Growth‐Factor Free Multicomponent Nanocomposite Hydrogels That Stimulate Bone Formation

Synthetic osteo‐promoting materials that are able to stimulate and accelerate bone formation without the addition of exogenous cells or growth factors represent a major opportunity for an aging world population. A co‐assembling system that integrates hyaluronic acid tyramine (HA‐Tyr), bioactive peptide amphiphiles (GHK‐Cu2+), and Laponite (Lap) to engineer hydrogels with physical, mechanical, and biomolecular signals that can be tuned to enhance bone regeneration is reported. The central design element of the multicomponent hydrogels is the integration of self‐assembly and enzyme‐mediated oxidative coupling to optimize structure and mechanical properties in combination with the incorporation of an osteo‐ and angio‐promoting segments to facilitate signaling. Spectroscopic techniques are used to confirm the interplay of orthogonal covalent and supramolecular interactions in multicomponent hydrogel formation. Furthermore, physico‐mechanical characterizations reveal that the multicomponent hydrogels exhibit improved compressive strength, stress relaxation profile, low swelling ratio, and retarded enzymatic degradation compared to the single component hydrogels. Applicability is validated in vitro using human mesenchymal stem cells and human umbilical vein endothelial cells, and in vivo using a rabbit maxillary sinus floor reconstruction model. Animals treated with the HA‐Tyr‐HA‐Tyr‐GHK‐Cu2+ hydrogels exhibit significantly enhanced bone formation relative to controls including the commercially available Bio‐Oss.


Introduction
There is an increasing need to develop strategies to facilitate bone repair and regeneration in and around missing or defective craniomaxillofacial regions, such as eyes, ears, noses, maxilla, mandible, and teeth. One area that requires particular attention, given its anatomical complexity and potential complications, is the posterior maxillary. Here, loss of teeth can result in adverse conditions including severe maxillary sinus pneumatization, which can also increase the risk of root tips, teeth displacement into the sinus cavity, or microbial contamination of the maxillary sinus lift. [1] Therefore, implantation of osseointegrated biomaterials has gained widespread attention in dentistry to replace missing or lost teeth with successful outcomes of complete or partial edentulism. [2] However, due to insufficient alveolar bone and irregular structure of the maxillary sinus in atrophied maxilla, accidental displacement of a dental implant into the maxillary sinus remains a common complication encountered in dental clinical practice. [3] Consequently, maxillary sinus floor promoters of tissue growth, their use is associated with a variety of critical complications, such as the need for large amounts of growth factors due to the rapid inactivation and clearance of the growth factors, [26] high-costs, and uncontrolled activity. [27] Thus, an ideal therapeutic hydrogel for maxillary sinus lift reconstruction/augmentation should also be biodegradable generating nontoxic products, be simple to implant, exhibit a short set time, be mechanically stable, and rapidly fill irregular anatomical volumes. [28] Furthermore, given the complex anatomy and physiology of bone, the hydrogel should hold multifunctional properties to recreate key features of the extracellular matrix (ECM) and stimulate cell types of interest. In this context, multicomponent self-assembly offers an attractive avenue to design hydrogels with multiple building-blocks, functionalities, and the molecular precision of self-assembly. [29,30] In this study, we have developed a three-component selfassembling system that integrates hyaluronic acid (HA), peptide amphiphiles (PAs), and Laponite (Lap). HA is a large ECM polysaccharide ubiquitous in tissues and organs that has been extensively used as a biomaterial due to its biocompatibility and biodegradability. [31] However, HA exhibits poor structural integrity and stability and consequently is usually chemically modified with, for example, tyramine (Tyr) [32] or hybridized with other biomaterials, such as hydroxyapatite [33] to increase its functionality. PAs are a class of self-assembling peptidebased building blocks with the intrinsic capacity to assemble into well-defined nanofibrous hydrogels. [34] PAs consist of: i) a hydrophobic tail that drives self-assembly, ii) a β-sheet forming amino acid sequence that stabilizes the assembled nanofibers through hydrogen bonds, and iii) a charged functional head group that facilitates solubility in aqueous environments. This platform can incorporate a spectrum of bioactive epitopes, which have been used to target regeneration of tissues, such as bone, [35] enamel, [36] cartilage, [37] and vascular. [38] However, self-assembling materials typically provide limited structural integrity, which has hindered their wide spread applicability. Lap is a 2D nanosilicate with anisotropic charge distribution, which has been exploited as an effective cross-linker and rheology modifier for hydrogels [39] for the delivery of drugs, growth factors, and antibodies. [40,41] Lap has been reported to promote cell adhesion and proliferation and can exert osteogenic effects on cells in vitro. [42] Consequently, Lap has been combined with macromolecules, such as DNA [43] or proteins [44] to fabricate hydrogels capable of promoting osteogenic differentiation in vitro [45,46] or bone regeneration in vivo in mice. [44] Here, we report the synthesis and characterization of a multicomponent self-assembling system that integrates the osteogenic properties of Lap, the signaling and nano fibrous structure of PAs, the proangiogenic properties of the GHK-Cu 2+ peptide, and the biocompatibility and instant gelation properties of Tyr-modified HA to fabricate an osteoinductive and osteoconductive hydrogel for bone regeneration. The system also takes advantage of both covalent (oxidative coupling) and noncovalent (electrostatic) interactions to generate a material that is both injectable and robust. The applicability of the materials was assessed in vitro using human mesenchymal stem cells (hMSCs) and human umbilical vein endothelial cells (hUVECs), and in vivo using a rabbit maxillary sinus floor reconstruction model.

Rationale of Design
Our approach enables the rationale design of complex and multifunctional hydrogels for bone regeneration (Figure 1a-d). We used HA to provide a rich and biocompatible ECM macromolecule, which was functionalized with Tyr (HA-Tyr) to control stability through enzyme-mediated oxidative coupling as previously described by Eglin and co-workers. [47] To further enhance the hydrogel's structural integrity and bioactivity, we designed PA molecules to coassemble with HA-Tyr through electrostatic interactions into nanofibers that exhibit the proangiogenic osteonectin fragment glycine-histidine-lysine (GHK-Cu 2+ ). [48] In addition, Laponite discs (Lap) were incorporated to provide a distinctive dual-charged structure (i.e., positive rim and negative face charge of the nanaosilicate disc) that would facilitate interaction with both the anionic charged HA-Tyr and the cationic PA (GHK-Cu 2+ ). It is noteworthy that the osteogenic effects of Lap can be an additional benefit. Finally, to facilitate temporal control of assembly and implantation, the hydrogels were designed to assemble instantly through; a) oxidative coupling of the phenolic moiety of HA-Tyr mediated by horseradish peroxidase and H 2 O 2 , which has been demonstrated to be biocompatible both in vitro and in vivo, [49] as well as b) electrostatic interactions between the cationic GHK-Cu 2+ , anionic HA-Tyr, and anionic/cationic Lap.

Synthesis and Characterization of the GHK-Cu 2+ PA
The PA molecule (GHK-Cu 2+ ) used in this study is a threedomain molecule with a hydrophobic tail (black), β-sheet forming amino acid residues (blue), and the therapeutic cationic tripeptide matrikine (pink) derived from osteonectin (C 16 H 31 CO-VVVAAAGHK) (Figure 1b). In order to present GHK as a complex of copper (the form in which it exists in the human body), we mixed an aqueous solution of copper (II) sulfate (CuSO 4 . 5H 2 O) (4 mg mL −1 ) with a GHK solution (2 wt%). The copper chelating ability (GHK-Cu 2+ ) was confirmed by electron paramagnetic resonance (EPR) spectroscopy ( Figure S1, Supporting Information). The circular dichroism (CD) spectrum of GHK-Cu 2+ demonstrated a β-sheet-like bisignate with a maximum and minimum at 198 and 220 nm, respectively ( Figure S2, Supporting Information), while transmission electron microscopy (TEM) confirmed self-assembly into the classical PA nanofibrous morphology measuring ≈10 nm in diameter and several microns in length ( Figure S3, Supporting Information).

Synthesis and Characterization of the Hydrogels
The multicomponent system (HA-Tyr-Lap-GHK-Cu 2+ ) was prepared in stages. First, a HA-Tyr solution (6 wt%) in horseradish peroxidase (HRP)-containing phosphate buffer saline (PBS) was Adv. Funct. Mater. 2020, 30,1906205  combined with exfoliated Lap (5 wt%). Separately, an aqueous solution of GHK-Cu 2+ (2 wt%) was combined with H 2 O 2 (0.75 × 10 −6 m). Upon mixing both solutions, instant gelation occurred by HRP mediated oxidative coupling of the phenolic moiety on HA-Tyr using H 2 O 2 as oxidant and by electrostatic interactions between the GHK-Cu 2+ with both Lap and HA-Tyr. The hydrogels HA-Tyr, HA-Tyr-GHK-Cu 2+ , and HA-Tyr-Lap were prepared as described in the Experimental Section and used as controls. In all cases, self-supported and robust hydrogels were formed ( Figure S4, Supporting Information). Due to the multicomponent nature of the hydrogels, these were prepared with fixed concentrations of both HA-Tyr [50] and Lap [45] that have been previously reported by others to exhibit suitable mechanical properties and bioactivity. We used fluorescence and attenuated total reflectance-Fourier transform infrared (ATR-FTIR) spectroscopies to investigate molecular interactions underpinning co-assembly of the various components in our hydrogels. The fluorescence emission spectra of HA-Tyr solution and hydrogels prepared by oxidative coupling were examined. Spectra were collected at an excitation wavelength (λ ex ) of 260 nm. The fluorescence spectra from HA-Tyr solution (0.5 wt%) depicted a weak broad emission maxima (λ em ) at 330 nm, which corresponds to the emission wavelength of a phenolic group ( Figure S5, Supporting Information). Upon oxidative coupling the spectra shows an intense emission maxima at 422 nm with a shoulder at 450 nm ( Figure S5, Supporting Information). The redshifted fluorescence emission is indicative of peroxidase-catalyzed oxidative coupling leading to dityramine/dityrosine bridge formation. [51] The FTIR spectrum of GHK exhibited a band at 3250 cm −1 due to N-H vibrational stretching of amide II, 2952 and 2850 cm −1 due to C-H stretching of alkyl groups, 1632 cm −1 due to CO stretching of amide I, 1540 cm −1 due to N-H stretching of amide II (aromatic) and 1230 cm −1 due to N-H stretching of amide III ( Figure S6, Supporting Information). For HA-Tyr xerogel, the characteristic absorption bands were observed at 3200, 1638, 1540, and 1020 cm −1 corresponding to the O-H vibrational stretching, CO stretching of amide I, N-H stretching of amide II, and C-O asymmetric vibrational stretching, respectively ( Figure S6, Supporting Information). The distinctive absorption band at 993cm −1 in the spectrum of Lap was attributed to the Si-O vibrational stretching. [45] In the spectrum of HA-Tyr-GHK-Cu 2+ xerogel, we observed a slight shift in the vibrational stretching frequency of the CO region to 1635 cm −1 , suggesting hydrogen bond with associated electrostatic interactions between HA-Tyr and GHK-Cu 2+ . Similarly, the spectrum of HA-Tyr-Lap xerogel shows that the CO and Si-O bands shifted to 1635 and 1000 cm −1 , respectively, which is indicative of hydrogen bond interactions between HA-Tyr and Lap. Interestingly, all the changes observed in the chemical environments of the functional groups (see HA-Tyr-GHK and HA-Tyr-Lap) were also revealed in the spectrum of the multicomponent HA-Tyr-Lap-GHK-Cu 2+ xerogels. Given the evidence of the peroxidase-mediated oxidative coupling of HA-Tyr and the electrostatic and hydrogen bond interactions provided by Lap and the cationic GHK-Cu 2+ , we reasoned that the synthesis of our multicomponent HA-Tyr-Lap-GHK-Cu 2+ hydrogels is based on orthogonal interactions between all the components.

Structural Properties of the Hydrogels
Scanning electron microscopy (SEM) observations revealed that the GHK-Cu 2+ hydrogel exhibited the classical PA nanofibrous architecture [34] (Figure 2a), while the HA-Tyr hydrogel exhibited a microporous morphology (Figure 2b) as previously described. [52] As expected, HA-Tyr-GHK-Cu 2+ hydrogels exhibited features of both of these hydrogels including nanofibers and microscopic pores (Figure 2c). The multicomponent HA-Tyr-Lap-GHK-Cu 2+ hydrogels revealed an architecture that combined and retained all these features ( Figure 2e) with Lap uniformly distributed on the surface of the fibers (Figure 2f), as evidenced by energy dispersive X-ray (EDX) analysis confirming the presence of Si, Mg, Na, and O ( Figure 2g). Although SEM examinations are not the ideal approach to describe in detail the structure of each hydrogel given the likelihood for structure disruption as a result of the sample preparation, the results provide a useful relative comparison between the different hydrogels.

Mechanical and Physical Properties of the Hydrogels
A major goal of our design was to provide a multifunctional bioactive hydrogel that could stimulate specific biological processes while being a robust and easy to use system. To test the structural and mechanical properties of the hydrogels, we performed a number of tests including compressive tests, stress relaxation measurements, swelling tests, and enzymatic degradation experiments.
Compression tests were performed on 5 mm diameter and 5 mm high cylindrical hydrogels. HA-Tyr hydrogels exhibited a Young's modulus of 25.03 ± 4.00 ( Figure 2h; and Figure S7, Supporting Information), in agreement with those of previous studies. [50] Interestingly, upon co-assembly with GHK-Cu 2+ the Young's modulus of the hydrogel increased slightly up to 28.18 ± 4.32 kPa (HA-Tyr-GHK-Cu 2+ ), which may result from an enhanced entanglement with the PA nanofibers and consequent slight decrease in porosity ( Figure 2c). Importantly, upon co-assembly with Lap, the hydrogels exhibited a significantly increase in Young's moduli up to 58.23 ± 7.8 kPa (HA-Tyr-Lap) and 63.11 ± 8.0 kPa (HA-Tyr-Lap-GHK-Cu 2+ ) (Figure 2h, Figure S7, Supporting Information), indicating that the dityramine moieties (HATyrTyrHA bonds) that result from enzymatic crosslinking of the tyramine form strong physical interfacial bonds with Lap. These results are in agreement with previous studies that have reported a Lap-induced increase in the stiffness of hydrogels [53] and evidence the potential of our hydrogels to be used as robust and bioactive hydrogel implants for bone regeneration.
It is well-established that hydrogel stiffness plays a crucial role in directing cell phenotype and that stiffer hydrogels (>30 kPa) are able to promote osteoblastic phenotypes. [54] While the Young's modulus of our multicomponent HA-Tyr-Lap-GHK-Cu 2+ hydrogel is higher (63.11 kPa), differences in stiffness measuring techniques and other hydrogel properties such as porosity and stress relaxation are important to consider. In addition to the effect of hydrogel stiffness on cell behavior, stress relaxation has also been shown to play a role in cell signaling by modulating ligand binding and cytoskeletal organization. [55] Stress relaxation tests revealed that all multicomponent hydrogels and HA-Tyr exhibited a similar relaxation profile of ≈55% at about 5 min (Figure 2i). This behavior is likely due to reversible transient molecular interactions and reorganization of the HA chains and/or PA nanofibers as well as release of entanglements driven by in situ formation of dityrosine bridges between adjacent tyramine moieties in the HA-Tyr chains. However, in Adv. Funct. Mater. 2020, 30,1906205  the initial the 10 s, HA-Tyr-Lap-GHK-Cu 2+ hydrogels exhibit a faster relaxation profile compared to all other hydrogels ( Figure 2i). This rapid hydrogel stress relaxation has been shown to promote proliferation and differentiation toward an osteoblastic phenotype of mesenchymal stem cells (MSCs). [55] These results demonstrate that the multicomponent covalent and noncovalent approach also generates a hydrogel relaxation profile with potential benefit for bone regeneration applications.
The propensity of hydrogels to absorb and retain water provides a measure of structural integrity and adaptability as well as capacity to enable critical nutrient and waste diffusion. Consequently, we conducted swelling experiments on freeze-dried hydrogels by immersing them in PBS and systematically calculating the percentage of water uptake at 0, 10, 20, 30, and 40 min. HA-Tyr hydrogels exhibited an exceptionally high swelling ratio of 1680 ± 94% after 10 min of immersion in PBS, which increased to 1840 ± 87% after 40 min (Figure 2j). Similarly, HA-Tyr-GHK-Cu 2+ hydrogels displayed a high swelling ratio of 1540 ± 57% which increased to 1802 ± 51%, after 40 min. In contrast, Lap-containing hydrogels exhibited lower swelling ratios of 1470 ± 45% (HA-Tyr-Lap) and 1350 ± 17% (HA-Tyr-Lap-GHK-Cu 2+ ) after 10 min of incubation, which increased to 1540 ± 35% (HA-Tyr-Lap) and 1580 ± 24% (HA-Tyr-Lap-GHK-Cu 2+ ) after 40 min (Figure 2j). This decrease suggests that despite the ionic and hydrophilic nature of Lap, its presence impedes water uptake. Since the degree of swelling ratio is inversely proportional to the crosslinking density of hydrogels, [56] we reasoned that both the physical crosslinking provided by the Lap discs and covalent crosslinking between the tyramine moieties (HATyrTyrHA bonds) significantly limit water uptake in HA-Tyr-Lap and HA-Tyr-Lap-GHK-Cu 2+ hydrogels. Nonetheless, all hydrogels exhibited a relatively fast and high level of equilibrium swelling ratio, returning to their original size upon rehydration (Figure 2j). This capacity may facilitate their use as geometrically customized hydrogels that can be dried, stored, and rehydrated prior to implantation.
Controlled degradation is another important parameter for bioactive implantable hydrogels aiming to balance stimulation of cell growth and adequate replacement of new tissue. While HA has been extensively pursued as a biomaterial, its use to fabricate robust scaffolds for tissue regeneration has been limited in large part by its susceptibility to rapid degradation in the presence of hyaluronidases both in vitro and in vivo. [57] Several approaches have been devised to enhance HA hydrogel stability to enzymatic degradation with varying degree of success. [58,59] Our design enables the possibility to combine covalent and noncovalent interactions as well as exploit synergistic interactions between the different components. To test this, hydrogels were prepared, immersed in aqueous solutions of hyaluronidase (20 U mL −1 ) at 37 °C, and their weight measured at various times points up to 50 d (Figure 2k). At 18 d, HA-Tyr hydrogels were found to be completely digested, while the multicomponent (HA-Tyr-GHK-Cu 2+ , HA-Tyr-Lap, and HA-Tyr-Lap-GHK-Cu 2+ ) hydrogels resisted full degradation up to 50 d (Figure 2k). In particular, Lap-containing hydrogels exhibited a significant decrease in degradation rate, which correlates with previous studies also reporting an enhanced Lapinduced stability in polymers [60] and biopolymers. [45] We speculate that the suppressed hydrogel susceptibility to rapid enzymatic degradation observed in HA-Tyr-Lap and HA-Tyr-Lap-GHK-Cu 2+ may be attributed to the strong affinity of Lap discs to proteins and enzymes through physical adsorption, [41] which would consequently limit hyaluronidase access to the HA backbone.

Viability and Proliferation of hMSCs
To test the applicability of the multicomponent hydrogels, we first conducted in vitro tests by culturing hMSCs on the materials. Using a live/dead assay, cytocompatibility was first assessed. As expected, HA-Tyr alone did not support cell attachment and spreading (Figure 3), likely as a result of lack of cell-binding motifs. [52] However, cells cultured on all multicomponent HA-Tyr-GHK-Cu 2+ , HA-Tyr-Lap, and HA-Tyr-Lap-GHK-Cu 2+ hydrogels adhered and appeared to exhibit a spread Adv. Funct. Mater. 2020, 30,1906205 (Figure 3e). We then examined the effect of the hydrogels on hMSC proliferation and found that cell number increased from day 1 to 7 on all multicomponent hydrogels, with greater increase on HA-Tyr-Lap-GHK-Cu 2+ compared to HA-Tyr-Lap and HA-Tyr-GHK-Cu 2+ . In addition, cell proliferation was higher on HA-Tyr-Lap-GHK-Cu 2+ prepared with the higher concentration of GHK-Cu 2+ (1 wt%) compared to the lower one (0.1 wt%) (Figure 4a), which is in agreement with previous work reporting the stimulating role of GHK on cell proliferation. [61]

Osteoblastic Differentiation of hMSCs
Next, we assessed the osteoinductive potential of the multicomponent hydrogels by growing hMSCs using culture media with and without osteoinductive agents and quantifying alkaline phosphatase (ALP) expression. In this case, we tested three different hydrogels comprising different concentrations of GHK-Cu 2+ . In osteoinductive media, up-regulation of ALP activity was observed on all hydrogels with highest expression at day 4 on the TCP control compared to day 9 for the multicomponent hydrogels, suggesting a delayed expression on the hydrogels ( Figure S9, Supporting Information). However, in the absence of osteoinductive media, a more indicative scenario of the inductive properties of the materials, ALP expression peaked for all tested substrates on day 9. In this case, the highest expression was observed on the HA-Tyr-Lap-GHK-Cu 2+ hydrogels containing the highest concentration of GHK-Cu 2+ compared to all other hydrogels and TCP control (Figure 4b). This result suggests that the multicomponent HA-Tyr-Lap-GHK-Cu 2+ hydrogel has the capacity to stimulate osteoblastic differentiation on hMSCs in the absence of osteoinductive  factors. Furthermore, the dose-dependent effect of GHK-Cu 2+ on ALP activity is in agreement with previous reports on the osteogenic effect of copper-free alginate-GHK hydrogels. [61]

Angiogenic Effect on hUVECs
Blood vessel formation is essential in bone regeneration. Therefore, we qualitatively assessed the in vitro proangiogenic effect of the multicomponent hydrogels on hUVECs. By comparing the morphology of calcein-stained hUVECs growing on the different materials, GHK-Cu 2+ -containing multicomponent hydrogels HA-Tyr-GHK-Cu 2+ and HA-Tyr-Lap-GHK-Cu 2+ were observed to trigger cell elongation and angiogenic sprouting with microcapillary-like structures by day 1 of culture ( Figure 4c_(iii-iv)). In contrast, cells seeded on HA-Tyr and HA-Tyr-Lap (without GHK-Cu 2+ ) maintained their normal endothelial phenotype (Figure 4c_(i-ii). In addition, a more complex vascular lumen [62] structure was also formed on the GHK-Cu 2+ -containing hydrogels after day 5 (Figure 4c_(vii-viii)). Such lumen structures were similar to those observed on the positive control (Matrigel + 50 ng mL −1 VEGF) (Figure 4c_(ix)) at day 5. These results suggest that the GHK-Cu 2+ -containing multicomponent hydrogels have proangiogenic properties.
The proangiogenic effects and ability of GHK to increase VEGF secretion have been previously associated with its binding to α 6 or β 1 integrin or both. [63] Also, a recent metabolomics pathway analysis of cells in alginate-GHK hydrogels revealed that the integrin linked kinase mediates the numerous biological functions of GHK tripeptide. [61] The formation of vascular system is mainly ensured by the emergence of new microcapillary from existing vessels (sprouting). [64]

In Vivo Assessment of the Hydrogels
To test the bioactivity and in vivo bone regenerative capacity of the hydrogels, experiments were conducted on a standard sinus rabbit model following established protocols. [65] The four hydrogel materials were tested with untreated animals serving as negative controls and animals treated with the commercial product Bio-Oss (Geistlich Ltd., Switzerland) as positive controls. Bio-Oss is a bone substitute material derived from deproteinized bovine bone marrow and has been widely used in regenerative dentistry with good success as a filler for maxillary sinus augmentation. [6] Animals were implanted with 50 µL of the hydrogels or positive control, sacrificed at either 6 or 12 weeks after implantation, and assessed qualitatively and quantitatively for bone formation and cytotoxicity.
All animals survived surgery and exhibited normal behavior during the implantation time and no signs of inflammation or infection were physically observed at the time of sacrifice. Cone beam computer tomography (CBCT) was used to qualitatively assess differences in the performance of the different materials. At 6 and 12 weeks, CBCT scans revealed new bone formation in the sinus of animals treated with the hydrogels and positive control (Bio-Oss) (Figure 5b). However, higher contrast likely associated with new bone formation was observed in animals treated with the multicomponent hydrogels, especially those treated with the HA-Tyr-Lap-GHK-Cu 2+ hydrogels. Interestingly, this enhancement also appeared to be larger than animals treated with the positive control at both 6 and 12 weeks' time-points. It is noteworthy that the radiation resistant area in the sinus of the positive control is partly due to the residual materials of the Bio-oss, which can be difficult to distinguish from the newly mineralized tissue. Nonetheless, the results suggest that the HA-Tyr-Lap-GHK-Cu 2+ hydrogels are capable of promoting new bone formation within the sinus.
To confirm this result, histological sections stained with hematoxylin and eosin (HE) were acquired from animals at 6 and 12 weeks after implantation. Animals implanted with the HA-Tyr-Lap-GHK-Cu 2+ hydrogels qualitatively exhibited greater amounts of newly formed bone compared to animals treated with all other hydrogels and controls at both weeks 6 and 12, evidenced by the presence of relevant cells and ossified tissue (Figure 5c). These histological sections were then used to quantify new bone formation within the sinus region by identifying and quantifying the areas within the sinus region exhibiting osteocytes and ossified tissue (Figure 5a). At both 6 and 12 weeks postsurgery, the results confirmed that animals receiving the multicomponent HA-Tyr-Lap-GHK-Cu 2+ hydrogels revealed the highest percentage of mean area of ossified tissue (40.37 ± 1.54 and 60.12 ± 2.80%, respectively) within the sinus of all tested groups including significantly higher than animals treated with the positive control (35.97 ± 1.54% and 37.56 ± 1.18%, respectively) (Figure 6a). The histological sections of animals treated with the HA-Tyr-Lap-GHK-Cu 2+ hydrogels also exhibited both osteoblasts and osteocytes within lacuna, further evidencing the presence of an active regenerative environment (Figure 6b). It is likely that these cells emerged from MSCs or preosteoblasts migrating from the local bone surface, periosteum, or the blood. By comparing these results with those of the other hydrogels (Figure 6a), we conclude that the enhanced bioactivity of HA-Tyr-Lap-GHK-Cu 2+ hydrogels may result primarily from the presence of Lap and GHK-Cu 2+ . These results correlate with the in vitro results, which show that Lap and especially GHK may have osteoinductive properties in the absence of growth factors. Vascularization is of upmost importance in bone regeneration. Closer examination of the histological sections revealed that animals treated with HA-Tyr-Lap-GHK-Cu 2+ exhibited a qualitatively higher amount of blood vessels compared to animals treated with HA-Tyr-GHK-Cu 2+ and the positive control (Figure 6c). Interestingly, blood vessels were much less prevalent in hydrogels that did not contain GHK-Cu 2+ , which is in accordance with the in vitro experiments as well as previous studies that have reported on the ability of GHK-Cu 2+ to promote formation of blood vessels in mice. [66] The in vivo experiments were also used to investigate the biosafety of the hydrogels through a systemic toxicity test. Histological sections of the liver, heart, spleen, lung, and kidney from the animals treated with the HA-Tyr, HA-Tyr-GHK-Cu 2+ , HA-Tyr-Lap, and HA-Tyr-Lap-GHK-Cu 2+ hydrogels were examined and presented no signs of inflammation or histological changes compared to control animals ( Figure S11, Supporting Information). These results suggest that the HA-Tyr-Lap-GHK-Cu 2+ hydrogels did not degrade into toxic by-products. We speculate that, beyond the molecular signaling and biocompatibility of the HA-Tyr-Lap-GHK-Cu 2+ hydrogel, its high osteo-promoting activity may also result from the inherent mechanical and physical properties of the hydrogel. Both the hydrogel's Young's modulus (Figure 2h) and stress relaxation profile (Figure 2i) exhibit values that have been reported to be beneficial for bone promoting applications. [54,55] Furthermore, the hydrogels display an enzymatic degradation profile (Figure 2k) that may have permitted sufficient signaling to migrating and surrounding cells while progressively degrading to enable new tissue formation. This capacity for timely scaffold degradation is key for optimum tissue regeneration. [67] Furthermore, beyond these beneficial molecular, chemical, and physical properties, the potential to easily manipulate and deliver the hydrogel represent key advantages to facilitate its clinical use and impact.

Conclusion
We have developed a practical and multifunctional self-assembling hydrogel biomaterial for bone regeneration applications. The material takes advantage of both covalent and noncovalent interactions to integrate HA, PAs, and Lap into a bioactive hydrogel with a spectrum of molecular, physical, and mechanical properties designed to promote bone regeneration as well as minimally invasive implantation. We demonstrate the capacity of the hydrogels to support cell growth and stimulate both osteoblastic differentiation and angiogenic sprouting of hUVECs in vitro as well as promote faster bone regeneration in a rabbit model compared to a commercially available goldstandard material. The current study introduces a new molecularly designed self-assembling material that stimulates bone formation without the use of exogenous growth factors and demonstrated its potential use in maxillary sinus reconstruction and other bone tissue regeneration procedures.

Experimental Section
Materials: Laponite XG was a generous gift from the laboratory of Professor Richard OC Oreffo and Dr. Jonathan Dawson, University of Southampton while HA-Tyr was synthesized as previously described elsewhere. [45] Dulbecco's Modified Eagle's Medium (DMEM), fetal bovine serum (FBS), PBS, penicillin, and streptomycin, horseradish peroxidase, hydrogen peroxide were purchased from sigma-Aldrich Inc. (UK). The Live/Dead assay kits (calcein AM and ethidium homodimer) were purchased from Thermo Fisher Scientific Ltd (UK).
Peptide Synthesis and Characterization: PAs were synthesized as previously described [68] by solid phase peptide synthesis on Liberty Blue-automated microwave peptide synthesizer (CEM Ltd, UK). The standard 9-fluorenylmethoxycarbonyl (Fmoc) protection chemistry on a 4-methylbenzhydrylamine (MBHA) Rink Amide resin (Novabiochem Corporation, UK) was employed. Amino acid couplings were performed using 4 mmol equivalent of Fmoc-protected amino acids (Novabiochem Corporation, UK), 4 mmol equivalents of 1-hydroxybenzotriazol (HOBT, Carbosynth Ltd, UK) and 6 mmol equivalents of N,N'-diisopropylcarbodiimide (DIC, Sigma-Aldrich Inc., UK) for 1 h. Fmoc deprotections were performed with 20% piperidine (Sigma-Aldrich Inc., UK) in N,N-dimethylformamide (DMF, Alfa Aesar Inc., UK). Following Fmoc removal from the final amino acid residue, the alkyl tail moiety (from palmitic acid, C 16 H 32 O 2 , Calbiochem Inc., UK) was conjugated to the free N-terminus. The alkylation reaction was accomplished by using palmitic acid (4 mmol), HOBT (4 mmol), and DIC (6 mmol) in DMF/dichloromethane. The reaction was allowed to proceed at room temperature for 4 h or until obtaining a negative Kaiser test. PA cleavage from the resin and deprotection of the side chains were carried out with a mixture of trifluoracetic acid (TFA, Sigma-Aldrich Inc., UK)/triisopropylsilane (TIS, Alfa Aesar Inc., UK)/water (95:2.5:2.5) for 3 h at room temperature. After filtration of the cleavage mixture, TFA was removed by rota-evaporation and the resulting solution was triturated with cold diethylether at −20 °C. The precipitate was collected by centrifugation, washed twice with cold diethylether, air-dried, dissolve in deionised water and lyophilized. The product was then purified using a preparative HPLC (Waters Ltd, USA) with reverse-phase Xbridge C18 column (Waters Ltd, USA) and water/acetonitrile (0.1% TFA) binary mobile phase. Hydrochloric acid (10 × 10 −3 m) was added to the HPLC fractions and rotavap to remove the residual TFA, leaving behind chloride as the PA counter ions. Finally, the PA was dialyzed against deionized water using 500 MWCO dialysis tubing (Spectrum Europe BV, The Netherlands) to remove salts, lyophilized to obtain a white fluffy pure PA. Circular dichroism (CD) was measured with Chirascan circular dichroism spectrometer (Applied Photophysic Ltd, UK) using quartz cell with 1 mm path length and the following parameters: data pitch-0.5 nm, scanning mode-continuous, scanning speed-100 nm min −1 , bandwith-2 nm and accumulation-5. All CD data were presented as ellipticity and recorded in millidegree (mdeg). CD spectra were obtained by signal integrating 3 scans, from 190 to 260 nm at speed of 50 nm min −1 . Data were processed by a simple moving average and smoothing method. Electron paramagnetic resonance spectroscopy was recorded on a Bruker EMX EPR equipped with a standard cavity, operating at X-band frequency using standard Wilmad quartz tubes at −80 °C. Transmission electron microscopy was performed on etched carbon-coated copper grids (Agar Scientific Ltd, Stanstead, UK) using JEOL 1230 TEM fitted with Morada CCD camera. Samples were stained with aqueous 2% uranyl acetate solution.
Hydrogel Preparation and Characterizations: The requisite concentration (6 wt%) of HA-Tyr was prepared in PBS containing 3 unit mL −1 of HRP and allowed to fully dissolve overnight at 4 °C. This concentration of Molecular Characterization of Co-assembly: IR spectra were recorded on a PerkinElmer ATR-FTIR spectrometer in the range of 4000-400 cm −1 at 2 cm −1 resolution. Fluorescence spectroscopy was carried out on a Hitachi F4500 spectrophotometer. Both excitation and emission slit width was 10.0 nm, scan speed was set to 200 nm min −1 .
Unconfined Compression and Stress Relaxation Testing: Elastic moduli Eo and stress relaxation properties of the hydrogels were measured from the compression tests of the hydrogels with an Instron 560 (Instron, Norwood, MA) using 10 N load cell to a compressive strain of 30% and a deformation rate of 1 mm s −1 . The cylindrical hydrogels (diameter = height = 5 mm) were swollen in PBS for 6 h prior measurements and the testing was carried out inside a flat-bottom petridish filled with PBS. Prior to the tests, an initial compressive contact of 0.01 N was applied to ensure a complete contact between the hydrogels and the petri-dish. No bulging of the side faces of the hydrogels was observed. The slope of the stress versus strain curve gives Eo. After the compression test, the strain was held constant for 5 min, while the load was recorded as a function of time. In order to calculate the stress, the force was divided by the area of the hydrogels in the undeformed state. Tests were carried out three times to ensure reproducibility.
Swelling Properties and Enzymatic Digestion of Hydrogels: In order to determine the swelling properties of the hydrogels, the wet hydrogels were first freeze-dried to a constant initial weight (Wi). The dried xerogels were then incubated in PBS at 37 °C. The wet weights (Wt) of the hydrogels were measured at various time intervals during incubation until there was no noticeable increase in the weight of the wet hydrogels. The percentage swelling of the hydrogels was calculated using the following equation (W t − W i )/W i × 100. Degradation rate of the hydrogels was characterized by incubating the cylindrical hydrogels in PBS (pH = 7.2) for 24 h and then treated with hyaluronidases (20 unit mL −1 ). The remaining weight of the hydrogels was measured up to 50 h after enzymatic treatment. The weight loss was computed using equation (W f − W i )/W i × 100, where W i and W f represent initial and final hydrogel weights, respectively.
Characterization of Microstructures: Microstructure of the hydrogels was observed using SEM. The hydrogels were frozen by liquid nitrogen and lyophilized to obtain dried samples. The dried gels were then sputtercoated with gold (10 nm thick) for 60 s. SEM micrographs of the dried xerogels were acquired on Inspect F50 (FEI Comp, the Netherlands). EDX spectroscopy was also used to analyze the dried gels coated with carbon to obtain elemental compositions of the dried xerogels.
In Vitro Experiments: Live/Dead assay-Live/dead assay was performed using a Live/Dead Cytotoxicity Kit (Thermo Fisher Scientific, UK) hMSCs (5000 cells mL −1 ) (Thermo Fisher Scientific, UK) seeded on various hydrogels at day 1, 3, and 7. Imaging was performed on an inverted confocal laser scanning microscope (CLSM, Leica, Germany). Proliferation-Cell proliferation was assessed by seeding 5000 hMSCs on various hydrogels. The cell seeded hydrogels were incubated at 37 °C under 5% CO2 condition. Cell proliferation was quantitatively measured at various time points (day 1, 4, and 7) using PrestoBlue reagent (PrestoBlue Cell Viability Reagent, Thermo Fisher Scientific, UK). Briefly, the spent media was removed from each well and the cells were incubated with PrestoBlue reagent solution (100 µL). The cells were incubated for 1 h at 37 °C, away from light. Florescence measurements were carried out with using excitation wavelength of 560 nm and an emission wavelength of 590 nm. Differentiation-Cell differentiation was assessed by seeding 20 000 hMSCs (passage 10) on various hydrogels with and without osteogenic media. Experiments were setup in 4 replicates. With osteogenic media, cells were first cultured with basal media and replaced with osteogenic media after 1 d. After each time point, cells were fixed using 4% formaldehyde for 10 min and washed thrice with sterile dH2O. Cells were incubated with 100 µL of SIGMAFAST BCIP/NBT reagent (Sigma, UK). ALP activity was spectroscopically quantified after days 4, 9, and 14 by measuring absorbance at 650 nm. Angiogenesis-hUVECs (5000) were seeded on various hydrogels. Cell-seeded hydrogels were incubated with supplemented endothelial cells growth media for 1 and 5 d. Cells were stained with calcein AM (2 × 10 −3 m) and optical images were acquired on CLSM (Leica, Germany).
In Vivo Maxillary Sinus Floor Reconstruction Procedure: Herein, 30 healthy adult New Zealand rabbits were randomly divided into two study groups (control and experimental groups) for observation at week 6 and week 12. General anesthesia was performed through injection of 3% sodium pentobarbital via the marginal ear vein (1 mL kg −1 ). After the disappearance of eyelash reflex, the hair in the surgery region was shaved and the region was sterilized by 1% Iodine volts. A 2.5 cm incision on the nasal skin along the midline was made and the periosteum was stripped to expose the nasal bone and nasoincisal suture line. Two round windows at each side of the midline were prepared using bone drill. The windows were 5 mm in diameter and located ≈20 mm anterior to the nasofrontal suture line and 10 mm lateral to the midline. During the osteotomy, sterile saline solution was injected to the drill for continuous cooling and the sinus membrane was carefully protected. The mucosa and periosteum were gently elevated with a periosteum elevator and 50 µL gel was prepared in situ within the space upon the surface of the bone. In the experimental groups, four types of gel, as subgroups, including HA-Tyr, HA-Tyr-GHK-Cu 2+ , HA-Tyr-Lap, and HA-Tyr-Lap-GHK-Cu 2+ was formed within the space. The space in the positive control group was filled with equal volume of Bio-oss and the space in blank control group was empty. The periosteum and skin were then sutured. Each group consisted of five animals.
Histological and Histomorphomeric Analysis: The animals were euthanized 6 and 12 weeks postsurgery by injecting sodium pentobarbital via the marginal ear vein (100 mg kg −1 ), and the sinuses were retrieved in blocks. The specimens were immediately fixed with paraformaldehyde (4%) for 48 h and demineralized by soaking them in 15% disodium ethylene diamine tetraacetate for 2 months and then washed in tap water for 15 min. All the specimens were trimmed and put into 70%, 80%, 90%, 95%, 100% ethanol step by step for gradually dehydrate, and finally embedded in paraffin. Longitudinal histological HE slices of 4 µm thickness were obtained to visualize the entire sinus. The sections were observed and digitally captured with a microscopic imaging system composed of a microscope (BX51, Olympus, Japan) and an image processing software, Cell Sens (Olympus, Japan). For the calculation of new bone area, Photoshop (Adobe Inc.) was first used to draw the outline of the new bone and ImageJ Fiji (developed by the National Institutes of Health and the Laboratory for Optical and Computational Instrumentation) was then used to calculate the new bone areas. Five slices were analyzed for each group. Fresh tissue blocks containing main organs including heart, liver, spleen, lung, and kidney were obtained and fixed with 4% paraformaldehyde to prepare HE stained slices analysis. The animal experimental protocol was approved by the Institutional Animal Care and Use Committee and the Ethics Committee of the School of Stomatology, Jilin University (Changchun, China).
Statistical Analysis: Statistical comparisons between groups treated with various hydrogels, blank (negative control), and Bio-oss (positive control group) were performed using 2-way ANOVA followed by Bonferroni tests if significant differences were observed using Graph Prism software (San Diego). Tukey's multiple comparison test was used for the cell proliferation and differentiation.

Supporting Information
Supporting Information is available from the Wiley Online Library or from the author.