Self-assembling biomimetic peptide hydrogel regulates tissue homeostasis to promote repair of persistent corneal epithelial defects

Abstract

Persistent corneal epithelial defects (PCED) remain a significant clinical challenge due to rapid tear turnover, blinking-induced shear stress, and a protease-rich ocular surface that limit the efficacy of soluble bioactive factors. We report a dual-functional self-assembling peptide hydrogel, mIGF-Gel (Biotin-^DFYIGSSSR), incorporating the adhesion motif YIGSR derived from laminin and the SSSR bioactive motif derived from insulin-like growth factor 1 (IGF-1) in a single design. The peptide forms a β-sheet-rich nanofibrous hydrogel with shear-thinning, viscoelastic, and self-recovering properties, enabling adaptation to the dynamic ocular surface. Compared with recombinant IGF-1, mIGF-Gel retains IGF-1-like bioactivity while providing enhanced proteolytic stability, ocular retention, and local delivery efficiency. Mechanistically, mIGF-Gel activates IGF-1 receptor-mediated PI3K/AKT signaling to promote corneal epithelial cell proliferation, migration, and survival. In an acute corneal injury model, topical mIGF-Gel accelerates re-epithelialization and improves epithelial repair quality. In a dexamethasone-suppressed delayed-healing model, once-daily administration achieves therapeutic outcomes comparable to three-times-daily IGF-1, while further enhancing epithelial stratification, basement membrane reconstruction, and early nerve repair. Collectively, these results establish mIGF-Gel as a bioactive, wound-adaptive peptide hydrogel that overcomes pharmacokinetic limitations and holds translational potential for the treatment of persistent epithelial defects, as demonstrated in both acute injury and a PCED-mimetic delayed-healing model.

Graphical abstract

A supramolecular peptide hydrogel is engineered through non-covalent self-assembly to function at the dynamic ocular surface. The material integrates bioactive motifs with proteolysis resistance and shear-adaptive viscoelasticity, enabling sustained receptor activation under tear clearance and blinking. This multifunctional platform demonstrates how rational materials design restores corneal epithelial homeostasis.

Highlights

• •Supramolecular hydrogel mimics IGF-1 and laminin for PCED repair.
• •D-amino acids confer proteolytic stability and prolonged ocular retention.
• •Therapy restores corneal epithelial and neuronal homeostasis in vivo.

1. Introduction

Persistent corneal epithelial defects (PCED) are pathological conditions characterized by delayed healing or persistent non-healing of corneal epithelial wounds [1,2]. The pathological features of PCED mainly include impaired corneal epithelial cell migration and proliferation, abnormal epithelial–stromal adhesion, and dysregulated tissue repair, accompanied by sustained local inflammatory activation, increased oxidative stress, and corneal nerve dysfunction or insufficient neurotrophic support [[3], [4], [5]]. These abnormalities can maintain corneal wounds in a prolonged non-healing state, thereby increasing the risk of corneal ulceration, stromal melting, and even perforation. As an ocular surface injury-related disorder, PCED shares certain pathological mechanisms with other ocular surface diseases, such as oxidative stress, inflammatory responses, and epithelial injury. Dry eye syndrome (DES), one of the most common superficial ocular surface diseases, is primarily driven by tear film dysfunction and chronic inflammation, and typically manifests as diffuse superficial punctate epithelial erosions [6,7]. In contrast, PCED is more prominently characterized by full-thickness loss of the corneal epithelium and failure of re-epithelialization. These distinctions suggest that therapeutic strategies for PCED should focus on promoting corneal re-epithelialization and improving the local wound-healing environment, rather than merely alleviating ocular surface symptoms [8,9].

While blood-derived therapies such as autologous serum have shown efficacy in promoting re-epithelialization, their clinical utility is severely restricted by logistical complexity, batch variability, and prohibitive costs [[10], [11], [12]]. To actively compensate for the loss of regenerative cues, bioactive proteins have been explored. Laminin, a basement membrane component containing the YIGSR motif, provides structural support for epithelial adhesion and migration, and also contributes to nerve growth [[13], [14], [15]], whereas insulin-like growth factor-1 (IGF-1) promotes epithelial proliferation, differentiation, and neuroregeneration through receptor-mediated signaling [[16], [17], [18]]. The coordinated and sustained supply of both components is essential for reconstructing the integrated regenerative microenvironment required for corneal epithelial healing. Despite their potent biological functions, protein-based therapies are fundamentally limited by short half-lives, poor ocular bioavailability, and insufficient resistance to mechanical stress [19,20].

Self-assembling peptide hydrogels represent a compelling solution, as they spontaneously form nanofibrous architectures that resemble native extracellular matrix and provide mechanical protection while enabling localized, sustained delivery, with well-defined functionality and enhanced resistance to degradation [[21], [22], [23]]. For instance, peptides like Nap-FFG assemble into stable β-sheet nanofibers through non-covalent interactions, providing supportive microenvironments and wound protection [19,20]. Concurrently, minimalist peptide motifs derived from full-length proteins provide an efficient path to recapitulate desired biofunctions while avoiding the intrinsic limitations of their parent proteins, including instability and complex folding requirements [13,24,25]. For example, the “SSSR” sequence, a core functional motif derived from IGF-1, has been demonstrated to enhance cell proliferation and suppress apoptosis without the instability associated with larger, unstructured protein fragments [26,27].

Peptide-based supramolecular systems, including self-assembling peptide hydrogels, have recently emerged as promising topical platforms for ocular surface disorders such as corneal injury, dry eye disease, and infectious keratitis. In these applications, peptide hydrogels have mainly been engineered to improve local delivery by prolonging ocular surface residence, enhancing drug bioavailability, and enabling sustained release or transcorneal transport of therapeutic cargos, thereby contributing to inflammation control and epithelial repair [[28], [29], [30]]. Despite these advances, most reported systems remain delivery-centered: the hydrogel matrix primarily serves as a local depot, whereas its intrinsic ability to instruct epithelial regeneration is limited. As a result, these materials are not ideally suited to address the multifactorial pathology of PCED, which requires coordinated regulation of epithelial adhesion, migration, survival, and resistance to apoptosis within a protease-rich and mechanically dynamic wound microenvironment. Therefore, an optimal self-assembling peptide hydrogel for impaired epithelial repair, should move beyond a passive carrier role and function as a bioactive, wound-adaptive matrix capable of actively remodeling the corneal epithelial repair microenvironment.

Here, we rationally designed a multifunctional self-assembling peptide, mIGF-Gel (Biotin-^DFYIGSSSR), by integrating two complementary bioactive motifs for cornea injury treatment (Scheme 1). The laminin-derived adhesion sequence YIGSR, which has been widely reported to promote epithelial cell adhesion and migration [[31], [32], [33]], was incorporated as an epithelial repair motif. In parallel, the IGF-1-derived motif SSSR, selected from the IGF-1 C-domain, was introduced to retain IGF-1-related bioactivity [27,34]. Together, these motifs were integrated into a self-assembling peptide scaffold to confer both hydrogel-forming capability and biological signaling functions. Incorporation of D-phenylalanine and biotin at the N-terminus further drives the assembly of the peptide into a stable β-sheet-rich nanofibrous network, conferring shear-thinning, viscoelastic, and self-recovering properties that enable adaptation to the dynamic ocular surface while enhancing local retention and proteolytic stability. Through this combination of molecular functional encoding and supramolecular assembly, mIGF-Gel not only serves as a local delivery platform but also provides intrinsic bioactivity to sustain epithelial repair processes. We first characterized its physicochemical properties and bioactivity in vitro, and then assessed its early wound-healing capacity using an acute corneal epithelial injury model. To more closely approximate clinically refractory PCED, we validated key in vivo findings in a dexamethasone-induced delayed-healing model, which better recapitulates the impaired epithelial repair that defines this condition. Collectively, these findings establish a design strategy for peptide hydrogels with both structural adaptability and regenerative bioactivity, and offer a new paradigm for the development of materials for ocular surface regeneration.

Scheme 1.

Molecular design and therapeutic mechanism of mIGF-Gel. (A) The rationally designed biomimetic peptide Biotin-^DFYIGSSSR, which fuses laminin-adhesive and IGF-1–bioactive motifs, self-assembles into the shear-thinning hydrogel mIGF-Gel, forming a β-sheet–rich nanofibrous architecture. (B) Schematic illustration of corneal homeostasis restoration. mIGF-Gel orchestrates re-epithelialization, basement membrane remodeling, and nerve regeneration, ultimately restoring corneal homeostasis in corneal injury model. Achieved by directly binding to the IGF-1 receptor and activating the PI3K/AKT signaling pathway.

2. Results and discussion

2.1. Preparation and characterization of mIGF-Gel

To experimentally validate this design, mIGF-Gel and five control peptides were synthesized and characterized, including a scrambled peptide (Biotin-^DFGIYSRSR), an SSSR-containing peptide (Biotin-^DFGIYSSSR), a YIGSR-containing peptide (Biotin-^DFYIGSRSS), Biotin-YIGSSSR, and Biotin-^LFYIGSSSR, and HR-MS and ¹H NMR analyses confirmed the successful synthesis and structural identity of peptide (Fig. S1–S6). These controls allowed us to distinguish the roles of motif composition, sequence order, and D-phenylalanine in peptide assembly and bioactivity.

Concentration-dependent gelation experiments showed that mIGF-Gel reproducibly formed a stable, transparent hydrogel at 0.4 wt% through a heat–cool cycle, whereas most control peptides either precipitated or remained turbid even at higher concentrations, indicating that the precise sequence arrangement is critical for stable hydrogel formation (Fig. 1A and S7). The self-assembly behavior of mIGF-Gel was further supported by a critical aggregation concentration (CAC) of 205 μM (Fig. S8). TEM revealed that mIGF-Gel assembled into uniform nanofibers forming an entangled three-dimensional network. Similar nanofibrous structures were observed for Biotin-^LFYIGSSSR, Biotin-^DFYIGSRSS, Biotin-^DFGIYSSSR, and Biotin-^DFGIYSRSS, whereas Biotin-YIGSSSR exhibited more bundled fibrous aggregates with less evident network organization (Fig. 1B and S9), suggesting that the introduction of phenylalanine plays an important role in promoting nanofiber formation. CD and FTIR analyses further confirmed that mIGF-Gel adopted a predominant β-sheet structure, as evidenced by its CD profile and an amide I band centered at 1637 cm⁻¹ (Fig. 1C–S10A, and S11). Although both mIGF-Gel and Biotin-^LFYIGSSSR formed β-sheet-rich nanofiber hydrogels, they exhibited markedly different assembly stability. Proteolytic degradation experiments showed that after 6 h, only 30.57% of the L-configured peptide remained, whereas 49.10% of the D-configured peptide was retained (Figure S10B). In addition, after standing for 12 h, the hydrogel formed by the L-configured peptide became obviously turbid, whereas the D-configured hydrogel remained transparent (Figure S10C-D), suggesting that D-phenylalanine is not required for self-assembly initiation but mainly contributes to improved hydrogel stability and proteolytic resistance.

Fig. 1.

Self-assembly properties and physicochemical characterization. (A) Optical and (B) TEM images of mIGF-Gel at 0.4 wt%. Scale bar: 200 nm. (C) CD spectrum of mIGF-Gel. (D) The G′ of mIGF-Gel at different concentrations (0.6, 0.5, 0.4, and 0.3 wt%). (E) Sol-gel transition and self-recovery behavior of mIGF-Gel. (F) Changes in G′ and G″ under gradually increasing strain. (G) Thixotropic and self-healing properties of mIGF-Gel evaluated by step-strain measurements. (H) Shear-thinning behavior of mIGF-Gel under variable shear rates. (I) Representative OCT images of rat corneas before and after hydrogel instillation and following blinking. Scale bar: 100 μm.

Given the dynamic mechanical environment of the ocular surface, we next evaluated the rheological properties of mIGF-Gel. The hydrogel exhibited concentration-dependent viscoelasticity, with the storage modulus (G′) consistently exceeding the loss modulus (G″), indicating a dominant elastic feature (Fig. 1D). Time-sweep measurements further confirmed the stability of the hydrogel network after gel formation (Fig. S12). Notably, mIGF-Gel was returned to its gel state even after mechanical disruption into a viscous sol, demonstrating thixotropy and self-healing behavior (Fig. 1E). Specifically, strain sweep analysis revealed a gel-sol transition at a critical strain of ∼56%, while strain-dependent measurements demonstrated reversible transitions under alternating high (100%) and low (0.1%) shear (Fig. 1F and G). Although the first cycle showed slightly higher G′ and G″ values than the following cycles, the subsequent cycles displayed stable and reproducible recovery without progressive modulus decay. This indicates that the hydrogel maintains strong reversible reconstruction ability upon repeated shear, despite possible initial network rearrangement after the first high-strain loading. This self-healing behavior likely arises from reversible supramolecular interactions, including β-sheet-mediated hydrogen bonding, π–π stacking, and hydrophobic interactions [35,36]. Crucially, mIGF-Gel displayed shear-thinning behavior (Fig. 1H), a property that ensures high viscosity at rest for retention and low viscosity under shear to minimize friction during blinking. Optical coherence tomography (OCT) imaging confirmed that mIGF-Gel could spread uniformly across the ocular surface before and after blinking, demonstrating good extensibility and highlighting the advantages of hydrogel-based ocular drug delivery (Fig. 1I). Taken together, these favorable rheological properties suggest that mIGF-Gel possesses good ocular surface spreading performance and retention capability, highlighting its potential advantages for topical ocular delivery.

2.2. Sustained IGF-1R activation via protease-resistant mimicry of mIGF-Gel

To validate signaling activation [17], we examined pIGF-1R levels in human corneal epithelial cells (HCECs) by western blotting. As shown by the immunoblot and corresponding densitometric quantification (Fig. 2A and B), mIGF-Gel markedly increased pIGF-1R levels compared with the control group, and the magnitude of this increase was comparable to that induced by recombinant IGF-1 protein. We further examined the phosphorylation of AKT, a key downstream effector. Under hydrogen peroxide induced oxidative stress, pIGF-1R expression was markedly reduced. In contrast, mIGF-Gel sustained IGF-1R activation and promoted AKT phosphorylation, demonstrating its ability to maintain bioactivity under pathological conditions (Fig. 2D–F). Although microscale thermophoresis (MST) indicated that mIGF-Gel binds IGF-1R with a moderate affinity (K_D = 185.7 nM) compared to recombinant IGF-1 (K_D = 14.9 nM) (Fig. 2C), this is significantly compensated by its superior proteolytic resistance. In Proteinase K assays, IGF-1 was degraded within 30 min, whereas ∼49.1% of mIGF-Gel remained intact after 6 h (Fig. 2G and H). Owing to their distinct molecular characteristics, mIGF-Gel and recombinant IGF-1 were evaluated by different analytical methods after identical proteolytic treatment; therefore, this comparison provides supportive, although not strictly equivalent, evidence for the superior proteolytic resistance of mIGF-Gel. This enhanced resistance may derive from the incorporation of D-amino acids and the β-sheet conformation formed by self-assembly, supporting the hydrogel to maximize the biological function of IGF-1 in the complex in vivo environment. This proteolytic stability may help preserve the bioactive motif of mIGF-Gel under complex biological conditions, thereby maintaining its capacity to engage IGF-1R and trigger downstream signaling at the cellular level.

Fig. 2.

Sustained IGF-1R Activation via Protease-Resistant Mimicry of mIGF-Gel. (A) Western blot analysis of pIGF-1R and IGF-1R expression levels in HCECs treated with different compounds. (B) Quantification of the pIGF-1R/IGF-1R ratio. (C) Microscale thermophoresis (MST) analysis of K_D values of mIGF-Gel and IGF-1 binding to IGF-1R. (D) Western blot analysis of pIGF-1R/IGF-1R expression in HCECs treated with different compounds after H₂O₂ induction. (E) Quantitative analysis of pIGF-1R/IGF-1R levels following H₂O₂ induction. (F) Quantitative analysis of pAKT/AKT levels following H₂O₂ induction. (G) Remaining mIGF-Gel hydrogel content over time following proteinase K digestion. (H) Remaining IGF-1 protein content over time following proteinase K digestion. All data are presented as the mean ± SD. Statistical significance was determined using one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001 and ns means not significant.

We further performed comparative proteomics to analyze protein expression changes induced by mIGF-Gel under H₂O₂-induced oxidative stress. Kyoto Encyclopedia of Genes and Genomes (KEGG) Pathway analysis indicated that the differentially expressed proteins were primarily involved in focal adhesion, PI3K-Akt signaling, regulation of the actin cytoskeleton, and apoptosis. This suggests that mIGF-Gel promotes repair by enhancing adhesion- and migration-related signaling while inhibiting apoptosis and other injury-associated responses (Fig. S13A). Consistently, gene ontology (GO) analysis highlighted the enrichment of proteins involved in cell cycle regulation, cell migration, and cell adhesion processes. These proteins were particularly associated with the PI3K-Akt signaling and integrin-mediated signaling pathways, further supporting the role of mIGF-Gel in enhancing cell–matrix interactions and basement membrane reconstruction, providing molecular insights into corneal epithelial regeneration (Fig. S13B). In summary, mIGF-Gel modulates key reparative signaling via IGF-1R activation, leveraging its exceptional protease stability to maintain signaling bioactivity and promote robust injury repair under oxidative stress.

2.3. In vitro biological activity of mIGF-Gel

2.3.1. Promotion of cell proliferation

To assess the cytocompatibility of mIGF-Gel with HCECs, cells were cultured in media supplemented with concentrations ranging from 10 to 400 nM. Cell Counting Kit-8 (CCK-8) assays revealed no detectable cytotoxicity below 400 nM (Fig. S14). Within this range, hydrogel promoted proliferation followed a bell-shaped dose-dependent curve, peaking at 50 nM (Fig. 3A). To mimic the pathological microenvironment of PCED, we induced oxidative stress in HCECs by exposure to H₂O₂ [[37], [38], [39]]. Co-treatment with mIGF-Gel or IGF-1 effectively counteracted the marked reduction in cell viability induced by H₂O₂, with the protective effect being most pronounced on day 3 (Fig. 3B and S15). Live/dead cell staining confirmed the protective effect of mIGF-Gel, comparable to IGF-1 (Fig. 3C). Moreover, analysis of the proliferation marker Ki67 revealed consistent results across Western blot and qPCR assays. H₂O₂ markedly downregulated Ki67 protein and mRNA expression, whereas mIGF-Gel restored both transcriptional and translational levels to near-normal, similar to IGF-1 (Fig. 3D and S16). Immunofluorescence imaging corroborated these findings, showing enhanced Ki67-positive nuclear signals in both the IGF-1 and mIGF-Gel groups (Fig. 3E and F). To further clarify the individual contributions of the YIGSR and SSSR motifs to the proliferative activity of mIGF-Gel, additional sequence control peptides were evaluated using the CCK-8 assay (Fig. S17). Both the YIGSR-containing peptide (Biotin-^DFYIGSRSS) and the SSSR-containing peptide (Biotin-^DFGIYSSSR) promoted HCEC proliferation compared with the scrambled control peptide (Biotin-^DFGIYSRSR). These findings suggest that both YIGSR and SSSR contribute to the biological activity of mIGF-Gel, and the dual-motif peptide showed a more favorable pro-proliferative effect than either single-motif control. Collectively, these results demonstrate that mIGF-Gel is non-toxic, effectively promotes corneal epithelial cell proliferation at relatively low concentrations, and safeguards against oxidative stress.

Fig. 3.

In vitro promotion of cell viability and proliferative capacity by mIGF-Gel. (A) Effects of mIGF-Gel on HCECs viability assessed by CCK-8 assay (n = 5). (B) Effects of mIGF-Gel on HCECs viability under H₂O₂-induced oxidative stress by the CCK-8 assay (n = 5). (C) Representative live/dead staining images of HCECs exposed to H₂O₂ and subjected to different treatments. Scale bars: 50 μm. (D) qPCR analysis of Ki67 mRNA expression in H₂O₂-induced HCECs after different treatments. (E) Representative confocal immunofluorescence images and (F) quantitative analysis of Ki67-positive epithelial cells (n = 3). Scale bars: 50 μm. All data are presented as the mean ± SD. Statistical significance was determined using one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001 and ns means not significant.

2.3.2. Suppression of apoptosis and oxidative stress

We next evaluated the anti-apoptotic efficacy of mIGF-Gel using TUNEL assays. HCECs were assigned to four groups: untreated control (DMEM), H₂O₂ treatment, H₂O₂ + mIGF-Gel co-treatment, or H₂O₂ + IGF-1 co-treatment. TUNEL assays indicated that H₂O₂ increased the apoptotic rate to 13.69%, which was reduced to 6.71% by mIGF-Gel (Fig. 4A and B). Flow cytometry using Annexin V-FITC/PI confirmed these findings: apoptosis was 14.28% with H₂O₂ versus 5.39% in control, but decreased to 8.56% with mIGF-Gel, comparable to IGF-1 (7.22%) (Fig. 4C and D). Flow cytometric analysis of DCFH-DA fluorescence showed a marked increase in intracellular ROS levels following H₂O₂ treatment (Fig. 4E). Quantitative analysis demonstrated that ROS levels increased to approximately 2.8-fold of the control group, whereas mIGF-Gel treatment reduced ROS levels by 44.73%, comparable to the reduction observed with IGF-1 treatment (48.26%) (Fig. 4G). This was further confirmed by representative fluorescence images (Fig. 4F). These findings demonstrate that mIGF-Gel exerts both anti-apoptotic and anti-oxidative effects.

Fig. 4.

In vitro protection against apoptotic, ROS, inflammation by mIGF-Gel. (A) Representative confocal immunofluorescence images and (B) quantitative analysis of TUNEL positive (green) HCECs. Scale bars: 50 μm. (C) Representative flow cytometry plots and (D) quantitative analysis of apoptosis in HCECs. (E) Flow cytometry histograms of ROS levels in HCECs following different treatments and (G) the corresponding quantitative analysis. (F) Representative confocal immunofluorescence images showing ROS (green) levels in HCECs after various treatments. Scale bars: 25 μm. qPCR analysis of (H) TNF-α, (I) IL-6, and (J) IL-1β mRNA expression in H₂O₂-induced HCECs after different treatments. (K) Immunofluorescence images of IL-1β (green) in HCECs following different treatments and (L) the corresponding statistical analysis. Scale bars: 25 μm. Nuclei was stained blue. All data are presented as the mean ± SD (n = 3). Statistical significance was determined using one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001 and ns means not significant.

2.3.3. Mitigation of inflammatory responses

Under oxidative stress, HCECs release pro-inflammatory cytokines including IL-1β, IL-6, and TNF-α, which critically modulate corneal wound healing [14,38]. qPCR analysis revealed that H₂O₂ exposure markedly upregulated IL-1β, IL-6, and TNF-α mRNA. Treatment with mIGF-Gel reduced their expression by 65.36%, 18.38%, and 43.43%, respectively, comparable to IGF-1 (Fig. 4H–J). Immunofluorescence confirmed reduced IL-1β expression in treated cells (Fig. 4K and L). These results indicate that mIGF-Gel effectively attenuates oxidative stress-induced inflammatory responses.

2.3.4. Enhancement of cell migration

Directed migration of epithelial cells is essential for re-epithelialization [12]. Using transwell and scratch assays, we demonstrated that mIGF-Gel robustly promotes HCEC migration. H₂O₂ stimulation markedly reduced the number of HCECs migrating to the lower chamber, indicating impaired cellular motility under oxidative stress. In contrast, treatment with mIGF-Gel significantly enhanced cell migration, exhibiting efficacy comparable to IGF-1 (Fig. 5A and B). Consistent results were observed in the scratch wound healing assay, where both mIGF-Gel and IGF-1 treatments notably accelerated wound closure, with no significant difference between the two groups (Fig. 5C and D). The lack of statistically significant differences between IGF-1 and mIGF-Gel suggests that current experimental models may not fully capture their distinct in vivo potentials. Alternative approaches may be needed to design cell experiments that more accurately replicate real ocular surface application scenarios, which warrants further investigation. Collectively, these findings demonstrate that mIGF-Gel effectively promotes corneal epithelial cell migration, thereby contributing to its pro-regenerative biological activity. Mechanistic analysis focused on laminin γ2 (LAMC2), a basement membrane component regulating adhesion, motility, and proliferation, as well as integrin β1 (ITGB1), a key adhesion receptor. Western blotting revealed that H₂O₂ markedly suppressed LAMC2 and ITGB1 expression. mIGF-Gel co-treatment effectively restored their expression, and notably outperformed native IGF-1 (Fig. 5I and S16). qPCR and immunofluorescence confirmed restoration of LAMC2/ITGB1 transcription and protein expression (Fig. 5E–K). Collectively, these findings indicate that mIGF-Gel drives migration by reinforcing basement membrane reconstruction and cell-matrix anchorage.

Fig. 5.

In vitro promotion of LAMC2/ITGB1 mediated epithelial migration by mIGF-Gel. (A) Representative images and (B) quantitative analysis of HCECs in lower chamber with transwell migration assay. Scale bars: 50 μm. (C) Representative image and (D) the corresponding statistical analysis of HCECs scratch assays after 12 h. Scale bars: 100 μm. (E) Immunofluorescence staining of LAMC2 in HCECs following different treatments and (F) the corresponding statistical analysis. Scale bars: 50 μm. (G) Immunofluorescence staining of ITGB1 in HCECs following different treatments and (H) the corresponding statistical analysis. Scale bars: 50 μm. (I) Representative protein bands of Ki67, LAMC2, and ITGB1 in different treatment groups. qPCR analysis of mRNA expression levels of (J) LAMC2 and (K) ITGB1 after different treatments. All data are presented as the mean ± SD (n = 3). Statistical significance was determined using one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001 and ns means not significant.

2.4. In vivo therapeutic effects of mIGF-Gel

2.4.1. Therapeutic activity of mIGF-Gel in the acute corneal epithelial injury model

The therapeutic effect of mIGF-Gel was first evaluated in a corneal epithelial debridement model with partial basement membrane disruption [40]. The model involves epithelial debridement with concurrent disruption of the underlying basement membrane and subepithelial nerve plexus, resulting in impaired epithelial adhesion and altered epithelial–neuronal homeostasis [41,42]. While this model retains intrinsic healing capacity and is useful for assessing early reparative responses after acute ocular surface injury, it alone does not recapitulate the persistent, refractory pathology characteristic of PCED. In the corneal debridement model, both the protein and mRNA levels of IGF-1 were markedly decreased following injury (Fig. S18), suggesting that depletion of endogenous IGF-1 may contribute to limited epithelial repair, a mechanism potentially relevant under more severe or chronic injury conditions. Fluorescein staining indicated that the mIGF-Gel treated group (94.10%) exhibited significantly faster wound closure at 48 h compared with the IGF-1 (86.08%) and PBS (76.88%) groups, achieving complete re-epithelialization by 72 h (Fig. 6B and C). OCT imaging further confirmed more rapid shortening of the epithelial defect length (Fig. 6D and E). Collectively, these findings indicate that mIGF-Gel promotes early corneal epithelial repair by accelerating re-epithelialization during the wound closure phase.

Fig. 6.

mIGF-Gel accelerates re-epithelialization via enhanced retention. (A) Schematic illustration of the experimental design and treatment workflow. (B) Representative images of corneal epithelial defects at indicated time points following fluorescein sodium staining after different treatments. (C) The healing rate of the corneal epithelium at different time points (n = 5). (D) The length of the corneal wound at different time points (n = 5). (E) Representative OCT images of corneal epithelial defects in rats treated with different compounds at each time point after injury. The red line segment indicates the length of the defect site. Scale bars: 250 μm. (F) Representative in vivo fluorescence images of RhB-IGF-1 and RhB-hydrogel administration in injured corneas (n = 3). All data are presented as the mean ± SD. Statistical significance was determined using one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001 versus the PBS group. ^###p < 0.001, ^####p < 0.0001 versus the IGF-1 group.

Considering the rapid spontaneous repair in this model, wound closure alone may not fully reflect the delivery-related advantages of mIGF-Gel in vivo. Ocular surface retention experiments showed that solution formulations were almost completely cleared within 5 min post-administration, whereas hydrogel formulations still exhibited fluorescence at 30 min and became undetectable by 40 min (Fig. S19). In vivo imaging system (IVIS) quantification revealed that the hydrogel group retained approximately 54% of its initial fluorescence at 30 min. Notably, mIGF-Gel remained on the ocular surface for up to 50 min, whereas the IGF-1 solution signal decayed rapidly (Fig. 6F and S20). For ocular surface drug delivery systems, longer retention is not necessarily preferable, as excessive retention may cause blurred vision, reduce patient compliance, and compromise ocular surface tolerance. Thus, effective local retention within an appropriate time window may be of greater practical relevance. Consistently, aqueous humor pharmacokinetics demonstrated that mIGF-Gel achieved 2.4- and 2.6-fold increases in maximum concentration (C_max) and area under the curve (AUC), respectively, compared with IGF-1 solution (Fig. S21). These findings suggest that mIGF-Gel accelerates wound closure while also prolonging ocular surface retention and enhancing local ocular delivery. These properties may be advantageous in more challenging injury settings, although further validation is needed.

Although mIGF-Gel and recombinant IGF-1 showed comparable effects in short-term in vitro assays, this result should be interpreted in the context of the experimental setting. In static cell culture, both treatments remain continuously available to HCECs, and therefore the advantages of hydrogel-mediated retention and proteolytic resistance are not fully reflected. These in vitro results primarily demonstrate that mIGF-Gel preserves IGF-1-like bioactivity, including activation of IGF-1R/AKT signaling and protection against oxidative injury. In contrast, the ocular surface represents a dynamic and protease-rich environment, where soluble IGF-1 can be rapidly diluted and cleared by tear turnover and blinking. Under these conditions, the self-assembled nanofibrous hydrogel provides additional benefits by stabilizing the bioactive peptide motifs, prolonging ocular surface residence, increasing local delivery efficiency, and reducing the required dosing frequency. Therefore, the therapeutic rationale of mIGF-Gel is based not on stronger acute cellular stimulation than free IGF-1 in vitro, but on its ability to maintain IGF-1-like bioactivity more effectively in the complex in vivo ocular surface microenvironment.

Histological analysis at 72 h post-injury revealed incomplete epithelial coverage in the PBS group, whereas both IGF-1 and mIGF-Gel groups achieved full re-epithelialization. Notably, the hydrogel-treated epithelium was thicker and more stratified, suggesting superior tissue reconstruction quality (Fig. 7A). Ki67 and TUNEL immunostaining showed sparse proliferating cells and abundant apoptotic cells in the PBS group (Fig. 7B, C, 7F, 7G), whereas mIGF-Gel treatment enhanced Ki67 signals and significantly reduced TUNEL-positive cells, indicating promotion of epithelial proliferation and inhibition of apoptosis. Further immunofluorescence analysis showed that mIGF-Gel reduced IL-1β expression to 49.8% of the PBS group (Fig. S22), suggesting effective attenuation of IL-1β–related inflammation and improvement of the local microenvironment [37,43]. Given that corneal epithelial repair depends on epithelial migration and adhesion [[44], [45], [46], [47]], we further examined the distribution of LAMC2 and ITGB1. LAMC2 exhibited continuous, linear distribution along the epithelial–stromal interface, with significantly higher intensity and continuity in the mIGF-Gel group compared with IGF-1 (Fig. 7D and H). ITGB1, normally polarized in basal epithelial cells, lost polarity in PBS and IGF-1 groups but was restored in the mIGF-Gel group (Fig. 7E and I).

Fig. 7.

mIGF-Gel promotes proliferative and adhesion responses alongside anti-apoptotic and anti-inflammatory effects to restore epithelial architecture. (A) H&E sections of corneas with different treatments. Scale bar: 25 μm. (B) Immunofluorescence staining of Ki67 (red) in corneas under different treatments and (F) the corresponding statistical analysis. Scale bar: 50 μm. (C) Immunofluorescence staining of TUNEL (green) in corneas under different treatments and (G) the corresponding statistical analysis. Scale bar: 50 μm. (D) Immunofluorescence staining of LAMC2 (red) in corneas with different treatments and (H) the corresponding statistical analysis. Scale bar: 50 μm. (E) Immunofluorescence staining of ITGB1 (red) in corneas with different treatments and (I) the corresponding statistical analysis. Scale bar: 50 μm. (J) Immunofluorescence staining of corneal nerve (green) in corneas with different treatments and (K) the corresponding quantification of the nerve occupation ratio in the central cornea. Scale bar: 1 mm. All data are presented as the mean ± SD (n = 3). Statistical significance was determined using one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001 and ns means not significant.

Western blot analysis further confirmed that mIGF-Gel more effectively reversed injury-induced downregulation of Ki67, LAMC2, and ITGB1, while suppressing IL-1β upregulation (Fig. S23–24). Collectively, these results suggest that mIGF-Gel may enhance corneal repair by simultaneously modulating proliferation and inflammation while improving basement membrane reconstruction.

Moreover, mIGF-Gel treatment upregulated PAX6 and CK12 expression, suggesting that it may help preserve corneal epithelial characteristics and support the differentiation of newly formed epithelium [48,49] (Fig. S25). Treatment with either IGF-1 or mIGF-Gel led to a marked increase in the nerve-positive area within the subepithelial region, with mean values reaching 28.27 ± 1.6% and 35.67 ± 1.7%, respectively, versus 10.67 ± 0.6% in the PBS group, suggesting a beneficial effect on early nerve repair responses (Fig. 7J and K). Together, these findings suggest that mIGF-Gel may contribute to early epithelial repair and support a microenvironment favorable for subsequent nerve recovery [9,50].

Overall, the in vivo activity of mIGF-Gel in this acute corneal injury model can be summarized in two aspects: first, its viscoelastic and shear-thinning properties prolong ocular surface retention and enhance intraocular drug exposure; second, even against a background of rapid spontaneous healing, it promotes proliferation, alleviates inflammation, and improves basement membrane-associated reconstruction. These findings suggest potential applicability under more complex pathological conditions. However, its therapeutic robustness requires further validation under impaired healing conditions. Based on this, we established a dexamethasone-suppressed delayed-healing model to systematically evaluate the advantages of mIGF-Gel in a compromised microenvironment more representative of PCED pathology.

2.4.2. Therapeutic activity of mIGF-Gel in the delayed-healing corneal defect model

Given the strong inherent tendency for spontaneous healing in mechanical corneal epithelial debridement models, we further established a dexamethasone-induced delayed-healing corneal epithelial defect model to evaluate the reparative capacity of mIGF-Gel in a challenging ocular surface microenvironment. Following epithelial debridement, dexamethasone was applied to inhibit re-epithelialization, thereby creating a more challenging compromised ocular surface microenvironment [2,51,52]. In this model, in addition to PBS, IGF-1, and mIGF-Gel groups, we included single-motif controls, scrambled peptide, and blank hydrogel groups to assess the effects of dual-motif combination, single-motif activity, and non-specific hydrogel effects on in vivo repair outcomes. Corneal wound closure kinetics, histological repair quality, and representative markers associated with epithelial proliferation, adhesion anchoring, nerve regeneration, and corneal sensation were systematically evaluated.

In the delayed-healing model, slit-lamp fluorescein staining revealed that the mIGF-Gel group exhibited faster wound contraction at days 1, 3, 5, and 7, achieving complete re-epithelialization by day 5. In contrast, single-motif controls, scrambled peptide, and blank hydrogel groups healed more slowly, reaching complete or near-complete closure by day 7, while the PBS group remained incompletely healed (Fig. 8B). Quantitative analysis showed that under a once-daily administration regimen, the mIGF-Gel group achieved a healing rate of 96.35 ± 1.06%, comparable to the three-times-daily IGF-1 treatment group (97.37 ± 2.30%) (Fig. 8C). H&E staining further demonstrated that the PBS group exhibited persistent epithelial discontinuity, whereas single-motif, scrambled peptide, and blank hydrogel groups partially promoted epithelial coverage, but the newly formed epithelium was thin, with poor stratification and organization. In contrast, mIGF-Gel not only achieved more complete epithelial coverage but also restored a more organized multilayered structure (Fig. 8D). These results indicate that mIGF-Gel can achieve IGF-1-comparable repair with lower dosing frequency under delayed-healing conditions, while further improving epithelial reconstruction quality.

Fig. 8.

Therapeutic activity of mIGF-Gel in the delayed-healing corneal defect model. (A) Schematic illustration of the experimental design and treatment workflow. (B) Representative images of corneal epithelial defects with different treatments. (C) The healing rate of the corneal epithelium with different treatments. (D) H&E sections of corneas with different treatments. Scale bar: 50 μm. Immunofluorescence staining of (E) Ki67 (red), (F) LAMC2 (green), and (G) ITGB1 (green) in corneas with different treatments. Scale bar: 50 μm. (H) Immunofluorescence staining of corneal nerve (green) with different treatments. Scale bar: 1 mm. Statistical analysis of (I) Ki67, (J) LAMC2, (K) ITGB1, (L) the cornea nerve occupation ratio, and (M) corneal sensitivity. All data are presented as the mean ± SD (n = 3). Statistical significance was determined using one-way ANOVA. ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001.

Notably, the pro-healing activity of mIGF-Gel cannot be simply attributed to the nonspecific physical coverage provided by the hydrogel matrix, but rather arises from the coordinated contribution of its dual bioactive motifs and self-assembled nanofibrous architecture. Physicochemical characterization of control groups supports this structure–activity relationship: the dual-motif Biotin-^DFYIGSSSR (mIGF-Gel), integrating YIGSR and SSSR, self-assembles into a transparent hydrogel predominantly adopting β-sheet secondary structures. Single-motif peptides, scrambled sequences, or blank hydrogels, however, failed to simultaneously satisfy stable gelation, good solubility, and effective motif presentation. Previous studies indicate that self-assembling peptide behavior is strongly sequence- and context-dependent, with both sequence background and environmental conditions dictating secondary structure, supramolecular assembly, and functional epitope presentation. Moreover, bioactive sequences such as SSSR exhibit spatially dependent effects when incorporated into scaffold networks. Therefore, the superior repair performance of mIGF-Gel over single-motif, scrambled, and blank hydrogels likely arises from the effective delivery and combined bioactivity of the dual motifs in a suitable assembled state, rather than from simple hydrogel coverage [53,54].

Ki67 immunofluorescence showed that on day 7 post-surgery, the mIGF-Gel group retained strong Ki67-positive signals in the corneal epithelium, comparable to the IGF-1 positive control, and significantly higher than single-motif, scrambled peptide, and blank hydrogel groups. This indicates that mIGF-Gel maintains sustained proliferative activity even at low dosing frequency (Fig. 8E and I). Further immunofluorescence analysis revealed significant recovery of LAMC2 and ITGB1 expression in the mIGF-Gel group, mirroring the IGF-1 group. Spatially, LAMC2 was continuously distributed along the interface between regenerated epithelium and underlying stroma, while ITGB1 localized specifically in basal epithelial cells. In other control groups, although partial recovery was observed, signal continuity and interface localization were notably weaker (Fig. 8F–G, 8J-K). Together with wound closure data, these findings suggest that the advantages of mIGF-Gel extend beyond achieving wound closure to continuously improving epithelial–basement membrane interactions and adhesion anchoring, supporting higher-quality tissue reconstruction.

Epithelial maturation, homeostasis, and nerve regeneration are also critical indicators of corneal repair in delayed-healing models. Supporting information shows that CK12 and PAX6 staining demonstrated improved trends in epithelial differentiation and homeostasis in the mIGF-Gel group. As CK12 is a key marker of mature corneal epithelium, and PAX6 is closely linked to epithelial identity maintenance and homeostatic regulation, these findings suggest that mIGF-Gel promotes the differentiation and functional restoration of the regenerating corneal epithelium (Fig. S26). Whole-mount corneal nerve staining further showed that the mIGF-Gel group exhibited a significantly higher β-tubulin-positive nerve area percentage at day 7 than the IGF-1, single-motif, scrambled peptide, and blank hydrogel groups (Fig. 8H and L). Correspondingly, corneal sensation tests indicated superior sensory recovery in the mIGF-Gel group (Fig. 8M). These results demonstrate that the therapeutic advantages of mIGF-Gel are not limited to surface wound closure but extend to more complete tissue restoration. Given prior reports that YIGSR-containing materials promote epithelial stratification and support nerve ingrowth [55], while SSSR motifs facilitate corneal epithelial repair [27], the improved advantage of mIGF-Gel in nerve and sensation recovery likely reflects the combined presentation of adhesion anchoring and migratory repair signals within a single material framework.

In summary, the superior repair efficacy of mIGF-Gel in delayed-healing conditions is not achievable by single motifs or non-specific hydrogels alone, but likely arises from the combined presentation of specific dual motifs in an appropriate assembled state. By simultaneously supporting migration, adhesion anchoring, nerve regeneration, and epithelial maturation, mIGF-Gel ultimately achieves more complete corneal repair in a compromised healing microenvironment.

2.5. Safety assessment of mIGF-Gel

To assess clinical translational potential, we systematically evaluated the biosafety of mIGF-Gel. CCK-8 assays revealed that cell viability of both corneal epithelial and stromal cells remained unaffected at concentrations below 400 nM, indicating excellent cellular tolerance (Fig. S27). Live/dead cell staining confirmed the protective effect of mIGF-Gel on human corneal stromal cells (HCSCs) under H₂O₂ exposure (Fig. S28). Subsequent ocular irritation testing further demonstrated that topical instillation of the hydrogel onto the ocular surface of rats elicited no signs of conjunctival hyperemia, secretion, or corneal opacity. Fluorescein staining confirmed no additional epithelial damage, with corneal transparency fully maintained (Fig. S29). Histopathological examination of major organs (i.e., heart, liver, spleen, lung, and kidney) by H&E staining revealed no inflammatory infiltration, necrosis, or structural abnormalities (Fig. S30). Although the present safety results support the good local and systemic biocompatibility of mIGF-Gel and suggest a low risk of long-term residual toxicity, the exact in vivo metabolic fate and final clearance pathway of the hydrogel were not directly investigated in the present study. Collectively, these findings verify the excellent biocompatibility and biosafety of mIGF-Gel both in vitro and in vivo, with no evidence of ocular or systemic toxicity. Combined with its physicochemical stability and potent reparative efficacy, the hydrogel represents a clinically translatable platform for safe and effective ocular surface regeneration.

3. Conclusion

In conclusion, mIGF-Gel is a dual-functional self-assembling peptide hydrogel that integrates laminin-derived adhesion and IGF-1-mimetic bioactive motifs within a single molecular framework. The resulting β-sheet-rich nanofibrous network provides shear-thinning, viscoelastic, and self-recovering properties, enhancing ocular retention, proteolytic stability, and sustained pro-repair stimulation under dynamic corneal conditions. Functionally, mIGF-Gel retained IGF-1-like bioactivity and promoted epithelial repair in vitro. Its pro-healing capacity was first demonstrated in an acute corneal epithelial injury model and further confirmed in a dexamethasone-induced delayed-healing model that more closely recapitulates the impaired epithelial repair characteristic of persistent corneal epithelial defects. Collectively, these results highlight mIGF-Gel as a structurally adaptive, regeneration-promoting peptide hydrogel that demonstrates potential in compromised corneal healing conditions and suggests broader applicability for ocular surface regeneration. This study establishes a modular design strategy for bioactive, wound-adaptive peptide hydrogels and offers a versatile platform for ocular surface regeneration.

4. Experimental section

4.1. Materials

The 2-chlorotrityl chloride resin and N-Fmoc-protected amino acids were purchased from GL Biochem Ltd. (Shanghai, China). Polyvinylidene difluoride (PVDF) membranes were purchased from Millipore (Darmstadt, Germany). Transwell chambers and confocal dishes were sourced from Corning (New York, USA). Crystal violet staining solution, ROS assay kit, proteinase K, blocking buffer, antibody dilution buffer, antifade mounting medium with DAPI, and recombinant human IGF-1 were obtained from Beyotime Biotechnology (Shanghai, China).

4.2. Synthesis and characterization of peptides

4.2.1. Synthesis of peptides

Standard solid-phase peptide synthesis served as the method for preparing all peptide derivatives, with 2-chlorotrityl chloride resin acting as the solid support and relevant N-terminal fluorenylmethoxycarbonyl (Fmoc)-protected amino acids were incorporated during the process. Peptides were cleaved off the resin matrix; thereafter, the crude products were harvested via filtration and subjected to vacuum drying. Purified peptides were obtained via preparative HPLC and then evaluated by HR-MS to confirm correct mass assignment and overall sample purity.

4.2.2. Transmission Electron Microscopy (TEM)

We dissolved a defined quantity of the peptide in PBS to a final concentration of 4 mg/mL, and added Na₂CO₃ to adjust the solution pH to 7.4. The peptide solution was heated at 100 °C for 3 min until a homogeneous solution was obtained and then cooled to room temperature to induce hydrogel formation. A 10 μL aliquot was deposited on a copper grid, with negative staining using uranyl acetate. After drying, the prepared samples were examined using a transmission electron microscope.

4.2.3. Circular Dichroism (CD) spectra

For CD measurements, 500 μL of peptide solution (1 mg/mL) in a quartz cell (path length, 0.2 cm) was scanned from 190 to 260 nm via a circular dichroism spectrometer (BioLogic MOS-450). Secondary-structure fractions were subsequently estimated via the DichroWeb online platform using the CONTINLL algorithm.

4.2.4. Critical Aggregation Concentration (CAC)

Laser light scattering measurements for CAC determination were carried out on a Brookhaven system (ZETAPALS/BI-200SM, USA). A series of ten sample solutions with descending concentrations were prepared in dust-free PBS, with each sample volume set to approximately 150 μL. For measurement, 100 μL aliquots of each solution were transferred to a scattering cuvette. The scattered light intensity of each solution was recorded in order from the highest to the lowest concentration. By plotting the scattered light intensity as a function of sample concentration, the CAC value was derived from the inflection point of the resulting curve.

4.2.5. Rheological tests

Hydrogel rheology was characterized on an Anton Paar rheometer. A PP25 parallel-plate (25 mm diameter) was used for tests. An oscillatory strain sweep was initially conducted at 1 rad s⁻¹. Time-dependent behavior was then monitored by a time sweep at 1 Hz under 1% strain. A frequency sweep was performed over the range of 0.1 to 10 rad s⁻¹ at a constant 1% strain to evaluate viscoelasticity. Thixotropy and self-healing characteristics were examined at 25 °C and 1 Hz by imposing a high-strain disturbance. In addition, a step-strain test of hydrogels was conducted by alternately applying 0.1% and 100% strain at 25 °C.

4.2.6. Fourier-Transform Infrared Spectroscopy (FTIR)

FTIR spectra were recorded using a Bruker TENSOR 37 Fourier-transform infrared spectrometer (Bruker, Germany). For sample preparation, 100 μL of 0.4 wt% peptide solution was rapidly frozen in liquid nitrogen and subsequently freeze-dried to obtain a powdered sample. The dried peptide powder was thoroughly mixed with potassium bromide and pressed into transparent KBr/peptide pellets for subsequent FTIR analysis.

4.2.7. Microscale Thermophoresis (MST)

The dye NT-647 NHS was used to fluorescently label recombinant human IGF-1R. The peptide was serially diluted and mixed with the labeled protein. Following a 30 min incubation at room temperature, each mixture was loaded into an MST capillary and analyzed on a Monolith NT.115 instrument. Binding affinities were quantified by fitting the thermophoresis data to obtain the equilibrium dissociation constant (K_D) for the peptide–IGF-1R interaction.

4.2.8. Chemical stability

Proteolytic stability of the peptide amphiphiles was evaluated in a Proteinase K challenge assay. Peptide amphiphile solution (2 mM) was combined with 2 mg/mL proteinase K at a 1:1 (v/v) ratio, followed by incubation at 37 °C. At designated sampling times, 50 μL of the sample was withdrawn and dissolved in 500 μL of methanol, filtered through 0.45 μm organic filter membrane, and detected by LC-MS. Degradation rates were determined by integrating the target peptide peak.

4.2.9. Stability assessment of IGF-1 protein

IGF-1 protein (500 μg) was reconstituted in 70 μL of PBS, and treated with proteinase K solution at 2 mg/mL, followed by incubation at 37 °C in a water bath. Aliquots of 10 μL were taken from the reaction mixture at each predetermined time point; these aliquots were then mixed rapidly and thoroughly with an equal volume of 2× loading buffer before being transferred to a metal bath for thermal denaturation. The denatured samples were separated by gel electrophoresis until the loading buffer completely migrated out of the gel. After staining with Coomassie Brilliant Blue and subsequent washing, the gel was observed visually.

4.3. In vitro cell experiments

4.3.1. In vitro cytotoxicity assays

For CCK-8 assay, HCECs and HCSCs were separately plated overnight in 96-well cell culture plates at 5 × 10³ and 2 × 10³ cells/well. The medium was replaced with 100 μL of mIGF-Gel peptide solution with varying concentrations, followed by incubation for 24, 48, and 72 h. The degree of cytotoxicity was determined according to the CCK-8 assay protocol. For the live/dead assay, cells (1 × 10⁶ cells/mL) were seeded in 6-well plates. Cells were treated for 24 h with four conditions: low-serum DMEM/F-12 medium (1% serum, control), H₂O₂ (50 μM), H₂O₂+IGF-1 (10 ng/mL), or H₂O₂+mIGF-Gel (50 nM), each at a 2 mL volume. After staining as instructed, fluorescent images were acquired by microscopy and quantified using ImageJ software.

4.3.2. Transwell cell migration assay

For the migration test, HCECs (5 × 10⁵ cells/mL) were seeded into the upper Transwell chambers in 24-well plates. The lower chambers contained low-serum DMEM/F-12 medium (1% serum, control), H₂O₂ (50 μM), IGF-1 (10 ng/mL), or mIGF-Gel (50 nM), followed by an 18 h incubation. The inserts were subjected to fixation with 4% paraformaldehyde (PFA) for 30 min, followed by a stain step using 1% crystal violet. Cells that had not migrated and remained attached to the upper side of the insert membrane were carefully removed. Migrated cells were imaged using an inverted microscope (Leica DMi8, Germany), and quantified using ImageJ software.

4.3.3. Scratch assay

HCECs were cultured in 6-well plates to allow monolayer formation. A sterile 200 μL pipette tip was used to create a linear scratch across the confluent cell layer to simulate a wound. Cells were treated with four conditions for 12 h. Images of the wounded area were captured at identical locations under a microscope. Wound closure was quantified by measuring the remaining gap width with ImageJ software.

4.3.4. Immunocytochemical staining

HCECs were cultured in confocal dishes overnight. After PBS washing, cells were exposed to four conditions for 24 h. Cells were then fixed with 4% PFA for 30 min. Primary antibodies against Ki67 (ABclonal, A20018), IL-1β (Abcam, ab9722), ITGB1 (Servicebio, GB115629-100), and LAMC2 (Santa Cruz, sc-28330) at a 1:500 dilution were applied overnight at 4 °C after permeabilization and blocking. Subsequently, the samples were incubated with secondary antibodies for 1 h at room temperature, and counterstained with DAPI for 10 min. Fluorescence micrographs were captured using an SP8 laser scanning confocal microscope and processed for quantitative analysis with ImageJ software.

4.3.5. Cell apoptosis assay

For the flow cytometry assay, HCECs in 6-well plates were treated for 6 h with four conditions. Cells were then harvested, stained following the protocol, and then analyzed on a flow cytometer. Data were processed using FlowJo. For the TUNEL assay, treated HCECs were fixed with 4% PFA, permeabilized with 0.1% Triton X-100, and incubated with TUNEL reaction solution in the dark for 1 h. Fluorescent images were acquired by a confocal microscope and quantified using ImageJ software.

4.3.6. Cell ROS

HCECs were inoculated in confocal dishes at 5 × 10⁴ cells/mL, followed by 24 h of intervention with four conditions. Post-treatment, the cells were subjected to PBS rinsing before being exposed to DCFH-DA, which was conducted at 37 °C for 30 min in a light-shielded container. After PBS washing, images were captured. For quantitative analysis, cells were also collected, processed, and subjected to flow cytometric detection, with all data analyzed by FlowJo software.

4.3.7. Western blot

For protein expression analysis, HCECs subjected to different treatments were lysed on ice using RIPA buffer containing protease and phosphatase inhibitors. The resulting cell lysates were centrifuged at 13,000 rpm and 4 °C for 15 min to collect the supernatant. Protein concentrations were measured with a BCA assay. SDS-PAGE was used to separate equal loads of protein, followed by electrophoretic transfer to PVDF membranes. The membranes were blocked and incubated overnight at 4 °C with the primary antibodies: Ki67 (ABclonal, A20018), LAMC2 (ABclonal, A1869), ITGB1 (ABclonal, A23497), IGF-1R (ABclonal, A0243), pIGF-1R (ABclonal, AP0367), AKT (ABclonal, A18675, 1:1000), and P-AKT (ABclonal, AP1453) at a 1:1000 dilution, and GAPDH (Boster, A00227-1, 1:50,000). After extensive washing with TBST, membranes were incubated with secondary antibodies for 2 h at room temperature. Finally, protein bands were visualized and analyzed for relative protein expression.

4.3.8. Real time-quantitative PCR (RT-qPCR)

For transcriptional analysis, total RNA was obtained from treated HCECs by a TriZol reagent kit (SBS, Beijing, China). The quality of the extracted RNA, in terms of concentration and purity, was assessed prior to downstream processing. Reverse transcription was carried out to generate cDNA, followed by quantitative PCR using a commercial PCR system (TransGen, Beijing, China). The 2^−ΔΔCt method was applied to calculate relative gene expression. Detailed primer sequences used in this assay are provided in Table S1.

4.3.9. Liquid Chromatography-Mass Spectrometry (LC–MS/MS) analysis

To characterize the differential protein expression of treated HCECs, proteins extracted from cells were quantified by BCA assay. Following reduction with DTT and alkylation with IAA, samples were digested using the FASP method. After buffer exchange into NH₄HCO₃, tryptic digestion proceeded overnight at 37 °C. Peptides generated from digestion were desalted on C18 stage-tips, lyophilized, and redissolved in 0.1% formic acid solution for mass spectrometry analysis. An EASY-nLC 1200 system online-connected to a Q Exactive HF-X mass spectrometer (both from Thermo Fisher Scientific) fitted with a nanoelectrospray ion source was used for peptide separation. Raw mass spectrometry data were analyzed using Spectronaut (v18). KEGG and GO enrichment analyses were conducted using R (v4.1.2, x64).

4.4. In vivo animal studies

Male Sprague-Dawley (SD) rats (200 ± 10 g) were obtained from SiPeiFu Biotechnology Co. Ltd. (Beijing, China). All animal-related experimental operations complied fully with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. These experimental procedures were authorized by the Ethics Committee of the School of Medicine, Nankai University, under the valid approval license (2025-SYDWLL-000644).

4.4.1. Corneal injury model

Prior to surgery, rats received general anesthesia via intravenous sodium pentobarbital (50 mg/kg), and the corneal surface was anesthetized with topical instillation of 0.5% proparacaine hydrochloride eye drops. An AlgerBrush II corneal debrider was employed to mechanically remove the corneal epithelial layer and the underlying basement membrane, followed by irrigation with sterile saline. Following model establishment, animals were randomly allocated to three experimental groups and treated topically twice daily: the PBS-treated control group, the IGF-1 eye drop group, and the mIGF-Gel hydrogel eye drop group (10 μL each). Corneal healing was evaluated by slit-lamp imaging and fluorescein staining. Rats were sacrificed, and eyeballs were collected for histological analysis on day 3 post-injury. For the delayed wound healing model, based on the above model, dexamethasone eye drops were administered daily to suppress wound healing. After injury, each group received the respective treatment as eye drops, and corneal healing was assessed on days 1, 3, and 7. On day 7, the eyes were harvested for histological analysis.

4.4.2. Measurement of IGF-1 level in the cornea

We determined endogenous IGF-1 content in rat cornea tissues using an enzyme-linked immunosorbent assay (ELISA) kit (Elabscience Biotechnology Co., Ltd.). Final concentrations were derived by mapping the OD of each sample against the standard curve.

4.4.3. In vivo precorneal retention assay

Fluorescein sodium-labeled IGF-1 solution and mIGF-Gel hydrogel were carefully delivered to the conjunctival sac of rat eyes at a volume of 10 μL per eye. The residence of fluorescent signals on the ocular surface was examined at designated time points using slit-lamp biomicroscopy under blue-light illumination. Furthermore, Rhodamine B (RhB) was incorporated into all formulations as a fluorescent marker. Precorneal retention was evaluated by in vivo fluorescence imaging, and subsequent image analysis was performed using ImageJ software to determine the retention efficacy.

4.4.4. In vivo pharmacokinetics assay

New Zealand White rabbits with experimentally induced corneal injury were randomly assigned to two groups. Rabbit eyes were topically treated with either mIGF-Gel hydrogel or IGF-1 solution at a dose of 50 μL per eye. At defined time points following administration, the aqueous humor samples (20 μL) were collected using a 29G insulin syringe. The collected samples were mixed with 80 μL of HPLC mobile phase and then centrifuged at 13,000 rpm for 10 min to remove cellular debris and impurities. To determine the drug concentration, the supernatant obtained from the aqueous humor was analyzed by HPLC.

4.4.5. Histological assessment and immunofluorescence

Eyeballs were harvested 3 days post-operation. Paraffin-embedded corneal sections (5 μm thick) were prepared and subjected to standard hematoxylin-eosin (H&E) staining. Alternatively, 10 μm frozen sections were prepared from eyes fixed in 4% PFA (4 h) for immunofluorescence assays. Sections were incubated overnight at 4 °C with primary antibodies against Ki67 (ABclonal, A20018), IL-1β (Abcam, ab9722), LAMC2 (Santa Cruz Biotechnology, sc-28330), ITGB1 (Servicebio, GB115629-100), PAX6 (Proteintech, 12323-1-AP), and CK12 (ABclonal, A19284) at a 1:500 dilution after permeabilization and blocking. After washing, the secondary antibodies were applied for 2 h at room temperature. Images were acquired using an SP8 confocal microscope and analyzed using ImageJ software. Whole-mount corneas were utilized for nerve staining. To reduce nonspecific antibody binding, the tissues were incubated in a blocking solution containing PBS, 0.1% Triton X-100, and 2% goat serum after fixation with 4% PFA. The whole-mount corneas were then incubated at 4 °C overnight with Tubulin β-3 (BioLegend, 802001, 1:200), followed by labeling with secondary antibody for 2 h. After radial cutting, nerve morphology was imaged by fluorescence microscopy, and nerve fiber coverage was quantified using ImageJ. For each sample, the central cornea was imaged under identical acquisition settings. Images were converted to grayscale, and a uniform threshold was applied to segment β-tubulin-positive pixels (the same threshold was used for all images in a batch). The nerve-occupied area fraction was calculated as the percentage of β-tubulin-positive pixels relative to the total area of the corneal epithelium in each sample. The mean value of three samples per group was used for statistical analysis with ImageJ.

4.4.6. Western blot

Rat corneas were harvested on day 3. Tissues were lysed in 200 μL of RIPA buffer (Solarbio, China), homogenized and sonicated on ice. The lysates were clarified via centrifugation at 12,000 rpm for 15 min at 4 °C, and supernatants were harvested for protein quantification using a BCA kit. We separated equal amounts of protein by SDS-PAGE and electroblotted them to PVDF membranes. After blocking, the membranes were probed overnight at 4 °C with primary antibodies against Ki67 (ABclonal, A20018), LAMC2 (ABclonal, A1869), ITGB1 (ABclonal, A23497), and IL-1β (Abcam, ab9722) at a 1:1000 dilution, and GAPDH (Boster, A00227-1) at 1:50,000 dilution. The membranes were probed for 1 h at room temperature with secondary antibodies. Following additional washes, the protein bands were visualized and analyzed.

4.5. Biosafety assessment experiments

The assessment of cytotoxic responses in vivo has been extensively described in previous methodological reports. Rats received topical administration of PBS or hydrogel into the right conjunctival sac twice daily. On day 7 post-administration, corneal integrity was assessed by staining with 5 μL of sodium fluorescein. Thereafter, corneal tissues together with major internal organs were excised and processed for histopathological analysis.

4.6. Statistical analysis

Statistical evaluations were conducted using GraphPad Prism 8 software (GraphPad Software, CA, USA). Experimental results are presented as means ± standard deviation from at least three independent biological experiments. Prior to conducting ANOVA, datasets were examined to ensure compliance with assumptions of normal distribution and equal variance. Statistical comparisons between groups were performed using one-way ANOVA or two-way ANOVA. Significant main effects were further determined using Tukey's honest significant difference (HSD) test and Sidak's multiple comparison test, respectively. A p-value of less than 0.05 was considered statistically significant.

Data availability statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Ethics approval and consent to participate

All animal-related experimental operations complied fully with the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. These experimental procedures were authorized by the Ethics Committee of the School of Medicine, Nankai University, under the valid approval license (2025-SYDWLL-000644).

CRediT authorship contribution statement

Mengyuan Shan: Conceptualization, Data curation, Investigation, Methodology, Writing – original draft. Yinghao Ding: Conceptualization, Data curation, Methodology, Writing – original draft. Ping Lu: Data curation, Methodology. Fuqi Deng: Data curation, Methodology. Le Chang: Project administration. Qin Zhou: Project administration. Xinyu Li: Project administration. He Tian: Project administration. Qimiao Wang: Investigation, Software. Yi Dong: Investigation, Software. Zhimou Yang: Funding acquisition, Supervision, Writing – review & editing. Jiaqing Wang: Funding acquisition, Supervision, Writing – review & editing. Yan Wang: Funding acquisition, Supervision, Writing – review & editing.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

The following is the Supplementary data to this article: