Tissue-specific granular extracellular matrix biomaterials drive differential responses in human adipose-derived mesenchymal stem cells

Introduction: : Articular cartilage lacks capacity for repair and focal defects can initiate matrix degradation in surrounding cartilage and bone, ultimately predisposing the joint to osteoarthritis (OA). OA is a debilitating disease afflicting millions[1]. Unfortunately, current treatments are limited by the availability of donor tissue and results in fibrocartilage that is mechanically inferior to native tissue[2]. There is a critical need for integrative cartilage repair strategies that restore the native organization and structure-function relationships of osteochondral tissues. Thus, we developed a novel biomaterial that incorporates granulated extracellular matrix (gECM) into a thiolated hyaluronic acid (HA) hydrogel. gECM particles are derived from human-sourced cartilage or bone, otherwise discarded or nonviable, to provide tissue-specific microenvironments and signaling. Our biomaterial is extrudable to conform to the size and shape of the defect and under physiological conditions, gECM crosslinks to available thiol groups on the HA hydrogel[3,4]. The mechanical properties of the gECM hydrogel are governed by the high packing density of gECM[4].
We recently demonstrated efficacy of porcine derived gECM hydrogels in large animal (goat) studies3 and are now poised for clinical translation. To meet regulatory approval requirements, we produced gECM hydrogels under current good manufacturing practice (cGMP) specifications and aim to demonstrate pre-clinical evidence of positive cellular responses. Thus, the objective of our study was to characterize the integration of human adipose-derived mesenchymal stem cells (hAD-MSCs) within gECM hydrogels. We assessed cell viability, mechanical properties, and gene expression of key differentiation markers over time to understand how these tissue-specific environments influence early MSC behavior.

Materials and

Methods: : gECM hydrogel fabrication: Human articular cartilage and bone tissue from 4 donors (AlloSource) were decellularized, lyophilized, pulverized and size-sorted based on particle diameter (cartilage < 250µm, bone 40-250µm) under cGMP conditions. 3D hydrogels were formed by mixing the gECM with thiolated HA (1% w/v) and hAD-MSCs. Cell Culture: We expanded hAD-MSCs, from 1 donor (AlloSource), on TCP to passage 4 (P4) and collected RNA from a portion as controls for qPCR. We encapsulated remaining hAD-MSCs into the composite hydrogel at 7.5x10^5cells/construct and cultured in osteochondral media (DMEM, 3% FBS, 1% ITS+Premix, 1X sodium pyruvate, 1X Pen/Strep, 50µg/mL ascorbic acid, 10mM β-glycerophosphate,100nM dexamethasone, 2.5µg/mL TGFβ3). We harvested constructs on days in vitro (DIV) 3, 14, and 21. Microscopy: We stained constructs with Calcein AM and ethidium homodimer for cell viability, fixed them with 4% PFA and labeled nuclei with DAPI. We acquired images using a 10x objective on a Nikon AIR confocal. Surface Area: We imaged constructs with a 0.75x objective on a Leica m80 and measured surface area using Otsu thresholding in ImageJ. Compressive Modulus: We performed unconfined, uniaxial compression testing using a Bose ElectroForce 5500. We compressed samples to 40% strain at 0.1%/sec and calculated the modulus from the linear portion of the stress-strain curve between 30%-40% strain. Gene expression: We isolated RNA from constructs, reverse transcribed to cDNA and preformed RT-qPCR. Data were normalized to GAPDH, and fold changes were calculated relative to P4 controls using the ∆∆Ct method.

Results, Conclusions, and Discussions:: Human AD-MSCs remained viable on cartilage- and bone-derived biomaterials over 21 days, forming confluent layers on the top and bottom surfaces and populating the full construct depth (Fig1A,E). Interestingly, cartilage and bone constructs showed distinct macroscopic changes over time. By DIV3, bone constructs developed a central concavity, and by DIV14, some samples completely folded inwards (Fig A). In contrast, cartilage constructs remained flat with minimal cratering (Fig E). These observations prompted us to quantify surface area, which revealed that both tissue constructs progressively contracted (Fig1B,F). Additionally, both biomaterials increased in stiffness over time (Fig1 C,G). By DIV21, the compressive moduli of bone constructs were 10x greater than cartilage constructs. Lastly, we found that hAD-MSCs on bone-derived constructs progressively upregulated COL1A2, COLX, and BGLAP while downregulating ACAN and SOX9 when compared to P4 hAD-MSCs prior to encapsulation (Fig 1D). Alternatively, hAD-MSCs on cartilage-derived constructs initially upregulated ACAN and SOX9 at DIV3 and DIV14 and expression decreased by DIV21 (Fig 1H).
Our results indicate that hAD-MSCs encapsulated in bone and cartilage gECM have tissue-specific responses. The maintained cell viability indicates effective nutrient diffusion throughout the constructs. Despite the densely packed ECM particles, there was adequate space for cell spreading and proliferation. The progressive contraction and stiffening imply cell-generated forces and newly deposited ECM. We observed the greatest differences in surface area and stiffness between DIV3 and 14, after which increased stiffness likely restricted further contraction driven by cells and ECM tension. The bone-derived constructs exhibited enhanced contraction and stiffening compared to the cartilage constructs. hAD-MSCs also exhibited differential gene expression with bone constructs upregulating markers of osteogenesis (COLX, COL1A2, BGLAP) while the cartilage constructs initially upregulated markers of chondrogenesis (ACAN, SOX). These differences are likely driven by tissue-specific signals from the gECM considering that both constructs received the same biochemical factors from the media. The change in gene expression with time may be influenced by newly synthesized ECM and cell-cell interactions rather than from the gECM. Future work will further define hAD-MSC differentiation by RNA-sequencing and histological analysis with additional cell and tissue donors.

Acknowledgements and/or References (Optional):: Acknowledgements: Bone, cartilage, and hAD-MSC donor tissue and cells came from AlloSource. AlloSource also provided cGMP facilities for gECM hydrogel fabrication. This work was supported by funding from NIH and Colorado Program for Musculoskeletal Research training grant T32.

References: [1] Deshpande B., et al., Arthritis Care Res, 2016. [2] Wegener, B., et al., Biotechnology and applied biochemistry, 2009. [3] Barthold, J., et al., Biofabrication, 2022. [4] Barthold, J., et al., Advanced Functional Materials, 2021.