Wearable electrodriven switch actively delivers macromolecular drugs to fundus in non-invasive and controllable manners

doi:10.21203/rs.3.rs-5120119/v1

Download PDF

Article

Wearable electrodriven switch actively delivers macromolecular drugs to fundus in non-invasive and controllable manners

https://doi.org/10.21203/rs.3.rs-5120119/v1

This work is licensed under a CC BY 4.0 License

Version 1

posted

You are reading this latest preprint version

Current treatments for fundus disorders, such as intravitreal injections, pose risks, including infection and retinal detachment, and are limited in their ability to deliver macromolecular drugs across the blood‒retinal barrier. Although non-invasive methods are safer, their delivery efficiency remains suboptimal (< 5%). We have developed a wearable electrodriven switch (WES) that improves the non-invasive delivery of macromolecules to the fundus. The WES system, which integrates an electrodriven drug delivery lens with a square wave generator, leverages electrical stimulation to enhance drug penetration through the sclera-choroid-retina pathway. In our study, WES achieved a delivery efficiency of 14% for immunoglobulin G, comparable to that of intravitreal injection (16%). Moreover, WES-enhanced anti-VEGF administration resulted in an 86% inhibition of choroidal neovascularization, and anti-PDL1 delivery inhibited choroidal melanoma growth more effectively than intravenous injections, with no adverse effects on ocular health. These findings suggest that WES holds transformative potential for the non-invasive treatment of chronic fundus diseases.

Health sciences/Diseases/Eye diseases/Macular degeneration

Biological sciences/Biotechnology/Protein delivery

Physical sciences/Nanoscience and technology/Nanoscale devices/Electronic devices

Physical sciences/Engineering/Biomedical engineering

Many fundus diseases, such as choroidal neovascularization and uveal melanoma, require invasive treatments because of the complex barriers within the eye. For example, choroidal neovascularization (CNV), which is often caused by wet age-related macular degeneration (AMD), trauma, and inflammation, leads to irreversible vision loss and blindness^1,2. Currently, anti-vascular endothelial growth factor (anti-VEGF) biomacromolecules, such as bevacizumab, are the most effective treatments for CNV, as they inhibit neovascularization^3–5. However, the presence of the blood‒retinal barrier (BRB) necessitates invasive intravitreal injection for the delivery of these macromolecular drugs^6–8. On the other hand, uveal melanoma, the most common primary fundus malignancy in adults^9,10, often requires invasive treatments such as intraocular injections or surgical removal because of the complex ocular barriers^11,12. However, repeated intravitreal injection are gloomily shadowed by the high risk of endophthalmitis infection, bulbar bleeding and retinal detachment in patients^13,14. These injections also have a short duration of action for small molecules and low target bioavailability for many protein- and gene-based drugs and nanomedicines^15–17. The inner limiting membrane (ILM) also limits access of intravitreal materials into the retina⁸. Alternative methods, such as subretinal and suprachoroidal injections, are more technically demanding and less frequently used^7,8.

To address these challenges, non-invasive drug delivery platforms are crucial for treating multiple posterior segment ocular diseases. Although eye drops or gels based on anionic polymers, membrane-penetrating peptides and exosomes have been developed as auxiliary delivery agents^18–21, they still have limitations in enhancing the intraocular penetration of macromolecular therapeutic agents and are not conducive to clinical dissemination because of their low drug utilization, easy clearance from the ocular surface and high systemic toxicity^22–23. Intriguingly, bioelectricity, a fundamental physiological signal, has shown promise in various disease interventions^24–27. Exogenous electrical stimulation has significant effects on tumour treatment^28,29, neurological disorders^30–32, and tissue repair^33,34. Iontophoresis, which leverages electrophoresis for drug penetration, offers a safe, reliable, and nontoxic method for responsive drug delivery^35,36. Studies have demonstrated the ability of iontophoresis to enhance small molecule drug penetration in the cornea, promoting treatments for anterior segment diseases^37,38. However, the ability of electrical stimulation to enhance intraocular drug delivery, especially in the fundus, remains unproven.

To fill this gap, we developed a wearable electrodriven switch (WES) designed to significantly enhance non-invasive therapeutic outcomes for fundus diseases (Fig. 1a). The WES system comprised an electrodriven drug delivery lens (EDDL) and a square wave generator (SWG) (Fig. 1b and Supplementary Fig. 1). The SWG, which was embedded on a compact flexible circuit board, generated a monophasic square pulse current (PC) signal from a direct current (DC) input via an NE555-type timer chip and the corresponding circuit design (Fig. 1c and Supplementary Fig. 2). This signal was then transmitted to the EDDL, where electrodes were embedded into a polydimethylsiloxane (PDMS) lens through cast-molding (Fig. 1d and Supplementary Fig. 3). WES-induced electrical stimulation facilitated a rapid increase of intracellular calcium ions (Ca²⁺), promoting the contraction of the perijunctional actomyosin ring (PAMR) and the redistribution of tight junction (TJ) and adherens junction (AJ) proteins in adult retinal pigment epithelial-19 (ARPE-19) cells. This mechanism significantly enhanced the delivery of macromolecular drugs to the fundus via the sclera-choroid-retina pathway. Crucially, the parameters of WES could be adjusted to control the degree and duration of cellular junction opening (Fig. 1e).

Additionally, WES-triggered iontophoresis accelerated drug distribution within the vitreous, surpassing the limitations of slow diffusion observed with intravitreal injection (Fig. 1f). The efficacy of WES has been demonstrated with various macromolecular drugs, including monoclonal antibodies, DNA origami, and extracellular vesicles. Notably, compared with intravenous injection, WES-enhanced ocular administration of anti-VEGF agents resulted in superior inhibition of CNV. Furthermore, WES-enhanced anti–programmed cell death ligand 1 (anti-PDL1) agent administration was more effective than intravenous injection in treating early-stage choroidal melanoma in mice, highlighting the platform's potential for melanoma control after excision (Fig. 1f). Long-term studies in mice confirmed the safety of WES, with no adverse effects on intraocular pressure or vision. Thus, our findings strongly supported the capacity of WES for efficient macromolecular drug delivery and its non-invasive therapeutic efficacy in treating CNV and choroidal melanoma.

The capability of WES to deliver macromolecular drugs to fundus

To ensure the safety of WES, we compared the effects of DC and PC signals on the cytotoxicity of ARPE-19 cells and human umbilical vascular endothelial cells (HUVECs). Our results revealed that 80 µA DC stimulation caused significant cytotoxicity in ARPE-19 cells after 10 minutes, whereas 80 µA PC (0.024 Hz) signals did not exhibit cellular toxicity (P = 0.0133) (Supplementary Fig. 4a). Similarly, DC caused significant cytotoxicity in HUVECs, while no toxicity under the PC treatment (P < 0.0001) (Supplementary Fig. 4a). This finding could be attributed to the fact that DC stimulation tends to induce adverse electrochemical reactions and rapid pH shifts near the electrodes, resulting in cellular damage²⁵. Conversely, varying the frequency of PC signals from WES did not result in significant cytotoxicity of ARPE-19 cells (Supplementary Fig. 4b). Furthermore, the MTT assay results confirmed the excellent cytocompatibility of WES in ARPE-19 cells and HUVECs (Supplementary Fig. 4c, d). Moreover, WES was comfortable and safe, causing no significant temperature changes in the eye or surrounding tissues (Supplementary Fig. 5).

We next used immunoglobulin G (IgG), a 150 kDa molecule like anti-VEGF and anti-PDL1, as a model drug to evaluate the ability of WES to deliver macromolecular drugs to the fundus. For visualization purposes, fluorescein isothiocyanate labelled lgG (FITC-lgG) solution was applied to the Au electrode surface of the EDDL and dried to form a drug storage layer. The EDDL loaded with FITC-IgG was then placed on the eyeballs of the experimental animals. By adjusting the stimulation time and frequency, we determined the optimal PC parameters. The results demonstrated that increasing the electrical stimulation time of WES significantly enhanced the delivery of FITC-IgG (Fig. 2a and Supplementary Fig. 6). Compared with EDDL, WES (0.73 Hz, 80 µA) markedly facilitated IgG delivery to the fundus with more than 5 minutes of WES (P = 0.0008) (Fig. 2a). Specifically, 10 minutes of WES at 0.024 Hz dramatically improved FITC-IgG penetration compared with 10 minutes at 0.73 Hz (P = 0.0204) (Fig. 2b and Supplementary Fig. 7). WES-enabled FITC-IgG even penetrated the retinal pigment epithelium (RPE) layer and reached the retina. Representative confocal fundus images from the WES group revealed stronger fluorescence signals than those from the intravitreal injection group did (P = 0.0021), highlighting the efficacy of WES in non-invasive macromolecular drug delivery to the fundus (Fig. 2c, d and Supplementary Fig. 8).

We also evaluated the ability of WES to deliver various macromolecular drugs to the fundus. As depicted in Fig. 2e, f, we designed 5-carboxytetramethylrhodamine-labelled rectangular DNA origami sheets (TAMRA-rDOSs) and DiD-labelled extracellular vesicles (DiD-EVs). Design and DNA Sequence of TAMRA-labelled strands was shown in supplementary information (Supplementary Fig. 9 and Supplementary Table 1, 2). The rDOSs had extended TAMRA fluorophores at their 5’-ends, with dimensions of ~ 80 nm (Fig. 2e). The diameter of EVs was ~ 100 nm (Fig. 2f). Both the rDOSs and EVs were well stabilized, with zeta potentials of approximately − 3.5 mV and − 17 mV, respectively (Supplementary Fig. 10). The confocal images in Fig. 2g showed strong fluorescence of TAMRA-rDOSs at the choroid and retina in WES-treated eyes, whereas intravitreal injection resulted in weak fluorescence (P = 0.0039) (Fig. 2h and Supplementary Fig. 11). Similarly, WES-treated eyes exhibited strong fluorescence from DiD-EVs, unlike the intravitreal injection group (P = 0.0056) (Fig. 2i, j and Supplementary Fig. 12). These results underscored the versatility of WES in delivering macromolecular drugs and nanomedicines to the fundus compared with intravitreal injection. The WES parameters of 0.024 Hz and 80 µA were used for subsequent experimental studies.

Quantifying the IgG concentration in WES-treated mouse eyes via enzyme-linked immunosorbent assay (ELISA) revealed a delivery efficiency of 14%, which was comparable to that of intravitreal injection at 1 hour (16% of delivery efficiency), with nearly 3% detected after 24 hours (Fig. 2k). Remarkably, the developed WES achieved a macromolecular drug delivery efficiency of up to 14% in fundus, which was about three times higher than other non-invasive methods^4,5. This result strongly demonstrated the feasibility of WES as a non-invasive macromolecule delivery platform with significant potential for treating fundus diseases. To further assess the primary distribution of IgG within the eye, we used rabbit eyes, which closely resemble the human eye structure (Fig. 2l). A higher concentration of IgG was delivered to the cornea by WES compared with intravitreal injection (P = 0.0013), probably because WES facilitated the permeation of IgG in the cornea. In contrast, only a small amount of IgG from intravitreal injection reached the cornea via the aqueous humor³⁹. However, there was no significant difference in scleral or intravitreal IgG levels between the WES and the intravitreal injection group (Fig. 2l). This result may be because the barrier properties of the ILM limit IgG permeability, resulting in lower levels of IgG entering the vitreous in the WES group and higher levels in the intravitreal injection group owing to passive diffusion³⁹. To sum up, these findings provided robust evidence that WES could significantly increase IgG delivery from the sclera to the choroid, retina and vitreous, enabling efficient non-invasive delivery of monoclonal antibodies.

The mechanism of WES to enhance macromolecular drug delivery to fundus

The TJ and AJ proteins between ARPE-19 cells formed an external blood‒retinal barrier with Bruch's membrane and choroid on the outer side of the retina, limiting the movement of macromolecules⁴⁰. Therefore, opening this barrier is essential for delivering macromolecular drugs to treat fundus diseases^41,42. The ability of WES to reversibly open the cellular barrier was explored at the cellular and tissue levels. We first investigated the effects of WES on TJ and AJ proteins in retinal epithelial cells via WES. Experimentally, ARPE-19 cells were incubated to form a compact monolayer with closely arranged AJ and TJ proteins, simulating the outer blood‒retinal barrier. We treated ARPE-19 cells with WES for 10 min and observed the expression and distribution of TJ-associated proteins, including zonula occludens-1 (ZO-1)⁴³, E-cadherin³⁹, occludin⁴⁴, and filamentous actin (F-actin)⁴⁵, via confocal laser scanning microscopy (CLSM) (Fig. 3a). CLSM images revealed that WES triggered a redistribution of ZO-1 and occludin, especially in the regions marked by red arrows, where discrete fluorescence signals indicated open cell gaps. In contrast, untreated cells presented a continuous fluorescence grid. WES-induced redistribution of connexin ZO-1 did not significantly decrease ZO-1 expression levels. WES also caused transient E-cadherin leakage, but sufficient E-cadherin remained at the cell membrane to reform AJ and TJ. The cellular barrier was fully restored after 24 h (Third column in Fig. 3a)

Given the importance of the cytoskeleton in maintaining cell morphology and regulating TJs and AJs⁴⁶, we explored the impact of WES on F-actin expression in the cytoskeleton. Immunofluorescence (IF) images revealed dense, continuous actin in untreated cells, whereas WES-treated cells displayed F-actin breakage and leakage (Fourth column in Fig. 3a).. Microscopic images in bright field of view further confirmed that the cell junctions of ARPE-19 cells were opened and substantially restored after 24 hours (last column in Fig. 3a). These results suggested that WES reversibly regulated ARPE-19 cell paracellular permeability by modulating the redistribution of TJ and AJ related protein.

We also studied the transepithelial electrical resistance (TEER) of monolayer ARPE-19 cells under different WES durations (2, 5, and 10 min) (Fig. 3b). WES for 2 and 5 minutes immediately decreased the TEER to approximately 79% and 73%, respectively, maintaining these levels for at least one hour before returning to 94% of the initial value within five hours. This finding indicated that short-term WES (< 5 min) could temporarily open ARPE-19 cell junctions and substantially restore them within a few hours. Ten minutes of WES prolonged junction opening, decreasing the TEER to 57% and restoring 80% of normal levels after 29 hours (Fig. 3b and Supplementary Fig. 13). This result demonstrated WES's unique advantage of controllable junction reshaping and restoration by adjusting the WES timing.

In vivo immunofluorescence staining of mouse eye sections revealed significantly weaker fluorescence of ZO-1 in the RPE monolayer after 10 minutes of WES than in untreated cells (P = 0.0009) (Supplementary Fig. 14). However, the fluorescence intensity increased after 24 h and was not significantly different from that of untreated cells (Supplementary Fig. 14). These findings reinforced the conclusion that WES promoted reversible cellular junction opening of RPE monolayer.

To quantify the reversible modulation of cellular junctions by WES, we measured the paracellular permeability of a dense ARPE-19 cell monolayer to FITC-IgG. Research has indicated that Rho/ROCK signalling and myosin light chain kinase (MLCK) activation synergistically phosphorylate the myosin light chain, promoting actomyosin ring contraction and junction disruption⁴⁷. We used ML-7, an MLCK inhibitor, to restore junction integrity. Compared with the untreated group, the WES group exhibited greater FITC-IgG permeability after 24 h (P = 0.0008), indicating enhanced paracellular junction permeability (Fig. 3c). Notably, compared with the WES alone group, the WES + ML-7 group presented partially inhibited IgG penetration (P = 0.0003) (Fig. 3d), suggesting that WES-induced junction barrier opening is reversible.

As reported in pervious literatures, the extracellular repeat structural domain of E-cadherin bound Ca²⁺ to form a rigid linear molecule, which enhanced adhesion between neighbouring cells and was involved in the formation of intercellular AJs^31,47. Therefore, extracellular Ca²⁺ depletion could contract F-actin and redistribute TJ and AJ proteins^31,47. Considering the presence of multiple calcium channels associated with electrical signalling⁴⁸, we explored the effects of WES on the free intracellular Ca²⁺ dynamics in ARPE-19 cells. Interestingly, WES significantly remodelled the cellular junctions near the positive electrode, with reduced junction opening as the electrical potential decreased (Supplementary Fig. 15). The fluorescence intensity of intracellular Ca²⁺ was significantly stronger at the P₁ position near the positive electrode than at the P₃ position away from the electrode (P = 0.0004), and the fluorescence intensity tended to decrease gradually as the distance from the electrode increased (Supplementary Fig. 16). This may be because extracellular Ca²⁺ in the near positive electrode were more susceptible to electrostatic repulsion and the cell membrane was more easily depolarised leading to a rapid increase in intracellular Ca²⁺ concentration⁴⁸. To sum up, these results confirmed the ability of WES to remodel local, controllable, and safe cellular connections. In addition, the increase in intracellular Ca²⁺ was strongly correlated with the duration of electrical stimulation of WES. WES for 2 min failed to induce free intracellular Ca²⁺ enrichment (Supplementary Fig. 17a, b). However, the intracellular Ca²⁺ concentration sharply increased after 5 min of WES (P < 0.0001), with no significant difference from that in untreated cells at 5 h (Supplementary Fig. 17c). Ten minutes of WES significantly increased the intracellular Ca²⁺ (P < 0.0001), which returned to control levels after 24 hours (Fig. 3e, f). It can be found that the trend of changes in intracellular Ca²⁺ concentration at different times of WES and the changes in cellular junctions all remain almost the same. These results strongly suggested that WES rapidly and significantly regulated extracellular Ca²⁺ influx, thus enabling reversible opening of the cellular barrier.

To explore whether WES induced an electrophoretic effect on drug delivery, we used an in vitro simulation with an agarose-hyaluronic acid mixture to mimic vitreous fluid⁴⁹. In the WES group, FITC-IgG was injected into the centre of the channel and the ivory-shaped Au electrodes of SWG were attached on both sides within the channel. Compared with passive diffusion, WES significantly increased the fluorescence area in the channel, indicating that WES promoted the electrophoresis of FITC-IgG (P = 0.0248) (Fig. 3g, h). Finite element model analysis of the drug concentration distribution in the eyeball further confirmed the ability of WES to accelerate drug movement through electrophoresis, resulting in a higher delivery rate than free diffusion (Fig. 3i). Detailed parameters of the theoretical simulations of WES-facilitated drug electrophoresis were shown in the Supplementary Table 3. Compared with passive intravitreal injection, WES-triggered iontophoresis also enabled the active delivery of biomolecules to the fundus via electrophoresis. Taken together, by triggering the extracellular Ca²⁺ influx, WES achieved reversible tight junction and adherens junction opening between retinal epithelial cells as well as cytoskeletal remodelling, which allowed macromolecular drugs to penetrate the outer blood‒retinal barrier.

Therapeutic effects of WES in a laser-induced CNV mouse model

Given the efficacy of WES in the non-invasive delivery of monoclonal antibody analogues, we evaluated the therapeutic effects of WES-based delivery of anti-VEGF (aV) on CNV in mice. We established a laser-induced CNV mouse model, a well-validated model for studying exudative AMD⁵⁰ (Fig. 4a). Five days after laser treatment, fundus fluorescein angiography (FFA) revealed strong sodium fluorescein fluorescence at CNV lesion sites, and optical coherence tomography (OCT) revealed significant lesion thickening, confirming the successful establishment of the CNV model (Fig. 4b and Supplementary Fig. 18).

We divided C57BL/6 mice into five groups (n = 6 per group): Untreated, WES, aV, WES + aV, and aV injection. The mice in the aV injection group received a single intravitreal injection of 2 µl aV (2.4 mg/kg) on day 1 of treatment. The mice in the WES group received 10 minutes of WES once every 4 days. The mice in the aV group were treated with aV (0.6 mg/kg) via an EDDL once every 4 days. The WES + aV group received non-invasive delivery of aV (0.6 mg/kg) via 10 minutes of WES once every 4 days. After a 14-day treatment cycle, quantification of CNV lesion areas revealed that the WES + aV group most effectively inhibited new vessel growth (P < 0.0001), outperforming the intravitreal injection group (P = 0.0089) (Fig. 4b, f and Supplementary Fig. 19). OCT analysis revealed that CNV lesion thickness was significantly lower in the WES + aV group than in the untreated group (P < 0.0001), with effects comparable to those of aV injection (P < 0.0001) (Fig. 4c, g). Visual function was evaluated via electroretinography (ERG) under scotopic conditions. Statistical analysis revealed that the A and B waves in WES-treated mice were significantly restored to normal ranges, whereas those in the control groups failed to recover visual function because of inferior CNV treatment effects (P < 0.0001) (Supplementary Fig. 20). Hematoxylin and eosin (H&E) staining confirmed significant CNV thickness in the untreated group, whereas the WES + aV group showed significant inhibition, which was comparable to the aV injection group (Fig. 4d and Supplementary Fig. 21). IF staining of CNV-bearing mouse retinas revealed that WES + aV treatment resulted in the most pronounced reductions in VEGF and CD31 expression (Fig. 4e and Supplementary Fig. 22). Statistical analysis revealed that, compared with the other groups, the WES + aV group presented the lowest fluorescence intensities of VEGF (P = 0.0253) and CD31 (P = 0.0054) (Fig. 4h, i). These results indicated that WES could non-invasively and effectively delivered aV to the choroid and retina, demonstrating superior CNV inhibition compared with injection.

Therapeutic effects of WES in choroidal melanoma

Inspired by the promising results of WES-enhanced aV delivery in CNV therapy, we evaluated the therapeutic efficacy of WES-based anti-PDL1 (aPDL1) delivery in choroidal melanoma, validating its versatility in non-invasive drug delivery for fundus disorders. A choroidal melanoma model was constructed by microinjecting luciferase B16F10 (Luc-B16F10) cells into the subretinal chamber of C57BL/6 mice (Fig. 5a). The mice were divided into five groups (n = 5 per group): (G1) Untreated: Untreated groups, (G2) aPDL1: daily 10 min EDDL containing aPDL1 (0.6 mg/kg), (G3) WES: daily 10 min WES, (G4) WES + aV: daily 10 min WES containing aPDL1 (0.6 mg/kg), and (G5) aPDL1 injection: aPDL1 intravenous injection (4.2 mg/kg) on day 0.

Intraperitoneal injections of D-luciferin potassium salt (10 mg/kg) followed by in vivo bioluminescence imaging at three-day intervals were used to assess treatment efficacy. The in vivo bioluminescence images of representative mice revealed that G4 had the lowest bioluminescence signal after 7 days, indicating effective tumour inhibition (P = 0.0237), whereas G5 was less effective (Fig. 5b, c and Supplementary Fig. 23). Slit lamp photography on the sixth day revealed significant tumour growth in all groups except G4, which showed no tumour accentuation (Fig. 5d). Histological analysis with H&E staining revealed fewer melanomas in eye slices from the G5 group than in those from the untreated group, but melanoma cells were still present, whereas the G4 group had no melanoma cells (Fig. 5d). The survival rate was significantly prolonged to 15 days in G4 compared with the other groups (P = 0.0113) (Fig. 5e). These results confirmed that WES could cross the blood‒retinal barrier, enabling effective choroidal melanoma treatment, whereas intravenous administration was hindered by ocular barrier complexities.

Subsequently, the immune response to choroidal melanoma after different treatments was investigated by flow cytometry analysis (Fig. 5f and Supplementary Fig. 24). Flow cytometry analysis revealed increased percentages of CD4⁺ and CD8⁺ T cells among CD3⁺ cells in WES-treated tumours (G4), indicating successful delivery of aPD-L1 to the tumour site and effective tumour inhibition (P < 0.0001) (Fig. 5g, h and Supplementary Fig. 25)⁵¹. The ELISA results revealed significant upregulation of IFN-γ in G4 compared with the other groups, and G4 had the highest levels of TNF-α (P < 0.0001) (Fig. 5i and Supplementary Fig. 26). These results strongly confirmed that WES effectively inhibited choroidal melanoma growth by efficiently delivering aPD-L1 to tumours and activating the immune response.

In vitro and in vivo safety assessment of WES

To ascertain the safety of WES in treating fundus diseases, we conducted in vitro and in vivo evaluations over 4 weeks. Healthy C57BL/6 mice were divided into four groups: Untreated, EDDL, SWG, and WES. The EDDL group wore the EDDL for 10 minutes daily. The SWG group wore the SWG for 10 minutes daily. The WES group wore the WES for 10 minutes daily. The sodium fluorescein drops on the corneas of the WES-treated mice remained clear and unbroken, as observed via slit lamp microscopy, indicating that no corneal damage occurred after 4 weeks (Fig. 6a). Intraocular pressure (IOP) measurements at 2 and 4 weeks revealed no increase in IOP due to WES treatment (Fig. 6b). We further assessed visual function via ERG. The representative ERG results indicated that the visual ability of the WES-treated mice remained normal (Fig. 6c). Statistical analyses of A- and B-waves confirmed that WES did not adversely affect visual function (Fig. 6d, e). The levels of the inflammatory markers IL-1β and TNF-α in the eyes of WES-treated mice did not indicate significant inflammation, supporting a favourable safety profile (Fig. 6f, g). The OCT images revealed no difference in retinal thickness between WES-treated and normal mice (Fig. 6h). H&E staining revealed that the eye structures of WES-treated mice remained intact and comparable to those of normal mice (Fig. 6i). Differences in organ structure between WES-treated and healthy mice were minimal (Supplementary Fig. 27). The blood biochemistry and blood analysis results revealed that all the parameters were within normal ranges, indicating long-term tissue safety (Supplementary Fig. 28).

We presented a lightweight WES device for effective, non-invasive macromolecular drug delivery to the ocular fundus. Powered by a lithium battery, WES converted DC to monophasic square pulse current signals for scleral. The brief WES periods induced rapid increase in intracellular Ca²⁺ concentration, resulting in the remodelling of Ca²⁺-dependent cellular junctions in ARPE-19 cells. WES triggered PAMR contraction and redistributed TJ and AJ proteins, enhancing drug delivery via the sclera-choroid-retina pathway. When controlled by stimulation time and voltage, WES effects were localized, ensuring safety. WES accelerated drug distribution in the vitreous, surpassing the slow diffusion of intravitreal injection. Our results demonstrated the superior efficiency of WES in delivering various macromolecular drugs, including monoclonal antibodies, rDOSs, and EVs. WES-enhanced anti-VEGF administration effectively inhibited CNV, outperforming intravitreal injection. Additionally, WES-enhanced anti-PDL1 therapy showed efficacy in non-invasive early mouse choroidal melanoma treatment. A 4-week safety study confirmed the non-invasive nature of WES and the lack of adverse effects on ocular and systemic health. This study confirmed the ability of WES to reversibly open the blood‒retinal barrier via extracellular Ca²⁺ influx and junction redistribution in ARPE-19 cells. The WES platform offered a promising approach for non-invasive and efficient treatment of chronic fundus diseases, presenting a new paradigm in ocular drug delivery.

Reagents and materials.

N-Hydroxysuccinimide (NHS)-ester labelled fluorescein isothiocyanate (FITC) (Thermo Fisher Scientific, catalogue no. 46410), Actin-Tracker Red-Rhodamine (Beyotime, catalogue no. C2207S), Fluo-4 AM (Beyotime, catalogue no. S1061S), anti-PDL1 (Bio X Cell Biotech Co., Ltd., catalogue no. BE0101), and anti-VEGF (MCE, catalogue no. HY-P9906) antibodies were used. The rat lgG ELISA kit (Elabscience, catalogue no. E-EL-R0518), anti-occludin (Abcam, catalogue no. ab216327), anti-E-cadherin (Proteintech, catalogue no. 20874-1-AP), anti-ZO-1 (Proteintech, catalogue no. 21773-1-AP), goat anti-rabbit lgG H&L (FITC) (Abways, catalogue no. AB0121), D-luciferin potassium salt (Abcam, catalogue no. ab143655), ML-7 inhibitor (MCE, catalogue no. HY-15417), anti-CD3–FITC (BioLegend, catalogue no. 100204), anti-CD4–APC (BioLegend, catalogue no. 100412), anti-CD8–PE (BioLegend, catalogue no. 100708), IFN-γ ELISA Kit (Thermo Fisher Scientific, catalogue no. 88-7314-88), and TNF-α ELISA Kit (Thermo Fisher Scientific, catalogue no. 88-7324-88) were used.

Fabrication of the WES

The SWG was obtained via a commercial flexible printed circuit fabrication process. The bill of materials includes passive components (li-ion battery, tactile switch, capacitors, resistors and diodes), two li-ion batteries (ML-621S/DN, Panasonic) served as power supply for the entire circuit, a tactile switch (TS2030E16055, HanElectricity) controlling the start-up and shut-down of the WES, a timer chip (NE555, IDCHIP) with precision timing circuits capable of producing accurate time delays or oscillation, a multilayer ceramic capacitor (CGA0805X5R476M100MT, HRE), two switch diodes (1N4148W, YAGEO), a multilayer ceramic capacitor (0805B104K160CT-C, Walsin) and two multilayer ceramic resistors (SC0805F1304F4ANRH, Sunway) for adjusting the frequency and low-to-high ratio of the output current signals. The substrate of the flexible circuit board is polyimide (22 mm×18 mm×0.2 mm). The ivory-shaped flexible electrodes of the SWG were then folded and embedded in a polydimethylsiloxane (PDMS) contact lens via the cast molding technique at room temperature to fabricate the EDDL. The connection area between the SWG and EDDL was further encapsulated by an elastomeric polyurethane coating.

Preparation of the FITC-lgG

The antibody was first prepared to a concentration of 10 mg/mL for the labelling reaction. Then the appropriate amount of NHS labelled FITC was mixed with the lgG antibody solution and ice-bathed for 2 hours. Finally, the FITC labelled lgG (FITC-lgG) was obtained after removing unlabelled NHS-FITC by ultrafiltration.

Preparation of the TAMRA-rDOSs

The rDOSs were obtained by reacting M13mp18 ssDNA scaffold (New England Biolabs, catalogue no. N4040S) with the short strand in 1 × TAE-Mg²⁺ buffer and annealing according to an established procedure, which specifically consisted of heating the mixture to 65°C and then cooling it to 25°C at a rate of 10 min/°C. The samples were then washed twice with 400 µL of TAE-Mg²⁺ buffer and centrifuged each time for 5 minutes at 2,000 × g. Post-centrifugation samples was analyzed by using 1% agarose gel electrophoresis and an Atomic Force Microscope (AFM) (Dimension Icon, Bruker). Subsequently, an excess of TAMRA-labelled chain was added to 100 µl along with purified rDOSs. And the reaction was carried out according to the established procedure: the mixture was heated to 45°C and then cooled to room temperature at a rate of 10 min/°C. Finally, the TAMRA-rDOSs were obtained by ultrafiltration.

Preparation of the DiD-EVs

Macrophages and human retinoblastoma cell (Y79) membranes were isolated by a reported protocol. Specifically, cells were collected and resuspended in phosphate buffer (pH = 7.4) supplemented with a protease and phosphatase inhibitor cocktail consisting of 0.25 M sucrose, 1 mM EDTA, and 20 mM HEPES-NaOH. The suspensions were sonicated at 100 W with alternating cycles of 5-second cooling and 5-second shear for 10 min until a clear lysate was A clear lysate was obtained. The resulting supernatant was centrifuged at 4000 rpm and 14800 rpm for 15 min at 4°C to remove intracellular impurities. The precipitate from the ultracentrifuge tube was then resuspended in deionised water containing 0.2 mM EDTA. Liposomes were prepared by thin film dispersion technique. DOPE, CHEMS, DSPE-PEG and DSPE-PEG-COOH in a molar ratio of 45:11.8:6:0.6 were dissolved in a chloroform/methanol mixture (9:1, v/v) and evaporated to form a thin lipid film. Ultrasound and hydration were used to generate liposomes in phosphate buffer. Then liposomes and cell membranes were combined using an extruder (Avanti Polar Lipids, Alabaster, AL) and extruded into polycarbonate membranes. The binding ratio of liposomes to cell membranes was 1:25:25 (w/w). The morphological of EVs was characterized by the transmission electron microscope (TEM) (Tecnai G2 F20 S-TWIN, FEI). DiD-EVs were harvested by dispersing the lipid film in the DiD pre-treated phosphate buffer.

In vitro cytotoxicity analysis

An MTT assay was performed to analyse the in vitro cytotoxicity of ARPE-19 cells and HUVECs. The cells were seeded in a 96-well plate at 5000 cells per well and incubated at 37°C for 12 hours. After different treatments and 24 hours of incubation, the cells were washed and cultured with fresh media containing 5 mg/ml MTT solution. After 3 hours of incubation with MTT solution, the medium was replaced with 100 µl of DMSO, and the mixture was incubated for 10 minutes. The absorbance at 570 nm was measured to determine the relative cell viability.

Quantitative analysis of the permeability of lgG in the eye

IgG penetration in mouse eyeballs was measured after 10 min WES or intravitreal injection at 1 and 24 hours. After the experimental mice were euthanised, the eyeballs were collected and cleaned with buffer solution. Supernatants from homogenized eyeball tissue were collected after centrifugation and detected via ELISA kits. IgG distribution in New Zealand white rabbit eyeballs was measured after 10 min WES or intravitreal injection at 1 hours. The rabbit eyeballs were washed and dissected into the cornea, choroid, and vitreous after treatment, and the IgG concentration was measured via ELISA kits.

Immunofluorescence analysis of tight junction and adherens junction proteins in ARPE-19 cells and eyeballs

ARPE-19 cells were seeded on a 12-well plate at 5×10⁵ cells per well and incubated at 37°C for 24 hours. Cells and mouse eyeballs, untreated or treated with WES for 10 minutes, were fixed in 4% paraformaldehyde at 37°C for 15 minutes and washed with PBS. The samples were permeabilized with 0.5% Triton X-100 in PBS at 37°C for 20 minutes, washed with PBS, blocked with 5% bovine serum albumin in PBS at 37°C for 30 minutes, and incubated with primary antibodies (anti-ZO-1, anti-E-cadherin, and anti-occludin) at 4°C overnight. The samples were then incubated with FITC-conjugated secondary antibodies at 37°C for 2 hours, stained with actin-tracker red-rhodamine and DAPI, and imaged via an LSM 800 confocal laser scanning microscope.

TEER measurement of the ARPE-19 cells

ARPE-19 cells were seeded at 5×10⁴ cells per well in the top chamber of a transwell to form a dense monolayer barrier. The cells were treated with WES for 2, 5, or 10 minutes, and the TEER values were measured over 30 hours.

TEER (ohm·cm²) = (R₁ − R₀) × 0.33

where R₁ is the TEER value with cells and R₀ is without cells. The top chamber membrane area was 0.33 cm².

Quantitative analysis of the permeability of FITC-lgG in monolayers of ARPE-19 cells

ARPE-19 cells were seeded at 5×10⁴ cells per well in the top chamber of a transwell. FITC-IgG was added to the top chamber, and the samples were treated via different methods. In the ML-7 group, the cells were pretreated with the ML-7 inhibitor overnight. In the WES group, the cells were treated with WES for 10 minutes. In the WES + ML-7 group, the cells were treated with WES for 10 minutes after pretreatment with the ML-7 inhibitor overnight. The fluorescence intensity of FITC-IgG was measured via a fluorescence spectrometer.

Intracellular Ca²⁺ imaging during WES

ARPE-19 cells were seeded on a 6-well plate at 2×10⁶ cells per well and incubated at 37°C for 12 hours. A total of 1 ml of Fluo-4 staining solution (assay buffer with 0.2% Fluo-4 AM and 0.2% solubility enhancer) was added to each well, and the mixture was incubated at 37°C for 30 minutes. The cells were treated with WES for different durations, and Fluo-4 AM fluorescence was imaged via an LSM 800 confocal laser scanning microscope.

In vitro electrophoretic ability analysis of WES

A solution containing 0.95 mg/ml agarose and 0.7 mg/ml hyaluronic acid was used to simulate vitreous fluid. DiD-EVs were used as large model drugs with WES for 10 minutes. Fluorescence images and statistical analysis of the mean fluorescence intensity of DiD-EVs were obtained via an animal living fluorescence imaging system.

Theoretical simulation of the potential distribution in cell culture dishes under WES

Theoretical simulations of the potential distribution in cell culture dishes under WES were conducted via COMSOL Multiphysics software with current modules. A constant current formed a stable potential distribution in the cell culture dish. Current module equations:

$$\:\nabla\:\bullet\:J={Q}_{\text{j},\text{v}}$$

$$\:J=\sigma\:E+\frac{\delta\:D}{\delta\:t}+{J}_{\text{e}}$$

$$\:E=-\nabla\:\bullet\:V$$

$$\:V={V}_{0}$$

where V denotes the electrical potential, E refers to the intensity of the electric field, J represents the current density, σ represents the material conductivity, ∇ represents the Hamiltonian.

Theoretical simulations of WES-facilitated drug electrophoresis

Theoretical simulations of the WES-facilitated drug electrophoresis were conducted via COMSOL Multiphysics software with current modules and substance transmission modules. The ocular drug delivery process by the WES was visualized by constructing a 2D eye model emulating the experimental device layout. The vitreous fluid was modelled as the eye's fluid environment, with WES in contact with the scleral surface. The dynamic substance penetration process was then simulated by the substance transmission module on the basis of the current module. The theoretical equation can be expressed as follows:

$$\:\frac{\partial\:{c}_{\text{i}}}{\partial\:t}+\nabla\:\bullet\:{J}_{\text{i}}={R}_{\text{i}}$$

$$\:{J}_{\text{i}}=-{D}_{\text{i}}\nabla\:{c}_{\text{i}}-{z}_{\text{i}}{u}_{\text{m},\text{i}}F{c}_{\text{i}}\nabla\:V$$

where J_i refers to the diffusion flux vector, c denotes the concentration, z represents the charge number, F refers to the Faraday constant, V is the electrical potential, D_i corresponds to the effective diffusion coefficient, and u_m,i denotes the effective mobility. The relationship between D_i and u_m,i can be expressed via the Nernst–Einstein equation:

$$\:{u}_{\text{m},\text{i}}=\frac{{D}_{\text{i}}}{RT}$$

where R represents the Moore gas constant and T represents the temperature.

The therapeutic effects of WES in CNV mice

A mouse model of laser-induced CNV was established via laser photocoagulation (370 mW, 60 ms) according to a standard protocol⁴⁸. CNV lesions were observed via an OCT system (Heidelberg, Germany) after the intraperitoneal injection of 2% fluorescein sodium (Alcon) (5 mL/kg). The mice were divided into five groups: Untreated, WES, aV, WES + aV, and aV injection. The CNV lesion area and thickness and the mean fluorescence intensity of VEGF and CD31 were calculated via ImageJ software. All the animal experiments were conducted in accordance with the animal experiment protocols approved by the Soochow University Laboratory Animal Center (approval number: syxk(su) 2021–0073). The animal experiments were approved by the animal welfare regulatory authorities.

The therapeutic effects of WES in choroidal melanoma mice

The choroidal melanoma model was created by injecting 2×10⁵ Luc-B16F10 cells into the subretinal chamber of C57BL/6 mice. After 2 days, the mice were divided into five groups: (G1) Untreated, (G2) 10 min EDDL containing aPDL1 (0.6 mg/kg) daily, (G3) 10 min WES daily, (G4) 10 min WES containing aPDL1 (0.6 mg/kg) daily, and (G5) aPDL1 intravenous injections (4.2 mg/kg) on day 0. The mice were injected intraperitoneally with D-luciferin potassium salt (10 mg/kg) and imaged via an in vivo bioluminescence imaging system at three-day intervals. Immune cells in choroidal melanoma tumours were studied by obtaining single-cell suspensions from mouse eyes at 6 days posttreatment and then stained with anti-CD3–FITC, anti-CD4–APC, and anti-CD8–PE antibodies. IFN-γ and TNF-α concentrations were measured via ELISA kits. Eyeballs were sectioned (8 µm thickness), stained with H&E, and imaged via light microscopy (DM 4000).

Statistical analysis

A paired two-tailed t test or one-way ANOVA was used for statistical analysis via GraphPad software. All the data are presented as the means ± SDs. The quantitative assessment of fluorescence intensity was performed via ImageJ, a commercial image analysis program.

Life Science Reporting Summary.

Further details regarding the experimental design can be found in the Life Science Reporting Summary.

Data availability

The data supporting the findings of this study are contained within the paper and its Supplementary information. Source data are included with this paper.

Acknowledgements

We thank Prof. Qian Chen (Sochoow University, China) for her general help and valuable suggestions.Y. H. discloses support for the research described in this study from the National Key R&D Program of China (2023YFB3208200), the National Natural Science Foundation of China [grant numbers 22393932, T2321005, 21825402], the Science and Technology Development Fund, Macau SAR [grant number 0002/2022/AKP, 0115/2023/RIA2], the Major Independent Research Project of Jiangsu Key Laboratory for Carbon-Based Functional Materials & Devices (grant number L421490022) and the Program for Jiangsu Specially Ap-pointed Professors to Professor Yao He, a project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions (PAPD), 111 Project and Collaborative Innovation Center of Suzhou Nano Science and Technology (NANO-CIC). H.Y.W. provided support for the research described in this study from the National Natural Science Foundation of China [grant number 22074101]. B. B. C. provided support for the research described in this study from the National Natural Science Foundation of China [grant number 22204117].

Author Contributions

X.Q., B.B.C., Z.W., X.H.S., H.Y.W. and Y.H. conceived and designed the research. X.Q., H.L.S. and H.Y.L. carried out most of the experiments and analysed the data. J. W. Z. performed additional experiments and characterizations. X.Q., H.L.S., Z.W., X.H.S., H.Y.W. and Y.H. wrote the manuscript.

Competing interests

The authors declare no competing financial interests.

Additional information

Supplementary Information is available in the online version of the paper.

Reprints and permissions information is available online at www.nature.com/reprints.

Correspondence and requests for materials should be addressed to X. H. Sun, Z. Wen, H. Y. W. or Y. H.

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

M. Fleckenstein, et al. Age-related macular degeneration. Nat. Rev. Dis. Primers. 7, 31 (2021).
J. Cehajic-Kapetanovic, et al. Bioengineering strategies for restoring vision. Nat. Biomed. Eng. 7, 387-404 (2023).
K. D. Falkenberg, et al. The metabolic engine of endothelial cells. Nat. Metab. 1, 937-946 (2019).
N. Ferrara. Vascular endothelial growth factor and age-related macular degeneration: from basic science to therapy. Nat. Med. 16, 1107-1111 (2010).
Y. Tian, et al. Reduction of choroidal neovascularization via cleavable VEGF antibodies conjugated to exosomes derived from regulatory T cells. Nat. Biomed. Eng. 5, 968-982 (2021).
S. Ling, et al. Lentiviral delivery of co-packaged Cas9 mRNA and a Vegfa-targeting guide RNA prevents wet age-related macular degeneration in mice. Nat. Biomed. Eng. 5, 144-156 (2021).
R. Yang, et al. Getting drugs across biological barriers. Adv. Mater. 29, 1606596 (2017).
E. Ramsay, et al. Selective drug delivery to the retinal cells: Biological barriers and avenues. J. Control. Release 361, 1-19 (2023).
C. C. McLaughlin, et al. Incidence of noncutaneous melanomas in the U.S. Cancer 103, 1000-1007 (2005).
B. Damato, et al. Genotypic profiling of 452 choroidal melanomas with multiplex ligation-dependent probe amplification. Clin. Cancer Res. 16, 6083-6092 (2010).
K. M. Komatsubara; R. D. Carvajal immunotherapy for the treatment of uveal melanoma: current status and emerging therapies. Curr. Oncol. Rep. 19, 45-56 (2017).
J. Yang, et al. Treatment of uveal melanoma: where are we now? Ther. Adv. Med. Oncol. 10, 1758834018757175 (2018).
G. P. Barry, et al. Retinal detachment after treatment of retinopathy of prematurity with laser versus intravitreal anti–vascular endothelial growth factor. Ophthalmology 128, 1188-1196 (2021).
M. G. Maguire, et al. Five-year outcomes with anti–vascular endothelial growth factor treatment of neovascular age-related macular degeneration: the comparison of age-related macular degeneration treatments trials. Ophthalmology 123, 1751-1761 (2016).
J. Shen, et al. Eyedrop-based macromolecular ophthalmic drug delivery for ocular fundus disease treatment. Sci. Adv. 9, eabq3104 (2023).
X. Zhao, et al. Antiangiogenic nanomicelles for the topical delivery of aflibercept to treat retinal neovascular disease. Adv. Mater. 34, 2108360 (2022).
M. Bohley, et al. A single intravenous injection of cyclosporin A–loaded lipid nanocapsules prevents retinopathy of prematurity. Sci. Adv. 8, eabo6638 (2022).
D. D. Nguyen, et al. Highly retina-permeating and long-acting resveratrol/metformin nanotherapeutics for enhanced treatment of macular degeneration. ACS Nano 17, 168-183 (2023).
M. Yang, et al. Engineering hibiscus-like riboflavin/ZIF-8 microsphere composites to enhance transepithelial corneal cross-linking. Adv. Mater. 34, 2109865 (2022).
W. Poon, et al. A framework for designing delivery systems. Nat. Nanotechnol. 15, 819-829 (2020).
Y. C. Kim, et al. Gelling hypotonic polymer solution for extended topical drug delivery to the eye. Nat. Biomed. Eng. 4, 1053-1062 (2020).
M. Herrera-Barrera, et al. Peptide-guided lipid nanoparticles deliver mRNA to the neural retina of rodents and nonhuman primates. Sci. Adv. 9, eadd4623 (2023).
A. Badia, et al. Repeated topical administration of 3 nm cerium oxide nanoparticles reverts disease atrophic phenotype and arrests neovascular degeneration in AMD mouse models. ACS Nano 17, 910-926 (2023).
Z. Liu, et al. Electroactive biomaterials and systems for cell fate determination and tissue regeneration: design and applications. Adv. Mater. 33, 2007429 (2021).
K. Krawczyk, et al. Electrogenetic cellular insulin release for real-time glycemic control in type 1 diabetic mice. Science 368, 993-1001 (2020).
C. Chen, et al. Micro triboelectric ultrasonic device for acoustic energy transfer and signal communication. Nat. Commun. 11, 4143 (2020).
G. Shim, et al. Bioelectric stimulation controls tissue shape and size. Nat. Commun. 15, 2938 (2024).
Y. Pan, et al. Electrical stimulation induces antitumour immunomodulation via a flexible microneedle-array-integrated interdigital electrode. Sci. Bull. 68, 2779-2792 (2023).
H. S. Venkatesh, et al. Electrical and synaptic integration of glioma into neural circuits. Nature 573, 539-545 (2019).
R. Polanía, et al. Studying and modifying brain function with noninvasive brain stimulation. Nat. Neurosci. 21, 174-187 (2018).
T. Kim, et al. Deep brain stimulation by blood–brain-barrier-crossing piezoelectric nanoparticles generating current and nitric oxide under focused ultrasound. Nat. Biomed. Eng. 7, 149-163 (2023).
S. M. Won, et al. Wireless and battery-free technologies for neuroengineering. Nat. Biomed. Eng. 7, 405-423 (2023).
H. Wu, et al. Accelerated intestinal wound healing via dual electrostimulation from a soft and biodegradable electronic bandage. Nat. Electron. 7, 299-312 (2024).
E. Karbassi, et al. Cardiomyocyte maturation: advances in knowledge and implications for regenerative medicine. Nat. Rev. Cardiol. 17, 341-359 (2020).
C. Moritz, et al. Noninvasive spinal cord electrical stimulation for arm and hand function in chronic tetraplegia: a safety and efficacy trial. Nat. Med. 30, 1276-1283 (2024).
F. Zhao, et al. A hydrogel ionic circuit based high-intensity iontophoresis device for intraocular macromolecule and nanoparticle delivery. Adv. Mater. 34, 2107315 (2022).
L. Jouve, et al. Conventional and iontophoresis corneal cross-linking for keratoconus: efficacy and assessment by optical coherence tomography and confocal microscopy. Cornea 36, 153-162 (2017).
E. C. O'Neil, et al. Iontophoretic delivery of dexamethasone phosphate for noninfectious, nonnecrotizing anterior scleritis, dose-finding clinical trial. Br. J. Ophthalmol. 102, 1011-1013 (2018).
F. Peng, et al. Nanoparticles promote in vivo breast cancer cell intravasation and extravasation by inducing endothelial leakiness. Nat. Nanotechnol. 14, 279-286 (2019).
B. Zhu, et al. Large enhancement in neurite outgrowth on a cell membrane-mimicking conducting polymer. Nat. Commun. 5, 4523 (2014).
M. Lamminsalo, et al. Extended pharmacokinetic model of the intravitreal injections of macromolecules in rabbits. part 2: parameter estimation based on concentration dynamics in the vitreous, retina, and aqueous humor. Pharm. Res. 37, 226 (2020).
P. Romani, et al. Crosstalk between mechanotransduction and metabolism. Nat. Rev. Mol. Cell. Biol. 22, 22-38 (2021).
S. Bilal, et al. ZO-1 protein is required for hydrogen peroxide to increase MDCK cell paracellular permeability in an ERK 1/2-dependent manner. Am. J. Physiol. Cell. Physiol. 315, C422-C431 (2018).
S. DuPont; S. A. Wickström. Mechanical regulation of chromatin and transcription. Nature Reviews Genetics 23, 624-643 (2022).
N. Gao, et al. Respiratory syncytial virus disrupts the airway epithelial barrier by decreasing cortactin and destabilizing F-actin. J. Cell Sci. 135, jcs259871 (2022).
E. Latorre, et al. Active superelasticity in three-dimensional epithelia of controlled shape. Nature 563, 203-208 (2018).
L. Li, et al. Fluoride disrupts intestinal epithelial tight junction integrity through intracellular calcium-mediated RhoA/ROCK signalling and myosin light chain kinase. Ecotoxicol. Environ. Saf. 257, 114940 (2023).
R. Fabbri, et al. Graphene oxide electrodes enable electrical stimulation of distinct calcium signalling in brain astrocytes. Nat. Nanotechnol. 19, 1344–1353 (2024).
S. S. Thakur, et al. Validation of hyaluronic acid-agar-based hydrogels as vitreous humor mimetics for in vitro drug and particle migration evaluations. Eur. J. Pharm. Biopharm. 148, 118-125 (2020).
V. Lambert, et al. Laser-induced choroidal neovascularization model to study age-related macular degeneration in mice. Nat. Protoc. 8, 2197-2211 (2013).
T. N. Yamamoto, et al. Developing neoantigen-targeted T-cell–based treatments for solid tumors. Nat. Med. 25, 1488-1499 (2019).

There is NO Competing Interest.

SIWearableelectrodrivenswitchactivelydeliversmacromoleculardrugstofundus.docx

Download PDF