Engineering targeted and responsive nanoparticles to modulate left ventricular remodeling after myocardial infarction

Introduction: An American will experience a myocardial infarction (MI) every 40 seconds, which often progresses to heart failure (HF). Sadly, current HF therapies are merely palliative, resulting in ≥ 37% mortality and hospitalization. Complicating this, patients presenting with HF are pathophysiologically diverse, which significantly limits the translation of “one-size-fits-all” therapies. Therapeutic approaches that target the molecular signatures of MI and release drugs in a manner proportional to local stimuli could revolutionize the field.

Fibroblast activation protein alpha (FAP) is a deleterious enzyme that is upregulated after MI and whose enzymatic activity drives the adverse extracellular matrix (ECM) remodeling that results in HF. Recent advances suggest that FAP is selectively expressed days after MI and that the inhibition of FAP prevents progression to HF. However, there are currently no drug therapies that target FAP for MI patients. Accordingly, we developed nanoparticles (NPs) that (i) encapsulate and (ii) locally target and deliver the highly specific FAP inhibitor, talabostat, and (iii) release talabostat in response to local FAP activity. We hypothesize that FAP inhibition by NPs loaded with talabostat will modulate remodeling after MI to reduce the onset of HF. This biomaterial strategy has many advantages in that the local delivery of talabostat will minimize off-target effects, the targeting and release is dependent on the patient FAP levels, and the intravenous administration of NPs is non-invasive.

Materials and

Methods: NPs were engineered to encapsulate talabostat by a charge-driven process (coacervation) between acrylated hyaluronic acid (AHA, synthesized through the reaction of HA with acrylic anhydride) and chitosan (Fig. 1a). Thereafter, NPs were crosslinked using FAP-sensitive peptides by a Michael addition reaction (Fig. 1a). FAP-sensitive peptides (GCNSGPSNCG), sourced from GenScript Biotech, were designed such that thiols in cysteines crosslink AHA, with GP being the FAP-sensitive sequence. To demonstrate that crosslinking NPs slows drug release, Cyanine 7 dye (Cy7) as a size-equivalent surrogate for talabostat was encapsulated to profile release using a Microplate reader. NPs were also designed to have an overall positive charge, such that negatively-charged, monoclonal IgG1 or FAP antibodies can be adsorbed to the surface. To evaluate NPs in vivo, an ischemia/reperfusion MI model was used in 8-week old Wistar rats (Charles River) by ligation of the left anterior descending artery for 30 minutes followed by reperfusion (IACUC approval number: 2871). Accordingly, the targeting efficiency of Cy7-containing NPs (with either IgG1 or FAP antibody) were evaluated 3 days after MI, when FAP levels have been shown to be elevated. NPs were intravenously administered by tail vein injections, followed by euthanasia after 1 hour for ex-vivo tissue imaging. For therapeutic studies, five groups including (i) sham, (ii) MI, and then MI with intravenously administered (after 3 and 7 days) (iii) empty FAP NPs; (iv) FAP NPs loaded with talabostat; and (v) the same amount of talabostat in saline. Functional and histological outcomes were measured after 4 weeks.

Results, Conclusions, and Discussions: Transmission electron microscopy (TEM) shows the formation of NPs ranging from 50-200 nm in diameter (Fig. 1a). Crosslinking NPs using a thiolated, FAP-sensitive peptide reduced Cy7 release compared to uncrosslinked NPs, and exposure of the crosslinked NPs to recombinant human FAP enzyme (to simulate varied FAP expression documented in patients) released Cy7 in a dose-dependent manner (Fig. 1b). Engineered NPs had an overall positive Zeta potential, such that adsorption of negatively-charged, monoclonal antibodies resulted in a dose-dependent decrease in NP net charge (data not shown). To verify adsorption of antibodies to NPs and whether FAP NPs are preferentially targeted to the heart over IgG NPs, adsorbed NPs were intravenously administered 3 days after MI, then euthanized 1 hour later (Fig. 1c). Preferential localization of FAP NPs to the infarcted heart suggested that FAP antibodies effectively targeted the NPs to the diseased region of the heart. To evaluate therapeutic efficacy, rats with MI were administered treatments on days 3 and 7, followed by echocardiography by day 30 after MI (Fig. 1d). Both ejection fraction and fractional shortening were reduced in the MI group compared to sham controls, consistent with heart failure progression (Fig. 1e). Remarkably, FAP NPs encapsulating talabostat not only outperformed empty NPs and talabostat in saline, it restored ejection fraction and fractional shortening to levels similar to sham controls (Fig. 1f-g).

Engineering FAP-responsive NPs that potently modulate HF after MI is a major advance for patient care because there is an unmet need for HF drug therapies. Excitingly, talabostat has been advanced to Phase II cancer clinical trials, making it a translatable candidate. As FAP is also implicated in the molecular progression of solid cancers affecting multiple organ-systems, the technology developed herein will have broad multi-disciplinary impact beyond cardiovascular disease and for the delivery of other small molecules. While different coacervate formulations, alone or within hydrogels, have been directly injected in the heart or tumors, coacervates have not been successfully engineered for intravenous administration for targeted drug delivery. Thus, this approach is highly innovative and impactful, given that injecting directly into organs (e.g. the heart) is challenging.

Acknowledgements (Optional): This study is funded by a grant from the National Institutes of Health (R01 HL167994) and a postdoctoral fellowship from the American Heart Association (https://doi.org/10.58275/AHA.24POST1187647.pc.gr.190896).