Gene Therapy for Cardiac Transplantation

Abstract

Gene therapy is an advanced treatment approach that alters the genetic composition of cells to confer therapeutic protein or RNA expression to the target organ. It has been successfully introduced into clinical practice for the treatment of various diseases. Cardiac transplantation stands to benefit from applications of gene therapy to prevent the onset of post-transplantation complications, such as primary graft dysfunction, cardiac allograft vasculopathy, and rejection. Additionally, gene therapy can be used to minimize or potentially eliminate the need for immunosuppression post-transplantation. Several animal models and delivery strategies have been developed over the years with the goal of achieving robust gene expression in the heart. However, a method for doing this has yet to be successfully translated into clinical practice. The recent advances in ex vivo perfusion for organ preservation provide potential ways to overcome several barriers to achieving gene therapy for cardiac transplantation into clinical practice. Optimizing the selection of the gene-carrying vector for gene delivery and selection of the therapeutic gene to be conferred is also crucial for being able to implement gene therapy in cardiac transplantation. Here, we discuss the history and current state of research on gene therapy for cardiac transplantation.

1. Introduction

Cardiac transplantation is the gold standard therapy for end-stage heart failure. The perfection of surgical interventions, development of modern immunosuppressive therapies, and implementation of rigorous transplant care protocols have contributed to better outcomes over the last several years [1, 2]. However, cardiac transplantation is limited by the number of available donor hearts, primary graft dysfunction (PGD), rejection of the heart, as well as by the side effects caused by immunosuppression therapy [3]. Gene therapy is an advanced treatment intervention that can potentially bridge the gap to overcome these common post-transplantation complications. The success of commercially available gene therapy interventions, such as Zolgensma for spinal muscular atrophy and Luxturna for Leber congenital amaurosis, demonstrates that gene therapy provides a viable treatment option for people who would otherwise suffer from diseases that have traditionally been thought of as impossible to treat.

Gene therapy works by altering the genetic composition of cells to confer therapeutic protein or RNA expression to the target organ. To date, it has been commercially used to treat spinal muscular atrophy, Duchenne muscular dystrophy, and for various types of ocular disorders [4]. There are currently many gene therapy clinical trials underway and growing in number (clinicaltrials.gov). Gene therapy based interventions have been studied for various cardiovascular diseases, such as coronary artery disease (CAD), heart failure (HF), and myocardial ischemia (MI) [5]. However, no intervention has yet been able to attain robust or long-term transgene expression in the heart in clinical practice. One promising intervention for HF was the AAV1-SERCA2a therapeutic which was evaluated in human clinical trials (CUPID, AGENT-HF, SERCA-LVAD). The trials, unfortunately, failed to demonstrate that the intervention led to a statistically significant difference in the primary endpoint of time to recurrent HF and secondary endpoint of time to first terminal events [6, 7, 8].

The heart is a complex target for gene therapy interventions due to its location in the body, the mechanical force of blood flow, endothelial barriers, cellular barriers, and the body’s immune response [9]. A cardiac graft being treated prior to transplantation is uniquely amenable to gene therapy as most of these traditional barriers of gene delivery to the heart can be overcome. Through gene therapy, a cardiac allograft can be engineered to express selected therapeutic genes that could prevent the onset of post-transplantation complications and potentially minimize or eliminate the need for traditional systemic immunosuppression medications [10, 11]. Gene therapy for heart transplantation, though attractive, has not been translated clinically.

There are major challenges that need to be overcome for gene therapy to be able to be applied for cardiac transplantation. One of them is that, despite major advances in the understanding of transplant immunology, there remains an incomplete understanding of the mechanisms of both rejection and tolerance. This includes the understanding of the details of regulatory cytokine networks, MHC-antigen interactions during the rejection process, and a complete understanding of co-stimulatory factors and their functions [12]. Another challenge is that most current gene delivery mechanisms confer a transient, low level of gene expression [13]. With the current understanding of gene therapy in the context, it also is unclear what is the optimal dose of the therapeutic transgene needed to confer an appreciable clinical effect. However, recent investigations describe methods of robust and global gene delivery to cardiac grafts that offer promise to overcome this challenge. Similarly, viral vector delivery systems pose risks to the host and allograft via eliciting undesired immune reactions, off-target gene delivery, and genome integration. With the recent success and clinical adoption of ex vivo normothermic perfusion, the possibility of gene delivery that is isolated to the cardiac graft is feasible and promising for translation into clinical practice. Ex vivo normothermic perfusion also provides the optimal environment for viral vectors to be able to attach and enter cardiac cells for efficient transduction. With the advances that have been made to address these challenges, it will not be long before we witness the successful application of gene therapy to cardiac transplantation.

To achieve a successful gene therapy intervention for cardiac transplantation, several components need to be addressed: disease or indication and therapeutic target, use of an appropriate animal to test the therapeutic, selection of the vector for gene delivery, and method for vector delivery. Here we review select post-transplantation complications and potential targets where gene therapy can be implemented to prevent them. We will also review translational animal models that have been developed for investigating gene therapeutic targets. Finally, we will discuss the different viral and non-viral vectors that can be used for gene delivery, the selection of promoters, and the different modalities that have been investigated for the delivery of vectors to cardiac grafts.

2. Disease and therapeutic targets for cardiac transplantation

There are various insults that a cardiac graft experiences prior to, during, and after transplantation. Early damage to the cardiac graft can happen during the brain or cardiac death of the donor, organ procurement, organ preservation time, the implantation procedure, or as a result of reperfusion injury. These points of insult to the cardiac graft can trigger both innate and adaptive immune responses that result in injury. Common complications that occur following transplantation include primary graft dysfunction (PGD), coronary allograft vasculopathy (CAV), and rejection.

2.1 Primary graft dysfunction

PGD is a leading cause of early mortality post-transplantation [14]. The diagnosis of PGD occurs in the first 24 hours following heart transplantation. It presents as severe ventricular dysfunction of the cardiac graft in the immediate post-transplant period, resulting in low cardiac output and hypotension despite the presence of adequate filling pressures [15]. Either the left, right, or both ventricles can be involved, and the severity of the dysfunction can range from mild to moderate to severe depending on the extent of circulatory support that is needed to maintain hemodynamic stability [16].

Numerous causative factors, starting from donor death to weaning the heart from cardiopulmonary bypass in the recipient, have been identified that contribute to the cause of PGD [17]. These factors relate to ischemic and ischemic-reperfusion injury of the cardiac graft. Additionally, the onset of systemic inflammatory response syndrome and the development of vasoplegic syndrome in the recipient have also been identified as significant causes [18]. Finally, the use of extended criteria donors, such as donation after cardiac death (DCD), has also been identified as a significant risk factor for PGD [19].

The treatment of PGD is primarily through supportive care. It is typically initially managed with the use of inotropic support using catecholamines and phosphodiesterase inhibitors. The next escalation in care is typically the use of an intra-aortic balloon pump, followed by the initiation of advanced mechanical support using extracorporeal membranous oxygenation.

2.2 Cardiac allograft Vasculopathy

Cardiac allograft vasculopathy (CAV) is a major cause of late heart graft failure [20]. It is characterized by diffuse and concentric narrowing of large epicardial and small intramyocardial arteries due to intimal fibromuscular hyperplasia, atherosclerosis, and vasculitis. As a result, the transplant recipient develops pathological changes within the donor blood vessels leading to a spectrum of diseases ranging from MI to HF. CAV is often unable to be diagnosed by coronary angiography and requires intravascular ultrasound for diagnosis.

The main driver of CAV is believed to be the immune system of the host. The intimal thickening seen in CAV results from an accumulation of smooth muscle cells (SMC) accompanied by the infiltration of T cells and macrophages which further contribute to intimal expansion [21, 22]. Yet CAV lesions characteristically stop at the suture line between the donor and the recipient. The endothelial lining of the vessels remains intact in CAV lesions suggesting that SMC injury may result from sterile inflammation as is seen during cold and warm ischemia effects and ischemia–reperfusion injuries [23].

Current treatments are based on vascular risk factor management and the use of statins and mTOR inhibitors (sirolimus and everolimus) to reduce the development of the disease. Percutaneous revascularization is used to treat focal obstructive coronary stenosis but repeat revascularization rates are high due to restenosis and disease progression [24]. However, patients who go on to develop allograft dysfunction require re-transplantation [25, 26].

2.3 Rejection and immunosuppression

Cardiac allograft rejection is among the most common causes of death in heart transplant recipients [1]. Acute rejection is categorized into hyperacute rejection acute cellular rejection (ACR), and antibody mediated rejection (AMR). Currently, recipients undergo routine screening for rejection with endomyocardial biopsies obtained by a bioptome. Hyperacute rejection is due to the presence of preformed host antibodies against the graft and portends an inevitable immediate immune rejection resulting in death [27]. ACR and AMR take longer to manifest and are thus amenable to potential gene therapy intervention and we will focus our discussion on these forms of rejection. To prevent rejection of the cardiac allograft, patients are treated with systemic multidrug immunotherapies. Multidrug immunosuppressive regimens currently used in human transplant recipients are associated with an increased risk of malignancy and opportunistic infections, a metabolic syndrome characterized by insulin resistance and dyslipidemia, and drug-specific toxicity [11].

2.4 Potential targets for gene therapy

An understanding of the different insults that the cardiac graft experiences during the different steps of transplantation helps to identify potential targets for gene therapy for cardiac transplantation. The cardiac graft endothelium is vulnerable to ischemic reperfusion injury. In this setting, leukocytes adhere to the activated endothelium. The complement system becomes activated, neutrophils migrate into the cardiac graft, subsequently followed by natural killer cells and macrophage infiltration. These early non-specific inflammatory reactions are then followed by alloimmune reactions that result in massive graft infiltration by dendritic cells, T-cells, B-cells, and macrophages. Donor-derived dendritic cells leave the cardiac graft and migrate to recipient lymph nodes and the spleen. There they present donor antigen to recipient T cells directly and trigger acute rejection.

2.4.1 Inflammatory targets

Many candidate genes that interfere with one of these inflammatory mechanisms have been investigated in the context of cardiac transplantation. One such gene is endothelial nitric oxide synthase (eNOS). eNOS produces nitric oxide which is vasoprotective. Delivery of eNOS into the donor heart attenuated ischemic reperfusion injury, leukocyte infiltration, and cardiac graft rejection in a rabbit model [28]. Similarly, superoxide dismutase (SOD) gene delivery into a donor heart attenuated ischemic reperfusion injury after organ preservation and transplantation in a rabbit model [29]. SOD functions as a free radical scavenger that neutralizes reactive oxygen species generated during ischemic reperfusion injuries. Another target, nuclear factor-kappa B (NFkB), is a transcription factor involved in the up-regulation of pro-inflammatory gene products. One possible therapeutic intervention is to block NFkB in endothelial cells to attenuate ischemia–reperfusion injury in the myocardium. Sakaguchi et al. blocked NFkB by using double-stranded oligodeoxynucleotides with a specific affinity for NFkB (NFkB decoy group) to transduce rat hearts utilizing HVJ envelope [30]. The hearts were then preserved for 16 hours in hypothermic preservation solution before being heterotopically transplanted into a recipient rat. What they found is that the intervention attenuated ischemic reperfusion injury after prolonged heart preservation in hypothermic solution. Another protein that is up-regulated during inflammation and serves as a potential target for gene therapy is heat shock protein-70 (HSP-70). It has an essential role in protein folding and translocation and as chaperones for intracellular proteins. HSP-70 has particularly been shown to be associated with protection against ischemia–reperfusion injury. Jayakumar et al. infused rat hearts using 1 mL of the gene vector solution then incubated the hearts on ice for 10 minutes before heterotopically transplanting them into a recipient rat [31]. 4 days after the intervention, the hearts were perfused on a Langendorff apparatus for 45 minutes followed by reperfusion for 1 hour. They found that post-ischemic recovery of mechanical function was greater in the treatment arm versus control, recovery of coronary flow was greater as well. The conclusion was that HSP-70 gene transduction protects both the mechanical and endothelial function of the cardiac graft.

2.4.2 Rejection targets

The most direct and immediate barrier to the success of cardiac transplantation is the recipient immune response. Currently, the most effective clinical therapy is lifetime immunosuppressive therapy. Knowledge about the immune response in transplantation has grown tremendously in recent years such that gene therapy can be used to intervene on different targets of the immune response. Both cell and antibody mediated effector mechanisms are responsible for acute rejection [32]. A strategy to protect the cardiac graft from recipient immune responses is through the delivery of genes that confer proteins to the graft that modulate host immune responses. These would include cytokines or soluble ligands. Qin et al. utilized a retrovirus and a plasmid delivery system to transfer genes that encode transforming growth factor beta-1 (TGF-β and interleukin 10 (IL-10) to a mouse myoblast and non-vascularized cardiac graft [33]. Grafts transduced with either of these genes had significantly prolonged survival when compared with the vector alone (39 days with IL-10 vs. 26 days with TGF-β vs. 12 days with vector alone). The therapeutic effect of transduced IL-10 and TGF-β1 has been demonstrated in follow-up investigations using different types of vectors [34, 35, 36].

Another point of gene intervention would be at the point of T-cell costimulatory activation. Cytotoxic T lymphocyte antigen-4 (CTLA-4) is a protein that modulates T-cell costimulatory activation. It becomes upregulated on T-cells upon T-cell activation. Gene delivery of a soluble CTLA-4 immunoglobulin fusion protein (CTLA4Ig) into the donor heart was associated with detectable CTLA4Ig serum levels 120 days after transplantation as well as long-term cardiac graft survival, >100 days in a rat model [37]. However, the expression of CTLA4Ig did enter systemic circulation causing some systemic immunosuppression in the rats. Another similar target is the programmed death-1 (PD-1) gene. It is expressed on activated T-cells, B-cells, and myeloid cells. When PD-1 binds one of its ligands, PD-L1 or PD-L2, it leads to the inhibition of activated T-cells. PD-L1 and PD-1 play an important role in both acute and chronic rejection of transplanted hearts in animal studies [38, 39, 40]. In rejecting human transplanted hearts, PD-L1 expression is decreased relative to PD-1 expression [41]. Gene delivery of soluble PDL1Ig fusion protein into the donor heart moderately prolonged cardiac graft survival in rats [42].

2.4.3 Cardiac ischemic disease targets

An additional example of gene therapy applied to treat cardiac disease involves targeting angiogenic gene therapy that facilitates neovascularization to augment blood flow in ischemic myocardium. These include vascular endothelial growth factor (VEGF), fibroblast growth factor (FGF), and hepatocyte growth factor (HGF). In particular, these targets have been assessed for treating ischemic disease caused by MI or congestive heart failure. Rosengart et al. delivered 4 x 10⁸ to 4 x 10¹⁰ particle units of an adenoviral vector encoding the VEGF gene to individuals undergoing bypass graft surgery and as the sole therapy to the experimental group via mini-thoracotomy. The intervention demonstrated no adverse events and there was symptomatic improvement in both groups [43].

The Angiogenic Gene Therapy (AGENT) clinical trials were the first randomized control trial studies investigating the benefits of stimulating coronary angiogenesis with gene therapy using FGF-4 [44]. FGF-4 was delivered using adenovirus administered by infusion into the coronary arteries of patients with chronic stable angina. AGENT evaluated incremental doses of 3 x 10⁸ to 1 x 10¹¹ particle units. The overall improvement in exercise treadmill time was similar for those in the treatment and the control arms. However, post-hoc analysis showed that when baseline neutralizing antibody titer was controlled for, patients with titers less than 1:100, 44% had increased their exercise treadmill time by more than 30%. In patients with titers greater than 1:100, only 7% had increased their exercise treadmill time by more than 30%. AGENT 2 investigated whether FGF-4 improved myocardial perfusion compared with placebo. A significant decrease in ischemic defect size was observed in the treatment arm (21% relative decrease) that was not observed in the placebo group. AGENT 3 and 4 were planned to determine the efficacy and safety of FGF-4 in the larger population, however, an interim review of the data demonstrated no differences in exercise treadmill time and therefore recruitment was stopped.

HGF as a therapeutic target has been evaluated in numerous studies. In the context of therapy for MI, Jin et al. investigated the long-term effects of HGF in a rat MI model [45]. Utilizing an adenoviral vector for delivery of HGR, the vector was injected directly into the infarct border zone immediately after permanent coronary ligation. 10 weeks post-intervention, there was no significant difference in the left ventricular ejection fraction, but capillary density was significantly higher in the treatment groups, whereas arteriole density was unchanged. Masahiro et al. describe the use of recombinant HGF delivered by HVJ envelope for prolonged cardiac graft preservation in rats during hypothermic storage [46]. The rationale for this choice is that HGF functions as an antiapoptotic factor in the heart. They concluded that the administration of HGF prevented myocardial apoptosis and improved cardiac function after prolonged myocardial preservation in hypothermic solution.

3. Animal models

Selection of an appropriate animal model for heart transplantation is critical to be able to translate a potential gene therapy intervention from the laboratory bench to the patient bedside. Numerous small animal models using rodents have been described where the heart is transplanted either heterotopically or orthotopically in the recipient animal. Similarly, there have been numerous large animal models described using pigs, sheep, and non-human primates (NHP). We will discuss examples of different types of small and large animal models in heart transplantation and in what instances an investigator may choose one over the other.

3.1 Heterotopic heart transplantation

Heterotopic heart transplantation (HHT) is when the transplanted heart is placed in an ectopic position inside of the recipient without the removal of the recipient’s native heart. Intra-abdominal HHT is primarily used to investigate transplantation biology and is also suitable for studying unloading induced changes in the heart [47]. The heart of a donor animal is explanted and subsequently transplanted into the abdomen of a recipient animal. To accomplish this the donor ascending aorta is anastomosed to the recipient infrarenal aorta and the donor pulmonary aorta is anastomosed to the recipient inferior vena cava. The result of this configuration is that the graft beats with reduced left ventricular filling while coronary perfusion is preserved. It offers several advantages over orthotopic transplantation in research applications such as technical simplicity, better accessibility for biopsies, and survival of the recipient even in cases of graft rejection [48].

The first heterotopic abdominal heart transplantation was published using rats by Abbott et al. [49] in 1964 and subsequently modified by Ono et al. [50]. The technique of the latter has been widely adopted as the standard HHT rodent model. Heterotopic heart transplantation in mice is more challenging than in rats, however, testing mechanistic hypotheses is more practical in mice given the greater diversity of genetic modifications available in mice. The advantage of using a small animal model is that they are less costly when compared to the cost of a large animal. A larger number of small animals can be used to assess and describe the effects of a therapeutic transgene. It also allows for several transgenes to be tested in parallel to study the differences in efficacy between them. The main challenge in using small animals is that the micro-surgical implantation technique is very challenging given their smaller size. Another aspect that makes this surgery more challenging to do in smaller animals is that they have a lower tolerance for blood loss. Because of this, it is especially important that there be minimal blood loss during the procedure and that the anastomoses be hemostatic at the time of procedure completion.

The advantage of large animals is that the results of the gene therapy intervention are able to be translated more quickly into clinical practice than are the results obtained from small animal studies. However, large animals are very costly to acquire and maintain in comparison to small models. In the setting of small primates, Minanov et al. positioned NHP hearts into the iliac fossa of primate recipients [51]. More recently this transplant configuration has been used to investigate interventions in xenotransplantation using a porcine heart transplanted into a baboon [52, 53, 54]. Porcine to porcine heart transplantation is also used in the research setting to investigate the immune system effects of cardiac transplantation as well as gene therapy interventions (Figure 1) [55, 56, 57]. This surgical research model is also amenable for modeling post-cardiac transplantation complications, such as CAV and rejection, without subjecting the animal to a high risk of morbidity or mortality [55, 58]. The recent success of a porcine to human xenotransplantation using genetically modified pigs to minimize rejection by the human immune system of the xenograft stresses the importance of the selection of the appropriate animal model. After procuring the heart, the xenograft was preserved utilizing an ex vivo perfusion device until the time of implantation. The experiments leading up to this milestone utilized the heterotopic heart transplantation model to establish the longevity of the graft against rejection [53, 59].

Figure 1.

Heterotopic heart transplantation in the intra-abdominal position in a large animal porcine model. The donor aorta is anastomosed to the recipient infrarenal aorta and the donor pulmonary artery is anastomosed to the recipient inferior vena cava.

3.2 Orthotopic heart transplantation

Orthotopic heart transplantation is when the transplanted heart is placed in the position of the recipient’s native heart. As such, the cardiac graft takes over providing the cardiovascular support of the recipient. This transplant configuration in research is most useful to investigate the cardiac graft’s overall ability to support the recipient following an administration of a new intervention. The pros of this design are that it most closely reflects clinical practice so one can investigate beyond the immunopathologic changes the heart undergoes after transplantation. This approach allows the investigator to determine whether an intervention permits the transplanted heart to perform its intended function to support the recipient’s cardiovascular system. This has been successfully described in porcine to porcine models, as well as in pig to baboon xenotransplantation models [60, 61, 62].

4. Vectors for gene delivery

Vectors for gene delivery comprise viral and non-viral vectors. Viral vectors are the more efficient of the two but are also associated with more side effects than non-viral vectors. Each type of viral vector confers different gene expression characteristics, such as the length of time for transgene expression and the intensity of transgene expression (Table 1). Additionally, when constructing the optimal vector for cardiac gene delivery consideration must be given to the selection of promoter. Constitutively active promoters, such as CMV or RSV promoters, confer broad tissue tropism and strong expression. However, cardiac-specific promoters, such as myosin heavy chain promoter, myosin light chain promoter, and troponin T promoter have been used to restrict transgene expression in the heart [63]. While the cardiac-specific promoters focus gene delivery to cardiac tissue, they confer weaker expression when compared with constitutively active promoters (Table 2).

Viral vector	Genetic material	Capacity	Transduction ability	Peak gene expression	Main advantages	Characteristics
Adenovirus	dsDNA	4.5–36 kb	Transduces both dividing and non-dividing cells.	1–7 days	Efficiently delivers genes to most tissues.	Short-term but highly efficient gene delivery. Can elicit a strong inflammatory response.
Adeno-associated virus	ssDNA	4.7 kb	Transduces both dividing and non-dividing cells.	2–4 weeks	Low immunogenicity. Broad but specific tropism.	Long-term gene expression. Low immunogenicity.
Lentivirus	RNA	8 kb	Transduces both dividing and non-dividing cells.	4–6 days	Can carry multiple transgenes. Persistent gene transfer in dividing cells.	Persistent gene expression in dividing cells. Low but potential risk of mutagenesis.

Table 1.

Summary of common viral vectors used in gene therapy.

Reference	Transduced gene	Therapeutic mechanism	Transduction method	Key findings/conclusions
Iwata et al. [28]	eNOS	Attenuation of ischemia–reperfusion injury	Lipid/DNA complex via intra-op coronary infusion	Allogeneic rabbit heart transplant model demonstrated that intramyocardial neutrophil and T-cell populations were halved in eNOS transduced hearts. NF-kB activation in microvascular endothelial cells and cardiomyocytes was significantly reduced.
Abunasra et al. [29]	SOD	Attenuation of ischemia–reperfusion injury	Ad via ex vivo perfusion	Heterotopic heart transplant model in rats demonstrated positive immunoreactivity for SOD and 86.8% +/− recovery of pre-ischemic left ventricular pressure.
Jayakumar et al. [31]	HSP-70	Attenuation of ischemia–reperfusion injury	HVJ envelope via ex vivo perfusion	Heterotopic heart transplant model in rats demonstrated greater post-ischemic recovery of mechanical function and greater recovery of coronary flow in HSP-70 treated mice.
Sakaguchi et al. [30]	NF-kB decoy	Attenuation of ischemia–reperfusion injury	HVJ envelope via ex vivo coronary infusion	Heterotopic heart transplant model in rats demonstrated introduction of NF-kB decoy into the nuclei of endothelial cells and cardiomyocytes. After 1 hour of reperfusion the NF-kB decoy group showed significantly higher degrees of recovery of left ventricular function.
Guillot et al. [37]	CTLA4Ig	Attenuation of T-cell costimulatory pathway	Ad via intramyocardial injection	Heterotopic heart transplant model in rats demonstrated indefinite graft survival (>100 days) and could be detected in the graft at least 1 year after gene transfer. Evident suppression of antibody production against donor alloantigens up to at least 120 days after gene transfer.
Dudler et al. [42]	PD-L1Ig	Attenuation of T-cell costimulatory pathway	Ad via ex vivo coronary infusion	Heterotopic heart transplant model in rats demonstrated a prolonged median survival time (17 days vs. 11 days). Also demonstrated a decreased number of CD4 cells and monocytes/macrophages infiltrating the graft.
Grines et al. [44]	FGF	Angiogenic therapy	Ad via intracoronary infusion	Randomized controlled trial that enrolled patients with chronic stable angina demonstrated improved exercise time on a treadmill for those treated with intervention and had a baseline time < or equal to 10 minutes. Intervention decreased the ischemic defect size. Larger efficacy studies failed to demonstrate significant differences in exercise time on a treadmill so the trial was stopped.
Rosengart et al. [43]	VEGF	Angiogenic therapy	Ad via intramyocardial injection	Phase I clinical study that enrolled patients with clinically significant coronary artery disease. There were no systemic or cardiac related adverse events related to vector administration. Coronary angiography and stress sestamibi scan showed improvement in the treated area. All patients reported improvement in angina class after therapy.
Jin et al. [45]	HGF	Angiogenic therapy	Ad via intramyocardial injection	Myocardial infarction model in rats demonstrated no significant difference in the left ventricular ejection fraction. It did observe increased capillary density in the treatment group.
Ryugo et al. [46]	HGF	Angiogenic therapy/Antiapoptosis	HVJ via cold static storage	Cardiac grafts procured from rats demonstrated that HGF treated hearts had a significantly higher recovery rate of left ventricular developed pressure. c-MET/HGF receptor expression was stronger in the treatment group.

Table 2.

Summary of investigations of gene therapy for cardiac transplantation.

4.1 Adenoviral vectors

Adenovirus (Ad) vectors have high transduction efficiency. They are able to transduce both quiescent and dividing cells and maintain epichromosomal persistence in the host cell [64]. Ad vectors also have a broad tropism profile and large packaging capacity (4.5-36 kb). They offer efficient transduction of cardiomyocytes. However, gene expression is transient, peaking 1–7 days after delivery and then diminishing until it ceases at about 2–3 weeks after transduction [65]. They carry double-stranded DNA. Their main disadvantage is the widely pre-existing viral immunity among the general population. Since Ad is strongly immunogenic it causes undesired immune responses in treated subjects [66]. In order to overcome this and improve their capacity, Ad vectors have undergone several generations of engineering.

The first generation of Ad vectors was designed by removing the E1A gene which makes it so the recombinant Ad is unable to replicate within the host cell [67]. With the deletion of this gene, complementary cell lines, such as HEK293, had to be designed to express E1A and E1B in order to produce the viral vector. The main disadvantages of the first generation of Ad were that de novo expression of Ad proteins could activate the host immune response and there was still the possibility of spontaneous homologous recombination between the vector and engineered E1 region from HEK293 that could generate replication-competent adenovirus [68].

In the second generation of Ad vectors, further early gene regions (E2a, E2b, or E4) of the vector were deleted to permit additional space for the transgenes. As in the first generation, the deleted genes needed to be complemented by engineered production cell lines. However, the deletion of these genes led to inefficient complementation of E2/4 with engineered cell lines thus negatively affecting viral vector amplification, resulting in lower titers. Another disadvantage was that the native Ad late genes that were still retained within the viral genome could trigger host immunogenicity and cellular toxicity [69].

Finally, the third generation of Ad vectors have all Ad viral sequences deleted except for the inverted terminal repeat sequences and packaging signal. As such, these are referred to as “gutless” or “high capacity” Ad vectors (HCAd). The production of HCAds in cell culture requires an adenoviral helper virus similar to the first-generation Ad vectors. Compared with the previous Ad vector generations, HCAds have reduced immunogenicity, prolonged transduction in the host cell, and a significantly larger transgene capacity [64]. Their large transgene capacity makes it so that multiple transgenes could be delivered. The main disadvantage of HCAds is the challenge of ensuring that the helper virus is eliminated from the final vector preparation.

4.2 Adeno-associated viral vectors

Adeno-associated viral (AAV) vectors were discovered as a contaminant of Ad preparations in 1965 [70]. They lack essential genes needed for replication and expression of their own genome. They are not known to cause any human diseases. AAV vector was first used in humans in 1995 to deliver the cystic fibrosis transmembrane regulator (CFTR) gene into a patient with cystic fibrosis using the AAV2 capsid [71]. Today, recombinant AAVs are the leading vectors for the delivery of gene therapies. The first recombinant AAV gene therapy product, Glybera, was approved by the European Medicines Agency to treat lipoprotein lipase deficiency in 2012. Five years later, Luxturna was approved as the first recombinant AAV gene therapy product in the United States [72, 73].

AAVs carry single-stranded DNA (ssDNA). However, the efficiency of AAVs are limited by ssDNA in that it needs to be converted to double-stranded DNA (dsDNA) prior to expression. This step is circumvented through the use of self-complementary vectors which package an inverted repeat genome that can fold into dsDNA without the requirement for DNA synthesis or base-pairing between multiple vector genomes [74]. Transgene expression peaks at around 2–4 weeks after delivery. AAVs can carry transgenes up to 4.7 kb in size.

There are 13 natural AAV serotypes. These have been isolated from laboratory Ad stocks and mostly from human or non-human primate origin [75]. Engineering or recombinant AAV capsids confer the vector the capability to transduce multiple tissue types. Recombinant AAVs are composed of the same capsid sequence and structure as found in wild-type AAVs. Recombinant AAVs encapsidate genomes that are devoid of all AAV protein-coding sequences and that have therapeutic genes designed in their place. The complete removal of viral coding sequences maximizes the packaging capacity of these AAVs and contributes to their low immunogenicity and cytotoxicity [73].

Capsid development approaches are based on rational design and directed evolution. The rational design was among the first approaches to improve vector capsids. This entailed adding peptide sequences onto the surface of the capsid to direct the tropism of the vector and deter immunological recognition [76]. While rational design allowed for the early development of specialized AAVs, a major limitation of that approach is that there oftentimes is insufficient knowledge regarding AAV cell surface binding, internalization, trafficking, uncoating, and gene expression. The basis of directed evolution is in the simulation of natural evolution. Capsid libraries are placed under selective pressure to yield genetic variants with specific biological properties and advantageous characteristics. This way directed evolution of the capsid does not require a prior understanding of the molecular mechanisms involved in the selection criteria [73].

Cell-type specific transgene expression, however, is conferred at the level of gene transcription by the promoters used in AAV vectors. The serotype AAV9 has been shown to have the highest cardiac gene transduction efficacy in mice and rats with either systemic or direct cardiac injection [77, 78]. Meanwhile, the serotype AAV6 has proven to be a more effective vector when injected into the myocardium of pigs and non-human primates [79, 80]. Piacentino et al. described a recombinant AAV serotype engineered via rational design, termed SASTG, which has extremely high-level cardiac transduction and tropism [81]. A challenge for AAV-mediated gene therapy is overcoming the negative effect that innate immunity has on transgene expression. Yet adaptive immunity to the capsid and the foreign transgene is the main factor for decreased efficacy. Notwithstanding, recombinant AAVs are accepted as the least immunogenic when compared to other viral vectors. Patients that have been exposed to AAV serotypes that gene therapy is based on will have a high chance of forming antibodies against the vector capsid [82]. One plausible way of removing these anti-AAV antibodies from the bloodstream is by using plasmapheresis [83]. Another described pre-treatment is the use of IgG-cleaving endopeptidases which reduce IgG antibodies from the serum [84]. Besides removing the neutralizing antibodies, investigators have also utilized rational design and directed evolution to develop AAV capsids that evade neutralizing antibodies [85, 86, 87, 88].

4.3 Lentivirus

Lentiviral vectors constitute a genus of the retrovirus family. They permit long-term transgene expression by integrating the delivered genes into the host genome and can carry transgenes up to 8 kb in size [89]. They can deliver single-stranded RNA to both dividing and non-dividing cells and display robust transduction efficiency [90]. A unique advantage of lentiviral vectors is the ability to express multiple genes from a single vector [91, 92]. Transgene expression peaks after 4–6 days. The immune response to lentiviral vectors is low but concerns remain about potential insertional mutagenesis and off-target gene expression [93]. They have a preference for targeting the coding regions of genes, carrying the risk of insertional oncogenesis [94]. Additionally, the vector lacks tropism for the heart, making it unideal for heart-specific delivery through in vivo delivery, however, may have a role in ex vivo delivery [95, 96].

4.4 Non-viral vectors

Naked nucleic acids allow for the delivery of large genes in high quantities. These include DNAs, mRNAs, micro RNAs, and siRNAs. However, the lack of protection from endonuclease degradation makes them unreliable with low cellular internalization of the transgene [97]. Additionally, naked nucleic acids have an uncondensed shape and polyanionic charge that does not allow for their efficient uptake into cells. The half-life of plasmid DNA is about 10 minutes following systemic injection into mice [98].

Nanoparticles have been developed to interact with nucleic acids to protect them from degradation and condense them into nano-sized complexes that can be internalized by cells. Two main types of nanoparticles being used in investigations are lipid-based and cationic polymer-based. Another modification that is being used to improve the uptake of naked nucleic acids by cells is through chemical modification to mRNA to reduce the activation of the immune system and improve the stability of the RNA. These modified mRNAs are attractive agents for short-term gene delivery to the myocardium [99].

4.5 Hemagglutinating virus of Japan envelope vector

Wild-type hemagglutinating virus of Japan (HVJ) was discovered in 1953 and is a member of the paramyxovirus family. The envelope of HVJ is composed of a lipid bilayer and two integral membrane glycoproteins, F and HN, that project from the viral surface [100, 101]. HVJ envelope vector is constructed by incorporation of plasmid DNA into inactivated HVJ-containing liposomes [102]. During the preparation of the envelope vector, HN and F are retained but all the genome inside of HVJ is removed. It has high efficacy to induce a molecule into a target cell by the strong action of fusing cells on its membrane. Additionally, the removal of all the virus genomes confers low immunogenicity to the vector and eliminates replication and viral gene expression in cells. It is in essence a “viral, non-viral hybrid vector” [101]. HVJ can be used to deliver DNA, RNA, and oligonucleotides efficiently both in vitro and in vivo. The genetic material is entrapped within the HVJ liposomes and directly introduced into the cellular cytoplasm by means of the fusion activity of HVJ and not by endocytosis.

5. Methods for delivery of vectors for cardiac gene therapy

Gene delivery to any organ is a challenging feat. Gene delivery to the whole cardiac allograft is an especially challenging task given numerous obstacles. In vivo physiologic barriers include the heart’s location in the body, the mechanical force of blood flow, endothelial barriers, cellular barriers, and the body’s immune response [9]. Additional barriers involve the limited spread of the vectors from the site of vector exposure to achieve widespread transgene expression as well as the lack of an effective procedure for delivering the vectors without causing significant injury to the cardiac tissue that also maximizes the exposure time to cardiac tissue to the vector.

5.1 Intramyocardial injection

Direct intramyocardial injection of the vector into the myocardium is one such technique for vector delivery. It is easy to perform the injections and could theoretically be performed during graft procurement or after cardiac transplantation. Guzman et al. described the use of this technique for the delivery of adenovirus injected through a 25-gauge needle into the cardiac apex [103]. The intramyocardial injection has also been described in a clinical trial where subjects underwent a thoracotomy with the injection of vascular endothelial growth factor-2 naked deoxyribonucleic acid. They found that the procedure is well tolerated and reported few major adverse cardiac events at 1 year [104]. The major limitation of this technique for cardiac transplantation is that it only allows for limited focal delivery and the inability to target deeper muscular structures of the heart, such as the septum. Additionally, it is challenging to keep all of the injected material inside of the myocardium, leading to leakage from the needle holes and causing injury to the heart [105].

5.2 Intracoronary infusion

Intracoronary infusion of the vector is another described technique. By this method, the vectors are infused directly into the coronary arteries and reach the target cells for transduction by transit through the coronary arterial tree. Intracoronary infusion can be achieved by several methods: coronary catheterization prior to procurement, in vivo coronary infusion through the cardioplegia catheter prior to explantation, and ex vivo coronary infusion.

Catheterization of the coronary arteries for delivery and infusion through the cardioplegia catheter at the time of the graft procurement allows for a more dispersed and homogenous distribution of transgene delivery than is achieved through intramyocardial injections. Generally, transgene expression is able to be observed along with the distribution of the coronary arteries [2]. Several disadvantages exist with these delivery techniques. One is the negative effect pre-existing coronary artery disease has on the ability of vectors to reach their cellular target. Another is that since the infusion of the vector is based on a single bolus delivery when using a catheter-based approach, there is a large amount of vector that is lost to the systemic circulation resulting in poor transduction efficacy of the heart and a significant amount of off-target transduction. Finally, transduction efficacy is hampered by the presence of circulating neutralizing antibodies in the recipient against viral vectors. Vector particles containing proteins that are similar to antigens that humans are exposed to following natural infection may be neutralized by antibodies upon injection in some humans because of pre-existing immunity [106].

5.3 Complete heart isolation by cardiopulmonary bypass

Administration of the vectors during cardiopulmonary bypass featuring complete heart isolation and continuous cardiac perfusion addresses the issues associated with the catheter-based intracoronary infusion. The technique for achieving this was described by Katz et al. using separate pumps for the systemic and cardiac circuits permitting continuous isolated arrested heart perfusion [8]. This allows for the vectors to be recirculated through the coronary circulation of the heart, allowing for additional opportunities for the vectors to attach to cells and achieve entry. However, cardioplegia arrest requires for the heart and circulation to be maintained at a cold temperature (4°C) which is not favorable for vector attachment and entry into the target cells [107].

5.4 Ex vivo perfusion

The procedure for cardiac transplantation offers a unique opportunity for gene delivery that does not exist for other indications for therapeutic intervention for heart disease. The cardiac graft is removed from the recipient and preserved for a period of time ex vivo until it is implanted into the recipient. During this time the heart can be treated in isolation, obviating the need for additional procedures on the donor or recipient and minimizing or potentially eliminating the risk of off-target transduction by the gene delivering vector. With the recent FDA approval of ex vivo perfusion devices for organ preservation during transplantation, the ability to deliver vectors via ex vivo coronary perfusion seems plausible for introducing gene therapy interventions to the cardiac allograft that confer global transgene expression to the whole graft. Currently, there are two methods of ex vivo heart perfusion: hypothermic (4°C) and normothermic (>32°C). Hypothermic ex vivo perfusion involves pumping a cold crystalloid solution into the coronary arteries of the arrested heart to deliver oxygen and nutrients while removing toxic metabolites [108]. Normothermic ex vivo perfusion maintains the donor heart in a warm, contractile, near physiologic state during transport from the donor to the recipient. The donor heart is arrested prior to being placed on the perfusion device and then prepared by cannulation of the aorta and pulmonary artery and ligation of the superior vena cava and inferior vena cava. The circuit is primed with 1–1.2 L of donor blood mixed with a crystalloid perfusate solution. The cannulated heart is reanimated by pumping oxygenated blood mixed with the perfusate solution that enters the aorta to perfuse the coronary arteries. The coronary sinus effluent then crosses the tricuspid valve and gets pumped by the right ventricle into the cannulated pulmonary artery [109].

Gene delivery to a whole cardiac graft has been described in both small and large animal models utilizing ex vivo perfusion methods. Kypson et al. described a successful adenovirus-mediated transfer of the marker genes LacZ and Luciferase that was accomplished by flushing the rat heart before performing implantation of the heart into the recipient rat [110]. Similarly, utilizing a pig model Bishawi et al. described a successful adenovirus-mediated transfer of the marker gene Luciferase that was accomplished by utilizing the Organ Care System^™ ex vivo perfusion device (TransMedics, Inc) [56]. The porcine heart was perfused ex vivo with normothermic, sanguinous perfusate containing the adenoviral-luciferase vector for two hours prior to implantation into the recipient pig (Figure 2).

Figure 2.

Schematic for delivery of viral vectors to a cardiac allograft using normothermic, sanguinous ex vivo perfusion. The heart and blood are collected from the donor (a). The blood is then washed to remove any vector neutralizing components from the donor serum (B). The cardiac graft is perfused on the ex vivo perfusion device (C) and the viral vectors are added to the perfusion circuit to transduce cardiac graft (D). After completion of the perfusion/transduction period, the cardiac allograft is transplanted into the recipient pig in the heterotopic, intra-abdominal position.

There are several advantages that make ex vivo normothermic, sanguinous perfusion the ideal platform for translating gene therapy into clinical practice. The ability to recirculate the perfusate through the coronary arteries multiple times over a prolonged period of time optimizes the chances the delivery vectors attach to the target cells and enter. Normothermic perfusion provides a favorable environment for viral vectors to be able to efficiently transduce cells, enabling receptor-mediated vector entry and optimizing the downstream processes of transductions [107]. The main obstacle to overcome with this vector delivery modality is the use of whole blood from the donor to make the circulating perfusate. The presence of preformed antibodies to different viral vectors could effectively neutralize the ability of the viral vectors to achieve cellular attachment. One successful intervention to overcome this is the addition of a blood washing step prior to adding the donor blood to the perfusion device and this way remove any neutralizing blood components [56].

6. Conclusion

Gene therapy for cardiac transplantation promises to transform clinical practice in the near future with cardiac grafts that are more robust and lasting than ever. However, in order to achieve its widespread adoption, there are various factors that need to be taken into consideration for how to achieve successful vector delivery and transgene expression to the cardiac graft. Here, we discussed several considerations such as choice of vector, choice of the therapeutic gene, and choice of vector delivery mechanism. Just as important is the selection of the appropriate animal model for determining the efficacy and therapeutic effect of a gene therapy construct. The successful translation of gene therapy interventions for cardiac transplantation can potentially minimize or eliminate the incidence of post-transplantation complications and the need for systemic immunosuppression therapy.