Mechanical anisotropy in a scaffold can be imparted by fiber alignment and organization. To date, the methods to generate aligned, anisotropic scaffolds can be classified into the following categories: electrospinning with a rotating collector, gap electrospinning, and 3D bioprinting. Brief descriptions of the main strategies and examples of each category are provided below.

Electrospinning utilizing a rotating collector permits the modulation of fiber alignment through alterations in the geometry and/or rotational speed of the collector. A rotating cylinder mandrel is the most commonly used method, although it does not provide the highest degree of alignment compared to other methods. In this method, the linear speed at the surface of the rotating drum (i.e., rotating velocity) should match the solvent evaporation rate. The kinematics of the mandrel are determined by the category of processing parameters, which further influence the arrangement of nanofibers (alignment, fiber size, etc.) on the collecting surface [64,65].

Achieving fiber alignment requires the careful selection of the processing conditions when using a cylinder rotating mandrel to achieve fiber alignment. First, the induction of fiber alignment occurs within a narrow range of the rotational speed (e.g., between 3.0 and 10.9 m/s) [65]. When the rotating speed is lower than the take-up speed of the fiber, randomly oriented fibers are formed on the drum. When the rotating speed is too high, the depositing fiber jet breaks, and this prevents continuous fibers from being collected [66]. Secondly, within this range, an increasing rotational speed results in more aligned nanofibers. The fiber alignment typically presents a normal distribution of the fiber angles, and the degree of anisotropy is determined by the histogram profile of the fiber angles on the sheet [58,67]. This feature can be viewed as an advantage because the myofibers/collagen fibers from the histological measurement of native myocardium exhibit the same pattern [68].

To further enhance the fiber alignment, some researchers have utilized a rotating disc. In this setup, the thin edge of the collector concentrates the electric field, permitting the deposition of highly aligned fibers thereon. The charged jet is restricted within the edge because the electrostatic field between the sharp edge point (+) and needle (−) becomes the strongest in this location. However, highly aligned fibers can only be formed in a relatively small region, and this severely limits the size of the scaffold that can be fabricated [69,70,71]. Like in the cylinder mandrel setup, one should note that the rotating speed not only affects the nanofiber alignment but also the fiber diameter and porosity and, ultimately, the bulk mechanical properties.

Additional modifications to the collector enable the replication of the 3D geometry of the tissue. For example, a 3D tube construct can be formed for vascular graft applications using a small-diameter rotating rod (<5 mm) [72,73]. This makes it possible to employ distinct polymers for different layers without the need for further assembly, which replicates native vessel characteristics [73]. Using a conical mandrel allows the fabrication of scaffolds with curvilinear microarchitectures that mimic heart valves [74].

Gap electrospinning induces aligned nanofibers by an applied electrical field. By applying a positive voltage to the polymer solution and a negative voltage to two neighboring plates separated by a gap, the fibers are deposited and stretched from one plate to the other due to the residual electrostatic repulsion between the plates. Numerous alterations have been made to the basic setup to achieve variations in the microarchitecture, and these were reviewed in depth by Robinson et al. [62]. However, the maximum length of nanofiber sheets has been limited to 10 cm, because large distances inhibit the jet crossing from one side to the other [75,76,77]. To overcome this limitation, Lei et al. recently applied a negative voltage to a U-shape collector and successfully produced long aligned fibers (up to 60 cm) [78,79].

Gap electrospinning offers significant benefits in producing controllable, aligned electrospun fibers. It is cost-effective since, in most configurations, no extra equipment is required beyond a typical electrospinning device. In addition, the fiber orientation and gradient of alignment can also be adjusted. However, there are a few drawbacks to the approach. The technology is restricted by the mesh thickness, as the residual charge increases with the mesh thickness. The rise in residual charge causes electrical repulsion and, consequently, a loss of fiber alignment [73]. Finally, since the highly aligned scaffold generally possesses low mechanical strength in the cross-fiber direction, the handling of the thin scaffold is challenging during the removal of the scaffold from the mandrel.

Three-Dimensional printing can also be used to induce fiber alignment in anisotropic scaffolds. There are two strategies to deposit aligned fibers: (i) direct depositing into a customized pattern to achieve the complex alignment of micro/nanofibers [80,81]; and (ii) the shear-induced alignment of threadlike nanofibers or the elongated deformation of injected components along the printing direction [82,83]. Cu et al. [84] printed a variety of designs featuring different fiber widths (100, 200, 400 μm); filling densities (20, 40, 60%); fiber angles (30°, 45°, 60°); and stacking layers (2, 4, 8 layers) to create anisotropic scaffolds compatible with cardiomyocytes. They claimed that the scaffolds accurately represented the transmural fiber alignment and curvature of murine left ventricles.

The advantages of this method include the simultaneous control over the micro-geometry and macro-architecture (such as fiber alignment), the feasibility of achieving a high resolution (~5–50 μm) in the fiber organization, and the proper cell density within the scaffolds [85]. It is important to note that hydrogels are often used and deposited as bioinks to enhance the bioactivity of the scaffold; recent 3D bioprinted cardiac scaffolds were reviewed by Wang et al. (see Table 1 in [86]).

The incorporation of anisotropy in tissue-engineered scaffolds not only replicates the structural features of native cardiac tissues but also allows for mechanistic studies that can improve the understanding of heart diseases. One important consideration in replicating tissue anisotropy is the fiber angle distribution. As described above, the ventricular wall exhibits a normal distribution of myofiber angles in the tissue sections, and this feature can be achieved by electrospinning with a cylindrical rotating mandrel [12]. In contrast, other approaches including 3D bioprinting generate uniformly aligned or grid structures of fibers that are absent in native tissues. The exact cause and consequences of the normal distribution of myofibers in a single section are not yet fully understood, but a biomimetic cardiac scaffold should consider this feature during scaffold fabrication. Moreover, multiple layers of sheets with varied main fiber angles can be produced either by electrospinning or by 3D bioprinting methods, replicating the myocardium or heart valves with layered, anisotropic characteristics. However, in native tissues, there is also a functional integration of aligned constituents across layers. The current engineering techniques have not been able to provide such in vivo bonding features between aligned layers [87].

The benefit of using or implanting an anisotropic scaffold for the whole organ function has been reported previously. Mathematical modeling and in vivo studies have shown that anisotropic scaffolds, compared to isotropic ones, enhanced the functionality of a diseased heart by improving depressed LV pump function and increasing systolic function without compromising the filling (diastolic function) [103,104]. Through mathematical modeling, Sallin et al. [105] further demonstrated the significance of myocardial fiber arrangement in the ventricular wall by promoting effective cardiac pumping. When the heart is modeled as an ellipsoid with myocardial fibers oriented in the circumferential (diseased) vs. longitudinal (normal) direction with a helical fiber organization, the ejection fractions are markedly different (30% vs. 60%) and represent those of failing and normal hearts, respectively. Chang et al. fabricated a 3D dual-ventricle bioscaffold with three layers, each with distinct helical arrangements. They showed that the cardiomyocytes (CMs) exhibited appropriate alignments in this scaffold, and the entire construct achieved the spatiotemporal control of excitation–contraction coupling. Additionally, their observation of an increased ejection fraction in the longitudinally aligned scaffold agreed with the results predicted from Sallin’s model. In this investigation, however, the mechanical behavior of the scaffolds did not match that of the native myocardium. The collagen fibers in the natural myocardium coil tightly at small strain rates and uncoil to become stiffer at high strains. In contrast, this 3D scaffold did not reproduce the nanoscale structure of collagen fibers, resulting in straight, bundled fibers that were linearly elastic throughout the strain range [106,107].

Anisotropic structures of native tissues, resulting from the aligned arrangement of ECM components or cells, play an essential role in carrying out and maximizing their direction-dependent physiological functions. Studies probing the cellular responses to anisotropic mechanical environment have been conducted by comparing the outcomes obtained from isotropic and anisotropic scaffolds. The first response of cells to aligned substrates is to change their shape and orientation. Cardiomyocytes cultured on (isotropic) plastic are oriented randomly. As a result, their contractile force is distributed in all directions. However, when cultured in anisotropic scaffolds, the CMs will adopt the fiber alignment and be properly positioned on the scaffold [108]. The elongated cell alignment in turn influences the contractile force as well as cell–cell and cell–matrix interactions. Aligned CMs are also more mature and exhibit a more physiological behavior than randomly distributed cells. For instance, Wanjare et al. [109] co-seeded human iPSC-derived cardiomyocytes (iCMs) and endothelial cells (iECs) onto electrospun polycaprolactone scaffolds with either a randomly oriented or parallel-aligned microfiber configuration. They showed that, in contrast to randomly oriented scaffolds, the aligned scaffolds led to iCM alignment along the microfiber direction and promoted iCM maturation by increasing the sarcomeric length and gene expression of myosin heavy chain adult isoform (MYH7). The maximal contraction velocity of iCMs on aligned scaffolds was significantly greater (3.8 m/s) than that on randomly oriented scaffolds (2.4 m/s). These outcomes demonstrate that anisotropic scaffolds promote CM maturation and contractility.

Other groups have examined the effect of matrix anisotropy on stem or progenitor cell function to elucidate cell mechanobiology and its regenerative potential for the heart. For instance, the role of matrix anisotropy in mesenchymal stromal cell (MSC) behavior and paracrine functions has been investigated. Matrix anisotropy has been shown to play a role in MSC morphology, differentiation fate, and other paracrine functions [110,111,112,113,114,115,116]. Recently, Nguyen-Truong et al. [97] examined the effect of RV tissue mechanics on the pro-angiogenic paracrine function of MSCs, concentrating on the combined effect of RV-like tissue stiffness and anisotropy. Using random and aligned PEUU electrospun scaffolds with the stiffness of normal RVs, they found that the MSCs cultured on the anisotropic group consistently exhibited a higher pro-angiogenic function than those cultured on the isotropic group, showing a positive influence of anisotropy on MSC paracrine function. However, this impact of anisotropy was lacking in the stiff scaffold groups resembling diseased RVs. These results highlighted the importance of the synergistic effect of matrix stiffness and anisotropy in the regulation of MSC function, which may lead to the mechanical conditions of MSC-based treatments for heart failure. Similarly, Allen et al. [117] investigated mouse embryonic stem cell differentiation toward CM regulated by substrate anisotropy. They showed that the cell alignment exhibited a gradient-based response (nonaligned, semi-aligned, and highly aligned) to substrate anisotropy and that an aligned substrate accelerated CM maturation to generate synchronous beating.

Inspired by biological tissues, the fabrication of crimped, extendable fibers is the main strategy to impart nonlinear elasticity on a biomaterial. One way to induce crimped fibers is by permanently lengthening the sheet along the main-fiber direction first and then returning the sheet back to the pre-stretched length. Meng et al. applied this method to electrospun scaffolds made with polylactocaprone (PCL), poly(lactic acid) (PLA), and poly(l-lactide-co-caprolactone) (PLCL), and they found that the mixture of the three was effective in the formation of crimped structures [119]. In the aligned PLCL scaffold, the fibrous sheet was stretched repeatedly, resulting in permanent elongation. Then, the entire sheet was positioned into the pre-stretched shape, treated with heated ethanol spray, and cooled down quickly to produce wavy nanofibers. This crimped fibrous structure was confirmed by SEM imaging, and the nonlinear elastic behavior was measured by uniaxial tensile mechanical tests. Interestingly, the same methodology failed to generate the crimped fiber structure in the randomly aligned PLCL scaffolds, and thus the nonlinear elastic behavior was absent in these scaffolds. However, using similar methods, Niu et al. electrospun tubular PLCL scaffolds with randomly aligned, axially aligned, and circumferentially aligned structures [120]. They reported nonlinear elastic behavior in all scaffolds. The nonlinearity of these scaffolds was compared and found to be similar to that of native blood vessels (porcine aorta ventralis).

Another way to produce crimped fibers is by controlled heating and/or chemical treatment, as briefly reviewed by Szczesny et al. [121] and Zhang et al. [122]. However, these methodologies often generate scaffolds with low porosity, which results in limited crimped fibers and poor cell infiltration. To improve these aspects, Szczesny et al. electrospun a dual poly-L-lactide (PLLA)/poly(ethylene oxide) (PEO) solution and heated the sheet between two glass slides, either before or after washing the scaffolds to dissolve PEO fibers, with or without poly(vinyl alcohol) (PVA) treatment to increase fiber bonding [121]. They found that only the wash-and-then-heat group exhibited nonlinear stress–strain behavior, whereas the PVA-treated scaffolds failed to present nonlinear elastic behavior. In addition, increased porosity has been found to promote the formation of crimped fibers. In the same study, the authors showed a potential link between porosity and the fiber crimping of the scaffold. Recently, Zhang et al. prepared nanofibrous PLCL/PEO scaffolds and found that the fiber crimping and nonlinear elastic behavior increased with an increase in mesh porosity [120]. This report was consistent with the previous finding of Szczesny et al.

Finally, certain materials may exhibit nonlinear behavior and can be used to fabricate scaffolds. For example, poly(glycerol dodecanedioate) (PGD) is a shape-memory, biodegradable elastomer that is linearly elastic at room temperature but has nonlinear elasticity at body temperature. Ramaraju et al. showed that the incorporation of the small intestinal submucosa (SIS) into the PGD sheets induced nonlinearity in the scaffolds. The mechanical properties of PGD can be tuned with native SIS by altering the thermal curing conditions used. The reason for the nonlinear elastic behavior is thought to be the void spaces formed during the incorporation of SIS sheets into PGD, but increasing the void spaces also decreases the stiffness of the scaffolds [123].

The nonlinear elasticity of matrices changes cell–matrix interactions by regulating cell adhesion, spreading, and signal transduction. Prior studies have shown that cells grown on fibrous ECM with mechanical nonlinearity perceive the mechanical signal distance to be far greater than those grown on synthetic linear elastic polymeric material [119,124,125]. Meng et al. showed that compared to the human umbilical vein endothelial cells (HUVECs) cultured on linear elastic scaffolds, the HUVECs cultured on nonlinear (aligned and crimped) PLCL scaffolds had a greater density of focal adhesions and a higher expression of focal adhesion proteins. This indicated a stronger cell–matrix interaction, which more effectively transduced mechanical signals. These cells also had an increased spreading area, thereby promoting the formation of an endothelial layer on the vascular scaffold. The cell proliferation rate on the nonlinear elastic scaffold was lower than that on the linear elastic scaffold, but it was attributed to the lower Young’s modulus in the nonlinear elastic scaffold [119]. In a separate study, Zhang et al. showed that a nonlinear elastic scaffold promoted HUVEC adhesion and proliferation despite the reduced stiffness of the scaffold. These cellular responses were attributed to the rough surface, increased porosity, and increased hydrophilicity of the nonlinear elastic scaffold rather than mechanical factors [122]. Liu et al. showed that the nonlinearity of the ECM regulated the organization of hASCs by preparing six gels with different concentrations and critical stresses. Finally, Niu et al. cultured HUVECs on nonlinear elastic tube scaffolds with three different fiber orientations (random, circumferential, and longitudinal alignment) [120]. They did not include linear elastic scaffolds as a control, and thus it remains unknown whether the cell proliferation is altered by nonlinear elastic properties.

Crimped fibrous scaffolds promote cell spreading and adhesion, but the effect on cell proliferation remains unclear. However, the mechanisms for altered cell responses are mostly attributed to the matrix topography (rough surface or porous structure) or surface chemistry (hydrophilicity) of the crimped fibrous scaffolds. Whether the mechanical behavior (nonlinear elasticity) is just a side product of the crimped fibers or directly affects the mechanical transduction of the cells is unknown. The exact role of the nonlinear elastic behavior of the substrate in the mechanical signaling pathway of cells should be investigated in future work.

Hydrogels are the most commonly used biomaterials for constructing viscoelastic substrates. Hydrogels can be classified based on the source of the polymers—natural ECM biopolymers (e.g., collagen or fibrin hydrogels); synthetic hydrogels (e.g., polyethylene glycol (PEG) or polyacrylamide (PAM) hydrogels); and naturally derived macromolecular hydrogels (e.g., alginate or chitosan hydrogels). Currently, the main approaches used to modulate the viscoelasticity of hydrogels include: (1) crosslinking polymers; (2) altering the polymer architecture, such as length and branching; (3) tuning the composition; and (4) altering the concentration of the polymer or polymer mixture [132].

Crosslinks in polymeric hydrogels can be physical (e.g., ionic or covalent) and can be static or dynamic. Vining et al. generated various alginate–collagen hydrogels via combined ionic and covalent crosslinking at different densities to tune the matrix viscoelasticity. Across a narrow range of moduli (0.25 kPa, 0.5 kPa, and 2.5 kPa), the equilibrium stress relaxation of the scaffolds was similar to that of the native ECM [133,134]. This parameter was increased significantly (>3000 s) by the addition of covalent crosslinks, which indicated a weakening of the viscoelastic behavior of the scaffold. Because ionic crosslinks are weaker bonds than covalent crosslinks and make it easier to induce frictional energy loss during deformation, the stress relaxation is more pronounced in ionically crosslinked hydrogels. Besides physically crosslinked hydrogels, hydrogels such as hydrazone, oxime, and thioester contain chemically crosslinked hydrogels with dynamic covalent bonds, creating a covalent adaptable network that possesses viscoelasticity. Morgan et al. tuned the mechanical properties of the oxidized alginate hydrogels by mixing with different ratios of dihydrazide (to form hydrazone) and bishydroxlamine (to form oxime) to alter the dynamic covalent crosslinks [135]. In general, the more oxime crosslinks, the stiffer the gel (larger storage modulus). A similar trend was found in the viscosity (loss modulus or relaxation time) of the gels. By changing the composition of crosslinks, the viscoelasticity can also be tuned. Richardson et al. synthesized a range of hydrazone crosslinked polyethylene glycol hydrogels [136]. By adjusting the ratio of alkyl-hydrazone and benzyl-hydrazone crosslinks, the average stress relaxation time of the hydrogels varied from hours (e.g., 4.01 × 10³ s) to months (e.g., 2.78 × 10⁶ s). Pauly et al. prepared agarose hydrogels containing proteoglycan mimetic graft copolymers with various polysaccharide side chains (dextran, dextran sulfate, heparin, chondroitin sulfate, and hyaluronan) [137]. Agarose gels have a strain-rate-dependent compressive modulus. When either the highly charged polysaccharide heparin or the neutral polysaccharide dextran is added to the gel, the modulus of the hydrogel is unmodified or reduced; however, when the heparin or dextran additive is included in the form of a proteoglycan-mimetic graft copolymer, the modulus is increased. The gels also exhibit stress relaxation behaviors with multiple time constants for relaxation that can be modulated by the structure and composition of the proteoglycan mimic additives.

While hydrogels are the main type of biomaterials used for viscoelastic studies in the literature, there are a limited number of studies investigating the viscoelastic property of synthetic scaffolds. For instance, the viscoelasticity of PCL scaffolds can be tuned by blending natural or synthetic components at different ratios. Kim et al. attempted to tune the viscoelasticity of PCL scaffolds by adding different concentrations of alginate. They showed that the fluidic viscosity of the scaffold increased by increasing the alginate weight fraction in the composites. The storage modulus (G′) of the blended scaffolds was higher than that of pure PCL scaffolds, and it was increased with an increasing alginate concentration (0.1 Pa to 40 Pa at 0–30 wt % of alginate) [42]. Moreover, Peter et al. reported the preparation of a wide range of viscoelastic polydimethylsiloxane (PDMS) scaffolds, and tuning viscoelasticity was achieved by changing the base:crosslinker ratio of Sylgard 184 and the ratio of Sylgard 184 and Sylgard 527 [40]. Increasing the ratio of Sylgard 184 and Sylgard 527 caused decreases in the storage modulus (G′) and loss modulus (G″) of the scaffolds. The use of synthetical biomaterials can overcome the limitations of most natural-material-based hydrogels, i.e., the achieved viscoelasticity range is relatively small and in a sub-physiological range (i.e., lower elasticity and viscosity than native tissues). Shamsabadi et al. used the microsphere sintering technique to fabricate scaffolds for bone tissue engineering using PCL and bioactive glass (BG) 58S5Z (58S modified with 5wt% zinc) [41]. The viscoelastic behavior of the 0% BG (scaffold with only PCL) and 5% BG samples was determined by performing compressive stress relaxation tests. The storage modulus for both samples increased with the frequency. The loss modulus of the 5% BG sample was higher only for frequencies <0.4 Hz. The smaller loss modulus for the 5% BG at higher loading rates indicated its lower viscosity, and because of this, its storage modulus remained nearly constant in this range. Mondesert et al. fabricated fibrous scaffolds with repetitive honeycomb patterns. The relaxation of the scaffolds was tested in directions D1 and D2 at a 15% strain [138]. The scaffolds exhibited a slight relaxation in both directions, showing that the viscosity of the material did not drastically influence the mechanical behavior. Hence, the viscous behavior of these scaffolds was neglected while analyzing their mechanical properties.

Recent pioneering work has revealed some new findings on the impact of substrate dynamic mechanical behavior (viscoelasticity) on various cellular behaviors, including cell morphology and spreading, migration, proliferation, differentiation, and ECM deposition.

Cell spreading is closely related to cell–matrix interactions, which affect the distribution of cell traction forces and mechanotransduction pathways and maintain the mechanical homeostasis of the cell. To examine how cell spreading is influenced by matrix viscoelasticity, Cameron et al. modulated the viscosity (the loss modulus) of polyacrylamide (PAM) hydrogels while maintaining the same elasticity (storage modulus) to study the spreading effect of hMSCs on these hydrogels [139]. Increasing the loss moduli significantly decreased the length of the focal adhesions (FAs), which affected the spreading of the cells. The smaller size of the FAs in hMSCs on more viscous substrates showed that the FAs were less mature and more transient, indicating that the hMSCs were more motile or actively spreading. An additional study with RGD (Arg-Gly-Asp)-coupled alginate hydrogels showed that viscoelastic hydrogels induced a larger spreading area of human MSC than elastic hydrogels while keeping the initial modulus or ligand density constant [140]. Scaffolds with increased creep better promoted the spreading of MSCs on a 2D culture [141]. Similar findings were observed in the 3D culture of MSCs. Enhanced creep led to the increased spreading and osteogenic differentiation of MSCs in the 2D culture, and the increased substrate stress relaxation promoted cell spreading and proliferation in the 2D culture and altered the cell morphology in the 3D culture [142]. In accordance with this, the promotion of cell spreading on various viscoelastic substrates has been reported in other cell types such as U2OS cells [140], myoblasts [143], and fibroblasts [142], in both 2D and 3D cell cultures. Moreover, substrate viscoelasticity also plays a regulatory role in cell migration, and substrates with faster stress relaxation promote the migration of cells such as myoblasts [143] and fibroblasts [142].

Both regulatory effects may be explained by focal adhesion (FA) formation and ligand clustering [128]. FA formation is probably the key mechanism through which the viscoelastic property of the substrate affects cell behaviors [140]. For instance, promoted FA formation was observed in hydrogels with faster relaxation (more viscoelastic). Chaudhuri et al. used hyaluronic acid and collagen I to form 3D hydrogels and found that the FA in MSCs was promoted by more viscoelastic hydrogels. The increased accumulation of β1 integrin, indicative of increased FA formation, was observed in the periphery of MSCs encapsulated in RGD-coupled ionically crosslinked alginate hydrogels with faster stress relaxation [142].

Viscoelastic matrices promote cell proliferation. Chaudhuri et al. showed that MSC proliferation was elevated in a PAM-alginate hydrogel with a faster relaxation rate [142]. Ryan et al. modified collagen hydrogels with insoluble elastin to induce prolonged stress relaxation (i.e., reduced viscosity), which resulted in lower proliferation and a more contractile phenotype of human smooth muscle cells (SMCs) [144]. Chao et al. seeded chondrocytes in chitosan-modified PLCL scaffolds with a viscoelastic property close to that of native bovine cartilage and observed that the cell proliferation was higher compared with that in unmodified (non-viscoelastic) scaffolds [145]. Peter et al. seeded preosteoblast cells (MC3T3-E1) on alginate-blended PCL scaffolds, and increased cell proliferation was found on viscoelastic scaffolds compared to pure PCL (low-viscoelasticity) scaffolds [42]. Finally, Tamate et al. showed that the proliferation of HeLa cells (cancer cells) was inhibited when the viscosity of the hydrogel was diminished [146]. The above studies all consistently demonstrated that substrate viscosity promotes cell proliferation in a variety of healthy and cancer cells.

The effect of substrate viscoelasticity on cell differentiation has been mostly studied in MSCs and the application of orthopedic tissue regeneration. For example, hydrogels with rapid stress relaxation induced the greater osteogenic differentiation of MSCs [147,148,149]. Viscoelastic hydrogels have also been successfully applied to regulate cell–cell and cell–matrix interactions for the differentiation and regeneration of bone and cartilage tissues with MSC spheroids [147,150]. The improved osteogenic differentiation of MSCs in faster relaxing (more viscoelastic) substrates has been related to mechanotransduction regulators such as the enhanced clustering of integrin ligands or stronger actomyosin contractility [142]. Li et al. prepared PAM hydrogels with different substrate stiffness to study cell proliferation. The substrate with slower stress relaxation drove the pro-inflammatory polarization of human bone-marrow-derived monocytes and their differentiation into antigen presenting cells, indicating an anti-inflammatory role of viscoelastic substrates [151].

ECM deposition is a key outcome in the regeneration of connective tissues including bone and cartilage. Chondrocytes encapsulated in scaffolds with similar viscoelasticity to native cartilage tissue displayed the greater deposition of a cartilage-like matrix composed of type 2 collagen and aggrecan and the lower expression of type 1 collagen [152]. MSCs encapsulated in a viscoelastic hydrogel consisting of an interpenetrating network of alginate and fibrillar collagen type I with interferon-γ (IFN-γ)-loaded heparin-coated beads suppressed the proliferation of human T cells [153]. However, the results showed that cell proliferation was independent of substrate stiffness and was more dependent on the crosslinking components of the hydrogel.