Gelatin methacryloyl hydrogel as an injectable scaffold with multi-therapeutic effects to promote antimicrobial disinfection and angiogenesis for regenerative endodontics

Abstract

Regenerative endodontics represents a paradigm shift in dental pulp therapy for necrotic young permanent teeth. However, there are still challenges associated with attaining maximum root canal disinfection while supporting angiogenesis and preserving resident stem cells viability and differentiation capacity. Here, we developed a hydrogel system by incorporating antibiotic-eluting fiber-based microparticles in gelatin methacryloyl (GelMA) hydrogel to gather antimicrobial and angiogenic properties while prompting minimum cell toxicity. Minocycline (MINO) or clindamycin (CLIN) was introduced into a polymer solution and electrospun into fibers, which were further cryomilled to attain MINO- or CLIN-eluting fibrous microparticles. To obtain hydrogels with multi-therapeutic effects, MINO- or CLIN-eluting microparticles were suspended in GelMA at distinct concentrations. The engineered hydrogels demonstrated antibiotic-dependent swelling and degradability while inhibiting bacterial growth with minimum toxicity in dental-derived stem cells. Notably, compared to MINO, CLIN hydrogels enhanced the formation of capillary-like networks of endothelial cells in vitro and the presence of widespread vascularization with functioning blood vessels in vivo. Our data shed new light onto the clinical potential of antibiotic-eluting gelatin methacryloyl hydrogel as an injectable scaffold with multi-therapeutic effects to promote antimicrobial disinfection and angiogenesis for regenerative endodontics.

---

1. Introduction

The loss of immature permanent teeth can have adverse effects in young patients, which involve poor jaw growth, malocclusion, and serious psychosocial harm.^1,2 The modern era of regenerative endodontics procedures (REPs) has emerged around the premises of promoting natural pulpal physiological functions of immature permanent teeth and favoring the termination of their development.^3,4 The procedure employs the classic tissue engineering triad, in which a fibrin clot obtained by lacerating periapical tissues (evoked bleeding) serves as a scaffold, while endogenous stem cells and growth factors migrate into the root canal space to support pulp–dentin complex regeneration.⁵ However, root canal disinfection is a major challenge in the success of REPs due to the presence of a polymicrobial infection that extends deep into the dentinal tubules and cannot be eliminated simply by debridement and necrotic tissue removal.⁶

Apical periodontitis caused by long-term root canal infection, as well as a root canal filled with necrotic debris and thick bacterial biofilm, has been documented in several studies. In addition, histologic examinations of extracted teeth have raised concerns about the efficacy of current disinfection strategies for REPs.^7–9 Importantly, it has been suggested that antibiotic concentrations used for disinfection are toxic to remaining viable cells in the canal and may also alter the bioavailability of endogenous growth factors, thus impairing tissue healing/regeneration.^10–14 Hence, there has been a significant effort dedicated to delivering low, yet effective concentrations of antibiotics for maximum disinfection with minimal detrimental effects on resident stem cells.^15–18

One of the most relevant reasons for regenerative procedure failure is lack of vascularization.¹⁹ Angiogenesis – the development of new blood vessels, is essential for supplying oxygen and nutrients, as well as clearing out dead cells and their toxic byproducts for optimal tissue regeneration by preventing fibrosis. Additionally, different cell types can communicate with one another by secreting signaling molecules and cytokines, ultimately leading to the regeneration of fully functional tissues.^4,20,21 However, the impact of antibiotics used for disinfection in REPs on angiogenesis remains to be elucidated. For example, minocycline (MINO), one of the components in the triple antibiotic paste (TAP), has long been recognized for its anti-angiogenic function.^22,23 In fact, our group previously reported the use of clindamycin (CLIN) as a replacement for MINO due to their broad-spectrum activity against endodontic bacteria in biodegradable polymeric nanofibers to support the localized release of biocompatible, yet effective antibiotic doses to treat endodontic infection and support angiogenesis.^12,24 Nonetheless, given the intricate anatomy of root canals, it is critical to design injectable drug delivery systems capable of easily flowing into these highly complex canals while supporting angiogenesis with highest disinfection ability to maximize the regenerative outcome.^16,17,25

Hydrogels and electrospun nanofibers have both been widely used in drug delivery and tissue engineering.^26,27 However, it is difficult to avoid burst drug release from monolithic matrices (hydrogel or nanofibers), and they only partially meet the mechanical and biological characteristics of an ideal extracellular matrix for tissue engineering.^26,28 In regenerative endodontics, the integration of hydrogels with antibiotic-eluting nanofibers can be used as temporary frameworks to ablate infection and support neo-tissue formation via migration of host cells from the root apex for complete and effective regeneration of the pulp–dentin complex.^29,30 Gelatin methacryloyl (GelMA), a photopolymerizable hydrogel, is frequently used for biomedical applications due to its high biocompatibility and biodegradability. It consists of modified amine-containing side groups of gelatin with methacrylamide and methacrylate groups. GelMA possesses various characteristics that make it ideal for biological interactions, including hydrophilicity, integrin-binding motifs, and matrix metalloproteinase (MMP) degradation sites.¹⁶ Like other hydrogels, GelMA can be easily customized with a range of additives, such as antibiotics, nanotubes, nanofibers, and biomolecules, among others.³¹ Meanwhile, poly(lactic-co-glycolic acid) (PLGA) is a well-characterized FDA-approved biocompatible and biodegradable polymer commonly utilized in the production of nanofibers for tissue regeneration and drug delivery.³²

Here, we developed a hydrogel system by incorporating antibiotic-eluting fiber-based microparticles in gelatin methacryloyl (GelMA) hydrogel to gather antimicrobial and angiogenic properties while prompting minimum cell toxicity. The engineered hydrogels (i.e., MINO- or CLIN-eluting microparticles modified GelMA) demonstrated antibiotic-dependent swelling and degradability while inhibiting bacterial growth with minimum toxicity in dental-derived stem cells. Our data shed new light onto the clinical potential of CLIN-eluting microparticles modified GelMA as an injectable antimicrobial scaffold with the ability to enhance the formation of capillary-like networks of endothelial cells in vitro and the presence of widespread vascularization with functioning blood vessels in vivo for applications in regenerative endodontics.

2. Materials and methods

2.1 Gelatin methacryloyl (GelMA) synthesis

GelMA synthesis was performed by the freeze-drying method, as previously reported.¹⁶ Briefly, 10 g of type A gelatin from porcine skin (300 bloom, Sigma-Aldrich, St. Louis, MO, USA) was dissolved in 100 mL of Dulbecco's Phosphate Buffered Saline (DPBS, Sigma-Aldrich) at 50 °C and mixed for 1 h. Then, 8% v/v methacrylic anhydride (MA) was added dropwise in the gelatin solution and stirred for 2 h. After the reaction was complete, the methacrylation was stopped by adding DPBS, and the solution was dialyzed against deionized water for 7 d at 45 °C using dialysis tubing (12–14 kDa) to remove unreacted MA and impurities. The solution was diluted with ultrapure water, filtered, and lyophilized (Labconco FreeZone 2.5 L, Labconco Corporation, Kansas City, MO, USA) to obtain porous white foam. The 10% (w/v) GelMA hydrogel was prepared by dissolving porous foam into PBS solution containing 0.5% (w/v) lithium phenyl-2,4,6-trimethylbenzoylphosphinate (LAP) as a photoinitiator at 50 °C.

2.2 Electrospinning of antibiotic-eluting fibers and chemo-morphological characterization

For electrospinning, polymeric solutions were prepared by dissolving 75/25 poly(lactic-co-glycolic acid) (PLGA) (inherent viscosity 0.55–0.75 dL g⁻¹, Lactel Biodegradable Polymers, Birmingham, AL) at 18% (w/v) in chloroform (Sigma-Aldrich). Solutions containing 15% minocycline (MINO, Sigma-Aldrich) and clindamycin (CLIN, TCI America Inc., Portland, OR, USA) were prepared according to the weight of PLGA and were added directly into the polymer solution overnight at room temperature (RT). A plastic syringe with a 27-gauge metallic needle containing 5 mL of pure PLGA and antibiotic-containing solutions was attached to an automatic syringe pump (KDS 200 Legato, Harvard Apparatus, Holliston, MA, USA) located at a 20 cm distance from the rotating mandrel. Electrospinning was carried out at 18 kV and the polymer solution was injected at 2 mL h⁻¹ and 1 mL h⁻¹ for MINO and CLIN, respectively. The electrospun mats were then placed in a vacuum desiccator for at least 48 h to remove any remaining solvent.

The morphology of pure and antibiotic-eluting fibers was studied by scanning electron microscopy (SEM, JSM-6390, JEOL, Tokyo, Japan). The average fiber diameter was calculated from approximately 100 random measurements from 4 distinct SEM micrographs using ImageJ analysis software (National Institutes of Health, Bethesda, MD, USA). The functional groups of all synthesized fibrous mats were examined using Fourier transform infrared spectroscopy in the attenuated total reflection mode (ATR/FTIR-4100, Thermo Fisher Scientific, Inc., Waltham, MA, USA) by collecting 64 scans with a resolution of 2 cm⁻¹.

2.3 Fabrication of antibiotic-eluting electrospun fibers microparticles (EM)

In this investigation, electrospun fiber microparticles were obtained by cryomilling to resolve the challenge of achieving uniform dispersion in the hydrogel. For this, we modified a previously published protocol by our research group.³³ The electrospun fibers were completely soaked in 6 mL of 10% GelMA, crosslinked for 60 s with a polywave LED curing light (Bluephase, Ivoclar, Amherst, NY, USA), and allowed to dry overnight in a fume hood. The 150 mg GelMA/electrospun fiber composite mixture was cryomilled according to our previous protocol.³³ Briefly, the blend was pre-cooled by liquid nitrogen for 2 min in the milling vial, followed by 15 min of milling, consisting of an alternating cycle of 1 min milling, separated by cooling intervals of 1 min, respectively. The electrospun fibrous microparticles (EM) collected were sieved (45 μm) to standardize the particle size and stored in a silica-filled desiccator at RT until further use.

2.4 Synthesis and characterization of EM-loaded hydrogel

The processed microparticles (45 μm) with and without antibiotics at two different concentrations (1% and 2.5% w/v) were added into 15% GelMA solution with 0.5% (w/v) photoinitiator (LAP). Five groups were formulated, namely Antibiotic-free microparticles modified GelMA (control), 1% and 2.5% CLIN-eluting microparticles modified GelMA, as well as 1% and 2.5% MINO-eluting microparticles modified GelMA. The aforementioned mixtures were individually poured into custom-made silicone molds (∅ = 6 mm, h = 2 mm) and then exposed to visible light (385–515 nm) for 30 s (Bluephase, Ivoclar). The resultant hydrogels were taken out of the molds and characterized via SEM and FTIR as mentioned above (Fig. 1).

Fig. 1 Schematic illustration of the overall synthesis step to obtain the multi-therapeutic injectable gelatin methacryloyl (GelMA) hydrogel.

2.4.1 Swelling and biodegradation. To determine the water retention capability of the engineered hydrogels, cylindrical specimens (n = 4/group) were incubated in DPBS at 37 °C for 24 h. The wet and dry weight of the specimens was determined before and after lyophilization, respectively. The changes in hydrogel weight between the wet (W_w) and dry (W_d) states were used to determine the volumetric swelling through the following equation:

Swelling ratio = (W_w − W_d)/W_d × 100

The biodegradation behavior of GelMA and GelMA modified with microparticles (n = 4) was followed for 21 d. Cylindrical specimens (∅ = 6 mm, h = 2 mm) were weighed (W₀) and incubated in 5 mL DPBS containing collagenase type A (1 U mL⁻¹) at 37 °C, with a fresh solution replaced every 3 d to maintain constant enzyme activity. The weight (W_t) of the specimens was measured using an analytical balance at predetermined time points. Using the equation below, degradation was obtained as a change in sample weight:

Degradation = (W_t/W₀) × 100

2.5 Drug release

The drug release of MINO- and CLIN-microparticles encapsulated in GelMA was measured using liquid chromatography-tandem mass spectrometry (LC-MS/MS) and ultra-high performance liquid chromatography-tandem mass spectrometry (UPLC–MS/MS), respectively, due to differences in detecting the presence of antibiotics in buffer solution. The UPLC system consisted of an Acquity Quaternary Solvent Manager, Sample Manager-FTN, and Column Manager, and the TUV Detector (Waters Corporation, Milford, MA, USA) was used to perform the analyses of minocycline. The separation of minocycline was carried out with an Acquity UPLC BEH C18 column (1.7 μm, 2.1 × 100 mm, Waters) at 40 °C, and the mobile phase was acetonitrile/potassium phosphate buffer (pH 2.5)/(25/75) at 0.3 mL min⁻¹ with an injection volume of 10 μL. The concentration of minocycline was detected by UV absorbance at 350 nm of wavelength. Liquid chromatography analysis was performed using the same system for clindamycin at 210 nm of wavelength. The system was controlled by the Analyst v.1.4.2 software from Applied Biosystems (Foster City, CA, USA) and the Acquity Console to control the UPLC.

2.6 Evaluation of antimicrobial action

The antimicrobial activity of drug-eluting microparticles modified GelMA was evaluated by means of agar diffusion and antibiofilm assays. Gram positive Actinomyces naeslundii (A. naeslundii, ATCC 12104), Enterococcus faecalis (E. faecalis, ATCC 19433) and Gram-negative Fusobacterium nucleatum (F. nucleatum, ATCC 25586), pathogens were used in agar diffusion assays. The strains were cultured overnight in Brain and Heart Infusion broth (BHI) at a cell density of 3 × 10⁸ CFU mL⁻¹. The bacterial suspension (100 μL) was swabbed uniformly over BHI agar plates to form a bacterial lawn. The aliquots (10 μL) obtained at predetermined points from hydrogels incubated in sterile PBS were inoculated on agar plates. A 2% chlorhexidine digluconate (CHX; positive control) and PBS (negative control) were similarly placed on the agar plates. The plates were incubated at 37 °C for 48 h, and the diameters of the clear zones of growth inhibition were measured.

Infected dentin biofilm model assay was carried out, as previously described.¹⁶ The use of dentin slices was approved by the Institutional Review Committee (protocol #1407656657). Dentin slices were infected with A. naeslundii at a density of 7.5 × 10⁷ CFU mL⁻¹ in BHI broth and cultured in an anaerobic chamber for 7 d to allow for biofilm development, with the media replaced every 2 d. Infected dentin slices (n = 6/group) were randomly divided into 4 groups: drug-free hydrogel (GelMA), 2.5% CLIN-eluting microparticles modified GelMA (2.5% CLIN), 2.5% MINO-eluting microparticles modified GelMA (MINO) and untreated biofilm (control). At day 7, dentin slices were gently washed with PBS to remove the non-adherent biofilm, followed by placement of 50 μL of each material crosslinked for 15 s on biofilm. The samples were incubated in an anaerobic chamber for 3 d and examined for CFU mL⁻¹ (n = 4), SEM (n = 3), and CLSM (n = 3).

For the CFU mL⁻¹, the dentin slices were rinsed with PBS to remove non-adherent bacteria. Then the disks were placed in centrifuge tubes containing 10 mL of saline and ultrasonicated to remove biofilm. A 10-fold serial dilution was prepared and poured on BHI agar plates. The plates were incubated in an anaerobic chamber for 24 h at 37 °C, and colonies were counted. For the SEM evaluation, specimens were rinsed with PBS before being fixed in 4% paraformaldehyde. After 24 h, dehydration was performed sequentially using different concentrations of ethanol. Before imaging, the dentin specimens were hexamethyldisilazane-dried overnight and sputter-coated with Au–Pd. Confocal microscopy was used to image the live and dead bacteria in biofilms. After 48 h of incubation, biofilms stained with SYTO 9 (excitation/emission 485/498 nm) and propidium iodide (excitation/emission 535/617 nm) were observed under the Leica SP2 CL5Mt (Leica Microsystems GmbH, Wetzlar, Germany). The experiment was repeated independently three times and the results were compared with the control groups.

2.7 Cell culture and cytotoxicity

To confirm the cytocompatible character of the engineered hydrogels with and without antibiotics, MTS assay was carried out in accordance with the guidelines of the International Standards Organization (ISO10993-5: cytotoxicity tests—in vitro methods). The specimens (∅ = 6 mm, h = 2 mm) were incubated in 5 mL of α-MEM media at 37 °C for 21 d. At predetermined time points, 500 μL of aliquots were collected from each vial and replaced with an equal amount of fresh media (α-MEM) to keep the extraction volume constant. Prior to cell exposure, collected aliquots were filtered through a 0.22 μm membrane. Stem cells from human exfoliated deciduous teeth (SHEDs) were seeded at a density of 2.5 × 10³ cells per well and allowed to adhere to the wells of 96-well plates. The collected extract (100 μL) with and without antibiotics was replaced after 24 h. Then, 20 μL of MTS solution (CellTiter 96 AQueous One Solution Reagent) was added, cells were incubated for 2 h, and absorbance at 490 nm (SpectraMax iD3, Molecular Devices, LLC) was measured against a blank column. As a positive control, SHEDs cultured in complete media without hydrogels were used.

2.8 Angiogenesis assay–capillary like tube formation

Human umbilical vein endothelial cells (HUVECs) were seeded in a Matrigel precoated 24-well plate at a density of 2 × 10⁵ cells per well in ECGM supplemented with or without 50 ng mL⁻¹ of vascular endothelial growth factor (VEGF) with a transwell insert containing drug-free GelMA (control), and drug-eluting GelMA hydrogels samples (∅ = 6 mm, h = 2 mm). After 24 h of incubation, the insert was removed, the culture was washed with PBS (×2) and stained with calcein (5 μg mL⁻¹), and tube formation was examined using light and fluorescence microscopy (BZ-X710, Keyence Corporation of America, Itasca, IL, USA). WimTube software (Onimagin Technologies SCA, Córdoba, Spain) was used to quantify the capillary-like tube structure.

2.9 In vivo studies

Both handling of the animals and surgical procedures were performed in accordance with the Animal Care and Use Committee approved by the Ethics Committee on Animal Experimentation at the University of Michigan (IACUC, protocol #PRO00008502). Engineered hydrogels with and without microparticles with antibiotics (∅ = 6 mm, h = 2 mm) were subcutaneously implanted into both sides of the dorsal of 6 week old Fischer 344 male rats (Envigo RMS, Inc., Oxford, MI, USA) weighing around 300–320 grams. Briefly, the rats were induced with isoflurane 4% (Piramal Critical Care Inc., Bethlehem, PA, USA) and maintenance of 2% buprenorphine (0.6 mg kg⁻¹) was injected subcutaneously. The surgical site was cleaned with povidone-iodine solution (Sigma-Aldrich). Four samples were implanted in the dorsum of the rats (four samples per group/animal).

2.9.1 Histological analysis. At days 3 and 7 post-implantation, specimens (hydrogel/surrounding tissues) were collected from the animals, following euthanasia. The specimens were fixed with 4% paraformaldehyde before being dehydrated in a series of increasing ethanol concentrations until pure ethanol was reached, embedded in paraffin, and sectioned at 5 μm-thick sections and stained with hematoxylin and eosin (H&E), according to standard procedure. The stained slides were observed under a light microscope with a digital camera (Nikon E800, Nikon Corporation, Tokyo, Japan).

2.9.2 Immunohistochemistry. For immunohistochemistry, after deparaffinization and rehydration, the sections were quenched for 5 min with 3% hydrogen peroxide and blocked for 2 h with 5% bovine serum albumin to decrease non-specific interaction. Subsequently, the tissue section was stained with CD-31 antibody (1 : 400, Abcam) at 4 °C overnight. All sections were rinsed with PBS and incubated with biotinylated goat anti-rabbit secondary antibody for 2 h. Following this step, the sections were incubated with diaminobenzene (DAB) (K3468; Dako, Agilent Technologies, Inc., Santa Clara, CA, USA). Hematoxylin was used as a nuclear counterstain and sections were dehydrated with gradient ethanol, soaked with xylene, and mounted using an aqueous media. Five randomly selected fields of view were examined for each group at each time point using a digital camera equipped with a light microscope (Nikon E800, Nikon Corporation).

2.10 Statistics

The statistical software SPSS for Windows Version 25 (IBM, Armonk, NY, USA) uses One-way analyses of variance (ANOVAs) with post-hoc tests of Tukey's to compare more than two groups. Means and standard deviations were calculated from numerical data.

3. Results and discussion

3.1 Fabrication/characterization of the engineered hydrogels

Hybrid hydrogels, which combine electrospun fibers within GelMA, were successfully developed using the proposed cryomilling process followed by microparticles dispersion as described here. First, we fabricated antibiotic-eluting electrospun fibers, and the fiber morphology was consistent with previous studies.³⁴ Representative SEM images of the processed PLGA and antibiotic-containing PLGA nano/microfibers and average fiber diameter are shown in Fig. 2(A), with the average fiber diameter decreasing with the presence of CLIN, when compared with PLGA and MINO. This decreased fiber diameter for CLIN, compared to MINO and PLGA, can be attributed to higher conductivity of the solution, because of the hydrochloride form of clindamycin.³⁵

Fig. 2 (A) SEM micrograph of electrospun PLGA fibers with MINO and CLIN, as well as microparticle of electrospun fibers in GelMA hydrogel. (B) Schematic and ¹H NMR of methacrylate substitution confirmed by presence of methacrylate vinyl group signal at δ = 5.2 ppm and δ = 5.6 ppm ad methyl group signal at δ = 1.8 ppm. (C). FTIR spectra of pure PLGA, pure antibiotics, GelMA, and cryomilled electrospun fibers with antibiotics.

Gelatin methacryloyl is a versatile hydrogel with RGD and matrix metalloproteinase (MMP)-degradable motifs, resulting in superior biocompatibility and enzymatic degradability for drug delivery and tissue engineering.³¹ It has been modified with diverse types of additives, such as nanoparticles, biomolecules, stem cells, and nano/microfibers.^16,36,37 As a result, for regenerative endodontics, a hybrid hydrogel that can support cell attachment, proliferation, and antibiotic release in a sustained manner, would be a promising solution. The ¹H NMR (Fig. 2(B)) confirmed inclusion of double bonds of acrylamide at 5.3 and 5.6 ppm, and peaks between 1.5 and 2 ppm were due to the methyl function from methacrylate, confirming the successful gelatin methylacrylation. The homogenous distribution of antibiotic-eluting microfibers in the hydrogel is essential. In this study, electrospun fibers were mixed with GelMA to increase its hydrophilicity, then cryomilled to obtain microparticles that could be uniformly dispersed within the hydrogel. Cryomilling is a solvent-free process that enables the milling of polymers below their glass transition temperature, making it easier for fracture to occur.^33,38 The obtained fine powder was easy to disperse into the hydrogel due to the presence of GelMA on the surface of the cryomilled microparticles. SEM images of cryomilled electrospun nanofibrous structures uniformly distributed in GelMA are shown in Fig. 2(C). Due to electrostatic and van der Waals forces between the particles, the cryomilled powders showed signs of agglomeration.³⁸ FTIR analysis of the PLGA and antibiotic-eluting fibrous microparticles are given in Fig. 2(D). Pure MINO and CLIN showed multiple absorption peaks in the range of 600–1650 cm⁻¹. The characteristic peaks for MINO and CLIN could be observed in the FTIR spectrum of cryomilled microparticles.

The swelling ratio and degradation rate of the engineered hydrogels are critical parameters to consider when evaluating the potential application for drug delivery and tissue engineering.^39,40Fig. 3(A) illustrates the mass swelling ratio of the of GelMA and antibiotic-eluting microparticles modified GelMA after 24 h. The presence of the hydrophilic functional group plays an important role in water uptake and swelling behavior.⁴¹ The results indicated that the addition of microparticles affected the swelling properties of the hydrogel. The swelling ratio increased significantly when 1% CLIN- and 2.5% MINO-eluting microparticles were added to GelMA. The developed hydrogel had a higher swelling ratio, as the number of hydroxyl groups was higher in the microparticles-modified hydrogels, which allowed more water to be absorbed inside the matrix. Notably, no significant differences were found between 2.5% CLIN and 1% MINO compared to pure GelMA. All hydrogels were biodegradable (Fig. 3(B)), while GelMA was not fully degraded after 14 d; GelMA loaded with antibiotic-eluting microparticles degraded at different times, depending on the antibiotics and their concentration. It must be noted that comparatively less of the carboxyl group was available to crosslink with the hydroxyl group for formation of a highly crosslinking network for GelMA loaded with microparticles. As a result, high water inflow diluted the hydrogen bonding network and washed out the uncrosslinked hydrogel from the electrospun microparticles modified GelMA hydrogel.

Fig. 3 (A) Swelling ratio of GelMA hydrogel-laden with different concentrations of antibiotics encapsulated in electrospun microparticles with 1% CLIN and 2.5% MINO showed a significant increase in the swelling ratio. (B) Degradation profile showing all the hydrogels are biodegradable; however, GelMA was not fully degraded after 336 h. (C) and (D) Release profile of CLIN and MINO at different concentrations (1% and 2.5%) from cryomilled electrospun microparticle–hydrogel matrix as a function of time.

3.2 Drug release

The release of drugs from polymer-based drug delivery systems is significantly influenced by factors such as the drug's aqueous solubility, as well as the diffusion and erosion of the polymeric network.⁴² The data in Fig. 3(C) and (D) show the difference in release profiles of cryomilled microparticles containing MINO and CLIN in GelMA. CLIN is more widely available than MINO, which could be due to differences in the drug's chemical structure, properties, and fiber diameter.⁴³ After 24 h, the CLIN group showed burst release; whereas, the MINO group showed slower release, which could be due to differences in the antibiotics’ interactions with the polymers. Furthermore, differences in the solubility of MINO and CLIN in polymer solution and aqueous media will affect drug release.⁴⁴ All of these factors are very likely to influence the location and availability of the drug within the fiber, as well as its release rate.

The Korsmeyer–Peppas (KP) model has been widely used to describe drug release kinetics from hydrogels encapsulated with drug-loaded microparticles.⁴⁵ In this study, KP model is likely to the describe release behavior as it accounts for both diffusion and polymer erosion, which are two key factors that play a significant role in drug release from polymeric drug delivery systems.⁴⁶ However, it is important to note that different drug delivery systems may require different mathematical models to accurately describe their release behavior. In some cases, other models such as zero-order or first-order kinetics may be more appropriate, depending on the specific properties of the drug and polymer used. Future work will investigate controlled release of drugs on polymeric matrix based on their properties (solubility, molecular weight, size, chemical structure, binding affinity, and concentration) to understand underlying mechanisms and improve drug delivery system design.

3.3 Antimicrobial activity

Antimicrobial efficacy against Actinomyces naeslundii, Fusobacterium nucleatum, and Enterococcus faecalis was determined by the presence or absence of zones of inhibition (Fig. 4(A)–(C)) of antibiotic-eluting microparticles modified hydrogel over time (aliquot collection). These bacteria were used due to their role in root canal infection; in particular, An and Fn, common microorganisms found in traumatized permanent immature teeth and during interappointment endodontic flare-ups, respectively.^47,48 Herein, the Ef showed complete resistance to CLIN-containing hydrogel, which agrees with previously published reports.^49,50 In contrast, An was susceptible to both CLIN and MINO-containing hydrogel. Interestingly, 14-d aliquot with CLIN showed a significantly higher inhibition zone compared to a respective concentration of MINO (Fig. 4(A)–(C)). Pure (drug free) GelMA and the negative control group (bacterial growth) had no effect on biofilm formation or cell viability, as evidenced by the high colony-forming unit (log¹⁰ CFU mL⁻¹) values. Meanwhile, all antibiotic-modified hydrogels demonstrated antibiofilm properties, resulting in lower CFU counts than the controls (Fig. 4(D)). As the antibiotic loading increased the mean inhibition zone, however, bacteria can still survive the intracanal drug due to their deep involvement in the dentinal tubules and canal irregularities.⁵¹ Therefore, from a clinical standpoint, the dentin slice model⁵² was used to evaluate the effect of injectable hydrogel on A. naeslundii biofilm and visualized using SEM and CLSM (Fig. 5). SEM imaging demonstrated the penetration of biofilm into the dentinal tubules of the tooth slices’ dentin specimens for the antibiotic-free group (pure GelMA). Both CLIN- and MINO-treated dentin specimens showed a bacteria-free dentin surface and tubules. CLSM showed the presence of penetration and live (green) bacteria on the dentin slice; however, a high proportion of dead bacterial cells were seen for the groups treated with CLIN and MINO-loaded GelMA. SEM and CLSM imaging of the dentin slice confirm the absence of bacteria in the dentinal tubes, suggesting that suitability of our injectable drug delivery system could help inhibit the colonization of biofilm on dentin. In sum, while the use of single-species bacteria has provided valuable insights into the activity of the developed hydrogel system against root canal infections, they may not accurately reflect the diverse microbial community that can play a significant role in inducing root canal infections. It is important to recognize the limitations of this approach and develop more comprehensive models that better represent the complex nature of biofilm-associated infections.

Fig. 4 Antimicrobial effect of hybrid hydrogels: the diameter of mean inhibition zones (mm) from the agar diffusion assays against 3 bacteria at days 1, 3, 7, 14, and 21. (A) Ef (B) Fn, and (C) An. Chlorhexidine (CHX) served as the positive control. (D) CFU counting was performed for A. naeslundii cells grown on dentin slices treated with hybrid hydrogels, having a negative control group consisting of untreated bacterial growth. (*Statistically significant compared to bacterial growth, p < 0.05.).

Fig. 5 Biofilm formation on dentin slice: representative SEM and CLSM images showing a significant reduction in the An biofilm formation on the surface and in the dentinal tubules. SEM shows bacterial cells on the surface of dentin slices and forming dense extracellular matrix covering the entire surface (red arrows indicate the presence of bacteria on the surface and inside the canals). A greater quantity of bacteria and a predominance of live (green) bacteria can be observed in GelMA via CLSM.

3.4 Cytotoxicity and angiogenesis

Impact of the engineered hydrogels on cell viability was assessed on SHEDs grown in a monolayer and exposed to different antibiotic concentrations for varying time durations for their potential use in REPs. MTS assay was performed following the ISO 10993-5 standards (Biological evaluation of medical devices) that require at least 70% cell viability in comparison to the non-toxic control.¹⁸ The results confirmed that the formulated hydrogels did not induce significant cytotoxicity, as cell viability was found to be more that 70% after 28 d of culture (Fig. 6). Both CLIN- and MINO-eluting microparticles modified GelMA, at all concentrations, increased the SHED's viability in shorter periods (1–7 d), compared to pure GelMA. After 14 d, SHED's viability was significantly reduced by MINO; however, 1% CLIN maintained positive effects on cell viability over 28 d, compared to pure GelMA. Cell viability was higher in this study compared to previous studies by our groups and others with CLIN, metronidazole, ciprofloxacin, or TAP nanofibers.^15,24,53 This difference may be due to the slow and sustained release of antibiotics and degradation products of polymer from the synthesized antibiotic-eluting microparticles.

Fig. 6 Cell viability. Effects of different concentrations of antibiotics released from the hydrogel on SHED's viability measured indirectly using MTS assay. The percentage of cell viability was normalized by the mean absorbance of SHEDs cultured in a tissue culture plate on day 1. Distinct letters indicate statistically significant differences between the groups when compared with the control (SHED cells). The results are presented as mean ± SD (n = 5).

Prevascularization of the root canal with a ghost capillary network can provide an optimal environment for successful pulp regeneration strategies.²¹ Thus, to understand the role of antibiotics to support angiogenesis, we carried out capillary-like tube formation assays using HUVEC cells, a well-established in vitro angiogenesis assay.⁵⁴ Notably, the release of CLIN from the hydrogel had no adverse effects on the formation of tubular networks of endothelial cells (Fig. 7(A)) in the presence of vascular endothelial growth factors (VEGF) indicative of angiogenesis stimulation. The angiogenic response was assessed by quantification (Fig. 7(B)) of the capillary-like tube network, which demonstrated no statistical difference in total tube length, numbers, branching points, and loop for the CLIN-eluting microparticles modified GelMA hydrogels compared to GelMA. However, hydrogels with MINO in the presence of VEGF showed marked inhibition of endothelial cell tubular networks and was not significantly different compared to the non-VEGF group. This inhibition is unlikely to be caused by drug toxicity, as MTS findings revealed that cell viability did not vary between 2.5% of MINO and 2.5% of CLIN after 24 h. One potential explanation is thought to be that MINO blocked the VEGF from attaching to the receptors on the cells, thus reducing angiogenesis. Based on cell viability, qualitative, and quantitative analysis of capillary-like tube formation assay, the engineered CLIN-eluting microparticles modified GelMA may support cell migration and cellular crosstalk, contributing to the establishment of an optimal microenvironment for dental pulp regeneration.

Fig. 7 Angiogenesis assay. (A) Representative light and fluorescence microscopy images of HUVECs grown on Matrigel treated with hybrid hydrogels showing formation of capillary-like structure after 24 h. The hybrid hydrogels with MINO shows lack of tube formation, even in presence of VEGF. (B) Quantitative analysis of morphologic features of a capillary-like network structure after treating HUVECs with different hybrid hydrogels using WimTube™ software. (*Statistically significant compared to GelMA, p < 0.05.)

3.5 In vivo biocompatibility, degradation, and angiogenesis

Subcutaneous implantation under the dorsal skin of rodents is a well-established method to evaluate biocompatibility of hydrogels.⁵⁵ Histological analysis of the explanted hydrogels and surrounding tissues was done to assess the degradation and immune response at 3 and 7 days post-implantation. Hematoxylin and eosin staining (Fig. 8(A)) showed a low-to-moderate level of inflammatory infiltrate, predominantly mononuclear with newly formed unorganized collagen fibrils and fibroblasts associated to GelMA and CLIN groups. There was moderate-to-intense inflammation related to the implantation of MINO-eluting microparticle modified GelMA, with the presence of macrophages, polymorphonuclear, and mononuclear cells. The presence of endothelial cells (blood vessels) was found to be significantly higher in CLIN than in the other groups, demonstrating its superior biocompatibility. Furthermore, there was no evidence of necrosis in the samples, indicating that the hydrogels were accepted by the host. Despite significant degradation (Fig. 8(B)), all hydrogels were identifiable after 7 d. This finding is consistent with previous research, which found GelMA hydrogels containing additives do not degrade completely after 1 week.^56,57 Moreover, as seen in vitro, an accelerated degradation rate hydrogels with MINO was observed in vivo. A previous study has evaluated the use of chitosan hydrogel encapsulating hydroxyapatite particles for regenerative endodontics in immature dog teeth. However, in that study, a significant portion of the canal space remained occupied by the scaffold even after regeneration.⁵⁸ It is important for the degradation rate of the scaffold to match the regeneration rate of the host tissue, which can be achieved by adjusting the ratio of microparticles and hydrogel and the concentration of GelMA.^16,31 Additionally, the presence of endodontic infection amplifies the levels of matrix metalloproteinase (MMP), which can enable the degradation of GelMA to support the remodeling of new tissue. Therefore, the degradation of GelMA with electrospun microparticles can be tuned to be compatible with the rate of new tissue formation.

Fig. 8 Histological analysis of hybrid hydrogels degradation and tissue infiltration in vivo. (A) Representative HE-stained images of hybrid hydrogels and surrounding tissues after days 3 and 7 post subcutaneous implantation. The staining shows a high number of mononuclear cell infiltrates after hybrid hydrogels implantation and blood vessels (black arrow) with murine erythrocytes. Dash lines represent the boundary between the hydrogel and tissue. (B) H and E show faster degradation of unmodified GelMA when compared to hybrid hydrogels. Hybrid hydrogels are demarcated by a dotted line.

Angiogenesis, or the formation of new blood vessels, is a crucial process in tissue regeneration because it provides oxygen and nutrients to support the growth and function of new tissue. Therefore, a supportive angiogenic environment after root canal disinfection is key to ensure successful tissue repair and regeneration.⁵⁹ Subcutaneous implantation of hydrogel is a common method to assess angiogenic efficacy during the initial stage of scaffold or drug development, since it has the lowest impact on animal welfare, is simple to perform, and is reproducible.⁶⁰ Inspection of the implantation (Fig. 9(A)) site during euthanasia revealed ingrowth of host vessels with a tube-like structure exhibiting a vascularized phenotype in the group with CLIN-containing hydrogel. The GelMA group showed no sign of inflammation or necrosis in the subcutaneous tissue, but MINO showed signs of inflammation and induration (or palpable bump-signs of tissue edema). We next examined new vessel growth (vascularization) at days 3 and 7 by immunohistochemistry with antibodies against the endothelial marker CD31 (also called PECAM-1), a marker for endothelial cells commonly used to indicate wound bed blood vessels.⁶¹ There were significantly more CD31-positive cells in CLIN and GelMA, with increased length and thickness, compared to sham and MINO (Fig. 9(B)). Despite our promising results, additional investigation, including but not limited to orthotopic model and long-term in vivo studies, are essential to understanding the overall ability of the developed hydrogel to provide a supportive angiogenic microenvironment for dental pulp tissue repair when compared to conventional disinfection methods. Nevertheless, our findings support the notion that CLIN can support and enhance the local recruitment of vasculature during REPs, which is critical for their integration and long-term survival and function.

Fig. 9 Angiogenesis at the biomaterial–tissue-interface. (A) The macroscopic image of implanted hybrid hydrogels illustrates the apparent vascularization (black arrow) for hydrogels containing CLIN compared to GelMA only and MINO. (B) The density of blood vessels (red arrows) was markedly greater in CLIN compared to GelMA, alone, and MINO, as indicated by the IHC staining of anti-CD31 staining.

4. Conclusions

In the present study, novel hybrid hydrogels modified with antibiotic-eluting microparticles were successfully developed by embedding cryomilled electrospun fibers into GelMA hydrogel. The results indicated that the hybrid hydrogels allowed for sustained release of antibiotics, while also being cell-friendly and antimicrobial. Importantly, hybrid hydrogels with CLIN-eluting microparticles supported the stimulation of endothelial cell functions, such as migration and tubular formation in vitro and increase in microvessel density in vivo, indicating the potential ability of CLIN to support angiogenesis. In sum, CLIN-eluting microparticles modified GelMA is a biocompatible scaffold suitable to promote antimicrobial root canal disinfection and support the recruitment of host cells and angiogenesis to overcome the major challenge of REPs for functional dental pulp tissue engineering. The findings of this study may inspire the development of other multi-therapeutic hydrogels with a range of properties for various biomedical applications.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

M. C. B. acknowledges the National Institutes of Health (NIH, National Institute of Dental and Craniofacial Research/NIDCR) (grants R01DE026578 and R01DE031476) and J. E. N. acknowledges the NIH/NIDCR Grant R01DE021410. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. J. S. R. was supported in part by a scholarship from the CAPES Foundation (Brazil).

References

1. O. Schierz, K. Baba and K. Fueki, J. Oral Rehabil., 2021, 48, 256–270.
2. P. Kaur, S. Singh, A. Mathur, D. K. Makkar, V. P. Aggarwal, M. Batra, A. Sharma and N. Goyal, J. Clin. Diagn. Res., 2017, 11, ZC05.
3. S. Kim, M. Malek, A. Sigurdsson, L. Lin and B. Kahler, Int. Endod. J., 2018, 51, 1367–1388.
4. V. Rosa, T. M. Botero and J. E. Nör, Int. Dent. J., 2011, 61, 23–28.
5. F. Banchs and M. Trope, J. Endod., 2004, 30, 196–200.
6. A. A. Endodontics, 2018.
7. J. Silujjai and P. Linsuwanont, J. Endod., 2017, 43, 238–245.
8. L. M. Lin, E. Shimizu, J. L. Gibbs, S. Loghin and D. Ricucci, J. Endod., 2014, 40, 291–295.
9. R. Žižka, T. Buchta, I. Voborná, L. Harvan and J. Šedý, J. Endod., 2016, 42, 724–729.
10. M. Conde, L. Chisini, R. Sarkis-Onofre, H. Schuch, J. Nör and F. Demarco, Int. Endod. J., 2017, 50, 860–874.
11. A. Diogenes and K. M. Hargreaves, J. Endod., 2017, 43, S95–S101.
12. N. Dubey, J. Xu, Z. Zhang, J. E. Nör and M. C. Bottino, J. Endod., 2019, 45, 882–889.
13. K. M. Galler, W. Buchalla, K.-A. Hiller, M. Federlin, A. Eidt, M. Schiefersteiner and G. Schmalz, J. Endod., 2015, 41, 363–368.
14. A. R. Diogenes, N. B. Ruparel, F. B. Teixeira and K. M. Hargreaves, J. Endod., 2014, 40, S52–S57.
15. M. T. Albuquerque, J. Nagata and M. C. Bottino, J. Endod., 2017, 43, S51–S56.
16. J. S. Ribeiro, E. A. Bordini, J. A. Ferreira, L. Mei, N. Dubey, J. C. Fenno, E. Piva, R. G. Lund, A. Schwendeman and M. C. Bottino, ACS Appl. Mater. Interfaces, 2020, 12, 16006–16017.
17. J. S. Ribeiro, A. Daghrery, N. Dubey, C. Li, L. Mei, J. C. Fenno, A. Schwendeman, Z. Aytac and M. C. Bottino, Biomacromolecules, 2020, 21, 3945–3956.
18. M. Bekhouche, M. Bolon, F. Charriaud, M. Lamrayah, D. Da Costa, C. Primard, A. Costantini, M. Pasdeloup, S. Gobert and F. Mallein-Gerin, J. Mater. Chem. B, 2020, 8, 8422–8432.
19. R. Gianni-Barrera, N. Di Maggio, L. Melly, M. G. Burger, E. Mujagic, L. Gürke, D. J. Schaefer and A. Banfi, Stem Cells Transl. Med., 2020, 9, 433–444.
20. M. A. Saghiri, A. Asatourian and N. Sheibani, Oral Surg. Oral Med. Oral Pathol. Oral Radiol., 2015, 119, 122.
21. W. L. Dissanayaka, K. M. Hargreaves, L. Jin, L. P. Samaranayake and C. Zhang, Tissue Eng., Part A, 2015, 21, 550–563.
22. H. J. Jung, I. Seo, B. K. Jha, S. I. Suh, M. H. Suh and W. K. Baek, Arch. Biochem. Biophys., 2014, 545, 74–82.
23. C.-H. Li, P.-L. Liao, Y.-T. Yang, S.-H. Huang, C.-H. Lin, Y.-W. Cheng and J.-J. Kang, Arch. Toxicol., 2014, 88, 659–671.
24. A. Karczewski, S. A. Feitosa, E. I. Hamer, D. Pankajakshan, R. L. Gregory, K. J. Spolnik and M. C. Bottino, J. Endod., 2018, 44, 155–162.
25. J. S. Ribeiro, C. K. Sanz, E. A. Münchow, N. Kalra, N. Dubey, C. E. C. Suárez, J. C. Fenno, R. G. Lund and M. C. Bottino, Dent. Mater., 2022, 38, 1507–1517.
26. J. Li and D. J. Mooney, Nat. Rev. Mater., 2016, 1, 16071.
27. M. Eslamian, M. Khorrami, N. Yi, S. Majd and M. R. Abidian, J. Mater. Chem. B, 2019, 7, 224–232.
28. J. Wu, Z. Zhang, W. Zhou, X. Liang, G. Zhou, C. C. Han, S. Xu and Y. Liu, J. Controlled Release, 2020, 320, 337–346.
29. L. H. Nguyen, M. Gao, J. Lin, W. Wu, J. Wang and S. Y. Chew, Sci. Rep., 2017, 7, 1–12.
30. A. Romero-Montero, P. Labra-Vázquez, L. J. del Valle, J. Puiggalí, R. García-Arrazola, C. Montiel and M. Gimeno, RSC Adv., 2020, 10, 30508–30518.
31. K. Yue, G. Trujillo-de Santiago, M. M. Alvarez, A. Tamayol, N. Annabi and A. Khademhosseini, Biomaterials, 2015, 73, 254–271.
32. O. S. Muddineti and A. Omri, Expert Opin. Drug Delivery, 2022, 19, 559–576.
33. E. A. Münchow, A. F. da Silva, E. Piva, C. E. Cuevas-Suárez, M. T. de Albuquerque, R. Pinal, R. L. Gregory, L. Breschi and M. C. Bottino, J. Mater. Chem. B, 2020, 8, 10797–10811.
34. A. Daghrery, Z. Aytac, N. Dubey, L. Mei, A. Schwendeman and M. C. Bottino, Colloids Surf., B, 2020, 111011.
35. T. Sangnim, S. Limmatvapirat, J. Nunthanid, P. Sriamornsak, W. Sittikijyothin, S. Wannachaiyasit and K. Huanbutta, Asian J. Pharm. Sci., 2018, 13, 450–458.
36. N. Dubey, J. A. Ferreira, A. Daghrery, Z. Aytac, J. Malda, S. B. Bhaduri and M. C. Bottino, Acta Biomater., 2020, 113, 164–176.
37. A. Paul, V. Manoharan, D. Krafft, A. Assmann, J. A. Uquillas, S. R. Shin, A. Hasan, M. A. Hussain, A. Memic and A. K. Gaharwar, J. Mater. Chem. B, 2016, 4, 3544–3554.
38. S. Ramesh, L. Lungaro, D. Tsikritsis, E. Weflen, I. V. Rivero and A. P. Elfick, J. Appl. Polym. Sci., 2018, 135, 46692.
39. Z. Guo, C. Yang, Z. Zhou, S. Chen and F. Li, RSC Adv., 2017, 7, 34063–34070.
40. A. Chyzy and M. E. Plonska-Brzezinska, Molecules, 2020, 25, 5795.
41. P. P. Sikdar, M. M. Uddin, T. M. Dip, S. Islam, M. S. Hoque, A. K. Dhar and S. Wu, Mater. Adv., 2021, 2, 4532–4573.
42. M. Akrami-Hasan-Kohal, M. Eskandari and A. Solouk, Colloids Surf., B, 2021, 205, 111892.
43. D. Painuly, U. Nisha, S. Arya and J. S. Krishnan, J. Drug Delivery Sci. Technol., 2019, 51, 454–463.
44. G. Buschle-Diller, J. Cooper, Z. Xie, Y. Wu, J. Waldrup and X. Ren, Cellulose, 2007, 14, 553–562.
45. D. Caccavo, Int. J. Pharm., 2019, 560, 175–190.
46. P. Trucillo, Processes, 2022, 10, 1094.
47. J. Y. Nagata, A. J. Soares, F. J. Souza-Filho, A. A. Zaia, C. C. Ferraz, J. F. Almeida and B. P. Gomes, J. Endod., 2014, 40, 778–783.
48. L. E. Chávez de Paz Villanueva, Oral Surg. Oral Med. Oral Pathol. Oral Radiol. Endod., 2002, 93, 179–183.
49. A. C. Das, S. K. Sahoo, A. S. Parihar, S. S. Bhardwaj, P. Babaji and J. G. Varghese, J. Family Med. Prim. Care, 2020, 9, 239.
50. K. G. Bhat, P. Khot, S. Patil, G. Pattar and S. Majukar, J. Oral Maxillofac. Pathol., 2019, 23, 231.
51. A. Dewi, C. Upara, T. Krongbaramee, P. Louwakul, T. Srisuwan and S. Khemaleelakul, Aust. Endod. J., 2021, 47, 273–280.
52. M. C. Bottino, M. T. Albuquerque, A. Azabi, E. A. Münchow, K. J. Spolnik, J. E. Nör and P. C. Edwards, J. Biomed. Mater. Res., Part B, 2019, 107, 1576–1586.
53. M. G. Sousa, M. R. Maximiano, R. A. Costa, T. M. Rezende and O. L. Franco, Expert Opin. Drug Delivery, 2020, 17, 919–930.
54. G. Carpentier, S. Berndt, S. Ferratge, W. Rasband, M. Cuendet, G. Uzan and P. Albanese, Sci. Rep., 2020, 10, 1–13.
55. G. Yang, Z. Xiao, H. Long, K. Ma, J. Zhang, X. Ren and J. Zhang, Sci. Rep., 2018, 8, 1–13.
56. B. H. Lee, H. Shirahama, M. H. Kim, J. H. Lee, N.-J. Cho and L. P. Tan, NPG Asia Mater., 2017, 9, e412–e412.
57. S. Heltmann-Meyer, D. Steiner, C. Müller, D. Schneidereit, O. Friedrich, S. Salehi, F. B. Engel, A. Arkudas and R. E. Horch, Biomed. Mater., 2021, 16, 065004.
58. P. J. Palma, J. C. Ramos, J. B. Martins, A. Diogenes, M. H. Figueiredo, P. Ferreira, C. Viegas and J. M. Santos, J. Endod., 2017, 43, 1279–1287.
59. M. A. Saghiri, A. Asatourian, C. M. Sorenson and N. Sheibani, J. Endod., 2015, 41, 797–803.
60. D. Ribatti, In Vivo Models to Study Angiogenesis, Academic Press, 2017.
61. Z. Zhang, F. Nör, M. Oh, C. Cucco, S. Shi and J. E. Nör, Stem Cells, 2016, 34, 1576–1587.

---