Multifunctionalized and Dual‐Crosslinked Hydrogel Promotes Inflammation Resolution and Bone Regeneration via NLRP3 Inhibition in Periodontitis

Alveolar bone resorption caused by bacteria‐induced periodontitis remains challenging due to sustained inflammation. Periodontal pathogens like Porphyromonas gingivalis launch the primed signal of NOD‐like receptor family pyrin domain‐containing 3 (NLRP3) inflammasome in macrophages; consequent overproduction of proinflammatory cytokines and reactive oxygen species (ROS) leads to tissue destruction. This provides potential targets for a new therapeutic strategy. Herein, a multifunctionalized and dual‐crosslinked hydrogel pGM/cPL@NI with NLRP3 inhibitor MCC950 loaded is prepared. Driven by the strategic functionalization of gelatin methacryloyl and ε‐poly‐lysine with phenylboronic acid and catechol, respectively, pGM/cPL@NI containing dynamic and photo‐crosslinking networks demonstrates superior mechanical strength and stimuli‐responsive behavior, as well as the overwhelmed performance in bacteria killing and ROS scavenging. Crucially, pGM/cPL@NI restores the compromised osteogenesis by specifically suppressing the proinflammatory cytokine cascade triggered by NLRP3 inflammasome activation and promoting anti‐inflammatory polarization of macrophages. Collectively, pGM/cPL@NI presents robust potential as an effective “cocktail therapy” by combining antibacterial, antioxidant, inflammation resolution, and tissue regenerative functions. The present study reveals the underlying mechanism of the bacterial‐immune‐regeneration cascade and provides an extended approach for periodontal tissue engineering.

Alveolar bone resorption caused by bacteria-induced periodontitis remains challenging due to sustained inflammation.Periodontal pathogens like Porphyromonas gingivalis launch the primed signal of NOD-like receptor family pyrin domain-containing 3 (NLRP3) inflammasome in macrophages; consequent overproduction of proinflammatory cytokines and reactive oxygen species (ROS) leads to tissue destruction.This provides potential targets for a new therapeutic strategy.Herein, a multifunctionalized and dual-crosslinked hydrogel pGM/ cPL@NI with NLRP3 inhibitor MCC950 loaded is prepared.Driven by the strategic functionalization of gelatin methacryloyl and ε-poly-lysine with phenylboronic acid and catechol, respectively, pGM/cPL@NI containing dynamic and photo-crosslinking networks demonstrates superior mechanical strength and stimuli-responsive behavior, as well as the overwhelmed performance in bacteria killing and ROS scavenging.Crucially, pGM/cPL@NI restores the compromised osteogenesis by specifically suppressing the proinflammatory cytokine cascade triggered by NLRP3 inflammasome activation and promoting anti-inflammatory polarization of macrophages.Collectively, pGM/cPL@NI presents robust potential as an effective "cocktail therapy" by combining antibacterial, antioxidant, inflammation resolution, and tissue regenerative functions.The present study reveals the underlying mechanism of the bacterial-immune-regeneration cascade and provides an extended approach for periodontal tissue engineering.inflammasome, a cytosolic protein complex associated with various disorders, determines the production of mature IL-1β and IL-18.Accumulative studies have demonstrated that ROS and pathogens such as Porphyromonas gingivalis trigger the activation of macrophage NLRP3 via either canonical or noncanonical pathways.Subsequently, apoptosis-associated speck-like protein containing a CARD (ASC) and pro-Caspase-1 are recruited to finalize the assembly of NLRP3 inflammasome. [7]ro-IL-1β and pro-IL-18 precursors are cleaved into mature IL-1β and IL-18 by activated Caspase-1, which also activates gasdermin D (GSDMD) and ultimately causes pyroptosis of macrophages, while proinflammatory cytokines are consequently secreted and generated cytokine cascade. [8,9]MCC950 is a well-studied small molecule that interacts with the Walker B motif to selectively limit the activation of NLRP3, it has been reported to have a favorable effect on inflammasome-linked diseases. [10]After transition from the inflammation to the regeneration phases, mesenchyme stem cells (MSCs) are recruited and exert crucial effects on bone regeneration. [11]The normal function of MSCs is affected by immunological microenvironments, MSC-macrophage interaction addresses vital effects on bone homeostasis. [12,13]The NLRP3 inflammasome not only increases osteoclast-dominated bone resorption and stimulation of the nuclear factor kappa B (NF-κB) pathway, but also compromises the MSCs' capacity for osteogenic differentiation. [14]This premise leads to the proposal and emphasis of "osteoimmunology" in resolving periodontal bone destruction.Targeted inhibition of macrophages NLRP3 inflammasome activation presents a unique approach for periodontitis treatments.
Current nonsurgical treatments (supragingival or subgingival scaling) are considered inefficient in severe periodontitis. [15]xogenous repair technologies, including as guided tissue regeneration and bone transplantation, are being developed to facilitate tissue regeneration in periodontitis. [16]However, significant limitations have been exhibited as previous studies demonstrated. [17]ecently, with the development of periodontal tissue engineering, functionalized bioscaffolds as endogenous repair strategies present extensive promise. [18,19]Extracellular matrix-like hydrogels are regarded as the superior ones among them owing to their plasticity, biocompatibility, and drug-delivery properties.Varieties of periodontium-adaptive hydrogels have been developed, with varying physiochemical characteristics and structural designs with or without bioactive molecules loaded. [12,20,21]ydrogels are powerful carriers of for drugs, cells, exosomes, or small molecules that can be engineered to be stimuliresponsive by introducing dynamic bonds. [22]However, the majority of simply dynamic-crosslinked hydrogels possess relatively weak mechanical strength, which leads to good injectability but subpar performance in facilitating hard bone formation. [23]oreover, the previous strategies tended to focus more on antibacterial treatments than on inflammation resolution, or attempted to achieve bone regeneration by simply introducing osteogenic factors without regulating the immune microenvironment.In particular, the significance of the NLRP3 inflammasome in periodontitis has not been addressed by any of those inflammation-targeted treatments.As treating periodontitis is a complex and multistage process, a therapeutic hydrogel system with favorable characteristics that responds to different stages is expected and essential.
Herein, we proposed a novel dual-crosslinked hydrogel system pGM/cPL@NI with enhanced mechanical strength and multifunction that satisfies various needs in periodontitis treatment.By functionalized modification of gelatin methacryloyl (GelMA) and ε-poly-lysine (ε-PL) with phenylboronic acid (PBA) and catechol, respectively, MCC950-loaded pGM/cPL@NI was formed by the combination of dynamic borate ester bond and photoinduced crosslinking under irradiation.Favorable antibacterial and oxidative properties were promisingly achieved by the inherent characteristics of ε-PL and catechol.Under ROS or acid-enriched environment, pGM/cPL@NI responsively releases MCC950 and inhibits NLRP3 inflammasome activation in macrophages, thereby attenuating proinflammatory polarization and the cytokine cascade.The present synergistic strategy provides a potential and viable scheme to orchestrate the infectionimmune-regeneration cascade, thus leading to superior effects in rescuing periodontal destruction (Figure 1).

Preparation and Gelation Behavior
Based on the complexity and multistage microenvironment of periodontal inflammation, biomaterial scaffolds for treating periodontitis are required for optimal tissue adaptability and composite functions. [24]Additionally, stimuli-responsive smart biomaterials draw widespread attention to satisfy various application fields; these tunable and dynamic crosslinking designs are efficient in drug delivery and tissue-adaptability due to advanced shear-thinning and self-healing behavior. [22,25]owever, the relatively weak mechanical strength of dynamic crosslinked materials prevents them from undergoing stable, long-term repair. [26]In periodontal application, injectability and stimuli-responsive activities were typically achieved by dynamic covalent bonds or physical bonds, but simple dynamic hydrogels consequently sacrifice their mechanical strength and stability, which limits their application in long-term periodontal treatment.Based on the perspective of previous studies, dualcrosslinked hydrogels exhibit superior mechanical strength and toughness above single-network hydrogels. [27]Hence, dual-crosslinked hydrogels with good injectability, mechanical properties, and responsive drug-release behavior were introduced in the present study.
The dual-crosslinked network of pGM/cPL@NI was completed by the combination of two functionalized precursors.Briefly, PBA and catechol groups were grafted onto gelatin and ε-PL to generate GelMA-PBA (pGM) and ε-PL-cat (cPL), respectively (Figure 2A).Subsequently, dual-crosslinked hydrogel pGM/cPL and pGM/cPL@NI were formed with or without NLRP3 inhibitor (NI) MCC950 loaded by the two-step crosslinking between GelMA-PBA and ε-PL-cat.As the formation of a borate ester bond requires the condition of pH > pKa, we successfully prepared the first crosslinked network by combing the precursor solution of GelMA-PBA and ε-PL-cat at pH 8.0, which represented the in situ hydrogel formation.The second network was dependent on photo-crosslinking of GelMA-PBA by free radical polymerization following blue light irradiation.The hydrogel networks of pGM/cPL and pGM/cPL@NI were theoretically extremely sensitive to ROS and acid because of the borate ester bond (Figure 2B).While the first dynamic network responsively collapsed, the second photo-crosslinked network remained stable and effectively releases embedded drugs without rapid destruction of the hydrogel structure.According to the results of a preliminary study to verify the suitable gelation time, the molar ratio of GelMA-PBA and ε-PL-cat was fixed at 1:1.Next, the gelation time of the hydrogel was investigated.The sol-gel transition of hydrogels was rapidly observed after the component mixture (Figure 2C); the first gelation occurred in 4.36 AE 0.46 s, followed by further photo-crosslinking in 38.4 AE 2.38 s (Figure S1A, Supporting Information).Then, the effect of various stimuli on the gelation behavior of pGM/cPL and pGM/cPL@NI was observed.The sol-gel transition was not successful in pGM/cPL and pGM/cPL@NI after pretreatment with H 2 O 2 or HCl prior to irradiation; however, the second photo-initiated crosslinking was unaffected to form a single-crosslinked hydrogels (Figure 2D and S1B, Supporting Information).

Physicochemical and Rheological Properties
First, the chemical structure of pGM/cPL was characterized by Fourier transform infrared spectrometer (FTIR) and 1 H NMR (Figure 1E,F), confirming the effective grafting of PBA and catechol groups.Following the reaction with hydroxycitric acid (HCA), the primary amine group in ε-PL-cat disappeared, and the presence of a benzene ring caused the C═O stretching vibration of the secondary amide to shift from 1660 to 1632 cm À1 .A significant absorption peak resulting from B─O stretching vibration emerged at 1335 cm À1 in gelatin-PBA after grafting of PBA.Then, as injectability is crucial for filling irregular defect regions and providing in situ tissue regeneration in periodontal treatments, we investigated the injectability and rheological behavior of hydrogels. [20,25,28]As shown in Figure 2G, pGM/cPL was formed and injected easily by an injector containing a mixture of GelMA-PBA and ε-PL-cat under both dry and moist environments.In order to define the effect of components' proportion on the physical properties of hydrogels, three concentrations (5% w/v, 7.5% w/v, and 10% w/v) of pGM/cPL hydrogel were synthesized by mixing 4% w/v ε-PL-cat with 10% w/v, 15% w/v, and 20% w/v of GelMA-PBA at a ratio of 1:1.Based on the rheological tests (Figure 2H), it was shown that the storage moduli (G') of pGM/cPL with varying proportions were consistently higher than the loss moduli (G") during the 0-300 s period, suggesting a successful sol-gel transition regardless of whether the hydrogels were irradiated or not.Meanwhile, the storage moduli of hydrogel also increased significantly when the proportion of GelMA-PBA increased.The hydrogels' storage moduli exceeded 1 kPa following photo-crosslinking, demonstrating the significant improved mechanical property generated by the dual-crosslinked network formation.The results of compression tests showed that the compression modulus of pGM/cPL increased with the proportion of GelMA-PBA increased (Figure S1C, Supporting Information).No significant difference was observed between 5% w/v of pGM/cPL (2.02 AE 0.02 kPa) and 7.5% w/v of pGM/cPL (3.33 AE 0.54 kPa), while 10% w/v of pGM/cPL (4.51 AE 1.84 kPa) exhibited significantly enhanced strength (p < 0.05).
The borate ester bond theoretically endowed the dualcrosslinked hydrogels with predictable dynamic cleavage properties under ROS or acid stimulation in the present study, so we testified the responsive behavior by the following experiments. [29]First, in vitro degradation behavior of pGM/cPL was detected in different conditions during the 28 d.As indicated in Figure 2J, pGM/cPL degraded slower than GelMA in a neutral buffering environment, which remained more mass after 28 d.However, the degradation rate of pGM/PL was significantly faster in acid or ROS-enriched environments in an H 2 O 2 concentration-dependent manner, further confirming the stimuli-responsive property.Moreover, the drug-release behavior of pGM/cPL@NI after MCC950 loaded was consistent with the results above (Figure 2K); model drug Rhodamine B showed sustained-release behavior within 40 d under normal environments.However, the release rate was significantly accelerated within 2 d when pH was 5.5 or in the presence of 0.1 and 1 mM H 2 O 2 .In summary, the results indicated that the dual-crosslinked pGM/cPL hydrogel presented more supportive potential in long-term retention and hard tissue formation for modified mechanical properties and drug-delivery patterns.

Biocompatibility and Antimicrobial Properties of Dualcrosslinked Hydrogels
Biocompatibility of hydrogels as bioscaffolds in periodontal tissue engineering, which also prevents hydrogels from immune rejection, is necessary because that satisfactory cell function is a prerequisite for periodontal rehabilitation. [30]To verify the effects of versatile hydrogels on cell viability, proliferation, and migration, primary rat bone marrow stem cells (rBMSCs) (Figure S2, Supporting Information) and murine fibroblast cells (L929) were employed as model cells in CCK-8, live/dead staining, or cell scratch assays.As illustrated in Figure 3A, live/dead staining indicated favorable cell viability following coculturing with pGM/cPL and pGM/cPL@NI.Following treatment with dual-crosslinked hydrogels, cells proliferated and disseminated more considerably during the 7 d culture period compared to the GelMA group, with tighter intracellular junctions observed.Similarly, from the results of CCK8 assays, viability and proliferation of rBMSCs in pGM/cPL and pGM/cPL@NI groups were slightly higher than those in the GelMA group on days 1 and 4, but no significant difference was observed between the groups (p > 0.05).After 7 d, the facilitation of pGM/cPL and pGM/ cPL@NI on rBMSCs proliferation remarkably evolved (p < 0.001) (Figure 3B).Furthermore, cell scratch assay revealed the prominent accelerated migration by treating with pGM/cPL and pGM/cPL@NI (Figure 3C).Following a 36 h period, the healing rate of these two groups reached 86.08 AE 2.05% and 86.17 AE 3.23%, respectively, which were 1.2-1.3times greater than those of the control and GelMA groups (Figure S3A, Supporting Information).Finally, we clarified the hemocompatibility through the hemolysis assay, both single-crosslinked and dual-crosslinked pGM/cPL with or without MCC950 encapsulated showed a hemolysis rate of less than 5%, which is sufficient to support long-term in vivo use (Figure S3B, Supporting Information).
Oral dysbacteriosis initiates the signal of periodontal inflammation; excessive periodontal pathogen like P. gingivalis induces host inflammation mainly through virulent lipopolysaccharide (LPS), continually resulting in periodontal tissue injuries. [31,32]ere, inspired by mussel foot protein, [33] ε-PL with natural antimicrobial activity was selected for solving infection without the need for extra supplements. [34]Moreover, we modified ε-PL with catechol groups to form ε-PL-cat, which endows pGM/cPL and pGM/cPL@NI with synergic bactericidal mechanism as shown in Figure 3D.Based on this, the antimicrobial activities of pGM/ cPL and pGM/cPL@NI were subsequently clarified against P. gingivalis (ATCC 33277) and Escherichia coli (E.coli, ATCC 25922).In accordance with the reduction of bacterial suspension concentration by measuring OD 600 (Figure 3E and S4A, Supporting Information), colony forming units (CFUs) of P. gingivalis and E. coli were both significantly inhibited by coculturing with pGM/cPL and pGM/cPL@NI (Figure 3F).Compared to the PBS and GelMA groups, the relative bacterial viability in pGM/cPL and pGM/cPL@NI groups was remarkably decreased (p < 0.001), with less than 15% of E. coli and 2% of P. gingivalis surviving in the presence of dual-crosslinked hydrogels (Figure 3G).Growth of E. coli was significantly suppressed after 2 h, while P. gingivalis exerted a similar pattern after 6 h (Figure S4B,C, Supporting Information).As for live/dead staining (Figure 3H), dead bacteria labeled by propidium iodide (PI) notably increased in two dual-crosslinked hydrogels groups.Notably, following interaction with pGM/cPL and pGM/cPL@NI, E. coli and P. gingivalis tended to cluster.We further investigated the morphologies of bacteria after certain interventions by scanning electron microscopy (SEM) (Figure 3I); P. gingivalis and E. coli cultured with pGM/cPL and pGM/cPL@NI displayed irregular shape, shrinkage, or broken membranes.Concurrently, clustered tendencies were also observed consistent with live/dead staining.Overall, the multifunctional hydrogels pGM/cPL and pGM/cPL@NI performed considerable killing ability against bacteria, providing a solid foundation for the management of periodontal chronic inflammation.By contrast with the control and GelMA groups, the underlying antibacterial mechanism is mainly attributed to the cation-induced cell/wall membrane lysis of ε-PL as previously described. [35]DOPA residues from catechol synergically kill microbe through autoxidation; however, this impact was reduced by the presence of free phenolic hydroxyl groups in pGM/cPL and pGM/cPL@NI following gelation.

pGM/cPL@NI Effectively Attenuated P. gingivalis-induced NLRP3 Inflammasome Activation and Proinflammatory Cytokines Overproduction in Macrophages
Accumulative evidence has shown the importance of macrophages in defending periodontal infection and launching regeneration. [4,36]Host inflammation responds to bacterial invasion; however, sustained and chronic infection results in uncontrollable inflammatory effects that could compromise periodontal homeostasis. [31,37]Pathogens like P. gingivalis pull the danger alarm as pathogen-associated molecular patterns (PAMPs), which induce the pyroptosis of macrophages via NLRP3 inflammasome activation.While macrophage pyroptosis facilitates the clearance of microorganisms to a certain extent, the secondary pro-inflammatory cytokine cascade causes extensive tissue destruction. [38]In addition, pyroptosis leads to the efflux of intracellular contents, which, in turn, provides nutrients for the survival of periodontal pathogens. [39]Hence, targeted inhibition of NLRP3 inflammasome provides a potential strategy for treating periodontitis.Based on this, we loaded NLRP3 inhibitor MCC950 into the hydrogel network to utilize its inflammation resolution function; the specific mechanism is described in Figure 4A.
To clarify the influence of P. gingivalis infection on macrophages, the bone marrow-derived macrophages (BMDMs) were cultured and characterized (Figure S5, Supporting Information), followed by the establishment of P. gingivalis-BMDM coculture models as previously described. [40,41]BMDMs morphologies and viability changed significantly as multiplicity of infection (MOI) increased (Figure S6A,B, Supporting Information).Then, enzyme-linked immunosorbent assay (ELISA) was implemented to detect the expression level of IL-1β in BMDMs supernatants after P. gingivalis infection.The results demonstrated that the extracellular IL-1β secretion was significantly increased after P. gingivalis induction in an MOI or infection time-dependent manner (p < 0.05) (Figure 4B).Western blots were performed to assess protein expressions NLRP3 and pro-IL-1β in whole cell lysates following P. gingivalis-BMDMs coculture (Figure 4C and S6C, Supporting Information).It was observed that P. gingivalis infection significantly promoted the upregulation of NLRP3 and pro-IL-1β, which initiated the increase of IL-1β secretion as ELISA results.
As pGM/cPL and pGM/cPL@NI with inherent bactericidal activities could cause the diversity in P. gingivalis counts, a bacteria-free stimulation protocol using LPS/Nigericin was performed to activate the NLRP3 inflammasome according to previous reports. [42]Meanwhile, to define whether P. gingivalisinduced IL-1β expression depended on NLRP3 activation, we treated BMDMs with NLRP3-inhibitor MCC950 before P. gingivalis-LPS/Nigericin stimulation.A significant increase in extracellular IL-1β levels was detected after P. gingivalis-LPS/Nigericin stimulation, which was then concentrationdependently inhibited by MCC950 (Figure 4D).CCK-8 assays were conducted to optimize the loading concentration of MCC950 in hydrogels.The results showed that within 24 h, 0-10 μм of MCC950 did not exhibit significant cytotoxicity, but cell viability in the 10 μм group significantly decreased after 48 h (Figure S6D, Supporting Information).
Activation of the NLRP3 inflammasome mediates cell pyroptosis, membrane permeability increases, and lactate dehydrogenase (LDH) is extracellularly released. [43]Labeling of damaged cell membrane by PI, pGM/cPL, and pGM/cPL@NI greatly restored the viability and permeability of cell membrane after stimulation (Figure S7, Supporting Information).In addition, BMDMs infected with P. gingivalis gradually secreted LDH as a result of ruptured membranes as time increased; however, hydrogels significantly reduced this increase in LDH release over a 4 h period.Additionally, pGM/cPL@NI loaded with MCC950 showed even more advanced inhibition than pGM/cPL (p < 0.001) (Figure 4E).The immunomodulatory effects of pGM/cPL and pGM/cPL@NI were further revealed by the following western blots, qRT-PCR, and ELISA.At the protein level, the results showed that P. gingivalis-LPS/Nigericin pretreatment remarkably up-regulated the expression of NLRP3 in accordance with P. gingivalis direct infection (p < 0.01) (Figure 4F,G).pGM/ cPL and pGM/cPL@NI hydrogel hardly induced the increase of NLRP3 in normal conditions, but pGM/cPL@NI was found to effectively downregulate NLRP3 in inflammatory conditions compared with MCC950-free pGM/cPL (p < 0.05).Furthermore, the subsequent activation of NLRP3, Caspase-1, and GSDMD─which are in charge of inflammasome assembly, immature precursor cutting, and secretion pathways supplying, respectively─is necessary for the secretion of mature proinflammatory cytokines including IL-1β and IL-18. [44]mRNA expression of Nlrp3, Caspase1, and Gsdmd were detected (Table S1, Supporting Information), and the pattern of the changes was in line with western blots.Following hydrogel treatment, as shown in Figure 4H, expression of Nlrp3, Caspase1, and Gsdmd all showed tendencies of considerable decline, particularly in the pGM/cPL@NI group.Significant differences between pGM/cPL and pGM/cPL@NI were discovered in the meantime (p < 0.001).P. gingivalis-induced cell pyroptosis upregulated NLRP3 inflammasome, which was followed by the production of proinflammatory cytokines and activation of NF-κB.Under this premise, we further evaluated the extracellular expression of proinflammatory cytokines including IL-1β, IL-18, and TNF-α (Figure 4I).Three cytokines secreted significantly less when treated with hydrogels; pGM/cPL@NI presented the most pronounced control.Interestingly, while significant differences were observed between pGM/cPL and pGM/ cPL@NI in IL-1β and IL-18 secretion (p < 0.01), there were no statistically significant differences in TNF-α between the two hydrogels.In general, pGM/cPL@NI represented the superior inhibition effect on NLRP3 inflammasome activation of pGM/cPL@NI compared to the inhibitor-free ones, which is promising in preventing cytokine cascade mediated by NLRP3.

pGM/cPL@NI Scavenged Intracellular Oxidative Stress and Promoted Anti-inflammatory Macrophage Polarization in Vitro
The overproduction of ROS in periodontitis is mostly caused by periodontal bacteria.The primary pathogenic factor of LPS exaggerates the intracellular ROS concentration and undermines the endogenous antioxidative system like catalase and superoxide dismutase.Overexposure to ROS inhibits the osteogenic differentiation of BMSCs in addition to stimulating the production of osteoclasts through activation of the NF-κB pathway. [45]Besides, ROS serves as one of the danger sensors that consequently triggers NLRP3 inflammasome assembling. [9]Therefore, reducing oxidative stress at damage sites is crucial for periodontal inflammation resolution and bone regeneration.Due to the modified catechol on ε-PL, the phenolic hydroxyl groups become the optimal hydrogen or electron donor for ROS, effectively scavenging high-concentration of ROS.To investigate the antioxidant activities of hydrogels, cells were exposed to either P. gingivalis-LPS or H 2 O 2 , and then 2 0 -7 0 -dichlorodihydrofluorescein diacetate (DCFH-DA) probe was used to label intracellular ROS by detecting DCF intensity.As results indicated (Figure 5A,B), both pGM/ cPL and pGM/cPL@NI maintained the redox hemostasis under normal conditions, suggesting that hydrogels by themselves were unable to cause distinct ROS generation.Under P. gingivalis-LPS or H 2 O 2 stimulation, intracellular ROS levels were dramatically increased in macrophages.After coculturing cells with pGM/cPL or pGM/cPL@NI, the fluorescent intensity of DCF decreased noticeably.According to flow cytometry results, after receiving pGM/cPL and pGM/cPL@NI treatments, the dramatically increased ROS expression was significantly scavenged by approximately 50% (p < 0.01) (Figure 5C).
Additionally, mitochondrial function is closely linked to redox homeostasis, which is reflected by the membrane potential changes.Studies have shown that ROS could increase the permeability of mitochondrial intima and guide intracellular Ca þ influx, impairing the normal function of mitochondria. [19,46]n the meantime, aberrant mitochondrial function, in turn, exacerbates oxidative stress and inflammatory cascades. [47]JC-1 probe distinguishes between the monomer or aggregate form by green or red fluorescence, denoting the decreased and normal membrane potential, respectively.Therefore, we implemented the JC-1 assay to evaluate the effect of hydrogels on safeguarding Raw 264.7's mitochondrial function in inflammatory conditions.As Figure 5D,E indicates, P. gingivalis-LPS-induced inflammation significantly reduced mitochondrial membrane potential and raised the ratiomonomers/aggregates by 24.59 AE 4.33 folds when compared to the control.Conversely, pGM/cPL and pGM/cPL@NI restored the membrane potential with more JC-1 aggregates existing (p < 0.001).Based on the results above, we hypothesized that pGM/cPL and pGM/cPL@NI combat the oxidative stress by synergistically performing antibacterial and antioxidant functions, which further effectively alleviates the damage to mitochondrial function caused by local ROS enrichment after periodontal infection.

Macrophage polarization manifests various functions in anti-infection and inflammation regulation because the balance between pro-(M1
) and anti-inflammatory (M2) macrophages is crucial for maintaining periodontal homeostasis. [4]Nonetheless, chronic and uncontrollable infection results in prolonged M1 polarization instead of M2 polarization of macrophages, suggesting a delayed transition from the anti-infection to the anti-inflammation stage.This could lead to the failure of tissue reconstruction.Hence, we investigated the potential of pGM/cPL and pGM/cPL@NI in polarization modulation.As flow cytometry presented, M2 phenotype marker CD206 was distinctly downregulated after P. gingivalis-LPS pretreatment (Figure 5F and S8, Supporting Information).By coculturing BMDMs with pGM/ cPL and pGM/cPL@NI, it was able to drastically reduce the inhibitory effect of P. gingivalisLPS on M2 polarization and encourage the cells to polarize from M0 type to M2 type (p < 0.01).Furthermore, by detecting protein expression with western blots and immunofluorescence (IF) staining (Figure 5G-K), CD206 and M1 phenotype marker iNOS were significantly downregulated and upregulated in P. gingivalis-LPS group, respectively (p < 0.01), while pGM/cPL and pGM/ cPL@NI tended to stimulate more conspicuous M2 polarization rather than M1 compared to P. gingivalis-LPS group (p < 0.001).Particularly, in contrast with pGM/cPL, pGM/cPL@NI exhibited superior modulation in preventing proinflammatory polarization of macrophages.It was speculated to be related to the targeted NLRP3 inhibition by MCC950, which diminished the chance of proinflammatory polarization driven by cytokines including IL-1β and IL-18.

pGM/cPL@NI Facilitated Osteogenic Differentiation of rBMSCs In Vitro
During periodontitis, sustained infection and subsequent inflammation are responsible for alveolar bone resorption.Chronic inflammation impedes the transition from the inflammation phase to the rehabilitation phase, rebalance of the disordered bone metabolism is urgently needed.To clarify the underlying influences, we cultured rBMSCs with pGM/cPL, pGM/cPL@NI or hydrogel extracts under normal or inflammatory conditions induced by P. gingivalis-LPS (Figure 6A).Considering that BMSCs-material surface contact usually exerts an important impact on their osteogenic differentiation, we first investigated the adhesion and spread behavior of rBMSCs cultured on hydrogels by cytoskeletal staining.As depicted in Figure S9 (Supporting Information), polygonal rBMSCs with lager intercellular space and less spread area were observed in the P. gingivalis-LPS group.In comparison, following pGM/ cPL and pGM/cPL@NI cocultivation, which offered satisfactory scaffolds base during osteogenesis, rBMSCs showed improved adhesion and spread behaviors, as seen by higher density and extended filopodia.Subsequently, in accordance with the previous studies, alkaline phosphatase (ALP) staining and alizarin red S (ARS) staining─which represent mineralization and late calcium deposition, respectively─were carried out to assess the osteogenic effects both qualitatively and quantitatively (Figure 6B and S10, Supporting Information). [12]The matrix mineralization in both the pGM/cPL and pGM/cPL@NI groups were significantly increased compared with the control group, suggesting the superior osteogenesis effectiveness of pGM/cPL and pGM/cPL@NI.Although osteogenic potential of rBMSCs was weakened after P. gingivalis-LPS induction, the results indicated the improved osteogenesis under inflammatory conditions by treating cells with pGM/cPL and pGM/cPL@NI.Furthermore, the relative expression levels of osteogenesisrelated genes were detected by qRT-PCR (Figure 6C).mRNA expression of osteogenesis-related genes (Runx2, Alp, Ocn, and Osx) were significantly upregulated in pGM/cPL and pGM/cPL@NI groups compared to those in the P. gingivalis-LPS group (p < 0.05).In accordance with the results above, expression of osteogenesis-related proteins (Runx2, OCN) showed similar trends in immunofluorescent staining (Figure 6D,E).Both pGM/cPL and pGM/cPL@NI hydrogels dramatically increased Runx2 and OCN expression when P. gingivalis-LPS was not present (p < 0.01).P. gingivalis-LPS significantly reduced osteogenesis, which was thereafter successfully restored by hydrogel intervention (p < 0.01).Nevertheless, no significant difference was observed between pGM/cPL and pGM/cPL@NI (p > 0.05).
Additionally, crosstalk between macrophages and BSMCs is essential for osteoimmunomodulation.Macrophages are considered to be indispensable for osteogenesis and bone homeostasis; those with abnormal function exacerbate the balance between bone regeneration and resorption. [48]Therefore, to define the indirect control of osteogenesis via macrophage modulation, rBMSCs were cocultured with macrophage-conditioned culture medium (Mφ-CM) preinduced by P. gingivalis-LPS and Nigericin to simulate NLRP3 inflammasome activation environment.In comparison with the control group, rBMSCs treated with inflammasome-activated Mφ-CM showed decreased calcium accumulation during osteogenic induction, according to ARS staining and semiquantitative analysis (Figure 6F and S8, Supporting Information).By the immunomodulation effect of pGM/cPL@NI, the conditioned medium significantly enhanced the matrix mineralization of rBMSCs (p < 0.001).runx2 and ocn were shown to be upregulated in qRT-PCR data following pGM/ cPL@NI intervention.However, there was no significant difference between P. gingivalis-LPS/Nigericin and pGM/cPL groups (p > 0.05) (Figure 6G).
Taken together, the multifunctional dual-crosslinked hydrogels pGM/cPL and pGM/cPL@NI are promising bioscaffolds for periodontal bone regeneration owing to the advanced control of BMSCs cytoskeleton and osteogenic differentiation.The RGD peptide sequence in GelMA-PBA exhibited advantageous cell adhesion qualities that facilitated cell attachments to hydrogels. [49]In addition, pGM/cPL and pGM/cPL@NI facilitated the expression of osteogenic-related genes and proteins due to strengthened stiffness consistent with the previous theories. [11]pecially, we found that pGM/cPL and pGM/cPL@NI exhibited similar effects on osteogenic differentiation when directly cultured with rBMSCs.However, in indirect coculture, pGM/ cPL@NI played a more outstanding role compared to the drug-free counterparts due to a variation in macrophage control by NLRP3 inhibitor MCC950.Inhibited NLRP3 activation in BMDMs led to less production of pro-inflammatory cytokines like IL-1β and IL-18 as previously demonstrated, preventing the osteogenic differentiation potential of rBMSCs from being compromised by cytokines.Moreover, M2-polarized macrophages (M2φ) are highlighted as osteogenesis promoter owing to enhanced secretion of osteoinductive cytokines (IL-10, BMP2) and reduced secretion of resorption-related factors (TNF-α, IL-6).Therefore, it was hypothesized that another underlying mechanism in modifying osteogenesis would be the differing effects of pGM/cPL and pGM/cPL@NI on macrophage M2 polarization.

pGM/cPL@NI Reduced Inflammatory Bone Loss in Experimental Rat Periodontitis In Vivo
The in vitro biocompatibilities of multifunctional dualcrosslinked hydrogels were confirmed to be outstanding, we further evaluated the in vivo biosafety.Subcutaneous implantation models were constructed in rats based on the usual method. [50]ollowing the 7-day implantation of pGM/cPL and pGM/ cPL@NI hydrogels, skin tissues, main internal organs, and fresh blood were collected for H&E staining and blood biochemical analysis.As shown in Figure 7A, scattered and few inflammatory cells infiltrated into interfaces between hydrogels and skin tissue with the absence of swelling, congestion, or necrosis of the surrounding tissue.No significant toxicity was observed in major internal organs (heart, liver, spleen, lung, and kidney) (Figure S11, Supporting Information).The biochemical analysis of blood revealed mild increases in immunocytes and platelet markers with little effect on erythrocytes and hemoglobin (Figure S12, Supporting Information).In general, the results above indicated satisfactory biocompatibility of dual-crosslinked hydrogels for in vivo use.
Then, the therapeutic effects of pGM/cPL and pGM/cPL@NI on periodontitis were clarified.We created rat experimental periodontitis models by combining silk ligature with P. gingivalis injection, as previously described, in accordance with the mechanism of periodontal inflammation, which holds that periodontal pathogens were the initiator of host inflammatory response and tissue resorption (Figure 7B). [44,51,52]Confirming by micro-CT and H&E staining, the periodontitis models were verified to be successfully established (Figure S13 and S14, Supporting Information).Following the induction of periodontitis, there was a substantial increase in the cementoenamel junctionalveolar bone crest (CEJ-ABC) distance to 0.92 AE 0.88 mm compared to the control group (p < 0.001) (Figure S15, Supporting Information).Subsequently, the rats were divided into five groups discriminated by PBS, pGM/cPL, or pGM/cPL @NI treatments.Considering the frequent oral activity and saliva flush, hydrogels were injected every 3 d.After 4 weeks of periodontitis induction and 4 weeks of hydrogels intervention, micro-CT showed significant intergroup differences (Figure 7C).Silk ligature and P. gingivalis infection caused distinct bone resorption in both the proximal and distal parts around the second molar, even reaching the apical 1/3.PBS treatment for 4 weeks barely reduced the bone loss, while bone regeneration was remarkably enhanced when pGM/cPL and pGM/ cPL@NI were introduced into the periodontitis area, resorption was confined to the cervical 1/3.By measuring the CEJ-ABC distances, the height of the alveolar bone was raised by approximately 0.3 mm in the pGM/cPL group (Figure 7D).More significantly, drug-loaded pGM/cPL@NI helped reduce the CEJ-ABC distance to 0.43 AE 0.02 mm, which was nearly 50% less than that of the control and PBS group (p < 0.01).Meanwhile, as Figure 7E indicates, the bone volume versus total volume (BV/TV) ratios were quantified based on micro-CT analyses, which was slightly increased with no statistical difference after PBS treatment in contrast to the periodontitis group (p > 0.05).pGM/cPL and pGM/cPL@NI both remarkably increased the bone mass.(p < 0.05).
Histological staining was further utilized for comprehensive analysis of inflammatory severity and regeneration.According to the results of H&E staining (Figure 7F), the junctional epithelium and alveolar bone presented normal attachment and height.However, both of them presented apical recession after periodontitis with formation of deep periodontal pockets, accompanied by abundant infiltration of inflammatory cells and destruction of interdental papilla.The undesirable periodontal condition was then notably improved in the hydrogels groups, especially in the pGM/cPL@NI group.Results of Masson trichrome staining then indicated the newly formed alveolar bone and collagen deposition were increased in pGM/cPL and pGM/ cPL@NI groups; the surface of new bone was relatively smooth and consecutive in contrast to that of the PBS group (Figure 7G and S16, Supporting Information).Based on these findings, we speculated that the superior effects in the pGM/cPL@NI group were driven by the responsive release of MCC950 that supplied favorable immunomodulatory function.Subsequently, to define the immunomodulation and osteogenic effects of hydrogels in vivo, the expression of CD206, IL-1β, and Runx2 was evaluated by IF staining and semiquantitative analysis (Figure 7H-K).Inflammatory status represented by IL-1β was unoptimistic after PBS treatments.By contrast, periodontal inflammation was significantly diminished after hydrogels application.Meanwhile, compared to the PBS group, fluorescence-labeled CD206 and Runx2 were widely distributed in the pGM/cPL and pGM/ cPL@NI groups, manifesting the positive regulation of periodontium toward tissue rehabilitation by treating with multifunctionalized hydrogels.MCC950-loaded pGM/cPL@NI exerted the excellent potential in modulation of these markers, giving rise to a more remarkable improvement in orchestrating inflammation resolution and bone regeneration.
As previously stated, excessive inflammation can substantially impede the repair of periodontal tissue.Consistent with previous in vitro studies, in vivo investigations demonstrated that targeted suppression of the NLRP3 inflammasome effectively reduced inflammatory stress and bone loss.Due to this, the dualcrosslinked hydrogels, particularly the MCC950 loaded ones, showed promising outcomes in terms of facilitating periodontal regeneration.We hypothesized that the antibacterial, ROSscavenging, and immunomodulation properties synergistically reinforced the bone-preventive effectiveness of pGM/cPL and pGM/cPL@NI.The pGM/cPL@NI system has enormous potential as an effective protocol to treat bacteria-induced periodontitis in the future due to its multifunctional and feasible design.

Conclusion
The present study developed a multifunctional drug-delivery hydrogel system pGM/cPL@NI with dual-crosslinked networks, which combines antibacterial, ROS scavenging, immunomodulation, and osteogenic differentiation promotion to exert favorable regenerative effects.The combination of dynamic and photoinduced crosslinking reinforced the mechanical strength and endowed pGM/cPL@NI with microenvironment-responsive release of NLRP3 inhibitor MCC950.Bactericidal and antioxidant effects are realized by inherent properties of ε-PL and catechol without introducing additional substances.Importantly, we highlighted the correlation between P. gingivalisinduced NLRP3 inflammasome activation and periodontitis progression, which is effectively blocked by MCC950 released from pGM/cPL@NI, further promoting the anti-inflammatory modulation of macrophages.Overall, our findings here reveal the promising potential of pGM/cPL@NI in periodontitis alleviation, which provides novel options for designing and applying periodontal regenerative biomaterials.

Experimental Section
Synthesis of GelMA-PBA and ε-PL-Cat: GelMA-PBA and ε-PL-cat were synthesized by the following procedures.First, gelatin (20 g), EDC (24 g), NHS (24 g), and PBA (13.3 g) were combined with 1000 mL of deionized water to create gelatin-PBA.After 3 days of stirring and dialysis, GelMA-PBA was created by reacting lyophilized gelatin-PBA (10 g) with methacrylic anhydride (0.6 mL) in 500 mL of deionized water.Meanwhile, ε-PL (1.0 g) was dissolved in deionized water (25 mL) to form 4% ε-PL, and then EDC (4.485 g) and NHS (0.535 g) were added and stirred.Subsequently, ε-PL-cat was prepared by dissolving HCA (0.855 g) with the solution above and stirring for 24 h.Finally, prepared GelMA-PBA and ε-PL-cat were lyophilized for further experiments.
H NMR: Samples (3-5 mg) were dissolved in deuteration reagent (D 2 O or CDCl3) and loaded into NMR tubes.The NMR structure was determined by a nuclear magnetic resonance spectrometer (Bruker, Germany) at room temperature; further analysis was completed by MestReNova (Mestrelab Research, Spain).
Rheological Test: 5% w/v, 7.5% w/v, and 10% w/v of hydrogels were prepared according to the method above; rheological tests were completed with a 25 mm diameter of stainless-steel parallel rotor by rotational rheometer (NETZSCH, Germany).G' and G" refer to the storage modulus and loss modulus, respectively.Dynamic strain sweeps from 0.1 to 10 rad s À1 are performed at room temperature.
Compression Test: The compression test was implemented to determine the mechanical strength of hydrogel.The hydrogels were prepared to cylindrical shapes and placed on the plate.By pushing the hydrogel to fracture using MTS E45.305 (MTS, USA) at room temperature, the pressure values were recorded and analyzed for calculating compression modulus.
In Vitro Degradation Behavior: In vitro degradation behavior of pGM/cPL hydrogel was investigated in different ROS and pH environments.Lyophilized pGM/cPL is weighed and marked as W 0 , and then immersed in PBS containing different concentrations of H 2 O 2 (0, 0.1, 1 mM) or different pH (pH 7.4 or 5.5) in an oscillator (37 °C, 150 rpm); PBS was refreshed every 3 d.The remaining weight of pGM/cPL was recorded as W t at each time point (3, 7, 14, 21, 28 d).The degradation rate of hydrogel was calculated by the following equation: Degradation rate ð%Þ ¼ W 0 ÀW t W 0 Â100% In Vitro Drug-Release Behavior: To study the drug-release behavior of pGM/cPL under different pH (pH 5.5 and 7.4) or ROS (H 2 O 2 = 0, 0.1, 1 mM) environments, Rhodamine B (Macklin, China) was selected as the model drug according to the similar molecular weight to MCC950.Rhodamine B (2 mg) powder was added to the pGM/cPL pregel solution before complete gelation.The drug-loaded hydrogel was immersed in PBS (10 mL) and placed onto an oscillator (37 °C, 70 rpm).10 mL of PBS was collected and replaced with fresh PBS at various time points.Finally, the absorbance of the sample solution was detected by a microplate reader (λ = 552 nm) (BioTek, USA).
Cell Isolation and Culture: BMDMs were harvested from 4 week old male C57BL/6 mice as previously described. [40]The mice were euthanized by cervical dislocation, followed by sterilization for 15 min.The bone marrow was then extracted by isolating and flushing the tibiae and femurs.Centrifuged the bone marrow with high-glucose Dulbecco's Modified Eagle Medium (DMEM medium) (Gibco, USA), the erythrocytes were lysed for 30s after the supernatant was removed, centrifuged again, and filtered by 70 μm cell strainers before the cells were seeded into cell culture dishes under compatible atmosphere (5% CO 2 , 37 °C).The culture medium contained 20 ng mL À1 macrophage colony stimulating factor (M-CSF, Novoprotein, China), 10% v/v fetal bovine serum (FBS, ExCell Bio, China), and 1% penicillin-streptomycin (Gibco, USA).After 7 d induction, the mature BMDMs were characterized and used in the following experiments.
rBMSCs were isolated from 4 week old SD rats.Separated and flushed the femurs and tibiae as methods above.Then, the bone marrows were harvested and centrifuged, followed by seeding into the cell culture dishes with α-MEM containing 10% v/v FBS and 1% penicillin-streptomycin.The rBMSCs were cultured to approximately 80% of confluence; cells of passage 3-5 were used in the following experiments.The rBMSCs were then characterized by alizarin red staining, ALP staining, and oil red O staining after osteogenic or adipogenic differentiation according to the previous methods.
In addition, Raw 264.7 and L929 cell line was received from Guangdong Provincial Key Laboratory of Stomatology and cultured in RPMI 1640 medium (Gibco, USA) containing 10% v/v FBS and 1% penicillinstreptomycin.
CCK-8 Assay: Influences of pGM/cPL and pGM/cPL@NI hydrogel on cell viability were determined by CCK-8 assays.Briefly, hydrogel extracts in α-DMEM complete medium were prepared according to ISO Standard 10993-12. [53]rBMSCs were cultured in 96-well plates with a density of 5 Â 10 3 cells in each well; the cells were treated with either GelMA, pGM/cPL, or pGM/cPL@NI extracts for consecutive 7 d after adherence.CCK-8 reagents (10 μL) were added and incubated for 2 h at certain time points, followed by absorbance measurement at 450 nm using a microplate reader.
Live/dead Staining: The rBMSCs (1 Â 10 5 mL À1 ) was cocultured with GelMA, pGM/cPL, or pGM/cPL@NI hydrogels extracts for 7 d.Calcein AM (2 μM) and PI (8 μM) were thoroughly mixed to prepare the dyeing solution, which labeled live (green) and dead (red) cells, respectively.After dyeing for 20 min at room temperature, the images were instantly collected by a FV3000 laser scanning confocal microscope (Olympus, Japan).
Cell Scratch Assay: L929 cells were cultured in 24-well plates (2 Â 10 4 cells per well), and monolayer cells were scratched with a 200 μL pipette tip to simulate the scratch.Then, cell debris was removed by PBS, L929 cells were treated with various hydrogel extracts and cultured at 37 °C for 36 h.Cell migration at different time points (6, 24, 36 h) was observed and photographed by a phase-contrast microscope (Zeiss, Germany).
Hemolysis Assay: Briefly, fresh blood of SD rats was collected in EDTA anticoagulant tube.Then, the collected blood was centrifuged at 3000 rpm for 15 min to deposit the hemocytes.200 μL of pGM/cPL or pGM/ cPL@NI was cocultured with hemocytes suspension at 37 °C for 4 h; PBS and Triton X-100 were set as negative and positive control, respectively.Samples were centrifuged after coculture; the absorbance of supernatants was analyzed by a microplate reader at 540 nm.The hemolysis ratio was calculated by the following equation: Hemolysis ratio ð%Þ ¼ OD hydrogels ÀOD PBS OD Triton À OD PBS Â 100%.Antibacterial Activity: P. gingivalis and E. coli were cultured according to the previous methods. [47,54]After 24 h or overnight culture of P. gingivalis and E. coli, respectively, bacterial suspensions were diluted into OD 600 = 0.05; 200 μL of pGM/cPL or pGM/cPL@NI was added into suspension; 200 μL of PBS was set as a negative control.200 μL of bacterial suspension (n = 3) was pipetted into 96-well plates at different timepoint (0, 2, 4, 6, 8, 10, 12 h) to measure growth curve and suspension absorbance in Gen5 software.For observing colony formation, bacterial suspensions were diluted to 10 8 CFU mL À1 ; bacterial suspensions (1 mL) were coincubated with 200 μL of PBS, pGM/cPL, or pGM/cPL@NI for 24 h (P.gingivalis) or overnight (E.coli).Then, 100 μL of bacterial suspension in each group was diluted and spread onto plates for colony counting.After coincubation with hydrogels, P. gingivalis and E. coli were centrifuged (3000 rpm, 5 min) and washed with 0.85% NaCl twice, then a live/dead Baclight bacterial viability kit (Invitrogen, USA) was used to detect bacterial viability according to the manufacturer's specification, and then images were instantly collected by a FV3000 confocal laser scanning microscope (Olympus, Japan).
Establishment of Bacteria-Cell Co-Culture Model: To investigate the effect of P. gingivalis infection on macrophage viability and NLRP3 inflammasome activation, we established the coculture model of P. gingivalis and BMDMs as previously described. [55]After adherence to BMDMs, P. gingivalis in the logarithmic phase was cocultured with BMDMs at different MOI (0, 50, 100, 200).BMDMs were washed with PBS and fixed by 4% paraformaldehyde (20 min, RT) after 4 h, and then cell morphology was observed by staining with crystal violet.Cell viability after infection was determined by CCK-8 assay as described above.
Flow Cytometry: Treated cells were digested by 0.25% trypsin and formed single-cell suspensions.Before incubation of targeted antibodies, Fc receptors on the cells surface were blocked by anti-CD16/CD32 antibody (MultiSciences Biotech, China) for 30 min at room temperature, and then FITC anti-mouse F4/80 (BioLegend, USA) and APC anti-mouse CD11b (BioLegend, USA) antibodies were incubated with cells for 25-30 min on ice in the dark.For intracellular antigen detection, cells were sequentially fixed by Fixation Buffer (BioLegend, USA) and permeabilized by Intracellular Staining Permeabilization Wash Buffer (BioLegend, USA) according to the manufacturer's instruction, and then PE anti-mouse CD206 antibody (BioLegend, USA) was incubated with cells for 25-30 min on ice in the dark.Cells were washed 3 times before being detected by a flow cytometer (Cytoflex, Beckman, USA).
Osteogenic Induction of rBMSCs: The osteogenic differentiation of rBMSCs was induced according to the previous methods.The ALP staining and ARS staining were used to evaluate the degree of matrix mineralization by NBT/BCIP staining kit (Beyotime, China) and alizarin red (Cyagen, China).The semiquantification of ARS was completed with 10% cetylpyridinium chloride (Sigma-Aldrich, USA) and analyzed by microplate reader.
ELISA: To determine the effect of P. gingivalis infection on IL-1β secretion in BMDMs, briefly, cell supernatants were collected and centrifuged at 3500 rpm for 15 min to remove cell debris or bacteria.Then, the protein expression was measured by Mouse IL-1 beta Uncoated ELISA kit (Invitrogen, USA) according to the manufacturer's instructions.To detect the effects of hydrogels on the expression of proinflammatory cytokines IL-1β, IL-18, and TNF-α, ELISA was used according to the above methods.As the bactericidal effects of pGM/cPL and pGM/cPL@NI hydrogels may cause the differences in bacteria count, BMDMs were costimulated with P. gingivalis-LPS (1 μg mL À1 ) and Nigericin (10 μM) instead of P. gingivalis infection.
LDH Release Detection: LDH activity in the cell supernatants was measured using LDH Cytotoxicity Assay Kit (Beyotime, China) according to the manufacturer's instruction.LDH release (%) was calculated by the following equation: LDH release ð%Þ ¼ LDH sample ÀLDH unstimulated LDH maximum À LDH unstimulated Â 100% qRT-PCR: Total RNA was extracted by TRIzol (Invitrogen, USA) after induction and then quantified with NanoDrop 2000 spectrophotometer (Thermo Scientific, USA).PrimeScript RT reagent Kit (TaKaRa Biotechnology, Japan) and TB Green Premix Ex Taq (TaKaRa Biotechnology, Japan) were used for reverse transcription and qRT-PCR according to the manufacturer's instruction.The primer sequences of Gapdh, Nlrp3, Caspase1, Gsdmd, Runx2, Ocn, Alp, and Osx were presented in the Supporting Information.
Western Blots: Total protein was extracted by radioimmunoprecipitation assay buffer (Elabscience, China) on ice for 20 min, followed by ultrasonic lysis for 30 s per sample.The lysates were centrifuged at 14000 rpm for 15 min, and then the supernatant was extracted and quantified by a BCA protein assay kit (Elabscience, China).Sodium dodecyl sulfatepolyacrylamide gel electrophoresis was performed to separate proteins with different molecular weights, and then the gels were transferred onto polyvinylidene fluoride membranes (Millipore, MA, USA).Membranes were blocked with 5% nonfat milk (Elabscience, China) for 1 h at room temperature and subsequently incubated with primary antibodies for pro-IL-1β (1:1000, Abcam, UK), NLRP3 (1:1000, Abcam, UK) or iNOS (1:1000, Abcam, UK) overnight at 4 °C.Membranes were washed with TBST (3 Â 15 min), and then incubated with a secondary antibody (HRP-conjugated AffiniPure Goat Anti-Rabbit IgG, Proteintech, China) for 2 h at room temperature.Protein images were collected on a gel imager system (ChemiDoc, Bio-rad, USA) and semiquantified with ImageJ software.
Detection of ROS and Mitochondrial Membrane Potential: ROS generation and changes in mitochondrial membrane potential were investigated by DCFH-DA and JC-1 staining, respectively.The staining process was implemented according to the manufacturer's instruction.For further confocal microscopy or flow cytometry, DCFH-DA (10 μM, Aladdin, China) was incubated with adherent Raw 264.7 or cell suspension for 20 min at 37 °C in the dark.JC-1 (Beyotime, China) working solution was used to combine high (red) or low (green) mitochondrial membrane potential by staining for 30 min at 37 °C in the dark.Images of ROS and JC-1 labeled cells were collected using a LSM780 laser scanning confocal microscope (Zeiss, Germany).
Animals: 4 week old SD rats were purchased from the Laboratory Animal Center of Sun Yat-sen University.All animal experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of Sun Yat-sen University (approval no.SYSU-IACUC-2022-000401).
For investigating the in vivo biocompatibility of hydrogels, subcutaneous implantation models were established according to previous methods.Briefly, depilation of the dorsal skin of the rats was completed after intraperitoneal anesthetization.Dorsal incisions on the double side were carefully made to isolate subcutaneous tissue, and then pGM/cPL or pGM/cPL@NI hydrogel was implanted.Surrounding skin tissues were collected for hematoxylin & eosin (HE) staining after 7 d of implantation.Meanwhile, the whole blood of SD rats was extracted after 7 d for biochemical analysis.
Then, we established the rat periodontitis models according to our previous study. [52]Briefly, SD rats were fully anesthetized with 10% chloral hydrate (0.35 mL per 100 g), 15 # -40 # nickel-titanium root canal files, and silk and stainless-steel ligature wires were used to induce periodontitis.Subsequently, P. gingivalis (20 μL, 1.0 Â 10 8 CFU mL À1 ) was injected subgingivally every 2 d around bilateral second molars for 4 times.The presence conditions of ligature wires were checked every 2 d to confirm the validity of periodontitis models.After 4 week induction, silk and stainless-steel ligature wires were removed.pGM/cPL (30 μL) or pGM/ cPL@NI (30 μL) were injected into periodontal pockets and irradiated to form complete hydrogels in situ every 7 d.
Micro-CT Analysis: After hydrogel treatments for 4 weeks, SD rats were euthanized, the maxillae were collected and fixed in 4% PFA for 24 h at 4 °C.Fixed maxillae were then placed in customized sample containers and scanned with high-resolution micro-CT (Scanco Medical AG, Switzerland).The scanning parameters were set at 9 μm, 70 kV, and 114 mA.3D models of samples were reconstructed in specific analysis software (VGStudio MAX 1.2.1, Germany).The distance between CEJ and ABC and BV/TV was measured.
Histological Analysis: Before histological analysis, the collected rat maxillae were fixed with 4% paraformaldehyde solution and decalcified by 10% EDTA (pH 8.0) after 3 weeks.Subsequently, the tissues were embedded and sliced into 5 mm thicknesses with a freezing microtome (Leica, Germany).H&E (Sigma-Aldrich, USA) and Masson-trichrome staining (Servicebio, China) were performed according to the manufacturer's instructions.
Immunofluorescent staining was implemented to observe periodontal osteogenesis or inflammation-related protein expression.The maxillae were fixed, decalcified, embedded, and sliced according to the previous methods.The sections were blocked with goat serum for 30 min at room temperature, followed by sequential incubation of primary antibodies for IL-1β (1:800, Servicebio, China), Runx2 (1:200, Proteintech, China), OCN (1:200, Proteintech, China), and secondary antibodies.Finally, the sections were counterstained with DAPI (Beyotime, China) for 5 min at room temperature.Stained sections were observed and analyzed using inverted fluorescent microscope (Zeiss, Germany).
Statistical Analysis: All statistical analyses in the present study are performed using GraphPad Prism 9.0 (GraphPad Software, USA).All experiments were replicated at least 3 times, by expressing with mean AE standard deviation (SD) after t-tests and one-way ANOVA analyses.Tukey's post hoc test was used for intergroup multiple comparison.The significance level was set at 0.05.*, **, and *** represent p < 0.05, p < 0.01, and p < 0.001, respectively.
(grant no.KF2022120102), and the Innovative Research Fund for the Teacher of Higher Education of China (grant no.2021XCL03).
Open Access funding enabled and organized by Projekt DEAL.

Figure 1 .
Figure 1.Schematic illustration showing the mechanism of multifunctional dual-crosslinked pGM/cPL@NI hydrogel promotes periodontal bone regeneration via immunomodulation.A) Synthesis diagram of dual-crosslinked hydrogel pGM/cPL@NI with NLRP3 inhibitor MCC950 loaded.MCC950 can be ROS or pH-responsive released due to the dynamic borate ester bond formed by PBA and catechol.B) Based on the complexed process of periodontitis, pGM/cPL@NI performed favorable therapeutic effects by exerting versatile functions including antibacterial, antioxidative, macrophage immunomodulation, and osteogenesis promotion.Focusing on the vital modulation of macrophage during inflammation and regeneration, pGM/cPL@NI ameliorates bacteria-induced macrophage abnormalities through a) NLRP3 inflammasome-targeted inhibition that achieved by NLRP3 inhibitor MCC950 and b) promotion of anti-inflammatory polarization (M2 polarization).

Figure 2 .
Figure 2. Synthesis and characterization of hydrogels.A) Synthesis procedure of hydrogels.GelMA-PBA and ε-PL-cat were first crosslinked by the dynamic borate ester bonds; photo-crosslinking after irradiation subsequently formed the second network of pGM/cPL and MCC950-loaded pGM/cPL@NI.B) Stimuli-responsive mechanism of pGM/cPL and pGM/cPL@NI based on cleavage of the borate ester bond.C) Photographs of synthetic process of dual-crosslinked pGM/cPL and pGM/cPL@NI hydrogels.D) Photographs of hydrogels' gelation behavior in the presence of H 2 O 2 .E,F) FTIR and 1 H NMR spectra of pGM/cPL.G) Photographs showing the injectability of hydrogel.H) Rheological tests of pGM/cPL before and after irradiation.I) Compression tests of hydrogels with different weight/volume (w/v) concentration.J) Degradation behavior of pGM/cPL under different pH (pH 7.4, or 5.5) and H 2 O 2 (H 2 O 2 = 0, 0.1, or 1 mм) environments compared to the GelMA.K) Cumulative release of MCC950 from pGM/cPL@NI under different pH (pH 7.4, or 5.5) and H 2 O 2 (H 2 O 2 = 0, 0.1, or 1 mM) environments.

Figure 3 .
Figure 3. Biocompatibility and antibacterial properties of hydrogels in vitro.A-C) live/dead staining, CCK-8 assay, and scratch assay were performed to observe cell viability and migration after hydrogels treatment (scale bar, 100 μm).D) Schematic diagram of antimicrobial mechanism of pGM/cPL and pGM/cPL@NI hydrogels.E) Absorbance of bacterial suspension after coculture with hydrogels.F,G) Colonies formation and relative viability of P. gingivalis and E. coli compared to control groups.H) Bacterial live/dead staining for bacterial viability detection (scale bar, 10 μm).I) Bacterial morphological changes observed by SEM (scale bar, 1 μm).Data are presented as mean AE SD. *p < 0.05, **p < 0.01, ***p < 0.001; ns means no statistical significance.

Figure 4 .
Figure 4. pGM/cPL@NI inhibited the NLRP3 inflammasome activation by releasing MCC950 in vitro.A) Schematic diagram showing the macrophage immunomodulation mechanism of MCC950-loaded pGM/cPL@NI.B) P. gingivalis-induced IL-1β secretion in an MOI or time-dependent manner.C) BMDMs exhibited increased NLRP3 and pro-IL-1β protein expression after co-culture with P. gingivalis.D) The expression of IL-1β was significantly enhanced by combined stimulation of P. gingivalis-LPS and Nigericin, which is inhibited by MCC950 in a concentration-dependent manner.E) LDH release assay.F,G) Expression of NLRP3 was effectively downregulated by pGM/cPL@NI.H) Detection of NLRP3 inflammasome-related genes (Nlrp3, Caspase1, Gsdmd) expression by qRT-PCR with or without hydrogels intervention.I) Extracellular expression of proinflammatory cytokines.Data are presented as mean AE SD. *p < 0.05, **p < 0.01, ***p < 0.001; ns means no statistical significance.

Figure 6 .
Figure 6.pGM/cPL and pGM/cPL@NI directly and indirectly facilitated the osteogenic differentiation of rBMSCs in vitro.A) Schematic diagram of various interventions to detect the effects of hydrogels on rBMSCs.B) ALP and ARS staining of rBMSCs after 7 day and 21 day osteogenic induction.C) mRNA expressions of osteogenesis-related genes in rBMSCs were detected by qRT-PCR (n= 3, mean AE SD).D) Protein expressions of Runx2 (green) and OCN (green) were detected by immunofluorescent staining, costaining with phalloidin (red) and DAPI (blue).Scale bar, 20 μm.E) Semiquantification of mean fluorescence intensity (MFI, n = 3, mean AE SD); F) Representative images of ARS staining of rBMSCs after culture with different Mφ-CM.G) mRNA expressions detection of Runx2 and Ocn in rBMSCs after cultured with different Mφ-CM by qRT-PCR (n = 3, mean AE SD).Data are presented as mean AE SD. *p < 0.05, **p < 0.01, ***p < 0.001; ns means no statistical significance.