Keratoconus, a progressive corneal thinning disorder, causes vision loss and affects approximately 1 in 2000 individuals. Corneal crosslinking (CXL) effectively halts its progression by creating covalent cross-links in the corneal stroma, improving biomechanical strength and resistance to enzymatic degradation. Conventional CXL (C-CXL) is time-consuming (≈1h), while accelerated CXL (A-CXL) protocols, using higher UVA intensities, aim for faster treatment. However, A-CXL's efficacy is limited by the rapid depletion of stromal oxygen, hindering cross-linking. Studies have shown that oxygen depletion occurs within seconds of UVA exposure at higher intensities, significantly reducing the biomechanical strength of the treated cornea. Current strategies to address hypoxia, such as oxygen supply devices and anterior chamber injections, have limitations due to the slow diffusion of oxygen into the stroma and short A-CXL treatment times. Therefore, an efficient oxygen supply method that rapidly generates oxygen within the stroma during UVA irradiation is urgently needed. This study explores the use of photocatalytic oxygen-generating nanomaterials as a novel approach to address this challenge. Graphitic carbon nitride (g-C3N4), a biocompatible material with photocatalytic properties, is investigated as a potential candidate. g-C3N4, particularly in the form of quantum dots (QDs), possesses suitable properties for this application: a simple synthesis process, non-toxicity, appropriate electron energy levels for water splitting to produce oxygen, stable physical and chemical properties, and absorption of UVA light used in CXL. Previous studies have demonstrated the use of g-C3N4 QDs as oxygen donors in cancer photodynamic therapy, but their potential in CXL has not been explored. This research investigates the application of g-C3N4 QDs as an independent photosensitizer and in combination with riboflavin (RF) to enhance A-CXL.

Numerous studies have demonstrated the effectiveness of corneal crosslinking (CXL) in halting or reversing the progression of keratoconus. The conventional CXL protocol involves removing the corneal epithelium, applying riboflavin, and then irradiating the cornea with UVA light. However, this process is time-consuming. Accelerated CXL (A-CXL) protocols have been developed to shorten treatment time by using higher UVA intensities. However, a critical limitation of A-CXL is the rapid depletion of oxygen in the corneal stroma, which significantly reduces the effectiveness of the cross-linking process. Research has shown that oxygen is essential for the photochemical reaction in CXL, and its depletion at higher UVA intensities limits the extent of cross-linking achievable. Various methods to improve oxygen supply during CXL have been explored, including oxygen delivery devices and injection into the anterior chamber. These methods, however, have limitations. This study proposes a novel approach by using photocatalytic nanomaterials to generate oxygen directly within the corneal stroma during the A-CXL procedure. The potential of graphitic carbon nitride (g-C3N4) quantum dots (QDs) as a biocompatible oxygen-generating material in this context is particularly promising, as indicated by their successful application in other photocatalytic processes.

This study involved the synthesis and characterization of g-C3N4 QDs and RF@g-C3N4 QDs composite photosensitizers. g-C3N4 QDs were synthesized using a modified method involving melamine, KOH, and NaOH, followed by annealing at different temperatures. The RF@g-C3N4 QDs composite was prepared by dispersing g-C3N4 QDs and RF in deionized water. Various techniques were used to characterize the synthesized materials, including transmission electron microscopy (TEM), high-resolution TEM (HRTEM), X-ray diffraction (XRD), Fourier-transform infrared (FTIR) spectroscopy, X-ray photoelectron spectroscopy (XPS), UV-visible spectroscopy, fluorescence spectroscopy, and zeta potential measurements. The photocatalytic oxygen generation ability of g-C3N4 QDs was evaluated using a dissolved oxygen analyzer, and the singlet oxygen generation ability was assessed using the chemical probe DPBF and electron spin resonance (ESR). In vitro biocompatibility studies were performed using human corneal epithelial cells (HCEC), human retinal microvascular endothelial cells (hRMEC), and retinal pigment epithelium cells (RPE-19), employing CCK-8 assay, Calcein-AM/PI double staining, and Annexin V-FITC/PI apoptosis assay. In vivo A-CXL evaluation was conducted on male New Zealand white rabbits. The corneal epithelium was removed, and the cornea was presoaked with different photosensitizers (PBS, RF, g-C3N4 QDs, RF@g-C3N4 QDs) before irradiation with UVA light at varying intensities and durations. The A-CXL effect was assessed through enzymatic digestion of collagen, stress-strain measurements, Corvis ST biomechanical analysis, and Optovue RTVue OCT measurements of central corneal thickness (CCT). In vivo biocompatibility was evaluated using slit lamp microscopy, specular microscopy, H&E staining, Alizarin Red S and Trypan Blue staining, TUNEL staining, routine blood analysis, biochemistry tests, and H&E staining of various organs. A modified Draize test was used to assess ocular irritation. Statistical analysis was performed using Graphpad Prism 8.0 software.

The synthesized g-C3N4 QDs exhibited excellent photocatalytic oxygen generation ability, with the optimal oxygen production observed at an annealing temperature of 350 °C. The g-C3N4 QDs also demonstrated a high singlet oxygen generation capability, as confirmed by DPBF assay and ESR. In vitro biocompatibility studies revealed excellent biosafety of g-C3N4 QDs, with high cell viability even at high concentrations. In vivo A-CXL evaluation showed that the g-C3N4 QDs and RF@g-C3N4 QDs groups had significantly enhanced A-CXL effects compared to the RF groups. The enzymatic digestion resistance, stress-strain measurements, and Corvis ST biomechanical parameters (ALL, A2V, HC-R, HC-DA) were superior in the g-C3N4 QDs group and comparable to the conventional RF CXL protocol. CCT measurements revealed faster corneal recovery in the g-C3N4 QDs and RF groups. In vivo biocompatibility studies showed minimal ocular surface irritation, endothelial cell preservation, and no significant pathological changes in various organs after A-CXL treatment. Blood tests revealed only temporary abnormalities in a few parameters (ALT, ALB, PLT) on days 1 and 15, which returned to normal by day 30. A-CXL using g-C3N4 QDs showed improved efficacy even in a hypoxic environment, highlighting its oxygen-generating capabilities. The study also found that higher UVA intensities (6 mW cm⁻² and above) are necessary to fully leverage the dual functions (oxygen generation and photosensitization) of g-C3N4 QDs during A-CXL.

This study demonstrates the potential of g-C3N4 QDs as a novel photosensitizer for enhancing A-CXL. The superior results of A-CXL with g-C3N4 QDs compared to traditional RF-based A-CXL, especially under hypoxic conditions, are attributed to the photocatalytic oxygen-generating properties of g-C3N4 QDs. This innovative approach addresses the limitations of current A-CXL techniques by providing an in situ oxygen source, thus overcoming the rapid oxygen depletion that typically restricts the efficacy of A-CXL. The observed enhanced corneal biomechanical properties (increased resistance to enzymatic digestion, higher ultimate stress and Young's modulus) suggest improved corneal stability and reduced risk of keratoconus progression. The excellent biocompatibility of g-C3N4 QDs further supports their clinical potential. These findings have important implications for improving CXL protocols, potentially leading to faster, more effective, and safer treatment of keratoconus and other corneal ectasias.

This study successfully developed a g-C3N4 QDs-based oxygen self-sufficient nanoplatform for enhanced corneal crosslinking. The g-C3N4 QDs demonstrated superior A-CXL efficacy compared to traditional RF-based methods, particularly in hypoxic environments, due to their efficient oxygen-generating capabilities. Excellent biocompatibility both in vitro and in vivo was confirmed, supporting the clinical potential of g-C3N4 QDs for A-CXL. Future research could focus on optimizing the g-C3N4 QDs formulation, exploring different UVA irradiation parameters, and conducting larger-scale clinical trials to assess long-term efficacy and safety.

The study primarily utilized a rabbit model, which might not perfectly replicate human corneal physiology and response to the treatment. The sample size in some of the in vivo experiments was relatively small, and larger-scale studies are needed to confirm the findings. Long-term follow-up studies are required to evaluate the durability of the enhanced corneal biomechanical properties achieved with g-C3N4 QDs-based A-CXL.