*3.2. PRP Modifies Transjunctional Currents (Ij) and Gap Junctional Conductance (Gj) in Myofibroblast Pairs*

Next, the transjunctional currents (Ij) were analyzed in cell pairs in different experimental conditions by the dual whole-cell technique. Most of the fibroblast pairs cultured in PM exhibited families of Ij current traces with a nearly heterogeneous time course. Typical tracings obtained from a not differentiated cell pair cultured in PM are depicted in Figure 4A.

A minority of the cell pairs cultured in PM (about 20%) showed current records with a symmetrical time course for negative and positive Vj (not shown), indicating the involvement of homotypic GJs. Moreover, they also showed a linear Ij-Vj plot, suggesting the presence of not-voltage-dependent connexons. By contrast, the remaining 80% of the cell pairs investigated showed a non-linear time course. In particular, only 40% of these cells showed a symmetrical voltage dependence for positive and negative Vj, suggesting the involvement of homotypic GJs and voltage-dependent connexons; 60% of the cells showed asymmetrical voltage dependence (Figure 4A). This may suggest the dominant presence of heterotypic GJs in control not differentiating fibroblasts.

We then analyzed the time course of the Ij evoked in myofibroblast pairs. Typical tracings of the Ij evoked in cell pairs cultured in DM for 48 h or 72 h are shown in Figure 4B,C, respectively. About 25% of these cells showed linear not-voltage-dependent Ij. Notably, about 75% of the cell pairs cultured in DM for 48 h exhibited a non-linear time course, suggesting the prevalent expression of voltage-dependent connexons. Only 33% of this kind of cell pairs showed an asymmetrical voltage-dependence suggesting a minor presence of heterotypic voltage-dependent GJs in differentiating myofibroblasts. In contrast, 67% of this kind of response was symmetrical for negative and positive Vj, suggesting that the majority of the GJs involved in this myofibroblastic population were voltage-dependent and homotypic.

**Figure 4.** Time course of the transjunctional currents Ij, recorded from fibroblast and myofibroblast pairs in the absence or presence of PRP. (**A**) Representative Ij tracings (in pA) recorded in response to a bipolar pulse protocol applied to a fibroblast pair cultured in proliferation medium (PM). Note the asymmetrical time course between the two voltage polarities with a linear response for positive Vj. (**B**,**C**) Typical asymmetrical and almost voltage-dependent Ij tracings recorded from (**B**) a myofibroblast pair cultured in differentiation medium (DM) for 48 h and from (**C**) a myofibroblast pair in DM for 72 h. (**D**,**E**) Representative Ij recorded from a cell pair grown in (**D**) DM with PRP for 48 h (DM + PRP 48 h) and (**E**) for 72 h (DM + PRP 72 h). Note the completely linear and symmetrical responses in the latter condition.

After 72 h of culture in DM, we could observe an increase of the percentage of cell pairs with symmetrical voltage-dependent responses (about 80% in 72 h versus 67% in 48 h), indicating a progressive increase in the number of myofibroblasts exhibiting voltage-dependent and homotypic GJs.

The voltage dependence of Ij was evaluated by plotting the mean values of instantaneous currents, Ij,inst, (Figure 5A,C,E) and the steady-state currents, Ij,ss, (Figure 5B,D,F) as a function of Vj (Ij-Vj plot).

**Figure 5.** Voltage dependence of the transjunctional currents Ij, recorded from fibroblast and myofibroblast pairs in the absence or presence of PRP. (**A**–**F**) Transjunctional current values normalized for cell capacitance (in pA/pF), recorded from all of the fibroblast pairs cultured in proliferation medium (PM, continuous line, *n* = 6), and in differentiation medium (DM) for 48 h (*n* = 9) and 72 h (*n* = 8) plotted versus Vj. Panels A, C, and E show the Ij,inst values whereas panels B, D, and F show the Ij,ss values. Note that the plots related to DM show different slopes. (**C**,**D**) Comparison between Ij-Vj plot obtained from cell pairs cultured in DM for 48 h (open symbols, *n* = 9) and from cell pairs cultured in DM + PRP for 48 h (filled symbols, *n* = 9). Adding PRP to the culture medium for 48 h, altered the Ij voltage dependence, causing an almost complete linearity with voltage both for Ij,inst and Ij,ss. (**E**,**F**) Ij-Vj plots related to 72 h treatments. The presence of PRP in DM for 72 h (DM + PRP 72 h, filled symbols, *n* = 14), strongly reduced the mean normalized current amplitude observed in DM 72 h (open symbols, *n* = 8) and altered the Ij voltage dependence, causing an almost complete linearity with voltage both for Ij,inst and Ij,ss. All values represent mean ± S.E.M. \* *p* < 0.05 (unpaired Student's t-test).

Notably, from a qualitative point of view, the Ij-Vj plots showed a different shape according to the different culture conditions, namely PM and DM 48 h and 72 h (Figure 5A,B). Ij currents recorded in PM showed the smallest amplitude and a scarce deviation from linearity, especially for negative Vj, showing a similar slope for both the Vj polarities (Figure 5A,B). In contrast, the plot related to DM 48 h was almost linear and smoother for positive Vj (Figure 5A,B). The plot related to DM 72 h showed a kind of shoulder becoming S-shaped (Figure 5A,B). For negative Vj the resulting Ij-Vj plots showed a different steepness compared to that observed for positive Vj, and it was similar at any time in culture in DM for 48 h and 72 h (Figure 5A,B). Again, Ij data showing this asymmetrical Vj -dependence are indicative of heterotypic GJ channels. The Ij evaluated both at the peak (Ij,inst) and at the steady-state (Ij,ss) showed a progressively more marked voltage-dependence as the time in culture in DM increased. Based on this observation, we suggest a major involvement of voltage-dependent connexons during the differentiation time. Of note, when cells were cultured in DM + PRP for 48 h, about 75% of the cell pairs showed a linear time course (Figure 4D), even if this kind of response was not perfectly symmetrical for negative and positive Vj for all of the cell pairs investigated. When the mean values of all the Ij,inst and Ij,ss recorded were plotted versus Vj, the relation resulted approximately linear and symmetrical over the entire voltage range, clearly indicating the prevalence of not voltage-dependent homotypic GJ channels (Figure 5C,D) in this culture condition. On the other hand, the cells cultured in DM + PRP for 72 h exhibited only not-voltage dependent Ij (100%) (Figure 4E). Indeed, they showed a linear response and a perfectly symmetrical time course for positive and negative Vj. Again, the Ij-Vj plot analysis showed the Ij linearity with Vj and a marked symmetry, strongly indicating the presence of not-voltage-dependent homotypic GJ channels (Figure 5E,F). Remarkably, for the largest voltage steps applied, the mean current amplitudes recorded from cells cultured in DM for 72 h were statistically different (*p* < 0.05; multiple unpaired Student's t-test) to those estimated in DM + PRP at the same time. Of note, the comparison of Ij-Vj plots in Figure 5E,F with those in Figure 5C,D indicated that the maximal recorded normalized mean amplitude both of Ij,inst and Ij,ss resulted smaller in DM + PRP 72 h than in DM + PRP 48 h. For instance, the mean Ij,ss value estimated for the +150 mV step pulse was 2508 ± 566 pA/pF in DM + PRP 72 h and 7801 ± 7006 pA/pF in DM + PRP 48 h (Figure 5C,D).

Therefore, the features of Ij observed in the cells cultured in DM + PRP, such as the symmetry of the time course and the linearity of Ij,inst and Ij,ss versus voltage plots lead us to suggest a lessened contribution of voltage-dependent connexons in this condition.

To test for the current really flowing through GJs, we added heptanol (1 mM), a regularly used GJ channel blocker [59], to the bath solution during the recordings. The Ij was evoked from a cell pair cultured in DM, and then records were acquired from the same cell pair. After that, heptanol was acutely added to the bath solution. The recorded currents were significantly reduced compared to those elicited without heptanol. The mathematical subtraction of these two sets of traces gave the heptanol-sensitive current that is the one flowing through the GJs. In contrast, heptanol added during the recordings to cell pairs cultured in DM in the presence of PRP usually caused only a slight reduction of the current amplitude, suggesting a minor number of functional GJs allowing the current flow. A typical experiment related to cell pairs cultured in DM or in DM + PRP for 72 h is shown in Figure 6A. Only the current amplitude obtained by applying two representative voltage pulses (+130 and −30 mV) is shown as an example. Similar results were systematically observed for the bulk of cell pairs investigated. Since the heptanol sensitive current had a very small value in DM + PRP 72 h, we suggest a very small amount of functional GJs in the cells cultured in this experimental condition (Figure 6B–D).

**Figure 6.** Evaluation of the current flowing through the GJs by acute addition of heptanol under different experimental conditions. (**A**,**B**) Evaluation of Ij,ss (pA) related to a typical cell pair cultured for 72 h in differentiation medium (DM, white bars on the left side of each graph) or to a cell pair cultured for 72 h in DM + PRP (grey bars on the right). For clarity, only the current amplitude values obtained in response to a representative voltage pulse to +130 mV is reported in A, and to −30 mV in B. The current value recorded from the myofibroblast pair (DM 72 h) resulted clearly reduced after the acute addition of heptanol (HEPT, 1 mM) to the bath solution (HEPT DM 72 h). The current flowing through the GJs at the steady-state (HEPT-SENS CURRENT DM 72 h) is obtained by subtracting the current values recorded in the presence of heptanol from those recorded in DM alone. In contrast, the current amplitude recorded from the cell pair cultured in DM + PRP after acute addition of heptanol (HEPT DM + PRP 72 h) showed a slight reduction compared to DM + PRP 72h, having the heptanol-sensitive current a very small value (HEPT-SENS CURRENT DM + PRP 72h). Similar results were systematically observed for any voltage step applied in the bulk of the cell pairs investigated, suggesting a very small amount of functional GJs expressed under PRP treatment. (**C**) Characteristic time course of Ij (pA) recorded from a cell pair cultured in DM (DM 72 h) that exceptionally showed voltage independent features. (**D**) The same current traces recorded after acute heptanol addition (HEPT DM 72 h). (**E**) Resulting current traces (HEPT-SENS CURRENT DM 72 h) obtained by subtracting currents in D from currents in C, and representing the small heptanol-sensitive flux through the GJs. Note the different ordinate scale in E.

Noteworthy, even in those cell pairs cultured in DM for 72 h that exhibited not-voltage-dependent Ij, we could still measure a heptanol-sensitive current, suggesting that also the not-voltage dependent GJ functionality was hampered by the GJ blocker.

Finally, we analyzed the conductive properties of GJs. Intercellular current flow in cell pairs of fibroblasts and myofibroblasts were also used to study the dependency of the gap junctional conductance, Gj, on Vj by means of the Gj-Vj plot analysis. The related results are shown in Figure 7.

Data points obtained from cell pairs cultured in PM (Figure 7A) showed a horizontal distribution for positive Vj, suggesting the involvement of not-voltage-dependent GJs, in contrast to the not-linear distribution observed for negative Vj values. This asymmetrical voltage dependence of the Gj can suggest the prevalence of heterotypic GJs in proliferating fibroblastic cell pairs. These data points obtained from the cell pairs cultured in DM for 48 h (Figure 7B) did not follow a merely symmetrical relationship, being almost linear for negative Vj and more bell-shaped for positive Vj. Again, this may be consistent with the expression of more than one Cx isoform in myofibroblasts (possibly assembling in different combinations compared to those observed in PM) and hence confirms the involvement of heterotypic GJ channels in this cell population. After 72 h in DM, the Gj-Vj plot showed more or less the same shape as 48 h, but the Gj values for positive Vj resulted higher, suggesting a major contribution of the voltage-dependent component. In contrast, Gj-Vj plots related to the cells cultured in DM + PRP both at 48 h and 72 h were symmetrical and linear in any case, suggesting a lack of voltage-dependent GJs. This result suggests the involvement of homotypic GJ channels in this cell population.

**Figure 7.** Voltage dependence of the transjunctional conductance Gj. (**A**,**B**) Voltage dependence of the transjunctional conductance obtained by plotting Gj,ss/Gj,inst versus Vj related to (**A**) proliferation medium (PM) condition (open squares, *n* = 6), (**B**) differentiation medium (DM) condition at 48 h (open squares, *n* = 9) and 72 h (open circles, *n* = 9). These data obtained in DM show an asymmetrical distribution that becomes more bell-shaped for positive Vj. The treatment for 72 h gave higher values compared to 48 h, although not statistically significant (*p* > 0.05, multiple unpaired Student's t-test). (**C**,**D**) Symmetrical linear distribution observed under the concomitant treatment in DM + PRP at (**C**) 48 h (filled squares, *n* = 9), and (**D**) 72 h (filled circles, *n* = 14). All values represent mean ± S.E.M. Error bars are visible if they exceed the symbol size.

#### *3.3. PRP Reduces Cx43 Expression and Increases Cx26 Expression in Di*ff*erentiated Myofibroblasts*

Since the electrophysiological experiments showed an increase of Ij and Gj functionality during myo-differentiation of fibroblasts, we cultured the cells in DM in the presence of the GJ channel blocker heptanol (1 mM) for 72 h, to test the effective involvement of GJs in myofibroblast generation. We first analyzed the expression of α-sma in this experimental condition. As assessed by Western blotting and confocal immunofluorescence analyses, the cells exposed to DM + heptanol exhibited a clear reduction of α-sma expression (Figure 1A,B,F,G,I) compared to control differentiated myofibroblasts cultured in DM, supporting a key role of the GJs in the acquisition of myofibroblastic phenotype. Of note, the cells cultured in DM + PRP showed a more robust reduction of α-sma then those cultured in DM + heptanol, likely suggesting that PRP-induced prevention of fibroblast myofibroblast differentiation involves the activation of multiple molecular mechanisms. The cells cultured in DM + heptanol + PRP showed a significantly reduced (*p* < 0.05) expression of α-sma as compared to that observed in the cells exposed to single treatment (i.e., DM + heptanol or DM + PRP).

Then, taking into consideration the increasingly marked voltage dependence of the Ij recorded during differentiation time, we performed experiments aimed to evaluate the expression of the typical Cx types forming voltage- dependent connexins, namely Cx43.

We found that Cx43 expression both at mRNA and protein levels significantly increased (*p* < 0.05) with time in the cells cultured in DM as compared to control cells in PM as judged by RT-PCR (Figure 8A) and Western blotting analyses (Figure 8B), respectively.

**Figure 8.** Cx43 expression and localization during fibroblast to myofibroblast transition and related PRP effects. Fibroblasts were induced to differentiate into myofibroblasts by culturing in differentiation medium (DM) in the presence or absence of PRP for 48 h and 72 h. The cells cultured in proliferation medium (PM) were used as control undifferentiated cells. (**A**) RT-PCR analysis of Cx43 expression in the indicated experimental conditions. Representative agarose gel is shown. The densitometric analysis of the bands normalized to β-actin is reported in the histogram. (**B**) Western Blotting analysis of Cx43 expression. Histogram showing the densitometric analysis of the bands normalized to α-tubulin. (**C**–**E**) Representative superimposed differential interference contrast (DIC, grey) and confocal fluorescence images of the cells immunostained with antibodies against Cx43 (green) and

counterstained with propidium iodide (PI, red) to label nuclei. Scale bar: 30 μm. Scale bar in the inset in D: 15 μm. Arrows indicate the localization of Cx43 at the membrane level of two adjacent cells. (**F**) Histogram showing the densitometric analysis of the intensity of the Cx43 fluorescence signal performed on digitized images in 20 regions of interest (ROI) of 100 μm2 for each confocal stack (12). Data shown are mean ± S.E.M. and represent the results of at least three independent experiments performed in triplicate. Significance of difference: \* *p* < 0.05 versus PM; ◦ *p* < 0.05 versus DM 48 h; # *p* < 0.05 versus DM 72 h; § *p* < 0.05 versus DM + PRP 48 h (One-way ANOVA followed by the Tukey post hoc test).

Confocal immunofluorescence analysis confirmed the increase of Cx43 expression in differentiated cells after 48 h (data not shown) and even more after 72 h of culture in DM as compared to control undifferentiated cells in PM (Figure 8C,D,F). Moreover, we demonstrated the protein localization either at the cytoplasmic level or at the cell membrane level of adjacent cells (Figure 8C,D). To confirm the key role of this Cx isoform in fibroblast to myofibroblast transition, we silenced the cells for the expression of Cx43 by specific siRNA before culturing them in DM for 72 h (Figure 9A).

These cells exhibited a significant reduction (*p* < 0.05) of α-sma expression (Figure 9B,C,D,E,I) compared to cell cultured in DM, suggesting that Cx43 was required for the differentiation process. Of note, the cells cultured in DM + PRP concomitantly to reduced α-sma, showed a significant reduction of Cx43 expression (Figure 8A,B,E,F). This outcome was consistent with the electrophysiological data showing the reduction of voltage-dependent responses in these cells. Notably, according to the results of the experiments achieved in cells cultured with heptanol (Figure 1), the cells cultured in DM + PRP showed reduced expression of α-sma with respect to the cells silenced for Cx43 cultured in DM (Figure 9 B–I). Cells silenced for Cx43 expression and exposed to DM + PRP exhibited reduced α-sma expression levels as compared to those observed in the cells exposed to single treatment (i.e., DM + siRNA or DM + PRP) (Figure 9 B–I).

Finally, even the occurrence of a not-voltage-dependent response in myofibroblast pairs (although in a minority) as well as of the increase of this type of response after treatment with PRP, we analyzed the expression of a typical Cx type forming not/scarcely voltage-dependent connexons, namely Cx26. Western blotting (Figure 10A,B) and confocal immunofluorescence (Figure 10C–H) analyses demonstrated that Cx26 expression significantly increased in the cells after culture in DM for 48 h with respect to undifferentiated cells cultured in PM (*p* < 0.05). By contrast, the cells cultured in DM for 72 h exhibited Cx26 expression levels comparable to those of undifferentiated cells. The cells cultured in DM for 48 h in the absence or presence of PRP exhibited comparable levels of Cx26 expression (*p* > 0.05). Of note, the cells cultured in DM + PRP for 72 h exhibited a slight but significant increase (*p* > 0.05) of Cx26 as compared to cells cultured in the absence of PRP (Figure 10A,B,F,G,H).

**Figure 9.** Effect of inhibition of Cx43 expression on fibroblast to myofibroblast transition and related PRP effects. Fibroblasts were silenced by specific Cx43-siRNA duplexes and cultured in differentiation medium (DM) for 48 h in the absence or presence of PRP. SCR-siRNA duplexes were used as an internal control. (**A**,**B**) Western Blotting analysis of (**A**) Cx43 and (**B**) α-sma expression. The densitometric analysis of the bands normalized to α-tubulin is reported in the histograms. (**C**–**H**) Representative confocal fluorescence images of the cells double immunostained with antibodies against Cx43 (green) and α-sma (red). Scale bar: 25 μm. **(I)** Histogram showing the densitometric analysis of the intensity of Cx43 and α-sma fluorescence signal performed on digitized images in 20 regions of interest (ROI) of <sup>100</sup> <sup>μ</sup>m2 for each confocal stack (12). Data shown are mean <sup>±</sup> S.E.M. and represent the results of at least three independent experiments performed in triplicate. Significance of difference: \* *p* < 0.05 versus DM; ◦ *p* < 0.05 versus DM + CX43 − siRNA; # *p* < 0.05 versus DM + PRP (One-way ANOVA followed by the Tukey post hoc test).

**Figure 10.** Cx26 expression during fibroblast to myofibroblast transition and related PRP effects. Fibroblasts were induced to differentiate into myofibroblasts by culturing in differentiation medium (DM) in the presence or absence of PRP for 48 h and 72 h. The cells cultured in proliferation medium (PM) served as control undifferentiated cells. (**A**,**B**) Western Blotting analysis of Cx26 expression. (**A**) Representative blot. (**B**)Histogram showing the densitometric analysis of the bands normalized toα-tubulin. (**C**–**G**)Representative superimposed differential interference contrast (DIC, grey) and confocal fluorescence images of the cells immunostained with antibodies against Cx26 (green) showing the cellular localization of the protein. Scale bar: 25 μm. (**H**) Histogram showing the densitometric analysis of the intensity of the Cx26 fluorescence signal performed on digitized images in 20 regions of interest (ROI) of 100 μm2 for each confocal stack (12). Data shown are mean ± S.E.M. and represent the results of at least three independent experiments performed in triplicate. Significance of difference: \* *p* < 0.05 versus PM; ◦ *p* < 0.05 versus DM 48 h; # *p* < 0.05 versus DM 72 h (One-way ANOVA followed by the Tukey post hoc test).

#### **4. Discussion**

In recent years, great attention has been paid to the identification of new therapeutic agents and treatments that may promote the repair/regeneration of damaged skeletal muscle. In such a context, several in vitro and in vivo studies provided evidence supporting the advantage of the use of PRP for muscle regenerative purpose [30,66]. In this line, we have recently demonstrated the capability of PRP to either stimulate proliferation and differentiation of myogenic progenitors, including satellite cells [58], or prevent the TGF-β1 induced differentiation of fibroblasts towards myofibroblasts [48,57].

These data led us to suggest that PRP, if properly administered along the cascade of events through which skeletal muscle repair/regeneration proceeds (which also includes the physiological fibrotic reparative response), could exert a double beneficial effect on the healing of injured muscle. This may consist in the direct activation of the resident cells effectors of muscle regeneration, responsible for the formation of new muscle fibers and, in parallel, in the modulation/prevention of an excessive fibrotic response, thus contributing to the recreation of a more hospitable and conducive microenvironment for muscle progenitor functionality and thus the promotion of tissue regeneration. Experiments are ongoing in our lab aimed to assess the effects of PRP on differentiated myofibroblasts, by evaluating the capability of this blood product to modulate their fate. The results should be of interest to support the anti-fibrotic action of PRP. However, the ability of PRP in antagonizing fibrotic signaling pathways is still an issue of debate. Some reports show limited effectiveness or even inefficacy of this blood-derived product in counteracting the skeletal muscle fibrotic response [30,42,44,67–75]. The great heterogeneity of the available PRP formulation, PRP dosage and application timing represent critical points that may account for the reported conflicting results concerning the effects of this blood product in the modulation of skeletal muscle tissue fibrosis.

Based on these considerations, studies aimed to support the anti-fibrotic effect of this blood product are needed, as well as researches focused on the identification of the cellular and molecular target mechanisms of PRP, underpinning its action.

#### *4.1. PRP Counteracts Myofibroblast Generation*

According to findings from our previous studies and other research groups [30,47,48,76,77], here we have confirmed the ability of PRP to counteract the core cellular process of the fibrotic response, namely differentiation of fibroblasts towards myofibroblasts induced by the pro-fibrotic agent TGF-β1, based on: i) morphological and biochemical analyses showing that the cells treated with TGF-β1 in the presence of PRP did not acquire a mature myofibroblastic phenotype; indeed they rather appeared more spindle-shaped and showed either a reduction of type-1 collagen expression and a lower expression of α-sma, that was also less organized in filamentous structure as compared to differentiated cells; ii) the novel electrophysiological recordings of the membrane passive properties and gap junctional functionality, showing the ability of PRP to modify such parameters with respect to those recorded in differentiated myofibroblasts. Particularly, in the present experiments we observed that differentiated myofibroblasts tended to have a more positive RMP and PRP treatment counteracted this occurrence. The RMP is always critical for cell function since any small alteration of its value can substantially change cell excitability, contractility, and other properties, such as cell migration [21]. The less positive membrane potential registered in the cells induced to differentiate in the presence of PRP may counteract the depolarization of myofibroblasts and hamper their contractility, leading to an altered functionality. In this regard, it was shown that depolarization causes enhancement of ventricular myofibroblast contractility [21]. In this view, the present findings may suggest that PRP can revert myofibroblast RMP towards a more 'dormant' condition, counteracting their full differentiation.

In addition, PRP action opposed the increase of Rm value observed in differentiating conditions, showing its ability to revert the effect on the resting conductive properties induced by TGF-β1. The Rm parameter, corresponding to the reciprocal value of the membrane conductance, Gm, gives an idea of the total resting ionic fluxes across the membrane; thus, its physiological relevance is strictly linked to cell excitability. Moreover, the Cm value assumed as an index of membrane surface area increased in the

cells induced to differentiate as compared to proliferating cells. This observation was in agreement with the morphological analysis showing that the cells tended to increase their size upon TGF-β1-induced differentiation. Both phenomena were counteracted by PRP. Notably, the latter data are in accordance with the electrophysiological results described in our previous report dealing with the anti-fibrotic potential of relaxin [23].

### *4.2. Role of GJIC and Cx43 in Myofibroblast Generation*

The main relevance of our study is the contribution toward defining the molecular and functional mechanisms regulating TGF-β1 induced fibroblast-myofibroblast transition, highlighting the role of GJs in this process as well as the involvement of voltage-dependent connexin isoform, namely Cx43. In particular, we found that the majority of differentiated myofibroblast pairs exhibited an enhancement of Ij amplitude in the course of differentiation, suggesting an increased functionality of GJs, especially of the voltage-dependent ones, with increasing exposure time to TGF-β1. The role of GJs in this differentiation process was confirmed by the use of heptanol, a common GJ blocker. When the cells were induced to differentiate in the presence of heptanol, they actually failed to acquire a myofibroblast phenotype, indicating an essential role of functional GJs in the promotion of fibroblasts differentiation towards myofibroblasts. These data are in good accordance with previous studies showing that the selective blockade of GJs downregulated myofibroblastic phenotype [78,79]. Therefore, it can be stated that GJIC is of crucial importance in our cell model to regulate the fibroblast transformation towards myofibroblast. However, a possible role of such intercellular communications in the functional coupling of mature myofibroblasts to coordinate their activity can also be speculated [18,51–54]. In fact, while it is well accepted that myofibroblasts are responsible for the reparative scar formation and contraction, it is not clear yet whether they act individually or behave synchronically [78]. In this regard, we can propose that myofibroblasts can, at least in part, act as a coordinated functional syncytium, thus that hindering intercellular communication may represent a therapeutic target in diseases characterized by an overabundance of these contractile cells. Another important point is the ability of myofibroblasts to interact, by GJs, with other resident cell types of tissue globally affecting the organ functionality [80–82]. The analysis of the transjunctional conductance, Gj, gave some interesting information. Usually, the higher the Gj, the faster the current flows from a cell to the adjacent one, resulting in faster propagation speed [83]. The estimated Gj is the overall result of the total number of GJ channels docked between cells, the single-channel conductance of each GJ channel, and their functional states (fully open, sub conductance, or closed states). The GJ functional states can be dynamically modulated by chemicals and transjunctional voltage. The transjunctional voltage-dependent gating is an intrinsic property in all characterized GJs. In this study, we found that the Gj in differentiated myofibroblast pairs showed a progressively more marked voltage-dependence, suggesting a prevalent expression of voltage-dependent Cxs in myofibroblasts. From a functional point of view, this may reflect the need for myofibroblasts to be coupled in response to stimuli that cause membrane potential alterations. However, myofibroblasts likely need to express a kind of Cx, such as Cx43, whose trafficking, half-life, and regulation (by phosphorylation) can be maximally modulated during differentiation to myofibroblast [84]. Corroborating this suggestion coming from electrophysiological records, here we found that myofibroblasts showed an increased expression of Cx43. Furthermore, we found that cells silenced for the expression of Cx43 did not exhibit a mature myofibroblastic phenotype when cultured in differentiation condition in the presence of TGF-β1, suggesting the requirement of Cx43 for fibroblast-myofibroblast transition. Collectively these data are in accordance with the findings of our recent study [48] demonstrating an upregulation of mRNA expression of Cx43 in TGF-β1 treated cells (i.e., myofibroblasts). As well they agree with the studies by Asazuma-Nakamura and co-workers (2009) [85] and by Paw and co-workers (2017) [86] showing that Cx43 positively regulated myofibroblastic differentiation of cardiac and bronchial fibroblasts, respectively. In parallel, other studies showed that Cx43 expression is largely modulated during wound repair, and the modulation of Cx43 expression and gap junctional communication can be beneficial to

wound healing [87]. This process can be definitely altered by modulating Cx43 expression: wound closure can be delayed when Cx43 is overexpressed or accelerated when the levels of epidermal Cx43 are reduced [88–91]. In line with this, the transient blockade of Cx43 functions has been shown to reduce fibrosis as well as to promote experimental wound healing [89], and the normal GJ functions of Cx43 seems to be important for normal fibroblast function [92]. In this regard, multiple clinical trials are investigating Cx43 modulators and specific peptides targeting the intracellular loop, and the C-terminal tail region of this protein appear promising [93]. Based on our present findings, showing an increase of Cx43 expression in differentiated cells not only at the plasma membrane level but also in the cytoplasm, it is worth mentioning that a GJ independent function of Cx43 during fibroblast to myofibroblast transition cannot be excluded [59,94]. Taking into account the reported channel-independent influence of Cx43 on cytoskeleton remodeling and cell migration, we may speculate a similar role [94] for stress fiber assembly during fibroblast- myofibroblast transition.
