Xeno-free bioengineered human skeletal muscle tissue using human platelet lysate-based hydrogels

Human immortalized muscle satellite stem cells (HuMSCs) [44] were provided by Dr Bénédicte Chazaud from the Institut NeuroMyoGène, Lyon, France. Primary muscle satellite stem cells were previously isolated from a healthy 14 year-old muscle biopsy and immortalized using CDK-4/hTERT expression. Cells expressing PAX7 were grown in Skeletal Muscle Basal Medium (PromoCell), containing skeletal muscle supplemental mix (#C-39365, PromoCell), 1% v/v penicillin/streptomycin (P/S, 10 000 U ml⁻¹, Thermo Fisher Scientific), and 10% v/v fetal bovine serum (FBS, Thermo Fisher Scientific). The growth medium added after cell encapsulation did not contain FBS but was supplemented with 1 mg ml⁻¹ of 6-amino-n-caproic acid (ACA, Sigma-Aldrich). The differentiation medium (DM) consisted of Dulbecco's modified eagle medium (DMEM), high glucose, GlutaMAX™ (Gibco, Thermo Fisher Scientific), 1% v/v Penicillin-Streptomycin-Glutamine (P/S-G, 100 X, Gibco, Thermo Fisher Scientific), 1% v/v Insulin-Transferrin-Selenium-Ethanolamine supplement (ITS-X, 100 X, Gibco, Thermo Fisher Scientific), and 1 mg ml⁻¹ of ACA.

The hydrogel casting platforms were designed using Fusion 360 software (Autodesk) as circular chips (8 mm diameter) with a rectangular well of 35 µl volumetric capacity, containing two T-shaped posts (diameter: 0.8 mm; height: 3.25 mm) (figure S1(a)). In this protocol, polydimethylsiloxane (PDMS) replicas of a 3D printed master mold were obtained by Ecoflex^TM (00-30, Smooth-On) negative intermediary molds (figure S1(c)) [45].

2.2.1. Fabrication of the master molds

2.2.2. Fabrication of the negative molds

2.2.3. Replica molding of PDMS platforms

2.3.1. Preparation of HUgel precursors

2.3.2. Cell encapsulation by HUgel casting in PDMS platforms

PDMS casting platforms were treated with 0.2% Pluronic® (F-127, Sigma-Aldrich) for 20 min to avoid hydrogel attachment to the PDMS well. After treatment, the platforms were washed three times with phosphate buffered saline (PBS) and sterilized with UV light. For cell encapsulation, HuMSCs were tripsinized and resuspended in PL. In these experiments, we worked with final cell densities of 0.5, 1, or 2.5 × 10⁷ cells ml⁻¹. The a-CNC water dispersion was placed in a sterile Eppendorf tube at the desired concentration and sonicated for 5 min. Then, the suspension was mixed with thrombin from human plasma (2 U ml⁻¹, Sigma-Aldrich) and CaCl₂ (10 mM, Sigma-Aldrich). The cell suspension in PL was thoroughly mixed in a 1:1 ratio with the suspension containing a-CNC, thrombin, and CaCl₂. Finally, a volume of 35 µl of the mixture was carefully placed inside each PDMS platform well. The samples were allowed to crosslink at 37 °C for 30 min before adding growth medium (day −2). After two days (day 0), the medium was switched to differentiation medium (figure 1(a)). Then, half of the medium was replaced every two days. The 3D culture was carried out under shaking conditions inside an incubator (37 °C, 5% CO₂).

Zoom In Zoom Out Reset image size

2.3.3. Immunofluorescence staining

2.3.4. Electrical pulse stimulation (EPS)

2.3.5. Imaging

2.3.6. Force measurements

The post deflections during tissue contractions were recorded and analyzed using Fiji. The force-displacement relationship for the posts was estimated using linear bending theory based on previously published protocols [48, 49], by the following formula:

$$\begin{equation*}F = \frac{{6\pi E{D^4}}}{{64{a^2}\left( {3L - a} \right)}} \cdot d = k \cdot d.\end{equation*}$$

The Young's modulus of the PDMS (E) was previously measured as 1.6 ± 0.1 MPa. Considering the geometry of the posts and dimensions, we calculated a spring constant (k) of 3.54 N m⁻¹. The spring constant was used to transform the recorded post deflections (d) into the force generated by the bioengineered 3D skeletal muscle tissues (F) (figure S1(b)).

2.3.7. RNA extraction, RT-PCR, and real-time PCR

2.3.8. Statistical analysis

Engineered skeletal muscle tissues require specific topographical and microenvironmental signals that favor cell alignment and fusion into myotubes. When scaffolds are made of compactable biomaterials, such as collagen, EHS matrices, or fibrin, the most common fabrication approach introduces tendon-like attachment points to provide uniaxial tension during ECM remodeling [50]. Following this strategy, fibrin-Matrigel® and fibrin-Geltrex^TM hydrogels have been cast in silicone molds with nylon frames or hooks as anchor points [24–27]. The resulting muscle tissue bundles needed to be manipulated and transferred to a different platform for stimulation and force measurements. Another widely used method involves cell-laden matrix compaction around a pair of flexible silicone posts [16, 19, 22–24]. Besides providing uniaxial tension, these hydrogel casting systems allow in situ tracking of post deflection due to tissue contraction. Furthermore, these deflections can be transformed into force measurements. Considering these advantages, we fabricated hydrogel casting platforms containing two flexible posts as anchor points (figure S1). The final PDMS platforms were fabricated by replica molding of 3D printed master molds, using Ecoflex^TM as reusable intermediary negative molds. Ecoflex^TM is a platinum-catalyzed silicone that is cured at room temperature, resulting in a highly stretchable material that can be applied as a fully deformable elastic mold [51]. The individual platforms were designed as circular chips that fit inside 48-well plates. Each platform has a rectangular casting well with a volumetric capacity of 35 µl and contains a pair of posts that direct tissue formation. Crucially, to ensure that tissues are retained under tension, we designed hook-like features on the top of each post.

Human PL-based nanocomposite hydrogels (HUgel) were fabricated following a previously published protocol based on the formation of fibrin-based scaffolds [42]. In this work, HuMSCs were encapsulated in HUgel by hydrogel casting in the two-post PDMS platforms. Briefly, a suspension of HuMSCs in PL was mixed with a-CNC water dispersions containing thrombin and calcium ions. We fixed the concentrations of thrombin (1 U ml⁻¹) and CaCl₂ (5 mM) according to the previous study by Mendes et al [42]. For the initial experiments, a-CNC was mixed at a final concentration of 0.5% w/v, and cells were encapsulated at a density of 5 × 10⁶ and 1 × 10⁷ cells ml⁻¹. We observed that the cell-laden HUgel matrix started to detach slowly from the walls of the PDMS well after two days of culture. Hydrogel compaction around the posts due to matrix remodeling by the encapsulated cells could be observed after four days (figure S1(d)). At this time, tissues containing 1 × 10⁷ cells ml⁻¹ were more compacted than those with fewer cells. Matrix remodeling and compaction around the posts is essential in these casting systems to implement uniaxial tension and induce self-organization during tissue formation [19, 52]. Consequently, cell density was increased to 2.5 × 10⁷ cells ml⁻¹ for future experiments.

The physical and biological properties of nanocomposite hydrogels with varying a-CNC content have been extensively characterized in previous studies [42, 43]. Significantly, researchers found that increasing a-CNC content hampers the typical fast densification of the fibrin matrix, which is referred to as clot retraction. In some tissue engineering applications (e.g. for space-filling of wounded tissues), extensive clot retraction is considered an undesirable hydrogel effect. Nevertheless, for skeletal muscle tissue formation, the compaction of the cell-laden matrix around the posts contributes to implementing uniaxial tension that facilitates the self-organization and alignment of myoblasts. At the same time, the matrix remodeling favors myoblast fusion into myotubes. Here, we modified the content of a-CNC to obtain long-lasting scaffolds with high clot retraction properties that promote skeletal muscle tissue formation (figure 1(b)). Encapsulating human skeletal muscle satellite stem cells in a high density of 2.5 × 10⁷ cells ml⁻¹ resulted in HUgel detachment from the wells of the PDMS casting platforms within a few hours of culture. The matrix compaction after 12 h was calculated as the percentage of reduction in tissue width (figure 1(c)). We found that the initial working concentration of 0.5% w/v a-CNC resulted in 25.7 ± 1.6% of matrix compaction, whereas the tissues without a-CNC presented extensive compaction of 66.4 ± 4.3%. However, the lack of mechanical reinforcement from a-CNC caused hydrogel degradation, so the samples broke and collapsed between 12–60 h after cell encapsulation. On the other hand, hydrogels with a low content of a-CNC (0.3 and 0.2% w/v) offered a high retraction from the beginning of the culture (33.6 ± 2% and 46.7 ± 1.1%, respectively) and were stable for at least two weeks.

PL gels are attractive scaffolding materials for stem cell encapsulation because PL contains structural proteins like fibrinogen, as well as growth factors and cytokines that stimulate cell growth and proliferation [38, 41]. However, PL gels alone do not retain these growth factors within the scaffold during culture. Therefore, the impact of a-CNC content on the sequestering and release profile of proteins from the HUgel scaffolds was previously investigated [42]. The results indicated that the a-CNC play an important role in growth factor retention within the HUgel scaffolds due to the formation of a protein corona on the a-CNC surface. This process is mainly driven by the non-covalent protein binding enabled by the nanoparticles surface chemistry [53]. Notably, nanocomposite hydrogels with high a-CNC content (at least 0.61% w/v) are optimal for hASCs cell encapsulation, where proliferation during cell culture is desired [42, 43]. On the contrary, in skeletal muscle tissue engineering, differentiation of muscle precursor stem cells requires that cells stop proliferating and exit the cell cycle in order to fuse and differentiate into multinucleated myotubes. In fact, all cell culture protocols for differentiating skeletal muscle cells involve removing growth factors and reducing the animal serum present in the basal media [36]. Thus, a microenvironment rich in growth factors could hinder adequate differentiation.

Considering the matrix compaction properties and the expected growth factor release profiles, we selected a-CNC content of 0.2% and 0.3% w/v for the following experiments with HuMSCs. The serum-free growth medium was replaced by serum-free differentiation media 48 h after cell encapsulation (day 0) (figure 1(a)). We analyzed the expression of myogenic genes over culture time and between samples with different a-CNC content (figures 1(d)–(i)). Tissues with 0.2 and 0.3% of a-CNC shared a common profile of myogenic gene expression without any significant differences. As expected, the expression of myogenin (MYOG), a key developmental regulator for inducing myotube formation [54], was increased after a few days in differentiation conditions (day 4) (figure 1(d)). This increase was followed by a reduction of MYOG expression on day 7, indicating myotube maturation with culture time [55]. Myotube maturation was observed by an increased expression of MYH7 (figure 1(e)), along with the specific striated muscle markers TNN1 (figure 1(f)), ACTN2 (figure 1(g)), and titin (TTN) (figure 1(h)). Furthermore, after seven days in differentiation media, the xeno-free bioengineered muscle tissues were stained for (SAA, green) (figures 2 and S2). The confocal images showed long, highly aligned, multinucleated myotubes expressing SAA.

Zoom In Zoom Out Reset image size

The functionality of bioengineered muscle tissues was assessed by their response to EPS. After seven days in differentiation conditions, we performed EPS using graphite electrodes fixed on a modified 12-well plate lid. Following the experiments, immunostaining of electrically stimulated and control (non-stimulated) tissues was carried out (figure 3). Confocal images of SAA stainings showed the presence of striations (white arrows in figure 3(a)), indicating myotube maturation in some of the myotubes. These striated myotubes were only present in electrically stimulated tissues. Nonetheless, we expect that extended culture periods combined with electrical stimulation training regimes would improve myotube maturation, obtaining uniformly striated myotubes.

Zoom In Zoom Out Reset image size

As described above, one advantage of hydrogel casting around two flexible posts is performing in situ force measurements. In these systems, characterization of the posts' mechanics allows converting the post deflections (caused by contractile tissues) into force measurements [19, 56]. Therefore, measurements are non-invasive and compatible with cell culture conditions.

Bioengineered muscle tissues were subjected to a frequency sweep EPS regime going from low to high frequencies, according to the scheme in figure 4(a). We did not observe any spontaneous contractions without electrical stimulation. However, the electrically stimulated bioengineered human skeletal muscles presented twitch or tetanic contractions depending on the applied frequencies (figures 4(c) and (d)). To compare tissue functionality between different conditions, force measurements were performed for HUgel formulations containing 0.2 (figure 4(e)) and 0.3% w/v a-CNC (figure 4(f)). Notably, increasing a-CNC content led to a significant force reduction, and the maximum tetanus contractile force with 0.3% a-CNC was approximately four times lower than with 0.2% a-CNC (figure 4(g)). Since there were no significant differences in myogenic gene expression between 0.2 and 0.3% a-CNC, other factors might play a more critical role in the measured contractile forces. For instance, the difference in matrix compaction could influence the myotube distribution and organization throughout the tissue, and this could result in different force transmission. Furthermore, it has been shown that higher stiffness of engineered muscle tissues can be correlated to lower tetanus contractile force due to cell-matrix interactions [57]. Given that HUgel formulations with 0.3% a-CNC present a higher theoretical stiffness [42], a stronger response from embedded myotubes could be required to achieve the same magnitude of post deflection than in formulations with lower a-CNC content.

Zoom In Zoom Out Reset image size

Altogether, these results highlight the potential of PL-based nanocomposite hydrogels to develop bioengineered skeletal muscle tissues as functional xeno-free in vitro models. Taking advantage of the tunable physical and biological properties of these nanocomposite hydrogels, we optimized HUgel's formulation to generate contractile human skeletal muscle tissues. As mentioned above, the maturation of these bioengineered tissues could be enhanced by increasing culture time and electrical stimulation training.

The application of this biomaterial for the 3D cell culture of human adipose stem cells has been explored in previous works [42, 43]. However, at present, HUgel has not been used to generate any tissue models. Furthermore, using this biomaterial would offer several advantages for skeletal muscle tissue engineering, such as the low batch-to-batch variability of its components and its xeno-free characteristics. In fact, there are no xeno-free skeletal muscle tissue models reported to date. For drug development applications, xeno-free tissue models could provide increased relevance by eliminating animal-derived materials and reagents that could reduce sensitivity to drug effects. Additionally, the nature of these nanocomposite hydrogels offers excellent potential to develop personalized tissues from patient-derived PL, which could be used for in vitro studies, and at the same time, could have a more straightforward translation from bench to bedside.