Hyaluronic acid-curcumin nanoparticles for preventing the progression of experimental autoimmune uveitis through the Keap1/Nrf2/HO-1 signaling pathway

Abstract

Globally, uveitis is a collection of intraocular inflammatory disorders that affect mainly the uvea, resulting in irreversible blindness and a heavy socioeconomic burden. Excessive autoimmune inflammation and oxidative stress are major drivers that contribute to the initiation and progression of uveitis. Nevertheless, current therapeutic methods for uveitis are limited and are accompanied by several serious adverse effects. Recently, nanotechnology-based antioxidant strategies have provided novel options for the treatment of ocular diseases. Although curcumin (CUR) has prominent antioxidant capacity and reactive oxygen species (ROS) scavenging ability, its low bioavailability and undetermined mechanisms limit its extensive application. This investigation demonstrated that esterified hyaluronic acid-curcumin nanoparticles (HA-CUR NPs) with superior aqueous dispersion exhibited exceptional antioxidant enzyme mimetic activity, incorporating superoxide dismutase (SOD), glutathione (GSH), catalase (CAT), and free radical scavenging ability. Further in vitro and in vivo experimental results validated the protective function of HA-CUR NPs against oxidative stress-induced damage and inflammatory responses, attenuated pathological progression, relieved microvascular damage, and regulated fundus blood flow in retinal vascular networks. This may be attributable to the specific ability of HA-CUR NPs to target the CD44 receptor and activate the Keap1/Nrf2/HO-1 signaling pathway, suggesting a potential mechanism. In summary, this study revealed that HA-CUR NPs, which are composed of a natural product and biomacromolecules with outstanding artificial antioxidant enzyme activities, may be novel agents for effectively and safely treating uveitis and other ROS-related diseases.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12951-024-03082-3.

Introduction

Uveitis is a primary cause of permanent blindness; it constitutes a significant financial burden worldwide [1, 2]. It is characterized by severe intraocular inflammation that mostly affects the uvea. There are two forms of uveitis: infectious and noninfectious. The latter type is always regarded as an autoimmune or autoinflammatory illness that is an ocular symptom of systemic disease or only a local ocular appearance [3, 4]. Different types of uveitis may be caused by different infectious agents. Owing to the presence of aberrant autoimmunity and excessive inflammation in noninfectious uveitis (NIU), the primary treatments for NIU at present consist of immunosuppressive and anti-inflammatory medications. These medications include corticosteroids and biological drugs that target antitumor necrosis factor-α (TNF-α) [1–5]. However, the long-term use of conventional treatments may generate many major negative effects, such as damage to the liver, infection, and large cost burdens for patients. These findings indicate that there is a significant gap in the clinical need for NIU therapy.

The response that occurs as a result of oxidative stress is generally considered to be a protective mechanism. However, the imbalance between the oxidation and antioxidant systems may lead to severe cellular and tissue harm, which is related to autoimmune uveitis [6, 7]. One of the consequences of severe oxidative stress is an increase in the nuclear factor kappa B (NF-κB) transcription factor, which consequently results in increased reactive oxygen species (ROS) levels. Oxidative stress is further exacerbated, which leads to an increase in inflammatory cytokines and chemokines [6]. Thus, investigating innovative medications that target excessive oxidative stress, particularly ROS, might provide more effective options for the treatment of uveitis.

Curcumin, a naturally occurring substance, has been shown in several studies to have the ability to treat severe autoimmune and inflammatory disorders [8, 9]. One possible explanation for this is their antioxidant and anti-inflammatory characteristics. Natural compounds, on the other hand, have limited bioavailability and detailed processes that have not been discovered [8, 10, 11]. These are the primary limits of the use of natural products in clinical applications. The combination of nanotechnology and natural products to manufacture nanoparticles has resulted in considerable improvements in the therapeutic effectiveness of natural products in treating inflammatory illnesses [12]. Owing to its natural ligand status as a ligand of the endocytic HA receptor CD44, which is expressed in a variety of cells, hyaluronic acid (HA), which is a necessary element of the extracellular matrix, has the potential for targeted drug delivery [13]. Recent developments in the targeting of CD44 receptors for HA-based nanoparticles have been reported to have the potential to improve the stability of nanoparticles, enhance their mobility and penetration, and control the release of medicines in the vitreous and retina [14].

This study reported the successful preparation of hyaluronic acid-curcumin nanoparticles (HA-CUR NPs) by esterification between the carboxyl group of HA and the hydroxy group of CUR and their potential application in uveitis treatment (Scheme 1). Significant free radical scavenging and the activities of three antioxidant enzyme mimetics, namely, SOD, GSH, and CAT, were shown by the HA-CUR nanoparticles, which suggests that they may be able to eliminate excessive amounts of ROS. The cellular absorption behavior of HA-CUR NPs was significantly observed in ARPE-19 cells, and these cells were found to be protected against oxidative stress, according to further study. In vivo tests revealed that HA-CUR nanoparticles were able to decrease oxidative stress and inflammation in animals suffering from experimental autoimmune uveitis (EAU). The administration of HA-CUR NPs attenuated pathological progression, relieved microvascular leakage, and regulated fundus blood flow in the retinal vascular networks of EAU mice. Further mechanistic research revealed that activation of the Keap1/Nrf2/HO-1 signaling pathway may affect the ability of HA-CUR NPs to alleviate oxidative stress injury in ARPE-19 cells and disease progression in EAU mice. This study resulted in the production of HA-CUR nanoparticles that have remarkable antioxidant enzyme-mimicking capabilities as well as remarkable radical-scavenging capabilities. Furthermore, we investigated the functions and possible mechanisms of these substances in the inhibition of uveitis progression.

Scheme 1.

Preparation process of HA-CUR NPs. HA-CUR NPs are nanoparticles with excellent water dispersion synthesized by esterification of hyaluronic acid and curcumin. These compounds demonstrated excellent free radical scavenging ability and mimetic activities of three antioxidant enzymes, namely, superoxide dismutase (SOD), glutathione (GSH), and catalase (CAT). In addition, they effectively inhibited inflammatory and antioxidant stress, alleviating tissue damage in EAU mice

Materials and methods

Materials

HA (403.31 Da), dimethyl sulfoxide (DMSO), dicyclohexylcarbodiimide (DCC), 4-dimethylaminopyridine (DMAP) and CUR were obtained from Aladdin Biochemical Technology Co. Ltd. (Shanghai, China). The product Complete Freund’s Adjuvant (CFA) was manufactured by Sigma‒Aldrich (St. Louis, Missouri, USA). Wuhan Procell Life Technology Co. Ltd. provided the adult retinal pigment epithelial cell line-19 (ARPE-19). Human IRBP (peptide 1169–1191) was supplied by Sangon Bioengineering Co., Ltd. (Shanghai, China). Pertussis toxin (PTX) was obtained from List Biological Labs, Inc. (Campbell, California, USA). Beyotime Biotechnology Co. Limited (Shanghai, China) supplied the superoxide dismutase (SOD), catalase (CAT), malondialdehyde (MDA), glutathione (GSH), oxidized glutathione (GSSG) and MTT assay kits. An ROS-staining solution was purchased from BeiBo (Shanghai, China). ELISA reagents were obtained from Shanghai Enzyme-Linked Biotechnology Co., Ltd., for the detection of TNF-α, IFN-γ, and IL-17. Primary antibodies for immunofluorescence, flow cytometry and western blotting were obtained from Abcam (Boston, Massachusetts, USA), Biochem International, Inc. (USA).

Statistical analysis: All the experiments were repeated at least 3 times. All the quantitative results are expressed as the means ± standard deviations. Each average represents the average of all experimental groups analyzed. Statistical significance analysis was performed via SPSS 22.0 and GraphPad Prism 6.0 software.

Synthesis of HA-CUR NPs

Esterification was employed to produce HA-CUR NPs [15, 16]. To activate the carboxylic group of HA, 400 mg of HA, 50 mg of DCC, and 100 mg of DMAP were combined with DMSO/H2O (V/V, 1:1) and blended for 2 h. The mixture was churned at 60 °C for 24 h after 74 mg of CUR was added to produce HA-CUR NPs. The mixture was dialyzed against deionized water for 72 h via a 3.5 kDa dialysis membrane. Finally, water-insoluble residues were eliminated by centrifuging the dialysis mixture at 15,000 rpm.

Characterization of HA-CUR NPs

Transmission electron microscopy (TEM, FEI Tecnai F20) was performed to acquire images of the micromorphology of the HA-CUR NPs. Dynamic light scattering (DLS) was employed to determine the average hydrodynamic dimensions of the HA-CUR NPs (Malvern Zetasizer nano ZS90). The infrared spectra of HA, CUR, and HA-CUR NPs were analyzed via a Thermo-Scientific Nicolet iS5 FTIR spectrometer. The UV‒vis absorption wavelengths of the particles were measured via ultraviolet‒visible (UV‒Vis) spectroscopy. ¹H NMR spectra were obtained via a ¹H NMR spectrometer (Bruker Avance NEO 600 MHz. Germany).

Radical scavenging and mimetic antioxidant enzyme activity assays

2,2′-Azinobis(3-ethylbenzothiazoline-6-sulfonate) (ABTS), 1,1-diphenyl-2-picrylhydrazyl (DPPH), and methylene blue (MB) assays, electron spin resonance (ESR) spectroscopy, and CAT mimetic activity tests were performed with reference to Gui et al. [17]. SOD mimicry was measured via Total Superoxide Dismutase Assay Kits with WST-8. GSH and GSSG reagents were used to test GSH mimic activity per the manufacturer’s instructions.

ARPE-19 cell culture

The cells were grown in cell-specific culture medium consisting of DMEM/F12 (PM150312), 1% P/S (PB180120), and 10% FBS (1642210–50). These cells were cultured in a humidified incubator at 37 °C and 5% CO₂.

In vitro MTT assays

MTT assays were used to test the cytotoxicity of HA-CUR NPs. APRE-19 cells were cultured in 96-well plates at a density of 8 × 104 cells mL − 1 for 24 h. Subsequently, the APRE-19 cells were cultured with different concentrations of HA-CUR NPs (0, 5, 15, 25, 35, or 45 µg/mL) for 4 h, 12 h, 24 h, 36 h, or 48 h, respectively. The cells were then gently rinsed twice with prewarmed PBS. After the supernatant was removed, 150 µl of DMSO was added to each well, and the 96-well plate was shaken at low speed for 10 min on a horizontal shaker. The relative viability of the cells was calculated by measuring the absorbance at 490 nm.

In vitro uptake behavior study

The cellular absorption behavior of HA-CUR NPs and free CUR was observed via fluorescence microscopy. APRE-19 cells were cultured in 12-well plates for 24 h at a density of 5 × 10⁴ cells/well. These cells were subsequently divided into six intervention groups. This was followed by the division of the cells into six intervention groups: the control group; 500 µmol/L H₂O₂-treated APRE-19 cells for 40 min; 5 µg/mL CUR-treated APRE-19 cells for 12 h; 500 µmol/L H₂O₂-treated APRE-19 cells for 40 min + 5 µg/mL CUR-treated APRE-19 cells for 12 h; 500 µmol/L H₂O₂-treated APRE-19 cells for 40 min + 5 µg/mL HA-CUR NP-treated APRE-19 cells for 12 h; and 500 µmol/L H₂O₂-treated APRE-19 cells for 40 min + 5 µg/mL HA-treated APRE-19 cells for 4 h + 5 µg/mL HA-CUR NP-treated APRE-19 cells for 12 h. Then, 4% paraformaldehyde solution was used to fix these cells. The detection of CUR was based on spontaneous fluorescence, and the detection of CD44 was performed through immunofluorescence technology.

In vitro ROS scavenging assay

The oxidation-sensitive fluorescent dye DCFH-DA was used to quantify intracellular ROS. The cells were cultured in 12-well culture plates at a density of 5 × 10⁴ cells/well for 24 h. This was followed by the division of the cells into five groups: the control group; APRE-19 cells treated with 500 µmol/L H₂O₂ for 40 min; APRE-19 cells treated with 500 µmol/L H₂O₂ for 40 min + 2 µg/mL CUR for 12 h; APRE-19 cells treated with 500 µmol/L H₂O₂ + 2 µg/mL HA-CUR NPs for 12 h; and APRE-19 cells treated with 500 µmol/L H₂O₂ + 5 µg/mL HA-CUR NPs for 12 h. The cells were immersed in 10 µmol/L DCFH-DA solution at 37 °C for 20 min. The solution was mixed upside down every 5 min to ensure that the cells and probe were in full contact. These cells underwent three rounds of washing in serum-free media before being stained with DAPI for 20 min, after which they were further washed with PBS three times. The intensity of the fluorescence intensity was noted, and the samples were photographed within 30 min via fluorescence microscopy.

Transcriptome analysis of the effects of HA-CUR NPs

APRE-19 cells (1 × 107) were cultivated for 24 h. These cells were grouped into the following three subsets according to different interventions: the control group, APRE-19 cells treated with 500 µmol/L H₂O₂ for 40 min, and APRE-19 cells treated with 500 µmol/L H₂O₂ for 40 min + 5 µg/mL HA-CUR NPs for 12 h. These cells were rinsed in PBS after the incubation media was removed and then transferred to RNase-free centrifuge tubes for RNA extraction and further analysis. Transcriptome analysis was conducted by BGI Genomics Co., Ltd. (Shenzhen, China).

EAU model induction and intervention

SPF female C57BL/6J mice (8 weeks, n = 40) were acquired from Wuhan Huayan Biotechnology Co., Ltd. During the one-week period of acclimatization, all the animals were provided a standard diet. The mice received 300 µg of human IRBP (1169–1191, PTARSVGAADGSSWEGVGVVPDV) in CFA (v/v, 1:1) via hypodermic injection and subsequently received 1 µg of PTX intraperitoneally to induce EAU. The mice were returned to their original environments subsequent to vaccination. To ensure the success of the modeling process, all the mice that were injected were identified via a slit lamp seven days after EAU induction. Rice was subsequently grouped into the following four subsets according to different treatments: control; EAU; EAU + low-dose HA-CUR NPs, which received an intravitreal injection of 15 µg/mL HA-CUR NPs for four consecutive days; and EAU + high-dose HA-CUR NPs, which received an intravitreal injection of 35 µg/mL HA-CUR NPs for four consecutive days. The dosage for each injection was 1 µL. The front section was disinfected with dilute povidone iodine before injection, and levofloxacin eye drops were used after injection to prevent infection. On the fourteenth day, these mice were examined for pathological changes via various methods and then sacrificed for further experiments.

Evans blue permeability assay

Anesthesia was followed by injection of an Evans blue dye solution through the superior vena cava for circulation for 30 min in the mice. The eyes were fixed in 4% paraformaldehyde for 2 h in the dark after being removed. Next, the retinas were separated layer by layer under a microscope in a culture dish containing a small amount of precooled PBS. The retinas were subsequently separated and transferred to glass slides, desiccated, sealed with neutral adhesive, and identified via fluorescence microscopy.

Fundus OCTA imaging

High-definition optical coherence tomography angiography (OCTA) images of multiple anatomical levels of the fundus were obtained using a Toward Pi Medical Technology Ltd. Beiming-Kun 400,000 ultrawide-angle full-field sweep OCT. All procedures followed the manufacturer’s directions, and the mice were routinely dilated after intraperitoneal anesthesia. The methods used to detect fundus blood flow in the mice included a 24 mm tomographic scanner, detailed quantification of retinal blood flow and structural analysis.

Immunohistochemistry

Frozen eye slices (10 mm) were prepared. The slices were immersed in anti-CD44 antibody solution (diluted 1:100) overnight at 4 °C, followed by immersion in fluorescently tagged secondary antibody solution. The slices were then stored at room temperature for 30 min. The samples were subsequently restained with DAPI and analyzed via a fluorescence microscope.

Assessment of MDA, SOD, IL-17, IFN-γ, and TNF-α in the retina and serum

The cornea, iris, lens, vitreous humor, cornea, and iris were completely removed. The retina is dissociated from the sclera in the region between the peripheral and optic papilla. To assess the expression of MDA, SOD, IL-17, IFN-γ, and TNF-α in the retina, fresh eye tissues were removed, stored at -80 °C, and then measured via ELISA. To obtain serum and plasma, a whole blood sample was collected from the orbital artery. The levels of MDA, SOD, IL-17, IFN-γ, and TNF-α in the serum were evaluated via ELISA.

Western blotting

Total protein was obtained from ARPE-19 cells and retinal choroidal tissue via RIPA lysis buffer. An ultrasonic cell pulverizer was used to crush ARPE-19 cells and retinal tissue to collect the supernatant proteins. A BCA protein kit was then used to quantify the protein levels. The proteins were further separated on a 10% SDS‒PAGE gel and subsequently transferred to PVDF membranes. The membranes were blocked with 5% milk and incubated overnight with primary antibodies, including anti-Keap1, anti-Nrf2, anti-HO-1, anti-NQO1, and anti-β-actin, at 4 °C, followed by incubation with HRP-conjugated antibodies for 60 min once again.

In vivo biocompatibility of HA-CUR NPs

The weights of the mice were recorded every two days. The hearts, livers, spleens, lungs, and kidneys of the mice in all the groups were collected for H&E staining on day 14 after the intravitreal administration of HA-CUR NPs. Whole blood samples were acquired to examine white blood cells (WBCs), red blood cells (RBCs), platelets (PLTs), the mean corpusular volume (MCV), and hematocrit (HCT). Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), and creatinine (SCR) levels were measured.

Results and discussion

Synthesis and characterization of HA-CUR NPs

HA-CUR NPs were synthesized via esterification between the carboxyl group of HA and the hydroxy group of CUR through a straightforward and dependable method. In particular, CUR was dissolved in a reaction system containing HA, DCC, and DMAP to form HA-CUR NPs. After 24 h of stirring, the resulting mixture turned dark, suggesting the effective combination of HA and CUR. The mixture was subsequently dialyzed and centrifuged to obtain HA-CUR NPs with superior aqueous dispersion. The micromorphology of the HA-CUR NPs was observed via TEM. As shown in Fig. 1a, HA-CUR NPs were generally in the form of uniformly distributed ultrafine particles with a size of 22.4 ± 8.7 nm. The average hydrodynamic dimension of the HA-CUR NPs was subsequently measured via DLS and was comparable to the result determined via TEM (Fig. 1b). We subsequently determined the structure of the HA-CUR NPs by measuring their ¹H NMR spectra. As shown in Figure S1, the peaks of HA-CUR at 1.5–2.0 ppm and 6.5–7.0 ppm were attributed to the acetyl peak of HA (- NHCOCH3) and the aromatic proton of CUR, respectively. Furthermore, the infrared intensities of the HA, CUR, and HA-CUR NPs were analyzed via FTIR to verify that the HA-CUR NPs were successfully synthesized. As shown in Fig. 1c, compared with the infrared spectra of HA and CUR, the apparent absorption peak at 3327 cm-1 was the absorption band of the -OH group in the curcumin structure; the obvious absorption peak at 2930 cm-1 was the absorption band of the double C = O group in curcumin; the new absorption peak at 1576 cm-1 was attributed to the vibration of the C = C within the benzene ring of curcumin; and the new absorption peak at 1732 cm-1 was the absorption band of the newly formed C = O double bond in the ester. Moreover, the HA-CUR NPs were dissolved in three solvents: water, ethanol and 0.9% NaCl. The UV‒vis absorbance spectra indicated that there was no marked change at 0 and 7 days (Figure S2). These results were consistent with the findings of Manju S, indicating the successful synthesis of HA-CUR NPs.

Fig. 1.

Characterization, ROS scavenging ability, and enzyme-mimicking activities of HA-CUR NPs. (a) Transmission electron microscopy images of HA-CUR NPs. (b) Dynamic light scattering data of HA-CUR NPs. (c) FTIR spectra of the HA, CUR, and HA-CUR NP samples. (d) UV‒Vis absorbance showing the ABTS radical scavenging ability of HA-CUR NPs. (e) UV‒Vis absorbance showing the DPPH radical scavenging ability of HA-CUR NPs. (f) UV‒Vis absorbance image illustrating the ability of HA-CUR NPs to scavenge ·OH to protect MB. (g) Illustration of the ROS scavenging process. (h) ESR spectra of DMPO indicating ·OH capture with or without HA-CUR NPs. (i) ESR spectra of TEMPO indicating single oxygen capture with or without HA-CUR NPs. (j) SOD-like activity of HA-CUR NPs. (WST-8 kit assay) (n = 3 for each group). (k) GSH-like activity of HA-CUR NPs. (n = 3 for each group). (l) CAT-like activity of HA-CUR NPs (n = 3 for each group)

Radical scavenging and antioxidant capacity of HA-CUR NPs

CUR has strong antioxidant properties, but its low aqueous solubility greatly limits its widespread application. Here, we explored the free radical scavenging ability and antioxidant capacity of HA-CUR NPs. First, the ABTS and DPPH radical assays, which are two radical models widely used for quantifying the antioxidant capacity of biological samples, were adopted to test the radical scavenging ability [18, 19]. The capacity of HA-CUR NPs to scavenge ABTS and DPPH radicals was evaluated by measuring the absorbances at 734 and 517 nm, respectively. As the concentration of HA-CUR NPs increased, the color gradually decreased, and decreased absorbance values were observed (Fig. 1d and e), suggesting the strong ABTS and DPPH radical scavenging abilities of the HA-CUR NPs. Furthermore, the OH radical (·OH) is a physiologically related radical and is involved in various types of cellular and histopathological damage. The ability of HA-CUR NPs to scavenge ·OH was assessed via the methylene blue (MB) assay. Our findings revealed that with the protection of HA-CUR NPs, MB remained blue or faded slightly, indicating that HA-CUR NPs had a strong scavenging ability against ·OH (Fig. 1f). The specific mechanisms of these three assays are schematically presented in Fig. 1g. In addition, 5,5-dimethyl-1-pyrroline N-oxide (DMPO) and 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO) were employed to trap •OH and single oxygen, respectively. The authors evaluated the capacity of HA-CUR NPs to scavenge free radicals produced by the Fenton reaction through electron spin resonance (ESR) spectroscopy. A decreased ESR amplitude was observed after treatment with the HA-CUR NPs, suggesting the efficient scavenging ability of the HA-CUR NPs for ·OH and single oxygen (Fig. 1h and i). In summary, these findings indicate that HA-CUR NPs may be effective ROS scavengers because of their strong antioxidant properties.

Mimetic activities of three antioxidant enzymes in the HA-CUR NPs

The effects of HA-CUR NPs on SOD, GSH, and CAT activity were further explored. SOD, an antioxidant enzyme that decomposes superoxide radicals into hydrogen peroxide and oxygen, has been demonstrated to exert remarkable therapeutic effects in uveitis [20]. Furthermore, GSH is a widely present tripeptide thiol and another important intracellular and extracellular antioxidant that plays many important roles in cellular signal transduction processes [21]. H₂O₂ can be eliminated from the body by decomposition into H₂O and O₂ by CAT, an iron porphyrin-binding enzyme that plays a vital protective role in organisms. As shown in Fig. 1j, k, and l, the HA-CUR NPs demonstrated SOD, GSH, and CAT mimetic effects in a dose-dependent manner, suggesting that the HA-CUR NPs have excellent antioxidant effects.

Cellular uptake behavior of HA-CUR NPs and in vitro cytotoxicity of HA-CUR NPs

The pathogenesis and progression of uveitis are closely associated with the impairment of ocular immune privilege, which may be caused by destruction of the integrity of the RPE [22]. Flourescence microscopy was employed to evaluate the ability of HA-CUR NPs to target ARPE-19 cells by identifying their internalization of HA-CUR NPs and free CUR pretreatment with or without H₂O₂. CUR is represented by the green fluorescence signal, whereas the CD44 receptor is represented by the red signal. ARPE-19 cells that were not treated with H₂O₂ assimilated minimal amounts of free CUR, as demonstrated in Fig. 2a, and the quantitative analysis results are shown in Fig. 2b and c. The cellular uptake behavior of HA-CUR NPs was more easily observed in ARPE cells pretreated with H₂O₂ than in those pretreated with free CUR. The following observations were made to ascertain whether the increased cellular absorption of HA-CUR NPs by ARPE-19 cells pretreated with H₂O₂ was a result of their ability to target the CD44 receptor. Initially, pretreatment of ARPE-19 cells with H₂O₂ led to greater expression of the CD44 receptor than that in untreated ARPE-19 cells. These findings indicate that the expression of the CD44 receptor may be promoted by severe oxidative stress. ARPE-19 cells were then treated with HA or HA-CUR NPs following H₂O₂ pretreatment. The absorption of HA-CUR NPs by cells clearly decreased, potentially as a result of the competitive binding of HA to the CD44 receptor, which inhibited NP-CD44 receptor binding. The results were further verified through flow cytometry (Fig. 2d). MTT assays revealed that the activity of ARPE-19 cells was reduced by incubation with higher concentrations of HA-CUR NPs (Figure S3a-b). The relative survival rate of the cells and the proportion of those with a normal morphology decreased (Figure S3c). In summary, these data indicate that the CD44 receptor targeting of HA-CUR NPs may promote the cellular uptake of HA-CUR NPs due to CD44 receptor overexpression in ARPE-19 cells caused by excessive oxidative stress.

Fig. 2.

(a) Fluorescence images showing CD44 and CUR expression in different groups. A: Control; B: 500 µmol/L H₂O₂-treated APRE-19 cells for 40 min; C: 5 µg/mL CUR-treated APRE-19 cells for 12 h; D: 500 µmol/L H₂O₂-treated APRE-19 cells for 40 min + 5 µg/mL CUR-treated APRE-19 cells for 12 h; E: 500 µmol/L H₂O₂-treated APRE-19 cells for 40 min + 5 µg/mL HA-CUR NP-treated APRE-19 cells for 12 h; F: 500 µmol/L H₂O₂-treated APRE-19 cells for 40 min + 5 µg/mL HA-treated APRE-19 cells for 4 h + 5 µg/mL HA-CUR NP-treated APRE-19 cells for 12 h. (b) Quantification of the fluorescence intensity of CD44 in different groups. (c) Quantification of the fluorescence intensity of CUR in different groups. (d, e) Flow cytometry analysis of the fluorescence intensity of CUR in different groups. Note: */#, P < 0.05; **/##, P < 0.01; ***/###, P < 0.001

HA-CUR NPs attenuated oxidative stress and inflammatory responses in ARPE-19 cells

Encouraged by their excellent cellular uptake, HA-CUR NPs were evaluated for their ability to safeguard ARPE-19 cells from oxidative stress caused by H₂O₂. DCFH-DA staining was conducted to visualize intracellular ROS after different interventions. The green fluorescence signal indicates CUR, and the red fluorescence signal represents ROS. Compared with those that were not treated with H2O2, ARPE-19 cells that were pretreated with H₂O₂ presented a red fluorescence signal, which suggested that the intracellular ROS levels were elevated, as demonstrated by Fig. 3a and the results of the quantitative analysis shown in Fig. 3b and c. When the cells were cocultured with free CUR, significant intracellular ROS and extremely low cellular uptake of CUR were detected. However, the red fluorescence signal decreased significantly, and the green fluorescence signal increased when H₂O₂-pretreated ARPE-19 cells were cocultured with HA-CUR NPs, suggesting that HA-CUR NPs can cause a dose-dependent reduction in ROS.

Fig. 3.

(a) Fluorescence images of ARPE-19 cells stained with DFCH-DA and 4′,6-diamidino-2-phenylindole (DAPI) after different treatments. A: Control; B: 500 µmol/L H₂O₂-treated APRE-19 cells for 40 min; C: 500 µmol/L H₂O₂-treated APRE-19 cells for 40 min + 2 µg/L CUR-treated APRE-19 cells for 12 h; D: 500 µmol/L H₂O₂-treated APRE-19 cells for 40 min + 2 µg/mL HA-CUR NP-treated APRE-19 cells for 12 h; E: 500 µmol/L H₂O₂-treated APRE-19 cells for 40 min + 5 µg/mL HA-CUR NP-treated APRE-19 cells for 12 h. (b) Quantification of the fluorescence intensity of CUR in different groups. (c) Quantification of the fluorescence intensity of ROS in different groups. (d) Flow cytometry quantification of ARPE-19 cells stained with DFCH-DA after different treatments. (e) Mean fluorescence intensity calculated from the flow cytometry data in d; (f) expression level of IL-1β in different groups; (g) expression level of IL-6 in different groups; (h) expression level of TNF-α in different groups. Note: */#, P < 0.05; **/##, P < 0.01; ***/###, P < 0.001

These findings were further verified via flow cytometry (Fig. 3d and e). Furthermore, H₂O₂-pretreated ARPE-19 cells exhibited markedly increased production of inflammatory cytokines, including IL-1β, IL-6 and TNF-α, which were downregulated by HA-CUR NPs but not by CUR (Fig. 3f, g and h). These results collectively indicate that HA-CUR NPs may mitigate ARPE-19 cell oxidative stress and the inflammatory response.

HA-CUR NPs attenuated pathological progression, relieved microvascular damage, and regulated fundus blood flow in vivo

EAU mice’s clinical symptoms and retinal tissue pathological changes were graded by severity according to previous standards [23, 24]. On the basis of the excellent cellular uptake behavior and ability of HA-CUR NPs to protect against oxidative stress damage in vitro, we researched the therapeutic efficacy of HA-CUR NPs in vivo (Fig. 4a). Compared with normal control rats, EAU mice presented significant inflammation in the anterior chamber and severe ciliary and conjunctival hyperemia, whereas EAU mice treated with an intravitreal injection of high HA-CUR NPs presented relieved symptoms (Fig. 4b). H&E staining revealed typical histopathological signs in the EAU model, including significantly increased retinal folds and increased inflammatory cells, which were markedly attenuated in the EAU model mice treated with the intravitreal injection of high concentrations of HA-CUR NPs (Fig. 4c). Clinical and histopathological scores were also assessed (Fig. 4d and e). The effects of HA-CUR NPs on the regulation of retinal vascularity and blood flow in vivo were also thoroughly investigated. Evans blue serves as a diazo dye that quickly and irreversibly binds to plasma ALB and therefore can be used as a highly efficient retinal angiographic agent. As shown in the retinal vascular network images obtained from Evans blue staining in Fig. 5a, massive vascular leakage sites were observed in the retinas of the EAU mice, and this leakage was markedly alleviated in the EAU mice treated with the intravitreal injection of high concentrations of HA-CUR NPs (Figure S4). Conversely, symptoms were not significantly ameliorated in EAU model mice following the intravitreal injection of low concentrations of HA-CUR NPs. Furthermore, immunohistochemistry was implemented to evaluate the targeting of RPE cells by HA-CUR NPs. As shown in Fig. 5b, HA-CUR NPs could be observed in the cytoplasm of RPE cells. The results revealed the colocalization of HA-CUR NPs and CD44-positive RPE cells.

Fig. 4.

(a) Illustration of the experimental procedure used to establish the experimental autoimmune uveitis (EAU) model. (b) Clinical manifestations of retinas in the different groups were observed via a slit lamp. The white arrows indicate anterior chamber inflammation; the yellow arrows indicate conjunctival and ciliary congestion. (c) HE-stained histopathological images of retinas from the different groups. (d) Clinical score based on the results in (b). (e) Histopathological score based on the results in (C). Note: *, P < 0.05; **, P < 0.01; ***, P < 0.001

Fig. 5.

(a) Images of retinas from the different groups observed by EB angiography. (b) Immunohistochemical images of an EAU retina stained with CD44 and 4′,6-diamidino-2-phenylindole (DAPI). The yellow arrow indicates the colocalization of HA-CUR NPs and CD44 on RPE cells

OCTA, a noninvasive fundus vascular imaging technology that is innovative, has been extensively employed to diagnose fundus disorders, including diabetic retinopathy and uveitis. Unlike traditional methods, OCTA does not require the intravenous application of dye, preventing fluorescein leakage [25]. We employed OCTA imaging to quantify the thickness and density of blood flow in each fundus layer in distinct groups of mice. Figure 6a shows the superficial blood flow (SBF), deep blood flow (DBF), choroidal capillary blood flow (CCBF), and whole retinal blood flow (WRBF) of the fundus of the various groups. In contrast to NC mice, EAU mice presented typical retinal vascular morphological anomalies. Conversely, the retinal vascular networks of EAU model mice were normal following intravitreal injections of both low and high concentrations of HA-CUR NPs. Figure 6b shows the OCTA fundus blood flow and thickness/volume maps. Compared with NC mice, EAU mice presented more impaired blood flow signals and uneven vessel morphology across the whole retinal vascular network, indicating significant vascular dysfunction in the fundus of the EAU mice. In contrast, these abnormal symptoms were alleviated in EAU model mice by intravitreal injections of both low and high concentrations of HA-CUR NPs. The quantitative analysis of the SBF, DBF, CCBF, and WRBF layers of the fundus, as illustrated in Fig. 6c and d, revealed a marked decrease in thickness in the EAU group compared with the NC group. Both the EAU mice treated with the intravitreal injection of low- and high-HA-CUR NPs subsequently exhibited remarkable improvements to relatively normal levels. These results indicate that HA-CUR NP treatment markedly attenuated pathological progression, relieved microvascular leakage, and regulated fundus blood flow; therefore, HA-CUR NPs serve as effective agents for retinal protection in EAU.

Fig. 6.

Optical coherence tomography angiography (OCTA) imaging, blood flow, and thickness of multiple layers of OCTA in the fundus. (a) Superficial blood flow (SBF) (a1-h1), deep blood flow (DBF) (a2-h2), choroidal capillary blood flow (CCBF) (a3-h3), and whole retinal blood flow (WRBF) (a4-h4) of the OCTA panorama in different groups. (b) OCTA fundus thickness/volume (A, C, E, G) and blood flow (B, D, F, H) maps in different groups. (c) Quantification of the thickness of the SBF, DBF, CCBF, and WRBF retinal layers in different groups. (d) Quantification of blood flow in the SBF, DBF, CCBF, and WRBF retinal layers in different groups. Asterisks indicate a significant difference. Note: *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001

HA-CUR NPs mitigated oxidative stress and reduced inflammatory responses

HA-CUR NPs have been shown to mitigate oxidative stress in vitro; thus, we investigated whether HA-CUR NPs exert similar effects in vivo. MDA is usually regarded as a marker for lipid peroxidation and oxidative stress [26], whereas SOD is a metalloenzyme that safeguards the organism against oxidative stress [27]. The redox state of uveitis is suggested to be imbalanced, as evidenced by the significantly higher intraocular MDA and reduced intraocular SOD in EAU animals than in normal controls (Fig. 7a-j). Fortunately, the administration of HA-CUR NPs dose-dependently decreased high MDA levels and elevated low SOD levels in EAU mice, whereas treatment with free CUR had a minimal effect on these indicators. Furthermore, given that IL-17, IFN-γ and TNF-α are characteristic inflammatory cytokines that participate in the pathogenesis and development of uveitis [28], we further examined the expression of these three cytokines in the different groups. Compared with NC mice, EAU mice presented markedly greater expression levels of intraocular IL-17, IFN-γ and TNF-α, which were decreased in a dose-dependent manner by treatment with HA-CUR NPs rather than free CUR. Furthermore, we quantified the expression of the CD44 receptor in response to the excessive oxidative stress that was induced in vitro (Fig. 7k, l). The CD44 receptor expression of the EAU mice exceeded that of the normal controls, suggesting a potentially more intense targeted interaction between HA and the receptor. The combination of CUR and HA may increase CUR bioavailability by targeting HA to the CD44 receptor, reducing oxidative stress and inflammatory responses, given that HA-CUR NPs are more efficacious than free CUR is. Considering that HA-CUR NPs are significantly more efficacious than free CUR is, one reasonable conclusion can be drawn that the combination of CUR and HA could markedly increase the bioavailability of CUR, which may be mediated by the specific targeting of HA to the CD44 receptor, thereby efficiently attenuating excessive inflammatory responses and oxidative stress.

Fig. 7.

Quantification of the serum MDA (a), SOD (b), TNF-α (c), IFN-γ (d), and IL-17 (e) levels and the intraocular MDA (f), SOD (g), TNF-α (h), IFN-γ (i), and IL-17 (j) levels in different groups. (k, l) Expression of the CD44 receptor in control and EAU rats. Note: */#, P < 0.05; **/##, P < 0.01; ***/###, P < 0.001

HA-CUR NPs exerted anti-inflammatory effects by activating the Keap1/Nrf2/HO-1 signaling pathway

By identifying the potential mechanism behind the prominent oxidative stress protection of HA-CUR NPs, we may be able to better understand the therapeutic benefits of HA-CUR NPs and optimize their clinical application. To further explore the molecular mechanism of HA-CUR NPs, we conducted transcriptome sequencing analysis on ARPE-19 cells treated with different interventions. The Molecular Signatures Database C2 database was used to perform GSEA on the DEGs, and the top 20 results are illustrated in Fig. 8a. The majority of genes whose expression was significantly upregulated by HA-CUR NPs were associated with inflammation, immunology, and oxidative stress, including REACTOME_KEAP1_NFE2L2_PATHWAY, a classic pathway that is involved in the antioxidative stress response and safeguards cells from oxidative damage. Additional enrichment analysis confirmed the activation of the Keap1/Nrf2 signaling pathway (Fig. 8b and c). A gene expression heatmap of the Keap1/Nrf2 signaling pathway revealed that several genes in the Keap1/Nrf2 signaling pathway, including GSTM5, GSTO2, MGST1, KEAP1, HMOX1, and NFE2L2, were upregulated by the HA-CUR NPs (Fig. 8d and e), which promoted the expression of anti-inflammatory and antioxidant-related proteins to resist oxidative stress damage. In addition, the top 5 potential signaling pathways and the top 6 DEGs revealed by transcriptome sequencing analysis are shown in Fig. 8f. As one of the most vital antioxidant axes in physiological and pathological processes, the activation of Nrf2 signaling could contribute to a series of antioxidant responses, thereby protecting cells against oxidative stress damage [29, 30]. Specifically, Nrf2 is a transcription element that promotes the transcription and expression of various major antioxidative enzymes, including heme oxygenase-1 (OH-1) and NAD(P)H quinone oxidoreductase 1 (NQO1), by binding to antioxidation stress reaction sites [30, 31]. In numerous cell lines, including RPE cells, natural products have been shown to alter the Keap1/Nrf2/HO-1 signaling pathway [32]. Notably, Nrf2 was found to be closely related to the development of uveitis, serving as a potential intervention target in the modulation of redox hemostasis [33]. Western blotting was used to determine whether HA-induced CUR NPs can safeguard ARPE-19 cells and EAU model mice from oxidative stress-induced damage by activating the Keap1/Nrf2/HO-1 signaling pathway. The protein expression levels of Keap1, Nrf2, HO-1, and NQO1 in ARPE-19 cells pretreated with hydrogen peroxide were substantially lower than those in the control group. However, the H₂O₂-induced effects on the Keap1/Nrf2/HO-1 signaling pathway were reversed by HA-CUR NPs in a dose-dependent manner, as illustrated in Fig. 9a and b. Furthermore, in contrast to those in the control group, the protein expression levels of Keap1, Nrf2, HO-1, and NQO1 in the EAU group were substantially lower. On the other hand, HACUR NPs were still capable of reversing the changes in protein expression levels in a dose-dependent manner (Fig. 9c and d). These findings indicate that Keap1/Nrf2/HO-1 induces the protective influence of the HA CUR NPs. These results suggest that the Keap1/Nrf2/HO-1 signaling pathway contributes significantly to the anti-inflammatory effect of the HA CUR NPs. Curcumin is a natural activator of Nrf2 [34]. Mechanistic studies have shown that curcumin induces conformational changes in Keap1 and the ubiquitination of Nrf2 by binding to Keap1Cys151 [35]. However, the effects of curcumin on the protein expression of Nrf2 in cells subjected to oxidative stress are not entirely consistent. Liu et al. confirmed that pretreatment with curcumin did not increase the expression of total Nrf2 in cells subjected to oxidative stress stimulation [36]. These findings are inconsistent with our findings. However, some studies have shown that curcumin can increase the protein expression of Nrf2 in cells subjected to various oxidative stresses [35, 37]. The concentration of curcumin used and the duration of its action may be key factors influencing the level of Nrf2 expression.

Fig. 8.

(a) The top 20 results of GSEA enrichment analysis of the DEGs via the Molecular Signatures Database C2 database. (b, c) Running enrichment score of the Keap1/Nrf2 signaling pathway. (d, e) Gene expression heatmap of the Keap1/Nrf2 signaling pathway regulated by HA-CUR NPs. (f) Protein levels of Keap1, Nrf2, HO-1, and NQO1 in ARPE-19 cells after different treatments. h, i) Protein levels of Keap1, Nrf2, HO‐1, and NQO1 in rats in different groups

Fig. 9.

Western blotting results in vitro and in vivo. (a, b) Protein levels of Keap1, Nrf2, HO-1, and NQO1 in ARPE-19 cells after different treatments. (c, d) Protein levels of Keap1, Nrf2, HO-1, and NQO1 in rats in different groups. Note: */#, P < 0.05; **/##, P < 0.01; ***/###, P < 0.001

Assessment of biocompatibility in vivo

The biocompatibility of HA-CUR NPs in vivo was assessed through H&E staining of important organs, including the heart, liver, spleen, lung, and kidney. As shown in Fig. 10a-j, no remarkable lesions, hemorrhage, or necrosis were detected in these major organs. No substantial changes in liver or renal function were observed in any of the mice in the present study, as determined by ALT, AST, BUN, and SCR in the serum. The WBC, RBC, PLT, MCV, and HCT were also measured and demonstrated no significant changes. No obvious lesions were found in the retinal tissue after the HA-CUR NPs were injected into the eye. (Figure S5a). We further examined three inflammatory cytokines to evaluate the toxicity of HA-CUR NPs to intraocular tissues. As shown in Figure S5b, the expression levels of IL-17, IFN-γ and TNF-α in the eyes of the mice injected with HA-CUR NPs did not significantly differ from those in the control group. Therefore, the HA-CUR NPs synthesized in the present study have the potential to serve as safe and efficacious agents for preventing the progression of EAU.

Fig. 10.

In vivo biocompatibility of HA-CUR NPs. (a) Images of H&E-stained heart, liver, spleen, lung, and kidney tissues from the different groups. (b-f) Quantification of white blood cells (WBCs), red blood cells (RBCs), platelets (PLTs), mean corpuscular volume (MCV), and hematocrit (HCT) in the different groups. (g-j) Serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), and serum creatinine (SCR) levels in the different groups

Conclusion

In summary, HA-CUR NPs, which were produced through esterification between the carboxyl group of HA and the hydroxy group of CUR, demonstrated excellent free radical scavenging ability and mimetic activity for three antioxidant enzymes, SOD, GSH, and CAT. ARPE-19 cells clearly assimilated HA-CUR NPs in vitro, which may be attributed to the specific ability of HA-CUR NPs to target the CD44 receptor. HA-CUR NPs were shown to protect ARPE-19 cells against H₂O₂-induced oxidative stress damage. In vivo experimental assays demonstrated that HA-CUR NPs can mitigate intraocular inflammation and oxidative stress in EAU mice. HA-CUR NPs attenuated pathological progression, relieved microvascular damage and regulated fundus blood flow in the retinal vascular networks of EAU rats. Further mechanistic research revealed that activation of the Keap1/Nrf2/HO-1 signaling pathway may contribute to the ability of HA-CUR NPs to alleviate the damage generated by oxidative stress in ARPE-19 cells and disease progression in EAU mice. Biocompatibility assays indicated that the HA-CUR NPs had no remarkable toxicity. HA-CUR NPs, which possess numerous intrinsic antioxidant enzyme-mimicking capabilities, have the potential to scavenge ROS effectively and mitigate EAU symptoms and development. Consequently, these compounds are viable clinical translational therapeutic agents for uveitis and other ROS-related disorders.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

All authors contributed to the article and approved the final manuscript. Weiwei Tang & Yun-Di Yi & Xiaomin Huang&Fan Cao: Methodology, Investigation, Writing – original draft. Xianwen Wang & Lei Shi & Zheng-Xuan Jiang & Li-Ming Tao: Design of the manuscript. Manli Deng: Collection of resources. Xianwen Wang & Weiwei Tang: Revision the manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (No. 82070986, 82171043), the Excellent Youth of Natural Science Research Projects in Anhui Province Universities (2023AH030060), the Key Project of Scientific Research in Colleges and Universities of Anhui Province (grant no. 2024AH050550) and Anhui Province Excellent Research and Innovation Team Project (2024AH010013). The authors would like to thank the Shiyan laboratory (http://www.shiyanjia.com) for their help in language polishing.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.