Injectable Multifunctional Drug Delivery System for Hard Tissue Regeneration under Inflammatory Microenvironments

Abstract

Engineering multifunctional hydrogel systems capable of amplifying the regenerative capacity of endogenous progenitor cells via localized presentation of therapeutics under tissue inflammation is central to the translation of effective strategies for hard tissue regeneration. Here, we loaded dexamethasone (DEX), a pleotropic drug with anti-inflammatory and mineralizing abilities, into aluminosilicate clay nanotubes (halloysite clay nanotubes (HNTs)) to engineer an injectable multifunctional drug delivery system based on photo-cross-linkable gelatin methacryloyl (GelMA) hydrogel. In detail, a series of hydrogels based on GelMA formulations containing distinct amounts of DEX-loaded nanotubes was analyzed for physicochemical and mechanical properties and kinetics of DEX release as well as compatibility with mesenchymal stem cells from human exfoliated deciduous teeth (SHEDs). The anti-inflammatory response and mineralization potential of the engineered hydrogels were determined in vitro and in vivo. DEX conjugation with HNTs was confirmed by FTIR analysis. The incorporation of DEX-loaded nanotubes enhanced the mechanical strength of GelMA with no effect on its degradation and swelling ratio. Scanning electron microscopy (SEM) images demonstrated the porous architecture of GelMA, which was not significantly altered by DEX-loaded nanotubes’ (HNTs/DEX) incorporation. All GelMA formulations showed cytocompatibility with SHEDs (p < 0.05) regardless of the presence of HNTs or HNTs/DEX. However, the highest osteogenic cell differentiation was noticed with the addition of HNT/DEX 10% in GelMA formulations (p < 0.01). The controlled release of DEX over 7 days restored the expression of alkaline phosphatase and mineralization (p < 0.0001) in lipopolysaccharide (LPS)-stimulated SHEDs in vitro. Importantly, in vivo data revealed that DEX-loaded nanotube-modified GelMA (5.0% HNT/DEX 10%) led to enhanced bone formation after 6 weeks (p < 0.0001) compared to DEX-free formulations with a minimum localized inflammatory response after 7 days. Altogether, our findings show that the engineered DEX-loaded nanotube-modified hydrogel may possess great potential to trigger in situ mineralized tissue regeneration under inflammatory conditions.

Graphical Abstract

1. INTRODUCTION

According to the World Health Organization, dental caries and periodontitis are the most prevalent oral diseases that affect a significant part of the worldwide population.¹ These conditions are prompted by the tissue invasion of pathogens, such as Streptococci, Lactobacilli, and Porphyromonas, that gradually destroy mineralized dental tissues.^2-4 Clinically speaking, the release of bacterial metabolism byproducts suppresses the host defense mechanism of tissues, creating an inflammatory microenvironment, which may disturb the regenerative potential of endogenous cells.^5,6 Consequently, a cell-friendly biomaterial capable of ablating inflammation and promoting osteo/odontogenic differentiation of progenitor cells would be a significant advancement to the field of minimally invasive regenerative dentistry.

In recent years, hydrogels have been investigated due to their injectability and capacity to cater multifunctionality through precise modification with biological cues and/or inorganic nanocontainers that can function as a carrier for therapeutic agents.⁷ Additionally, hydrogels offer adjustable biodegradation, good biocompatibility, and a tunable 3D porous polymeric network key in cell adhesion, proliferation, and phenotype expression.^8-12 In this way, gelatin methacryloyl (GelMA) is considered an excellent candidate for mineralized tissue regeneration, as it can be perfectly injected into small, irregular areas, acquiring suitable mechanical properties upon in situ photo-cross-linking.^13,14

It is known that the successful achievement of dentin and periodontal tissue regeneration lies in the management of the current local inflammation; thus, itis extremely rational to devise a drug delivery system with anti-inflammatory action.^15,16 Lipopolysaccharide (LPS) is an endotoxin component present in the membrane of bacteria commonly found in deep caries or periodontitis, leading to pulp inflammation and bone resorption by the activation of pro-inflammatory cytokines.^17-19 Accordingly, pleiotropic drugs capable of amplifying the oste/odontoogenic differentiation ability of stem cells under inflammatory conditions have been deemed a suitable approach to creating drug delivery systems and/or biomaterials for mineralized tissue regeneration.²⁰ For example, dexamethasone (DEX), a synthetic glucocorticoid commonly utilized to treat inflammation,²¹ features a mechanism of action related with the early shift expression of anti-inflammatory cytokines, which helps in reducing pain and guiding the cellular response.²² Additionally, this molecule has proven to positively boost mesenchymal stem cell (MSC) differentiation in odonto/osteogenic phenotypes.^21,23-26 Nonetheless, local administration of DEX may induce toxic effects, depending on its concentration, thus impairing the aforesaid differentiation ability.^21,27

To overcome these limitations, the development of drug delivery systems has been considered crucial to modulate the desired cell phenomenon adjusted to the intended use.^28-31 Among available systems, halloysite clay nanotubes (HNTs) have been proven to be an adequate platform to promote sustained drug delivery in association with GelMA.^14,32 HNTs are naturally formed by the exfoliation process of aluminosilicate minerals (Al₂Si₂O₅(OH)₄·nH₂O). This process results in a bilayered hollow tubular structure with a positively charged inner surface afforded by alumina composition (Al─OH) and a negatively charged external surface consisting of silicon dioxide (SiO₂). Such features provide a hydrophilic and hydrophobic ligand capacity for HNTs^33-35 that facilitates functionalization with small therapeutic molecules such as DEX for controlled drug release.³⁶ Recent studies have reported on the promising ability of HNTs to function as a drug carrier platform to enhance bone regeneration,^37-39 stimulate cartilage formation,^40,41 target specific cancer cells,^42,43 and accelerate wound healing.^44,45 However, the literature still lacks studies that investigate strategies for hard tissue regeneration mediated by a drug delivery system composed of DEX and HNTs. In this investigation, we postulate that devising a drug delivery system that leverages the established pleiotropic capacity of DEX with the well demonstrated biocompatibility of GelMA and the controlled release benefits of drug loading into Halloysite could pave the way toward an innovative therapeutic strategy to amplify the regeneration of oral mineralized tissues by resident stem cells under inflammatory environments. Therefore, this work reports the fabrication of an injectable GelMA-based drug delivery system capable of augmenting the differentiation capacity of stem cells from human exfoliated deciduous teeth under inflammatory stimulus and to promote in vivo bone regeneration as a cell-free platform.

2. EXPERIMENTAL METHODS

2.1. Materials and Chemicals.

Methacrylic anhydride, type A gelatin from porcine skin with ~300 g bloom strength, l-ascorbic acid, Triton X-100, cetylpyridinium chloride, β-glycerophosphate disodium salt hydrate, Alizarin Red S, and poly(vinylpyrrolidone)–iodine complex were procured from Sigma-Aldrich (St. Louis, MO, USA). Dialysis tube membrane (Spectro/Pot, 12–14 kDa), phosphate-buffered saline (PBS), absolute-200 proof ethanol, l-glutamine (200 mM), fetal bovine serum (FBS), penicillin–streptomycin (5,000 U/mL), minimum essential Eagle-alpha-modified medium (α-MEM), and ethanol anhydrous were obtained from Thermo Fisher Scientific (Waltham, MA, USA). Halloysite nanotubes (HNTs, Halloysite, Dragonite HP) were kindly given by Applied Minerals (New York, NY, USA). Dexamethasone (TCI America, Portland, OR, USA), SensoLyte pNPP alkaline phosphatase assay kit (AnaSpec, Fremont, CA, USA), collagenase A (Hoffman La-Roche AG, Basel, Switzerland, CHE), CellTiter 96 AQueous One Solution Reagent (Promega Corporation, Madison, WI, USA), isoflurane (Piramal Critical Care, Bethlehem, PA, USA), Dulbecco’s phosphate-buffered saline (DPBS, Gibco Invitrogen Corporation, Grand Island, NY, USA), lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP L0290, TCI America), and normal saline solution (Baxter International, Deerfield, IL, USA) were obtained from the aforementioned suppliers.

2.2. Gelatin Methacrylate (GelMA).

GelMA fabrication was carried out as reported elsewhere.¹⁴ 10 g (w/v) of gelatin type A from porcine skin was solubilized into 100 mL of DPBS at 50 °C. Following that, gelatin methacrylation was accomplished by dripping methacrylic anhydride into the solubilized gelatin under stirring conditions. After 2 h, warm DPBS was added, and the prepared solution was transferred into a dialysis tube that remained in deionized (DI) water at 45 ± 5 °C for 7 days to remove methacrylic acid and nonreacted anhydride. Of note, throughout the dialysis process, the DI water was changed twice a day. The GelMA solution was filtered, followed by lyophilization (Labconco Corporation, Kansas City, MO, USA) for 1 week to obtain the GelMA porous foam.

2.3. Fabrication/Characterization of DEX-Loaded Nanotube-Modified GelMA.

The unique morphology of the HNTs was visualized by transmission electron microscopy (TEM, Tecnai BioTWIN, FEI Company, Hillsboro, OR, USA) following standard TEM nanoparticle sample preparation.¹⁴ Next, dexamethasone (DEX, TCI America, Inc.) loading into the nanotubes was performed as described by our group.¹⁴ DEX in powder form was utilized to obtain 10% and 20% DEX solutions (w/v) in ethanol anhydrous. Previously sieved (<45 μm) nanotubes (2.5 g) and 10 mL of the respective DEX solutions were centrifuged, vortexed for 20 s, and sonicated (Ultrassonic bath CPX5800; Thermo Fisher Scientific, Inc.) for 120 min. The mixture was then subjected to vacuum conditions (Hi-Temp Vacuum, Thermo Fisher Scientific, Inc.) for 1 h at 23 °C, vortexed for 1 h, and then once again submitted to the same vacuum conditions.¹⁴ Finally, the HNT/DEX solutions were washed and centrifuged (3000 rpm) for 10 min and then stored at 37 °C for 1 week until fully dried.¹⁴ Next, the mixture was sieved at 45 μm, and then, DEX-loaded nanotubes were obtained (hereafter referred to as H-D10% and H-D20%, respectively) for further use.¹⁴ Fourier transform infrared spectroscopy in the attenuated total reflection mode (ATR-FTIR, Nicolet iS50, Thermo Fisher Scientific, Inc.) was done to determine the drug’s presence in DEX-loaded nanotubes (64 scans with spectra between 4000 and 400 cm⁻¹) as previously described by our group.¹⁴

To obtain DEX-loaded nanotube-modified hydrogels, lyophilized GelMA (20% w/v) was dissolved in DPBS at 50 °C. Next, two different concentrations (2.5% and 5%, w/v) of pristine as well as DEX-loaded nanotubes (H-DEX10% and H-DEX20%) were added to the GelMA solution and kept under stirring conditions to enhance the nanotubes’ dispersion. LAP was used as the photoinitiator (0.05% w/v).¹⁴ The following GelMA formulations (GelMA [G], G+2.5%HNT [G+2.5H], G+5.0%HNT [G+5.0H], G+2.5%HNT-DEX10% [G+2.5H-D10], G +5.0%HNT-DEX10% [G+5H-D10], G+2.5%HNT-DEX20% [G +2.5H-D20], and G+5.0%HNT-DEX20% [G+5H-D20]) were made to fabricate hydrogel samples for the various assays reported in this work. Briefly, 100 μL of each GelMA-based solution was pipetted in silicone (CutterSil Putty PLUS, Kulzer Dental–North America, South Bend, IN, USA) molds (8 mm diameter × 3 mm thick) followed by photo-cross-linking for 60 s with a light-emitting diode (LED) device (Bluephase, Ivoclar-Vivadent, Amherst, NY, USA).

2.4. Drug Release.

To determine the amount of drug released from the DEX-loaded nanotube-modified GelMA, the distinct hydrogel samples (n = 4/group) formulated by incorporating 2.5% and 5% amounts of H-D10% and H-D20% were submerged in 1.5 mL of degradation solution (DPBS and 1 U/mL collagenase A; Hoffman La-Roche AG) followed by incubation at 37 °C for up to 14 days. At predetermined time points, 500 μL of supernatant was collected and frozen at −20 °C. Fresh enzymatic solution was replenished with the same amount collected at each time point and every 48 h.¹⁴ The amount of DEX released from the GelMA hydrogels was determined by a UV–vis spectrophotometer (SpectraMax iD3, Molecular Devices, LLC, San Jose, CA, USA) at a 242 nm wavelength. DEX concentrations were established by comparison with a standard curve plotted with known concentrations of DEX.

2.5. Hydrogel Swelling and Enzymatic Degradation.

The swelling rate and degradation profile for all the hydrogels (i.e., G, G +2.5H, G+5.0H, G+2.5H-D10, G+5H-D10, G+2.5H-D20, and G+5H-D20) were determined as described in detail by us.¹⁴ In brief, wet weight was obtained by the immersion of samples in DPBS, and the dry weight was reached after the lyophilization process. Swelling rate (%) was defined after 24 h, while the degradation profile was estimated up to 28 days. Data were calculated using the respective swelling rate (%) and degradation ratio (%) as detailed by Luo et al.⁴⁶

2.6. Biomechanical Characterization.

To determine the biomechanical properties (i.e., compressive strength and modulus), cylindrical-shaped (n = 5/group; 8 mm diameter × 3 mm thick) samples were made. Briefly, 150 μL of the various hydrogels was placed into silicone molds and photo-cross-linked for 60 s.¹⁴ After 24 h of incubation in DPBS, the samples were gently dried (Kimberly-Clark Corporation) and uniaxial compression testing (expert 5601, ADMET, Norwood, MA, USA) was performed.¹⁴ The slope of the linear region of the stress–strain curves (0–10% strain) was used to calculate the compressive modulus.¹⁴

2.7. In Vitro Biological Studies of the DEX-Loaded GelMA Hydrogels.

MSCs from human exfoliated deciduous teeth (SHEDs) have been selected for this study considering the self-renewal and multidifferentiation ability of this lineage. Also, several studies have reported the effective capacity of these cells in inducing the regeneration of mineralized tissues (e.g., bone and dentin).^47-50 Previously isolated and characterized SHEDs have been generously donated by Dr. Jacques Nör (University of Michigan, School of Dentistry, Ann Arbor, MI, USA).^51,52 The cells were grown in α-MEM supplemented with 10% FBS, 1% l-glutamine, and 1% penicillin–streptomycin at 37 °C and 5% CO₂ as previously described. Cells at passages 4–6 were used in all the experiments. Two experimental designs were performed, as given below.

To analyze the biological compatibility and functionality (i.e., bioactivity, induction of odonto/osteogenic differentiation) of GelMA-based formulations, cell viability, alkaline phosphatase (ALP) activity, and mineral deposition assays were performed as described below. SHEDs (1 × 10⁴ cells/sample) were seeded in one drop (15 μL) on top of the hydrogels in sextuplicate and incubated for 30 min at 37 °C, corresponding to the cell adhesion period on the hydrogel samples. Subsequently, 500 μL of α-MEM was then added, and the cell/hydrogel constructs were cultured for 21 days. Cells seeded directly on 48-well plates (Corning, NY, USA) were used as the negative control (SHEDs) and considered to be 100% of the cell parameters.

2.7.1. Cell Viability.

At predetermined time points (up to 21 days), a total of 60 μL of the MTS Assay (Promega Corporation) was placed in each well containing 500 μL of complete medium, followed by incubation for 2 h at 37 °C. Then, 100 μL from each well was transferred into the wells of a 96-well plate in triplicate. The absorbance was determined at 490 nm using a plate reader (SpectraMax iD3, Molecular Devices LLC., San Jose, CA, USA).

2.7.2. ALP Activity.

For this assay, cell/hydrogel constructs or cells on plates (n = 6) were cultured in osteogenic medium (α-MEM containing FBS, 50 μg/mL ascorbic acid, and 10 mmol/L β-glycerophosphate). At predetermined time points (7, 14, and 21 days), the samples were incubated for 10 min at 4 °C with a lysis buffer (Triton X-100; Sigma-Aldrich). The cell/hydrogel constructs were subjected to manual disruption. After centrifugation (10 000g at 4 °C for 15 min), the supernatant was transferred to a vial, and alkaline phosphatase activity was detected using the SensoLyte pNPP kit (AnaSpec). ALP activity was determined by data normalization with the total protein obtained by the Pierce BCA protein assay kit (Thermo Fisher Scientific) after incubation of the samples for 30 min with the working reagent and measurement of absorbance at 562 nm (SpectraMax iD3).

2.7.3. Mineral Deposition.

Alizarin Red staining (Sigma-Aldrich) was performed on 21-day-old cultures in osteogenic medium. The cell/hydrogel constructs or cells on plates (n = 6) were incubated for 1 h at 4 °C in 70% ethanol. After that, samples were washed with DI water followed by incubation with Alizarin Red staining (40 mM, pH 4.2) for 15 min. The samples were washed 5× in DI water. For cells on plates, images of mineralization nodules were captured (BZ-X 710; Keyence Corporation of America, Itasca, IL, USA). To quantify mineralization, samples were disrupted in 10% cetylpyridinium chloride and absorbance was read at 570 nm (SpectraMax iD3). For cell/hydrogel constructs, hydrogel samples with no cells were employed as a background control and supernatant was obtained by centrifugation.

To evaluate the potential of selected DEX-loaded nanotube-modified hydrogel formulations as a potential drug delivery strategy under an inflammatory environment, SHEDs (1.5 × 10⁴ cells/well) were cultured on the bottom of 24-well tissue culture plates and incubated for 24 h at 37 °C in complete α-MEM. The medium was then replaced by osteogenic medium with or without 10 μg/mL lipopolysaccharide (LPS) from E. coli (Sigma-Aldrich). The aforesaid LPS concentration was carefully selected from a systematic dose–response assay aiming to induce a degenerative stimulus on SHEDs capable of reducing its mineralized matrix deposition to ca. 50% without cell toxicity (Supporting Information and Figure S1A). To do this, Transwell inserts (24-well, 8 μm pore size, Corning) were adapted in the same plate where the cells were cultured, and GelMA hydrogel samples modified or not with 5%HNT and 5%HNT-DEX10% were individually placed in the upper chamber of the Transwells to mimic a direct pulp capping clinical scenario. The sets (SHEDs + Transwells + hydrogels) were incubated for 21 days, and in vitro biological assays were performed to determine cell viability (MTS, 1, 7, 14, and 21 days), ALP activity (SensoLyte, 7, 14, and 21 days), and mineral deposition (Alizarin Red, 21 days). Cells cultured in osteogenic medium with and without LPS were used as positive and negative controls, respectively.

2.8. Biocompatibility and Bone Regeneration Studies.

All animal work was performed under an approved Institutional Animal Care Committee (IACUC, protocol #PRO00008502, University of Michigan). Rats were housed in standard cages in groups of three and allocated in a room with monitored temperature (23 ± 2 °C) and humidity (55 ± 10%) and a 12 h dark–light cycle. Water and food were available ad libitum. Therefore, the animals were standardized to allow for the evaluation of biological parameters. Determination of the sample size was estimated using previously similar research studies^14,53-55 that investigated in vivo bone formation and biocompatibility of the formulated materials after implanting in critical-size calvarial defects and subcutaneously in rats. Briefly, 4 rats per group were used for each methodology based on these studies, which demonstrated a 50% difference among the experimental and control groups, regarding an alpha error of 0.05 to detect significant differences.

2.8.1. Subcutaneous Implantation Model.

Twelve Fischer 344 rats (6 week-old male, 300–320 g, Envigo RMS, Oxford, MI, USA) were utilized to investigate the effects of the formulated GelMA hydrogels on the inflammatory response and overall in vivo biocompatibility. All surgical procedures were carried out as described by our group.¹⁴ Implantation of the fabricated GelMA hydrogels in dorsal subcutaneous pockets in the rats (n = 4/group) was performed under induced general anesthesia and maintained with the inhalation of isofluorane (Piramal Critical Care Inc., Bethlehem, PA, USA) as recently reported by us.¹⁴ Polyethylene tubes (1.5 mm diameter × 10 mm length; Braintree Scientific, Braintree, MA, USA) filled with different experimental groups, GelMA (G), GelMA+5%HNT (G+5.0H), and GelMA+5% HNT-DEX10% (G+5.0H-D10), were implanted in subcutaneous pockets located on the dorsal skin of the rats. After 7 days, euthanasia was performed using CO₂ inhalation, and the samples, including adjacent tissue, were retrieved and processed for histology.⁵⁵ Hematoxylin and Eosin (H&E) staining was performed to evaluate the presence of luminal structures containing red blood and inflammatory cells under light microscopy (Nikon E800, Nikon Corporation, Tokyo, Japan).^3,55 The results were compared with unfilled polyethylene tubes (SHAM, control).

2.8.2. Bone Regeneration: Critical Size Rat Calvarial Defects.

Eight 6 week-old male Fischer 344 rats (Envigo RMS) weighing ~300–320 g were used to determine the regenerative potential of the DEX-loaded nanotube-modified hydrogels. The surgical procedures were carried out as described by Dubey et al.^3,55 Bilateral defects (5 mm diameter) were performed, and animals were randomly assigned to SHAM (negative control, defect only), GelMA (G), G+5.0%HNT (G+5.0H), and G +5.0%HNT-DEX10% (G+5.0H-D10). The three distinct hydrogel formulations were injected into the created defects and photo-crosslinked in situ for 60 s (Bluephase, Ivoclar-Vivadent). The animals were euthanized 6 weeks postimplantation by the conventional CO₂ protocol, and the skulls were collected and fixed in 4% paraformaldehyde (TCI America). Bone regeneration was assessed by microcomputed tomography (micro-CT, Scanco μCT 100 Medical AG, Scanco Medical AG, Brüttisellen, Switzerland) and histologically by H&E and Masson’s trichrome (MT) staining as previously described.^3,55 Micro-CT analysis was performed with samples in 360° rotation using 70 kV, 25 μm voxel size, and 114 μA monochromatic X-rays, and an exposure time at an average of 3 frames per 500 ms. Images were reconstructed at 1 μm scale using Scanco Medical software. Reconstructed 3D images were traced to the original defect circumference (i.e., the region of interest, ROI). The ROI of each sample around the defect was evaluated for bone volume (BV), tissue volume (TV), and bone volume fraction (BVF, BV/TV). Next, following decalcification, the samples were embedded in paraffin, cut into 5 μm thick sections, stained using H&E and MT to identify osteoid (red) and mineralized bone tissue (blue). Photomicrographs were taken using a digital camera coupled to a light microscope at distinct magnifications.

2.8.3. Statistics.

Data are shown as mean ± standard deviation. One- or two-way ANOVA was used to detect differences among groups, followed by Tukey’s posthoc test. The significance level was set at 5%.

3. RESULTS AND DISCUSSION

Our findings demonstrated a controlled release delivery system for dexamethasone from an injectable GelMA-based nanotube-modified hydrogel, illustrating the potential of this system for mineralized tissue regeneration. Overall, DEX-loaded nanotube-modified GelMA hydrogel, more specifically, the G+5H-D10 formulation (GelMA+5.0%HNT-DEX10%), showed suitable mechanical properties, biodegradability, cytocompatibility with MSCs (SHEDs), and in vivo biocompatibility, and it also supported bone regeneration in vivo.

3.1. DEX-Loaded Nanotube-Modified GelMA Hydrogels.

TEM images of the aluminosilicate clay (Halloysite, HNTs) nanocontainers utilized in the fabrication of the DEX-loaded nanotube-modified hydrogels are shown in Figure 1A. A typical tubular morphology with a defined open lumen measuring around 150 to 500 nm in length and approximately 35 to 75 nm in diameter can be seen (Figure 1A). According to the literature, at a pH below 8.5, HNTs exhibit a positively charged inner surface (i.e., lumen) and a negatively charged external surface, making negatively charged drugs, such as DEX, preferentially loaded inside the lumen.⁵⁶ Here, to obtain DEX-loaded nanotubes, DEX was dissolved in 100% ethanol, since it has been demonstrated that DEX loading is greater at a higher ethanol-to-water ratio as DEX solubility increases. DEX loading/impregnation into the nanotubes was assessed by FTIR. Analysis of pristine HNTs revealed peaks at 3750 cm⁻¹ related to the Al₂─OH stretching vibration and an asymmetrical strong peak at 1100 cm⁻¹, respective to the Si─O bond. DEX powder exhibited a smooth band at 3387 cm⁻¹, indicating stretch vibrations of −OH and −NH₂, as well as at 1100 and 1200 cm⁻¹ corresponding to C─C and C─O─C bonds, respectively. When the DEX-loaded nanotubes (HNT-DEX10% and HNT-DEX20%) were evaluated, the small peaks at 1660 and 1670 cm⁻¹, assigned to the amide I band, suggested DEX conjugation with the nanotubes (Figure 1B).

Figure 1.

Morphological and chemical analyses of aluminosilicate clay Halloysite nanotubes (H) loaded or not with dexamethasone. (A) Representative TEM micrograph of the pristine nanotubes. High magnification TEM micrograph demonstrating the well-defined open lumen and overall morphology of the nanotubes employed in this study (scale bars = 100 nm). (B) FTIR spectrum of the nanotubes (H), dexamethasone powder (D), and DEX-loaded nanotubes (H-D10 [10%] and H-D20 [20%]).

HNTs have proven to be a suitable strategy for promoting sustained drug release when associated with GelMA, since they can be fairly homogeneously dispersed into the hydrogel structure and do not interfere with the overall hydrogel porous microstructure.¹⁴ In Figure 2A, the scanning electron microscopy (SEM) images demonstrate that the porous architecture of GelMA was not drastically affected by HNT-DEX10% incorporation, regardless of the loading amount (2.5% vs 5%). On the other hand, GelMA modified with HNT-DEX20% revealed a heterogeneous porous matrix. High magnification SEM images suggest the presence of HNTs dispersed throughout the hydrogel surface (Figure 2A). Veerabadran et al.⁵⁶ first described the encapsulation of DEX into the HNTs and demonstrated a 75 times longer release of HNT-DEX compared to the DEX microcrystals when the samples were immersed in water at pH 7.4. In this investigation, DEX solutions at 10% and 20% were used to load DEX into HNTs prior to DEX-loaded nanotube incorporation at two distinct amounts (2.5% and 5%) into GelMA to refine the most promising drug delivery candidate for preclinical in vivo analyses to determine biocompatibility and bone regeneration. Overall, DEX release occurred in a dose-dependent fashion.

Figure 2.

(A) Morphological characterization of GelMA-based hydrogel formulations. Representative SEM images of hydrogels’ cross sections to reveal the internal microstructure. Note the well-dispersed DEX-loaded nanotubes in the nanotube-modified GelMA-based formulations (yellow arrows). (B) Mean DEX release over 14 days (336 h) for DEX-loaded nanotube-modified hydrogels after incubation at 37 °C with an enzymatic solution (PBS + 1U/mL collagenase A). Data represent mean ± SD (n = 6); p < 0.05.

Figure 2B shows the release profiles of DEX-loaded nanotube-modified GelMA formulations. A steady increase in DEX release was observed for all groups up to 15 h, which was then sustained for G+5.0H-D20 and G+2.5H-D20 groups until day 7, followed by a significant reduction of DEX liberation by day 14. Meanwhile, the G+2.5H-D10 and G+5.0H-D10 formulations experienced a sustained reduction in the amount of DEX release up to day 7 with minimal release by day 14. Taken together, the initial fast release of DEX might be related to its bound presence at the outer surface of the incomplete rolled nanotubes, which leads to the exposure of some inner alumina layer.⁵⁶

3.2. Degradation, Swelling, and Biomechanical Analyses of DEX-Loaded Nanotube-Modified GelMA.

Besides biological functionalities, the design of the hydrogels to act as drug delivery systems should also consider the biomaterials’ parameters, such as degradation, swelling, and biomechanical properties compatible with the intended clinical application. Indeed, the gradual degradation of biomaterials is key to supporting stem cells’ differentiation and should match the rate of new tissue formation.⁵⁷ In this study, the effect of DEX-loaded nanotube incorporated into the GelMA hydrogel was investigated by mass loss and swelling experiments. The degradation profile of the engineered hydrogels was assessed up to 28 days in an incubation medium containing collagenase. Our GelMA formulation (20%, w/v) featured a controlled degradation rate (G), even under enzymatic challenge, with a mass loss of 24.6%, 30.5%, and 38.5% at 14, 21, and 28 days, respectively (Figure 3A). This is an important feature, since a previous study in our group demonstrated that GelMA at 15% started to degrade immediately after incubation in a similar enzymatic challenge with the total mass loss detected after 21 days.¹⁴ Thus, the more stable degradation profile seems to be related to the GelMA concentration, in agreement with previous studies.¹⁴ Overall, nanotube incorporation, regardless of DEX presence, increased GelMA degradability in a dose-dependent fashion. The greater degradability values were associated with groups modified with 5%HNT (i.e., G+5H, G+5H-10D, and G +5H-20D). Nonetheless, it is important to note that all groups remained stable up to 7 days with a mass loss varying from 13% to 30% at 14 days (Figure 3A). Cidonio et al.⁵⁸ found that clay nanoparticles (Laponite) interfered with GelMA’s degradation, where an increased soluble fraction was observed. The authors demonstrated that 5%, 7.5%, and 10% GelMA concentrations featured an intense degradation profile after the incorporation of 0.5 to 1 wt % Laponite. An enhanced soluble fraction occurred regardless of Laponite concentration and increased swelling ratio. Indeed, Ribeiro et al.¹⁴ detected an increased swelling ratio for 15% GelMA after incorporation of 1–5 wt % HNTs. Meanwhile, in the present study, 20% GelMA modified with nanotubes revealed no significant alteration on the swelling ratio, irrespective of HNT concentration and DEX presence (Figure 3B). This indicates that the original structure of GelMA was maintained with no alteration in the polymer network after the interaction between the hydroxyl groups present on the HNT surface with the hydroxyl groups of GelMA.^59,60 Therefore, one may conclude that GelMA at 20% seems to be an adequate drug delivery platform when using nanotubes as the carrier, since it displayed a suitable degradation profile over 28 days with formulations containing DEX showing nearly 80% mass loss, which in turn, could be of significant importance to support mineralized tissue regeneration in vivo.

Figure 3.

Enzymatic hydrogel degradation, swelling ratio, and mechanical properties of the formulated hydrogels. (A) In vitro hydrogel degradation after enzymatic challenge. Mass loss (%) was determined over 28 days. (B) Swelling ratio of the distinct hydrogels was measured upon 24 h of PBS incubation at 37 °C. (C) Compressive modulus (kPa). Different letters denote statistical differences among the experimental groups and the control (one-way ANOVA, Tukey’s test, n = 6; p < 0.05). (D) Representative stress–strain curves for all groups.

Recent investigations have demonstrated that the incorporation of optimal amounts of nanoparticles into the hydrogel’s composition represents a viable approach to enhance compressive modulus.^50,61,62 According to Figure 3C, the addition of the nanotubes modified or not with DEX enhanced the compressive modulus of GelMA, except for G+2.5H, which featured no differences from unmodified GelMA (G). Thus, the tested formulations maintained a constant degree of methacrylation, even with the addition of HNTs into the GelMA framework without intervening negatively in the mechanical behavior (Figure 3D).

3.3. Biological Characterization of GelMA HNT-DEX Composite.

To assess the cytocompatibility and bioactive potential of GelMA formulations, MSCs from pulp tissue (SHEDs) were seeded onto photo-cross-linked hydrogels. Cells seeded directly onto the plates were used as a negative control for all in vitro experiments. An MTS-based cell viability assay demonstrated that nanotube incorporation with or without DEX did not lead to cell toxicity over 21 days (Figure 4A). Importantly, at 21 days, a considerable increase in cell viability (%) was detected for all groups in comparison with the control. It seems that the cell interaction mediated by GelMA is a consequence of its composition, since gelatin presents arginine–glycine–aspartic acid sequences that favor cell adhesion and proliferation.^63,64 Moreover, the compatibility and increased proliferative potential of fibroblasts, osteoblasts, and pulp cells on GelMA have been demonstrated.^63,64

Figure 4.

In vitro biological assessment of GelMA-based hydrogel formulations. (A) Cell viability over time of SHEDs seeded on GelMA-based hydrogel samples (MTS assay). Cell viability (%) was normalized to SHEDs cultured on tissue culture plate at day 1 (100%). (B) ALP activity in SHEDs after 7 and 14 days. At day 14, G+5.0H, G+2.5H-10D, and G+5.0H-10D promoted significant upregulation of ALP expression compared to the control.

To assess the odonto/osteogenic potential of the DEX-loaded nanotube-modified hydrogels, SHEDs were cultured on the different formulations and assayed for ALP activity and mineralized matrix formation. The analysis of ALP expression has been used to determine cell differentiation, since ALP is a highly activated enzyme in the early stages of mineralized tissue formation.⁶⁵ In this study, when SHEDs were cultured in direct contact with G+5H, G+2.5H-D10, and G+5H-D10, significantly higher ALP activity was evidenced as compared with control groups (SHEDs only and unmodified GelMA, G) at 14 days with lower ALP activity being detected at 7 days in comparison to the negative control (Figure 4B). This effect denotes the lower ALP activity expression of SHEDs at the initial stages with the peak being observed at 14 days of cell culture, as previously demonstrated in the literature for pulp-derived cells.^50,66-68 Nevertheless, the bioactive potential was confirmed by the Alizarin Red assay as enhanced mineralized matrix deposition occurred at 14 and 21 days for cells cultured on G+2.5H-D10 and G+5H-D10 with a considerable difference when compared to the control and G groups. The highest mineralized matrix deposition was seen at 21 days for SHEDs cultured on G+2.5H-D10 and G+5H-D10 (Figure 5). The osteogenic potential of DEX-loaded nanotube-modified hydrogels at the lowest DEX concentration (i.e., 2.5H-D10 and 5.0H-D10) may be explained by the lower amount of DEX release in comparison to groups containing nanotubes modified with DEX20%: 2.5H-D20 and 5.0H-D20. According to the literature, depending on the DEX concentration, MSCs can undergo adipogenic, chondrogenic, or myogenic differentiation.⁶⁹ A gradual and sustained liberation of low doses of DEX through different drug release systems has proven to be key in promoting osteogenic differentiation of MSCs by upregulating RUNX2, COL-1, BMP-2, OCN, and ALP gene expression.⁷⁰

Figure 5.

Alizarin Red assay for GelMA-based hydrogel formulations. (A) Quantitative Alizarin Red staining showing mineralized matrix deposition in SHEDs seeded on the hydrogels at 14 and 21 days. All experimental groups led to higher mineralization compared to the control at day 21. Different letters denote statistical differences among the experimental groups and the control group (one-way ANOVA, Tukey’s test, n = 6; p < 0.05). (B) Exemplary images of the Alizarin Red assay at 14 and 21 days.

3.4. DEX-Loaded Nanotube-Modified Hydrogel as a Multifunctional Drug Delivery Strategy.

From a clinical standpoint, exposed pulps and damaged periodontium are often under inflammatory conditions, since stem cells from the dental pulp and periodontal ligament recognize bacterial components released to elicit innate immune tissue response by triggering the Toll-like receptors 2 and 4 (TLR2 and TLR4).⁷¹ This process leads to pro-inflammatory mediators’ expression, such as interleukin (IL)-1β, IL-17, and TNF-α cytokines, which are quickly liberated in infected or damaged tissues;^72,73 nevertheless, these cytokines at high concentration can inhibit the stem cells’ regenerative ability since they downregulate the expression of osteo/odontoblastic phenotype, reducing or even impairing mineralized matrix deposition.^74-78

To assess the potential of the most promising DEX-loaded nanotube-modified GelMA formulation (i.e., G+5.0-D10) as a multifunctional therapeutic platform to amplify mineralized tissue regeneration under an inflammatory microenvironment, we performed an in vitro indirect contact assay using a Transwell to house the hydrogel while assessing the effects of DEX release on LPS-treated SHEDs cultured on the bottom of the tissue culture plate. Recent studies have demonstrated the relevance of low intensity pulp inflammation on the activation of MSC-mediated tissue regeneration. The authors reinforced the efficient in vitro LPS treatment to mimic the in vivo damage in cells to stimulate self-repair.^79,80 Therefore, on the basis of our pilot studies, we performed the aforementioned assay using an optimized LPS concentration capable of reducing the odonto/osteogenic differentiation of SHEDs without negatively interfering with cell viability (Supporting Information). According to Figure S1A, all LPS concentrations tested were not cytotoxic; a small decrease in viability by 13.6% and 16% was detected in those groups incubated with 10 μg/mL LPS at 14 days and 0.025 μg/mL LPS at 21 days, respectively. LPS concentrations of 0.1, 1.0, and 10 μg/mL significantly decreased the differentiation capacity of SHEDs according to the data from the Alizarin Red assay (Figure S1B,C).

The positive effect of G+5H-D10 on SHEDs was observed on day 1 when cell viability increased by 37% in comparison to the control (LPS−). A lack of significant differences was observed for the other groups at all time points, demonstrating that LPS treatment did not influence cell viability in the presence of GelMA-based hydrogels (Figure 6A).

Figure 6.

Cell viability and bioactivity of GelMA-based hydrogel formulations on LPS (10 μg/mL)-stimulated SHEDs. The assay was performed by individually placing the formulated hydrogels on a plastic Transwell insert with permeable support to allow the facile transport of released drug into the underlying cellular monolayer. (A) Cell viability analysis over 21 days using the MTS assay (normalized to SHEDs cultured on a flat-bottom tissue culture plate at day 1 without the plastic insert). Different letters denote statistical differences among the experimental groups and the control group (SHEDs) in each time point (two-way ANOVA, Tukey’s test, n = 6; p < 0.05). (B) ALP activity of SHEDs over time presented an increasing trend with G+5.0H-D10 (LPS−) and G+5.0H-D10 (LPS−) showing the highest activity at 14 and 21 days. Different letters denote statistical differences among the experimental groups and the control group (SHEDs) at day 1 (two-way ANOVA, Tukey’s test, n = 6; p < 0.05).

According to our data, DEX-loaded nanotube-modified GelMA (i.e., G+5.0H-D10) increased the odonto/osteogenic potential of SHEDs when cultured in the presence of the proposed drug delivery system. In detail, significant increases in ALP activity were detected at 14 and 21 days for the G+5.0H-D10 group in comparison to all other groups, irrespective of LPS treatment. This positive effect was also observed for the Alizarin Red assay with an improvement in mineral deposition in comparison to the control of ~50.5% in the absence of LPS. A more intense effect was detected for LPS-treated cells with an increase of 119.4% in comparison to the control (LPS+). When we compared these groups with plain GelMA (G), we detected a 70.5% and 83.8% increase for G+5H-D10 in the absence and presence of LPS, respectively. As expected, LPS treatment promoted a significant reduction in differentiation capacity of SHEDs seeded on the bottom of the tissue culture plate (control), validating the experimental model used. In the control (LPS+), a significant reduction of 32% and 65% on ALP activity (Figure 6B) and mineralized matrix deposition (Figure 7A), respectively, was observed in comparison to the control (LPS−) at 21 days.

Figure 7.

Alizarin Red assay of GelMA-based hydrogel formulations on LPS (10 μg/mL)-stimulated SHEDs using a Transwell insert. (A) Quantitative Alizarin Red data showing the highest mineralization for G+5.0H-D10 (LPS−) and G+5.0H-D10 (LPS−). Different letters denote statistical differences among the experimental groups and the control group (SHEDs) (one-way ANOVA, Tukey’s test, n = 6; p < 0.05). (B) Exemplary images from the Alizarin Red assay at 21 days.

Nevertheless, it is interesting to note that, in the presence of GelMA, irrespective of nanotube incorporation, LPS had no significant impact on ALP activity and mineralized matrix deposition, as no considerable differences were noted between the LPS+ and LPS− homologous groups. Therefore, it seems that GelMA had a protective effect on LPS-stimulated SHEDs in vitro, which is probably due to RGD sequences’ release from its structure that positively modulates cell behavior.^81-85 The RGD motifs remain active and available to interact with cells after the reaction of methacrylic anhydride in gelatin,⁸⁶ and these sequences can be released into the culture medium as GelMA suffers hydrolysis.¹⁴

Biomaterials prepared with ECM proteins, such as GelMA, have recently mimicked the natural regulatory function of the matrix in the immune system, favoring the longevity and functionality of implants. According to a recent study, digested soluble ECM fragments are capable of modulating cell behavior and exhibiting immunomodulatory function during tissue remodeling, thus playing a role in homeostatic control.⁸⁷ In the present study, the engineered DEX-loaded nanotube-modified hydrogel (G+5.0H-10D) demonstrated a positive effect on LPS-treated cells likely due to the release pattern mediated by the aforementioned hydrogel formulation, where higher DEX amounts were released initially, accompanied by a progressive reduction in DEX concentration over time (Figure 2B). Qi et al.⁸⁸ demonstrated that a two-stage release pattern of DEX obtained by gelatin nanoparticles was essential to promote TNFα downregulation on RAW 264.7 cells, along with osteogenesis induction on MC3T3-E1 cells, by upregulating the expression of Runt-related transcription factor 2 (RUNX2), collagen Type I (COL-1), osteocalcin (OCN), and ALP and favoring mineral deposition. From a clinical standpoint, others have found that DEX improves mineralized dentin bridge deposition in situations of pulp exposure in vivo with the maintenance of pulp vitality.^89-91

3.5. Biocompatibility and Bone Regeneration: in Vivo Models.

To evaluate the biocompatibility of DEX-loaded nanotube-modified GelMA, polyethylene (PE) tubes filled with GelMA (G), GelMA+5.0%HNT (G+5.0H), or GelMA+5.0% HNT-DEX10% (G+5.0H-D10) were subcutaneously implanted in rats (Figure 8A).

Figure 8.

(A) Engineered GelMA hydrogel formulations, biocompatibility in vivo. Detailed presentation of the surgical procedures involved in the implantation of the various hydrogel formulations in the dorsal subcutaneous pocket in rats. (B) HE-stained images of the tissue section of the implant area at 7 days postimplantation. Note the formation of a delicate, fibrous capsule (C) at the polyethylene tube opening with the subjacent connective tissue exhibiting small, congested blood vessels (yellow arrows), collagen fibrils (black arrows), areas of edema (stars), and fibroblasts distributed among the mononuclear inflammatory cells.

After 7 days, the host tissue response at the tubes’ opening was evaluated in comparison to SHAM (i.e., implantation of the unfilled PE tubes). Representative H&E staining images are shown in Figure 8B. A slight infiltration of the inflammatory cells, primarily situated around the edges of the PE tubes, was detected for G+5.0H and G+5.0H-D10 hydrogels. Of note, a thin fibrous capsule was observed between the hydrogel and the adjacent connective tissue, which exhibited several dilated and congested blood vessels and a few mononuclear inflammatory cells. Such as shown in the control group (PE), in G+5.0H, connective tissue with large areas of edema, scattered inflammatory cells, and blood vessels were observed adjacent to the thin, fibrous capsule formed in contact with the G+5.0H inserted into the tube. Also, an ingrowth of blood vessels was found at several areas around the edges of both the G+5.0H and G+5.0H-D10 groups, indicating material biocompatibility with angiogenesis. The outcomes of this study are in accordance with recent research supporting GelMA biocompatibility.^92,93

To determine in vivo the regenerative capacity of the DEX-loaded nanotube-modified hydrogel (i.e., G+5.0H-D10), the same GelMA-based formulations implanted subcutaneously were selected for implantation in critical-size calvarial defects in rats (Figure 9A).

Figure 9.

(A) Representative images of the clinical and surgical procedures of the critical-size calvarial defects in rats, followed by the injection of 150 μL of the GelMA-based formulations (i.e., GelMA [G], GelMA+5.0%HNT [G+5.0H], and GelMA+5.0%HNT-DEX10% [G+5.0H-D10]) into the defects and in situ photo-cross-linking for 60 s with an LED light-curing device. (B) Exemplary micro-CT images revealing bone formation 6-weeks postsurgery of the various GelMA-based formulations (scale bar = 1 mm). (C) Quantitative assessment of bone regeneration parameters for each treatment. Of note, the G+5.0H-D10 formulation outperformed the other groups in all evaluated parameters (*p < 0.05; mean ± SD, n = 4).

Micro-CT analysis shows a statistically significant increase in total bone volume for both nanotube-modified GelMA groups when compared to the SHAM and unmodified GelMA (G) controls (Figure 9B,C). However, the highest bone volume occurred in the GelMA hydrogel containing dexamethasone (G +5.0H-D10). When total bone volume was normalized by total volume, it was determined that G, G+5.0H, and G+5.0H-D10 presented values greater than the SHAM group with the latter featuring the highest volume. For histological analyses, H&E and Masson’s Trichrome (MT) staining were used. Fibrous connective tissue and a limited amount of new bone were observed in the negative control (SHAM) (Figures 9B and 10). The initial stage of new bone formation was detected at the edges of the defect in the unmodified GelMA (G) and nanotube-modified hydrogels; however, inflammatory cells and collagen fibers were more frequently observed in defects treated with G. Meanwhile, in defects treated with G+5.0H and G+5.0H-D10, new bone was well integrated from the original edges of the defect and small osseous islands were also noticed. Cidonio et al.⁵⁸ demonstrated that GelMA containing aluminosilicate clay nanoparticles enhanced the mineralization potential of bone marrow stem cells, and this effect was improved by DEX supplementation in the culture medium. The positive osteogenic effect of DEX-free nanotube-modified GelMA (G +5.0H) may be attributed to the increased hydrogel compressive modulus upon the HNTs’ incorporation (Figure 3C), thus agreeing with previous studies that assessed the role of hydrogel stiffness on cell behavior; while stiffer hydrogels increased cell spreading and osteogenic differentiation, softer hydrogels were prone to form cell aggregates that resulted in adipogenic differentiation.⁹⁴

Figure 10.

Histological assessment of the regenerative capacity of the GelMA-based hydrogel formulations. (A) HE-stained images and (B) MT-stained images show the bone regeneration response at the defect level in the different experimental groups after 6 weeks in vivo. The initial stage of the new bone formation at the edges of the defect in the G+5.0H-D10 group was considerably improved compared to the other groups.

In the present study, lithium acylphosphinate salt (LAP) was selected as the photoinitiator for GelMA, since it is capable of absorbing light in the visible region.⁹⁵ This biomaterial was injected into the bone defects and then photo-cross-linked with a LED curing device. The interesting data obtained using this in vivo protocol demonstrated the biocompatibility and possible clinical applicability of the proposed DEX-loaded nanotube-modified GelMA hydrogel. Moreover, mineralized tissue was present in the center of the bone defects filled with G+5%HNT and G+5%HNT-DEX (Figures 9B and 10), demonstrating that cells from the surrounding tissue were able to migrate to the material surface and infiltrate deeply into the GelMA porous architecture. This feature is essential for the success of tissue engineering strategies, since the biomaterial should act as a permeable template to provide a temporary ECM for neotissue formation.^96,97 Thus, the engineered drug delivery system composed by GelMA and HNTs loaded with 10% DEX seemed to be a suitable approach for mineralized tissue regeneration, as it was able to recruit adjacent cells and induced their differentiation and local maturation.

4. CONCLUSIONS

In this study, a cytocompatible, biodegradable, DEX-loaded nanotube-modified GelMA hydrogel was formulated for the sustained release of DEX for mineralized tissue regeneration under inflammation. When one takes into consideration that the DEX release led to a similar ALP expression and mineralized nodule formation in LPS-stimulated and nonstimulated SHEDs, it indicates that DEX exerted a protective role against inflammation. The proposed hydrogel induced a favorable tissue response marked by the small presence of inflammatory cells in subcutaneous implantation and demonstrated enhanced mineralized tissue regeneration in a critically sized defect model. Altogether, our findings show that the proposed injectable DEX-loaded nanotube-modified multifunctional hydrogel is considered a promising drug delivery system for potential applications in mineralized tissue regeneration under inflammation. Nevertheless, further investigations should focus on improving the anti-inflammatory potential of this system and on improving the clinical translation of this novel technology for clinical application. Further in vivo studies focusing on more specific application sites, such as pulp and periodontal tissues under inflammation, will improve our understanding of this system as a potent biomaterial to improve mineralized tissue regeneration in critical situations.

Supplementary Material

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsabm.1c00620.

Dose-response assay of LPS on SHEDs treatment; experimental methods to select LPS concentration; results of cell viability (%) and mineralized matrix deposition (%) to induce degenerative stimulus on SHEDs (PDF)