Electric communication between cells is mediated by bursts of the action potential and gap junctions provide the low resistance pathway for its cell-to-cell propagation. In addition gap junctions mediate flux of small molecules that regulate normal tissue development and tissue patterning. It is, therefore, reasonable to hypothesise that gap junctions can play a central role in the electromechanical incorporation of cardiac grafts. Autologous skeletal myoblasts are an attractive source for cardiac transplants because of their immune privileges, availability and non-tumourogenicity (reviewed in [21
]). However, when skeletal myoblasts differentiate into myotubes, they permanently lose expression of the major gap junctional protein, connexin 43, and, thus, do not have the apparatus for gap-junctional coupling. This is most likely to be the reason why, after transplantation of skeletal myoblasts, the newly developed graft has not been observed beating synchronously with the heart tissue [22
]. There was no intercellular transfer and electrical coupling between the cells developed from the transplanted myoblasts and the host cardiac myocytes [23
]. It was also reported that human patients with engrafted skeletal myoblasts suffered from ventricular tachycardia [24
]. Therefore, in pursuit of improved cardiac integration of skeletal muscle grafts we modified primary skeletal myoblasts by overexpression of the main gap-junctional protein connexin 43.
We have chosen a retroviral MLV vector for delivery and overexpression of the connexin 43 transgene into skeletal myoblasts because retroviral vectors are able to integrate stably into the genome of the transduced cells and thus to provide long-term expression of a transgene [26
]. Indeed, in our study, expression of connexin 43 and the EGFP marker in the transduced cells did not subside after 12 weeks of continuous passaging in tissue culture. However, long term gene expression in vitro
can be only tentatively projected to long term gene expression in vivo
, where transgene expression shutdown events are common [27
]. Therefore, one should consider the possible need of 'topping up' connexin 43 gene expression in the transplanted tissue in vivo
. As the engrafted cells do not actively divide, the choice of the 'topping up' vector is limited to cell division independent vectors, for example lentiviral HIV-based vectors. In this context employment of an MLV vector at the ex vivo
transduction step is beneficial because it allows subsequent use of efficient and MLV-compatible lentiviral vectors for additional transduction of the grafts in vivo
. If lentiviral vectors were used for initial transduction ex vivo
, there would be a possibility of a reduced efficiency of the 'topping up' transgene delivery in vivo
because of a CD4-independent superinfection interference of the resident lentiviral vector with the incoming one ([28
], reviewed in [29
Absence of connexin 43 in adult muscle can be due to shutdown of the native connexin 43 promoter during myoblast differentiation. We chose the MLV LTR and the CMV promoters to drive expression of human connexin 43 cDNA because these viral promoters are known to be able to direct transcription in adult muscle and therefore they are unlikely to shut down due to changes in the balance of transcription factors in the course of myoblast differentiation. A tandem arrangement of the two promoters in the MLV-CX43-EGFP vector could reduce the chances of transcription silencing after transplantation.
The EGFP marker of the generated viral vector MLV-CX43-EGFP was useful for the purification of the transduced cells by FACS and it can also be useful at the post-grafting stage to track the transplants in vivo.
Our observations show that confluent cultures of primary myoblasts can stay alive for at least a month in medium supplemented with FCS (and considerably longer in medium without serum). This property of primary myoblasts was in stark contrast to the transformed rat L6 myoblasts, which often died in a week after achieving confluency. Thus, fortuitously, the number of culture passages (and, therefore, cell divisions) required for our manipulation of primarily myoblasts (magnetic sorting, retroviral transduction, preparative FACS) was lower, than the number of passages needed for analogous manipulations with a permanent myoblast cell line L6.
To increase the yield of connexin 43 transduced skeletal myoblasts from a single muscle biopsy it is important to achieve a high efficiency of transduction by the MLV vector. However, preparations of amphotropic MLV commonly contain infection inhibiting substances, which reduce maximal transduction efficiencies without reduction of end-point virus titres [17
]. It is, therefore, important to optimise conditions for high efficiency of transduction. In our experiments we have shown for the first time that a combination of viral vector concentration on the plastic surface using the virus-binding protein RetroNectin®
] and transduction in the presence of lipid polycation Transfectam®
] is particularly effective for transduction of primary myoblasts by an amphotropic MLV vector. The achieved efficiency of transduction (38.30 ± 0.89%) can be further increased by: 1) improving the viral vector titre, for example by virion production at 32°C [30
]; 2) additional concentration of the viral vector, e.g. by using magnetic nanoparticles [31
] or low speed centrifugation [19
]; 3) increasing transduction competence of the recipient cells, for example by phosphate starvation of the myoblasts [17
] or boosting the myoblast division rate using growth factors [32
We have demonstrated that connexin 43 transduction of skeletal myoblasts and ensuing connexin 43 overexpression significantly improves propagation of action potential (measured as FAP activation rate) in co-culture of cardiac myocytes and skeletal myoblasts in vitro
. Enhanced gap junction formation between connexin 43 transduced skeletal myoblasts and cardiac myocytes is the most likely mechanism involved. This conjecture is supported by the results of Reinecke et al
] who reported that transplantation of genetically engineered myoblasts, which were designed to express connexin 43 during differentiation, resulted in close apposition of the skeletal myotubes and the host cardiac myocytes.
Skeletal myoblasts are non-differentiated muscle cells and, unsurprisingly, we did not observe any FAP activation in pure cultures of skeletal myoblasts, whether overexpressing connexin 43 or not. Thus, although fluorescent dye transfer occurred to a significantly greater extent in connexin 43 transduced skeletal myoblasts, improved gap-junctional communication in these cells did not result in FAP generation. We registered only a limited FAP activation in co-cultures of non-transduced skeletal myoblasts with cardiac myocytes. However, with connexin 43 overexpression in the skeletal myoblasts, the FAP activation rate in the co-cultures of skeletal myoblasts and cardiac myocytes was significantly enhanced, and was close to the FAP activation rate in pure cultures of cardiac myocytes.