Bio-based composite hydrogels for biomedical applications

Hydrogels are polymeric networks based on hydrophilic macromonomers that are able to retain large amounts of water [1–5]. They generally exhibit excellent biocompatibility and their mechanical properties can be designed in such a way that they match those of many biological tissues. Furthermore, their soft and rubbery nature minimizes inflammatory reactions of the surrounding cells and tissues. Hydrogels are consequently used for a wide range of biomedical applications, including controlled drug delivery and tissue engineering. They can be physically crosslinked by noncovalent interactions, chemically crosslinked by covalent bonds, or crosslinked by a combination of both.

Hydrogels may be classified as natural or synthetic, depending on the source of the constituting polymers. Synthetic polymers allow for a high control over the macromonomer structure and properties, but their synthesis often relies on petroleum-derived chemistry, refinery, and engineering processes which may have a negative impact on the environment [6]. In particular, since the late 1990s, the polymer industry has faced two serious problems: global warming and depletion of fossil resources. One solution to address these problems is to use sustainable resources instead of fossil-based resources. Biomass feedstock can be converted into raw materials for polymer production, and the resulting polymers are called bio-based polymers. Natural polymers formed by plants, microorganisms, and animals constitute an important class of bio-based polymers [7]. Such naturally derived biomass polymers offer great promise for reducing the environmental impact of the polymer industry and for realizing a bio-based and sustainable society. In addition to the environmental benefits, bio-based polymers also offer a high level of biocompatibility and biomimicry, thereby providing biophysical and biochemical properties which are able to induce the correct biological response for both in vitro and in vivo biomedical applications [8, 9]. Moreover, these natural macromolecules are biodegradable and many of them also possess intrinsic antibacterial and anti-inflammatory properties. The present review focuses on bio-based hydrophilic polymers, including naturally derived biomass polymers that have been slightly chemically modified, such as methylcellulose.

Composite materials consist of two or more constituent materials with markedly different physicochemical properties [10]. When the components are combined in a single entity where the individual components remain distinct and separate, the resultant composite material can exhibit different characteristics from the individual components and offer synergistic properties. Importantly, composite materials may also perform several tasks via ingenious combinations of multiple functional capabilities [11]. Early hydrogels were mostly based on a single (co)polymer and often designed to perform only one task [2]. Over the past few decades, hydrogel research has focused increasingly on composite hydrogels which offer more opportunities to mimic the composite nature of biological tissues.

This review discusses recent approaches aimed at producing composite bio-based hydrogels for biomedical applications. First, an overview is given of the bio-based polymers that are treated in this review, followed by descriptions of the use of polymeric micro- or nanoparticles embedded in hydrogels, as well as the use of electrospun fibers in hydrogels to impart desired properties. Subsequently a separate chapter is dedicated to bio-based hydrogels incorporating cellulose fibrous nanostructures with remarkable properties. Lastly, the author's vision on the future of this rapidly evolving area of research is presented.

Several bio-based polymers have been explored for the preparation of hydrogels and micro- or nano-sized functional structures such as particles and fibers. These polymers, generally polysaccharides or proteins, are derived from plants, microorganisms or animals and offer the advantages of biodegradability and inherent biocompatibility. Table 1 presents an overview of the most important polysaccharides and protein-based polymers that are addressed in this review. They will be briefly discussed below.

2.1. Polysaccharides

2.1.5. Hyaluronic acid

2.2. Protein-based polymers

Hydrogels are attractive materials for the controlled release of drugs because of the possibility to incorporate therapeutic agents in their water-swollen network. Hydrogels have been used for localized and sustained release as a strategy to decrease the number of drug administrations, to protect the drug from degradation and to allow for therapeutic drug concentrations for prolonged times [1]. Hydrogels were shown to be very useful for the controlled release of water-soluble compounds, but the hydrophilic nature of their matrices makes them unsuitable for the incorporation and release of hydrophobic drugs [49]. Typically, such poorly water-soluble agents require encapsulation in micro- or nanosized particles to achieve controlled release in biological systems. This review will focus on two types of polymeric particles: micelles and microspheres. Micelles are nano-sized colloidal particles with a core–shell structure, which are spontaneously formed by the self-assembly of amphiphilic macromolecules in a block-selective solvent into nanoaggregates above a certain concentration, the critical micelle concentration [50]. Microspheres are microparticles that consist of a solid and homogeneous polymer matrix [51]. In order to obtain a controlled and sustained delivery of hydrophobic drugs from hydrogel-based devices, various researchers have followed the strategy to load the drugs in particles, followed by loading of these particles in the hydrogel matrix. For example, Kundu et al reported on a micelle/hydrogel composite system based on gellan gum for oral drug delivery [52]. By conjugating alkyl chains to the gellan gum backbone via an etherification reaction, amphiphilic copolymers were obtained. In aqueous solution these copolymers self-assembled into spherical micelles, which were able to encapsulate simvastatin, a hydrophobic drug used in the treatment of hypercholesterolemia. Composites were prepared by incorporating micelles in a native gellan gum solution, which was subsequently crosslinked by adding aluminium chloride resulting in ionic interactions between aluminium ions and the negatively charged gellan gum chains. Simvastatin-loaded native hydrogels exhibited a lower degree of swelling than the composite hydrogels containing simvastatin-loaded micelles, which was ascribed to the hydrophobicity of the dispersed drug in the monocomponent hydrogel. The in vitro simvastatin release was slower from native hydrogels than from composite hydrogels due to the combined effects of low hydrogel swelling and low drug solubility. The relatively fast release from the composites was explained by a mechanism in which the hydrogels swell, disintegrate and then release the drug-loaded micelles. Following oral administration in rabbits, drug-loaded micelle/hydrogel composite beads were able to decrease cholesterol blood levels to a higher extent than drug-loaded native hydrogel beads.

In addition to having cholesterol-lowering effects, simvastatin has been suggested as a potential agent to promote bone growth. Yan et al prepared micelle/hydrogel composites for the controlled release of simvastatin to achieve local bone regeneration (figure 1) [53]. Micelles were prepared from maltodextrin, a polysaccharide produced from starch by partial hydrolysis, whereas hydrogels were based on carboxymethyl chitosan. Esterification between maltodextrin and palmitic acid resulted in amphiphilic micelle-forming polymers, which were subsequently treated with an oxidizing agent in order to cleave a number of glycosidic bonds of the polymer backbone, yielding free aldehyde groups. Mixing aldehyde-functionalized micelles, aldehyde-functionalized maltodextrin and amine-bearing carboxymethyl chitosan resulted in the formation of a hydrogel network via Schiff base formation. As the micelles participated in the crosslinking reaction, they were anchored to the hydrogel network via chemical bonds. The gelation time could be tuned from 40 to 80 s by adjusting the degree of aldehyde functionalization of the micelles. Rheological measurements showed that the storage modulus of composites containing chemically anchored micelles was twice as high compared to control composites containing physically encapsulated micelles. Direct loading of simvastatin in the hydrogel resulted in phase separation between the hydrophobic drug and the matrix, as evidenced by a large number of visible drug aggregates and an opaque appearance (figure 1). In contrast, hydrogels with simvastatin-loaded micelles showed good transparency, demonstrating the encapsulating effect of the micelles. The presence of chemical crosslinks between the micelles and the hydrogel network resulted in a more sustained release of micelles and their encapsulated simvastatin from the matrix compared to the control system with physically embedded micelles. In vitro experiments showed that the prolonged release of simvastatin from composites with chemically linked micelles facilitated the osteogenic differentiation of mouse osteoblast precursor cells.

Zoom In Zoom Out Reset image size

A composite delivery system may also solve the issue of fast, uncontrolled dug release from particles alone [54]. A more sustained overall drug release may be achieved via (1) release of drug-loaded particles from the hydrogel and subsequent release of the drug from the particles or via (2) release of the drug from the entrapped particles and subsequent drug diffusion through the polymer matrix of the hydrogel. Which mechanism dominates will depend on various parameters such as the size of the hydrogel mesh and the particle, the hydrophilicity of the drug, as well as the chemical composition of the particle and hydrogel. Zhu et al reported on a composite microsphere/hydrogel system for the controlled release of the hydrophilic drug berberine hydrochloride, which has anti-cancer, anti-inflammatory and antibiotic effects [55]. The drug was loaded via an emulsification method into alginate microspheres, which were crosslinked with calcium ions. The drug-loaded microspheres (10–15 µm in size) were subsequently entrapped in a glutaraldehyde-crosslinked carboxymethyl chitosan hydrogel. Whereas the drug release into PBS (pH 7.4) from the alginate microspheres or the chitosan hydrogel alone showed a large burst release, berberine hydrochloride was released at a nearly constant, zero-order rate from the composite system. When the pH of the release medium was decreased to 1.2, the release rate increased, which was ascribed to neutralization of the pendant COO^- groups in the composite system, resulting in expulsion of the drug solution and hydrogel shrinkage. In addition to having advantageous effects on the drug release behaviour, the presence of microspheres increased the compressive strength of the hydrogels by 50%.

Cong et al reported on a pH-responsive composite system consisting of chitosan micelles embedded in alginate hydrogel beads (150 µm) for the oral delivery of the hydrophobic drug emodin, having antioxidant and antimicrobial activities, to the intestine [56]. Micelles were prepared by mixing calcium chloride and chitosan solutions, leading to intrachain contraction of the cationic, amine-bearing chitosan chains by chloride anions and the formation of a crosslinked unimolecular micelle structure. Micelle-loaded hydrogel beads were obtained via dropwise addition of a micelle-containing sodium alginate solution into a calcium chloride solution. Transmission electron microscopy showed that the micellar diameter increased from 80 nm in aqueous solution to 100–200 nm when incorporated in hydrogel beads. The size increase was ascribed to the formation of a polyelectrolyte layer on the surface of the micelles. This layer was thought to originate from electrostatic interactions between carboxyl groups on alginate chains not participating in the hydrogel crosslinking and amino groups on chitosan chains not participating in the micellar crosslinking. Release experiments were performed by placing the micelles or micelle/hydrogel composites consecutively in simulated gastric fluid (pH 1.2, 2 h), simulated small intestinal fluid (pH 6.8, 3 h) and simulated colon fluid (pH 7.2, 4 h) to mimic the conditions in the gastro-intestinal tract. The emodin release from micelles alone was characterized by a burst release (80% release within 2 h) due to dissociation of the micelles at acidic pH caused by strong electrostatic repulsion among the protonated amine groups. In sharp contrast, little burst release at pH 1.2 (<20% for most formulations) was observed when the micelles were incorporated in the alginate hydrogel beads. The protective effect of the hydrogel regarding micelle degradation and drug release was maintained when the composites were transferred to pH 6.8 as the cumulative release increased only marginally during 3 h. At pH 7.4 however the cumulative release increased to 80% within 4 h. This was explained by degradation of the hydrogel caused by strong ionic repulsions between the alginate chains at pH 7.4, leading to release of the micelles from the hydrogel and subsequent release of the drug from the micelles.

Composites of polymeric micro- or nanoparticles embedded in hydrogels may also facilitate the simultaneous delivery of multiple drugs with different solubility. This is important to enable combination therapy, which has shown several advantages compared to monotherapy, such as reduced side effects [57] and lowering of drug resistance [58]. Zhong and co-workers prepared a dual drug delivery system by incorporating polylactide (PLA) microspheres in calcium crosslinked alginate hydrogel beads [59]. PLA is a widely investigated biodegradable polyester that can be derived from renewable natural resources [60]. The hydrophobic drug glycyrrhetinic acid (GA), having potential antiviral and antibacterial properties, was encapsulated in the PLA microspheres, whereas albumin was loaded in the hydrogel as a hydrophilic model protein drug. In this way, a more sustained GA release and a less significant burst were achieved compared to release from PLA microspheres alone. In vitro release experiments showed that albumin was released within 15 d, including an initial burst, whereas the delivery of GA was more sustained, with a release for multiple weeks. The release of albumin largely accelerated after increasing the PLA/alginate ratio, whereas at the same time, the release of GA decreased. Independent control over the individual drug releases, which is essential for successful combination therapy, was, however, not reported.

For biomedical applications, the use of in situ forming hydrogels is preferred over preformed hydrogels since there is no need for surgical interventions as gelation can take place under physiological conditions upon injection. Furthermore, the initial fluidic nature of the precursor solution ensures proper shape adaptation and biological components as well as particles can be incorporated in the hydrogel by simple mixing with the precursor polymer solution [61, 62]. Chen et al designed an injectable composite hydrogel of gelatin microspheres in chitosan-alginate hydrogels for the delivery of the anti-cancer drug 5-fluorouracil (figure 2) [63]. The hydrogel was crosslinked via Schiff base reaction between amino groups on carboxyethyl chitosan and aldehyde groups on oxidized alginate. The gelation time decreased from 150 to 75 s when drug-loaded gelatin microspheres were added to the hydrogel precursor solution, which was attributed to microsphere co-crosslinking via reaction of amino groups on the gelatin with aldehyde groups on the oxidized alginate. In addition to in situ gel formation, the composite hydrogel also showed self-healing behaviour, which was ascribed to the dynamic, reversible character of the Schiff base formation under selected conditions. The composite hydrogel showed a controlled release of 5-fluorouracil during 5 weeks, which was more sustained than the release from the hydrogel or the microspheres alone. Interestingly, incorporating magnetic Fe₃O₄ nanoparticles in the gelatin microspheres allowed for acceleration of the drug release by applying a magnetic field. The same group also reported on injectable hydrogels consisting of chitosan microspheres in carboxymethyl chitosan hydrogels for the release of albumin [64].

Zoom In Zoom Out Reset image size

Whereas the above examples focused on controlled drug delivery, bio-based particle/hydrogel composites have also been employed for other biomedical applications. Tissue engineering applies the principles of biology and engineering to the development of functional substitutes for damaged tissue [65]. From a biological point of view, hydrogels are attractive candidates for tissue engineering constructs because they can provide an aqueous 3D environment for cells, simulating the natural extracellular matrix. Also, viscoelastic properties of hydrogels can be tuned to influence cell fate and activity. Hybrid approaches offer unprecedented opportunities to mimic biological tissues with biomaterials [66, 67]. In particular, the incorporation of nano- or microparticles in hydrogels may enhance their mechanical properties and their biological performance. To mimic the complex nature of tissues, additive manufacturing has emerged as a valuable technology to generate bioengineered 3D structures [68]. In this approach, termed biofabrication, biological structures for tissue engineering are created by a computer-aided manufacturing process for patterning and assembling living and non-living materials with a prescribed 3D organization. The inks for biofabrication are typically based on hydrogels. However, 3D printed hydrogel scaffolds often suffer from low printing accuracy and poor mechanical properties because of their soft nature and tendency to shrink. To address this challenge, Zhang et al prepared composite scaffolds by incorporating silk microparticles in chitosan-based bioinks (figure 3) [69]. A bioprinter was used to extrude the composite bioink layer-by-layer, followed by manually pipetting NaOH solution onto the printed filaments to solidify the construct via crosslinking of the chitosan chains. In comparison with pure chitosan scaffolds, the addition of silk microparticles resulted in a 5-fold increase of the compressive modulus as well as an improved printing accuracy and a higher resistance against degradation. To better understand the improvement in mechanical properties resulting from the presence of silk microparticles, the scaffold stiffness data was fitted to existing empirical models for the prediction of the performance of particle-polymer composites. It was found that the silk microparticles led to a modulus enhancement that matched those predicted for longer fibers, confirming that the particles act as very efficient reinforcing agents in chitosan scaffolds. The composite hydrogel scaffolds showed no cytotoxicity and supported adherence and growth of human fibroblast cells.

Zoom In Zoom Out Reset image size

Whereas the number of publications on 3D printed hydrogels for tissue engineering is rapidly increasing, research concerning drug delivery from these constructs is still very limited, especially in the case of 3D printed bio-based composite hydrogels. Poldervaart et al developed a hydrogel scaffold for bone tissue engineering based on calcium crosslinked alginate [70]. The bone growth factor BMP-2 was loaded in glutaraldehyde-crosslinked gelatin microspheres, which were in turn dispersed in the alginate bioink prior to biofabrication. Using this approach a more sustained BMP-2 release was achieved in vitro in comparison with direct incorporation of BMP-2 in the alginate hydrogel precursor. The BMP-2 retained its bioactivity after biofabrication as demonstrated by in vitro osteogenic differentiation of embedded goat multipotent stromal cells. The release of BMP-2 from the 3D printed hydrogels significantly stimulated bone formation in vivo, but the loading of BMP-2 in the gelatin microspheres instead of direct loading in the hydrogel did not have an influence on the bone growth in the scaffolds.

An emerging topic in the field of bio-based, particle-containing hydrogels is the development of nanocomposite hydrogels consisting of a polymer matrix incorporating inorganic nanoparticles. The interactions between the nanoparticles and the hydrogels, such as coordination bonds and electrostatic interactions, may give rise to unique composite hydrogel properties. For example, Bian and coworkers prepared composite hyaluronic acid hydrogels for the simultaneous release of bioactive ions and small molecule drugs [71]. Mixing pamidronate-functionalized hyaluronic acid, pamidronate and MgCl₂ in PBS yielded self-assembled pamidronate-Mg nanoparticles thanks to the effective coordination between pamidronate and Mg²⁺. These nanoparticles functioned as a crosslinker to stabilize the hyaluronic acid network, resulting in the formation of pamidronate-Mg-hyaluronic acid nanocomposite hydrogels within 30 s, which allowed for in situ injection in bone defects. The release of Mg²⁺ ions from the hydrogels promoted both osteogenic differentiation of encapsulated human mesenchymal stem cells and activation of alkaline phosphatase. The latter catalysed the dephosphorylation of dexamethasone phosphate, a prodrug of dexamethasone, and therefore promoted the release of dexamethasone from the hydrogels to further stimulate osteogenesis. Thanks to this positive feedback circuit, the nanocomposite hydrogels promoted bone formation in vivo in rabbits. The same group reported on bisphosphonate-Mg-hyaluronic acid nanocomposite hydrogels, which also allowed for injection and encapsulation of cells [72].

Nejadnik et al reported on nanocomposite hydrogels based on reversible bonds between calcium phosphate nanoparticles and bisphosphonate-functionalized hyaluronic acid [73]. The injectable nanocomposites demonstrated adhesiveness to Ca²⁺ containing mineral surfaces such as hydroxyapatite and enamel. These non-covalently crosslinked composites showed robust yet biodegradable upon in vitro and in vivo testing and displayed bone interactive capacity as evidenced by bone ingrowth into the material.

Whereas micelles, microspheres and inorganic nanoparticles have been frequently used for the preparation of composite bio-based hydrogels, as discussed above, solid polymeric nanoparticles have not yet been explored in combination with bio-based hydrogels. Compared to nm-sized micelles, solid polymeric nanoparticles are potentially more resistant against disintegration as micelles have a dynamic nature and may destabilize prematurely resulting from dilution below the critical micelle concentration [74]. Compared to µm-sized polymeric microspheres, solid polymeric nanoparticles have a higher mobility in biological tissues and may potentially cross biological barriers by virtue of their smaller size [75]. In view of these advantageous features, future research should therefore also focus on bio-based composite hydrogels containing solid polymeric nanoparticles. Emphasis should thereby be placed on the biocompatibility of such systems since concerns have been raised regarding the in vitro and in vivo toxicity of solid polymeric nanoparticles [76].

The high water content of hydrogels endows them with various advantageous properties, including biocompatibility as well as the possibility to embed and release biologically active molecules. However, using hydrogels for mechanically demanding tasks is challenging, as the large volume fraction of water may result in weakness, compliance and brittleness. In order to obtain systems with improved mechanical properties without compromising cellular viability, composites of hydrogels and reinforcing materials, most often fibers, have been developed. Electrospinning is most commonly used to produce fibrous materials because it is a straightforward and effective method to prepare thin fibers with diameters ranging from micrometres to a few nanometres at low costs and high production rates. This electrostatic processing method uses a high-voltage electric field to form solid fibers from a polymeric fluid stream delivered through a millimetre-scale nozzle [77].

This section focuses on recent examples of bio-based hydrogels containing mainly non-cellulose fibers; the emerging field of nanocellulose-containing bio-based hydrogels will be covered separately in the last section of this review. The reader is referred to a number of excellent reviews concerning hydrogels with incorporated electrospun fibers that focus on specific application areas such as tissue engineering [77–79], osteochondral regeneration [80] and wound dressings [81].

Since the mechanical performance of hydrogel scaffolds is often inferior to that of native tissues, the development of mechanically robust hydrogels is especially important to provide optimal performance in tissue engineering applications. The group of Oyen prepared fiber-reinforced hydrogels using gelatin and alginate to mimic the microstructure and composition of the extracellular matrix of soft tissue (figure 4) [82]. Gelatin was electrospun into nanofibers and infiltrated with alginate solution, which was ionically crosslinked with calcium ions. The composites were further crosslinked using 1-ethyl-3-(-3-dimethylaminopropyl) carbodiimide/N-hydroxysuccinimide (EDC/NHS) chemistry to induce covalent bond formation. Composites contained fibers that were in a single layer or stacked in multilayer laminates with varying fiber orientations. Mechanical tests showed that the tensile strength of native alginate hydrogels (approximately 100 kPa) could be enhanced by up to two orders of magnitude by adding gelatin fiber mats. The larger the fraction of fibers in the direction of the applied load, the stiffer and stronger the composite hydrogels were. Without fiber reinforcement, alginate hydrogels exhibited brittle failure at small strains with a tear toughness of 40 J m⁻². Incorporation of gelatin fibers as multilayer laminates with different fiber orientations for each layer increased the toughness by two orders of magnitude compared with the unreinforced hydrogel. This was explained by a fiber-induced crack-deflecting mechanism which dissipates energy and locally reduces stresses. In addition, interfacial debonding was stated as energy dissipating mechanism that contributes to the increase in toughness. The composite gelatin fiber/alginate gel scaffolds closely matched the mechanical properties and transparency of porcine cornea, showing their potential for corneal tissue engineering [83].

Zoom In Zoom Out Reset image size

Other recent examples of fiber-reinforced bio-based composites for tissue engineering include PLA fiber/chitosan hydrogel [84], PLA fiber/hyaluronic acid-graft-alginate hydrogel [85] and collagen fiber/alginate hydrogel [86] systems.

The preparation of mechanically robust, injectable hydrogels can be challenging because the hydrogel precursors may diffuse away from the site of injection, thereby decreasing the effective polymer concentration. In addition, the in vivo setting limits the application of highly reactive, toxic crosslinking agents and end groups. With the goal of enhancing the mechanical properties of injectable gelatin hydrogels, Poveda-Reyes and co-workers investigated two methods for the fabrication of reinforcing fibers made from PLA (figure 5) [87]. One method consisted of manufacturing electrospun meshes followed by milling, whereas the second method involved precipitating a PLA solution in a highly turbulent non-solvent bath. In both cases, fibers of approximately 50 µm in length and 1 µm in diameter were obtained. To prepare composites, the fibers were dispersed in a solution of tyramine-grafted gelatin which was subsequently enzymatically crosslinked using hydrogen peroxide and horseradish peroxidase. At low concentrations (<1.5 w/v%) no limitations regarding injectability were observed and the high crosslinking rate (gelation time <1 min) prevented the precipitation of the fibers in the hydrogel precursor solution. A triple increase in the Young's modulus of the gelatin hydrogel was obtained when electrospun, milled fibers were incorporated, which was explained by a reduction in the water permeability and flow in the hydrogels. In contrast, no enhancement in Young's modulus was observed when precipitation-fabricated fibers were embedded. This was ascribed to the heterogeneous morphology and rough surface of the fibers leading to the formation of fiber aggregates, which do not influence water flow in the hydrogel. Cells were incorporated in the composites by adding them to the gelatin-fiber dispersions prior to crosslinking. It was found that the injectable systems hydrogels are not cytotoxic and induce steady cell growth.

Zoom In Zoom Out Reset image size

Mirahmadi et al reported on thermosensitive silk/chitosan hydrogel composites for cartilage tissue engineering [88]. Composite solutions were prepared by incorporating either chopped silk fibers or electrospun silk fiber mats in aqueous chitosan chloride solutions containing sodium glycerophosphate. Increasing the temperature from ambient to body temperature resulted in the formation of composite hydrogels. The thermosensitive gelation behavior was ascribed to e.g. increased hydrophobic interactions and reduced chitosan solubility at higher temperatures [89]. The presence of silk fibers significantly increased the compressive modulus of the chitosan hydrogels, with a more pronounced effect for silk fiber mats compared to dispersed silk fibers. This was attributed to the higher surface-to-volume ratio of the nanostructured fibers in the mats, resulting in more interfacial interactions. Chondrocytes were incorporated in the constructs by adding them to the hydrogel solutions before temperature-induced gelation. It was found that the collagen content in silk fiber mat/chitosan hydrogel composites was higher than in the native chitosan hydrogel and in composite hydrogels containing dispersed fibers, suggesting a correlation between mechanical properties and collagen deposition. Although the mechanical and biological properties of the silk fiber/chitosan hydrogel composites make them promising for tissue engineering, the reported gelation time of 30 min is likely too slow for clinical application.

Regev et al prepared composite hydrogels by dispersing electrospun albumin fibers in aqueous gelatin solutions, which were crosslinked by aldehyde-functionalized dextran at 37 °C [90]. The flow parameters of the hybrid hydrogel precursor solution were similar to those of the native hydrogel precursor solution. The incorporation of albumin fibers increased the elastic modulus of the native gelatin/dextran formulation by 40% and decreased the gelation time by 20%, which was ascribed to additional Schiff base reaction between the amine groups of albumin and the aldehyde groups of the modified dextran. In vitro studies showed that the composite albumin/gelatin/dextran hydrogels are biocompatible and support fibroblast growth.

In addition to enhancing the mechanical properties of native, single component hydrogels, the incorporation of electrospun fiber mats in hydrogels can also be exploited to control the behaviour of incorporated cells. This was demonstrated by Yang et al, who embedded highly aligned electrospun PLA nanofiber meshes in collagen hydrogels via a bottom-up, layer-by-layer assembly process, resulting in highly organized composite hydrogels (figure 6) [91]. The fibers dictated the orientation of the cytoskeleton of individual cells in a precise manner, allowing for control over the orientation of a cell population throughout the thickness of the hydrogel. The addition of fibers could also be used to direct the cell phenotype and the secretion of matrix proteins.

Zoom In Zoom Out Reset image size

Although bio-based polymers generally show good biocompatibility when they are applied as macroscopic biomaterials, the nanotopological features of bio-based polymeric fibers, including diameter and surface morphology, may negatively influence cell behaviour. This was demonstrated by Shoichet et al, who reported that the presence of electrospun collagen fibers in hyaluronic acid/methylcellulose hydrogels was deleterious to neural stem/progenitor cell (NSPC) proliferation and differentiation [92]. In contrast, composites with synthetic poly(-caprolactone-co-lactide) fibers maintained NSPC viability and guided cellular differentiation to neurons and oligodendrocytes, similar to native hydrogels. The inability of incorporated collagen fibers to support differentiation of NSPCs was attributed to their fine, fragmented and tangled structures. This demonstrates that the nanotopological characteristics of fibers should be taken into account as an important parameter when designing and evaluating bio-based fiber/hydrogel composites for tissue engineering applications.

Bioadhesives and sealants for wound closure and healing applications are promising alternatives when other techniques, such as stapling or suturing, are impractical or inefficient. Pinkas and Zilberman investigated the effect of cellulose fibers on the properties of a hydrogel formulation based on a combination of gelatin and alginate which was crosslinked with a water-soluble carbodiimide [93]. The incorporation of fibers (average length of 300 µm) significantly increased the cohesive properties of the hydrogel, most notably the burst strength i.e. the maximum pressure the hydrogel could withstand. This was ascribed to penetration of the gelatin and alginate formulation into pores of the cellulose fibers, enabling a good fiber-matrix interaction. Although the adhesive properties of the composite hydrogel were not tested under biologically relevant conditions, in previous studies the non-reinforced gelatin-alginate hydrogel showed good bonding strength to porcine skin as well as a high biocompatibility [94, 95], suggesting the potential of these fiber-reinforced composite hydrogels for bioadhesive and sealant applications.

In the field of controlled drug release, obtaining extended, uniform release with little initial burst has been challenging, especially for macromolecular therapeutics such as antibodies. To address this challenge, the preparation of electrospun fiber/hydrogel composite materials is an attractive approach, since the electrospun fiber mat may provide an additional diffusion barrier that can alter the release kinetics [96]. Elia et al prepared composites of electrospun silk fibroin mats and hydrogels formed by crosslinking thiol-modified hyaluronic acid with poly(ethylene glycol)-diacrylate [97]. Release experiments were performed using fluorescein isothiocyanate-dextran (FITC-dextran) as a model compound in the presence of the enzyme hyaluronidase, which is capable of degrading hyaluronic acid. Low molecular weight FITC-dextran diffused freely and quickly out of the highly porous hydrated gel, whereas high molecular weight FITC-dextran was released more slowly and involved both diffusion and enzymatic degradation of the hydrogel. To further evaluate the feasibility of the composites for controlled drug delivery, several anti-inflammatory drugs such as dexamethasone and prednisolone were dissolved in the precursor solutions prior to crosslinking and their in vitro release was measured in the absence of enzymes. All drugs exhibited first order release kinetics and complete release was achieved in several hours or several days, depending on the drug size and hydrophilicity. The observed release rates from the composite hydrogels were lower than those from hydrogels lacking a fiber mat, indicating that the drugs were retained for a longer period of time in the composite systems. This was ascribed to direct drug entrapment within the silk fibers or to swelling of the fibers, thereby hindering the release of drug near the silk by acting as a diffusion barrier. In addition to the relatively fast release of anti-inflammatory drugs, the composites provided a sustained release of vascular endothelial growth factor, a signal protein, for several weeks.

Natural plant and wood fibers can be processed via mechanical, chemical and enzymatic treatments to yield micro- and nanoscopic fibrils with altered mechanical properties. If the diameter of the fibrils is decreased to the nanoscale, they may exhibit remarkable reinforcing effects thanks to the presence of crystalline regions with a very high mechanical strength. Generally two types of such cellulose-based nanostructures are distinguished: cellulose nanocrystals and cellulose nanofibrils (figure 7). Cellulose nanocrystals are elongated, rigid rod-shaped particles with a diameter of 2–20 nm and a length of 100–600 nm [98]. They have been described in literature with various terms, including nanowhiskers and nanocrystalline cellulose. Cellulose nanofibrils, on the other hand, are more flexible and longer (diameter of 10–40 nm, length>1 µm). Cellulose nanofibrils have also been referred to in literature as nanofibrillated cellulose, microfibrillated cellulose or microfibrils, despite their nanoscale dimensions.

Zoom In Zoom Out Reset image size

Because of their high mechanical strength, biocompatibility and biodegradability, both cellulose nanocrystals and cellulose nanofibrils have been incorporated in hydrogels to efficiently impart mechanical reinforcement as well as various other desirable properties. Nanocellulose concentrations as low as 5 wt% have been shown to significantly enhance the mechanical properties of calcium crosslinked alginate hydrogels, whereas higher loadings did not lead to further improvements [100]. In addition to vegetal sources, nanocellulose can also be obtained from certain gram-negative bacteria such as Gluconacetobacter xylinus. Bacterial cellulose consists of a 3D network of interconnected cellulose nanofibers and is characterized by high strength and purity [101].

The field of nanocellulose-containing hydrogels has been reviewed a number of years ago by De France et al [102]. The following section focuses on selected, recent examples of the use of nanocellulose in bio-based composite hydrogels for biomedical applications.

Thanks to their reinforcing effects, cellulose nanostructures have been frequently embedded in hydrogels to create substitutes for intervertebral discs. Pereira et al prepared composite hydrogels by dispersing cellulose nanocrystals (100 nm in length) in gellan-gum precursor solutions [103]. The addition of phosphate buffered saline induced gelation by reduction of gellan gum helix repulsion via coordination of sodium and potassium ions with carboxylate groups on the helices. The critical gel concentration of gellan-gum decreased in the presence of cellulose nanocrystals. This was explained by the entanglement of cellulose nanocrystals between gellan-gum helices resulting in molecular bridging and chain crosslinking. The presence of cellulose nanocrystals enhanced the mechanical properties of the hydrogels such as the storage modulus and ultimate stress under compression. Furthermore, the composites matched the Young´s modulus (50 kPa) of native human intervertebral disc tissue. All hydrogels exhibited a slow degradation rate (5%–10% mass loss after 30 d), which was sufficient to allow free motion of bovine annulus fibrosus cells in the constructs. Cells cultured in the composite hydrogels showed a physiologically more relevant cell morphology (elongation and spreading) compared to cells cultured in the native gellan-gum hydrogels, which kept a round morphology after 14 d.

Similar promising results for intervertebral disc regeneration were obtained by Doench et al for physically crosslinked chitosan hydrogels reinforced with cellulose nanofibrils [104], thereby exploiting the excellent compatibility of chitosan with cellulose [99, 105]. Notably, ex vivo experiments performed in a pig vertebral unit model showed that the implantation of the composite hydrogel in a disc defect contributes to the restoration of the disc biomechanics, thereby approaching the functionality of a healthy disc.

In an attempt to improve the mechanical properties of nanocellulose-containing chitosan hydrogels for intervertebral disc regeneration and to better control their degradation, the group of Ulery applied a dual crosslinking technique by combining ionic crosslinking (using calcium phosphate) with covalent crosslinking via Schiff base formation (using genipin), resulting in a semi-interpenetrating network structure of nanocellulose and dual crosslinked chitosan [106]. As expected, ionic crosslinking, covalent crosslinking and the incorporation of cellulose nanocrystals all had an incremental effect on the compressive strength, reaching values comparative to human vertebral bone for some formulations (250–300 kPa). Composites with a very high crosslink density showed an even higher compressive strength but also brittleness and a fast degradation, making them less suitable for intervertebral disc tissue regeneration. Although this system allows for a high control over the composite hydrogel properties thanks to the various reinforcement strategies (dual crosslinking as well as the incorporation of nanocellulose), the observed cytotoxicity of the hydrogel, which may be related to the use of genipin as a covalent crosslinker, needs to be addressed for clinical translation.

A different example of a robust semi-interpenetrating network was reported by Hossen et al [107] who incorporated cellulose nanofibrils in a solution of methacrylate-functionalized carboxymethyl cellulose (figure 8). Subsequent UV irradiation led to photopolymerization of the methacrylate groups and, after freeze-drying, highly porous structures. The observed high resistance against degradation and collapse of the porous structure was ascribed to the presence of cellulose nanofibrils and chemical crosslinks, but also to hydrogen bonding between the two cellulose species present in these composite hydrogels.

Zoom In Zoom Out Reset image size

Nanocellulose has also gained interest for 3D printing of hydrogels because of its ability to improve shape reliability while printing. Most research to date focused on composites of cellulose nanofibrils or cellulose nanocrystals in alginate-based bioinks. Sultan and Mathew reported on a composite bioink consisting of cellulose nanocrystals in an alginate/gelatin hydrogel (figure 9) [108]. The composite bioink exhibited shear thinning behavior, as the viscosity decreased by four orders of a magnitude after increasing the shear rate to values typically encountered during 3D printing. The native alginate/gelatin formulation showed no shear thinning behaviour, suggesting that this formulation itself is not printable. The 3D printed structures were immersed in calcium chloride solution to ionically crosslink the alginate and subsequently in a glutaraldehyde solution to covalently crosslink the gelatin. SEM and x-ray scattering studies showed that the cellulose nanocrystals align preferably in the printing direction, which may be useful for guided cell growth during tissue regeneration.

Zoom In Zoom Out Reset image size

Abouzeid et al prepared scaffolds via 3D printing of an alginate bioink that was already partially crosslinked with calcium ions in order to prevent collapse of the filaments [109]. The prepared scaffolds were fully Ca²⁺ crosslinked immediately after the printing process to guarantee the long-term rigidity and stability of the hydrogel. Rheological experiments showed that the viscosity recovery for native alginate hydrogels was only 16% of the initial value, whereas it increased to 66% upon addition of cellulose nanofibrils to the alginate. Biomimetic mineralization of the scaffold was performed through immersion in simulated body fluid, mimicking the inorganic composition of human blood plasma. The deposition of hydroxyapatite was confirmed using SEM and infrared spectroscopy, showing the potential of cellulose nanofibrils/alginate composite bioinks for 3D bioprinting of bone tissue.

Gatenholm et al investigated the feasibility of cellulose nanofibrils/alginate composite bioinks loaded with human chondrocytes for cartilage tissue engineering [110, 111]. It was shown that the bioink was non-cytotoxic to the cells and that the bioprinting process had no influence on cell vitality. Cell-loaded, Ca²⁺-crosslinked hydrogel constructs exhibited excellent size and shape stability and induced an increase in cell viability and proliferation in vitro. Furthermore, the 3D printed composite hydrogels supported redifferentiation of chondrocytes and neo-synthesis of cartilage-specific extracellular matrix components, as demonstrated by gene expression, immunohistochemical and biochemical analyses. In another study, cellulose nanofibrils/alginate and cellulose nanofibrils/hyaluronic acid composite bioinks were loaded with chondrocytes and pluripotent stem cells and subsequently bioprinted [112]. Low proliferation and phenotypic changes away from pluripotency were observed for cellulose nanofibrils/hyaluronic acid bioinks. However, in the case of cellulose nanofibrils/alginate bioinks pluripotency was preserved and cartilaginous tissue as well as largely increased cell numbers were observed in the composite scaffolds after 5 weeks.

Nanocellulose-containing bio-based hydrogels have been frequently investigated for tissue engineering, as exemplified above; comparatively much less publications deal with controlled drug delivery from such composites. Xu et al reported on the fabrication of cellulose nanocrystal/chitosan hydrogels for the controlled delivery of the hydrophilic respiratory drug theophylline [113]. Cellulose nanocrystals were first oxidized using sodium periodate to obtain dialdehyde nanocellulose. Then, chitosan was crosslinked via Schiff base formation using dialdehyde nanocellulose in different weight ratios to fabricate cellulose nanocrystal/chitosan hydrogel composites. When the chitosan percentage in the hydrogel was increased, the isoelectric point shifted towards a higher pH, likely as a result of a higher amount of free amino groups. The lowest degree of swelling for a given formulation occurred at the isoelectric point, where the electrostatic repulsion between the ionic charges in the network is minimal. The crosslinked density decreased with decreasing nanocellulose concentration, resulting in an increased degree of swelling and hence a hydrogel morphology with larger pores. The theophylline release profiles showed a burst release during the first few hours, after which the drug release levelled off. The release rate decreased as the dialdehyde nanocellulose content in the composites increased, resulting from a higher amount of crosslinks in the networks. Due to a higher swelling ratio of the hydrogels in an acid environment, the cumulative drug release at pH 1.5 (85%) was significantly higher than the cumulative drug release at pH 7.4 (25%), showing the potential of these composite hydrogels as gastric drug delivery systems.

Following a similar approach, Zheng et al prepared giant network composite hydrogels by reacting dialdehyde nanocellulose with gelatin via Schiff base formation [114]. Upon incorporation of the nanocellulose, the compression strength of the hydrogel increased to 1.6 MPa, 40 times that of the pure gelatin hydrogel.

Only a limited number of publications report on bio-based hydrogel composites containing cellulose nanofibers obtained from bacteria. For example, Park et al incorporated bacterial nanocellulose in alginate hydrogels and showed that these composites are potential candidates for cell encapsulation thanks to the 3D fibrous structure of the cellulose, which resembles the extracellular matrix [115]. Yan et al demonstrated that the addition of bacterial nanocellulose to alginate/collagen hydrogels resulted in improved compressive strength as well as reduced hydrogel swelling and degradation [116]. This was ascribed to the fibrillar meshwork structure of bacterial nanocellulose and the formation of intermolecular hydrogen bonding among the components of the composite hydrogel. Interestingly, cellulose-producing bacteria offer the possibility for in situ biosynthesis of nanocellulose/hydrogel composites by supplementing the bacterial culture medium with hydrogel precursors [117, 118].

5.1. Unique advantages and potential limitations of composite hydrogels over monocomponent hydrogels

Important progress has been made in the field of bio-based hydrogels for biomedical applications during the last decades. This review clearly shows the interest of incorporating an additional component in such hydrogels, as the resulting composite materials present markedly improved, synergistic properties. In particular, the incorporation of micelles or microspheres in hydrogels facilitates the simultaneous, controlled release of multiple drugs with different solubilities, including hydrophobic drugs. Furthermore, the inclusion of electrospun fibers in hydrogels enhances their mechanical properties, enabling their use as robust tissue engineering scaffolds. Nanocellulose, an emerging class of materials encompassing both nanowhiskers and nanofibrils, has shown particularly effective in reinforcing hydrogels at remarkably low concentrations due to the presence of crystalline cellulose regions with a very high mechanical strength. In addition to facilitating desired drug release profiles and structural properties, particles and fibers have also been used to improve the biological performance of composite hydrogels by optimizing the micro-environmental conditions for cells.

However, despite the vast progress that has been achieved in the field, the state-of-the-art is insufficient to translate bio-based composite hydrogels into clinically viable biomedical solutions. First of all, most of the bio-based composite hydrogels have only been evaluated in vitro. To enable clinical translation of these promising systems, the therapeutic effects should be studied in vivo together with their long-term biodegradability, toxicity and accumulation. Attention should also be given to the foreign body response (FBR), which is an immune-mediated reaction to implanted materials where a series of wound-healing processes and inflammatory events result in fibrosis [119]. This can affect the performance and durability of biomaterials that require use during prolonged periods of time, as the thick isolating fibrotic layer may cause pain and ultimately lead to device failure. Although natural polymers are regarded as biocompatible, they have been shown to generate fibrotic responses. Future research concerning bio-based composite hydrogels should therefore also adopt modern strategies to minimize the FBR. Examples of such strategies are optimization of physical material properties (e.g. the surface topography), the incorporation of certain active agents in the hydrogels (e.g. anti-inflammatory drugs or anti-fibrotic agents) and the functionalization of the material with bioactive moieties such as the RGD peptide [119, 120].

Secondly, many of the composite hydrogels discussed in this review concern systems that have to be formed into their final shape before implantation. However, in situ hydrogel formation is an important requirement in view of the clear advantages such as facile administration, improved patient comfort and the straightforward incorporation of therapeutic agents as well as particles or fibers. Although in situ composite hydrogel formation has been reported in a number of cases, the crosslinking chemistry often relies on Schiff base formation, which is undesirable in an in vivo setting as it may interfere with biological processes. Future research should focus on a significant and fast response to environmental gelation triggers such as temperature (for physical crosslinking) or the use of bioorthogonal reactive groups, which do not interfere with biological processes upon covalent crosslinking. Thirdly, the influence of the size of the incorporated structures (particles or fibers) and the influence of the morphology of the hydrogel on the composite properties have not been investigated systematically. This should be addressed in future research as these characteristics will likely impact important properties of the composite hydrogels, such as the mechanical, degradation and biological behavior. Lastly, although the field of bio-based composite hydrogels is rapidly evolving, as demonstrated in this review, the number of reports on systems displaying spatial and temporal control over drug release and/or cell behavior is still too limited. In this regard, 3D printing will be a valuable tool since it enables the fabrication of gradient composites with a defined architecture and regional differences, which potentially allows for a high level of control over drug release and cellular responses.

Promising in vivo results combined with fast in situ formation via physical gelation, as reported by some groups [71, 73], show that the research concerning bio-based composite hydrogels is moving in the right direction, but increased research efforts are needed before clinical translation can be fulfilled. Nevertheless, the author believes that thanks to their inherent biocompatibility, their versatile nature as well as the increasing demand for products from renewable resources, bio-based composite hydrogels will likely increasingly replace single component, synthetic hydrogels in biomedical applications in the years to come.

The author gratefully acknowledges Dr Tatiana Budtova for critical reading and her suggestions for improvement of the manuscript.