myocardial matrix in vivo

Representative H&E images of the myocardial matrix with 0.05% glutaraldehyde (A), 1 mM genipin (B), 120 mU/mL transglutaminase (C), or 1 mg/mL doxycycline (D) injected into healthy myocardium. Injection sites were made up of elongate spindle to branching cells suggestive of degenerative/regenerative myocardial fibers, plus usually minimal inflammatory infiltrate consisting of lymphocyte-type small mononuclear cells. Only the glutaraldehyde crosslinked hydrogel injection site showed marked lymphocytic infiltrate and foreign body giant cells (arrows). scale bars = 100 μm.

Figure 7. Doxycycline reduces myocardial matrix degradation in vivo

A) H&E image of a heart section showing the injection site. B) Alexa Fluor 568 (red) labeled myocardial matrix is visible fluorescently. C) Quantification of the amount of fluorescence remaining at one-week after injection. Gen: genipin, TG: transglutaminase, Dox: doxycycline; * p < 0.05 compared to unmodified hydrogel controls; scale bar = 1 mm