Evaluation of Tyrosine Kinase Inhibitors Loaded Injectable Hydrogels for Improving Connexin43 Gap Junction Intercellular Communication

Abstract

Myocardial infarction (MI) is a leading cause of death globally and loss of cardiomyocytes plays a critical role in the pathogenesis of heart failure. Implicated in this process is reduced gap junction intercellular communication due to remodeling of Connexin43 (Cx43). We previously identified that intraperitoneal injection of the Pyk2 inhibitor PF4618433, reduced infarct size, maintained Cx43 at the intercalated disc in left ventricle hypertrophic myocytes, and improved cardiac function in a MI animal model of heart failure. With the emergence of injectable hydrogels as a therapeutic towards regeneration of cardiac tissue after MI, here we provide proof of concept that release of kinase inhibitors from hydrogels could have beneficial effects on cardiomyocytes. We developed an injectable hydrogel consisting of thiolated hyaluronic acid and P123-maleimide micelles, that can incorporate PF4618433 as well as the Src inhibitor Saracatinib and achieved sustained release. Using neonatal rat ventricular myocytes in the presence of a phorbol ester, endothelin-1, or phenylephrine to stimulate cardiac hypertrophy, release of PF4618433 from the hydrogel had the same ability to decrease Cx43 tyrosine phosphorylation and maintain Cx43 localization at the plasma membrane as when directly added to the growth media. Additional beneficial effects included decreases in apoptosis, the hypertrophic marker Atrial natriuretic peptide (ANP), and serine kinases upregulated in hypertrophy. Finally, presence of both PF4618433 and Saracatinib further decreased the level of ANP and apoptosis than each inhibitor alone, suggesting a combinatorial approach may be most beneficial. These findings provide the groundwork to test if tyrosine kinase inhibitor release from hydrogels will have a beneficial effect in an animal model of MI induced heart failure.

Graphical Abstract

1. Introduction

Gap junction channels composed of Connexin43 (Cx43) mediate electrical coupling and impulse propagation in the normal working myocardium. Following a myocardial infarction (MI), the cardiac muscle sarcoplasm from the surviving ventricular myocytes (epicardial border zone), experience many abnormalities, including reduced gap junction intercellular communication due to changes in Cx43 expression, phosphorylation state, and increased relocalization from the intercalated disc to the lateral membrane^1–7. Reduced gap junctional coupling causes instability of action potential duration and refractory period, effects seen in the epicardial border zone⁸.

Left ventricle hypertrophy following a MI is an adaptive response to increased biomechanical stress⁹. Initially, heart mass increases to normalize wall stress and retain normal cardiovascular function, accompanied by an increase in Cx43 expression and gap junction intercellular communication¹⁰. However, the increased cardiac mass and sustained overload eventually lead to contractile dysfunction and heart failure¹¹. As left ventricle hypertrophy becomes severe, propagation velocity decreases, which correlates with reduced Cx43 expression, altered phosphorylation which affects channel regulation and protein partners interactions, and translocation from the intercalated disc to the lateral membrane^12–14. While Cx43 lateralization is an almost ubiquitous response to cardiac pathology^{8, 15, 16}, the molecular mechanisms of this remodeling are ill-defined. Based upon data from our lab and others, a model was put forth describing a mechanism for laterization involving Src and Pyk2 phosphorylation of Cx43 causing dissociation of ZO-1 (binds F-actin), β-tubulin, and Drebrin (binds F-actin)^{7, 17–19}. Initially, ZO-1 unhooks from Cx43 causing the plaque to increase in size^{20, 21}. Then as it is no longer anchored by the cytoskeletal network, Cx43 moves away from the region of high concentration at the intercalated disc to the lateral membrane^{7, 22}. Lateralization is a key factor that reduces conduction leading to reentrant arrhythmias^{8, 23}.

Unlike other attempts to decrease Cx43 remodeling^{13, 24, 25}, inhibition of Pyk2 activity in an animal model of heart failure using the small molecule inhibitor PF4618433 maintained Cx43 at the intercalated disc in left ventricle hypertrophic myocytes and improved cardiac function²⁶. Additionally, although the acute time frame was not mechanistically investigated, PF4618433 also reduced the infarct size²⁶. PF4618433 is a diaryl urea small molecule inhibitor that binds Pyk2 in an allosteric binding pocket distinct from the ATP pocket to help minimizing off-target activity²⁷. In comparison to classical Pyk2 inhibitors, PF4618433 had superior overall selectivity (>35 kinases tested) and improved potency to other allosteric binding pocket Pyk2 inhibitors with an IC₅₀ of 637 nM²⁷. Pyk2 is activated following MI in human left ventricle implicating a novel potential target for therapy in patients with heart failure²⁶. Of note, the Src inhibitor Saracatinib was also included in this study because Src directly activates Pyk2¹⁴. In an animal model, Saracatinib decreased Src activity to the sham level, raised Cx43 expression at the scar border and distal ventricle, and had a higher level of the Cx43 P2 isoform (correlates with promoting gap junction intercellular communication)¹³. While no decrease in lateralization was observed, this still led to ~50% improvement in conduction velocity and lowered arrhythmia inducibility¹³.

Current available treatments of MI include pharmaceutical therapy, medical device implants, and organ transplants, however they all have their inherent limitations^28–30. Injectable hydrogels have emerged as a promising translational therapeutic towards regeneration of cardiac tissue after MI. These hydrogels have the capacity to encapsulate cells and/or locally release therapeutics after injection through a syringe or catheter^{31, 32}. Several formulations for injection into cardiac tissue have been demonstrated as safe in animals and are in clinical trials^{33, 34}. Recent hydrogels engineered for cardiac tissue to reverse the adverse myocardial microenvironment include matrix metalloproteinase-responsive hydrogels, reactive oxygen species-scavenging hydrogels, and immunomodulatory hydrogels³⁵. Hydrogels have been designed to improve electrical signal conduction and vascularization within the infarct area to promote the repair of cardiac function. Finally, 3D-printed hydrogels, which can achieve personal customized cardiac tissue via printing of intact cardiac structures, were developed to address the current shortage of heart donors³⁶. Here, we developed a micelle crosslinked hydrogel and tested the release of Pyk2 and Src inhibitors that have shown to be beneficial after a MI (in acute ischemia and in late stage left ventricle hypertrophy), however the mode of delivery was intraperitoneal injection. For this in vitro proof of concept study, we tested three different reagents phorbol 12-myristate 13-acetate (PMA), endothelin-1 (ET-1), and phenylephrine. ET-1 and phenylephrine are activated during ischemic injury and heart failure^37–40. All three induce cardiac cell hypertrophy and re-expression of cardiac-specific fetal genes^{14, 41–46}. We anticipate that our hydrogel system will be an improved delivery vehicle because the amount of inhibitor needed will be less, targets only the cardiac tissue, and allows for slow release over a long period of time.

2. Materials and methods

2.1. Synthesis of thiolated hyaluronic acid (HA-SH) and P123-maleimide (PM)

The HA-SH polymer conjugate was prepared by conjugating cysteamine (Sigma) to HA (290 kDa, Bloomage Biotech) using 1-ethyl-3-(3-dimethyl aminopropyl) carbodiimide (EDC)/N-hydroxysuccinimide (NHS) as the coupling agent (Fig. 1)⁴⁷. In brief, 200 mg of HA (0.5 mmol) was dissolved in 20 mL of de-ionized (DI) water and was activated after adding 384 mg of EDC (2 mmol, Oakwood Chemical) and 115 mg of NHS (1 mmol, Oakwood Chemical) under constant stirring. Next, 230 mg of cysteamine (2 mmol) was poured into the above mixture. Then the pH of the solution was adjusted to 5.0 using a 1N HCl (Fisher Chemical) solution. The reaction mixture was stirred at room temperature (RT) for 24 hr, transferred to a dialysis bag with a 6–8 kDa molecular weight cut-off (MWCO, Spectrum) and dialyzed against acidic DI water (with the addition of 1 mM tris(2-carboxyethyl)phosphine or TCEP at pH 3) for 3 days at RT, with the water changed every 12 hr. The purified conjugate solution was then neutralized with sodium hydroxide solution (1N NaOH, Fisher) and freeze-dried in a benchtop lyophilizer (model FreeZone, Labconco) for 2 days to acquire the purified HA-SH polymer conjugate. The conjugates were stored at −80°C.

Figure 1.

Schematic illustration of the micelle crosslinked injectable hydrogels for dual TKI delivery to cardiomyocytes.

The PM conjugate was synthesized in two steps. In the first step, 580 mg of Pluronic P123 (0.1 mmol, Sigma) and 122 mg of 4-dimethylaminopyridine (DMAP, 1 mmol, Sigma) were dissolved in 2 mL of 1,4-dioxane (Sigma) with 139 μL of triethylamine (TEA). After stirring at RT for 15 min, 200 mg of succinic anhydride (2 mmol, Sigma) in 2 ml of 1,4-dioxane was added to the Pluronic P123 solution⁴⁸. The mixture was stirred at RT for 24 hr and then the organic solvent was removed in vacuum. The residual sample was precipitated in ice-cold ether (EMD Millipore) twice. The precipitate was dissolved in water and then freeze-dried to obtain the white powder of dicarboxylated Pluronic P123 (P123-succinic acid). In the second step, after dissolving 100 mg of P123-succinic acid (17 μmol) in 5 mL of ice-cold dichloromethane (DCM), 23 mg of 1-Hydroxybenzotriazole (HOBt, 170 μmol, Oakwood) in 100 μL of dimethyl sulfoxide (DMSO, Fisher) was added dropwise and the mixture was stirred for 10 min. Next, 15 mg of N-(2-aminoethyl)maleimide hydrochloride (85 μmole TCI) in 100 μL of DMSO and 10 μL of TEA were added respectively. After stirring for another 15 min, 32.6 mg of EDC (170 μmol) in 300 μL of DMSO was added to the solution and the final mixture was stirred at RT for 24 hr. DCM was removed by vacuum and the residual was solubilized in water. The reaction mixture solution was dialyzed against water for 3 days and the purified PM was lyophilized and kept at −20°C before use.

2.2. Proton nuclear magnetic resonance (¹H NMR) analysis of polymers

¹H NMR characterization was performed on a 500 MHz Bruker NMR system and analyzed with Topspin 4.0 software. The polymer conjugate was dissolved in D₂O (Acros Organic) at 5 mg/mL for NMR acquisition, with the chemical shifts referring to the solvent peak of D₂O at 4.78 ppm at 25°C. The substitution degree of HA-SH was determined from the ratio of the integral of ethyl protons from the conjugated cysteamine (between 2.2~2.4 ppm, 4H, -C₂H₄) to the integral of the HA methyl proton peak (at 1.9 ppm, 3H, -CH₃). The grafting ratio of PM was calculated from the ratio of the integral of the alkene protons from the attached maleimide group (between 6.7~6.9 ppm, 2*C₂H₂) to the integral of P123 methyl proton peak (~1.1 ppm, 70*CH₃).

2.3. Preparation and rheological characterization of HA-PM hydrogels

An equal volume of 5% (w/v) PM dissolved in water was quickly mixed with either 1% or 2% (w/v) HA-SH in water to prepare the HA-PM hydrogels and the two hydrogels were referred to as the 5:1 and 5:2 HA-PM hydrogel, respectively⁴⁹. To investigate the gelation behavior, rheological studies were conducted. The rheological properties of the HA-PM hydrogels were characterized at 37°C using a Discovery HR-2 rheometer (TA Instruments) with several different protocols⁵⁰. In each study, around 0.2 mL of total hydrogel precursor solution was mixed in an Eppendorf tube and thus quickly transferred into the geometry gap between the 20 mm parallel plate and the base plate. The time sweep method was carried out with a fixed angular frequency of 2π rad/s (=1 Hz) and 10% strain to measure the storage modulus (G’) and loss moduli (G”) over time. Frequency sweeps from 0.1 to 10 Hz were performed at 10% constant strain to validate the hydrogel viscoelasticity. Flow sweep study was conducted on the hydrogels with a linearly ramped shear rate from 1 to 100 s^–1, and the viscosity recorded at different rates.

2.4. Swelling and degradation study

Each piece of hydrogel (50 μL) was incubated in 1 mL of PBS (Fisher, pH 7.4) in a 1.5 mL Eppendorf tube at 37°C. At each time point, hydrogel was taken out from the PBS, weighed, and then transferred into 1 mL of fresh PBS⁵⁰. Weight of each hydrogel sample (W_s) was determined after removing excess PBS from the hydrogel surface. The initial weight was set as W₀. Hydrogel swelling was monitored over 24 hr and calculated with the equation: W_s/W₀×100%. The degradation ratio was calculated using the equation: (W₀-W_s)/W₀×100%. Each group had three replicates.

2.5. Primary cardiomyocyte culture and MTT study

The animal procedure was approved by the institutional animal care and use committee at the University of Nebraska Medical Center. Primary cardiomyocytes (neonatal rat ventricular myocytes, NRVMs) were isolated from day 1–3 neonatal rat hearts using the Pierce Primary Cardiomyocyte Isolation Kit (>90% pure) per manufacturer protocol (catalog number: 88281)¹⁴. NRVMs were plated at a seeding density of 2.5 × 10⁵ cells/cm² in complete DMEM supplemented with 10% FBS and 1% pen-strep at 37°C in a 5% CO₂ incubator for 24 hr. The medium was then replaced with complete DMEM containing Cardiomyocyte Growth Supplement for 3 days. A piece of hydrogel (50 μL) was then added to each well of cells. The control group contained no hydrogel. Each group had 3 replicates. After 2 days incubation, the cytotoxicity was determined using MTT (Sigma) assay with the absorbance measured at 540 nm in a microplate reader (BioTek)¹⁴. The absorbance value from each group was normalized to the average absorbance value in the control group.

2.6. Preparation of P123 micelles loaded with dual tyrosine kinase inhibitors (TKIs)

Both Src and Pyk inhibitors are hydrophobic and show limited water solubility. PM micelles were applied to solubilize the drugs. The micelles were prepared by thin film hydration method⁴⁹. Generally, both 5% (w/v) PM conjugate was first dissolved in methanol. TKI stock solution was also prepared in methanol. The PM methanol was then added to aliquot amount of TKIs in an Eppendorf tube. After complete mixing, the methanol was evaporated in vacuum and a thin film was formed on the tube surface. Next an equal volume of DI water was added to hydrate the film for at least 15 min at RT. The mixture was sonicated for another 15 min and the non-encapsulated TKIs were removed by centrifugation at 10,000 g for 5 min. The size and polydispersity (PdI) of the micelles encapsulated with each drug was determined by dynamic light scattering measurement (Malvern). The non-encapsulated TKIs were dissolved in 100 μL of methanol and injected into high performance liquid chromatography (HPLC) to determine the drug concentration (method described below). The encapsulation (EE) was calculated by the equation: EE = (Weight of feeding drug - Weight of non-encapsulated drug) / Weight of feeding drug × 100%.

2.7. FTIR characterization

The polymer precursor, blank hydrogel, free drug, drug loaded micelle and drug loaded hydrogels were analyzed using a Nicolet Is50 FT-IR spectrometer (Thermo Scientific) in the range 500–4000 cm⁻¹. The data was recorded with the OMNIC software. The drug/polymer/lyophilized hydrogel powder was mixed with KBr and subjected to scanning.

2.8. Scanning electron microscopy analysis

The HA-SH were crosslinked to the drug loaded micelles to prepare TKIs loaded hydrogels. The micro-structure of the freeze-dried HA-PM hydrogel was studied by scanning electron microscopy (SEM, FEI Quanta 200). The freshly prepared drug loaded HA-PM hydrogels were completely lyophilized and then sprayed on a thin film of gold. The sample was then transferred to the SEM instrument for examination.

2.9. HPLC analysis

The release of both drugs from the TKI loaded hydrogels was determined by the dialysis method in PBS solution. The released amount of individual drug at each time point was simultaneously quantified by the Agilent 1260 HPLC at 250 nm equipped with a diode array detector by injecting 30 μL of the releasate into the HPLC instrument. A Poroshell 120 EC-C-18 column (3.0 × 150 mm, 2.7 μm particle size) was used for drug separation and quantification at 30°C and 0.5 mL/min flow rate¹⁴. The mobile phase consisted of an aqueous phase (A) containing 0.1% (v/v) formic acid and an organic phase (B) containing acetonitrile. A gradient method was used, in which the mobile phase started as 80% A and 20% B and linearly changed to 20% A and 80% B in 8 min. The mobile phase ratio was held consistent between 8 to 10 min and was switched to 5% A and 95% B by the12^th min. It steadily increased to 80% A and 20% B by the 15^th min. A standard curve was generated with known concentrations of TKIs (1–10 μg/mL) prepared in water/ACN (50:50 v/v). The percentage of drug release was calculated by dividing the accumulative amount of released drugs to the total encapsulated drugs at the beginning.

2.10. Pyk2 and Src inhibitor treatment

After purification and seeding (see MTT section above for more details), NRVMs were pre-treated with PF4618433, Saracatinib, or both (either dissolved in medium or incorporated within hydrogels) for 3 hr and then either PMA, ET-1, or Phenylephine was added for 1 hr or 24 hr. Of note, the NRVMs were placed in the bottom of a live imaging dish and the hydrogel was added to the outer ledge, not in direct contact with the cells.

2.11. Western blots

NRVMs were rinsed 2x with cold 1x Tris-buffered saline (TBS) and then homogenized in lysis buffer (50 mM Tris-HCl [pH 7.4], 150 mM NaCl, 0.2% SDS, Complete protease inhibitor, and PhosSTOP) on ice for 30 min¹⁴. Protein concentration was quantified by BCA assay (Pierce). A total of 30 μg protein lysate was resolved by SDS-PAGE and transferred to PVDF (EMD Millipore) membrane. Blots were blocked in 5% BSA in TBST (1x TBS, 0.1% Tween 20) for 1 hr at RT and incubated with indicated primary antibody in blocking buffer overnight at 4°C. Blots were washed 3× 10 min with TBST, and then incubated for 1 hr at RT with secondary antibody. Blots were then washed again 3× 10 min with TBST, detected using the SuperSignal West Femto Maximum Sensitivity (ThermoFisher) per manufacturer protocol, and imaged with an Invitrogen iBright Imager. Quantifications were done using NIH ImageJ software using a minimum of three independent replicates.

The following antibodies were used in this study: the α-phospho-Cx43 (Y247) antibody was custom produced in rabbit from a peptide designed and chemically synthesized by LifeTein. Additional antibodies used in this study include α-Src (#2123S), α-phospho-Src (Y416) (#6943S), α- protein kinase C (PKC) (#2056), α-phospho-PKC (S660) (#9375), α-Erk1/2 (#4695), α-phospho-Erk1/2 (T202/Y204) (#4370), α-JNK (#9252), α-phospho-JNK (T183/Y185) (#4668), α-Pyk2 (#3480S), α-Cleaved Caspase 9 (#9505S), α-rabbit-Alexa488 (#4412S), α-mouse-Alexa647 (#4410S), α-rabbit-HRP (#7074), and α-mouse-HRP (#7076) purchased from Cell Signaling; α-phospho-Pyk2 (Y579/580) (#44–636G) purchased from ThermoFisher; ANP (sc-515701) purchased from Santa Cruz; α-Cx43 (#C6219) purchased from Sigma; α-phospho-Cx43 (Y265) (ab193373), α-phospho-Pyk2 (Y402) (ab4800), and α-alpha actinin (ab90421) purchased from Abcam. DAPI (#5748) was purchased from Tocris Bioscience.

2.12. TUNEL Assay

The apoptosis of NRVMs was measured using TUNEL Assay Kit (Roche) per manufacturer protocol (catalog number: 11684817910)⁵¹. Primary cardiomyocytes were pre-seeded on live imaging dishes coated with Fibronectin. Cells were fixed with 4% PFA for 1 hr and then rinsed with PBS and incubated with permeabilization solution (0.1% Triton X-100 in 0.1% sodium citrate) for 2 min on ice. Then the cells were rinsed with PBS and incubated with TUNEL reaction mixture at 37°C for 60 min, followed by incubation with DAPI for 10 min. The results were observed and quantified by confocal fluorescence microscopy.

2.13. Immunofluorescence

NRVMs were pre-seeded on coverslips coated with Fibronectin. Cells were fixed with 4% PFA for 30 min and then washed with TBST and blocked (1x TBS with 2.5% horse serum and 0.2% Triton) for 1 hr at RT. Cells on the coverslips were incubated at 4°C overnight with primary antibodies. The next day, coverslips were washed 3× 10 min with 1x TBST, incubated with secondary antibodies for 1 hr at RT, stained with DAPI (100 ng/mL) for 10 min, and then washed 3× 10 min with 1x TBST. Coverslips were mounted on a drop of SlowFade anti-fade (Life Tech), sealed with clear nail polish, and imaged.

2.14. Confocal imaging

All cell immunofluorescence images were acquired on a Zeiss LSM 800 Confocal system using appropriate numerical aperture objectives and appropriate filter sets.

2.15. Statistical Methods

All data were analyzed by using GraphPad Prism 8.0 and presented as the mean ± SD. Statistical analysis performed in GraphPad Prism 8.0 were either one-way ANOVA with a Neuman-Keuls post-hoc analysis or Student’s t-test where appropriate. P-values <0.05 were considered statistically significant.

3. Results

3.1. Synthesis and characterization of HA-PM hydrogels.

A micelle crosslinked hydrogel system was developed to deliver two TKIs simultaneously to the NVCMs, thereby to synergistically influence Cx43 gap junction intercellular communication (Fig. 1). Two polymer precursors were first synthesized before the fabrication of the hydrogels.

The HA was conjugated with the cysteamine by EDC/NHS mediated coupling and ¹H-NMR result revealed a roughly 20% grafting ratio (Fig. 2A). The hydroxyl terminal of the Pluronic P123 was first activated by the succinic anhydride at high efficiency (~90% determined by ¹H-NMR). The P123 dicarboxylate was further coupled with the N-(2-aminoethyl) maleimide and the resulting PM showed a ~45% grafting ratio (Fig. 2B).

Figure 2. NMR characterization of synthesized hydrogel precursors.

A) ¹H NMR spectra of HA and HA-SH. B) ¹H NMR spectra of P123, P123-succinic acid, and P123-maleimide.

PM was dissolved in water at 5% and HA-SH was prepared at two concentrations: 1% and 2%, in water, respectively. The hydrogels were fabricated by mixing the two polymer conjugates together using the thiol-ene click reaction. Both 1% and 2% HA-SH formed hydrogels with the 5% PM rapidly (Fig. 3A). The time sweep testing indicated the storage modulus (G’) of both hydrogels plateaued within 8 min at 37°C. The 5:1 hydrogel showed a G’ of 140 ± 10 Pa and loss modulus (G”) of 15 ± 5 Pa while the 5:2 hydrogel showed a G’ of 355 ± 25 Pa and G” of 30 ± 10 Pa. Both hydrogels maintained their elastic behavior over the tested frequency range (0.1 – 10 Hz) as shown by the continual G’ > G” (Fig. 3B). FTIR further confirmed the two major polymer components in the hydrogel (Supplemental Fig. 1). The hydrogels possessed shear-thinning ability as confirmed by the sharp viscosity decrease when the shear rate ramped from 1 s⁻¹ to 100 s⁻¹ (Fig. 3C). The viscosity of the 5:2 hydrogel decreased sharply from 160 Pa-s at 1 s⁻¹ shear rate to 10 Pa-s at 100 s⁻¹. Meanwhile, the viscosity of the 5:1 hydrogel dropped from 74 Pa-s at 1 s⁻¹ to 1.3 Pa-s at 100 s⁻¹. The hydrogels were injectable and could be smoothly extruded through a 29 G needle (Supplemental Video 1). The compressive stress of both hydrogels was also characterized, and the recovery study also proved good injectability of the hydrogels (Supplemental Fig. 2).

Figure 3. Characterization of injectable HA-PM hydrogels.

A) Time sweep rheological testing indicating the storage modulus (G’) and loss modulus (G”) change of two hydrogels with different precursor ratio over time. (B) Frequency sweep rheological. (C) Viscosity of the two hydrogels with shear rate increasing from 1 s⁻¹ to 100 s⁻¹. The inserted image indicated that a dye-colored hydrogel was smoothly injectable through a 29 G needle. D) Swelling ratio of the two hydrogels over 24 hr. E) Degradation profile of two hydrogels over 21 days. F) Cyto-compatibility of the hydrogels determined by MTT study (n=4; **, p<0.01).

The 5:2 hydrogel showed more swelling ratio compared to the 5:1 hydrogel. This was probably caused by the higher amount of hydrophilic HA component in the 5:2 hydrogel (Fig. 3D). The 5:2 hydrogel was also relatively more stable than the 5:1 hydrogel in PBS (Fig. 3E). The 5:1 hydrogel degraded ~65% in 21 days while the 5:2 hydrogel only degraded ~25% in the same period. Finally, the cyto-compatibility of the hydrogels was evaluated on the NVCMs and the result implied no obvious cytotoxicity on the cells for the 5:2 hydrogel but nearly 15% decrease for the 5:1 hydrogel based on the MTT study (Fig. 3F). We decided to use the 5:2 hydrogel for further studies due to its better stability and cytocompatibility.

3.2. In vitro drug release from the TKIs loaded hydrogels.

We found Pluronic P123 could effectively encapsulate two TKIs, i.e., Saracatinib and PF4618433, presumably due to the micelle formation. On the other hand, Pluronic F-127 was able to encapsulate Saracatinib but not PF4618433, making it unsuitable for our purposes. In the 5% PM solution, we were able to dissolve up to 4 mg/mL of Saracatinib and 2 mg/mL of PF4618433, achieving an encapsulation efficiency of over 95%. However, our previous study indicated Saracatinib at 4 mg/mL was just as effective as 0.2 mg/mL to improve Cx43 gap junction intercellular communication¹⁴. Therefore, we prepared the drug loaded PM micelles with 0.2 mg/mL Saracatinib and 2 mg/mL PF4618433. We compared the particle size and PdI of micelles loaded with different drugs using dynamic light scattering. The results showed that PF4618433 loaded micelles had comparable size with the Saracatinib loaded micelles, although the PdI of the PF4618433 micelles was significantly smaller than the Saracatinib micelles (Fig. 4A, B). This was probably because the PF4618433 was more hydrophobic than the Saracatinib and could be packed more tightly in the hydrophobic cores of the micelles. When both drugs were loaded, the dual drug loaded micelles showed similar size and PdI compared to the PF4618433 loaded micelles. The encapsulated drugs did not have adverse effect in the hydrogel formation.

Figure 4. Characterization of drug loaded micelles and drug release from the micelles and hydrogels.

A, B) Dynamic light scattering analysis of different drug loaded micelles. C) SEM image of HA-PM hydrogel loaded with both drugs. D) Representative HPLC chromatography of dual-drug separation under the optimized gradient mobile phase condition. E) Dual drug release profile from the hydrogels over a 14 day period. SA, Saracatinib; PF, PF4618433.

The G’ and G” of the dual drug loaded micelle crosslinked hydrogels were comparable to the empty micelle crosslinked hydrogels. Noteworthy, the 5:2 hydrogel showed relatively smaller pores compared to the 5:1 hydrogel due to increased crosslinking degree (Fig. 4C). No obvious drug precipitates/aggregates were found in the SEM image of the hydrogels, indicating successful encapsulation and dispersion of the drugs in the micelles and hydrogels. This is also proved by the FTIR spectra of dual drug loaded hydrogel, which showed the characteristic peaks of dual drugs, validating the presence of both drugs in the hydrogel (Supplemental Fig. 3). HPLC was used to monitor the drug release, which could determine the concentrations of both drugs at the same time from the releasate (Fig. 4D). Saracatinib was released faster from the hydrogels than PF4618433 because it is more hydrophilic. It showed a burst release of around 43% on day 1 and an average of 92% release on day 14 (Fig. 4E). In contrast, there was only 12% of PF4618433 released on day 1 and an average of 32% was released on day 14.

3.3. Hydrogel release of the Pyk2 inhibitor PF4618433 has the same effect on PMA treated NRVMs as when directly added to the media.

Due to the lower release from the hydrogel of PF4618433 when compared to Saracatinib, here we initially tested if the biological effect of the PF4618433 would be similar to when the inhibitor was directly added to the media¹⁴. NRVMs were pretreated with PF4618433 and then exposed to PMA, which we and others have shown activates Pyk2 and Src leading to increased tyrosine phosphorylation of Cx43 as well as down regulation of gap junction intercellular communication^{14, 52}. Western blot data demonstrates that PF4618433 directly added to the media or released from the hydrogels reduced the activation of Src and Pyk2 in a similar manner (Fig. 5A, B). The level of phosphorylation on Cx43 residues Y247 and Y265 was also equally reduced by both delivery methods of PF4618433. The data suggests PF4618433 released from the hydrogel can influence signaling pathways and Cx43 regulation in NRVMs.

Figure 5. Comparing the effect of PF4618433 (PF) when released from the hydrogel to when directly added to the media on tyrosine kinase activity and Cx43 regulation.

A) NRVMs were pre-treated with PF (5 μM) in medium or in hydrogel (10 μg) for 3 hr and then treated with PMA (300 nM, 60 min). Lysate from each group was Western blotted. Antibodies used are labeled on the left of each panel. B) Protein levels (PF in medium, top; PF in hydrogel, below) were quantified by analyzing scanned blots using ImageJ software Phosphorylated protein levels were normalized to total protein. Data are representative of three independent experiments (one-way ANOVA, n=3, *P < 0.05, **P < 0.01, ***P < 0.001).

3.4. Hydrogel release of the Src inhibitor Saracatinib or Pyk2 inhibitor PF4618433 maintains Cx43 at the plasma membrane in PMA treated NRVMs.

As a control, presented is immunofluorescence data demonstrating that 60 min of PMA treatment causes internalization of Cx43 from the plasma membrane (Fig. 6A). Sarcomeric α-actinin was used as a cardiac cell marker. Release of the Src inhibitor Saracatinib from the hydrogel increased the level of plasma membrane Cx43, however a significant amount still remained intracellular. The Pyk2 inhibitor PF4618433 was more effective than Saracatinib at maintaining Cx43 at the plasma membrane. The combination of Saracatinib and PF4618433 (Dual) had no additional effect than PF4618433 alone. These observations in NRVMs were identical when both TKIs were directly added to the media prior to the addition of PMA¹⁴.

Figure 6. Effect of Saracatinib and PF4618433 (PF) release from the hydrogel on cellular localization of Cx43.

A) NRVMs cultured in 12-well plate were pre-treated with hydrogel containing Saracatinib (1 μg/well), PF4618433 (10 μg/well), or both (Dual, 3 h) and then treated with PMA (300 nM, 60 min). A representative fluorescent image is shown for each group (red, cardiac marker sarcomeric α-actinin; green, Cx43; blue, DAPI; white arrows, Cx43 in the junctional fraction at the plasma membrane; yellow arrows, internalized Cx43). B) Quantification was the percentage of total fluorescent intensity of Cx43 at GJ plaque to total Cx43. (one-way ANOVA, *P < 0.05, ***P < 0.001).

3.5. Hydrogel release of the Src inhibitor Saracatinib or Pyk2 inhibitor PF4618433 reverses PMA induced hypertrophy in NVCMs.

PMA activates the fetal gene program (e.g., ANP) and increases cardiomyocyte size via PKC signaling^{41, 42}. Here we investigated if Saracatinib and/or PF4618433 could reverse the hypertrophic response to PMA. Immunofluorescence data demonstrates that 24 hr of PMA treatment causes an increase in ANP expression and cell area (Fig. 7A, B). The release of Saracatinib or PF4618433 from the hydrogels had a similar effect to decrease the level of cells expressing ANP induced by PMA treatment back to the level of control no PMA. Interestingly, when both inhibitors were released together (Dual), they statistically had a lower level of ANP than either Saracatinib or PF4618433 alone (albeit still not different from control no PMA). The release of Saracatinib or PF4618433 had a similar effect to decrease the cell area compared to PMA treatment group, with the combination of both have no additive effect.

Figure 7. Effect of Saracatinib and PF4618433 (PF) release from the hydrogel on PMA mediated hypertrophy.

A) NRVMs cultured in 12-well plate were pre-treated with Saracatinib (SA, 1 μg/well), PF4618433 (PF, 10 μg/well), or both (Dual, 3 hr) and then treated with PMA (300 nM, 24 hr). A representative fluorescent image is shown for each group (red, cardiac marker sarcomeric α-actinin; green, ANP; blue, DAPI). B) Cell surface area and percentage of total fluorescent intensity of ANP to DAPI were quantified using Image J software. C) Lysate from each group was Western blotted. Antibodies used are labeled on the left of each panel. D) Protein levels were quantified by analyzing scanned blots using Image J software, with normalization of protein expression to the control lane (value set arbitrarily as 1). Phosphorylated protein levels were normalized to total protein. Data are representative of three independent experiments (one-way ANOVA, n=3, *P < 0.05, **P < 0.01, ***P < 0.001).

Next, signaling pathways activated by PMA were evaluated to see which may have been downregulated to contribute to the decreased ANP level and cell size. Western blot data demonstrates that 24 hr of PMA treatment activates PKC, ERK, and JNK (Fig. 7C, D). The release of Saracatinib from the hydrogel decreased the activity of PKC (but not to control no PMA level) and had no effect on ERK and JNK. Conversely, PF4618433 decreased the activity of PKC, ERK, and JNK, with PKC and ERK returning to the control no PMA level. The combination of both Saracatinib and PF4618433 (Dual) had no additional effect on PKC, ERK, and JNK activity than PF4618433 alone.

3.6. Hydrogel release of the Src inhibitor Saracatinib or Pyk2 inhibitor PF4618433 reverses ET-1 and phenylephrine induced hypertrophy and apoptosis of NVCMs.

ET-1, a vasoconstrictor peptide secreted from vascular endothelial cells as well as by cardiomyocytes, induces hypertrophy of cardiomyocytes through activating phospholipase C, PKC, Erk1/Erk2, and upregulation of c-Fos, c-Jun, and JNK^{53, 54}. Immunofluorescence data demonstrates that 24 hr of ET-1 treatment causes an increase in ANP and cell area (Fig. 8A, B). The release of Saracatinib or PF4618433 from the hydrogels had a similar effect to decrease the level of cells expressing ANP back to the level of control no ET-1. Like the treatment with PMA, when both inhibitors were released together (Dual), they statistically had a lower level of ANP than either Saracatinib or PF4618433 alone. The release of either Saracatinib or PF4618433 had a similar effect to decrease the cell area. However, when both inhibitors were combined (Dual), there was no enhanced effect as was seen for ANP. Western blot data demonstrates that 24 hr of ET-1 treatment activates Src, Pyk2, PKC, ERK, and JNK (Fig. 8C, D). Release of Saracatinib or PF4618433 form the hydrogels reduced the activity of Src, Pyk2, PKC, JNK and ERK all back to the control level. The combination of Saracatinib and PF4618433 (Dual) had no additional effect than PF4618433 alone.

Figure 8. Effect of Saracatinib and PF4618433 release from the hydrogel on ET-1 mediated hypertrophy.

A) NRVMs cultured in 12-well plate were pre-treated with Saracatinib (SA, 1 μg/well), PF4618433 (PF, 10 μg/well), or both (Dual, 3 hr) and then treated with ET-1 (100 nM, 24 hr). A representative fluorescent image is shown for each group (red, cardiac marker sarcomeric α-actinin; green, ANP; blue, DAPI). B) Cell surface area and percentage of total fluorescent intensity of ANP to DAPI were quantified using Image J software. C) Lysate from each group was Western blotted. Antibodies used are labeled on the left of each panel. D) Protein levels were quantified by analyzing scanned blots using Image J software, with normalization of protein expression to the control lane (value set arbitrarily as 1). Phosphorylated protein levels were normalized to total protein. Data are representative of three independent experiments. E) The apoptosis of NRVMs was detected using TUNEL assay. F) The TUNEL-positive cells were quantified using Image J software. Data are representative of three independent experiments (one-way ANOVA, n=3, *P < 0.05, **P < 0.01, ***P < 0.001).

The TUNEL assay detects DNA breakage by labeling the free 3ʹ-hydroxyl termini. Considering that genomic DNA breaks occur during apoptosis, we utilized TUNEL staining as a means to quantify apoptotic cell death induced by ET-1. Immunofluorescence results demonstrate that 24 hr of ET-1 treatment causes an increase in NRVMs undergoing apoptosis (Fig. 8E). Release of Saracatinib or PF4618433 from the hydrogels decreased the level of apoptosis, however not back to the control level. The combination of Saracatinib and PF4618433 (Dual) had a greater effect than either inhibitor alone as the amount of apoptosis returned to the control level without ET-1.

Phenylephrine, a selective α₁-adrenergic receptor (α₁-AR) agonist, is widely used as a stimulus for hypertrophy in cardiomyocytes by activating a number of the same signalling pathways as ET-1^{55, 56}. Using the TUNEL assay, immunofluorescence data demonstrates that 24 hr of phenylephrine treatment causes an increase in NRVMs undergoing apoptosis (Fig. 9A, B).

Figure 9. Effect of Saracatinib and PF4618433 release from the hydrogels on phenylephrine mediated hypertrophy.

A) NRVMs cultured in 12-well plate were pre-treated with Saracatinib (SA, 1 μg/well), PF4618433 (PF, 10 μg/well), or both (Dual, 3 hr) and then treated with phenylephrine (100 nM, 24 hr). The apoptosis of NRVMs was detected using TUNEL assay. The TUNEL-positive cells were quantified using Image J software (B). C) A representative fluorescence image of apoptosis-related protein (cleaved caspase 9) is shown for each group (red, cardiac marker sarcomeric α-actinin; green, cleaved caspase 9; blue, DAPI). D) Percentage of total fluorescence intensity of cleaved caspase 9 to DAPI was quantified using Image J software. (one-way ANOVA, *P < 0.05, **P < 0.01, ***P < 0.001).

Release of Saracatinib or PF4618433 from the hydrogels decreased the level of apoptosis, however the level did not return to the control level without phenylephrine. The combination of Saracatinib and PF4618433 (Dual) had a stronger effect than either inhibitor alone as the amount of apoptosis returned to the control level without phenylephrine.

To validate the apoptosis results, we conducted an additional test using another apoptosis marker, Caspase-9, which belongs to the cysteine aspartic acid protease (caspase) family⁵⁷. By employing an antibody that recognized activated Caspase-9 (cleaved version), we observed similar results in the immunofluorescence data as those obtained from the TUNEL assay. The release of Saracatinib or PF4618433 from the hydrogels resulted in a reduction in the level of apoptosis. However, this reduction did not reach the level observed in the control group without phenylephrine. The combined use of Saracatinib and PF4618433 (Dual) exhibited a stronger effect compared to either inhibitor used individually, returned to the control level without phenylephrine.

4. Discussion

MI is a leading cause of death globally and loss of cardiomyocytes plays a critical role in the pathogenesis of heart failure⁵⁸. With fewer cardiomyocytes, the heart is unable to sustain efficient contraction. Cardiomyocyte death leads to a series of pathological remodeling through the accumulation of fibrous tissue⁵⁹. The lack of electrical connection between the healthy myocardium and islands of intact cardiomyocytes in fibrotic matrices support asynchronous ventricular contraction, resulting in progressive functional decompensation⁶⁰. Several current heart failure therapies aim to promote cardiac repair by modifying immune cell infiltration and extracellular matrix composition, promoting angiogenesis and heart regeneration, and reducing cardiac hypertrophy^61–65.

An additional therapeutic to improve cardiac function would be to decrease the initial cardiomyocyte cell death in order to maintain proper impulse propagation. Implicated in this process is Cx43, as within 1 hour post-MI, expression is reduced and localization becomes lateralized at the infarct border zone; the same phenotype occurs in later in the disease during left ventricle hypertrophy^{8, 66 67}. Cx43 has been shown to be a promising therapeutic target in cardiovascular disease^{66, 68, 69}. For example, the peptide αCT1, which mimics the Cx43 residues involved in binding the ZO-1 PDZ2 domain, has anti-arrhythmic effects, as well as cardioprotective properties in mouse hearts subject to ischemia reperfusion injuries^{70, 71}. Inhibition of Pyk2 activity using the small molecule compound, PF4618433, reduced infarct size, maintained Cx43 at the intercalated disc in left ventricle hypertrophic myocytes, and improved cardiac function in an animal model²⁶. The Src inhibitor Saracatinib, raised Cx43 expression, improved conduction velocity, and lowered arrhythmia inducibility¹³. In this study, we identified mechanisms by which PF4618433 and Saracatinib elicit these positive responses post-MI. After PMA or ET-1 stimulation, PF4618433 and Saracatinib decreased expression of ANP. Plasma levels of ANP (and brain natriuretic peptide) and are elevated immediately post-MI as well as in heart failure and are predictors of ventricular dysfunction and mortality^72–74. We speculate that PF4618433 and Saracatinib are influencing the transcription factor GATA4, which is involved in ANP gene expression⁷⁵. This observation is consistent with the decreased cell area observed in the presence of PF4618433 and Saracatinib as ANP is a biomarker for hypertrophy⁷⁶.

PF4618433 had a greater effect than Saracatinib to decrease the activation of PKC, ERK, and JNK after PMA treatment, while these TKIs had a similar decreased effect decreases after ET-1 treatment. A number of studies have placed PKC activation upstream of Pyk2 activation^{77, 78}. However, one study identified that Pyk2 and PKCα can mutually influence each other to regulating Nanog and Oct4 transcription factors and the stemness phenotype in triple negative breast cancer⁷⁹. Conversely, Pyk2 is upstream of JNK and ERK activation and mediates its response through the adaptor proteins Crk and Grb2, respectively⁸⁰. Similar to Pyk2, Src directly activates PKCδ in many models of cardiac ischemia and hypertrophy and is upstream of JNK and ERK activation^81–83. Observations from the treatment with phenylephrine were similar to that of ET-1 (and PMA), in that there was an increase in apoptosis and cell area. These were reversed in the presence of PF4618433 or Saracatinib. The presence of both PF4618433 and Saracatinib further decreased the level of ANP and apoptosis than each inhibitor alone, suggesting a combinatorial approach may be most beneficial.

Based on these characteristics, PF4618433 and Saracatinib hold promise as potential therapeutics for enhancing cardiac function in humans. In support of this, we have presented evidence that Pyk2 is activated following MI in the human left ventricle²⁶. While PF4618433 was proven safe in an animal model of MI induced heart failure, PF4618433 was applied via intraperitoneally injection²⁶. Unfortunately, this route of administration may not be optimal for directly targeting the left ventricle at the necessary concentrations during the acute phase post-MI to prevent cardiomyocyte death, nor for the continuous application required over an extended period to support the hypertrophic heart²⁶. An injectable hydrogel-based delivery system holds promise as a more viable option. These hydrogels have been specifically designed for cardiac repair, with a focus on restoring electrical signaling propagation across the infarct region^{36, 84}. These hydrogels offer multiple functionalities, including mechanical support, improvement of the extracellular matrix, enhanced conductivity, scavenging of reactive oxygen species, and reduction in the expression of inflammatory cytokines^{35, 85, 86}. Here we provide the first step towards the proof of concept that injectable hydrogels, consisting of HA-SH and PM micelles, could also be used to release kinase inhibitors to decrease cell death by promoting intercellular communication. PM micelles can effectively incorporate both PF4618433 and Saracatinib and achieve sustained release, which cannot be accomplished by using Pluronic F-127 micelles. In addition, PM micelles act as a crosslinking agent for HA-SH, facilitating the formation of injectable hydrogels with shear-thinning properties. This hydrogel system can be delivered locally to the myocardium through transcatheters, allowing for targeted delivery of the dual TKIs. The next step in this study will be to revisit the MI model of heart failure were the hydrogel containing PF4618433 and Saracatinib is applied to the myocardium supplied by the left anterior descending artery and distal to this region that undergoes hypertrophy.

6. Conclusions

We developed an injectable hydrogel consisting of HA-SH and PM micelles, that can be injectable, incorporate PF4618433 and Saracatinib, and achieve sustained release. The significance of this study is that PF4618433 had the same effect released from the hydrogel as directly added to the media to reduce Src and Pyk2 activity as well as reduce tyrosine phosphorylation of Cx43 residues associated with a decrease in gap junction intercellular communication. Additionally, the presence of both PF4618433 and Saracatinib further decreased the level of ANP and apoptosis than each inhibitor alone, suggesting a combinatorial approach may be most beneficial. This delivery system holds promising for further intracardiac, pericardial or epicardial injection to treat MI.