Recent advances and challenges on application of tissue engineering for treatment of congenital heart disease

Heart, a complex organ for tissue engineering

Unlike other human tissue, cardiac tissue is a more complex tissue considering not only because of its mechanical and structural function but also due to its electrical properties as well. The human heart mainly consists of cardiomyocytes, functions as a blood pump which is regulated by the electrical signal generated by the pacemaker cells in the sinoatrial nodes. This signal is directed and spread through the atrioventricular node to Purkinje fibers, and this is highly important for the direction of blood flow (Files & Boucek, 2012). The diastolic and systolic function of the heart is necessary to be synchronized and adjusted according to the body’s needs. The isolation of this signal from the rest of cardiac tissue is as important as this electrical signal by itself. This importance is achieved by the extracellular matrix (ECM) of heart which is also responsible for mechanical support and endurance (Files & Boucek, 2012). Aside from the functional aspects briefly mentioned above, the structural aspects (the cardiomyocyte and its ECM) are other factors for strong consideration in the success of TE. To yield an engineered tissue with the best functionality, this engineered tissue must be similar in every sense to the native tissue. The ECM of the heart is mainly a complex mesh of structural elements such as cardio fibroblasts and collagen fibrils, and non-structural elements such as proteoglycans, glycosaminoglycans, and glycoproteins (Kaiser & Coulombe, 2016; Rienks et al., 2014) among other components. Repairing the heart using materials which do not have or do not comply with the above characteristics and cannot work in harmony with the host-heart, will result in a non-efficient functioning heart accompanied by a series of complications. The concept and act of repairing the heart using various engineered techniques has evolved over the years from the use of artificial heart valves and grafts to bioprosthesis, and currently forward to the use of biomaterials and scaffolds cells.

Artificial heart valve and grafts

Charles Hufnagel was the first to experiment on animals with an artificial valve which he designed in the 1940’s (Gott, Alejo & Cameron, 2003). A few years later, the same type of valve was transplanted into humans (Gott, Alejo & Cameron, 2003). Nonetheless, Hufnagel’s valve required changes which were succeeded by Harken-Soroff and later by Starr-Edwards ball valve (Gott, Alejo & Cameron, 2003). Following several improvements, the latest version is made of pure carbon as a lighter, smoother material for blood flow and more durable in comparison to other materials (Zilla et al., 2008). Due to the position of the valve, there is a high transvascular pressure which leads to ‘impact wear’ and ‘friction wear’ (Zilla et al., 2008). Further complications with the use of valve replacements include inflammation around the prosthesis and calcification of the valve itself (Brown, 2005). The main disadvantage of artificial valves is the thromboembolic risk which leads to lifetime treatment with anticoagulants. This type of treatment involves various complications and brings different risks to patient (Zilla et al., 2008). Similarly, to valves, artificial vascular grafts have been used for many years as a surgical treatment for CHD (Chaikof, 2007). Nevertheless, they provoke an inflammatory response, and they are much less flexible than the body’s natural tissue (Torok, 2015). Originally porous fabric knitted of Dacron and polytetrafluoroethylene (PTFE) was used for stenosis treatment. Later alterations in the porosity of fabric were introduced to prevent the material’s corruption (Torok, 2015). The first attempt to combine biomaterials with patient’s cells was made by Wesolowki and colleagues by “preclotting the graft with the patient’s blood” (Torok, 2015). However, future attempts to produce successful results created many doubts about the actual function of cells following citation on clot (Torok, 2015). Nonetheless, both artificial heart valves and conduits require continuous use of anticoagulants. Remarkably, they do not grow with the patients’ heart as patients with CHD are highly likely to require surgical intervention when they are an infant or a child (Torok, 2015; Ratner et al., 2004).

Bioprosthetics: allograft, xenografts, and autografts

Allograft heart valves and arterial grafts are collected from deceased humans. In contrast, xenografts are harvested from porcine and bovine animals including heart valves and carotid arteries (Zilla et al., 2008; Perry et al., 2003; Brown, 2005). Allografts and xenografts came into the picture as an alternative to artificial valves and conduits. Their main advantage is that they do not require lifetime treatment with anticoagulants (Zilla et al., 2008; Hibino et al., 2010; Schmidt & Baier, 2010). Animal tissue is treated with glutaraldehyde (Hibino et al., 2010). Glutaraldehyde is a five-carbon bifunctional aldehyde used to stabilize tissue to protect from chemical and enzymatic degradation and maintain “its mechanical integrity and natural compliance” (Hibino et al., 2010). Also, treatment is necessary to reduce immunogenicity of the xenograft by decellularization and sterilization of the tissue (Hibino et al., 2010). Like xenograft, allografts also need treatment before transplantation and can even be cryopreserved (Dodson, 2014). Despite the great advantage, a number of complications are related to bioprosthetic grafts related to their preparation (Hibino et al.,  2010).

The risk of cytotoxicity leading to inflammation, as well as the partial loss of mechanical properties of tissue, have been reported (Hibino et al., 2010). Moreover, calcification is often observed in infants and children with bioprosthetics. Many efforts are being made to find out an alternative treatment for heart bioprosthetics therapy. However, there is a controversy regarding their efficiency (Hibino et al., 2010).

In a study carried out by Homann and colleagues, the outcomes of 25 years using allografts and xenografts for reconstruction of the right ventricular outflow tract showed 66% survival at 10- years’ follow up. Furthermore, patients with allografts had a mean reoperation-free interval time of 16 years in contrast to the xenograft recipients which this interval time is 10.3 years (Dodson, 2014). Allografts may present better outcomes, but they are not in abundance like xenografts. Therefore, many studies are concentrating on the development of tissue valves and vascular grafts created by stem cell-seeded on artificial or natural scaffold (Perry et al., 2003).

The relatively recent “CorMatrix” patch fabricated from the decellularized porcine small intestinal submucosa extracellular matrix (SIS-ECM) mainly composed of collagen, elastin, glycan, and glycoproteins have been introduced into cardiac surgery. SIS-ECM has not only been used in animal models for cardiac surgery (Dodson, 2014; Homann et al., 2000; Files & Boucek, 2012; Kaiser & Coulombe, 2016; Rienks et al., 2014; Brown, 2005; Chaikof, 2007; Ratner et al., 2004; Hibino et al., 2010; Schmidt & Baier, 2010; Kurobe et al., 2012; Johnson et al., 1998; Ruiz et al., 2005), but also in human studies, for cardiac and vascular reconstructions such as augmentation of the tricuspid valve (Wainwright et al., 2012), vascular repair of ascending aorta, aortic arch, right ventricular outflow tract, pulmonary artery, valvular reconstruction (Scholl et al., 2010), and closure of septal defects (Quarti et al., 2011). The study by Witt et al. (2013) reported a small risk of stenosis when SIS-ECM is used in the reconstruction of the outflow tracts and great vessels. Interestingly, the SIS-ECM has effectively proved the function in the high-pressure vessels (Quarti et al., 2011). The pitfall of this study, however, was the short follow up period.

A very common surgical practice for CHD is the use of pericardial patches for repairing the septal defect (Torok, 2015). The autologous pericardium is the best choice for infants as it is free, it does not provoke any immune-response, and it is sterile. Even though it requires some preparation before application, autologous pericardium creates less fibrotic tissue in comparison to Dacron (Torok, 2015). Allograft pericardium is available, but quite a few risk factors are associated with its use (Torok, 2015).

Biomaterials and scaffolds for tissue engineering

Generally, scaffolds work as a primary base for cells to enhance and produce relevant tissue. The scaffolds should have specific morphological, functional, and mechanical properties to support cells survival and differentiation (Feric & Radisic, 2016). Biomaterials used to produce scaffolds should be made of components which will accommodate the above characteristics and create a friendly cell microenvironment (Feric & Radisic, 2016). The previously mentioned characteristics apply for all different types of engineered tissue, and the goal is to mimic host tissue in the best possible way. With regards to artificial and bioprosthetic cardiac correction choices, scaffolds and biomaterials should contain various properties such as being biodegradable, biocompatible, flexible and durable and absence of immunogenicity and calcification. Due to the variety of sizes of patients’ hearts and defects, designed scaffolds should have various sizes and be able to grow and adapt to the heart (Schmidt & Baier, 2010). The biomaterial should allow neo-vascularization for adequate oxygenation of the tissue, create minimal scarring tissue and thrombotic risk, as the latter could lead to life treatment with anticoagulants (Witt et al., 2013; Miyagawa et al., 2011). Furthermore, these biomaterials scaffolds should be bioactive, meaning they should enhance cellularization in vitro and in vivo, and optimize cell efficiency and degrade at a desirable rate (Miyagawa et al., 2011). What’s more, biomaterials and scaffolds should be in abundance and cost-effective, as high cost could restrict development and use of it in TE as a routine therapeutic choice (Avolio, Caputo & Madeddu, 2015). The most common biomaterials for cardiac and vascular TE used today are synthetic, and natural polymers (Files & Boucek, 2012), and the electrospinning technique has been proven to be the most efficient way to produce scaffolds with these biomaterials (Williams, 2004).

Scaffoldless cell sheet

Another technique, which is independent of scaffolds, has been developed based on the cells’ ability to connect via cell-to-cell junction proteins and create ECM (Carrier et al., 1999). The cells are cultured in normal conditions at 37 °C in a temperature-responsive polymer cultures dish. When the culture temperature conditions change, the cells detach from the polymer culture dish as one cell sheet (Carrier et al., 1999). This technique was developed to avoid inflammatory reactions and fibrotic deposits in the area of graft where scaffold was placed following degradation (Carrier et al., 1999). A study has shown that contractile chick cardiomyocyte sheets could function effectively around rat thoracic aorta when applied on host myocardium (Witt et al., 2013). This cell sheet could synchronize within 1 h of implantation with the host tissue (Witt et al., 2013). Similar results have been shown in 3D structures using several cell sheets aiming to create a thick cardiac patch (Yamato & Okano, 2004). The number of sheets is limited, as more than three exhibits poor vascularization. However, the combination of endothelial cells and cardiomyocytes is being examined to promote vascularization before implantation (Carrier et al., 1999). Table 3 summarizes the pros and cons of the various materials and biomaterials used in tissue engineering.

Stem cells for tissue engineering

An equally important point in choosing a suitable biomaterial, scaffold or scaffoldless cell sheet, is the choice of the most appropriate cell types suitable for the TE. Stem cells (SCs) as a known cell source possesses the ability to differentiate toward Cardiac Muscle Cells/cardiomyocytes (CMCs), smooth muscle cells (SMCs) and endothelial cells (ECs), can regenerate cardiac tissue. Based on these properties, they play a key role in TE field. The currently used SCs in TE are summarized in Table 4.

Embryonic stem cells

Embryonic stem cells (ESCs) are one of the cell sources which are used in TE approaches. ESCs are derived from the inner cell mass of preimplantation blastocyst (Boroviak et al., 2014). They can differentiate into all different cell types of three germ layers. Human ESC (hESC) could be a good candidate for cardiac tissue engineering. In a study conducted by Landry et al., hESC-derived cardiomyocytes (hESC-CMC) showed very good phenotype including myofibril alignment, density, morphology, contractile performance and gene expression profile which, however, was only confirmed after 80–120 days in vitro culture (Lundy et al., 2013). Various groups apart from Landry and colleagues conducted studies to confirm the successful differentiation of ESC to cardiomyocytes as presented in a review by Boheler and colleagues in 2002 (Boheler et al., 2002). Duan et al. (2011) investigated how native cardiac ECM could affect hESC differentiation. This group processed porcine hearts to collect digested cardiac ECM which then mixed with collagen to create a hydrogel for cell cultivation purposes. The cultured hESCs on biomaterials comprised of 75% native cardiac porcine ECM and 25% hydrogel with no additional growth factors have shown a great differentiation with cardiac troponin T expression and contractile behavior, compared to the hydrogel with a smaller ratio of native ECM. Based on various studies, ESCs could be a good option for cardiac TE (Zimmermann, Melnychenko & Eschenhagen, 2004). Additionally, some factors such as ethical concerns, provoked immunogenicity, and risk of tumorigenesis make the ESCs a very controversial choice of cell source for TE (Schmidt & Baier, 2010).

Induced pluripotent stem cells

Another type of SCs which are used in TE is induced pluripotent stem cells (iPSCs). The iPSCs are somatic cells which are reprogrammed to behave like ESC and show the same properties. The iPSCs can differentiate into all three germ layers (Takahashi et al., 2007; Yu et al., 2007). Takahashi and his group was the first group who was able to reprogram the somatic adult cells like fibroblasts to iPSCs using viral vectors to introduce four key factors OCT4, SOX2, c-Myc, and KFL-4 to fibroblasts (Takahashi et al., 2007). This method was used to reprogram fibroblasts to embryonic-like cells and from this state differentiate them into a relevant type of cells (Takahashi et al., 2007). Ludry et al. have also shown that human iPSC-derived cardiomyocytes (hiPSC-CMCs) present the same characteristics as hESC-derived cardiomyocytes in long-term in vitro culture (Lundy et al., 2013). Lu and colleagues presented the successful repopulation of decellularized cadaveric mouse heart with hiPSC-derived multipotential cardiovascular progenitors (Lundy et al., 2013). They also demonstrate that the heart ECM promotes proliferation, differentiation and myofilament formation of CMs from the repopulated hiPSC-derived cells. Furthermore, they have checked the electrical coupling of these cells and also examined the constructive ability of repopulated heart using electrocardiogram which presented arrhythmia. Lu et al. have further examined the effects of pharmacological agents on repopulated heart and observed remarkable responses (Zimmermann, Melnychenko & Eschenhagen, 2004). This model is explored as an option to personalized medicine concerning drug testing/discovery (Zimmermann, Melnychenko & Eschenhagen, 2004; Bosman et al., 2015). Specifically, for CHD which presents such a variety of profiles, individual patient-specific human models’ development could help to understand how each patient would respond to existing pharmaceutical treatments. Even though it may not be possible to create actual organ heart models with individual clinical features of the disease (Caputo et al., 2015).

Nonetheless, like ESC, iPSC has demonstrated tumorigenesis (Schmidt & Baier, 2010). Ieda and colleagues showed a direct transdifferentiation of fibroblasts to functional cardiomyocytes using three key factors, Gata4, Mef2c, and Tbx5, within a very short time and suggested that direct reprogramming could reduce the risk of tumorigenesis (Ieda et al., 2010). Still, using viral vectors for reprogramming procedure is problematic and involves various risks (Schmidt & Baier, 2010). Therefore, today more different ways for iPSC production are being used and investigated to find out safer and more effective alternatives for this reprogramming procedure (Tarui, Sano & Oh, 2014; Lüningschrör et al., 2013). In the event this problem is solved, iPSC could be the safest type of cell sources for TE as they will not provoke any immune-response and cell harvesting procedure to produce iPSCs is not life-threatening for the patients. Moreover, iPSC raises less ethical concerns in comparison to ESC or fetal SC.

Prenatal, perinatal, and postnatal stem cells

Prenatal, perinatal and postnatal SCs are other cell sources used in TE, which include chorionic villi derived multipotent SCs, amniotic fluid-derived SCs (AFSCs), umbilical cord blood derived-endothelial progenitor cells (UCB-EPCs) and umbilical cord- or cord blood-derived- multipotent SCs (Webera, Zeisbergera & Hoerstrup, 2011). UCB progenitors, like endothelial progenitor cells (EPCs), have distinctive proliferative properties in comparison to other cells sources (Murohara et al., 2000). This category of SCs is exceptionally important as the child’s own SCs could be used for heart TE, for CHD patients who are diagnosed before birth. Immunogenicity or an additional procedure to harvest autologous SCs from the infant or child could be avoided in this procedure. Furthermore, it has been proven that AFSCs do not form teratomas in contrast to ESCs and iPSCs (De Coppi et al., 2007). All categories of these cells have been investigated with remarkable results on engineered valves and vascular grafts (Webera, Zeisbergera & Hoerstrup, 2011; Schmidt et al., 2006; Schmidt et al., 2004; Schmidt et al., 2007; Yao, 2016; Petsche Connell et al., 2013). This type of SCs is not applicable to adults who have been diagnosed later in life with CHD.

Adult stem cells

Cardiac progenitor cells

Cardiac progenitor cells (CPCs) or also known as Cardiac Stem Cells (CSCs) are a type of cells which are found in the adult heart and express stem cell factor receptor kinase c-kit+, also shown to be negative for the markers of blood/endothelial cell lineage (Beltrami et al., 2003). CPCs have only been in the spotlight for about 15 years now with the credit given to Beltrami and colleagues as they demonstrated the self-renewing and multipotent characteristics of these cells; differentiating into all three different cardiogenic cell types which are cardiomyocytes, smooth muscle cells and endothelial cells (Beltrami et al., 2003). More recently, a study carried out by Vicinanza and colleagues helped to lay to rest the controversies about c-kit+ cardiac cells by demonstrating that only a small fraction of the c-kit+ adult cardiac stem cells possess the tissue-specific progenitor properties (Vicinanza et al., 2017). The study was carried out in adult Wistar rats in which an experimental acute myocardial infarction (MI) was induced and the border of the infarcts were injected with GFP+c-kit+ for the first group, “GFP-expressing CD45-c-kit+ CSCs (CSC^GFP)” for the second group, and the placebo group were injected with PBS. The study showed that CSC^GFP only significantly reduced apoptosis and hypertrophy of the cardiomyocytes, it also significantly reduced scarring and improved ventricular functions (Vicinanza et al., 2017). The study further demonstrated that about 10% of the overall c-kit+ cardiac cells are CD45-c-kit+ , and only about 10% of these are clonogenic and multipotent. Therefore, it was inferred that only about 1–2% of the total c-kit+ myocardial cells have a demonstrable multipotent CSC phenotype (Vicinanza et al., 2017). These c-kit+ CD45-(deletion of which also renders the cells CD31-, CD34-) CPCs were also shown to express to some degree Sca-1, Abcg2, CD105, CD166, PDGFR- α, Flk-1, ROR2, CD13, and CD90 (Vicinanza et al., 2017). The number of CPCs have however been shown to be significantly higher in neonates but dramatically decreases after the age of 2 (Chaikof, 2007; Bosman et al., 2011). Thus, these CPCs in combination with a suitable scaffold, could be the answer to treat a number of CHD, as CPCs could be collected during palliative surgery or via endomyocardial biopsy (Saxena, 2010; Torok, 2015).

The potential SCs and biomaterials for TE in CHDs are represented in the Fig. 1 below. Also, Fig. 2 below shows the promising strategies for the treatment of CHDs.