Collagen-Based Medical Device as a Stem Cell Carrier for Regenerative Medicine

Abstract

Maintenance of mesenchymal stem cells (MSCs) requires a tissue-specific microenvironment (i.e., niche), which is poorly represented by the typical plastic substrate used for two-dimensional growth of MSCs in a tissue culture flask. The objective of this study was to address the potential use of collagen-based medical devices (HEMOCOLLAGENE^®, Saint-Maur-des-Fossés, France) as mimetic niche for MSCs with the ability to preserve human MSC stemness in vitro. With a chemical composition similar to type I collagen, HEMOCOLLAGENE^® foam presented a porous and interconnected structure (>90%) and a relative low elastic modulus of around 60 kPa. Biological studies revealed an apparently inert microenvironment of HEMOCOLLAGENE^® foam, where 80% of cultured human MSCs remained viable, adopted a flattened morphology, and maintained their undifferentiated state with basal secretory activity. Thus, three-dimensional HEMOCOLLAGENE^® foams present an in vitro model that mimics the MSC niche with the capacity to support viable and quiescent MSCs within a low stiffness collagen I scaffold simulating Wharton's jelly. These results suggest that haemostatic foam may be a useful and versatile carrier for MSC transplantation for regenerative medicine applications.

Keywords: biocompatibility; medical device; paracrine activities; regenerative medicine; stem cell niche.

Figure 1

Composition of HEMOCOLLAGENE^®. Spectra obtained by Fourier transform infrared (FTIR) spectroscopy showing similarities of HEMOCOLLAGENE^® foam (red line) with type I collagen (black line).

Figure 2

Structure of HEMOCOLLAGENE^®. (A) Representative scanning electron microscopy image (SEM, scale bar indicates 1 mm); (B) Hematoxylin-Eosin-Saffron (HES) staining of paraffin-embedded HEMOCOLLAGENE^® cross-section (scale bar indicates 500 μm); (C,D) Pore distribution obtained by mercury intrusion porosimetry showing the mean radius on the incremental curve and the pore threshold on the cumulative curve.

Figure 3

Mechanical features of HEMOCOLLAGENE^®. Force-displacement responses obtained using indentation tests at 1 mm/min (n = 4) and 10 mm/min (n = 4). The Young’s modulus has been identified using the Hertz model. No significant difference was observed between the two different testing speeds.

Figure 4

Cell viability over the time of the study. (A,B): Histogram reflecting WST-1 assay and DNA quantification, respectively; and (C): Flow cytometry results obtained after Zombie^® labelling. Kinetic study performed after 4, 7, and 10 days of culture showing live and non-proliferating WJ-MSCs cultured within HEMOCOLLAGENE^® foam (n = 6, Mann and Whitney test).

Figure 5

Cell distribution and apoptosis after 10 days of culture. (A,B) Haematoxylin-eosin-saffron (HES) and Masson’s trichrome staining of paraffin-embedded cellularized HEMOCOLLAGENE^® cross-sections, respectively. HES staining showing nuclei in blue (yellow arrow) and collagen in orange. Masson’s trichrome staining, showing collagen in green and nuclei in brown (yellow arrow); and (C) cleaved caspase-3 immunohistochemistry showing few apoptotic cells (red arrow) within HEMOCOLLAGENE^®. (Scale bars indicate 40 μm).

Figure 6

Cell morphology after 10 days of culture. (A,B) Over and closer (rectangle) scanning electron microscopy views (scale bars indicate 100 and 50 μm, respectively), highlighting cell distribution within pores (white arrows); (C,D) Over and closer Imaris 3D views of cytoskeleton-labelled cells (scale bars indicate 30 and 10 μm, respectively), showing actin bundles and highly-elongated cell morphology.

Figure 7

Cell paracrine activity over the time of the study. (A,B) Histograms of IL-6 and IL-8 ELISA quantification, respectively, showing non-significant production of related proteins; (C–E) Histograms of IL6, CXCL8, and VEGFA qRT-PCR analysis, showing non-significant regulation of IL6 and CXCL8, and significant VEGFA down-regulation (n = 6, Mann and Whitney statistical test, p < 0.05 for four versus 10 days of culture).

Figure 8

Cell phenotype over the time of the study. (A–C) Histograms of NT5E (CD 73), THY1 (CD 90), and ENG (CD 105) qRT-PCR analysis, respectively, showing non-significant regulation of MSCs markers; (D) WJ-MSC migration on plastic; (E) WJ-MSC morphology after passage; and (F) flow cytometry results obtained after Zombie^® labelling of amplified cells (n = 6, Mann and Whitney statistical test).

Figure 9

Micro-indentation tests performed on HEMOCOLLAGENE^® foam. (a) Global view of the indentation test on a typical specimen (10 × 10 × 10 mm); (b) Zoom of the spherical indenter (radius = 0.75 mm).