Iron-Responsive Deferasirox Release from Covalently Grafted Hydrogels on Microwires Extends Iron Chelation Time Scales

Abstract

Implantation of intracortical microelectrodes generates free iron species that exacerbate pro-inflammatory responses by producing reactive oxygen species through the Fenton reaction. This work reports a covalently grafted, two-layer poly(ethylene glycol)-dimethacrylate (PEGDMA) hydrogel on polyimide-insulated Pt/Ir microwires for local delivery of the iron chelator deferasirox (DFO). The polyimide insulating layer was methacrylated to enable covalent anchoring of a PEGDMA550 interlayer, followed by a DFO-loaded PEGDMA3400 outer layer. Scanning electron microscopy and optical microscopy of the hydrogel-coated microwires revealed a uniform coating morphology with dispersed DFO aggregates embedded within the matrix. In buffer at 37 °C, DFO release followed diffusion-controlled kinetics and fell below the limit of detection by ultraperformance liquid chromatography by day 9. The iron-chelation properties of the DFO-loaded microwires were investigated using Fe(EDTA) as a free-iron mimic. When a high concentration of Fe(EDTA) was used (10 mM), the Fe(DFO)₂ complex remained detectable for up to 19 days, which expands well beyond the detection limit of DFO under iron-free conditions (9 days). Energy-dispersive X-ray spectroscopy showed time-dependent iron accumulation within the hydrogel, consistent with in-matrix complexation that perturbs local transport and prolongs chelation efficacy. Collectively, the coating functions as a mechanically integrated, responsive reservoir whose pharmacokinetics are governed by the iron concentration rather than initial drug loading. By coupling chelation capacity to local iron availability, the presented hydrogel coating establishes a generalizable route to extend therapeutic lifetimes of functionalized chronic neural electrodes.

Graphical Abstract

1. INTRODUCTION

Implantable microelectrodes provide a direct interface with the central nervous system, enabling precise electrical communication with neuron networks. Such interfaces can restore motor function through functional electrical stimulation of a patient’s own limbs, facilitate control of robotic prostheses in brain–computer interface (BCI) applications, and deliver targeted electrical stimulation for the treatment of neurological disorders.^1–4 However, the foreign body response (FBR) to implanted microelectrodes compromises their long-term reliability and stability, leading to progressive degradation of electrophysiological signal quality and shifts in stimulation thresholds. Insertion of a microelectrode causes acute tissue damage characterized by vascular rupture, hemorrhage, and activation of neuroinflammatory pathways. During chronic implantation, persistent mechanical irritation perpetuates this injury and leads to recurring exacerbation of secondary inflammatory responses. As a consequence, the FBR at the microelectrode–tissue interface induces persistent neuroinflammation,^5,6 promotes neuronal loss due to the resulting hostile microenvironment,⁷ and leads to glial encapsulation of the implanted devices.⁸ This glial encapsulation increases tissue impedance and progressively degrades signal quality,^9,10 establishing the foreign body response as a central barrier to long-term BCI performance by driving chronic inflammation, neuronal loss, and ultimately leading to device failure.

By disrupting the cerebral vasculature, implantation of microelectrodes permits erythrocytes to enter the brain parenchyma, where they undergo hemolysis, releasing hemoglobin that is subsequently degraded into heme and free iron.¹¹ The liberated ferrous ions can react with hydrogen peroxide via the Fenton reaction to generate reactive oxygen species (ROS).¹² These ROS not only amplify neuroinflammation by further activating microglia and other immune cells but also contribute to corrosion of the implanted electrodes.¹³ Previous in vivo studies from the Prasad group have demonstrated that free iron plays a significant role in oxidative stress following intracortical microelectrode implantation and that iron chelation can mitigate these effects.^11,14 In these studies, animals treated with iron chelators exhibited reduced oxidative stress and tissue damage at the micro-electrode–tissue interface, along with improved functional performance over a 4-week implantation period.¹⁴

Efforts to extend the functional lifespan of intracortical neural implants have focused on pharmacological strategies that reduce the FBR through either systemic or local drug delivery.¹⁵ In systemic delivery, drugs are administered orally, intravenously, or subcutaneously, after which they are absorbed and distributed throughout the body via the bloodstream.¹⁶ A major limitation of systemic delivery for brain targets is the blood–brain barrier (BBB), as systemically administered drugs must cross it to reach the implant site.¹⁷ However, the BBB is primarily permeable to low-dielectric, nonpolar molecules, which tend to have poor solubility in blood, thereby inherently restricting the amount of drug that can be delivered to the target tissue.^18,19 Another limitation of systemic delivery is the nonspecific distribution of the drug throughout the body, rather than its localization at the implant site. For example, in patients with thalassemia intermedia, the iron chelator deferoxamine has been administered subcutaneously to manage disease symptoms.²⁰ Clinical studies have shown that this treatment leads to a significant reduction in serum ferritin levels and a marked increase in iron excretion.²¹

Local drug delivery offers a means to bypass the BBB and administer therapeutics directly to the implant site. This approach enables precise control over drug dosage at the target tissue, consequently reducing the risk of systemic side effects associated with overadministration. Functionalization of microelectrodes with soft polymer scaffolds is an appealing strategy for local drug delivery and has shown promise in mitigating the FBR after implantation. For example, conductive polymer coatings on cochlear implants have been successfully used to target the delivery of neurotrophic factors.²² The Cui group exploited the internal cavity of carbon nanotubes (CNTs) to increase drug-loading capacity by incorporating multiwalled CNTs within a polypyrrole (PPy) matrix, creating nano-reservoirs for the encapsulation and controlled release of dexamethasone (Dex).²³ Dex released from these PPy/CNT-coated electrodes effectively attenuated lipopolysaccharide-induced microglial activation, demonstrating strong potential for promoting stable long-term neural interfaces.²⁴ We recently investigated a localized drug-delivery strategy in which a poly(ethylene glycol)-dimethacrylate (PEGDMA)-based hydrogel was applied around Pt/Ir microelectrode arrays and loaded with the anti-inflammatory corticosteroid Dex to mitigate implantation-induced neuroinflammation.²⁵ In vivo studies demonstrated that implantation of these Dex-loaded hydrogel–coated electrodes resulted in reduced expression of proinflammatory genes (IL1a, IL1b, IL6, and TNFα) at 48 h and 7 days postimplantation, compared to unmodified polyimide-coated Pt/Ir microelectrode arrays.

Polymeric hydrogels have recently gained attention as promising platforms for localized, sustained delivery of iron-chelating agents to treat conditions such as oxidative stress and iron overload.^26,27 Their biocompatibility and tunable release kinetics²⁸ make them ideal candidates for iron chelating agents, which often suffer from poor pharmacokinetics or off-target toxicity when delivered systemically.^29–31 Various strategies have been employed to deliver iron chelators, including the encapsulation of deferoxamine in chitosan-alginate hydrogels,³² PEG-based hydrogels,³³ and composite systems that embed drug-loaded nanoparticles.^34,35 Star-shaped PEG-deferoxamine conjugates and thermosensitive HA-F127 hydrogels have shown particularly favorable release profiles and safety in vivo.³³ In a recent example, a ROS-responsive hydrogel released deferoxamine in response to oxidative triggers, mitigating tissue damage and improving behavioral outcomes in a traumatic brain injury model.³⁶ Despite these advantages, iron-chelating hydrogels have not been applied to microwire neural electrode arrays, presenting a clear opportunity to reduce iron-mediated oxidative stress and enhance implant longevity. More broadly, nanoarchitectonics offers a powerful strategy for engineering hydrogels with tunable structure–function relationships by controlling organization across molecular, nanoscale, and mesoscale dimensions.^37–39 Through targeted nanoarchitectural design, hydrogels can be engineered to exhibit adaptive mechanical, chemical, and biological responses, making them particularly well suited for biointerfaces, soft electronics, and dynamic tissue-mimetic materials.^40–42 In this context, we present the development of a hydrogel-functionalized microelectrode platform designed for the release of the iron chelator deferasirox (DFO). Through a two-stage deposition process, we demonstrate the reliable fabrication of a thin, homogeneous DFO-loaded hydrogel on the insulating polyimide (PI) layer of Pt–Ir wires, forming a structurally robust interface with the underlying insulation. The loading capacity of DFO in the hydrogel can be tuned through modifying the drug concentration of the deposition solution, allowing for a versatile and reliable method to tune the amount of drug delivered. DFO-loaded hydrogel on microwires exhibits appreciable release over 10 days when incubated at 37 °C in HEPES buffer. Using Fe(EDTA) as a free iron mimic, we demonstrate the chelation properties of the DFO-loaded hydrogel microwires over 20 days. Increasing the concentration of the free iron mimic in solution from 1 mM to 10 mM significantly prolongs the detection of the chelated Fe(DFO)₂ complex to 20 days. This duration extends well beyond the complete depletion of DFO observed in hydrogelfunctionalized microwires under iron-free buffer conditions, which occurs by day 10. We attribute these findings to changes in the chemical microenvironment within the hydrogel matrix that can slow down the release of the embedded DFO aggregate.

2. MATERIALS AND METHODS

2.1. Materials

All chemicals were used as received unless otherwise specified. All air-sensitive reagents were handled in a Braun LABmaster pro glovebox under Argon atmosphere. Commercially available polyimide coated Pt/Ir microwires (70% Pt – 30% Ir) from California Fine Wire Company were used for all experiments. Unfunctionalized wires used in this study are denoted as “Pristine”. HPLC-grade solvents were utilized for all experiments unless otherwise specified. HPLC vial microinserts for drug release studies were purchased from Thermo Fisher Scientific. All solvents and chemicals used for synthesis or analysis were purchased from VWR unless otherwise specified. PEGDMA M.W. 550 was purchased from Sigma-Aldrich, and PEGDMA M.W. 3400 was purchased from Thermo Fisher Scientific. Deferasirox was purchased from AmBeed.

2.2. Instrumentation

HPLC analysis was performed using a Waters Acquity UPLC/UV/QDA-MS instrument using a ACQUITY UPLC BEH C18 Column, 130Å, 1.7 μm, 2.1 mm × 50 mm column utilizing a gradient of 60% H2O 40% MeOH to 40% H2O 60% MeOH. Optical microscopy images were obtained using a Zeiss TE 2000 Widefield microscope. Scanning electron microscopy and energy dispersive X-ray spectroscopy was performed using a FEI HELIOS NanoLab 650 SEM/FIB equipped with an Oxford EDXS system. UV/vis analysis was performed using a Cary 5000 UV/vis/NIR Spectrometer. Fourier transform infrared spectroscopy in ATR mode was performed using a Nicolet iS50 FT-IR spectrometer equipped with an iS50 ATR. Nuclear magnetic resonance spectroscopy was performed using a Bruker Avance III 500 MHz NMR spectrometer.

2.3. Synthesis of Covalently Functionalized Microwires

To facilitate functionalization of the polyimide coated microwires, pristine microwires were threaded through a rubber septum and mounted in a glass reactor. For each step, four 7 cm wires are threaded through a rubber septum and fastened onto the glass vial. After the reaction is completed, the septum is removed from the vial and the suspended wires washed with the reaction solvent. Wires are dried and stored under vacuum in a desiccator prior to proceeding to the next step. Methodology adapted from our previous work.²⁵

Step 1: Surface Hydrolysis.

To a vial loaded with four 7 cm wires, 40 mL of 2 M KOH solution is added and stirred for 1 h. The wires are then removed from the solution and washed with a continuous flow of 500 mL of DI water for 5 min before drying under vacuum in a desiccator.

Step 2: Esterification.

To a vial loaded with four 7 cm wires, 100 mg of EDC-HCl is added and dissolved in 35 mL of acetone. Then, 1.5 mL of DIPEA is added and the solution is stirred. After 15 min, 3.6 mL of TEG is added, and the solution is further stirred for 4 days. The wires are then removed and washed with 200 mL of acetone and 500 mL of water and then dried under vacuum in a desiccator.

Step 3: Methacrylation.

To a vial loaded with four 7 cm wires, 4 mL of methacrylic anhydride is added, followed by 34.5 mL of acetone and stirred. Then, 1.5 mL of DIPEA is added and the solution is kept out of light and stirred for 24 h. The wires are then removed and washed with 200 mL of acetone and 500 mL of water and then dried under vacuum in a desiccator for at least 24 h prior to hydrogel coating.

2.4. Synthesis of Drug-Loaded Microwires

Drug-Loaded Microwires Are Synthesized through a Two-Step Process.

To prepare the wires for coating, methacrylated microwires are mounted under tension in a parallel arrangement onto a small wooden circular support structure, which acts as a mechanical frame to ensure the wires remain taut throughout the subsequent hydrogel-coating processes. A primer coat of PEGDMA550 is applied to modulate the hydrophobicity of the surface as described in Section 3.2 of the main text, allowing for more homogeneous interfacing with the second, drug-containing coating consisting of PEGDMA 3400 and DFO.

Primer Coat.

To a 40 mL vial, 200 μL of PEGDMA550 is mixed with 230 μg of 2-hydroxy-4′-(2-hydroxyethoxy)-2-methylpropiophenone (Irgacure 2959) delivered in 10 μL of ethanol and sonicated for 1 min. The mixture is uniformly applied to the mounted wires as a thin, homogeneous layer by performing two slow, sweeping strokes with a fine-tip brush under constant pressure and digital microscope. The wires are irradiated in a Rayonet photoreactor operating at 350 nm for 6 h with a distance of ~3 cm. The wires are then removed from the photoreactor and placed under vacuum in a desiccator overnight to remove volatiles trapped in the deposited polymer coatings. This constitutes the primer precoat.

Drug-Loaded Coat.

To a vial, 225 mg of PEGDMA3400 is added and dissolved in 375 μL of ethanol, assisted by sonication. Then, 230 μg of Irgacure 2959 (0.1% w/w based on polymer mass) is delivered in 10 μL of ethanol and dissolved. Subsequently, 81 mg (18.9 wt % by total solvent mass) or 162 mg (37.8 wt % by total solvent mass) of DFO is added to the vial, followed by 125 μL of HPLC water. The mixture is sonicated until a homogeneous dispersion of DFO is achieved and then applied to the mounted wires using the same procedure as for the primer coat. The wires are placed under vacuum in a desiccator overnight to remove residual volatiles. Then, the wires are irradiated at 350 nm for 8 h with a distance of ~3 cm. The wires are then removed and stored under vacuum in a desiccator. This constitutes the drug-loaded hydrogel coat.

Using this approach, batches of 7–8 wires at a time could be coated on the same hoop, providing around 50 cm of final drug-coated microwires per batch treatment.

2.5. Swelling Ratio Determination

Swollen hydrogel-coated microwire platforms were mounted on a glass slide using tape and imaged by optical microscopy after 15 min of incubation with 1 mL of 2 mM HEPES buffer at pH 7.4 and 37 °C. Swelling ratio was determined from side-view images of one wire region before and after swelling using ImageJ’s area tool.⁴³ For each condition, the coating area was manually outlined, and the calculated area of the underlying polyimide coated wire was subtracted to isolate the coating contribution. The area-based swelling ratio (eq 1) was calculated as

$$Q_{\text{area}}=\frac{A_{\text{swollen}}-A_{\text{wire}}}{A_{\text{dry}}-A_{\text{wire}}}$$ (1)

To estimate the volumetric swelling (assuming isotropic swelling of the hydrogel in all three dimensions, eq 2), the following relationship was applied:

$$Q=Q_{\mathit{\text{area}}}^{3/2}$$ (2)

The volume swelling ratio was used to calculate the polymer volume fraction in the swollen state (eq 3):

$$v_{2,s}=\frac{1}{Q}$$ (3)

This value represents the fraction of swollen hydrogel volume occupied by the polymer network and was used to estimate the mesh size for different drug-loading conditions.^44,45

2.6. Standard Drug Release Study

Drug-loaded microwires consisting of either 18.9 wt % DFO or 37.8 wt % DFO conditions were prepared for the drug-release study. Drug-loaded microwires were segmented into 20 1 cm pieces and arranged into a bundle. Nine HPLC vials containing 150 μL microinserts were prepared and labeled by day of drug release. To begin the drug release study, the microinsert is filled with 100 μL of 2 mM HEPES buffer at pH 7.4 in the vial corresponding to day 1 of release. The bundle is then slowly lowered into the microinsert until completely submerged in liquid, and the vial is capped and placed in an incubator at 37 °C for 24 h. After 24 h, the bundle is removed from the day 1 microinsert and placed into the day 2 microinsert containing 100 μL of fresh buffer and again incubated at 37 °C for 24 h. The day 1 vial is then sealed and stored at 4 °C. This process was repeated for up to 9 days of release and was performed in triplicate for each condition.

2.7. Iron Chelation Drug Release Study

Drug-loaded microwires consisting of either 18.9 wt % DFO or 37.8 wt % DFO conditions were prepared for the iron chelation drug release study. Drug-loaded microwires were segmented into 20 1 cm pieces and arranged into a bundle. Nineteen HPLC vials containing 150 μL microinserts were prepared and labeled by day of drug release or iron condition. Microwire bundles of either 18.9 and 37.8 wt % drug were incubated at 37 °C with either 1 or 10 mM Fe(EDTA) in 2 mM HEPES buffer pH 7.4. To begin the iron chelation drug release study, the microinsert is filled with 100 μL of 2 mM HEPES buffer at pH 7.4 in the vial corresponding to day 1 of release. The bundle is then slowly lowered into the microinsert until completely submerged in liquid, and the vial is capped and placed in an incubator at 37 °C for 24 h. After 24 h, the bundle is removed from the day 1 microinsert and placed into the day 2 microinsert containing 100 μL of fresh buffer and again incubated at 37 °C for 24 h. The day 1 vial is then sealed and stored at 4 °C until analysis. This process was repeated for up to 19 days of release and was performed in triplicate for each condition (18.9 wt % DFO with 1 mM Fe(EDTA), 18.9 wt % with 10 mM Fe(EDTA), 37.8 wt % with 1 mM Fe(EDTA), and 37.8 wt % with 10 mM Fe(EDTA)).

3. RESULTS AND DISCUSSION

3.1. Synthesis of Functionalized Microwires

To enable robust covalent attachment between the hydrogel host matrix and the PI-coated microwire surfaces, we developed a surface modification protocol for PI coatings using mild solvents and conditions to install reactive tethers (Figure 1A). Prior methodologies relied on harsher organic solvents, such as formic acid, chloroform, and dimethylformamide, which can compromise polyimide coatings.²⁵ In contrast, our approach employs milder solvents that preserve polyimide integrity, allowing effective surface functionalization while minimizing damage.

Figure 1.

(A) Surface functionalization scheme for Pt/Ir microwires coated with polyimide (PI). Step 1: Base hydrolysis of PI coating with KOH introduces carboxylic acid groups. Step 2: Coupling of triethylene glycol (TEG) using EDC provides hydroxylated PEG-like chains for increased hydrophilicity and further functionalization. Step 3: Methacrylate groups are installed via esterification with methacrylic anhydride (MA), enabling subsequent hydrogel grafting. (B) Hydrogel formation and drug loading of functionalized wires. A PEGDMA550 precoat is photografted onto the methacrylated PI surface, followed by a second layer of PEGDMA3400 loaded with the iron chelator DFO. (C) Schematic of iron-responsive release kinetics. Drug-loaded microwires maintain a release profile that slows upon exposure to increasing concentrations of iron chelates, attributed to the formation of Fe(DFO)₂ complexes that retain cargo in the hydrogel and slow passive diffusion over time.

The surface is first treated with 2 M KOH for 1 h, promoting base-catalyzed hydrolysis of imide groups and generating surface-bound carboxylic acids (Figure 1A, Step 1). Microwire surfaces are then coupled with triethylene glycol moieties via EDC-mediated esterification (Figure 1A, Step 2), providing hydrophilicity to the surface and enabling subsequent functionalization of the surface with methacrylate through reaction with methacrylic anhydride (Figure 1A, Step 3). This strategy not only covalently modifies the wire’s surface chemistry but also enables the introduction of polymerizable groups that can nucleate hydrogel formation via photo-polymerization. These covalent tethers are essential for a mechanically robust interface, as hydrogel coatings that delaminate can shear off during insertion and subsequently fail under brain micromotion, compromising drug delivery and recording performance.^46–48

Functionalized microwires were evaluated for PI coating degradation at each step via optical microscopy analysis of at least three 1 cm representative segments of microwire (Figures S1–S4) and scanning electron microscopy (SEM) (Figures S13–S16). Degradation, defined as cracking, swelling, and delamination, was infrequent, and batches exhibiting any such damage were discarded. Furthermore, ATR-FTIR spectra were collected from five 1 cm microwire segments after each functionalization step to confirm the successful installation of targeted functional groups (Figures S62–S68).

3.2. Synthesis of Drug-Loaded Hydrogel Microwire Platforms

To enable local delivery of the iron chelator DFO, we engineered a two-layer hydrogel system on methacrylated microwires using photopolymerizable PEGDMA-based formulations. Uniform coating without delamination or beading is critical for mechanical stability under implantation; however, early formulations utilizing only PEGDMA 3400 and DFO exhibited erratic beading behavior (Figures S8 and S9), consistent with an interfacial mismatch between the hydrophilic hydrogel precursor solution and the more hydrophobic polyimide insulation layer.

To address this issue, we employed a two-layer photo-polymerization approach. Our strategy relies on an intermediate polymer junction, which offers compatibility with both the hydrophobic insulator and the hydrophilic hydrogel matrix. First, a thin interfacial layer composed of 200 μL of PEGDMA550 with 223 μg Irgacure 2959 (~0.1 wt % relative to polymer mass) was deposited and polymerized under 350 nm light for 6 h to promote adhesion between the insulating layer and the subsequent drug-containing PEGDMA3400 outer layer. This lower molecular weight PEGDMA forms a cross-linked, hydrophobic network that bridges the functionalized polyimide’s inherent surface hydrophobicity, thereby reducing beading.

After the initial polymerization, a second layer was then applied by depositing a precursor solution of 225 mg PEGDMA3400, 223 μg Irgacure 2959 (0.1 wt % relative to polymer mass), and either 81 mg (18.9 wt % by total solvent mass) or 162 mg (37.8 wt % by total solvent mass) DFO. The photoinitiator was employed at 0.1 wt % relative to PEGDMA to minimize cytotoxicity while ensuring a reproducible gelation across batches.^49–51 PEGDMA3400 was chosen for outer coating due to its relatively high hydrophilicity and low elastic modulus (25 kPa), which is within an order of magnitude of cortical tissue (0.1–1 kPa), providing mechanical compliance suitable for soft tissue engineering.²⁵

Control of methacrylation and hydrogel attachment was visually confirmed through optical microscopy imaging (Figure 2A,B), which demonstrate uniform coverage and consistent coating morphology across the wire surface before and after swelling in buffer. SEM images at high accelerating voltages revealed the presence of DFO rod-like aggregates residing within the hydrogel matrix, which are not present in hydrogel-coated microwires, excluding DFO (Figure 2C). Lyophilized SEM imaging of drug-loaded microwires provided further qualitative validation of robust hydrogel coverage and morphological uniformity. Since freeze-drying imposes large interfacial tension from the rapid evaporation of solvent and can contribute to cracking and delamination in weakly adhered hydrogels, retention of an intact coating further supports the mechanical stability of the multilayered hydrogel microwires (Figure 2D).^52,53 This two-layer approach improves coating uniformity and mechanical integration by incorporating a more hydrophobic base layer (PEGDMA550) to anchor the more hydrophilic, DFO-loaded PEGDMA3400 outer layer.

Figure 2.

(A) Optical microscopy images of methacrylated microwires before covalent hydrogel attachment. (B) Optical microscopy images of (i) 18.9 wt % DFO loaded microwires in the dry state, (ii) 37.8 wt % DFO loaded microwires in the dry state, (iii) 18.9 wt % DFO loaded microwires after 15 min swelling in 2 mM HEPES buffer at pH 7.4 and 37 °C, and (iv) 37.8 wt % DFO loaded microwires after 15 min swelling in 2 mM HEPES buffer at pH 7.4 and 37 °C. (C) SEM images at 30 keV accelerating voltage of 37.8 wt % DFO loaded microwires with zoom-in of DFO aggregates. (D) SEM images at 15 keV accelerating voltage of 18.9 wt % drug loaded microwires (left) and 37.8 wt % drug loaded microwires (right) after lyophilization following 24 h of incubation in 2 mM HEPES buffer at pH 7.4 and 37 °C.

3.3. Macroscopic Microwire Platform Properties

The functionalized microwire platforms were characterized using SEM and optical microscopy to evaluate the dry-state morphology and hydrogel swelling behavior. SEM analysis of pristine Pt/Ir microwires coated with unmodified polyimide revealed a consistent average width of 59.0 μm (STD = 1.7 μm). Following methacrylation of polyimide surface, the average width increased negligibly to 60.2 μm (STD = 1.9 μm), consistent with the deposition of small molecule linkers and preserved structural integrity of the insulating layer (Figure S22).

Deposition of the PEGDMA550 interfacial hydrogel layer produced a uniform, conformal coating with an average dry width of 63.3 μm (STD = 2.4 μm), as observed by SEM, indicating excellent and reproducible surface coverage and minimal surface heterogeneity. Application of the outer PEGDMA3400 hydrogel layer containing either 18.9 or 37.8 wt % DFO further increased the average dry width to 79.2 μm (STD = 9.8 μm) and 92.3 μm (STD = 19.6 μm), respectively (Figure S23), reflecting progressive increases in hydrogel volume with drug content.

To characterize the hydrogel morphology in the hydrated state, optical microscopy was employed after 15 min of immersion in 2 mM HEPES buffer at pH 7.4. Using eqs 1–3, an area-based swelling ratio of ~4.82× was calculated for the 18.9 wt % microwires and ~3.38× for the 37.8 wt % microwires using area-based image analysis by comparing identical dry and hydrated regions of hydrogel-coated microwires (Figure S12). Additional images of dried and swollen hydrogels for 18.9 and 37.8 wt % DFO-loaded hydrogels are included in the Supporting Information (Figures S10 and S11) and confirm the reproducibility of the engineered materials across multiple batches. Mesh size (ξ) was estimated from swelling ratios using the Peppas–Merrill adaptation of Flory–Rehner theory, a widely adopted model for PEG-based hydrogels (eq 4).^44,45,54

$$\xi =v_{2,\text{s}}^{-1/3}{\left(\frac{2{\text{C}}_{\text{n}}{\overline{M}}_{\text{c}}}{M_{\text{r}}}\right)}^{1/2}l$$ (4)

Assuming a PEG backbone with a Flory characteristic ratio^55–57 of C_n = 4, molecular weight of repeating unit of M_{r =} 44 g/mol, polymer bond lengths (using the average between C–O and C–C bond lengths) of l = ~1.5 Å, and molecular weight-average between cross-linking sites of ${\overline{M}}_{\text{c}}=3400\text{g}/\text{mol}$, the estimated mesh sizes were ~8.19 nm for the 18.9 wt % drug loaded microwires, and ~6.85 nm for the 37.8 wt % drug loaded microwires. The smaller mesh size observed for the 37.8 wt % formulation is consistent with prior PEG-based networks in which increasing hydrophobic content suppresses equilibrium swelling. For example, PEG–PCL-diacrylate hydrogels exhibit lower water uptake and reduced swelling as PCL (hydrophobic) fraction increases.^58,59 Doubling the hydrophobic drug content increases the effective solids fraction and reduces water uptake, lowering the apparent mesh size.

3.4. DFO Release Profile of Hydrogel-Coated Microwires

Drug release profiles over time were established using microwires coated with either 18.9 wt % DFO or 37.8 wt % DFO hydrogel formulations. For each condition, a bundle of 20 1 cm microwires was incubated in 100 μL of 2 mM HEPES buffer (pH 7.4) at 37 °C, using 150 μL glass HPLC vial microinserts to ensure full solvent contact of the microwire bundle and to mimic physiological conditions. At 24 h intervals, microwire bundles were removed and transferred to fresh buffer, while the remaining supernatant was collected for analysis. After 9 days, release aliquots were evaporated at 60 °C for 24 h, and then washed five times with 105 μL of 4.8% DMSO in ethanol solution to solubilize DFO due to its inherent low solubility in aqueous solution (K_sp = 1 × 10⁻⁵ mol/dm³ at 25 °C).⁶⁰ This methodology ensured rigorous and accurate quantification of both soluble and aggregated DFO in the supernatant.

DFO concentrations were quantified by UPLC-MS over a 9-day period (Figure 3). After day 9, the amount of released DFO from the hydrogel functionalized wire was not detectable for both conditions, and the experiment was terminated. All release studies were performed in triplicate to ensure reproducibility. For the 18.9 wt % DFO formulation, release followed an exponential decay model consistent with Fickian diffusion. Curve fitting yielded C₀ = 204 μM/cm and k = 0.562 day⁻¹, confirming that drug release is governed by passive release through a fixed matrix. The 37.8 wt % formulation displayed a comparable release rate constant (k = 0.487 day⁻¹), yet released approximately 6.5 times more drug, despite only a 2-fold increase in the initial drug concentration in the deposition solution. The fitting information can be found in Table S3. The amount of drug incorporated in either coating was determined by cumulative release over 9 days, and values were normalized to 1 cm of coated microwire, providing a loading capacity of ~9.4 μg/cm for 18.9 wt % drug loaded microwires, and ~60.6 μg/cm for 37.8 wt % drug loaded microwires. This nonlinear scaling suggests partial saturation of the hydrogel’s solubilization capacity, with additional DFO potentially present as microphase-separated aggregates. Similar values of the rate constant k across both formulations further support diffusion-limited, first-order release kinetics. Once incorporated into the hydrogel, drug release is driven primarily by diffusion mechanisms, where the flux is dictated by the local concentration gradient and the hydrogel’s mesh size, cross-linking density, and swelling behavior, rather than by the total mass of drug loaded.

Figure 3.

(A) Individual trials for drug release studies performed on 18.9 wt % drug loaded microwires over 9 days. (B) Individual trials for drug release studies performed on 37.8 wt % drug loaded microwires over 9 days. (C) Average drug release (n = 3) profile for 18.9 wt % drug-loaded microwires with the calculated exponential fit and calculated overall percent released. (D) Average drug release (n = 3) profile for 37.8 wt % drug-loaded microwires with the calculated exponential fit and calculated overall percent released.

SEM images obtained after 10 days of incubation (Figures S28 and S29) confirmed that the structure of the hydrogel remained intact, with no evidence of delamination or morphological degradation. Surface features and approximate coating thickness remained consistent with the predrug release, drug-loaded microwires. These findings indicate that the plateau in drug release arises from depletion of available drug cargo, rather than degradation of the hydrogel matrix. A quantitative comparison to representative literature systems is provided in Table S4. Although reported conditions vary, many particulate/micellar platforms exhibit hours-to-few-days release plateaus (~0.3–4 days),^61–65 whereas our covalently grafted PEGDMA coatings sustain release over a multiday window (~9 days). Together, these results support our PEGDMA-based hydrogel coatings as robust, diffusion-controlled delivery vehicles in a device-integrated, surface-confined format that is well suited for localized release from electrode interfaces.

3.5. Iron Chelation Drug Release Study

To assess the efficiency of iron chelation and determine whether DFO release can be modulated by extracellular iron, DFO-loaded microwires were incubated in buffered media supplemented with a free iron carrier (Fe(EDTA)). At physiological pH, unbound Fe(III) is poorly soluble, whereas in vivo, it is maintained in solution by high-affinity chelators such as transferrin.⁶⁶ To model this effect, we introduced Fe(EDTA), a stable and fully soluble source of Fe(III) complex,⁶⁷ at high (10 mM) and low (1 mM) iron loads, into the release medium. Upon binding Fe(III), the iron-chelating drug (DFO) forms the Fe–DFO complex (Fe(DFO)₂), which is more water-soluble than unchelated DFO. Indeed, unchelated deferasirox is soluble up to 10 μM in water, while, by our calibration curve, we demonstrate solubility up to 1 mM in HEPES buffer conditions (Figure S32). We therefore hypothesize that complexation occurring within the hydrogel would perturb the innate diffusion properties of DFO and extend the release window.

Twenty 1 cm DFO-loaded microwires were bundled and incubated in 2 mM HEPES buffer at pH 7.4 containing either 1 mM or 10 mM Fe(EDTA). Formation of the Fe–DFO complex in the supernatant was quantified by UV–vis single wavelength analysis at 550 nm for 19 days (Figure 4). We note that the low concentrations of the Fe–DFO complex detected (in the μM range), compared to the mM concentrations of the free iron mimic Fe(EDTA), can be attributed to the high association constant of EDTA with Fe(III) in 100% water solution (K_a ≈ 10^25.1 M⁻¹),⁶⁸ which outcompetes the formation of the Fe–DFO complexes. However, the detection of the Fe–DFO complex under these conditions further underscores the effectiveness of this platform to chelate even strongly bound, stable iron species.

Figure 4.

Iron chelation drug release profile tracking the formation of Fe(DFO)₂ per cm of wire for bundled 18.9 or 37.8 wt % microwires after daily incubation in either 1 mM or 10 mM Fe(EDTA) in 2 mM HEPES buffer pH 7.4. Error bars represent standard deviation between three separate trials. 20 cm of drug-loaded wires was bundled for each trial.

Relative to iron-free conditions shown in Figure 3, where DFO fell below quantifiable levels by day 9, the presence of the free iron carrier, Fe(EDTA), substantially prolonged the detection of the Fe–DFO complex in solution. Notably, the free iron carrier concentration dictated the duration of release, such that both 18.9 and 37.8 wt % formulations showed near constant complex formation under 1 mM free iron carrier conditions until day 13, followed by a decline. In contrast, the 10 mM free iron carrier conditions supported sustained complex formation with no release plateau within 19 days. Although early time concentrations varied slightly with initial drug loading, the continued detection of the Fe–DFO complex after 13 days (in the presence of 10 mM Fe(EDTA)) indicates that the initial drug concentration within the hydrogel has minimal influence on chelation efficacy. Instead, other factors are likely governing long-term performance.

It is important to note that only the Fe–DFO complex was quantified spectroscopically in these experiments, and not the total mass of free DFO. Due to the striking differences in solubility between DFO, the Fe–DFO complex, and chemical moieties present in the drug release, as well as in-line UPLC iron chelation, in-tandem quantification of DFO and the Fe–DFO complex could not be accomplished via UPLC-MS without a high degree of relative error. As such, we do not directly report the full extent of release from the hydrogel. However, given that the initial drug loading was held constant between 1 mM and 10 mM Fe(EDTA) conditions for both the 18.9 and 37.8 wt % DFO formulations, the observed differences in the Fe–DFO complex concentration and release duration directly reflect the influence of environmental iron availability on complex formation. As free iron carrier penetrates the coating, in-gel formation and local accumulation of the Fe–DFO complex can partially occupy or constrict transport pathways, effectively reducing diffusion (“clogging”). This effect would scale with the concentration of free iron carriers, explaining why 10 mM Fe(EDTA) conditions sustain Fe–DFO complex detection longer than 1 mM conditions.

3.6. EDS Analysis of Microwire Platforms and Mechanistic Insight

To confirm the hypothesis that iron-containing species accumulate within the PEGDMA hydrogel matrix and in turn influence the DFO release dynamics and iron chelation efficacy, energy-dispersive X-ray spectroscopy (EDS) was performed on drug-loaded microwires exposed to 2 mM HEPES buffer at pH 7.4 and 37 °C with 1 mM and 10 mM Fe(EDTA) at 1, 10, or over a 20 -days of incubation period. Coatings were imaged in the dry state at high accelerating voltages to excite Fe Kα lines with sufficient sampling depth. For each wire and time point, elemental maps were registered with secondary electron images to delineate the hydrogel region, and Fe Kα areal counts were normalized to C, O, Ir, and Pt to account for variations in coating thickness and composition.

Elemental mapping at 30 kV accelerating voltage revealed minimal iron content in both hydrogel formulations on day 1 across all microwires incubated with free iron carriers Fe(EDTA) conditions (Figure 5A,B). However, a substantial increase in iron adsorption was observed for day 10 and day 20 microwires, particularly in the 10 mM Fe(EDTA) conditions, with the 18.9 wt % DFO-loaded wires accumulating the highest relative iron content (Figure 5A,C). Notably, while the iron content within the matrix increased for both 1 mM and 10 mM Fe(EDTA) conditions, the rate and magnitude of accumulation were substantially greater at the 10 mM free iron carrier concentration, aligning with the sustained detection of the Fe–DFO complex in solution. The 18.9 wt % conditions at day 20 showed the greatest incorporation of iron, consistent with the greater mesh size of 8.6 nm calculated for the 18.9 wt % DFO hydrogel in Section 3.3.

Figure 5.

(A) EDS electron images and Fe (kα1) mapping for microwires with 18.9 wt % drug incubated in 10 mM Fe(EDTA) with 2 mM HEPES buffer pH 7.4 for either 1 (top), 10 (middle), or 20 days (bottom). (B) EDS electron images and Fe (kα1) mapping for microwires with 37.8 wt % drug incubated in 10 mM Fe(EDTA) with 2 mM HEPES buffer pH 7.4 for either 1 (top), 10 (middle), or 20 days (bottom). (C) Bar graph of relative wt % and standard deviation for each EDS map for microwires with 18.9 wt % drug. (D) Bar graph of relative wt % and standard deviation for each EDS map for microwires with 37.8 wt % drug. Scale bars = 50 μm.

The detection of iron species at day 20 in the hydrogel matrix for the 1 mM Fe(EDTA) experiments (Figure 5C,D), while the Fe–DFO complex in solution is no longer detected, is indicative of physisorption of iron species in the hydrogel matrix and is assumed to be a baseline. Furthermore, it is important to note a marginal variation of the relative Fe wt % detected by EDS between day 10 (0.08%) and day 20 (0.11%) for the 1 mM Fe(EDTA)/18.9 wt % DFO experiment, while the detection of the solvated Fe–DFO complex between these two time points is dramatic; day 9 = 6 μM/cm and day 20 = 0 μM/cm. An identical trend is highlighted in Figure 5D for the 1 mM Fe(EDTA)/37.8 wt % DFO experiment. This observation suggests that the Fe(DFO)₂ complex detected in solution after day 9 is unlikely to originate from complexes trapped in the hydrogel matrix which are slowly diffusing out. If this were the case, the decrease in solvated complex concentration would be witnessed in tandem with a decrease in the relative wt % of iron species in the hydrogel matrix. A similar argument can be made by comparing the Fe wt % for the 10 mM Fe(EDTA) experiments for both DFO loadings. Indeed, although a 6-fold increase in iron wt % is observed between days 10 and 19 for the 18.9 wt % DFO-loaded microwire, the concentration of solvated Fe–DFO complex remains nearly unchanged over the same interval, with values of 3.9 μM on day 10 and 3.6 μM on day 20. The increased accumulation of [Fe] species in the hydrogel matrix under high concentrations of the free iron carrier suggests a model in which the formation of the Fe–DFO complex within the hydrogel perturbs the local chemical environment of the hydrogel matrix, ultimately impacting the diffusive behavior, and therefore chelation effectiveness, of pristine DFO.

To conceptually correlate the experimental data derived from drug-release profiles with complementary coating characterization (SEM and SEM-EDS), we depict in Figure 6, a schematic illustration of the release behavior in buffer versus iron-containing media. This schematic is intended as a qualitative, data-guided framework to aid interpretation of experimentally observed trends rather than a direct microscopic visualization of transport pathways. Under standard drug release conditions (no iron), the UPLC-MS quantified supernatant concentration decreases with time and is well described by an exponential decay for both loadings (Figure 3C,D), with the cumulative percent release approaching completion over the same window (~9 days). Accordingly, Figure 6A schematically represents diffusion-driven depletion of free DFO from the hydrogel coating into solution over time, with higher early release followed by progressively lower release as the DFO is depleted. Figure 6A also includes an “insoluble” fraction to qualitatively reflect the solubility-limited behavior in aqueous media. Postrelease SEM images collected are used qualitatively after drying to assess coating morphology, showing persistent hydrogel-derived coating features in both standard and iron-containing media (Figures S28 and S29 and S34 and S45).

Figure 6.

Qualitative conceptual schematic summarizing the proposed drug release behavior under (A) iron-free and (B) iron-supplemented conditions. This model is consistent with (i) rapid loss of detectable drug under iron-free release conditions (Figure 3), (ii) sustained detection of the Fe–DFO complex in solution as a function of Fe(EDTA) incubation concentration (Figure 4), and (iii) time-dependent accumulation of Fe within the hydrogel coating by SEM-EDS (Figure 5C,D). Not to scale.

In iron-containing media, UV–vis monitoring shows that the Fe–DFO complex is detected in the supernatant at substantially lower concentration than in buffer-only media and, at higher Fe(EDTA) concentration, is maintained at an approximately steady level over an extended period (~20 days) (Figure 4). This experimentally observed shift from a decaying release profile in the absence of iron-containing media (Section 3.4, Figure 3) to a sustained, lower-concentration profile in the presence of iron-containing media (Section 3.5, Figure 4) is represented schematically in Figure 6B by the Fe–DFO complex formation and time-dependent retention/partitioning between the hydrogel coating and supernatant. Consistent with this representation, SEM-EDS quantification shows that the coating-associated Fe signal increases from day 1 to days 10 and 20 for both 18.9% and 37.8% DFO-loaded matrices (Section 3.6, Figure 5), supporting inclusion of progressive iron accumulation in the schematic. Taken together, the UPLC-MS release kinetics in the absence of iron-containing media, UV–vis monitored release behavior of the Fe–DFO complex, and SEM/SEM-EDS observations provide the experimental basis for the qualitative schematic depicted in Figure 6.

Taken together, the EDS and UV–vis data indicate that the chemical microenvironment within the hydrogel matrix of the DFO-loaded microwire critically influences long-term iron chelation efficacy, and, perhaps more importantly, that the release profile of unchelated DFO alone can be misleading and may overlook important behaviors. Indeed, our results indicate that the main driver is the concentration of free iron carriers, as evidenced by the strong contrast of chelation efficacy between 1 mM and 10 mM. This free iron responsiveness provides a foundation for engineering adaptive, stimuli-responsive, drug delivery systems, where therapeutic release can be selectively extended in environments containing high iron loads, such as sites of neuroinflammation or injury.

4. CONCLUSIONS

Herein, we present the development of a responsive polyimide-based implant coating capable of modulating release dynamics in response to external iron load. By leveraging established surface functionalization strategies, we attach covalently cross-linked hydrogels onto microwires and successfully load them with the iron-chelating drug, DFO. Standard drug release profiling in buffer reveals a Fickian diffusion-based release that is ultimately limited by the poor aqueous solubility of unchelated DFO. Remarkably, under iron-rich conditions, we observed an extension in the drug release lifetime, with higher Fe(EDTA) concentrations supporting sustained delivery of the Fe–DFO complexes for longer than 18 days. This effect is not simply due to lower diffusion but rather reflects a mechanistic shift in the release pathway, where in matrix complexation and retention dictates the release profile. EDS studies confirm that iron accumulates within the hydrogel overtime, correlating with prolonged release and suggesting that the formation of the drug-iron complex modulates both the solubility and matrix compatibility of the cargo.

Together, these findings establish a drug delivery platform whose pharmacokinetics can be tuned by the local biochemical environment without altering the coating’s composition. The system responds to iron concentrations, potentially enabling site-specific adaptive drug release behaviors based on inflammatory state, disease progress, or systemic iron levels,⁶⁹ and exhibits extended release time scales relative to representative benchmarked chelator-delivery systems. Future work will focus on evaluating the long-term biocompatibility and therapeutic efficacy of these coatings in vivo. Specifically, studies will aim to investigate the persistence and spatial distribution of inflammatory markers, such as IL-4, IL-2α, and TNFα, following cranial implantation in rodent models. These investigations will further elucidate how dynamic, stimuli-responsive coatings affect inflammatory signaling and support the rational design of adaptive implant systems.

Supplementary Material

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsami.5c23344.

Experimental details and additional information on functionalization, hydrogel formation, characterization, and analysis (PDF)