Gelatin methacrylate microspheres for controlled growth factor release

Abstract

Gelatin has been commonly used as a delivery vehicle for various biomolecules for tissue engineering and regenerative medicine applications due to its simple fabrication methods, inherent electrostatic binding properties, and proteolytic degradability. Compared to traditional chemical cross-linking methods, such as the use of glutaraldehyde (GA), methacrylate modification of gelatin offers an alternative method to better control the extent of hydrogel cross-linking. Here we examined the physical properties and growth factor delivery of gelatin methacrylate (GMA) microparticles (MPs) formulated with a wide range of different cross-linking densities (15-90%). Less methacrylated MPs had decreased elastic moduli and larger mesh sizes compared to GA MPs, with increasing methacrylation correlating to greater moduli and smaller mesh sizes. As expected, an inverse correlation between microparticle cross-linking density and degradation was observed, with the lowest cross-linked GMA MPs degrading at the fastest rate, comparable to GA MPs. Interestingly, GMA MPs at lower cross-linking densities could be loaded with up to a 10-fold higher relative amount of growth factor than conventional GA cross-linked MPs, despite the GA MPs having an order of magnitude greater gelatin content. Moreover, a reduced GMA cross-linking density resulted in more complete release of bone morphogenic protein 4 and basic fibroblast growth factor and accelerated release rate with collagenase treatment. These studies demonstrate that GMA MPs provide a more flexible platform for growth factor delivery by enhancing the relative binding capacity and permitting proteolytic degradation tunability, thereby offering a more potent controlled release system for growth factor delivery.

Keywords: Gelatin; Growth factor delivery; Methacrylate; Microsphere.

Figure 1. Gelatin methacrylate characterization

A) Schematic of methacrylate substitution of the primary amines of gelatin. B) H¹ NMR spectra was recorded for unsubstituted gelatin and GMA with 1:3, 2:3, and 1:1 mol methacrylic anhydride: mol unsubstituted amines on gelatin. The MA modification of lysine residues with increasing methacrylic anhydride addition can be confirmed by the continuous decrease in the lysine signal at δ = 2.9 ppm (y), and increase in the methacrylate vinyl group signal at δ =5.4 ppm and 5.7 ppm (x) and methyl group signal at δ=1.8 ppm. C,D) Degree of methacrylate substitution was also determined for the GMA formulations via a Fluorescamine assay normalized to unmodified gelatin determined as 0% substitution.

Figure 2. Gelatin microparticle morphology and sizing

A) i–iv) Phase and v–viii) scanning electron microscopy images of GMA and GA cross-linked MPs indicate round morphology and smooth surfaces. Ai–iv) Scale bar: 100 μm. Av–viii) Scale bar: 10 μm. B) Coulter Counter size analysis of MPs indicate that no differences were found between MP sizes despite different extents of MA substitution. GMA MPs can be fabricated with a greater proportion of smaller sized MPs and lower quantities of large sized MPs than can be achieved through GA formulations. *: denotes statistical significance, n≥3, p<0.05.

Figure 3. Microparticle gelatin composition and degradation

A) Gelatin content per volume of MPs was determined for each of the 4 gelatin MP formulations. The GA MPs had the highest gelatin content and for the GMA formulations, increasing the degree of methacrylation appeared to decrease the gelatin content. B) Degradation profiles were also determined via incubation of the MPs with collagenase. Increasing the methacrylation level resulted in an increase in MP degradation time, and GA MPs were comparable to the lowest methacrylated MP formulation in degradation profile. Degradation kinetics below 12 hours are expanded in C) for clearer visualization. The 15% MA and GA formulations were the most degraded by 12 hours, degrading completely by 48 hours. The higher methacrylated MPs were completely degraded by 96 hours. n≥3, *: denotes statistical significance with p<0.05 *: 90% MA vs. 15% MA and GA. ^: 15% MA vs. 50% MA. #: 50% MA vs. 15% MA and GA. +: 90% MA vs. 50% MA.

Figure 4. MP mechanical properties and mesh sizes

A) Elastic moduli were determined for MPs, which were conjugated to a glass coverslip and analyzed via AFM. The GA MPs and highest methacrylated MPs (90% MA) were significantly stiffer than the two lower methacrylated MPs (15% and 50%) by almost a magnitude. B) Swelling studies of MPs determined that all GMA MPs had a greater ability to swell in a hydrated environment than the GA MPs, with the 15% MA MPs swelling to a greater extent than all other MPs. C) Mesh sizes of the MPs were determined based on their elastic moduli and swelling ratios. The 15% and 50% MA MPs had larger mesh sizes than the GA MPs, as well as the highest methacrylated MPs (90% MA). E: elastic modulus. q: swelling ratio. ξ : mesh size. n = 15, *: denotes statistical significance with p<0.05.

Figure 5. Microparticle growth factor loading capacity

Gelatin MPs were incubated with six concentrations of A) BMP4 and B) bFGF to determine the growth factor binding capacity. A) GA MPs bound less BMP4 than all GMA formulations at the lower loading concentrations (50 and 100 ng/mg) and bound less than the two lowest GMA formulations at the highest loading concentrations (200 and 300 ng/mg). B) While GA MPs bound minute BMP4, even at the highest loading concentration, the lowest GMA MPs had a peak binding efficiency at 99.6%. C) The GA MPs also bound less BMP4 than all GMA formulations at the lowest and highest loading concentrations (50,100 ng/mg and 200,300 ng/mg). D) bFGF loading efficiency peaked at 100 ng/mL for all MP formulations. n≥3, symbols denote statistical significance with p<0.05. *: GA vs. all GMA MPs, #: GA vs. 15% MA and 50% MA, $: GA vs. 15% MA, &: 15% MA vs 90% MA, %: 50% MA vs 90% MA.

Figure 6. BMP4 and bFGF release from microparticles

The MPs were incubated with 10 ng/mg MPs of BMP4 (A–C) and 50 ng/mg MPs bFGF (D–F), and release kinetics was obtained over the course of 170 hours. A) Passive release of BMP4 was similar between all MP formulations. B) Greater release of BMP4 in 100 ng/mL collagenase was observed in the GA MPs compared to the 50% MA MPs at 60 hours. C) Larger BMP4 burst release at 3 hours was observed in the GA formulation compared to the higher methacrylated MPs with 1 μg/mL collagenase treatment. D) Less bFGF was released passively in the 50% MPs compared to the GA MPs after 25 hours. E) Greater bFGF was released from the GA MPs compared to the 50% MA MPs by 170 hours following treatment with 100 ng/mL collagenase. F) High collagenase treatment resulted in no observable differences in bFGF release between MP formulations. n≥3, symbols denote statistical significance with p<0.05. *:GA vs. 50% MA, %: GA vs 15% MA, $: GA vs 90% MA, #: 50% MA vs 15% MA, &: 50% MA vs 90% MA.

Figure 7. Time point specific collagenase-mediated BMP4 and bFGF release from MPs

A, C, E) BMP4 release at 3, 25, and 170 hours. Besides the initial release at 3 hours, GA MPs under protease treatment did not release any more BMP4 compared to passive release. On the other hand, the release of BMP4 from the higher methacrylated MPs, 50% MA and 90% MA, increased with collagenase treatment at all time points assayed, and this effect was sustained up to 170 hours. B, D, E) bFGF release at 3, 25, and 170 hours. No differences were observed between bFGF passive release and proteolytic-mediated release of GA MPs. However, an increase in bFGF released by the 50% MA MPs and 90% MA MPs were increased with collagenase treatment up to 25 hours and 170 hours, respectively. The three additional time points can be viewed in Supplemental Figure 2. *: denotes statistical significance, n≥3, p<0.05.