Fast Lithium-Releasing Phosphate Glasses Promote the Resolution of Inflammation

Abstract

Delay in the resolution of inflammation has disastrous implications for tissue regeneration, but local anti-inflammatory signals could accelerate this process. To test this, phosphate-based bioactive glasses were fabricated to release lithium, an anti-inflammatory ion, at a range of ionic concentrations. Lithium release was contingent on the aluminum molar ratio within the glass formulation. Glasses that released lithium faster exerted greater anti-inflammatory effects on activated macrophages. These effects resulted from the inhibition of GSK3β activity and the promotion of CD206 expression. This study demonstrates the therapeutic potential of anti-inflammatory phosphate glasses in resolving inflammation in the regenerative environment.

1. Introduction

A delay in inflammation resolution carries major implications for regenerative therapies, leading to repair by scar tissue instead of functional tissue regeneration. , Tissue damage immediately begins to recruit immune cells. Upon arrival at the injury site, immune cells may encounter necrotic cells, tissue debris, and microbial invaders, all of which are local cues eliciting an inflammatory response. Several studies have reported improved regeneration of osteal, vascular, chondrogenic, dermal, and neuronal tissues by promoting the resolution of local inflammation. − However, the translation of these approaches to clinical practice is hindered by the use of recombinant cytokines such as interleukin-4 (IL-4) and interleukin-10 (IL-10). Recombinant cytokines have inherent unpredictability (i.e., cascading effects potentially leading to side effects), a short half-life, molecular instability, and immunogenicity risks (e.g., from sourcing, manufacturing, and quality control processes). , Taken together, these concerns contribute to high production costs, delaying translation to clinical use. Therefore, alternative approaches to modulating local inflammation are necessary.

Bioactive glasses are composed of inorganic ions and can serve as a local source of bioactive ions. The classic formulation, 45S5 Bioglass, is highly effective in promoting bone regeneration due to the initial release of calcium and phosphate. , Notably, substitutions or additions of elements may stimulate additional biological responses. For instance, cobalt-substituted bioactive glasses activate cellular pathways involved in angiogenesis, while strontium substitutions modulate osteoblast and osteoclast activities. , However, silicate-based glasses such as the original 45S5 Bioglass are made of a highly stable glass network resistant to hydrolysis and, thus, offer low solubility. Additionally, silicate-based glasses are capable of forming a hydroxycarbonate apatite layer on the glass surface, impeding further dissolution and ion release. As such, a therapeutic dose of bioactive ions cannot be maintained. Thus, these features render silicate-based glasses less ideal for the local delivery of bioactive ions.

Alternatively, phosphate-based bioactive glasses are entirely soluble and can sustain a linear release of bioactive ions until total dissolution, which can be tuned to a desired rate or biological effect. For instance, Salazar et al. demonstrated that adding different ratios of oxides to the phosphate glass network could slow down or speed up the rate of dissolution. Adding aluminum oxide improved the network stability and chemical durability of phosphate glasses by cross-linking phosphate chains within the network, thereby slowing down dissolution. , This strategy has been successfully applied to tune the dissolution rate of several phosphate glasses, such that tunable phosphate glasses have become a useful degradable biomaterial. ,

Another advantage of phosphate over silicate glasses is the basic phosphate unit comprising the glass network. Phosphorus is a pentavalent element, bonding with four oxygen atoms and forming a double bond with one of these atoms. As a result, nonbridging oxygens are abundant within the glass network and can be exploited to form ionic bonds with bioactive cations. Here, this chemical characteristic is exploited to develop an experimental bioactive glass capable of releasing anti-inflammatory ions.

Among the medically safe bioactive cations, lithium is known for its anti-inflammatory effects. − For instance, lithium ions serve as selective inhibitors of glycogen synthase kinase-3 beta (GSK3β) by replacing the magnesium ion present at the protein-binding site, thereby directly reducing GSK3β stability and activity. , This effect is considered the primary mechanism behind the anti-inflammatory properties of lithium. − Lithium-releasing phosphate glasses were fabricated with increasing ratios of aluminum oxide to generate different dissolution rates. It was hypothesized that lithium-releasing phosphate glasses could provide localized, timely anti-inflammatory ions and that the appropriate concentration of lithium could resolve inflammation. The effects of the glass formulation on solubility and lithium release were characterized, and the anti-inflammatory effect of lithium-releasing glasses was tested in vitro and in vivo using models that recreate acute inflammation.

2. Materials and Methods

To study the release and biological effects of lithium ions (Li⁺) derived from our experimental phosphate-based bioactive glasses, four glass formulations containing Li₂O were prepared. Increasing molar concentrations of Al₂O₃ were used to modify the glass dissolution. Ion release was quantified, and the anti-inflammatory ability of glass particles was tested in vitro and in vivo against inflammatory models.

2.1. Preparation of Bioactive Glass Particles

Four bioactive glasses were prepared by melting 42P₂O₅–16CaO–(42–x)­Li₂O–xAl₂O₃ at different molar ratios, where x = 0, 2, 4, or 6. Table S1 provides the nominal glass compositions (Table S1). The components were melted in a platinum crucible for 1 h at 1100 °C and quenched on a stainless-steel plate. The cullet was ground in an alumina ball mill rotary system (Nitto Kagaku, Nagoya, Japan), and the resulting glass particles were sterilized with ethylene oxide gas and stored in dry cabinets at room temperature until use.

2.2. Characterization of Physical Properties

To determine the size distribution of the ground bioactive glass, particles were suspended in ethyl alcohol. The particle size in the suspension was quantified using a laser scattering particle size distribution analyzer (Partica LA-960 V2; HORIBA, Ltd., Kyoto, Japan).

To observe the particle microstructure and the elemental distribution within that structure, a field emission scanning electron microscope equipped with an energy-dispersive X-ray spectrometer (FE-SEM/EDS) (JSM-F100, JEOL, Tokyo, Japan) was used, and specimens were observed at an accelerating voltage of 15 kV.

To evaluate the amorphous state of glass particles, X-ray diffraction patterns were obtained using a diffractometer (Empyrean, Malvern Panalytical Ltd., Worcestershire, UK), calibrated with a standard pure silicon sample and operating with a CuKα radiation source at a wavelength (λ) of 0.154060 nm, at 45 kV and 40 mA, in the 2θ range from 5° to 80°.

2.3. Ion Release and Weight Loss Fraction

To measure the concentrations of lithium (Li⁺), calcium (Ca²⁺), phosphate ([PO₄]^3–), and aluminum (Al³⁺) ions released from the different glass compositions, eluates were produced. Ten milligrams of glass particles were added to Falcon tubes containing 10 mL of ultrapure water (Direct-Q 3 UV; Merck Millipore, Molsheim, France). The tubes were kept on an orbital shaker under 100 rpm agitation at 37 °C. Every 24 h, the tubes were centrifuged at 2700 × g for 10 min before removing 5 mL of eluate, replacing it with 5 mL of fresh ultrapure water, and then returning them to incubation. Experimental days 1, 3, 5, 7, 10, 14, 21, and 28 were selected to quantify the ionic concentrations and weight loss fraction of the glass particles. On each experimental day, the entire contents of the Falcon tubes were collected and filtered through a 0.22 μm syringe-driven filter. The filtered eluates were analyzed using inductively coupled plasma optical emission spectroscopy (ICP-OES) (iCAP 7000 Duo; Thermo Fisher Scientific, MA, USA), where the presence of Li⁺, Ca²⁺, [PO₄]^3–, and Al³⁺ ions was quantified in relation to a standard calibration curve. To measure the weight loss fraction, the remaining glass particles trapped in the filter were dried at 60 °C for 48 h and weighed.

2.4. Cell Culture and Inflammation

To recreate an inflammatory environment in vitro, the macrophage cell line RAW 264.7 (RIKEN BioResource Center, Tsukuba, Japan) was used. A potent inflammatory response was provoked using the M1 activation protocol for RAW 264.7 cells, as characterized by Orecchioni et al. Macrophages were cultured in Dulbecco’s Modified Eagle Medium (DMEM) (043-30085; FUJIFILM Wako, Osaka, Japan), supplemented with 10% fetal bovine serum (Thermo Fisher Scientific) and 1% penicillin-streptomycin (Nacalai Tesque, Kyoto, Japan). Macrophages were seeded at a density of 1 × 10⁵ cells/well in 12-well plates and allowed to attach for 24 h. Lipopolysaccharides (LPS) from Escherichia coli serotype O55:B5 (L2880; Sigma-Aldrich, MO, USA) were added to the culture medium at a concentration of 100 ng/mL, together with recombinant mouse interferon-gamma (IFNγ) (575302; BioLegend, CA, USA) at a concentration of 20 ng/mL, and incubated for 24 h. Inflamed macrophages subjected to the M1 activation protocol were defined as M1 macrophages.

2.5. Treatment with Bioactive Glass Particles

To test the anti-inflammatory abilities of our experimental bioactive glasses, the M1 macrophages described above were treated with each glass formulation. A transwell system was used to expose the M1 macrophages to Li⁺ released by glass particles, with unexposed macrophages serving as untreated controls. The macrophages were washed with phosphate-buffered saline (PBS), and a Transwell insert (Corning, AZ, USA) with a permeable membrane with a pore size of 0.4 μm was added to each well. For glass treatment conditions, the insert was filled with 1 mL of fresh medium containing glass particles at 1 mg/mL; for untreated controls, no glass particles were added. For both treatment and control conditions, the bottom well was filled with an additional 1 mL of fresh culture medium without glass particles. Only the bottom well medium was replaced every 48 h.

2.6. Expression of Inflammation-Related Genes by RT-qPCR

To quantify the expression of inflammatory and anti-inflammatory markers using macrophages following treatment with each bioactive glass formulation, RT-qPCR was performed. Experimental days 1, 3, and 7 were selected to collect the macrophages. The mRNA expression levels of inflammatory markers nitric oxide synthase 2 (Nos2), CD86 antigen (Cd86), and CD80 antigen (Cd80); the anti-inflammatory markers arginase (Arg1), mannose receptor C type 1 (Mrc1), and early growth response 2 (Egr2); and the housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (Gapdh) were detected using their corresponding primers (Table S2) (TaqMan; Thermo Fisher Scientific) and the Cells-to-Ct Kit (Thermo Fisher Scientific). The fold change in mRNA levels of Nos2, Cd86, Cd80, Arg1, Mrc1, and Egr2 relative to Gapdh was calculated using the 2^–ΔΔCt method, normalized to the control condition on day 1.

2.7. Immunofluorescence Staining

To observe macrophages expressing CD206 following treatment with each bioactive glass formulation, immunofluorescence staining was conducted. The macrophages were fixed with 4% paraformaldehyde for 10 min, permeabilized with 0.1% Triton X-100 for 10 min, and blocked with 0.5% bovine serum albumin for 1 h at room temperature. Cells were labeled with CD206 (E6T5J; Cell Signaling Technology, MA, USA) rabbit antimouse monoclonal antibody at a 1:200 dilution ratio and incubated at 4 °C for 18 h, followed by labeling with fluorescently conjugated goat antirabbit monoclonal antibody (Alexa Fluor 488; Thermo Fisher Scientific, MO, USA) at a 1:2000 dilution ratio for 1 h at room temperature. Cell nuclei were stained with Hoechst 33342 (Thermo Fisher Scientific), and F-actin was stained with rhodamine phalloidin (Thermo Fisher Scientific).

2.8. Secretion of Inflammatory Mediators

M1 macrophages are involved in a range of paracrine signals that exacerbate the inflammatory response. To assess whether glass-treated macrophages had lost this ability to intensify inflammation, the secretion of inflammatory mediators was investigated on day 7. Cell culture supernatants were collected and centrifuged at 1500 × g for 10 min at 4 °C to remove cell debris. Supernatants were immediately used in ELISA assays (Mouse M1/M2 Cytokines Multiplex ELISA Kit; Arigo Biolaboratories, Hsinchu, Taiwan) to determine the expression of interleukin-6 (IL-6), tumor necrosis factor α (TNFα), IL-4, and IL-10.

A Griess reaction system (Promega, WI, USA) was used to determine the levels of nitrite, a stable breakdown product of nitric oxide.

2.9. Phagocytosis of Synthetic Targets

Because macrophages engaged in phagocytosis secrete large amounts of inflammatory cytokines, phagocytic activity was quantified as an indicator of the macrophage inflammatory state. Latex beads (LB30; Sigma-Aldrich) were used as synthetic targets for phagocytosis. Measuring 3 μm in diameter, the beads were prepared according to the methodology proposed by Botelho. For a detailed description of this method, see Supporting Information: Phagocytosis of Synthetic Targets.

To observe the phagocytosed beads, macrophages were processed for fluorescence microscopy. Briefly, macrophages were fixed with 4% paraformaldehyde for 10 min, the cell nuclei were stained with Hoechst 33342, and F-actin was stained with rhodamine phalloidin. A confocal laser microscope (TCS SP8; Leica, Wetzlar, Germany) was used to acquire images of the cell nuclei in the blue channel, of F-actin in the green channel, and of beads in the differential interference contrast (DIC) channel. DIC uses orthogonally polarized light to increase the contrast in transparent samples, such as macrophages cultured in vitro. The high contrast enables a clear distinction between successfully phagocytosed beads and externally bound beads that were not phagocytosed. , At least 100 cells were counted for each experimental condition. The phagocytic index (1) and the phagocytic efficiency (2) were determined according to the following formulas:

$$Allphagocytosedbeads/allcells$$ 1

$$(Cellscontainingatleast1bead/allcells)\times 100$$ 2

2.10. Western Blot

Possible cascades explaining the anti-inflammatory effects of our experimental glasses were investigated. Macrophages were provoked by the M1 activation protocol and treated with glass particles, as described above. After 6 h of treatment, the phosphorylation states of GSK3β, signal transducer and activator of transcription 1 (STAT1) and signal transducer and activator of transcription 3 (STAT3) were assessed using Western blot. The following rabbit antimouse primary antibodies were used at a 1:1000 dilution ratio: GSK3β, phospho-GSK3β Ser9, phospho-STAT1 Tyr701, and phospho-STAT3 Tyr705 (Cell Signaling Technology). HRP-conjugated anti-β-actin (Proteintech, Opelstr, Germany) was used as the loading control. Goat antirabbit IgG HRP-conjugated antibody (Cell Signaling Technology) was used as the secondary antibody. Bands of interest were detected via enhanced chemiluminescence (SuperSignal, Thermo Fisher) using the iBright Imaging System (Thermo Fisher Scientific), and band peak intensity was measured using ImageJ. For a detailed description of the Western blot method, see Supporting Information: Western blot.

2.11. Anti-Inflammatory Effect In Vivo

To test the anti-inflammatory ability of our experimental glass particles in vivo, a mouse model of a full-thickness skin injury was used. This model promotes the accumulation of granulation tissue rich in inflammatory macrophages. All protocols for animal experiments were approved by the Institutional Animal Care and Use Committee of the Osaka University Graduate School of Dentistry (Approval No. R-04-006-0).

Fifteen nine-week-old male C57BL/6 mice (CLEA, Tokyo, Japan) were used in this study. The animals were anesthetized with a mixture of 0.3 mg of medetomidine (Domitor; Nippon Zenyaku Kogyo, Tokyo, Japan), 4 mg of midazolam (Dormicum; Maruishi Pharmaceutical, Osaka, Japan), and 5 mg of butorphanol (Vetorphale; Meiji Seika Pharma, Tokyo, Japan) per kilogram of body weight, administered intraperitoneally. A full-thickness excisional wound was made on the dorsal skin, extending through the panniculus carnosus, using a 4 μm-diameter biopsy punch. A donut-shaped silicone splint was placed around the wound and fixed with silk sutures to prevent closure by contraction and to promote the accumulation of granulation tissue. An injectable peptide hydrogel (PuraMatrix, Corning) was used as a carrier for bioactive glass particles. 10 mg of powdered glasses was combined with 0.5 mL of hydrogel via vortexing for 2 min. Wounds were treated with 150 μL of bioactive glass particles embedded in hydrogel, and the carrier alone was used as a treatment control.

Wounds and surrounding tissues were harvested on day 7 following treatment. Mice were anesthetized as described above and euthanized by cervical dislocation. Samples were fixed in 10% buffered formalin (FUJIFILM Wako) for 24 h, embedded in paraffin, and cut into 5 μm-thick sections. Sections were stained with hematoxylin and eosin (HE), Masson-Goldner staining, and chromogenic immunohistochemistry (IHC) staining for the detection of CD206⁺ cells. For a detailed description, see Supporting Information: Immunohistochemistry; Wound Area.

2.12. Statistical Analysis

All statistical analyses were carried out using GraphPad Prism (GraphPad Software; Boston, USA). Analysis of variance (ANOVA) was conducted, and comparisons using Tukey’s multiple comparisons test or a two-tailed Student’s t-test were performed where appropriate. A p < 0.05 was defined as statistically significant.

3. Results

3.1. Li⁺ Release Gradually Delayed by Increasing Al₂O₃ Content

Ion release and glass dissolution were inversely related to the Al₂O₃ content in the glass formulation. The four fabricated Li⁺-releasing glasses were labeled Li42, Li40, Li38, and Li36, reflecting their decreasing molar concentrations of Li₂O (Figure S1A and Table S1). Glass particles had a mean size of 13 μm (Figure B) and an amorphous network structure (Figure C), confirming the absence of a crystalline phase. FE-SEM/EDS images showed an even elemental distribution of aluminum, phosphorus, and calcium within the particle structure (Figure D). X-rays emitted by lithium have low energy levels; therefore, it was not possible to detect lithium using EDS. However, since phosphates provide nonbridging oxygens for ionic binding with lithium, it can be inferred that lithium’s distribution follows that of phosphorus.

1.

Characterization of Li⁺-releasing glasses. (A) Representative images of the four Li⁺-releasing glasses. (B) Size distribution of glass particles. (C) X-ray diffraction patterns showing the absence of a crystalline phase. (D) Scanning electron microscopy (SEM) and element mapping of particles, showing the element distribution of aluminum (Al), phosphorus (P), and calcium (Ca). Scale bar: 10 μm.

Nonetheless, the presence of Li⁺ was confirmed by ICP-OES measurements, which tracked the concentration of Li⁺ in the glass eluates over time (Figure ). Li42 and Li40 reached peak release on day 1, at concentrations of 337.6 μg/mL ± 3.2 and 247.6 μg/mL ± 4.9 μg/mL, respectively (Figure A). Li38 reached a peak release of Li⁺ on day 3 (138.9 μg/mL ± 7.1 μg/mL) and Li36 on day 5 for (90 μg/mL ± 1.9 μg/mL) (Figure A). The release of calcium ions (Ca²⁺) and phosphate ([PO₄]^3–) peaked on day 1 for all glasses except Li36, which contained the highest molar percentage (mol %) of Al₂O₃, delaying peak release until day 5 (Figure B,C). The addition of Al₂O₃ to the glass formulations also delayed the glass solubility; on day 28 of water immersion, Li36 still maintained approximately 2% of its original weight (Figure E). However, Li38 and Li40, with a smaller mol % of Al₂O₃, were completely solubilized by day 7. Li42, which did not contain Al₂O₃ (Figure D), retained less than 5% of its weight at the end of day 1 (Figure E).

2.

Ion release profiles of Li⁺-releasing glasses in water. (A) Lithium (Li⁺), (B) calcium (Ca²⁺), (C) phosphate ([PO₄]^3–), and (D) aluminum (Al³⁺). (E) Weight loss fraction over time.

3.2. Fast Li⁺ Release Reverses Macrophage Inflammatory Phenotype

Among the four different ion release profiles from each glass formulation, a faster release of Li⁺ rather than a delayed release reduced the Cd80 mRNA levels and increased CD206 protein levels. Macrophages activated with LPS and IFNγ were treated with Li⁺-releasing glasses for up to 7 days (Figure A). On day 1, a significant reduction in the expression of inflammatory markers Nos2 and Cd80 was observed compared with the control condition without glass particles (Figure B). The expression levels of Cd86, a hallmark of macrophage inflammatory activation, were significantly lower on days 3 and 7 (Figure B). Li⁺-releasing glasses also stimulated the upregulation of Arg1 and Mrc1, both markers of anti-inflammatory macrophages (Figure C). The Mrc1 protein product CD206 was further investigated via immunofluorescence staining. Macrophages treated with Li⁺-releasing glasses exhibited an increased ratio of CD206⁺ cells relative to the control in a dose-dependent manner (Figure D,E).

3.

Fast Li⁺ release reverses macrophage inflammatory phenotype. (A) Schematic diagram of the M1 activation and Li⁺-releasing glass treatment protocols. (B) qRT-PCR analysis of mRNA expression levels for inflammatory markers and (C) for anti-inflammation markers (ANOVA, Tukey’s multiple comparisons, mean ± sd, n = 5, *p < 0.05). (D) Immunofluorescence staining of CD206⁺ cells following treatment with Li⁺-releasing glasses and (E) semiquantitative analysis of the ratio of CD206⁺ cells (ANOVA, Tukey’s multiple comparisons, data shown in mean with 95% confidence intervals; bars represent range from min to max values, n = 15, *p < 0.05). Scale bar: 25 μm.

We next attempted to identify the specific Li⁺ concentration in the cell culture media at each time point. ICP-OES quantification of ions released into DMEM was conducted using the same Transwell system to replicate the conditions of cell culture, including media changes every 48 h but without adding any cells. For Li42, Li⁺ release began within the first 30 min of immersion in DMEM (1.8 μg/mL ± 0.2 μg/mL) (Figure S1). The Li⁺ concentration peaked on day 1 for all glasses, with Li42 releasing the highest amount of Li⁺ (33.3 ± 0.9 μg/mL).

Additionally, the biocompatibility of glass particles was assessed without inflammatory stimuli (Figure S2). All glass formulations are biocompatible, showing no cytotoxicity to macrophages or to osteoblast-like cells (MC3T3-E1) (Figure S2A,B). Both cell types were able to proliferate in the presence of glass particles (Figure S2A,B).

3.3. Factors Exacerbating the Inflammatory Response Curbed by Li⁺-Releasing Glasses

Paracrine signaling, which allows macrophages to recruit other inflammatory cells and exacerbate inflammation, was restricted by Li⁺-releasing glasses. Treatment with Li42 and Li40 significantly reduced IL-6 secretion, a major inflammatory mediator (Figure A). Conversely, the levels of the anti-inflammatory cytokine IL-4 were significantly increased by all Li⁺-releasing glasses (Figure A).

4.

Factors exacerbating the inflammatory response curbed by Li⁺-releasing glasses. (A) Fold change in cytokine levels present in the supernatant on day 7 following treatment with Li⁺-releasing glasses (n = 3). (B) Concentration of nitrite following treatment with Li⁺-releasing glasses (n = 6). (C) Differential interference contrast and fluorescence staining of macrophages engaged in phagocytosis. White arrows indicate external beads, and black arrows indicate phagocytosed beads (scale bar: 25 μm). (D) Corresponding semiquantitative analysis of the phagocytic activity (n = 5) (ANOVA, Tukey’s multiple comparisons, mean ± sd, *p < 0.05).

Nitric oxide is produced by inflammatory macrophages and promotes vasodilation, allowing immune cell migration to the site of inflammation. The presence of nitric oxide in the supernatant was measured by its more stable breakdown product, nitrite. Treatment with all glasses significantly reduced the total amounts of nitrite until day 3 (Figure B).

Inflammatory macrophages are particularly efficient at phagocytosis of pathogens; when engaged in phagocytosis, they secrete large amounts of inflammatory paracrine signals. Therefore, macrophage phagocytic activity was evaluated (Figure C). Both the phagocytic index and phagocytic efficiency were significantly reduced in macrophages treated with Li⁺-releasing glasses (Figure D).

3.4. Li⁺-Releasing Glasses Inhibit GSK3β Activity

Li⁺ is a GSK3β inhibitor, and our experimental Li⁺-releasing glasses retained this ability. Active GSK3β is phosphorylated at serine 9, becoming inactive (p-Ser9-GSK3β) and unable to perform catalytic functions. Therefore, the phosphorylation state of GSK3β was investigated. Western blots of proteins extracted from macrophages treated with the different Li⁺-releasing glasses for 6 h showed no change in the expression of total GSK3β (Figure A). However, Li42 and Li40 significantly increased the relative expression of p-Ser9-GSK3β (Figure B) compared with the control condition and the ratio of p-Ser9-GSK3β to total GSK3β (Figure C), indicating GSK3β inactivation.

5.

Li+-releasing glasses inhibit GSK3β activity. (A) Protein blots derived from macrophages following 6 h of treatment with Li+-releasing glasses. (B) Relative expression of p-Ser9-GSK3β and (C) GSK3β phosphorylation expressed as a ratio of total GSK3. Relative expression of STAT1 and STAT3, determined by their active form (D) p-STAT1 and (E) p-STAT3 (ANOVA, two-tailed Student’s t-test, mean ± sd, n = 5, *p < 0.05). (F) Schematic illustration of the anti-inflammatory effects of Li+-releasing glasses via GSK3β/STAT1/3 signaling.

Li42 and Li40 showed a fast Li⁺ release, leading to a swift inactivation of GSK3β within 6 h. Therefore, we investigated the downstream effects on STAT1 and STAT3 transcription factors, which exhibit a quick turnover rate and are essential for the expression of inflammation-related genes. STAT1 and STAT3 were significantly downregulated by Li42 and Li40 (Figure D,E). These results suggest a pathway for the anti-inflammatory effects of Li42 and Li40, initiating with GSK3β inactivation and subsequent interruption of STAT1/3 signaling (Figure F).

3.5. CD206 Expression Increased in Tissue Macrophages

Splinted skin injuries spontaneously build up granulation tissue; however, when treated with Li42 and Li40, wounds exhibited less granulation tissue and more advanced re-epithelization in comparison to treatment with the carrier only (Figure A). Masson–Goldner staining revealed that collagen fibers in wounds treated with the carrier only were deposited in parallel bundles, indicative of scarring (Figure C). By contrast, a weave-like organization of collagen fibers was observed in wounds treated with Li⁺-releasing glasses, indicative of tissue repair (Figure B). Finally, a significant increase in CD206 expression (Figure C,D) and a reduction in the wound area (Figure E) were observed for Li42- and Li40-treated wounds in comparison to carrier-only treatment.

6.

In vivo anti-inflammatory ability of Li⁺-releasing glasses. (A) HE staining of skin injuries 7 days after treatment with Li⁺-releasing glasses. Dashed lines indicate granulation tissue area, and black arrows indicate re-epithelization (scale bar: 1 mm). (B) Masson–Goldner staining of the collagen fibers. Dashed lines indicate injury limits; black arrows indicate collagen deposited in parallel bundles; and pound signs indicate weave-like collagen fibers within the granulation tissue (scale bar: 100 μm). (C) Immunohistochemistry for CD206. Black arrows indicate CD206⁺ cells (scale bar: 100 μm). (D) Corresponding semiquantitative analysis of CD206⁺ area. (E) Corresponding semiquantitative analysis of HE stained wound area (ANOVA, two-tailed Student’s t-test, mean ± sd, n = 3, *p < 0.05).

4. Discussion

Inflammation is an essential step preceding regeneration; however, a delay in resolving inflammation hinders the regenerative process. Inflammation is built up by cascading events and complementary mechanisms that recruit additional inflammatory cells and exacerbate the response. Given this physiology, it is not surprising that early treatment with anti-inflammatory signals reduces inflammation. Here, we found that fast Li⁺ release, rather than a delayed release, exerted significant anti-inflammatory effects in vitro and in vivo.

An increase in mol % of Al₂O₃ reduced glass solubility (Figure D,E) and delayed Li⁺ release (Figure A and S1). A comparable effect has been reported by Moreau et al. using the system 50Li₂O–xAl₂O₃–(50–x)­P₂O₅, where an increasing Al₂O₃ mol % generated stronger bonds between phosphate chains. The effect of Al₂O₃ on glass solubility is explained by Al–O forming stronger bonds than P–O, with Al–O–P bonds being stronger than P–O–P bonds. − The energy necessary to break the glass network is therefore increased, creating a more interconnected and stable structure. However, Al₂O₃ above 5 mol %, as Moreau et al. reported, initiated crystallization, such that the material was no longer a glass. , In our study, Li36 contained the highest mol % of Al₂O₃ (6 mol %), and X-ray diffraction patterns indicated that crystallization did not occur in our glass system (Figure C). Our observation can be explained by the presence of CaO in our system, by a higher mol % of phosphates, and by the higher melting temperature applied here: all factors reported to influence glass crystallization. −

The glass formulation with the fastest release of Li⁺ (Li42) lacked Al₂O₃ and dissolved almost completely by day 1 (Figure E). Zhang et al. produced a similarly fast-releasing phosphate glass in the system 45P₂O₅–55Li₂O but reported high toxicity due to an initial burst of phosphate ions. Here, toxicity was not observed (Figure S2), which is possibly explained by the differences in the maximum concentrations measured. In this study, a maximum of ∼80 μg/mL of phosphate was measured by day 1 (Figure S1), while Zhang et al. reported 500–3000 μg/mL.

In vitro, the toxic dose of Li⁺ reportedly lies above 100 mM (∼700 μg/mL), which is well over the maximum concentration observed in our cell culture experiments (33 μg/mL for Li42; Figure S1). Conversely, the therapeutic dose of Li⁺ ranges from 0.5 to 20 mM (3–138 μg/mL). In mechanistic studies, 10–20 mM (69–138 μg/mL) of Li⁺ has been reported as anti-inflammatory. , However, in a seminal pharmacological study, Klein and Melton observed that Li⁺ reduced the catalytic activity of GSK3β at 2 mM (13.8 μg/mL). Indeed, activation of the Wnt/β-catenin pathway was reported to promote osteogenesis at 5 mM (34 μg/mL) and cementogenesis at 2.3 mM (16 μg/mL). , The observations in this study and the literature indicate that the therapeutic dose of Li⁺ should vary according to the intended application. Therefore, instead of a single optimal dose, there is a range of useful Li⁺ dosages. In this study, 7.5 μg/mL (Li40) was the lowest Li⁺ dose to achieve an anti-inflammatory effect with statistical significance.

Several reports have described divergent cellular responses to Li⁺ treatment across different cell types; however, a consistent observation is the reduction in oxidative stress in the context of inflammation. − All Li⁺-releasing glasses had the ability to significantly reduce Nos2 expression (Figure B) and nitrite levels (Figure B), which are indicators of oxidative stress, mirroring findings from previous reports. Oxidative stress is characterized by abundant reactive oxygen species, which promote vasodilation and the recruitment of inflammatory cells. − Therefore, reducing oxidative stress is desirable and represents an important advantage of our Li⁺-releasing glasses.

CD86 and CD206 are considered markers of macrophage phenotype during tissue repair. While the expression levels of these markers alone do not represent an accurate measure of the cellular phenotype, they do indicate a gradual change from inflammatory to anti-inflammatory behavior. , In this study, macrophages treated with Li⁺-releasing glasses exhibited a reduced level of Cd86 expression (Figure B) during the early days of treatment, with CD206 subsequently increasing (Figure C–E). Additionally, the attenuation of functional aspects of the macrophage inflammatory phenotype, such as phagocytosis and IL-6 production (Figure ), provided further evidence of the anti-inflammatory actions of Li42 and Li40.

As mentioned above, lithium is a direct, reversible inhibitor of GSK3β. ,,, Lithium increases the phosphorylation of GSK3β at serine 9, whereby phosphorylated-serine 9 GSK3β (p-Ser9-GSK3β) is unable to carry out catalytic functions, and, therefore, remains inactive. , As expected, Li42 and Li40 could significantly inhibit GSK3β, as shown by the increased levels of p-Ser9-GSK3β (Figure A,B). Notably, the M1 macrophages also exhibited an increased p-Ser9-GSK3β compared with that of unprovoked macrophages (Figure A). This can be explained by the physiological negative feedback mechanisms triggered in response to LPS stimulation, such as the PI3K/Akt pathway. − In M1 macrophages, Akt (or protein kinase B) is fully activated and can phosphorylate several downstream targets to balance the inflammatory response, including GSK3β. Therefore, a baseline for GSK3β phosphorylation is commonly observed in M1 macrophages; however, treatment with Li42 and Li40 significantly increased the level of GSK3β phosphorylation (Figure B,C). This further supports the application of Li42 and Li40 as therapeutically useful anti-inflammatory phosphate glasses.

GSK3β has numerous downstream targets that may relate to the anti-inflammatory effects observed in this study. Given that the highest significant anti-inflammatory effects were observed for the fast release of Li⁺namely, Li42 and Li40 (Figure A and S1)we focused on the GSK3β targets that exhibit short half-lives and relatively quick renewal rates. STAT1 and STAT3 are activated upon receptor binding to IL-6 and IFNγ, which transmit the inflammatory signal to the nucleus. , Active STAT1 and STAT3 may have a half-life as short as 15 min, and, therefore, require constant reactivation. , Here, STAT1 and STAT3 activation was significantly reduced by treatment with Li42 and Li40, as shown by the reduced amounts of p-STAT1 and p-STAT3 (Figure A,D,E). Moreover, this effect was observed as early as 6 h after treatment with Li⁺-releasing glasses. Taken together, these results indicate that the anti-inflammatory effects of Li42 and Li40 derive from their ability to inactivate GSK3β and consequently interrupt the replenishment of STAT1 and STAT3 (Figure F), accelerating the resolution of inflammation.

In addition, histological observations revealed functional tissue repair rather than scarring in wounds treated with Li42 and Li40, as reflected by fewer inflammatory infiltrates (Figure A) and a weave-like organization of collagen fibers (Figure B). Tissue repairrather than scarringwas further corroborated by the abundance of CD206⁺ cells within the granulation tissue (Figure C,D), which are largely considered a marker of inflammation resolution. In this study, approximately 3 mg of Li⁺-releasing phosphate glasses was applied to the surgical site. It is expected that the optimal concentrations of phosphate glasses may vary for different target tissues. However, this study sets a reference point from which to narrow the concentration range.

Here, we present an experimental bioactive glass with fast anti-inflammatory action. Li42 and Li40 are promising candidates, providing therapeutically useful doses of Li⁺. They represent an alternative strategy for controlling inflammation in regenerative therapies, offering a lower production cost and reduced health risks compared to current recombinant cytokines. Slowly degrading bioactive glasses may have suitable biomedical applications, such as in pathological and chronic wounds, which require extended healing times. Follow-up studies should investigate whether stable, long-term ion release is beneficial in such situations. Additionally, other bioactive ions could be utilized in a similar manner to achieve different biological effects, either simultaneously or at various stages of tissue repair.

5. Conslusion

Experimental phosphate-based glasses showing a fast Li⁺ release profile (Li42 and Li40) exerted anti-inflammatory effects in vitro and in vivo. The presence of Li⁺ in the microenvironment blocked GSK3β activity and STAT1/3 transcription, interrupting pro-inflammatory signaling. These biomaterials have potential applications as local sources of anti-inflammatory ions to accelerate inflammation resolution.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsabm.5c00128.

• Nominal molar percentages of glass formulations; RT-qPCR probes; additional information on the phagocytosis assay, Western blot, and immunohistochemistry; ion release in the Transwell system; biocompatibility of Li⁺-releasing glasses (PDF)

This work was supported by the Japan Society for the Promotion of Science (award numbers 24K19917 and 22K20989 to Abe GL; award number 23K27770 to Imazato S).

The authors declare no competing financial interest.