Self-generating electricity system driven by aqueous humor flow and trabecular meshwork contraction motion activated BCKa for glaucoma intraocular pressure treatment

Abstract

Primary open-angle glaucoma (POAG) is the most common form of glaucoma and the leading cause of irreversible vision loss and blindness worldwide. Intraocular pressure (IOP) is the only modifiable risk factor, and prompt treatment to lower IOP can effectively slow the rate of vision loss due to glaucoma. Trabecular meshwork (TM) cells can maintain IOP homeostasis by correcting and adjusting the resistance to aqueous humor outflow in response to sustained pressure changes. TM cells’ function is reduced, and membrane ion channels are impaired in POAG. The dysfunction of Large conductance Ca²⁺-activated K⁺ (BKCa) plays a central role in the pathogenesis of POAG. In this work, we targeted MXene nanoparticles (MXene-RGD) with piezoelectric response to TM cells in a 3D model of glaucoma in vitro as well as in the rabbit Transient Ocular Hypertension (OHT) Model in vivo. MXene-RGD gives the TM electromechanical transfer properties, while the self-enhancing and self-generated electricity properties of the TM are determined by the aqueous humor flow rate and the size of the deformation of the TM. MXene-RGD is nontoxic, as illustrated by a cell toxicity study and histological examination. In a 3D in vitro model of high-pressure glaucoma, whole-cell patch-clamp confirmed that piezoelectric stimulation turns on BKCa, which reduces the volume of the cell. MXene-RGD was injected into the anterior chamber with minimal trauma, i.e., anterior chamber injection, and specifically targeted to TM cells. The OHT model in vivo confirmed the potential IOP-lowering ability of MXene-RGD. We evaluated the ion channels involved in the reduction of IOP by MXene-RGD by pre-treatment with a BKCa channel blocker (iberiotoxin, IbTX) and a voltage-gated Ca²⁺channel blocker (nifedipine). Quantitative qPCR analysis showed that MXene-RGD inhibited the upregulation of mRNA expression levels of the myofibroblast marker α-smooth muscle actin (α-SMA) and the inflammatory response marker interleukin-6 (IL-6) induced by IOP. Histology confirmed that MXene-RGD attenuated IOP-induced proliferation and collagen production in the TM. Taken together, we present for the first time a minimally invasive surgical approach for targeting TM cells for POAG by utilizing piezoresponse nanomaterials to target BKCa to repair or awaken the ability of TM cells to regulate IOP homeostasis on their own.

---

New concepts

In this manuscript, we present for the first time a minimally invasive surgical approach for targeting TM cells for POAG by utilizing piezoresponse nanomaterials to target BKCa to repair or awaken the ability of TM cells to regulate IOP homeostasis on their own. We targeted MXene nanoparticles (MXene-RGD) with piezoelectric response to trabecular meshwork cells in a 3D model of glaucoma in vitro as well as in the rabbit Transient Ocular Hypertension Model in vivo. With its piezoelectric and self-generated electrical properties, MXene-RGD is a great mechanoelectric transducer that activates the BKCa channels with an appropriate exogenous voltage. This regulates cell volume and contraction, reduces IOP, and prevents excessive extracellular ECM deposition caused by high IOP. We believe that MXene-RGD could be the best alternative to traditional ophthalmic formulations as a novel, more effective, and safer nanomedicine.

Introduction

Glaucoma is the leading cause of irreversible blindness, of which primary open-angle glaucoma (POAG) is the most common type.^1,2 Elevated intraocular pressure (IOP) is the only known modifiable risk factor for glaucoma,³ and IOP reduction is the only effective treatment.⁴ The major outflow pathways of aqueous humor (AH) include the trabecular meshwork (TM) and the Schlemm's canal (SC). TM cells actively regulate AH outflow resistance, thereby maintaining physiologic IOP and modulating IOP homeostasis.⁵ During POAG, TM cell function decreases, and it loses its ability to regulate IOP homeostasis,⁶i.e., IOP homeostatic imbalance. TM dysfunction can lead to increased intraocular pressure and optic nerve damage, ultimately leading to glaucoma⁷ (Scheme 1A).

Scheme 1 Schematic illustration of the role of piezoresponse nanomaterial (MXene-RGD)-based targeting of BKCa on the homeostatic regulation of intraocular pressure in glaucomatous trabecular meshwork cells.

The three commonly used ways to lower IOP are drugs, lasers, and surgery. Currently, IOP-lowering drugs either inhibit the production of AH or promote AH outflow in the uveoscleral pathway, and very few drugs target the trabecular meshwork pathway. Rho-associated protein kinase inhibitor (ROCKi) is a new class of IOP-lowering drug in the world that targets the TM cells, reducing IOP by decreasing the contractility of the TMC and relaxing the TMC.^8,9 Although ROCKi has been approved by the FDA, its promotion and application have been limited by obvious side effects⁴ such as conjunctival hyperemia, conjunctival hemorrhage, cornea verticillate, blepharitis, and allergic conjunctivitis.¹⁰ As a result, there has been no breakthrough in targeting TM cell function to treat POAG. TM cells phagocytose pigments, cellular debris, metabolic wastes, and reactive oxygen species that are shed in the AH.¹¹ Surgical removal or ablation of the trabecular meshwork leads to a deficiency of antithrombotic substances in the AH, such as tissue plasminogen activator (tPA), which in turn causes blockage of the scleral collector channels or anterior aqueous veins.¹² Therefore, ignoring the function of TMCs to passively unblock AH is not an ideal way to lower IOP. Targeting the TMC, repairing or awakening the TM cell's own ability to regulate IOP homeostasis, is a more optimal choice for treating glaucoma.

It has been shown that TM can regulate the outflow resistance and thus IOP by changing the shape, volume, and contractile properties of TM cells.¹³ TM cell volume is influenced by the activities of the Na⁺–K⁺–2Cl⁻ cotransporter, the Na⁺/H⁺ exchanger, and K⁺ and Cl⁻ channels.¹⁴ Large conductance Ca²⁺-activated K⁺ (BKCa) channels have been shown to regulate TM cell volume and contractility, as well as the aqueous humor outflow facility¹⁵ (Scheme 1A). It has been shown that TM cells widely express BKCa and predominantly express BKCa β4 and α1 subunits (the α subunit is encoded by the KCNMA1 gene, and the β subunit is encoded by the KCNMB4 gene), and animal experiments have confirmed that interfering with the activity of BKCa significantly reduces the rate of aqueous humor outflow in mice.¹⁶ The study points to alterations in membrane-associated ion channels as one of the earliest and most fundamental molecular responses of cells against stress injury.¹⁷ Luis et al. indicated that POAG is a membrane ion channel disorder and that BKCa dysfunction plays a central role in the pathogenesis of POAG.¹⁸ Activation of BKCa may be a very promising strategy to modulate IOP and treat early glaucoma.

Ionic conductors that regulate cell volume and shape have been suggested to play an important role in TM cell volume regulation.¹⁹ However, the presence of multiple biological barriers in the human eye, as well as the existence of a short window of availability of the drug, leads to an increased need for frequent daily dosing regimens, which may result in concomitant side effects and even reduced patient compliance.^20–23 This demonstrates the significant unmet need for current POAG treatment strategies, as well as the urgent need to develop a more effective and safer glaucoma treatment strategy that achieves more reliable IOP control.

Recently, nanomedicines have emerged as the best alternative to conventional ophthalmic formulations due to the advantages of increased barrier permeability, sustained drug release, tissue targeting, and reduced systemic absorption of titrated drugs.²⁴

BKCa channels can be activated by changes in membrane potential and intracellular calcium concentration.²⁵ Piezoelectric nanomaterials have been proposed as mechanical–electric transducers to convert mechanical forces into locally generated electric fields.^26,27 There is some experimental evidence that piezoelectric stimulation is capable of modulating ion channel currents.²⁸

MXenes, as a novel two-dimensional nanomaterial composed of transition metal complexes, can be formed by MAX Ti₂AlC etching in the form of three to seven extended detached Ti–C layers, which promotes unrestricted electron transfer between the layers and thus exhibits high electronic conductivity.²⁹ Ti₃C₂T_x-MXene has proved to be a good conductor of electricity³⁰ and the most popular material for constructing conductive networks for conductive hydrogels.³¹ MXene has transition metal (Ti) carbide cores with high electrical conductivity and functional surfaces with reactive chemical groups, such as –O and –OH³² to create specific surface charge distributions. These surface terminations make MXene hydrophilic. This can reduce the internal resistance of the hydrogel and have charge trapping and rapid charge transfer ability.³³ MXene hydrogel has been used to create flexible electronic devices with self-powered properties.³⁴ Functional groups on the surface of MXene break the inversion symmetry of the lattice structure and thus have piezoelectric properties. The cyclic strain can excite a stable oscillating piezoelectric voltage and current output.³⁵

In this work, we targeted MXene nanoparticles (MXene-RGD) with piezoelectric response to trabecular meshwork cells in a 3D model of glaucoma in vitro as well as in the Transient Ocular Hypertension Model in vivo. With its piezoelectric and self-generated electrical properties, MXene-RGD is a great mechanoelectric transducer that activates the BKCa channels with an appropriate exogenous voltage (Scheme 1B). This regulates cell volume and contraction, reduces IOP, and prevents excessive extracellular ECM deposition caused by high IOP. MXene-RGD could be the best alternative to traditional ophthalmic formulations as a novel, more effective, and safer nanomedicine.

Results and discussion

The characterization of MXene

The surface of Ti₃C₂ nanosheets is enriched with –OH groups, which facilitates the reaction with silane coupling agents to aminate MXene.³⁶ RGD peptides antagonize αvβ3 and αvβ5 integrins, which are readily expressed by cells of the TM.³⁷ To target nanoparticles (MXene-RGD) to HTM cells, amine-MXene was functionalized with the αvβ3 specific ligand cyclo (Arg-Gly-Asp-D-Phe-Cys). Fig. 1A and D shows the bright field transmission electron microscope (TEM) of the lamellar MXene and MXene-RGD. The morphology of the MXene and MXene-RGD was characterized by field-emission scanning electron microscopy (FE-SEM), which shows a two-dimensional structure with lateral sizes of 1–3 μm (Fig. S1A, ESI†). Based on the higher magnification image (Fig. S1B, ESI†), the thickness of one single layer MXene and MXene-RGD can be estimated to be around 1–2 nm. The selective area electron diffraction (SAED) pattern (Fig. 1B and E) demonstrates a good single crystallinity of the MXene and MXene-RGD nanosheets. Energy dispersive X-ray (EDX) spectroscopic mapping (Fig. 1C) reveals a uniform distribution of Ti and C elements in the MXene nanosheets. Besides, EDX spectroscopic mapping clearly demonstrates the uniform distribution of C, Ti, N, O, and Si elements in MXene-RGD nanosheets (Fig. 1F), respectively. It can be seen that the rich functional groups on the surface of MXene can help to construct a protective layer of homogeneous silanization. The Fourier transform infrared spectroscopy (FTIR) spectra of MXene-RGD and MXene showed obvious bands at 1665 cm⁻¹ (amide I, NC = O) due to the contribution of the peptide molecules, further confirming the successful surface linkage of RGD (Fig. 1G). As shown by the XRD patterns, the occurrence of MXenes is evidenced by the presence of 3 peaks: an intensive peak at 2θ = 5.66 and 2 weaker ones at 13.1, and 60.6, corresponding to the planes with Miller indexes of (002), (004), and (0016), respectively, in the lamellar structure of MXene. All the diffraction peaks were well-matched with the standard pattern for MXene (JCPDS no. 17-438). No changes in the XRD results were observed after grafting RGD onto MXene (Fig. 1H). XPS analysis was used to confirm the covalent grafting of the RGD peptide onto the MXene. The existence of C, N, Ti, O, and Si elements is identified by the XPS spectrum of the as-synthesized MXene-RGD (Fig. 1I), and this result is consistent with EDS mapping (Fig. 1F(i)–(v)). The O 1s spectrum (Fig. 1J) exhibits four peaks at 528.9, 530.1, 531.0, and 531.9 eV, which are associated with the Ti–O, Si–O, C–O, and C=O orbitals, respectively. The C=O and C–O orbitals are derived from the RGD peptide and oxygen-containing functional groups on the Ti₃C₂T_x surface, the Ti–O traced the oxidation of MXene, and the Si–O was due to the silanization by APTES. Simultaneously, the saturated end bonds (C–N) are traced to the amination reagent, and new covalent bonds (N–O) are formed during amination (Fig. 1K). The TGA data (Fig. 1L) showed the amount of grafted RGD was approximately 5.6 wt% weight loss from 55 °C to 700 °C, further confirming the successful modification of RGD peptide. The changes on zeta potential of MXene and MXene-RGD indicated a successful conjugation of MXene and RGD due to the positive potential of RGD (Fig. 1M). The dynamic light scattering determines that the average hydrodynamic diameters of MXene and MXene-RGD in aqueous solutions were around 636 and 1683.4 nm, respectively (Fig. 1N and O).

Fig. 1 Characterization of the MXene and MXene-RGD. A low-magnification image of the MXene (A) and MXene-RGD (D) nanosheets acquired by TEM. A SAED pattern of the MXene (B) and MXene-RGD (E) nanosheets. EDX maps of the MXene (C) and MXene-RGD (F) nanosheets. The FTIR spectra (G), XRD pattern (H), X-ray photoelectron spectroscopy (XPS) (I), and zeta potential (M) of MXene and MXene-RGD nanosheets. High-resolution XPS spectra of O 1s (J) and N 1s (K) for MXene-RGD. (L) TGA curves of MXene and MXene-RGD under nitrogen. Particle-size distribution of MXene (N) and MXene-RGD (O).

The construction and characterization of 3D human trabecular mesh in vitro model

In order to develop effective glaucoma treatments, a better understanding of the behavior of human TM cells and the response of various biological interactions within the TM three-dimensional (3D) complex is needed.³⁸ Traditional 2D models lack the 3D complexity of natural TM. 3D culture models are designed to simulate appropriate cell–cell and cell–environment interactions, providing the complex biochemical and physical signals found in in vivo tissue structures.^39,40 Ideally, 3D cell culture matrices can reconstruct extracellular cell matrix (ECM) features to better mimic the in vivo environment and serve as good models for in vitro cellular studies. Therefore, we first conducted some research to simulate the 3D environment to better study TM.

Hydrogels provide a relatively good simulation of the natural tissue environment and therefore serve as good models for in vitro cellular studies, and they are heavily used in tissue engineering applications. Copolymerized acrylamide (AM) and gelatin methacryloyl (GelMA) hydrogels are widely used as scaffolds in tissue engineering due to their ability to promote cell adhesion, proliferation, and growth by providing favorable cell–matrix interactions.⁴¹ Flexible PAM chains enhance brittle gelatin networks. To further increase the 3D complexity of our model, hyaluronic acid (HA) was also included in the hydrogel,⁴² from which we prepared a 3D human trabecular mesh AM/GelMA/HA hydrogel as an in vitro model.

To investigate whether piezoelectric stimulation affects the behavior of TM cells. We mixed MXene into the hydrogel at different concentrations (Fig. 2A). To prepare chemical-crosslinked hydrogels, the methacryl groups were introduced into the gelatin and HA main chains. In the HA–MA spectrum, all the HA signals are present plus a signal at 1723.32 cm⁻¹, which was associated with the stretching of the carbonyl of the ester group (Fig. 2B). We also examined the replacement of amino groups on the gelatin side chains by methacrylic anhydride using FTIR spectroscopy (Fig. 2C). We show peaks at 1636 cm⁻¹, 1539 cm⁻¹, and 1236 cm⁻¹, indicating the formation of C=O stretching (amide I), N–H bending (amide II), and C–N stretching plus N–H bending (amide III), respectively. In Ti₃C₂T_x MXene, Ti and C atoms are arranged alternately in the order of Ti/C/Ti/C/Ti, and the T_x end functional groups are distributed on the surface of the layered MXene,⁴³ which provided additional hydrogen bond (H-bond) binding sites and hence promoted the crosslinking of the hydrogel (Fig. 2A). In AM/GelMA/HA, secondary crosslinking can also be initiated between MXene, amino groups on the PAm chain, hydroxyl groups on HA and GelMA, and carboxyl groups on the HA chain. The photographs of the series of MXene hydrogels are shown in Fig. 2D. It can be seen that the hydrogel changed from a transparent gel to a uniform black gel with the increase of MXene content. In addition, it can be seen in the photos that there is no black agglomerate in the hydrogel, indicating that the MXenes are uniformly cross-linked or dispersed in the hydrogel. In this work, hydrophilic MXene was incorporated into the hydrogel network, altering the initial hydrophilic/hydrophobic balance of the backbone. Fig. 2E shows the swelling ratios of the hydrogels. The swelling ratios of hydrogels slightly increase with the addition of MXenes (0.5%). All hydrogels swelled rapidly and reached equilibrium within 12 hours. The degradation rate of hydrogel affects cell growth.⁴⁴ To prove the applicability of hydrogels for HTM tissue engineering, the degradation of hydrogels with different MXene concentrations in PBS solution was observed over a period of 28 days (Fig. S2, ESI†). Under the same conditions, the degradation rate of AM/GelMA/HA 0% MXene hydrogels is slower than that of AM/GelMA/HA 4% MXene hydrogels. Both hydrogels showed a slow degradation rate. The hydrogel was only 6.5% (0% MXene) or 7.2% (4% MXene) degraded after 28 days. From this, it is concluded that hydrogel has applicability and can be used for a long time. Fig. 2D shows the FESEM images of the fractured cross-sections of freeze-dried hydrogels. The observed general morphological trend is that as the MXene content increases from 0% to 8%, the pore size first increases and then decreases. This also explains the change in the swelling ratio (Fig. 2E), but still allows for a large amount of water to be stored and transferred. Fig. 2D also shows the Ti element evenly distributed in the hydrogels analyzed by EDX elemental mapping. The chemical element data obtained by EDX for all hydrogel groups, including the atom (%) and weight (%) percentage of C, O, and Ti, are presented in Fig. 2D.

Fig. 2 Characterization of the hydrogel. (A) Schematic illustration of the preparation and synthesis process of AM/GelMA/HA hydrogel. FTIR spectra of (B) hyaluronic acid (HA), methacrylated hyaluronic acid (HA–MA), and (C) gelatin, gelatin methacryloyl. (D) Representative images, FESEM images, and the EDX mapping images of the Ti in the hydrogels. (E) Swelling ratio of hydrogels (n = 4).

Mechanical–electric conversion properties of the MXene composite 3D human trabecular mesh in vitro model

Many studies have combined MXene nanomaterials with biopolymers to synthesize new nanocomposites with better mechanical properties, electrical conductivity, and self-powered properties.^45,46 Ti₃C₂T_x MXene has excellent properties such as abundant interlayer ion diffusion paths, electron-conducting carbide core layers, and interlayer ion storage sites, which can be used to construct mechanically reliable hydrogel electrolyte-based supercapacitors.⁴⁷

MXene, i.e., anisotropic crystals, inside the hydrogel can generate an electric dipole when mechanically squeezed. The electric dipole is also called piezoelectricity.⁴⁸ When the hydrogel is subjected to a certain direction of external force, the internal electrode phenomenon occurs; at the same time, the surface produces the opposite sign of the charge; when the external force is withdrawn, the hydrogel is restored to its uncharged state; when the direction of the external force is changed, the polarity of the charge is also changed; and the amount of charge produced by the force is directly proportional to the magnitude of the external force⁴⁹ (Fig. 3A).

Fig. 3 Mechanical–electric conversion property tests of the hydrogel. (A) Operating principles of hydrogel piezoelectric response. (B) The triboelectric mechanism based on microchannels of the hydrogel. (C) Changes in AM/GelMA/HA 4% MXene hydrogel capacitance during multiple presses. (D)–(G) The open-circuit voltage of the hydrogel for different doping concentrations of MXene nanosheets. (I)–(M) Photograph of an LED lit by hand tapping of the hydrogel.

MXene nanosheets are characterized by abundant interlayer ion diffusion paths, electronically conductive carbide core layers, and interlayer ion storage sites that allow for the insertion and diffusion of a large number of ions.⁵⁰ The abundant surface functional groups in MXene nanosheets provide additional hydrogen bonding sites, which promote cross-linking of the hydrogel and facilitate good contact with the hydrogel electrolyte. Therefore, the surface of MXene nanosheets can be regarded as a microchannel, which is filled with molecular chains and a large amount of water. When water rubs against the MXene nanosheets, the surface will carry negative charges, while water near the surface will produce positive charges, forming a double electric layer and a microchannel that is in equilibrium without external vibration.⁵¹ When pressure is applied to the hydrogel, the microchannels are compressed, and water will flow out of the microchannels. The flowing water will push the positive counterbalance ions outside from the microchannels of the electric double layer but retain negative charges on the surface of the MXene (Fig. 3B).

The adjacent MXene sheets can act as the conductive layers of a microcapacitor, and the polymer between the MXene sheets acts as the dielectric of the microcapacitor. AM/GelMA/HA hydrogels can be used as an ideal electrode material for building supercapacitors with excellent performance. The whole system can be considered a bulk capacitor junction formed by the series–parallel connection of numerous microcapacitors (Fig. 3B).

The MXene doping concentration was studied to optimize the hydrogels’ output performance (Fig. 3D–H). When the hydrogels were pressed with a 20-gram weight, they were able to generate an open-circuit voltage of 20–40 mV (Fig. 3G and H). In contrast, in hydrogels without MXene nanosheets, 0% MXene did not generate a voltage under the same manual tapping (Fig. 3D), whereas the 0.5% and 1% MXene hydrogels could only generate a very low voltage (Fig. 3E and F). The output performance of the hydrogels is similar when the MXene doping concentration increases from 4% to 8% (Fig. 3G and H). This decreased performance is likely due to an agglomeration of excessive MXene nanosheets leading to longer electron transfer paths, which greatly suppresses the ion transport in the hydrogel. In addition, the porosity also affects the output performance of the hydrogel.⁵² As the MXene content increases from 4% to 8%, the pore size decreases (Fig. 2D). The decrease in pore size and spacing intensifies the electrostatic shielding effect of the pores, hindering further adsorption of conductive ions, and leading to a reduction in the electrical output of hydrogel.⁵³ It is also clear that with no MXene of AM/GelMA/HA hydrogels was just around 4.01 μS cm⁻¹, whereas the conductivity almost doubled (7.92 μS cm⁻¹) once 4% MXene nanosheets were incorporated (Table S3, ESI†). The obvious increase in electrical conductivity was attributed to the formation of a well-connected MXene network with superior electron transport ability.

Next, we measured the change in capacitance at cyclic pressures. Cyclic pressure tests indicated that the AM/GelMA/HA 4% MXene hydrogel could quickly respond to changes in pressure and return to its original state when the pressure was released (Fig. 3C). The decrease in C/C₀ after cyclic pressure is due to the continuous fracture and surface abrasion of the contact interface during “contact–separation”, which has an adverse effect on the stability of power generation.⁵⁴

Then, the hydrogels attempted to light up several light-emitting diodes (LEDs), whose circuit diagram is shown in Fig. 3I–M. LEDs were easily illumined by directly tapping the hydrogel with hands (Fig. 3L, M, and Videos S1, S2, ESI†), revealing great application potential for the 4% and 8% hydrogels in low-frequency mechanical energy harvesting.

Several studies have reported the successful in vitro stimulation of neuronal cells by a variety of piezoelectric nano-transducers with a range of voltages^27,55 that have been shown to be effective, safe, and efficient.⁵⁶ Piezoelectric stimulation has been used to activate voltage-gated membrane channels in neuronal cells, and the resulting voltages are compatible with the values required to activate voltage-sensitive channels (Fig. 3G).²⁷ These results indicated that AM/GelMA/HA 4% MXene hydrogel has a high mechano–electrical conversion efficiency, making it more potent for generating cell-scale electrical stimulations to cells.⁵⁷

Glaucomatous 3D human trabecular meshwork in vitro disease model

To engineer an in vitro HTM model that recapitulates the pathology of glaucoma, we inoculated HTM cells onto AM/GelMA/HA hydrogels for 3D cell culture. We first examined cell viability by live/dead staining. We demonstrated that those hydrogels are non-cytotoxic (Fig. 4A).

Fig. 4 3D model of human trabecular meshwork of glaucoma. (A) Live/dead assay images of cell culture on hydrogels with 0% and 4% MXenes. (B) Schematic illustration of high-pressure cultivation model. Morphological changes in 3D-TM cells. The effect of TGF-β2 on human TM morphological changes (C), cell contraction (D), mRNA expression (E) (α-SMA, COL1A1, and FN), and cell volume (n = 25) (F). The effect of oxidative stress on human TM morphological changes (G), cell contraction (H), mRNA expression (I) (α-SMA, COL1A1, and FN), and cell volume (n = 25) (J). The effect of elevated pressure on human TM morphological changes (K), cell contraction (L), mRNA expression (M) (α-SMA, COL1A1, and FN), and cell volume (n = 25) (N). Scale bar = 50 μm.

Because of POAG's multifactorial etiology, determining the pathogenesis is complex. Choosing an in vitro model with a “closer in vivo” microenvironment and structure can identify biomarkers involved in the early stages of glaucoma (e.g., dysfunction of trabecular meshwork cells to regulate their volume) and contribute to developing new therapeutic approaches. The bioengineered 3D TM cell structures mimicked the TM structure found in vivo conditions in terms of their spatial distribution, ECM synthesis, and secretion, as well as their responsiveness after elevated pressure-induced mechanical strain and drugs, i.e., TGF-β2 or steroids. We used three experimentally induced glaucoma models in our study.

It has been shown that levels of TGF-β2 are elevated in eyes of glaucomatous patients compared to age-matched normal eyes.⁵⁸ We first explored the effects of TGF-β2 on human TM cells in the context of AH efflux. In our studies, cells were exposed to 5 ng mL⁻¹ TGF-β2 for 24, 48, and 72 h. Labeling of F-actin with phalloidin showed that TGF-β2 induced CLAN formation compared to untreated controls (Fig. 4C). We noted significant contraction of the gel after 24 h of TGF-β2 treatment compared to control cells, which were further contracted by 48 and 72 h of treatment (Fig. 4D). After administration of 5 ng mL⁻¹ of TGF-β2 for 48 h, quantitative qPCR analysis showed that the mRNA expression levels of the myofibroblast marker α-smooth muscle actin α-SMA, ECM synthesis-related genes Col-I (COL1A1), and fibronectin (FN) were significantly increased in HTM cells (Fig. 4E). The results showed a significant increase in cell volume after 24 h of TGF-β2 treatment, which was restored after a further 48 and 72 h of treatment (Fig. 4F). This result suggests that short-term induction of TGF-β2 did not cause impaired regulation of trabecular meshwork cell volume.

Oxidative stress (OS) in the TM has been identified as a major contributor to glaucoma pathology,⁵⁹ and H₂O₂ is the most widely used agent for inducing oxidation in TM cells in vitro.⁶⁰ In order to study the effects of oxidative stress on TM, which is one of the main causes of TM damage, TM cells were treated with daily doses of 2 h of 500 μM H₂O₂, and in the remaining time (22 h), they were subjected to recovery.^61,62 Treatment for 72 hours with H₂O₂ (daily doses of 2 h of 500 μM H₂O₂) induced cell–cell adherence into unrecognizable clusters, elongation of cells, and distorted cell morphology (Fig. 4G). The contraction assay revealed that H₂O₂-treated trabecular meshwork cells had no obvious contraction effect on the gel (Fig. 4H). There was a significant decrease in cell viability upon treatment with H₂O₂. Within 48 hours of treating the cells as described above, HTM cells all showed an increase in the nucleus (Fig. 4G) and cell volume size (Fig. 4J), and OS-treated HTM cells showed up-regulation of α-SMA, COL1A1, and FN as compared to controls as analyzed by qPCR (Fig. 4I). After further treatment for 72 h, the expression levels of the relevant genes and cell volume were all down-regulated, which may be related to H₂O₂-induced cell death in HTM cells.⁶³

Elevated intraocular pressure (IOP), accompanying abnormalities in aqueous humor outflow resistance, is found in the majority of glaucoma. We therefore investigated the effects of sustained high IOP on HTM molecules, morphology, and extracellular matrix formation. We performed high-pressure stimulation of HTM using a previously reported high-pressure culture model²⁷ (Fig. 4B). Phalloidin-stained F-actin in HTM cells showed an increase in cellular microactin fibers with increasing pressure, arranged radially or in parallel, and the cells had significantly larger nuclei. Elevated pressure (60, 80 mmHg) activated HTM cells, involving contraction, and upregulation of genes encoding the secreted ECM (Fig. 4M), but only pressure (80 mmHg) had a significant effect on gel contraction (Fig. 4L). TM cells at a pressure of 40 mm Hg had elevated COL1A1 expression, but no significant effect on either α-SMA or FN expression (Fig. 4M). The cell volume of both groups of cells at pressures of 10 mm Hg and 20 mm Hg did not change significantly from the control group after 24 hours. As the pressure increased, the cell volume became significantly larger (Fig. 4N). The above results suggest that trabecular meshwork cells can only bear pressure below a certain level. They may be destroyed structurally or impaired functionally by pressure above this level of the high-pressure cultivation model. The high-pressure cultivation model demonstrated the main feature of impaired volume regulation in trabecular meshwork cells.

In vitro study of piezoelectric stimulation activating BKCa channels in trabecular meshwork cells

BKCa dysfunction plays a core role in the pathogenesis of POAG. HTM cells can grow or decrease their volumes adaptively, regulating the AH to go out. Consider the dysfunction of the conventional drainage of the TM cells if they lose the capacity to regulate their volume, avoiding the right passage of intraocular flux among them. The regulation of the opened-closed processes of voltage-dependent ion channels (BKCa) affects the volume regulation of cell.¹⁸ Voltage-dependent ion channels (BKCa) can be activated by membrane potentials (external voltage or electrical stimulation) and intracellular calcium ions.²⁵ Therefore, we inoculated HTM cells onto AM/GelMA/HA 4% MXene hydrogels in a high-pressure culture model to study the effect of piezoelectrical stimulation of the trabecular meshwork. At pressures of 40–80 mmHg, trabecular meshwork cell volume regulation was impaired, and recovery to the initial volume was not possible within 24 hours (Fig. 5A). The pressure (40 mmHg–80 mmg)-induced cell volume was significantly reduced by pressing the hydrogel (20 g, 0.05 Hz, 30 min) (Fig. 5A).

Fig. 5 Piezoelectric stimulation improves BKCa dysfunction and regulates trabecular meshwork cell volume. (A) Piezoelectrical stimulation abolishes the elevated pressure-induced increases in trabecular meshwork cell volume. (B) BKCa current recorded in a representative HTM cell with the voltage protocol shown in the inset was inhibited by the BKCa blocker iberiotoxin (IbTX). BKCa current recorded (C) and I–V relationships of BKCa current (D) in HTM cells in Scaffolds for different treatment. (E) Confocal fluorescence microscopy images of piezoelectrically stimulated HTM cells. (F) I–V relationships of BKCa current of piezoelectric stimulation of elevated pressure-induced trabecular meshwork cells in the presence of calcium channel inhibitors.

To determine that the effects of piezoelectrical stimulation are occurring through activation of the BKCa channel, we used a whole-cell membrane clamp to record the changes in electrical properties in a three-dimensional culture. Fig. 5B shows the whole-cell macroscopic current recorded in representative cells. The oscillatory current was inhibited by the BKCa channel blocker iberiotoxin (200 nmol L⁻¹, Fig. 5B), indicating a typical BKCa current. BKCa current was greater in normal cells compared to pressure-treated trabecular meshwork cells (Fig. 5C). The oscillatory current recovered somewhat after piezoelectric stimulation (Fig. 5C). Fig. 5D illustrates the voltage–current (I–V) relationships of the normalized BKCa current in HTM with different treatments. The current density (at −80 to +60 mV) of HTM cells treated with elevated pressure (80 mmHg) is lower than that of normal HTM cells (P < 0.05 or P < 0.001). These results suggest that BKCa currents are downregulated in HTM cells with elevated pressure treatment. Piezoelectric stimulation can partially restore BKCa current magnitude (at −80, −40 to +60 mV, P < 0.05 or P < 0.001) (Fig. 5D).

It is well recognized that Ca²⁺ sparks are localized Ca²⁺ signals that activate BKCa channels to mediate cell relaxation.⁶⁴ Piezoelectric-mediated plasma membrane depolarization can cause Ca²⁺ influx.²⁷ We, therefore, examined the effects of piezoelectrical stimulation on intracellular Ca²⁺ dynamics in HTM cells loaded with the rapid Ca²⁺ indicator dye fluo-4 AM.

Time-lapse Ca²⁺ imaging on AM/GelMA/HA 4% MXene hydrogel cultures demonstrated the cells were successfully activated by piezoelectric stimulation (Fig. 5E). The time-lapse video of Ca²⁺ imaging performed on AM/GelMA/HA 4% MXene hydrogel cultures is available as ESI† (Video S3).

Therefore, in the following studies, we wanted to determine if the effects of Ca²⁺ sparks on cell volume were additive in the presence of piezoelectrical stimulation. Voltage-sensitive channels, which are expressed in a variety of cell types, can mediate the entry of ions into excitable cells.⁶⁵ Piezoelectric stimulation can alter the cell membrane potential, leading to the opening of voltage-gated ion channels, particularly Ca²⁺ voltage-gated channels.⁶⁶ TM cells express voltage-dependent L-type Ca²⁺ channels that influence intracellular Ca²⁺ concentration.^67–69 To investigate the ion channels involved in the observed Ca²⁺ transients, whole-cell membrane clamp experiments were performed in the presence of blockers of either of the voltage-gated Ca²⁺ channels (nifedipine). The addition of nifedipine before piezoelectric stimulation significantly inhibited the activation of BKCa currents (at 0 to +60 mV, P < 0.01 or P < 0.001) (Fig. 5F). Here, we show that MXene-induced piezoelectric stimulation of Ca²⁺ influx via voltage-gated Ca²⁺ channels can alter local Ca²⁺ signaling, generating membrane potential hyperpolarization and thereby activating BKCa channels.

In vivo study of MXene-RGD in the transient ocular hypertension model

In vitro, we used AM/GelMA/HA hydrogel to simulate the artificial trabecular mesh, and the 4% MXene incorporation resulted in a high electromechanical conversion efficiency of the trabecular mesh model and stable piezoelectric stimulation of the trabecular mesh cells under in vitro cyclic pressure. To replicate the same model in vivo as in vitro, we targeted MXene with piezoresponse MXene to the trabecular mesh tissue, and the trabecular mesh formed a composite with the targeted MXene equivalent to an AM/GelMA/HA 4% MXene hydrogel (Fig. 6A).

Fig. 6 Effects of MXene-RGD in the transient ocular hypertensive rabbit model. Schematic diagram of piezoelectric responsive MXene-RGD targeting trabecular meshwork (A) and operating principles of MXene-RGD piezoelectric response (B). (C) Representative images and confocal images of MXene-RGD. (D) Cytotoxicity of MXene-RGD after the incubation with HTM cell lines for 24 h. (E) Confocal microscopy pictures of HTM cells incubated with MXene-RGD. (F) Confocal microscopy pictures of trabecular meshwork after 24 hours of injection of MXene-RGD into the anterior chamber of rabbits. (G) Histological examination of the heart, liver, kidney, spleen, cornea, and retina. (H) Intraocular pressure (IOP) time course. (I) mRNA levels of α-SMA and IL-6 were determined by RT-PCR. Histological examination (J) and Masson trichrome staining (K) of the trabecular meshwork in each group.

The trabecular mesh is subjected to many forms of biotic forces during the outflow of house water. In addition to hydrostatic pressure and fluid shear, the trabecular mesh is also subjected to mechanical stresses due to stretching, compression, and torsion from daily activities such as rubbing and blinking the eyes, which can result in a 20–105% change in the dimensions of the trabecular mesh.⁷⁰ The trabecular meshwork dilates when IOP increases and AH flow increases, and when AH is discharged, the pressure peaks decay, the trabecular meshwork retracts to its original position, and the cycle repeats.⁷¹ In the same principle as in vitro hydrogel piezoelectricity, deformation of the trabecular mesh causes MXene to generate a piezoelectric signal; with the larger the deformation. the larger the charge (Fig. 3A and 6A). In the same mechanism as the friction electricity of in vitro hydrogel microchannels, the friction of aqueous humor against MXene nanosheets also generates friction electricity signals ((Fig. 3B and 6B)). Because of the change in flow velocity of the AH and the magnitude of the trabecular mesh motion deformation under IOP stimulation, the trabecular mesh has the property of self-enhancement and self-generated electricity after MXene-RGD targeting to trabecular mesh tissues.

For ease of observation, we labeled MXene with FTIC (Fig. 6C). The cytotoxicity profiles of MXene-RGD were evaluated. No obvious cytotoxicity was observed at a high concentration of 10 mg mL⁻¹ after coincubation for 24 h, indicating a relatively high biocompatibility of MXene-RGD (Fig. 6D). HTM cells were incubated with MXene-RGD for 24 h. After removing non-bound nanoparticles and thorough washing, targeted nanoparticles were observed to bind to the HTM cell through confocal laser scanning microscopy (Fig. 6E). After 24 hours of injection of MXene-RGD into the rabbits' anterior chamber, particles labeled with FTIC were observed at the trabecular meshwork (Fig. 6F).

Two days after the injection of MXene-RGD into the anterior chamber, rabbits were sacrificed by the injection of excess barbiturate in the ear vein. After the eyeball was removed, the white rabbits were dissected, and the rabbits' hearts, liver, kidneys, cornea, and retina were removed to observe the toxic effects of MXene-RGD on themselves. Similar to the control group, MXene-RGD groups had no pathological changes in the heart, liver, kidneys, cornea, or retina, and the tissue remained normal, demonstrating that MXene-RGD is biocompatible (Fig. 6G). We first confirmed that MXene does not affect normal rabbit intraocular pressure (Fig. 6H).

We used a transient ocular hypertension (OHT) model induced by injection of 100 μL of 5% hypertonic saline into the vitreous to address the potential IOP-lowering ability of MXene-RGD (75 μL, 10 mg mL⁻¹ MXene-RGD in the anterior chamber). IOP rose from 15.025 ± 0.842 mmHg at baseline to 41.76 ± 4.73 mmHg after hypertonic saline injection. In the MXene-RGD experimental set, IOP rose from 16.145 ± 1.68 mmHg at baseline to 34.69 ± 2.63 mmHg after hypertonic saline injection (Fig. 6H). The MXene RGD group accelerated the regulation of intraocular pressure and returned to normal intraocular pressure values (17.918 ± 4.134 mmHg) at 90 minutes. We evaluated the ion channels involved in the reduction of IOP by MXene-RGD by pre-treatment with a BKCa channel blocker (iberiotoxin, IbTX) and a voltage-gated Ca²⁺channel blocker (nifedipine). The IOP-lowering activity of MXene-RGD was suppressed by IbTX and nifedipine pre-treatment (Fig. 6H). We also state here that stimulation of Ca²⁺ influx via voltage-gated Ca²⁺ channels by MXene-induced piezoelectric stimulation can alter local Ca²⁺ signaling, generating membrane potential hyperpolarization and thereby activating BKCa channels. Quantitative qPCR analysis showed that MXene-RGD inhibited the upregulation of mRNA expression levels of myofibroblast marker α-smooth muscle actin (α-SMA) and inflammatory response marker interleukin-6 (IL-6) induced by IOP (Fig. 6I).

To determine whether MXene-RGD can alleviate the effects of IOP in the TM, we compared the results of HE staining (Fig. 6J) and Masson's trichrome staining (Fig. 6K) of TM sections in each group. The control group showed the structure of normal TM under the HE staining. Normal trabecular meshwork tissue consists of irregular, porous laminar material, resembling a porous sponge structure. Trabecular meshwork gradually aggregated after stimulation with high IOP, and the space between trabecular columns decreased. MXene-RGD regulated the aggregation of trabecular meshwork, and trabecular meshwork organization was sparser when compared with that in the high IOP group (Fig. 6J). It has been reported that IOP can trigger hyperplasia, fibrosis, and collagen production in vitro. The results of Masson trichrome staining confirmed that MXene-RGD attenuated IOP-induced collagen production (Fig. 6K).

Experimental

Materials

The materials, antibodies, primers, and cells used in this study were purchased from varying suppliers. MXene Ti₃C₂(–OH) was the product of Beike 2D Materials Co., Ltd (Beijing, China); alginate with an average viscosity of 85 mpa s was purchased from Qingdao Hyzlin Biology Development Co., Ltd (Qingdao, China); methacrylic anhydride (MA), ammonium persulfate (APS), paraformaldehyde (PFA), and N-isopropylacrylamide were obtained from Sigma-Aldrich Co. (St. Louis, MO, USA); trizol from Thermo Fisher Scientific (Waltham, MA, USA); and DEPC from Biotopped (Beijing, People's Republic of China). Antibodies for immunohistochemistry assays were obtained from Boster Bio. Fluo-4 AM (calcium ion fluorescent probe, 2 mM) was the product of Beyotime Biotechnology (Shanghai, China).

Acquisition of materials

Polysaccharide syntheses. The methacrylated hyaluronic acid (HA–MA) was prepared by reacting a 1% w/v solution of hyaluronic acid with a 20-fold excess of methacrylic anhydride and adjusting the pH to 8 with NaOH. The solution was allowed to react for 24 hours at 48 °C. The modified hyaluronic acid was purified via dialysis against MQ water (MW 5–8 kDa) for 48 h to remove excess methacrylic anhydride, and the final product was recovered by lyophilization.

GelMA was synthesized from the reaction of methacrylic anhydride and gelatin; specifically, 10% (w/v) gelatin solution was prepared in phosphate buffered saline (PBS; pH 7.4, 10 mM) at 60 °C; methacrylic anhydride (520 μL) and anhydrous ethanol (1.0 mL) were solubilized in the gelatin solution and stirred at 42 °C in the dark for 4 hours, which was then subjected to dialysis (4000 Da cut off) against MQ H₂O for 3 days at 60 °C.

MXene-RGD. The 5 mg of MXene were dispersed in ethanol (10 mL), followed by the addition of (3-aminopropyl) triethoxysilane (APTES, Sigma-Aldrich, Shanghai, China) (100 μL) dropwise and reflux for 12 h at room temperature to obtain amino-modified MXene (MXene-NH₂), which was then dispersed into 1 mL of ethanol after washing twice with ethanol. Add 16 μL 40 mmol of FITC solution to the reaction system and react for 5 hours at 4 °C, protected from light. After washing twice with ethanol, it was dispersed into 25 mL of phosphate buffer solution (PBS). Then cRGD (MCE, Shanghai, China) (0.1 mmol) was added to MXene-NH₂ dissolved in PBS and stirred for 24 hours at RT. The final product of MXene-RGD was collected by centrifugation after washing with deionized water for three times.

Preparation of the hydrogel

The AM/GelMA/HA was prepared by a chemical cross-linking strategy. Firstly, AM–GelMA–HA–MA/MXene solution was prepared by dissolving 0.5 g GelMA, 1 g AM, and 0.02 g HA–MA into 10 mL MXene solution with various concentrations (0, 0.5, 1, 4, 8%) under magnetic stirring for 1 h at 65 °C. Secondly, add 4% APS to the mixed solution and pour it into a mold. React at 65 °C for 15 minutes to form a cross-linked network. And the hydrogels were coded as AM/GelMA/HA 0% MXene, AM/GelMA/HA 0.5% MXene, AM/GelMA/HA 1% MXene, AM/GelMA/HA 4% MXene, and AM/GelMA/HA 8% MXene, respectively.

Characterization

MXene and MXene-RGD characterization. The nanoparticles were sonicated to obtain a stable 1 mg mL⁻¹ nanoparticle dispersion. The obtained dispersion was characterized through transmission electron microscopy (TEM, Zeiss 902). The TEM samples were prepared by dropping several drops of the sample, diluted in DI water, onto a copper grid, and drying in air. Moreover, particle size distribution and Z-potential were analyzed with a Nano Z-Sizer 90 (Malvern Instrument). Thermogravimetric analysis (TGA) was carried out using a Netzsch STA 449F3 thermal analyzer with a heating rate of 10 K min⁻¹ under nitrogen. The XRD patterns of the MXene and MXene-RGD were recorded on a Rigaku D/MAX 2500PC powder XRD instrument with Cu K_α radiation operating at 40 kV and 200 mA, in the scanning range of 5°–80° with at a scan rate of 8° min⁻¹. Field-emission scanning electron microscopy (FE-SEM) images were obtained using an SEM system (Model S-4800, Hitachi, Japan). The chemical compositions of the samples were further analyzed using high-resolution X-ray photoelectron spectroscopy (XPS) (ESCALAB250Xi, ThermoFisher).

Fourier-transform infrared spectroscopy (FTIR). To confirm the structure of HA, HA–MA, gelatin, GelMA, MXene and MXene-RGD FTIR was performed by using a Shimadzu FTS400 spectrometer (Shimadzu Corporation, Kyoto, Japan).

Field emission scanning electron microscopy. To better reserve the microstructure of hydrogels prepared above, swollen equilibrated hydrogel samples were frozen in liquid nitrogen and then freeze-dried under vacuum at −50 °C for at least 24 h until all the water was sublimed. For imaging, hydrogel samples were placed on double-sided graphite tape, attached onto a metal surface, and sputter-coated with gold for 10 s. We then imaged the cross-sectional morphology of hydrogels using FE-SEM (HITACHI S4300-SEM; Hitachi Corporation, Tokyo, Japan) with secondary electrons at an acceleration voltage of 1.0 kV, coupled with energy-dispersive X-ray (EDX) analyzer.

Equilibrium swelling ratio. The prepared hydrogels were cut into 10 mm lengths and put into a 15 mL centrifuge tube in 10 mL of PBS, and then the tube was held at 37 °C for different time points (1, 2, 3, 4, 5, 6, 7, 12, 24, 48 h). The swelling ratio of the hydrogels was calculated using the following equation.
where W_s stands for the swollen weight of the polymer hydrogel (g), which is recorded at regular time intervals after the excess surface water was carefully adsorbed with filter paper; W_d represents the initial weight of the dried hydrogel (g). Four samples were tested for each group (n = 4).

Degradation properties of hydrogel. The degradation study of hydrogel was carried out in phosphate-buffered saline (PBS) at 37 °C. The hydrogel was put into a six-well plate, and 3 mL of PBS was added to each well and replaced once a day. At the set time points, the hydrogel was removed and blotted gently with filter paper to remove surface water. The hydrogels were freeze-dried to obtain dry weight (W_d). The degradation percentage of the hydrogels at different points was calculated as the equation:
where W_i is the initial weight of the hydrogel.

Conductivity test. The conductivity of hydrogels was measured by an LCR meter (VICTOR 4091C) with applied voltage of 1 V and measuring frequency of 1 kHz. The conductivity (K) is calculated by the following equation.

K = L/RA

in which R is the resistance and L and A are the length and cross-sectional area of hydrogel samples, respectively. Four samples were tested for each group (n = 4).

Mechanical–electric conversion property tests of hydrogel. To investigate the piezoelectric output performance of the hydrogel, the hydrogel (30 mm long, 20 mm wide, and 5 mm thick) is connected to the Keithley 6514 electrometer through copper wires. The output voltage induced by the compressive deformation of the hydrogel when pressed by 20 g weights was tested by the electrometer.

In vitro experiments on HTM cells

Cell culture. The HTM cells (Human Trabecular Meshwork Cells) were purchased from PriCells (Wuhan, China). Cells were cultured in DMEM/F12 medium supplemented with 15% (v/v) FBS, 1% (v/v) 100 U mL⁻¹ streptomycin, and 1% (v/v) 100 U mL⁻¹ penicillin. The cells were maintained at 37 °C in a humidified 5% CO₂ atmosphere.

Hydrostatic pressure apparatus. Pressure apparatus were performed using a previously reported method.⁷² 1 × 10⁶ cells were seeded evenly into hydrogels of different concentrations, and the cell–hydrogel constructs were kept in an incubator for 1 day to allow complete adhesion of the cells to the hydrogel. The hydrogel of implanted cells was put into a 10-mL glass culture bottle, and the culture medium was filled into the bottle and sealed with a rubber stopper. The syringe was injected with sterile air through a three-way tube, which was connected to a manometer to measure the pressure, and four experimental groups were set up with pressurization of 1.33, 2.67, 5.33, and 10.67 kPa (1 kPa = 7.5 mmHg), and no pressurization was set up as the control group.

CCK-8. HTM cells were seeded in 24-well plates at a density of 1 × 10⁵ cells per well for 24 h, followed MXene (10 mg mL⁻¹) being immersed in the medium. After 24 h of culturing, 100 μL of well completed medium with CCK-8 solution (10 μL) was incubated in the incubator for 2 h. After shaking the well plate gently, the absorbance value at 450 nm was measured on a multi-functional microplate reader. The cell survival rate was figured by using the following relation.

Relative cell viability (%) = OD_experiment/OD_control × 100.

Optical density (OD) values were calculated from the average of three different experiments.

Live/dead staining. HTM cells were collected and resuspended in culture medium at a density of 5 × 10⁶ cells per mL. In culture dishes, 1 × 10⁶ cells were seeded evenly into hydrogels of different concentrations, and the cell-scaffold constructs were kept in an incubator for 1 day to allow complete adhesion of the cells to the scaffold. Calcein–AM and propidium iodide (PI) solutions were used to perform live/dead staining of the cells. Optical analysis was performed by fluorescence microscopy. Three replicates were used for each sample (n = 3).

Hydrogel contraction assay. 1 × 10⁶ cells were seeded evenly into hydrogels, and the cell-scaffold constructs were kept in an incubator for 1 day to allow complete adhesion of the cells to the scaffold. The hydrogels were pressure-stimulated or drug-treated, and the area of the gels was measured and analyzed after 24, 48, and 72 h using Image J software. The area of the gels containing untreated TM cells was used for normalization, and relative changes were shown as a plot chart.

Measurement of cell volume. Cell volume measurements were performed using the modified protocols of Williamet et al.²⁷ 1 × 10⁶ cells were seeded evenly into hydrogels of different concentrations, and the cell-scaffold constructs were kept in an incubator for 1 day to allow complete adhesion of the cells to the scaffold. Prior to any treatments, cells were loaded with the fluorescent dye calcein–AM in DMEM/F12 at 37 °C in a 5% CO₂ incubator for 60 min to ensure a stable baseline. Images were taken without treatment; this served as the experimental control. Images of cells were taken at different time periods after the pressure stimulation or electrical stimulation or application of the drugs. Cell volume was calculated from image J.

Real-time PCR. Real-time (RT) polymerase chain reaction (PCR) was performed on a 7500 real-time PCR system (Applied Biosystems; Thermo Fisher Scientific) with the Fats Start Universal SYBR Green Master (Rox; Hoffman-La Roche Ltd, Basel, Switzerland) with the following cycling conditions: 5 °C for 2 min, 95 °C for 10 min, 95 °C for 15 s, 60 °C for 1 min, and then step 3 initiated for 40 cycles. The primer sequences for reverse transcription polymerase chain reaction (RT-PCR) are shown in Tables S1 and S2 (ESI†). Gene expression was normalized with the GAPDH expression level. RT-PCR analyses were performed in triplicate for 3–5 independent repeats.

Electrophysiology recordings of HTM cells in hydrogel with different treatments. In culture dishes, 1 × 10⁶ cells were seeded evenly into 500 μm-thick hydrogel slices of different concentrations, and the cell-scaffold constructs were kept in an incubator for 1 day to allow complete adhesion of the cells to the scaffold. The cell scaffolds were transferred to a recording chamber and continuously perfused with HBSS containing 150 mM NaCl, 5 mM KCl, 1.7 mM CaCl₂·2H₂O, 1 mM NaH₂PO₄, 0.9 mM MgSO₄, 5 mM glucose, 10 mM HEPES (pH 7.4) with NaOH, and maintained at 37 °C with a temperature control system. HTM cells were visualized with a microscope. Borosilicate glass micropipettes were pulled with a horizontal puller to an access resistance of 4–7 MΩ. Single micropipettes were filled with intracellular solutions containing 120 mM KCl, 9.9 mM EGTA, 1 mM NaH₂PO₄, 2.6 mM CaCl₂·2H₂O, 10 mM HEPES, and 0.9 mM MgSO₄·7H₂O (pH 7.2) with NaOH. Leak current and residual capacitative current were subtracted with a P/N protocol where N is 4. Cells were clamped at −60 mV, and depolarizing pulses were applied in 10 mV steps to evoke outward K⁺ currents. Outward potassium currents were blocked with a specific BKCa blocker, iberiotoxin (IbTX, 100 nM).

Calcium transient assay. Ca²⁺ imaging was performed during stimulation by 20 g pressing, with or without piezoelectric hydrogel, taking advantage of Fluo-4 AM Ca²⁺-sensitive fluorescence dye. Before pressure stimulation, HTM cells were stained with Fluo-4 AM (Invitrogen, 1 mM in DMEM/F12 for 30 min at 37 °C), washed twice with PBS, and incubated with HEPES-supplemented (25 mM) phenol red-free DMEM/F12. Fluorescence time-lapse imaging was performed with the Olympus Fluoview FV1000 confocal microscope (Olympus Corporation, Toyko, Japan), and the obtained images were processed using Image J.

Animal models for in vivo studies

Transient ocular hypertension model. All animal experimental procedures and husbandry were approved by the Experimental Animal Welfare Ethics Committee of Harbin Medical University via the protocol #2023042. New Zealand rabbits, 3- to 5-months old and weighing 1.5 to 2.0 kg, were purchased and acclimatized for 1 week prior to the experiments. These rabbits were then randomly divided into four groups, each with 6 rabbits, for the transient IOP elevation model with different post-treatments, i.e., the ones without injection, injected with MXene-RGD (10 mg mL⁻¹, 0.075 mL), injected with MXene-RGD and BKCa blocker, iberiotoxin (IBTX, 1 μM), injected with MXene-RGD and L-type calcium channel blocker, nifedipine, 10 μM, named as Transient Ocular Hypertension Model group, MXene-RGD group, MXene-RGD + IbTX group, and MXene-RGD + nifedipine group. For each group of 6 rabbits, six eyes (the right one of each rabbit) were used for treatment, while the other eyes were used for control. The transient IOP elevation model (OHT model) was obtained by injection of 0.1 mL of sterile hypertonic saline (5%) into the vitreous bilaterally in locally anesthetized rabbits with one drop of 0.2% oxybuprocaine hydrochloride.⁷³ IOP was measured before and after surgery with a Schiötz Tonometer. In this model, ocular hypertension was reached in 10 min and then slowly decreased for 1.5 hours. Compounds were instilled into the lower conjunctival pocket. All compounds were administered prior to saline injection, and IOP was measured at the beginning of the experiment to determine basal IOP, followed by measurements at 10-minute intervals.

Histology. Tissue samples were cleaned with 1× PBS and fixed with 4% paraformaldehyde for 24 hours, then dehydrated and embedded in paraffin. Serial sections (20 μm thick) were cut, dehydrated, and stained with hematoxylin and eosin stain for light microscopy examination. Collagen deposition was assessed with Masson staining. ImageJ software was used for quantitative analysis.

Statistical analysis

All statistical analyses were performed using GraphPad Prism 5. The results were presented as the means and standard deviation. Comparisons between two treatments were made using the Student's t-test (two-tailed, unequal variance), and comparisons between multiple treatments were made using the analysis of variance (ANOVA), where ‘*’, ‘**’, and ‘***’ indicate p values < 0.05, 0.01, and 0.001, respectively.

Conclusions

Aiming at the key factor of IOP homeostasis regulation (trabecular meshwork cells) and the target of rapid response to pressure changes (BKCa), we have constructed MXene nanoparticles (MXene RGD) that can target the trabecular meshwork. They were injected into the anterior chamber with minimal trauma, i.e., anterior chamber injection, and specifically targeted to TM cells. Because of the change in flow velocity of the AH and the magnitude of the trabecular mesh motion deformation under IOP stimulation, the trabecular mesh has the properties of self-enhancement and self-generated electricity after MXene-RGD targeting to trabecular mesh tissues. It is capable of generating appropriate exogenous voltages, thereby altering the TMC membrane potential. It is also able to alter local Ca²⁺ signaling by stimulating voltage-gated Ca²⁺ channels and inducing Ca²⁺ influx, which in turn generates membrane potential hyperpolarization. Through these two pathways, MXene can activate BKCa channels. Furthermore, it can repair or awaken the TM cells' ability to regulate IOP homeostasis, providing a minimally invasive surgical method for targeting TM cells in the treatment of POAG.

Author contributions

Conceptualization: T. W. M., W. R. Q., W. H. Y. Methodology: S. Y. Y.,W. C., W. H. Y.,W. G. F. Investigation: W. R. Q., Z. Y. J., Y. Z. Q., L. Y. F., C. J. W. Visualization: Y. Z. Q. Supervision: W. R. Q., W. C. Writing – original draft: W. R. Q. Writing – review & editing: T. W. M.

Data availability

All relevant data are within the manuscript and its additional files.

Conflicts of interest

The authors declare no competing interests.

Acknowledgements

This work was supported by the China National Key R&D Program (project number 2022YFA1604502, received by T. W. M.) and the Postdoctoral Fellowship Program of CPSF under grant number GZC20242213.

Notes and references

1. R. George, S. Panda and L. Vijaya, Eye, 2022, 36, 2099–2105.
2. C. Garudadri, S. Senthil, R. C. Khanna, K. Sannapaneni and H. B. L. Rao, Ophthalmology, 2010, 117, 1352–1359.
3. W. M. Dismuke, C. C. Mbadugha and D. Z. Ellis, Am. J. Physiol.: Cell Physiol., 2008, 294, C1378–C1386.
4. H. Jayaram, M. Kolko, D. S. Friedman and G. Gazzard, Lancet, 2023, 402, 1788–1801.
5. S. Naik, A. Pandey, S. A. Lewis, B. S. S. Rao and S. Mutalik, Eur. J. Pharmacol., 2020, 881, 173–208.
6. T. S. Acott, J. A. Vranka, K. E. Keller, V. Raghunathan and M. J. Kelley, Prog. Retinal Eye Res., 2021, 82, 100897.
7. L. Rodriguez-Estevez, P. Asokan and T. Borrás, Gene Ther., 2020, 27, 127–142.
8. A. P. Tanna and M. Johnson, Ophthalmology, 2018, 125, 1741–1756.
9. E. Berrino and C. T. Supuran, Expert Opin. Ther. Pat., 2019, 29, 817–827.
10. J. Wu, J. Wei, H. Chen, Y. Dang and F. Lei, Curr. Drug Targets, 2024, 25, 94–107.
11. W. D. Stamer and A. F. Clark, Exp. Eye Res., 2017, 158, 112–123.
12. S. Kumar, S. Shah, H. M. Tang, M. Smith, T. Borrás and J. Danias, PLoS One, 2013, 8, e72447.
13. D. A. Ryskamp, A. M. Frye, T. T. Phuong, O. Yarishkin, A. O. Jo, Y. Xu, M. Lakk, A. Iuso, S. N. Redmon and B. Ambati, Sci. Rep., 2016, 6, 30583.
14. C. H. Mitchell, J. C. Fleischhauer, W. D. Stamer, K. Peterson Yantorno and M. M. Civan, Am. J. Physiol., 2002, 283, 315–326.
15. D. Soto, A. Escalada, M. Morales, X. Gasull, J. Pales, N. Comes, E. Ferrer, C. Solsona and A. Gual, Invest. Ophthalmol. Visual Sci., 2004, 45, 3650–3661.
16. J. A. Bertrand, M. Schicht, W. D. Stamer, D. Baker, J. M. Sherwood, E. Lütjen-Drecoll, D. L. Selwood and D. R. Overby, Invest. Ophthalmol. Visual Sci., 2020, 61, 41.
17. G. Du, L. Zhang, X. Liu, J. Hao, J. Zhu, J. Wu, H. Chen, W. Zhang and Q. You, Exp. Biol. Med., 2020, 245, 180–189.
18. L. Nino-de-Rivera, D. Cervera and P. Castillo-Juarez, Ocular Hypertension-The Knowns and Unknowns, IntechOpen, 2021.
19. X. Gasull, M. Castany, A. Castellanos, M. Rezola, A. Andrés-Bilbé, M. I. Canut, R. Estévez, T. Borrás and N. Comes, Sci. Rep., 2019, 9, 5392.
20. L. Dasso, T. Al-Khaled, S. Sonty and A. A. Aref, Clin. Ophthalmol., 2018, 12, 1939.
21. A. P. Tanna, H. Esfandiari and K. Teramoto, J. Glaucoma, 2020, 29, e41–e43.
22. C. E. Wisely, K. C. Liu, D. Gupta, A. N. Carlson, S. G. Asrani and T. Kim, Am. J. Ophthalmol., 2020, 217, 20–26.
23. H. Hara, N. Toriu and M. Shimazawa, Cardiovasc. Drug Rev., 2004, 22, 199–214.
24. Z. Zhai, Y. Cheng and J. Hong, Acta Biomater., 2021, 125, 41–56.
25. W. M. Dismuke and D. Z. Ellis, J. Ocul. Pharmacol. Ther., 2009, 25, 309–314.
26. X. Wang, J. Song, J. Liu and Z. L. Wang, Science, 2007, 316, 102–105.
27. A. Marino, S. Arai, Y. Hou, E. Sinibaldi, M. Pellegrino, Y.-T. Chang, B. Mazzolai, V. Mattoli, M. Suzuki and G. Ciofani, ACS Nano, 2015, 9, 7678–7689.
28. J. Kubanek, J. Shi, J. Marsh, D. Chen, C. Deng and J. Cui, Sci. Rep., 2016, 6, 24170.
29. F. Liu, A. Zhou, J. Chen, J. Jia, W. Zhou, L. Wang and Q. Hu, Appl. Surf. Sci., 2017, 416, 781–789.
30. Y. Tian, Y. An and B. Xu, Nano Energy, 2022, 101, 107556.
31. Y. Zhang, J. Zou, S. Wang, X. Hu, Z. Liu, P. Feng, X. Jing and Y. Liu, Composites, Part B, 2024, 272, 111191.
32. S. Lai, J. Jeon, S. K. Jang, J. Xu, Y. J. Choi, J.-H. Park, E. Hwang and S. Lee, Nanoscale, 2015, 7, 19390–19396.
33. Y. Zhou, W. Deng, J. Xu and J. Chen, Cell Rep. Phys. Sci., 2020, 1, 100142.
34. J. Zou, X. Jing, Z. Chen, S. J. Wang, X. S. Hu, P. Y. Feng and Y. J. Liu, Adv. Funct. Mater., 2023, 33, 2213895.
35. D. Tan, C. Jiang, N. Sun, J. Huang, Z. Zhang, Q. Zhang, J. Bu, S. Bi, Q. Guo and J. Song, Nano Energy, 2021, 90, 106528.
36. L. Yue, L. Chen, X. Wang, D. Lu, W. Zhou, D. Shen, Q. Yang, S. Xiao and Y. Li, Chem. Eng. J., 2023, 451, 138687.
37. D. Gagen, M. S. Filla, R. Clark, P. Liton and D. M. Peters, Invest Ophthalmol. Visual Sci., 2013, 54, 5000–5011.
38. B. Adhikari, B. S. Stinson, M. J. Osmond, M. B. Pantcheva and M. D. Krebs, Ind. Eng. Chem. Res., 2021, 60, 17417–17428.
39. D. Cawkill and S. S. Eaglestone, Drug Discovery Today, 2007, 12, 820–825.
40. J. Lee, M. J. Cuddihy and N. A. Kotov, Tissue Eng., Part B, 2008, 14, 61–86.
41. R. Wang, S. Deng, Y. Wu, H. Wei, G. Jing, B. Zhang, F. Liu, H. Tian, X. Chen and W. Tian, Colloid Interface Sci. Commun., 2022, 49, 100632.
42. M.-S. Guo, Y.-Y. Wu and Z.-B. Liang, Mol. Vision, 2012, 18, 1175.
43. M. Malaki and R. S. Varma, Adv. Mater., 2020, 32, 2003154.
44. J. Cai, H. Zhang, Y. Hu, Z. Huang, Y. Wang, Y. Xia, X. Chen, J. Guo, H. Cheng and L. Xia, J. Nanobiotechnol., 2022, 20, 460.
45. B. Anasori, M. R. Lukatskaya and Y. Gogotsi, Nat. Rev. Mater., 2017, 2, 1–17.
46. X. Luo, L. Zhu, Y. C. Wang, J. Li, J. Nie and Z. L. Wang, Adv. Funct. Mater., 2021, 31, 2104928.
47. J. Peng, M. Zhou, Y. Gao, J. Wang, Y. Cao, W. Wang, D. Wu and Y. Yang, J. Mater. Chem. A, 2021, 9, 25073–25085.
48. M. Pohanka, Int. J. Electrochem. Sci., 2017, 12, 496–506.
49. C. Chen, S. Zhao, C. Pan, Y. Zi, F. Wang, C. Yang and Z. L. Wang, Nat. Commun., 2022, 13, 1391.
50. X. Jin, J. Wang, L. Dai, X. Liu, L. Li, Y. Yang, Y. Cao, W. Wang, H. Wu and S. Guo, Chem. Eng. J., 2020, 380, 122475.
51. H. Zhang, J. Yu, X. Yang, G. Gao, S. Qin, J. Sun, M. Ding, C. Jia, Q. Sun and Z. L. Wang, ACS Nano, 2020, 14, 3461–3468.
52. M. M. Rastegardoost, O. A. Tafreshi, Z. Saadatnia, S. Ghaffari-Mosanenzadeh, C. B. Park and H. E. Naguib, Nano Energy, 2023, 111, 108365.
53. Y. Song, H. Wu, X. He, C. Fang, Q. Song, M. Chen, Z. Liu, Y. Lu, B. Yu and T. Liu, ACS Appl. Mater. Interfaces, 2024, 16, 50630–50639.
54. J. Jiang, Q. Guan, Y. Liu, X. Sun and Z. Wen, Adv. Funct. Mater., 2021, 31, 2105380.
55. G. G. Genchi, L. Ceseracciu, A. Marino, M. Labardi, S. Marras, F. Pignatelli, L. Bruschini, V. Mattoli and G. Ciofani, Adv. Healthcare Mater., 2016, 5, 1808–1820.
56. A. Marino, G. G. Genchi, V. Mattoli and G. Ciofani, Nano Today, 2017, 14, 9–12.
57. C. Li, C. Xiao, L. Zhan, Z. Zhang, J. Xing, J. Zhai, Z. Zhou, G. Tan, J. Piao, Y. Zhou, S. Qi, Z. Wang, P. Yu and C. Ning, Bioact. Mater., 2022, 18, 399–408.
58. P. Agarwal, A. M. Daher and R. Agarwal, Mol. Vision, 2015, 21, 612.
59. A. Izzotti, S. C. Sacca, M. Longobardi and C. Cartiglia, Invest Ophthalmol. Visual Sci., 2009, 50, 5251–5258.
60. P. Sahay, A. Rao, D. Padhy, S. Sarangi, G. Das, M. M. Reddy and R. Modak, Invest Ophthalmol. Visual Sci., 2017, 58, BIO106–BIO113.
61. S. C. Saccà, S. Tirendi, S. Scarfì, M. Passalacqua, F. Oddone, C. E. Traverso, S. Vernazza and A. M. Bassi, ALTEX, 2020, 37, 265–274.
62. A. Poehlmann, K. Reissig, P. Schönfeld, D. Walluscheck, A. Schinlauer, R. Hartig, W. Lessel, T. Guenther, A. Silver and A. Roessner, J. Cell. Mol. Med., 2013, 17, 1619–1631.
63. Y. Shen, Y. Zhu and F. Rong, Mol. Med. Rep., 2020, 22, 1605–1612.
64. J. E. Brayden and M. T. Nelson, Science, 1992, 256, 532–535.
65. Y. Kong, F. Liu, B. Ma, J. Duan, W. Yuan, Y. Sang, L. Han, S. Wang and H. Liu, Adv. Sci., 2021, 8, 2100962.
66. B. C. Heng, Y. Bai, X. Li, Y. Meng, X. Zhang and X. Deng, Smart Mater. Med., 2022, 3, 4–11.
67. O. Yarishkin, T. T. Phuong, F. Vazquez-Chona, J. Bertrand, J. van Battenburg-Sherwood, S. N. Redmon, C. N. Rudzitis, M. Lakk, J. M. Baumann and M. Freichel, Front. Immunol., 2022, 13, 805076.
68. O. Yarishkin, M. Lakk, C. N. Rudzitis, J. E. Searle, D. Kirdajova and D. Križaj, Vision Res., 2024, 224, 108487.
69. K. Steinhausen, F. Stumpff, O. Strau, H. Thieme and M. Wiederholt, Exp. Eye Res., 2000, 70, 285–293.
70. C. Xin, R. K. Wang, S. Song, T. Shen, J. Wen, E. Martin, Y. Jiang, S. Padilla and M. Johnstone, Exp. Eye Res., 2017, 158, 171–186.
71. K. Gao, S. Song, M. A. Johnstone, Q. Zhang, J. Xu, X. Zhang, R. K. Wang and J. C. Wen, Invest Ophthalmol. Visual Sci., 2020, 61, 21.
72. F. Jiang, H. Wei and Y. Lu, Chin. J. Ophthalmol., 1997, 33, 410–412.
73. C. Lanzi, L. Lucarini, M. Durante, S. Sgambellone, A. Pini, S. Catarinicchia, D. Łażewska, K. Kieć-Kononowicz, H. Stark and E. Masini, Int. J. Mol. Sci., 2019, 20, 981.

---

Footnotes

† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4mh01004c

‡ Ruiqi Wang and Haiying Wei contributed equally to this work.

---