Inclusion of Chondroitin Sulfate Into a Gelatin Hydrogel Shifts Local and Global Mechanical Behavior and Fibrochondrogenic Phenotype for Applications in Insertional Tissue Engineering

TL;DR

Adding chondroitin sulfate to a gelatin hydrogel changes its mechanical properties and cell behavior, making it better for tissue engineering applications like tendon-to-bone repair.

Contribution

The study introduces a new method of incorporating chondroitin sulfate into hydrogels to enhance both mechanical and biological performance for tissue engineering.

Findings

Incorporating chondroitin sulfate increases mesenchymal stem cell metabolic activity and osteo-tendinous differentiation.

CS inclusion alters stress–strain behavior in hydrogel zones linking tendon and bone collagen scaffolds.

Free CS incorporation versus covalent tethering of oxidized CS leads to different biological and mechanical outcomes.

Abstract

Glycosaminoglycans (GAGs) like chondroitin sulfate (CS) influence both mechanical properties and biological signals within the tissue microenvironment. CS modifications have been prevalent in a range of biomaterial design strategies, particularly those with a focus on wound healing. Here, we investigate the impact of CS incorporation within a thiolated gelatin (Gel-SH) hydrogel previously established as a promising biomaterial for tendon-to-bone entheseal repair, reporting a dual biological and mechanical effect. We show that CS inclusion increases mesenchymal stem cell metabolic activity and osteo-tendinous differentiation patterns in the Gel-SH biomaterial. Additionally, we demonstrate that inclusion of CS into a Gel-SH hydrogel insertional zone used to link dissimilar tendon and bone specific collagen scaffolds induces favorable local changes in stress–strain behavior. We further show that the mode of incorporation, free incorporation of CS versus covalent tethering of oxidized CS (CSO), clearly impacts these observed effects. Overall, these results highlight promising new motifs to modulate Gel-SH hydrogels for greater promotion of enthesis-associated behavior in resident hMSCs; further, they offer broad insight into design strategies and key considerations for modification of multicompartment materials, namely in consideration of incorporation methods and on the interplay of mechanical and biological properties.

Figures

A](#F2)). Additionally, despite increasing the density of the hydrogel, inclusion of CS or CSO reduced hydrogel stiffness, as seen in representative stress–strain diagrams ([Figure 2B](#F2)). This is also visualized in best-fit exponential approximations of the stress–strain curves, where the incorporation of CS or CSO resulted in a significantly lower best-fit exponential factor, with a trend for further reduction in stiffness of CSO-modified hydrogels versus CS-modified hydrogels (*p* < 0.10).

A](#F5),[B](#F5)). Local digital image correlation indicated that the incorporation of CS or CSO led to higher magnitude strain concentrations more localized to the tendon compartment or tendon–Gel-SH boundary when the scaffold was subjected to tensile loading ([Figure 5C](#F5)).

Full text

Introduction

1 |

Tissue engineering efforts commonly use biomaterial design motifs to introduce structural, mechanical, and biochemical signals that influence cell activity. Chondroitin sulfate (CS) is a glycosaminoglycan (GAG) comprising repeated linear disaccharide units of glucuronic acid and N-acetylgalactosamine, with variants possessing sulfation at different carbon sites [1]. It is extremely versatile within the body, capable of binding to a variety of core proteins to form proteoglycans crucial to the central nervous system (phosphocan), fibrillogenesis (decorin, biglycan), and cartilage and skin tissue development (aggrecan) [2, 3]. As a component of cell-surface proteoglycans, CS can also interact with key growth factors associated with wound healing and inflammation, including fibroblast growth factors (FGFs) and transforming growth factors (TGFs) [4]. Furter, it contributes to tissue mechanics and homeostasis through generation of high fixed charge density and osmotic swelling pressure, granting cartilage significantly enhanced compressive strength and load-bearing capacity [5]. This wide degree of physiological impact has motivated the use of CS in approaches for drug delivery and biomaterials tissue engineering, particularly for applications in wound healing and musculoskeletal treatment, such as osteoarthritis treatment Condrosulf [6–10]. However, as a primary material, CS is prone to rapid degeneration in water due to its high solubility, motivating instead its conjugation with or incorporation within more stable polymers via functionalization of its abundant hydroxyl and carboxyl functional groups [11, 12]. Such incorporation within engineered hydrogels can enhance chondrogenic gene expression and cartilage-specific matrix production, namely collagen II synthesis [11, 13, 14]. One promising incorporation route utilizes sodium periodate to oxidize native CS, cleaving vicinal glycols on a glucuronic acid unit to form a dialdehyde [15]. These new aldehyde groups can readily bind to free amines, which are found in many native tissues and naturally derived materials, through Schiff base reactions, covalently tethering the GAG [16, 17]. In addition, these aldehyde groups can also quickly crosslink with free thiol groups to form hemithioacetals, providing an additional avenue of crosslinking [9, 18]. These results suggest incorporation of oxidized CS is a facile and efficient means of tethering CS within naturally derived biomaterials.

We have previously described a thiolated gelatin (Gel-SH) hydrogel for use in a triphasic multicompartment biomaterial designed to improve regenerative healing of the tendon-to-bone interface [19, 20]. The inclusion of a hydrogel zone is used to functionally integrate dissimilar tendon- and bone-specific collagen scaffolds to form a triphasic material with increased resistance to tensile fracture. More recently we showed the Gel-SH biomaterial supports human mesenchymal stem cell (hMSC) viability and presents a permissive environment for hMSC differentiation toward a fibrochondrogenic phenotype in response to both local material properties as well as in response to cues from cells in the neighboring tendon- and bone-specific regions [20, 21]. Yet, while the current multicompartment design may provide a level of fibrochondrogenic guidance to hMSCs, it does not promote relevant extracellular matrix deposition or other functional markers of mature fibrochondrocyte differentiation at the accelerated time scales desirable for a regenerative material. Given the biological significance of CS for chondrogenic promotion, we sought to incorporate oxidized CS within the interfacial Gel-SH construct via both free amine and free thiol groups present on its backbone.

In this manuscript we describe efforts to use two methods of CS incorporation within an interfacial hydrogel, namely inclusion of freely suspended, unaltered CS within the Gel-SH hydrogel versus direct crosslinking of oxidized CS (CSO) onto the Gel-SH backbone. We hypothesized that successful CS incorporation could induce pro-fibrochondrogenic or enthesis-associated behavior in resident hMSCs while also improving local mechanical properties of a full triphasic scaffold containing a Gel-SH interfacial zone. We report significant shifts in hMSC behavior following the inclusion of CS/CSO in both monolithic and triphasic Gel-SH variants, as well as significant alteration to mechanical properties of the respective biomaterials. These findings provide new insight into the design of multicompartment biomaterials and the use of a hydrogel-based insertional zone to aid rotator cuff entheseal injury repair.

Materials and Methods

2 |

Gel-SH Synthesis

2.1 |

Gel-SH was synthesized based on previously reported methods [20]. Briefly, bovine skin derived type B gelatin (1 w/v%, CAS: 9000-70-8, Sigma-Aldrich #G9391, Lot 089K0040) was dissolved in deionized water at 50°C until a clear solution was obtained [19, 20, 22–24]. Traut’s reagent (2-iminothiolane-HCl, CAS: 4781-83-3, ThermoFisher Scientific Chemicals Inc. #PI26101) was added at an approximate 2:1 M ratio to free amine groups on the gelatin backbone. The pH was then adjusted to 7.0 (FiveEasy pH Meter, Mettler Toledo) and the solution was stirred at room temperature for 20 min. Subsequently, the pH was lowered to 5.0 and stirring was continued at room temperature for 2 h. The reaction mixture was then transferred into 12–14 kDa MWCO dialysis tubing (08–667E, Spectrum Spectra/Por 4 RC Dialysis Membrane Tubing, Fisher Scientific), and dialyzed against 5 mM HCl for 24 h, followed by 1 mM HCl for an additional 24 h. Following dialysis, the solution was frozen at −20°C for at least 24 h before lyophilizing (FreeZone 2.5L Benchtop Freeze Dryer, Labconco Corporation) and storage at −20°C.

Chondroitin Sulfate Oxidation

2.2 |

CSO was synthesized through reaction of CS and sodium periodate. CS (0.84 w/v%, Spectrum Chemical Mfg. Corp #C1610) was first fully dissolved in deionized water, and an 80% molar equivalent of sodium periodate was added slowly to initiate oxidation [9]. The reaction was carried out under dark conditions with continuous stirring at room temperature for 4 h. To quench residual periodate, 5 mL of ethylene glycol was added, and the mixture was stirred for an additional hour. The oxidized product was then transferred to 8 kDa MWCO dialysis tubing and dialyzed against deionized water for 4 days, with 12 total water changes during the dialysis period. The dialyzed solution was subsequently frozen at −20°C and lyophilized for 7 days, yielding a white, foam-like CSO product.

Hydrogel Preparation

2.3 |

Gel-SH hydrogels were prepared according to established methods, while CS and CSO modified hydrogels were fabricated with slight modifications [20]. A hydrogel precursor solution containing 3.5 wt% Gel-SH, 10 mM hydrogen peroxide (CAS 7722-84-1, H325–500, Fisher Scientific Inc.), 5 mM tyramine (T90344–5G, Aldrich Chemistry), and 5 U/mL horseradish peroxidase (PI31490, Thermo Fisher Scientific Chemicals Inc) was prepared in a solution of Dulbecco’s Phosphate Buffered Saline (DPBS, 21–030-CV, Corning). For hydrogels containing CS or CSO, an additional 3.5 wt% CS or CSO was added to the solution. Precursor solutions were loaded into prepared molds appropriate to produce either a monolithic Gel-SH hydrogel or a triphasic scaffold containing non-mineralized and mineralized scaffold compartments linked by a Gel-SH hydrogel transition (details in subsequent section) [21]. Monolithic hydrogels were allowed to crosslink at room temperature in molds for approximately 45 min, after which they were freeze-dried using a Genesis 25XL freeze-dryer (VirTis), first cooling the suspension to −10°C at a rate of 1°C/min and holding for 2 h, followed by heating to 0°C at 1°C/min while reducing the chamber pressure to 0.2 Torr [20].

Triphasic Scaffold Fabrication

2.4 |

Collagen Slurry

2.4.1 |

Collagen-glycosaminoglycan (Col-GAG) precursor slurries were prepared following previously established protocols [25–27]. To fabricate the bone-specific mineralized collagen slurry, purified fibrillar bovine collagen (1.9 w/v%, Collagen Matrix Inc. #IPFC20N.01) was homogenized within a mineralization buffer (0.1456 M Phosphoric Acid; CAS: 7664-38-2, 7732-18-5, Fisher Scientific #A242; 0.037 M Calcium Hydroxide, CAS: 1305-62-0, Sigma-Aldrich) in a jacketed cooling vessel maintained below 10°C. Once a uniform collagen slurry was achieved, a CS solution prepared in mineralization buffer (0.84 w/v%, Spectrum Chemical Mfg. Corp #C1610) was gradually added in small aliquots. Subsequently, a salt solution composed of calcium hydroxide (0.64 w/v%, CAS: 1305-62-0, Sigma-Aldrich) and calcium nitrate tetrahydrate (0.39 w/v%, CAS: 13477-34-4, Sigma-Aldrich) was incorporated in a similar manner.

For the tendon-specific non-mineralized collagen slurry, a similar protocol to the mineralized slurry was conducted, with slight modifications. Purified fibrillar bovine collagen (0.5 w/v%, Collagen Matrix Inc. #IPFC20N.01) was homogenized in acetic acid buffer (0.05 M, CAS: 64-19-7, Fisher Scientific #A38–212). CS sodium salt (0.044 w/v%, Spectrum Chemical #C1610), dissolved in identical buffer, was added incrementally to the collagen solution to prevent aggregation. Both slurries were homogenized using a T25 digital ULTRA-TURRAX disperser fitted with a S25N-25G-ST Dispersing Tool (IKA Works) and stored at 4°C overnight prior to degassing and lyophilization.

Precursor Loading and Lyophilization

2.4.2 |

Triphasic scaffolds were fabricated according to previous work [19, 21, 28]. Collagen slurries and hydrogel precursor solutions were loaded into custom polytetrafluoroethylene (PTFE) molds with a 1/4-inch-thick copper side using 3D-printed polylactic acid (PLA) phase dividers to enable stratified loading of all three components. Each PTFE well (30 mm × 14 mm × 6 mm) was filled with 1050 μL non-mineralized collagen suspension adjacent to the copper plate, 1050 μL mineralized collagen suspension on the opposite side, and 450 μL of hydrogel precursor solution in between. Non-mineralized slurry was loaded against the copper side of each well to promote unidirectional heat transfer, which allows for anisotropic structure alignment in that region. Immediately after loading, the PLA divider was removed to allow diffusion and integration of the three phases at room temperature for 1 h. Following this period, the loaded molds were then freeze-dried using a Genesis 25XL freeze-dryer (VirTis) as previously described above in Section 2.3. Scaffolds were trimmed post-lyophilization into desired dimensions (28 mm × 6.5 mm × 5 mm) using razor blades and cutting jigs.

Material Rehydration and Secondary Crosslinking

2.5 |

After fabrication, scaffolds were sterilized via 12-h ethylene oxide treatment in an AN74i Anprolene gas sterilizer (Anderson Sterilizers Inc.) followed by a minimum of 48 h ventilation [29, 30]. Prior to in vitro cell work, scaffolds were rehydrated and crosslinked based on established protocols [21, 31]. Scaffolds were distributed on six-well plates and soaked in 100% ethanol to optimize material conditions for aqueous infiltration. Ethanol was then replaced with phosphate-buffered saline (PBS) and incubated for 1 h. Crosslinker 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC, Sigma-Aldrich) was coupled with stabilizing agent N-hydroxysulfosuccinimide (NHS, Sigma-Aldrich) to introduce bonding between carboxylic acid and primary amine functional groups in a PBS solution, which was prepared at a 5:2:1 M ratio of EDC:NHS: COOH based on the carboxylic acid groups expected on the collagen backbone [31]. Scaffolds were incubated in this crosslinking solution at 37°C and 5% CO_2_ on a shaker for 90 min. Following crosslinking, scaffolds were washed with PBS and then basal hMSC culture media comprising low glucose Dulbecco’s Modified Eagle Medium (DMEM) with glutamine (School of Chemical Sciences Cell Media Facility, University of Illinois Urbana-Champaign) supplemented with mesenchymal stem cell fetal bovine serum (10 v/v%, Gemini Bio Products) and antibiotic–antimycotic solution (1 v/v%, ThermoFisher Scientific) at 37°C and 5% CO_2_ for approximately 48 h.

Mechanical Testing

2.6 |

Hydrogel Compression Testing

2.6.1 |

Hydrated monolithic hydrogels were mechanically tested by unconfined compression using an Instron 5943 Mechanical Testing System fitted with a 5-N electromechanical load cell [20, 31, 32]. Samples were removed from solution, blotted to remove excess surface liquid, and compressed at a constant 0.5 mm min^−1^, or roughly 10% strain min^−1^. Stress–strain curves were generated based on measured hydrogel height and initial contact surface area. A best-fit exponential curve ([Stress] = A * [Strain]^b^) was generated for each sample with the exponential parameter b compared between groups as a representative of material compliance.

Scaffold Tensile Testing

2.6.2 |

Hydrated triphasic scaffolds were mechanically tested via uniaxial tensile testing using a Psylotech Micro Test System fitted with a 100-N electromechanical load cell [21, 28, 33]. Prior to hydration, triphasic scaffolds were embedded using a previously described potting method: encapsulation within custom 3D-printed PLA endblocks using a 4:1 ratio of a two-component silicone (RTV615, Momentive Specialty Chemicals Inc.), which was cured at 60°C overnight [19]. Samples were stretched at a rate of 1 mm min^−1^ until fracture, generating a stress–strain curve that was used to calculate tensile elastic modulus (slope of linear elastic region of the stress–strain curve), ultimate tensile stress (maximum stress observed prior to fracture), and toughness (total area under the stress–strain curve from beginning to fracture) [28, 33].

Digital Image Correlation

2.6.3 |

Immediately prior to tensile testing, triphasic scaffolds were dusted with silicon carbide powder to allow for adequate contrast visualization and tracking (P2800 grit, SIC-240-P1, Dace Technologies). During tensile testing, samples were imaged at a rate of one photo per second. Following the conclusion of tensile testing, speckled images were analyzed using VIC-2D software (Correlated Solutions Inc.) [28, 33]. Analysis was run at a subset size of 39 and step size of 6 to calculate local strain within the direction of displacement, e_xx_. Heat maps of this raw data were generated to visualize areas of local strain accumulation and magnitude.

Cell Culture

2.7 |

hMSCs at passage 4 (P4), derived from bone marrow of a 20-year-old African American female donor with no known pre-existing conditions (Lot #310263, RoosterBio Inc.), were thawed in a 37°C water bath. Cells were evenly distributed into T175 tissue culture flasks and expanded in RoosterNourish-MSC media (RoosterBio Inc.) until reaching approximately 80% confluency [20, 34, 35]. Cells were then lifted and counted to prepare a defined cell suspension for orbital seeding, following previously established protocols [20]. Hydrated scaffolds placed in ultra-low attachment (ULA) six-well plates received a defined volume of cell suspension to achieve a seeding density of 600,000 passage 5 (P5) hMSCs per scaffold. The ULA plates were incubated on a shaker at 100 rpm for 6 h at 37°C and 5% CO_2_. After seeding, scaffolds were transferred to fresh ULA tissue culture plates and maintained in mesenchymal stem cell growth media (as described above in Section 2.5) for up to 21 days, with media changes every 3 days.

Metabolic Activity

2.8 |

Relative cell metabolic activity within triphasic scaffolds (n = 6) was assessed using AlamarBlue Cell Viability Reagent (Thermo Fisher Scientific). Prior to staining, scaffolds were gently rinsed in PBS within a new ULA six-well plate. Scaffolds were then incubated in a 9:1 mixture of basal hMSC growth media to AlamarBlue reagent for 90 min on shaker at 100 rpm, under standard incubation conditions: 37°C and 5% CO_2_. Following incubation, AlamarBlue assay solution was transferred to a 96-well plate for analysis. Fluorescence was measured in triplicate at 540/580 nm using an F200 spectrophotometer (Tecan).

Gene Expression Analysis

2.9 |

RNA was extracted both from cell suspensions at Day 0 and from seeded scaffolds at Day 7 and Day 14 using previously established methods [20, 30]. The number of samples analyzed varied depending on the type of interfacial hydrogel incorporated. For cell suspensions, total RNA was isolated using the RNAqueous-Micro Total RNA Isolation Kit (Thermo Fisher Scientific), following the manufacturer’s instructions precisely. For scaffold-seeded samples, each triphasic construct was carefully sectioned into ~2 mm fragments using sterile razor blades to ensure reproducible and compartment-specific sampling. Each segment was labeled and transferred to Phasemaker tubes (ThermoFisher Scientific #A33248). The interfacial hydrogel region was processed using TRIzol Reagent (ThermoFisher Scientific #15596026) followed by chloroform treatment (Sigma-Aldrich #151831) according to previous protocols [30]. Subsequently, RNA clean-up and DNase I treatment were performed using the RNA Clean & Concentrator Kit (Zymo Research #R1014) with the recommended buffer system to ensure removal of contaminants and genomic DNA. Final RNA concentrations were quantified using a NanoDrop Lite spectrophotometer (ThermoFisher Scientific), and purified RNA samples were stored at −80°C until further analysis.

Gene expression in triphasic scaffolds was evaluated using reverse transcription quantitative polymerase chain reaction (RT-qPCR). Purified RNA samples were reverse transcribed into complementary DNA (cDNA) using the QuantiTect Reverse Transcription Kit (Qiagen #205313) on an S100 thermal cycler (Bio-Rad). Quantitative PCR was performed using 65 ng of cDNA per sample with TaqMan Fast Advanced Master Mix (ThermoFisher #4444556) and gene-specific TaqMan primers (Table 1). All reactions were carried out in duplicate using a QuantStudio 7 Real-Time PCR System. Gene expression was analyzed using the ΔΔCT method, where threshold cycle (CT) values were normalized to the housekeeping gene RPL13A [36–38] and to the average values at a reference timepoint (D0, unless otherwise noted) of corresponding interfacial hydrogel samples across experimental groups.

Statistics and Data Visualization

2.10 |

Raw data was characterized and assessed for significance using R (Version 4.3.3) via RStudio (Version 2024.12.1+563). First, data was screened for statistical outliers, defined as any point falling outside 1.5 IQR (interquartile range). Outliers were discounted for any further data characterization and analysis, though included in the data set for visualization. Screened data was assessed for normality and homogeneity of variance using the Shapiro–Wilk and Levene tests, respectively. Any results below a significance threshold of 0.05 resulted in a rejection of the null hypothesis. Data determined to be normal and of homogeneous variance was assessed via ANOVA and Tukey’s post hoc. Normal data without homogenous variance was tested by Welch’s ANOVA and a Games-Howell post hoc. Data with homogenous variance and a non-normal distribution was tested by Kruskal–Wallis and Dunn (Benjamini-Hochberg) post hoc, and data determined to be neither normally distributed nor homogenous in variance was tested via Welch’s Heteroscedastic F-test (trimmed means and winsorized variances) followed by Games-Howell post hoc.

All data-based figures were generated through OriginPro 2024 (Origin Lab Corporation). For box-and-whisker plots, boxes depict the 0.25–0.75 percentile range, with a line and a square symbol showing the data median and mean, respectively. Whiskers denote data within the 1.5 IQR range. Regarding significance, a lone asterisk denotes the significance of one group from others from the same timepoint. An asterisk with branches depicts significant differences between the two indicated groups.

Results

3 |

Validation of Chondroitin Sulfate Oxidation

3.1 |

We validated the oxidation of CS through spectral analysis. Samples appeared largely similar across wavelengths, with the exception of a noticeable peak for CSO samples not seen in standard CS samples (Figure 1). This peak occurred near the 1740 cm^−1^ wavelength, the region typically attributed to the C O bond in an aldehyde group, indicating successful production of CSO.

Chondroitin Sulfate Incorporation Significantly Alters Hydrogel Material Properties

3.2 |

We then incorporated CS and CSO into Gel-SH hydrogels at an equal weight percent to the Gel-SH. The incorporation of either CS or CSO increased the smoothness of the hydrogel constructs as seen in representative SEM imaging (Figure 2A). Additionally, despite increasing the density of the hydrogel, inclusion of CS or CSO reduced hydrogel stiffness, as seen in representative stress–strain diagrams (Figure 2B). This is also visualized in best-fit exponential approximations of the stress–strain curves, where the incorporation of CS or CSO resulted in a significantly lower best-fit exponential factor, with a trend for further reduction in stiffness of CSO-modified hydrogels versus CS-modified hydrogels (p < 0.10).

Chondroitin Sulfate Incorporation Into Gel-SH Hydrogels Affects hMSC Behavior

3.3 |

We next evaluated the biological effects of CS and CSO incorporation on hMSC activity. After 7 days of culture, hMSCs displayed significantly lower metabolic activity in hydrogels modified with CS relative to those in the base Gel-SH. A similar trend was also observed for CSO modified Gel-SH hydrogels (Figure 3). After 7 days of culture, hMSCs also differentially expressed several genes associated with cartilage, fibrocartilage, and the enthesis in response to different hydrogel environments (Figure 4). Cellular expression of COL1A1 was expressed significantly higher in hydrogels supplemented with CS compared to those with CSO. In other genes, such as SPP1, SOX9, and COL10A1, trends of increased expression of these genes were observed in the CS group compared to the base Gel-SH. These effects were not observed in the CSO group, which showed no change or even trends of downregulation versus conventional Gel-SH hydrogels.

Chondroitin Sulfate Incorporation Within an Interfacial Gel-SH Hydrogel Alters Triphasic Scaffold Mechanical Properties

3.4 |

We next investigated the impact of CS or CSO modified Gel-SH hydrogels in a triphasic format containing dissimilar non-mineralized and mineralized collagen scaffolds linked by a Gel-SH entheseal hydrogel. Incorporation of CS into the Gel-SH component of a triphasic scaffold resulted in trends of increased ultimate tensile stress (UTS) and toughness, whereas incorporation of CSO led to a trend of tensile modulus reduction (Figure 5A,B). Local digital image correlation indicated that the incorporation of CS or CSO led to higher magnitude strain concentrations more localized to the tendon compartment or tendon–Gel-SH boundary when the scaffold was subjected to tensile loading (Figure 5C).

Chondroitin Sulfate Incorporation Within an Interfacial Gel-SH Hydrogel Alters hMSC Activity in a Triphasic Scaffold

3.5 |

We finally evaluated hMSC behavior for 21 days across triphasic scaffolds fabricated with a hydrogel entheseal zone formed from either the standard Gel-SH hydrogel or Gel-SH hydrogels supplemented with CS or CSO. Interestingly, incorporation of CS within the Gel-SH interfacial region resulted in a trend of increased metabolic activity, with little to no change observed with the incorporation of CSO compared to the base Gel-SH en-theses (Figure 6).

Lastly, we assessed shifts in gene expression differences at Day 7 and Day 21 for hMSCs cultured within triphasic scaffolds with either a Gel-SH interface or one supplemented with CS or CSO (Figure 7). Compared to a base Gel-SH interface, hMSCs in the CS-supplemented interface significantly upregulated expression of RUNX2 at both Day 7 and Day 21, with trends of upregulation in COL1A1 at Day 7 and Day 21, as well as MMP3 on Day 7. In contrast, hMSCs in a CSO-supplemented interfacial region significantly downregulated both SCX and SOX9 compared to those in a base Gel-SH interface, though upregulation trends at Day 7 for SCX and RUNX2 were observed.

Discussion

4 |

The incorporation of additional extracellular matrix components, vitamins, minerals, and other supplements to tissue engineered scaffolds and hydrogels can significantly enhance instructive guidance toward specific stem cell lineages or responses [25, 30, 39–42]. GAGs in particular have garnered significant interest in this regard due to their influence on cell behavior through modulation of extracellular matrix mechanics and small biomolecule signals, and GAG incorporation has led to significant advances in tissue engineered products for wound healing and orthopedic applications [43, 44]. However, while we have extensively explored modifications to our well-established classes of collagen scaffold, our recently described Gel-SH platform for enthesis tissue engineering has not been similarly investigated. This construct offers a permissive environment for the fibrocartilage patterns observed across the tendon-to-bone enthesis [20, 21], but it would greatly benefit from increased speed and potency of pro-chondrogenic cues to resident hMSCs. CS suggests a first opportunity for modifying the Gel-SH hydrogel, as it is already incorporated within our tendon- and bone-specific collagen scaffolds and is widely established in cartilage tissue engineering [45–47]. However, many methods for the incorporation of GAGs exist, some preserving the component’s natural structure and others modifying it to induce crosslinking or tethering, which can improve integration at a potential cost in bioactivity [48, 49]. This balance of mechanics and biologics is especially crucial within a multicompartment biomaterial, where the material properties of the interfacial region can significantly alter both bulk properties and local mechanoresponse. As a result, we considered the incorporation of CS via two approaches, unmodified or oxidized, to investigate either free incorporation or covalently tethered incorporation, respectively, into the hydrogel network.

We first demonstrated significant differences in physical properties following the free (CS) or tethered (CSO) incorporation into the Gel-SH construct. SEM analysis suggested differences in material surface morphology. Interestingly, both incorporation methods resulted in an evidently softer construct, despite these incorporations increasing the overall weight percentage of the material. Further, CSO incorporation appeared to reduce this response to a greater degree than the CS. Others have reported an increase in compressive modulus in gelatin-based systems incorporating CS or CSO, making the changes we observed unlikely to be occurring due to charge- or crosslinking-based effects of CS/CSO themselves [50, 51]. However, a key difference between our construct and other hydrogel models is the lyophilization, re-hydration, and then secondary crosslinking processes to which our hydrogels are subjected [20]. CSO may also have introduced competition for thiol groups on the gelatin backbone, limiting initial crosslinking of the gel. Though CS and CSO incorporation significantly affected physical properties of monolithic hydrogels, incorporation within a triphasic scaffold was not as substantial at a bulk level. Here, while triphasic biomaterials containing CS and CSO Gel-SH hydrogel interfaces also exhibited trends toward reduced modulus, they additionally displayed beneficial shifts in the distribution of local patterns of tensile strain. Tensile deformation is a critical concern for entheseal regeneration as the rotator cuff sees repeated patterns of tensile stretching. We observed that incorporation of CS or CSO into the Gel-SH interface induced more concentrated pockets of strain in the tendon-specific region of the multicompartment. This is likely due to the higher overall weight percentages of the CS and CSO supplemented interfaces, which may have inhibited diffusion of the hydrogel during fabrication and limited compartment integration. However, increased tensile strain in the tendon compartment may be a favorable finding as tenocyte differentiation and activity is highly dependent on tensile strain in both regenerative and developmental settings [52–57].

CS incorporation into Gel-SH hydrogels also influenced subsequent cellular activity. Notably, CS incorporation boosted cellular expression of several genes of interest for the enthesis region. Here, CS added to monolithic hydrogels increased fibrochondrogenic progenitor markers SOX9 and SCX [58–60], extracellular matrix marker COL10A1, which is associated with hypertrophic chondrocytes typically found within the mineralized fibrocartilaginous region of the enthesis [61–63], and COL1A1, a key marker in fibrocartilage development as well as osseous and tendinous tissue [63, 64]. When incorporated into triphasic scaffolds, hMSCs within the interfacial region displayed a similar response to CS in COL1A1 expression. The addition of CS also promoted early (Day 7) SOX9 expression and later (Day 21) expression of SCX and the transcription factor RUNX2 associated with osteogenesis and hypertrophic chondrocytes [63, 65]. Broadly, these findings provide strong evidence that the incorporation of soluble, unmodified CS within a Gel-SH hydrogel promotes a pro-enthesis phenotype in hMSCs. Interestingly, incorporation of modified CSO had a largely negative effect on enthesis-relevant gene expression patterns. hMSCs in CSO-incorporated hydrogels showed downregulated COL1A1 and COL10A1. hMSCs in triphasic constructs containing a CSO-functionalized Gel-SH insertion displayed downregulated SCX and SOX9 as well as reduced ACAN and RUNX2. Of further note, the impact on hMSC metabolic activity with CS and CSO incorporation was highly dependent on whether the cells resided in monolithic hydrogels or within triphasic scaffolds, with CS incorporation resulting in significantly reduced metabolic activity within monolithic constructs and trends of reduced activity in triphasic scaffolds. The significant differences in the response to CSO or CS incorporation may stem from the chemical changes CSO enforces upon the Gel-SH base. Others also documented negative effects of highly oxidized CSO on cellular viability and chondrogenic induction, which could further contribute to this behavior [49]. Such findings motivate subsequent efforts to varying degrees (wt%, oxidation extent) of CS modification to the Gel-SH hydrogel as well as projects to assess the influence of CS incorporation on biomolecule sequestration and release. Further, the significant chemical and biological environmental differences between a monolithic hydrogel and a triphasic scaffold with a hydrogel interface clearly impact the effects of CS or CSO incorporation, highlighting the importance of utilizing triphasic scaffolds in future work. Sulfated GAGs such as CS have previously been used to change the bioavailability of exogenously added biomolecules or endogenously produced autocrine factors [66–70]; while not the subject of this work, future efforts may be able to exploit CS-modified Gel-SH hydrogels to enhance the quality and kinetics of hMSC fibrocartilage differentiation and matrix remodeling.

Conclusions

5 |

Here we describe the impact of incorporation of CS into a Gel-SH hydrogel platform for fibrocartilaginous enthesis engineering. Both unmodified and CSO significantly alter the mechanical properties of the base hydrogel, increasing the compliance of monolithic hydrogels and reducing the ability of a Gel-SH hydrogel interface to transfer tensile loads across a multicompartment material. Incorporation of unmodified CS induced increased expression of key fibrochondrogenic transcription factors and matrix markers at both early and extended timepoints, suggesting the formation of an environment more conducive to a fibrochondrogenic lineage. These results highlight an improved hydrogel material for the promotion of fibrocartilage promotion and enthesis tissue engineering, offering new insight into avenues and future considerations for biomaterial enhancement via the incorporation of bioactive components.