Targeted nanoscale therapeutics for myocardial infarction

Abstract

Nanoscale therapeutics have promise for the administration of therapeutic small molecules and biologics to the heart following myocardial infarction. Directed delivery to the infarcted region of the heart using minimally invasive routes is critical to this promise. In this review, we will discuss the advances and design considerations for two nanoscale therapeutics engineered to target the infarcted heart, nanoparticles and adeno-associated viruses.

Graphical Abstract

Introduction

In 2017, ischemic heart disease affected almost 126 million people globally and led to the death of nearly 9 million people, making it the leading cause of mortality in the world¹. While ischemic heart disease represents various clinical conditions, one of its primary manifestations is myocardial infarction (MI). Though techniques have been developed that aid in revascularization and mitigation of further tissue damage after an MI, these practices do not promote cardiac repair. Biomaterial platforms and gene therapy have emerged as exciting regenerative medicine methods for treating the heart post-MI. The administration of therapeutics to the heart can be achieved in various ways. Intramyocardial injection can be performed by surgically opening the chest but can also be done via a transendocardial injection catheter. This method is not a good option for administration to the heart during the acute phase of MI as the heart is more prone to arrhythmias and the wall is at higher risk of rupture^2–6. Another option is intracoronary infusion, which is performed via catheterization. While these catheter-based techniques are less invasive, intracoronary infusion allows for less control of material dissemination and transendocardial injections are technically challenging⁷.

The current clinical standard of care for patients during the acute phase of MI includes angioplasty, stenting, and administration of pharmaceuticals⁸. While these methods are useful in initiating reperfusion and reducing oxygen demand in the heart, they are not able to address imbalances within the heart that arise due to ischemic damage. There is a need for therapies that can be safely administered to patients while also delivering novel therapeutic payloads that are capable of addressing physiological issues that are present on the cellular level. Current bioengineering-based strategies in the form of cardiac patches or injectable biomaterials^{9, 10} have shown therapeutic efficacy but are more suited towards patients in the sub-acute and chronic phases of MI as they require surgical or transendocardial delivery.

Ideally, administration of therapeutics post-MI should be less invasive or non-invasive, as accomplished through methods like intravenous (IV) injection or oral ingestion. However, systemic administration of therapeutics often faces limitations in efficacy due to off-target effects and short circulation times. To further complicate these issues, many of the therapeutics are not water-soluble, limiting their physiological application in vivo^{11, 12}. These issues go beyond small molecules. Cardiovascular medicine has become increasingly focused on the use of biologics (miRNAs, therapeutic transgenes, siRNAs, etc)^13–15. These payloads face their own and parallel issues with the aforementioned small molecule drugs—limited efficacy with off-target effects and short circulation times. To overcome these challenges, drugs and biologics can be packaged into various delivery platforms that have been modified to accumulate at the site of interest, decreasing the necessary dose and mitigating off-target effects.

Properly designed nanoparticles and adeno-associated viruses (AAVs) are both nanoscale platforms that have been shown to improve the delivery of various therapeutic payloads to the heart after MI. During the ischemic period, there is a decrease in flow-related shear stress on luminal endothelial cells—resulting in abnormal endothelial-dependent relaxation and enhanced permeability^16–19. Nanoparticle delivery to the heart relies on this permeability effect for extravasation into infarcted tissue, but retention must be improved through the use of a specific targeting mechanism. While it is yet to be completely understood how AAVs selectively target the heart, certain serotypes appear to transduce cardiac tissue more effectively²⁰. With the use of enhanced targeting, AAV transduction can be more precisely controlled and directed to the damaged heart. In this review, we will discuss the current state of the field and utility of both nanoparticles and AAVs as nanoscale therapeutics that can be administered minimally invasively to treat MI (Fig. 1). Specifically, we will focus on the design elements that are taken into consideration when developing these nanoscale therapies.

Figure 1: Nanoscale therapeutics used for cardiac applications.

Both nanoparticles and AAVs of various compositions and serotypes can be modified for improved delivery to the heart post-MI.

Nanoparticles that target the infarcted region of the heart

Nanoparticle therapeutics have emerged as a novel way to revitalize therapeutics that have limited in vivo application due to poor solubility or deleterious side effects. Incorporating therapeutic payloads into a nanoparticle system has been shown to increase circulation time and improve therapeutic retention in preclinical models^{21, 22}. While the preclinical research and clinical trial landscape for nanoparticle therapeutics has been overwhelmingly ruled by cancer applications²³, the field is becoming more diverse. Nanomedicine for cardiovascular applications is newer than in oncology but is seeing rapid development and growth²⁴. Since 2016, two nanoparticle platforms have been approved by the FDA for imaging applications post-MI²⁵ but there have not been any FDA approved nanoparticle platforms for treatment of the infarcted heart. Nanomedicine was initially seen as a way to increase drug circulation time and reduce systemic toxicity. However, as the field has advanced, materials have become increasingly complex with more functionalization to aid in site-specific delivery. Though this increased in design complexity has led to a slowdown in clinical progress²⁴, it will allow for the development of more effective nanoparticle platforms that can traffic through the body safely and accumulate in the infarcted region for optimal efficacy. Here we will discuss nanoparticle targeting methods into two main categories, designs that take advantage of the immune response and those that are created to interact with cardiac-specific features (Table 1).

Table 1:

Targeted nanoparticle therapies for MI^a

Authors	Material	Particle Size	Animal Model	Target	Dosage Timeline	Payload	Outcomes	Ref.
Scott et al. (2009)	Liposome conjugated with anti-P-selectin antibodies	180 ± 35 nm	Myocardial infarction (permanent occlusion, rat)	P-selectin	Immediately after injury	VEGF	Targeted VEGF resulted in improved LV function and increased number of anatomical and perfused vessels.	30
Nguyen et al. (2015)	Polynorbornene backbone, peptide sequence conjugated	15–20 nm	Myocardial infarction (ischemia-reperfusion, rat)	MMPs	1-day post injury	N/A	Study was done as proof of concept showing nanoparticle localization in the infarcted region of the heart	31
Dong et al. (2017)	RGD modified, PEGylated solid lipid nanoparticles	110.5 nm	Myocardial infarction (permanent occlusion, rat)	αvβ3 integrin	Immediately post-injury	Puerarin	RGD-modified nanoparticles demonstrate selective accumulation in the heart compared to controls, longer circulation half-life, and significant reduction in infarct area,	33
Yao et al. (2020)	Mesoporous silica nanoparticles coated with mesenchymal stem cell membranes	~110 nm	Myocardial infarction (ischemia-reperfusion, mouse)	ICAM-1	Every day for 3 days post-injury	miR-21	Membrane-coated NPs resulted in increased left ventricular ejection fraction and fractional shortening as well as reduced scar size and cellular apoptosis compared to a non-membrane coated control.	41
Cheng et al. (2014)	Iron core nanoparticles conjugated to anti-CD34 and anti-MLC antibodies	95.7 ± 14.5nm	Myocardial infarction (ischemia- reperfusion, rat)	Endogenous bone marrow derived stem cells (CD34+) and injured cardiomyocytes (MLC+)	10 minutes post-injury	N/A	Antibody targeted nanoparticles resulted in an increase in the amount of viable tissue and improved left ventricular ejection fraction. These trends were further improved with the application of a magnetic field to enhance nanoparticle targeting.	44
Vandergriff et al. (2018)	Cardiac stem cell-derived exosomes conjugated to cardiac homing peptide	~95 nm	Myocardial infarction (ischemia-reperfusion, rat)	Hypothesized to be alpha-B crystalline	1-day post injury	Cardiac stem cell-derived exosomes	Targeted exosomes increased ejection fraction, viable tissue, and cardiomyocyte proliferation while also decreasing infarct size.	47
Zhang et al. (2018)	Lipid-polymeric nanoparticles modified with TPP and TPGS	140 nm	Myocardial infarction (permanent occlusion, rat)	Mitochondria	Every other day for 2 weeks post-injury	Tanoshine II-A	TPP/TPGS targeted nanoparticles showed higher drug accumulation in the heart and reduced infarct size	49
Zhang et al. (2019)	PLGA-PEG-SS31	54 nm	Myocardial infarction (ischemia-reperfusion, rat)	Mitochondria	5 minutes before reperfusion	Cyclosporin A	Significantly increased localization in the heart compared to a non-targeted control, decreased cardiomyocyte apoptosis and infarction area.	51
Xue et al. (2018)	PEGylated dendrigraft poly-L-lysine dendrimers modified with an AT1 targeting peptide	~50 nm	Myocardial infarction (permanent occlusion, mouse)	AT1 receptor	1-day post injury	Anti-microRNA-1	AT1 modified dendrimers significantly reduced miR-1 expression, cellular apoptosis, and infarct size.	53

Targeted to immune response

Targeted to cardiac features

^aVEGF: vascular endothelial growth factor, PEG: polyethylene glycol, TPP: triphenylphosphonium, TPGS: D-α-tocopheryl polyethylene glycol 1000 succinate, AT₁: angiotensin II type 1

While nanoparticles have increased circulation time compared to freely administered small molecules and biologics, they are still subject to off target accumulation, specifically in satellite organs such as the kidneys and liver; organs designed to clear such materials, making this task inherently difficult²⁶. Generally, their potential systemic effects remain a limiting factor in the clinical translation of nanoparticles. However, with thoughtful material design, it is possible to improve site-specific delivery as is seen in several cases we will highlight.

To design vehicles that target the heart, it is crucial to understand the underlying biology during and following MI (Fig. 2). The upregulation of certain enzymes and cytokines as well as the influx of immune cells into the infarcted region of the heart represent unique biological signatures that can be leveraged and utilized as markers for targeting. In addition, production of reactive oxygen species (ROS) has been shown to have a significant role in the pathogenesis of vascular damage¹⁶ and leukocyte chemotaxis²⁷ to the area of injury. Neutrophils are the first inflammatory cell to infiltrate into the infarct with macrophages of various subsets following at a later stage²⁸. Matrix metalloproteinase (MMP) production is upregulated in the heart within 10 minutes after occlusion²⁸. Out of the many types of MMPs, MMP-2 and MMP-9 are specifically upregulated during the acute phase of MI and remain at high levels of expression for months²⁹. Accompanying this intense inflammatory response, there is a decrease in the pH of the heart following MI by at least a unit, from 7.4 to ~6.4²⁸. These alterations in the heart microenvironment can serve as inspiration for targeted design of therapeutics.

Figure 2: The inflammatory response during the acute phase of MI.

Hallmarks of this response include recruitment of neutrophils to the infarct via adhesion molecules P selectin and vascular cell adhesion protein 1 (VCAM-1). Additionally, matrix metalloproteinases (MMP) 2 and 9 are upregulated and begin to degrade the collagen in the heart. Apoptotic cardiomyocytes release inflammatory cytokines such as IL-6 that activate inflammatory cascades in fibroblasts, immune cells, and vascular cells.

Nanoparticles that target hallmarks of inflammation and wound healing

To determine features of the inflammatory response, we will specifically focus on physiological abnormalities in heart that are a response to hypoxia, such as increased protease activity, acidity, and vasculature damage. The most common way of imparting a targeting mechanism to a nanoparticle is via surface functionalization. Scott et al. utilized this design method by decorating the surface of immunoliposomes (180 ± 35 nm) with antibodies that bind P-selectin. P-selectin, a cell adhesion molecule, exhibits increased expression in the vasculature of the infarcted region and has been used to localize the delivery of VEGF post-MI³⁰. Using a rat model of MI, animals were injected with their treatment via the tail vein immediately after induction of the ischemic injury. This targeted method resulted in a significant increase in fractional shortening and improved systolic function compared to systemic free VEGF treatment, demonstrating the efficacious advantages of utilizing targeted therapies. In addition to functional improvement, a significant increase in number of vessels and number of perfused vessels in the heart compared to an untreated control was noted after 4 weeks. However, the vessels in the free VEGF group were not quantified.

Another robust example of surface functionalization is through the incorporation of an enzyme-responsive peptide sequence to the polymer backbone of the nanoparticle. Nguyen et al. designed peptide-polymer amphiphiles composed of a hydrophobic polymer backbone (polynorbornene) followed by the conjugation of a hydrophilic peptide sequence (GPLGLAGGWGERDGS) that is recognized and cleaved by MMP-2 and MMP-9³¹. Once cleaved, the hydrophobic core of the nanoparticle is exposed, causing aggregation from the nanoscale (15–20 nm) to the micron-scale. This morphological switch enhances nanoparticle retention in the infarct. Nanoparticles administered via IV injection one day post-MI in a rat model showed selective aggregation of particles in the heart only when the MMP-cleavable peptide is conjugated to the polymer backbone (Fig. 3). This was validated through the use of a non-responsive control peptide sequence that did not lead to nanoparticle aggregation after exposure to MMPs. While methods such as these for targeting areas of inflammation do lead to MI targeting, they are still subject to off-target accumulation due to the fact that they are not cardiac-specific and there can be other areas of active inflammation in the body.

Figure 3: Enzyme responsive nanoparticles accumulate in the infarcted region of the heart.

(A) H&E images of hearts at various time points. (B) α-actinin (green) stained hearts following IV nanoparticle injection shows accumulation and retention of rhodamine-labelled particles (red) in infarcted areas (white boxed region) up to 28 days post-injection compared to a non-responsive (NR) nanoparticle control (bottom row). (C) Magnified boxed region from column B. Reproduced from ref. 31 with permission from WILEY - V C H VERLAGGMBH & CO. KGAA., copyright 2015.

Targeting the α_vβ₃ integrin, which is upregulated in endothelial cells during angiogenesis³², is another technique that has been used to target the heart. The cyclic peptide sequence arginyl-glycyl-aspartic acid (RGD) was used to improve cell anchoring, specifically in cardiac applications. Dong et al. designed solid lipid nanoparticles (SLN)³³ covered in a layer of polyethylene glycol (PEG), a surface modification known to improve the circulation time³⁴, and modified the surface of the particle with the RGD peptide sequence to enable improved cell-surface binding (110.5 nm). The small molecule puerarin (PUE), a reactive oxygen species scavenger^{35, 36}, was encapsulated within the core of the nanoparticle. Following induction of MI via ligation of the left descending coronary artery in rats, the therapeutic efficacy and biodistribution of the RGD/PEG-PUE-SLN was compared to IV administration of saline, free drug, drug encapsulated in plain SLN, and drug encapsulated in PEGylated SLN administered immediately after injury. The RGD/PEG-PUE-SLN treated animals demonstrated increased nanoparticle accumulation in the heart compared to satellite organs and significantly decreased infarct size compared to all other treatment groups. More recently, RGD has also been used to successfully deliver miR-133 to the heart using polyethylene glycol-polylactic acid nanoparticles. It was shown to increase miR-133 levels in the infarcted area and improve heart function³⁷, demonstrating the versatility and function of RGD in targeting the heart acutely post-MI. However, RGD can also be used to target other regions in the body where there is active wound healing or angiogenesis and is therefore not only cardiac-specific^38–40.

Stem cell membranes have been used as another method for targeting the infarcted heart⁴¹. Yao et al. designed mesoporous silica nanoparticles camouflaged with mesenchymal stem cell membranes to deliver microRNA-21, which is involved in cardiogenesis and cardiac regeneration (~110 nm). These stem cell membranes cloak the nanocomplexes from immune clearance while also imparting exosome-like qualities to this platform. Specific integrins on the membranes of mesenchymal stem cells can bind to the overexpressed intercellular adhesion molecule-1 (ICAM-1) on endothelial cells⁴² and injured cardiomyocytes⁴³. One day post-MI in a mouse model, these nanocomplexes were injected via the tail vein every day for three days. Whole organ imaging post-administration demonstrated strong accumulation of membrane-coated NPs in the heart (Fig. 4). Functionally, administration of membrane-modified nanoparticles delivering microRNA-21 resulted in a significant increase in ejection fraction and fractional shortening as well as a reduction in myocardial scar size and cellular apoptosis compared to a non-targeted control. There was also a significant increase in Ki67 positive cardiomyocytes in the injured region of MI hearts in response to membrane-cloaked particles, suggesting cardiomyocyte proliferation. While this method shows promise in targeting the infarct, it relies upon the upregulation of ICAM-1, which is a consequence of inflammation and is not solely cardiac-specific. To improve targeting capability, one can consider designing nanoparticle platforms that interact with biological signatures that are unique to the heart.

Figure 4: Improved localization of mesoporous silica nanoparticles (MSN) in the heart with the addition of cell membranes (CM).

(A) Cy-3 labeled miRNA delivered using CM-modified MSNs resulted in higher accumulation in the heart compared to an MSN alone and a scrambled miR sequence (miR-sc). (B) CM-miR-MSN demonstrated enhanced blood circulation compared to controls. Reproduced from ref. 41 with permission from Elsevier, copyright 2020.

Targeting protein and cellular features within the infarct

To expand on inflammatory response targeting mechanisms that have been previously discussed, there is also a growing list of ways in which nanoparticle platforms are being designed to localize to features that are present in the heart. By incorporating multiple targeting moieties, Cheng et al. designed a “nanomatchmaker” comprised of an iron core that is capable of linking therapeutic cells to injured cells through the use of antibody conjugation to the nanoparticle surface (95.7 ± 14.5nm)⁴⁴. This platform (called MagBICE₂) uses anti-CD34 and anti-myosin light chain (MLC) antibodies to tether endogenous bone marrow derived stem cells (CD34) to injured cardiomyocytes (expressing MLC). Nanoparticles were IV injected 10 minutes after injury in a rat model of MI. To enhance nanoparticle targeting to the heart, some animals were subjected to a magnetic field that interacted with the nanoparticle’s iron core. Animals were then harvested after 4 weeks to evaluate therapeutic efficacy. MagBICE₂ demonstrated significantly higher iron intensity and CD34⁺ cells in the injured myocardium compared to animals treated with an iron nanoparticle control, Feraheme (FH). In addition, nanoparticle treatment increased the amount of viable tissue, as evaluated via Masson’s trichrome, and improved left ventricular ejection fraction. These trends were further improved when a magnetic field was applied to enhance nanoparticle localization. Magnetic field targeting further enhanced cardiac functional improvement over nanoparticle administration alone, demonstrating the utility of the iron core design. Overall, this platform is interesting in that the targeting moiety also acts as a mechanism for therapy, resulting in a simple but therapeutically efficient design. One can also see how this platform could be easily adapted to bear other antibodies, making it more versatile.

In vivo phage display is a technique that has emerged as a method for discovering tissue-specific targeting methods⁴⁵. A phage library is injected in vivo and allowed to circulate for a certain amount of time at which point, the tissue of interest is collected, and the phages are isolated and amplified to be injected again. This process is repeated several times to isolate the phages that are known to localize to the site of interest. These phages can then be sequenced and used as targeting moieties. Kanki et al. completed this process for the ischemic left ventricle in a rat model of ischemia reperfusion with a Ph.D.-C7C phage library injected 10 minutes post-injury⁴⁶. Three peptide motifs that demonstrated preferential binding to the ischemic myocardium compared to uninjured animals were identified. By conjugating these sequences to the recombinant fluorescent protein SUMO-mCherry, they assessed the affinity of each sequence to the ischemic myocardium and found that the sequence CSTSMLKAC localized best to the heart after MI. This sequence bears similarity to titin, a cytoskeletal protein that binds to alpha-B crystalline, thereby protecting titin from ischemic damage. Though the exact mechanism by which this peptide sequence targets is unknown, it is postulated that it is through interactions with alpha-B crystalline. This “cardiac homing peptide” sequence has been used to aid in targeting IV injection of exosomes to the infarcted heart 24 hours post-MI⁴⁷. Both in vitro and in vivo, conjugation of the peptide sequence to exosomes was achieved through the use of a DOPE-NHS linker. When comparing the efficacy of exosomes conjugated to the cardiac homing peptide versus a scrambled peptide sequence, there was a significantly higher ejection fraction and amount of viable myocardium in addition to decreased infarct size (Fig. 5). Not only did the addition of a targeting peptide increase localization to the area of interest, it also allowed for a significant increase in cellular uptake of the exosomes, as demonstrated in vitro. The exploration of peptide sequences that target the heart has also been done by another group using an M13 phage library⁴⁸. Using the same biopanning method of multiple rounds of injection, isolation, and amplification, they discovered the peptide sequence, APWHLSSQYSRT. Similar to the cardiac homing peptide, this sequence also shows improved localization and internalization in the heart in vivo and could also be used to aid in targeted therapeutic delivery to the heart.

Figure 5: Exosomes (XO) modified with the cardiac homing peptide (CHP) improve heart function and reduce scar size.

(A) CHP-XO treatment leads to ejection fraction improvement over time and (B) significant improvement compared to PBS and an exosome conjugated to a scrambled peptide sequence control (SCR-XO). (C) Representative M mode echocardiography images. (D) Masson’s trichrome stained hearts show that CHP-XO treatment (E) decreases infarct area and (F) increases viable tissue. Reproduced from ref. 47 with permission from Ivyspring International Publisher, copyright 2018.

During ischemic injury, the mitochondrial membrane is damaged in cardiomyocytes. This leads to mitochondrial dysfunction and, ultimately, cardiomyocyte death that is mediated by the opening of the mitochondrial pore complex in the inner mitochondrial membrane. Zhang et al.⁴⁹ developed lipid-polymeric nanoparticles that enable mitochondrial-targeted delivery of tanoshine IIA, a cardioprotective medicinal herb⁵⁰. These lipid nanoparticles were modified with the triphenylphosphonium (TPP) cation to target mitochondria and D-α-tocopheryl polyethylene glycol 1000 succinate (TPGS) to prolong blood circulation and enhance cellular uptake (~140nm in diameter). Following permanent occlusion of proximal left coronary artery, rats were injected via the tail vein every other day for 14 days. Targeted nanoparticles with the TPP cation showed significantly higher drug accumulation and reduced infarct size in the heart compared to non-targeted nanoparticle and free drug controls.

In another attempt to leverage the unique phenotype of mitochondria post-MI, Zhang et al. developed polyethylene glycol nanoparticles (54 nm) that had been modified with the Szeto-Schiller peptide sequence (SS31) to enable targeted delivery of Cyclosporin A (CsA), an immunosuppressive drug, to the heart after MI when administered via IV injection 5 minutes prior to reperfusion⁵¹. The SS31 sequence specifically localizes to the inner mitochondrial membrane via interacting with cardiolipin. In vitro, increased cellular uptake of SS31-modified nanoparticles was demonstrated with hypoxia reoxygenated H9c2 cells. The damaged mitochondrial membrane enabled concentrated accumulation of SS31 particles and delivery of the therapeutic payload. In a rat model of acute MI, it was demonstrated that administration of nanoparticles loaded with CsA resulted in significantly less cellular apoptosis and reduced infarct size compared to CsA alone. This is an excellent example of how nanoparticles could be used to improve the bioavailability and efficacy of small molecule drugs. While this method of targeting works, it is not specific to only the heart; cardiolipin is found in the membranes of all mitochondria. Though mitochondrial damage is strongly present in the heart post-MI, it may not be exclusive to only the heart and platforms that are designed to target this abnormality could still be subject to off target effects.

Targeting the angiotensin II type 1 (AT₁) receptor following MI has previously been shown to be a successful technique for directing nanoparticle delivery to the heart, as it is upregulated in the infarcted following MI⁵². By attaching an AT₁ targeting peptide to PEGylated dendrigraft poly-_L-lysine dendrimers (~50 nm), Xue et al. were able to deliver the anti-miR-1 antisense oligonucleotide to inhibit cardiomyocyte apoptosis⁵³. One day post-MI, mice were injected with the dendrimer “nanovector” therapeutic or various controls via the tail vein. The nanovectors bearing the AT₁ targeting peptide demonstrated strong accumulation in the heart starting as soon as one-hour post-MI, as well as significantly reduced miR-1 expression compared to a non-targeted control. In addition, at 7 days post-MI there was a significant reduction in TUNEL positive cells, indicating a reduction in cellular apoptosis, as well as a reduction in infarct size in response to treatment with the AT₁ receptor-targeted dendrimers. Having been used as a target for multiple nanoscale platforms, the AT₁ receptor appears to be a promising target for cardiovascular applications.

Viral vectors for therapeutic delivery after MI

In addition to nanoparticles, viral vectors have also proven themselves as promising nanoscale vehicles for delivering therapeutic transgenes post-MI. AAVs are ~25 nm, making them similar in size to other nanoparticle platforms⁵⁴. Much like nanoparticles, they are versatile and can be engineered to improve localized delivery to the heart. In this next portion of the review, we will focus on the application and design of viral vectors for MI. AAVs are single-stranded DNA viruses that mainly remain episomal and do not integrate into the host genome⁵⁵ and have been shown to induce long lasting transduction in the heart⁵⁶. From a clinical standpoint, AAVs have the most translational potential; other viral vectors such as lentiviruses and adenoviruses illicit a stronger immune response and their delivered genetic material is incorporated into the host’s genome⁵⁵. In comparison, AAVs are replication defective while still maintaining their infectious qualities. They are able to deliver genes to post-mitotic cells and illicit a less aggressive immune response⁵⁷. For this reason, we will focus on AAVs in this review.

AAVs are members of the parvoviridae family. More specifically they belong to a subset of this family known as the dependovirus or “gutless” viruses that were discovered in the 1960s as a contaminant of cells that had been treated with adenoviruses⁵⁷. They are unable to replicate without the presence of a helper virus. This is typically either an adenovirus or herpes virus. AAVs are less immunogenic, though immunosuppressants are still used to mitigate adverse immune response reactions. The main mechanism by which AAVs trigger the immune response is via neutralizing antibodies that can be pre-existing or developed in response to AAV treatment, and cytotoxic T-cell effectors that could hinder long term efficacy of gene therapy⁵⁸. Various AAV serotypes have also demonstrated specificity to certain tissues—leading to better localization and transduction efficiency when the correct serotype is used for its corresponding tissue type⁵⁹. It has been previously determined that AAV1 and AAV6 are suitable for cardiac gene transfer when administered via intramyocardial, intrapericardial, or intracoronary administration whereas AAV8 and AAV9 are have been shown to achieve strong cardiac transduction when injected IV⁶⁰. The exact mechanism for cardiac tropism in certain AAV serotypes is not yet understood⁶⁰. Overall, there is a need for improving localization, reducing off-target transduction, and mitigating the immune response of these nanoscale therapeutics. Here we will discuss various targeting techniques for improving AAV therapies in the heart (Table 2).

Table 2:

Cardiac applications of AAVs^a

Authors	AAV serotype	Animal Model	Method of Administration	AAV modifications	Dosage Timeline	Transgene	Outcomes	Ref.
Muller et al. (2006)	“De-targeted” AAV-2	Healthy mice	IV injection	Cardiac-specific promoter MLC1.5 fused to CMV, mutations of heparin binding motifs	One-time injection	Luciferase	De-targeted AAV-2 lead to a 100-fold increase in transduction in the heart compared to the liver.	72
Pacak et al. (2008)	AAV-9	Newborn mice	IV injection	Cardiac-specific promoters: CMV, Des, α-MHC, MLV-2, and cTnC	One-time injection	LacZ	Use of tissue specific promoters mitigated off-target transduction. Highest levels of expression were observed in the CMV, Des, and α-MHC promoter groups, but α-MHC was the most cardiac-specific.	71
Prasad et al. (2011)	AAV 1, 2, 6, 8, and 9 (transduction optimization), AAV-9 (efficacy study)	Myocardial infarction (ischemia-reperfusion, mouse)	Direct intramyocardial injection	Cardiac-specific promoter: Cardiac troponin promoter	4 weeks prior to injury	EcSOD	AAV-9 demonstrated the most efficient transduction in the adult mice and reduced infarct size compared to an AAV-9/GFP control.	62
Greenberg et al. (2016)	AAV-1	Human patients with heart failure	Intracoronary infusion	N/A	One-time infusion (exact scenario depended on patient)	SERCA2a	AAV-1/SERCA2a did not reduce recurrent events related to heart failure or mortality compared to a placebo.	69
György et al. (2017)	AAV-9	Healthy mice	IV injection	Extracellular vesicles associated with AAV capsid (e-AAV)	One-time injection	Luciferase or GFP	Ex-AAVs were 136 times more resistant to neutralizing antibodies and lead to increased transduction while also lessening the number of viral particles necessary for administration.	84
Shin et al. (2018)	TANNylated AAV-9	Myocardial infarction (ischemia-reperfusion, rat)	IV injection	Tannic acid conjugation	Immediately post-injury	GFP	Tannic acid binds to the extracellular matrix and has increased uptake in myoblasts. In vivo results demonstrate significantly higher GFP expression in the heart compared to non-TANNylated control AAV-9.	74

^aAAV: adeno associated virus, MLC: myosin light chain, CMV: cytomegalovirus immediate-early gene promoter, α-MHC: alpha myosin heavy chain, EcSOD: extracellular superoxide dismutase, SERCA2a: sarcoplasmic/endoplasmic reticulum Ca²⁺ ATP

AAVs in the heart

Though many types of viral vectors have been explored as gene therapy candidates in the heart, AAVs are the most promising⁶¹. Compared to adenoviruses and lentiviruses where the bulk of preclinical publications came out in the early 1990–2000s, AAV application for MI is actively being pursued. Prasad et al. tested various AAV serotypes (1, 2, 6, 8, and 9) expressing luciferase to assess their ability to successfully transduce the left ventricle of mice⁶². Through this study they were able to demonstrate that direct intramyocardial injection of AAV-9 resulted in the highest expression of luciferase in the heart compared to the other serotypes tested at all time points (4, 7, 14, 28, and 42 days post-injection). With this information, they then went on to demonstrate in a mouse model of MI that administering AAV-9 expressing extracellular superoxide dismutase (EcSOD) was successful in decreasing infarct size by 50%. However, it should be noted that AAV administration occurred 4 weeks prior to induction of the MI. EcSOD has previously been demonstrated to protect the heart against MI—though efficacy has been unreliable from study to study⁶³. By packaging this enzyme into an AAV, the delivery is perhaps more stable and robust.

In another study, Gabisonia et al. used AAV-6 to deliver miR-199a 10 minutes post-injury and showed increased cardiac repair when injected directly along the infarct’s border zone in a porcine model of acute MI⁶⁴. This microRNA was previously demonstrated to stimulate rodent cardiomyocytes to re-enter the cell cycle⁶⁵. In addition to robust expression of miR-199a compared to controls, animals treated with AAV6-miR-199a also saw significantly increased cardiomyocyte proliferation as well as decreased fibrosis and infarct size one-month post-injection. Adding to this, ejection fraction was significantly increased, and end-systolic volume significantly decreased in AAV6-miR-199a animals. This is particularly noteworthy as decreased end-systolic volume is one of the best predictors of survival post-MI⁶⁶. One caveat to the success of this study is the fact that around 2 months after AAV injection, there was an increase in spontaneous death in the animals that received AAV6-miR-199a. In their analysis, it is acknowledged that this was most likely due to the long term, uncontrolled expression of miR-199a, which in high amounts can cause sudden cardiac death. This complication could potentially be resolved by optimizing AAV dosing.

AAVs in the clinic

Though AAVs have had much preclinical success in large animal models^{67, 68}, they have faced difficulty in translation due to the prevalence of neutralizing antibodies in patients. For example, the CUPID2 trial (NCT01643330) delivered AAV1 with the transgene SERCA2 to chronic heart failure patients via intracoronary infusion, but the study failed as it was discovered that there were no benefits to the therapy versus a placebo⁶⁹. This unsuccessful application of AAVs in the heart can be explained by a few elements within the study that were not optimally designed. As previously discussed, it is now known that certain serotypes, such as AAV8 and AAV9, have demonstrated improved transduction of cardiac tissue when administered via intracoronary infusion. Patients also tend to have lower levels of neutralizing antibodies to these serotypes, making them more amenable for successful transduction⁶⁰. Moving forward, it could be beneficial to engineer and apply viral vectors much like we do for nanomedicine. In addition, further research has illuminated other relevant genes of interest that are dysregulated after MI such as small ubiquitin-like modifier 1, S100A1, protein phosphatase 1 inhibitor-1, and G protein-coupled receptor kinase-2⁷⁰ By designing AAVs to more effectively traffic through the body to the heart while taking advantage of their natural infectivity, a more efficient gene therapy platform could be discovered. Many groups have begun to employ more complex designs of AAVs and seen improved cardiac transduction and reduced off-target accumulation.

AAVs designed for targeted delivery to the heart

Much like nanoparticles, AAVs have been engineered to preferentially target the heart or transduce cardiac tissue. Increasing accumulation in the site of interest will allow for lower doses, potentially mitigating the immune response and off-target transduction. Previous studies have shown improved transduction in cardiac tissue under the use of tissue-specific promoters. Pacak et al. surveyed five promoters for their activity in vivo: cytomegalovirus immediate-early gene promoter (CMV), human desmin (Des), human alpha-myosin heavy chain (α-MHC), rat myosin light chain 2 (MLC-2), and human cardiac troponin C (cTnC)⁷¹. Each promoter was attached to LacZ for visualization in tissue sections, flanked by terminal repeats of AAV2, and packaged into AAV9 capsids. Four weeks after IV injection into healthy newborn mice, it was observed that the highest levels of expression were observed in animals who received CMV, Des, and α-MHC. Biodistribution analysis revealed that while Des and CMV had overall higher transduction efficiency, α-MHC was the most cardiac specific. Overall, the use of tissue specific promoter sequences showed preferential transduction in cardiac tissue compared to other satellite organs (brain, skeletal muscle, lung, liver, spleen, kidney, and small intestine)—as confirmed via qRT-PCR. This study demonstrates the ability to design control mechanisms into AAVs, allowing for more desired transgene activity in the heart following systemic injection.

By combining transcriptional and cell surface targeting, Müller et al. aimed to increase AAV specificity and efficiency. They made use of the tissue-specific promoter 1.5kb cardiac myosin light chain (MLC1.5) fused to the cytomegalovirus immediate early enhancer (CMV) with luciferase as the reporter gene⁷². In addition to this, they also investigated “de-targeting” of AAV2 to the liver by inducing mutations at the heparin binding domains known to be a prerequisite for transduction in the liver by engineering AAV-2 capsids with a double mutation in the heparin-binding motif at R484 and R585⁷³. IV injection of this mutant AAV-2 in healthy mice lead to a 100-fold increase in cardiac transduction over hepatic transduction. In doing this, they were also able to further eliminate heparin binding motifs as a key player in cardiac tissue transduction. This multi-level approach to mitigating off-target transduction is unique in that it has both a mechanism for increasing favorable cell interactions but also engineers a method for reducing unwanted transduction in the liver, a satellite organ that is known for accumulation of systemically injected therapeutics. They also compared the transduction of cardiac tissue using AAV-6. Though AAV-6 demonstrated higher overall cardiac gene transfer, there was an increase in non-specific hepatic interactions compared to the mutant AAV-2. Overall, the data suggests that the combination of transcriptional (CMV_enhMLC1.5) and transductional (heparin-binding motif mutations) targeting modifications resulted in a significant increase in gene transfer to the heart. This combinatorial technique has potential for improving the efficacy of systemically injected AAVs.

Beyond tissue specific promoters and capsid engineering, Shin et al. investigated the benefits of conjugating tannic acid to AAVs for improved delivery to the heart⁷⁴. Tannic acid is a flavonoid derived from plants that binds to components of the extracellular matrix, such as elastin and collagen, without interacting with the glycocalyx. In a rat model of myocardial ischemia-reperfusion injury, it was demonstrated that TANNylated proteins were able to pass through the endothelium and bind to the extracellular matrix within the myocardium when injected immediately post-MI. This phenomenon relies on the enhanced permeability and retention effect that is present during the acute phase of MI. In this study, tannic acid was conjugated to GFP, AAV-9, and bFGF and delivered via IV injection to the heart immediately after ischemic injury. The mechanism by which tannic acid is able to bind to the extracellular matrix (ECM) was explored via turbidimetric assays to demonstrate tannic acid’s affinity for collagen and elastin, major components of the ECM, as compared to heparan sulfate or hyaluronic acid, components of the glycocalyx. This association with the ECM allows for increased cellular uptake of TANNylated proteins by myoblasts. As opposed to previously discussed papers, this study demonstrates not only localization, but also transduction efficiency. In vivo results showed that TANNylation of AAV-9 loaded with a GFP transgene resulted in a significant increase in GFP expression in the infarcted region heart compared to administration of AAV-9 alone (Fig. 6).

Figure 6: TANNylation increases AAV transduction in the heart.

Injection of TANNylated AAV9 leads to significantly increased GFP expression in the infarcted myocardium (white outlined tissue) compared to AAV9 alone. Reproduced from ref. 74 with permission from Springer Nature, copyright 2018.

The future of targeted nanoscale platforms

The use of IV injection for administering therapeutics to the heart is simpler and more easily translatable than approaches that involve open-heart procedures or the use of catheter-based systems. The field of nanomedicine offers platforms that could be particularly applicable following MI. As customizable as these nanoscale platforms are, the main concern of IV injections, off target accumulation and reduced therapeutic efficacy, cannot be overlooked. To overcome this, methods for targeting the heart have been explored. As discussed, there are many different techniques for targeting—most center around leveraging the pathophysiology of the microenvironment of the infarcted heart. Common design criteria focus on interacting with the inflammatory response or proteins and cellular markers that are primarily found or upregulated in the heart. While these methods have proven successful in preclinical studies, they can still suffer from issues with off-target accumulation, as many of the targets are not solely unique to the heart. This is particularly problematic with systems that target hallmarks of the immune response as many of these features could allow for localization in other areas of inflammation. Patients with co-morbidities such as arthritis or cancer may not be eligible for platforms that are targeting upregulation of the immune response and hallmarks of inflammation. In addition, many NP platforms depend on the enhanced permeability of the tissue, which is only temporarily present in the infarcted heart⁷⁵, meaning the timeline for delivery is limited. As a result, many of these nanoparticle systems are only applicable for acute MI. Moreover, this permeability can go both ways, with materials struggling to be retained, unless targeting can be achieved via a morphological transition of the material³¹. To advance nanomaterial therapeutic strategies, it could be beneficial to engineer nanoparticles that have a dual-targeting mechanism. Dual targeted mechanisms have been explored for other cardiovascular applications and have shown improved efficacy compared to single targeted counterparts^{35, 76}. Nanoparticle platforms that are triggered by more than one stimulus have the potential to improve localization in the heart and are an area of nanomedicine research worth pursuing.

Gene therapy in the heart is not a new concept, but the use of AAVs as viral vectors has more recently gained popularity as a platform for therapy following MI. As a standalone therapeutic, AAVs have proven their ability to successfully transduce both dividing and quiescent cells. In addition to this, we have also discussed specific examples of AAV delivery of a therapeutic transgene following MI. Though there have been promising results, these come with certain reservations. AAV delivery alone, especially systemic delivery, is still subject to many of the limitations we discussed for nanoparticle platforms: off-target accumulation, and poor retention at the site of interest. Taking this into consideration along with subpar clinical results to date, important future considerations should be made with regards to the method of administration and AAV design. This is currently being explored by research in the field focused on the use of tissue-specific promoter sequences for the heart, inducing mutations in motifs that bind to the liver, and co-delivery of AAVs with biomaterial scaffolds made out of alginate⁷⁷, fibrin^{78, 79} and synthetic polymers^{46, 80–83}. At this point, the clinical translation of AAVs is also hindered by neutralizing antibodies that render the viral vectors useless. For AAV delivery specifically, the biggest issues that need to be addressed center around finding ways to minimize the cytotoxic T-cell response and neutralizing antibody interactions. We have already seen efforts to address the latter by associating AAVs with extracellular vesicles (ex-AAVs). György et al. found that ex-AAVs were 136 times more resistant to neutralizing antibody interactions in vitro, allowing for increased transduction with a decreased number of viral particles administered⁸⁴. Other techniques for mitigating neutralizing antibody interactions and improving gene delivery include AAV PEGylation⁸⁵ and directed evolution⁸⁶. These are a strong example of the kind of engineering design that needs to be put into developing successful viral vector platforms as therapeutics. By creating solutions to neutralizing antibodies, it would be possible to imagine AAVs as a method to deliver therapeutic transgenes multiple times with high efficacy.

The use of targeting moieties to direct systemically injected therapeutics to the site of interest has been able to improve the therapeutic efficacy of various payloads loaded into nanoparticles or viral vectors, two types of nanoscale therapeutics. While therapeutic impact is the most important factor to consider, it is also crucial to assess the biodistribution of targeted therapeutics to ensure off target accumulation is mitigated. Though some of the examples discussed have included the necessary controls to demonstrate improvement over a non-targeted equivalent, there is still a need for more rigorous biodistribution quantification. With AAV delivery, PCR can be used to evaluate the transduction of various satellite organs. With administration of nanoparticle platforms, biodistribution methods are not standardized. Some nanoparticle platforms will be tagged with a fluorescent marker to allow for detection, however, the means for detecting these nanoparticles is highly variable. In vivo imaging system (IVIS) is a common method for measuring fluorescence in both living animals and excised tissues⁸⁷. Other groups have adopted a homogenization method for detecting fluorescence in tissue samples^{26, 88–90}. Overall, there is a need for the development of more standardized and high-resolution methods for detecting nanoparticle distribution in satellite tissues. This is particularly pertinent to platforms that claim targeting capability, since there is currently a lack of quantitative biodistribution in the field.

For both AAV and nanoparticle therapeutics, it is important to continue to explore new methods for targeting the heart. As discussed, there have been more efforts to increase cardiac-specific targeting via in vivo phage display and surface modification with binding moieties such as RGD or AT₁ targeting peptides, but even these methods are not perfect and could benefit from added specificity. Moving forward, it could be beneficial to engineer and apply viral vectors similarly to how we do for nanomedicine. As discussed, a few groups have begun to employ more complex designs of AAVs and seen improved cardiac transduction and reduced off-target accumulation.

Nanomedicine has become a powerful tool in cardiovascular research, allowing for improved circulation time and efficacy of systemically administered therapeutics in early preclinical studies. As the field evolves and grows, we can expect to see further improvements in mitigating off-target effects of therapeutic payloads through the clever use of physiological triggers and more cardiac-specific targets. Though the pathway to regulatory approval may be slower due to the increasing complexity in material design, it is worth the trade-off for optimized therapeutic impact. Overall, there are many promising nanoscale platforms on the horizon that could make minimally invasive delivery of nanoscale therapeutics to MI patients a reality.