The main finding of our study was that the transfection of the TBX3 plasmid and the addition of the small molecule SB431542 can lead to the differentiation of AD-MSCs into CPLCs, either given individually or combined with both of them. But in this study, the best results were shown in the group transfected with plasmid TBX3. Compared to the SM and TBX3 + SM groups, treatment using only TBX3 gene transfection resulted in cells expressing higher levels of the gene and protein markers of CPLCs. Gene transfection works by inserting a plasmid, TBX3, into the nucleus with the help of lipofectamine. Transfection of the plasmid was conducted on day 3 of 15-day differentiation, the stage before MSCs committed to becoming cardiac progenitor cells. MSC differentiation into cardiomyocytes has a shorter stage due to their nature as multipotent cells. Cardiac differentiation of MSCs starts with multipotent cells (day 0–4), cardiac progenitor cells (day 5–6), immature cardiomyocytes (day 7), and mature cardiomyocytes (after day 7) [23].

TBX3 is specifically expressed in the atrioventricular (AV) conduction system, and loss-of-function and gain-of-function experiments have demonstrated that TBX3 is required for cardiac conduction system development and homeostasis [24]. Study conducted by Zhao in 2020 [25], in which overexpression of the TBX3 gene was found to induce a reduction in the expression profile of working cardiomyocytes to become pacemaker cells. Vedantham [26] also described the function of TBX3 as a transcriptional repressor in pacemaker cells by silencing the expression of genes associated with force-generating cardiomyocytes and indirectly promoting the expression of genes in the pacemaker program. The TBX3 gene can also reduce intercellular coupling, reduce LK1 density to activate cardiac diastolic depolarization, and reprogram working cardiomyocytes [10]. Mohan et al. explored whether the AV conduction system is affected by TBX3 dose reduction through the characterization of electrophysiological properties and morphology of heterozygous TBX3 mutant (TBX3+/−) mouse hearts. They found PR interval shortening and prolonged QRS duration, as well as atrioventricular bundle hypoplasia after birth in heterozygous mice. The TBX3+/− AV nodes showed increased expression of working myocardial gene programs (mitochondrial and metabolic processes, muscle contractility) and reduced expression of pacemaker gene programs (neuronal, Wnt signaling, calcium/ion channel activity). Furthermore, they identified TBX3-dependent regulatory DNA elements active in the AV conduction system and validated the functionality of these elements. Deletion of a regulatory DNA elements drives expression of Cacna1g in the cardiac conduction system [27].

TBX3 not only suppresses differentiation but also induces the action of the HCN family, Cx45 and Cx30, and suppresses the action of Cx40 and Cx43 [10]. The results of the qRT-PCR analysis carried out on connexin genes such as Cx30, Cx40, and Cx43 showed the same results as research conducted by Zhao in 2020 [28], which demonstrated an increase in the expression of the Cx30 gene as well as a decrease in Cx40 and Cx43 gene expression. Cx30 contributes to the slowing of impulse propagation from the AV node and limits the maximum heart rate carried from the atria to the ventricles, thereby preventing rapid conduction of potentially worsening hemodynamics to the ventricles [26, 28]. In addition to this, Cx30 also integrates all cardiac pacemaker cells with different intrinsic frequencies [29].

HCN4 is the most expressed ion channel in the SA node. During development, HCN4 is initiated in the cardiac crescent and is progressively reduced and retained in the SA node during differentiation and in the adult heart. Inhibition of HCN4 regulation leads to lower expression of TBX3, SHOX2, BMP4, and Cacna1g in SA node development. This proves that an increase in TBX3 will affect HCN4 in the SA node, accompanied by the pacemaker phenotype that is still maintained [30]. In addition, cultured embryonic hearts of the transgenic mouse line which were treated with Bmp2, resulted in induction of endogenous AV canal genes (Tbx2, Tbx3) and reduction of chamber myocardium markers (Nppa, Nppb) [31]. The increase in HCN1 and HCN3 expression in our study can be explained by Ragunathan’s study [15], which revealed that when HCN4 is highly expressed by TBX3, HCN1 and HCN3 will be highly expressed as well [32].

The Ca2+ ions play a role in the regulation, flexibility, and contractility of the heart. The entry of Ca2+ ions through the channel not only plays a role in initiating cardiac excitation and contraction but also affects several crossing pathways that can change the membrane potential [29]. KCNN is an ion channel that plays an important role in Ca2+ exchange and the regulation of cardiac excitability [32]. One of the sub-families of KCNN is KCNN4. KCNN4 is an intermediate type of calcium-activated potassium channel that plays an important role in functional activity in the pacemaker [30]. Study conducted by Kleger [33], which showed an increase in the expression of Cx30 and KCNN4 in pacemaker-like cells. Raghunathan [15] also noted that there was an increase in the expression of specific pacemaker-like cell genes, including TBX3, KCNN4, Cx30, and BMP2.

Nodal inhibitors work by inhibiting the nodal protein, which is a protein that is secreted and binds to membrane proteins such as serine/threonine kinase receptors type I and II [34]. After activating the receptor, signal induction is then continued by the Smad2/3 protein, which in turn activates gene transcription along with the FoxH1 and Smad4 proteins, transcribing the nodal gene itself, the Lefty gene, and the PITX2 gene [35]. PITX2 protein activity can reduce the expression of the SHOX2 transcription factor, which plays a role in increasing TBX3 gene expression. From this pathway, we can see that the effect of adding nodal protein inhibitors does not directly increase TBX3 gene expression. It's different if we use gene transfection treatments, which have a direct effect on increasing TBX3 gene expression.

The combination of nodal inhibitor and TBX3 transfection did not show better results; it is also possible because of the accumulation of nodal inhibitor concentrations that were given continuously during the time of differentiation. This makes the cell viability decrease, as happened in the first batch. However, the results of the three treatment groups showed positive differentiation of MSCs into pacemaker-like cells. So, it is possible that adding small molecules is sufficient to increase TBX3 gene expression and differentiation of MSCs into pacemaker-like cells, although by transfecting the gene, the results of TBX3 gene expression were higher than other methods.

Electrophysiological examination using a patch clamp was also carried out to prove there were differences in the differentiation of the three treatment groups compared to the cardiomyocyte group. This difference was evident from the absence of spontaneous action potentials in the cardiomyocyte group, whereas in the TBX3, SM and TBX3 + SM groups, spontaneous action potentials were seen. This is similar to the study by Zhao et al. [28] that transfected TBX3 to iPSCs, also study by Sergei Yechikov et al. [5] that addded small molecules to iPSCs. From the morphological analysis of the stimulated action potential, it can be seen that the morphology of the action potential of the cardiomyocyte group looks more like an atrial action potential while in the treatment groups looks more like a pacemaker cell action potential. This is also in line by comparing the ratio of APD90 to APD50 where it was found in the cardiomyocyte group that the average APD90/APD50 ratio was 1.91 ± 0.05 which according to Rajamohan et al. [22] is the ratio of atrial action potentials, while in the treatment group the ratio of APD90/APD50 between 1.4 and 1.7 which is the ratio of the action potential of the pacemaker cells. Thus the results of the electrophysiological examination can provide support from the results of genetic expression examinations and protein analysis that there has been a change in the differentiation process from MSCs to typical cardiomyocytes of pacemaker cells in TBX3, SM, and TBX3 + SM group.

The study has limitation as we used single donor of AD-MSCs obtained from liposuction which was a stored biological material in the Stem Cells and Tissue Engineering (SCTE) laboratory, IMERI, Faculty of Medicine, University of Indonesia. The use of one donor sample in this study was by intention to minimize variability in this initial study and to save the cost. Glass et al. [36] studied age effect on gene expression by examining expression profiles in adipose tissue from 856 female twins aged from 39 to 85 years old. They found that the genes had lower levels of expression (50.8%) with age in adipose tissue. It was why we chose 19 years old healthy female as the source of AD-MSCs. However, the use of single cells sample is still sufficient for studies of gene expression differentiation [37]. Yet, we successfully produce high purity MSCs.

In conclusion, this study indicated that transfection of transcription factor TBX3 is capable to initiate differentiation of human AD-MSCs into CPLCs.

The first step, we currently are developing a bigger animal model with complete heart block that mimicking degenerative process. The animal model will serve as recipient of our CPLCs product. Dose effect correlation will be studied. Secondly, phase one clinical trial in limited subjects with partially dependent to pacemaker device will give important information of any AV conduction improvement. In the future, the cells might be used as a therapy to correct AV conduction abnormalities.