Gelatin-based Hydrogel Degradation and Tissue Interaction in vivo: Insights from Multimodal Preclinical Imaging in Immunocompetent Nude Mice

Abstract

Hydrogels based on gelatin have evolved as promising multifunctional biomaterials. Gelatin is crosslinked with lysine diisocyanate ethyl ester (LDI) and the molar ratio of gelatin and LDI in the starting material mixture determines elastic properties of the resulting hydrogel. In order to investigate the clinical potential of these biopolymers, hydrogels with different ratios of gelatin and diisocyanate (3-fold (G10_LNCO3) and 8-fold (G10_LNCO8) molar excess of isocyanate groups) were subcutaneously implanted in mice (uni- or bilateral implantation). Degradation and biomaterial-tissue-interaction were investigated in vivo (MRI, optical imaging, PET) and ex vivo (autoradiography, histology, serum analysis). Multimodal imaging revealed that the number of covalent net points correlates well with degradation time, which allows for targeted modification of hydrogels based on properties of the tissue to be replaced. Importantly, the degradation time was also dependent on the number of implants per animal. Despite local mechanisms of tissue remodeling no adverse tissue responses could be observed neither locally nor systemically. Finally, this preclinical investigation in immunocompetent mice clearly demonstrated a complete restoration of the original healthy tissue.

Keywords: Autoradiography ex vivo; Biomaterials; Computed tomography; Magnetic resonance imaging; Optical imaging; Positron emission tomography..

Figure 1

Hydrogel implantation and visualization. (a) Synthesis scheme of gelatin-based hydrogels G10_LNCO. Gelatin was reacted with 3- or 8-fold excess of L-lysine diisocyanate ethyl ester (LDI) related to amine-groups in gelatin for G10_LNCO3 and G10_LNCO8 (upper left panel). Confocal microscopy images reveal different surface structures of the different hydrogels (upper right panel). Implantation workflow is shown in the lower panel. First mice skin is disinfected, second incision and skin pocket is formed, third pre-swollen hydrogel is implanted, fourth incision is closed by the use of spray-plaster. (b) Representative axial MRI images on day 1 and day 35 after implantation. After drawing a volume of interest around the material (red sphere) and applying a threshold (middle panel) the volume of the hydrogels could be calculated. Hematoxylin & Eosin (H&E) staining confirmed difference in degradation of G10_LNCO3 and G10_LNCO8 35 days after implantation (right panel).

Figure 2

Quantification of hydrogel degradation and local MMP-activity. Degradation behavior of G10_LNCO3 and G10_LNCO8 could be observed noninvasively using dedicated small animal MRI with a specialized T2 measuring sequence. Volume was quantified after applying a threshold. Volumes are given as % of initial volume. Degradation behavior of G10_LNCO3 and G10_LNCO8 in (a) single-implanted (n=6-7) and (b) double-implanted animals (n=10). In vivo quantification of MMP activity for (c) single-implanted and (d) double-implanted animals. Mean + s.d. * G10_LNCO3 vs. G10_LNCO8, ^# G10_LNCO3 vs. Sham, ° G10_LNCO8 vs. Sham, * p < 0.05, ** p < 0.01.

Figure 3

[¹⁸F]FDG PET imaging and quantification. (a) Maximum intensity projection, coronal (0-60 min p.i., upper panel) and transversal projection (mid panel, 30-60 min p.i.) of dynamic PET experiments with [¹⁸F]FDG coregistered to MRI projection (lower panel) 14 days after implantation of either G10_LNCO3 (left panel) or G10_LNCO8 (mid panel) as well as of both G10_LNCO3 and G10_LNCO8 (right panel). (b) Mean SUV of [¹⁸F]FDG (30‑60 min p.i.) on day 14, 21, and 35 after hydrogel implantation. (c) Autoradiography of double-implanted mice 35 days after implantation (the upper panel). Transversal section is shown in the lower panel. Left side shows autoradiography. Whole animal cryo-sections are shown on the right side. Mean + s.d. * p < 0.05, ** p < 0.01.

Figure 4

Histological quantification. Histological quantification of positive-stained area compared to counterstaining of cell nuclei of single-implanted animals. Time course for (a) CD68 as pan-macrophage marker, (b) CD206 as M2-macrophage marker, (c) Ki67 for proliferation, (d) S100A4 as fibroblast marker, (e) CD31 as blood vessel marker, and (f) VEGFR-2 for new born blood vessels (n=3 different animals). Quantification was applied using color thresholds at mosaic images of whole, centric 10 µm slices of the hydrogel implant and the surrounding tissue. Mean + s.d. * G10_LNCO3 vs. G10_LNCO8, ^# G10_LNCO3 vs. Sham, ° G10_LNCO8 vs. Sham p < 0.1.

Figure 5

Angiogenesis at the biomaterial-tissue-interface. Representative histological images of angiogenic blood vessels (VEGFR-2) and all blood vessels (CD31) are shown at the first and last time point of detection around the hydrogels (a) 7 days (left) and 21 days (right) after implantation for G10_LNCO3 and (b) 14 days (left) and 35 days (right) after implantation for G10_LNCO8.

Figure 6

Serum MMP-levels. Time course of (a) MMP-2, (b) MMP-3, (c) MMP-8, (d) proMMP‑9, and (e) MMP-12 serum concentrations of single-implanted mice (n=4 different animals). Mean + s.d.