Reaching out: junctions between cardiac telocytes and cardiac stem cells in culture

Abstract Telocytes (TCs) were previously shown by our group to form a tandem with stem/progenitor cells in cardiac stem cell (CSC) niches, fulfilling various roles in cardiac renewal. Among these, the ability to ‘nurse’ CSCs in situ, both through direct physical contact (junctions) as well as at a distance, by paracrine signalling or through extracellular vesicles containing mRNA. We employed electron microscopy to identify junctions (such as gap or adherens junctions) in a co‐culture of cardiac TCs and CSCs. Gap junctions were observed between TCs, which formed networks, however, not between TCs and CSCs. Instead, we show that TCs and CSCs interact in culture forming heterocellular adherens junctions, as well as non‐classical junctions such as puncta adherentia and stromal synapses. The stromal synapse formed between TCs and CSCs (both stromal cells) was frequently associated with the presence of electron‐dense nanostructures (on average about 15 nm in length) connecting the two opposing membranes. The average width of the synaptic cleft was 30 nm, whereas the average length of the intercellular contact was 5 μm. Recent studies have shown that stem cells fail to adequately engraft and survive in the hostile environment of the injured myocardium, possibly as a result of the absence of the pro‐regenerative components of the secretome (paracrine factors) and/or of neighbouring support cells. Herein, we emphasize the similarities between the junctions described in co‐culture and the junctions identified between TCs and CSCs in situ. Reproducing a CSC niche in culture may represent a viable alternative to mono‐cellular therapies.


Introduction
Regenerative medicine and specifically stem cell therapies show the greatest promise in correcting the extensive ventricular scarring and massive cell loss following ischaemic events in coronary heart disease. Different cell types have been investigated and even considered for regeneration of the injured heart in animal models and humans, of which notably: embryonic stem cells(ESCs), adipose-derived stem cells, mesenchymal stem cells, bone marrow-derived cells and CSCs [1][2][3][4][5][6][7][8][9]. Unfortunately, the results have been unconvincing and sometimes contradictory, due mainly to less than ideal cell delivery methods and poor cell engraftment and survival [10][11][12][13][14][15][16][17][18]. The very low survival rate is suggested to be caused by the notoriously 'hostile environment' of the infarcted region, characterized by ischaemia, a heightened inflammatory response, lingering pro-apoptotic signals and by the absence of an extracellular matrix [19,20]. In addition, successful stem cell engraftment appears to be dependent on the presence of various paracrine factors (e.g. growth factors, cytokines, chemokines) and extracellular vesicles, part of the so called 'secretome' released by stem cells and/or other neighbouring cells within CSC niches [14,[21][22][23][24]. It, therefore, becomes obvious that understanding the signalling mechanisms between support and stem cells, or even replicating the architecture of CSC niches in vitro prior to transplantation, is key for improving current therapies [25].
Among the cell types recently suggested to be involved in cardiac homoeostasis and regeneration are TCs [26][27][28][29]. Telocytes are distinct interstitial cells found in most organs, characterized by the presence of lengthy extensions -telopodes (Tp) [30][31][32][33][34]. Indeed, our group has shown that TCs form a complex network within the myocardium, communicating with adjacent cells both through direct physical contact as well as by means of paracrine signalling [35,36]. Cardiac TCs shed at least three different types of extracellular vesicles in situ [36,37] and in culture [38]. In an experiment involving Cy5-labelled oligoRNAs, these vesicles were successfully transferred to stem cells in culture, indicating a potential similar mechanism for the transfer of regenerative factors [39]. It is plausible that stem/progenitors cells are 'nursed' by TCs in CSC niches, thereby sustaining a continuous cardiac renewal process in the adult mammalian heart [35,36,40,41].
In this work, we demonstrate that TCs form 'atypical' junctions with stem cells not only in tissue but also in culture.

Cell culture
Cardiac TCs were isolated from the hearts of 3-month-old Wistar rats and cultured as described previously [34,39]. Laboratory animals were handled in accordance with the 'Victor Babeș' Institute Ethics Board guidelines. Rat CSCs were a kind gift from Prof. Piero Anversa (Brigham and Women Hospital, Boston, MA, USA). This cell line is described elsewhere [42,43].

Transmission electron microscopy
Cells cultured for 24 and 48 hrs, were fixed and embedded in epoxy resin (Agar 100) as described previously [38]. Ultra-thin sections (~60 nm) were then obtained using a diamond knife, double stained with 1% uranyl acetate and Reynolds lead citrate and finally visualized using a Morgagni 268 TEM (FEI Company, Eindhoven, The Netherlands) at 80 kV. Digital electron micrographs were recorded using a MegaView III CCD. Image processing was done using iTEM-SIS software (Olympus, Munster, Germany). Samples originating from tissue [41] were used for comparison.

Results
Telocytes were identified in culture ( Fig. 1) based on the presence of very long extensions (Tp), with lengths varying from 100 to 200 lm [38]. Telopodes were frequently observed in apparent contact with either CSCs or other TCs (Fig. 1). Heterocellular contacts markedly increased in number after 48 hrs of culture and this corresponded to an increase in the number of CSCs (Fig. 1B) in the presence of TCs.
Telopodes originating from different TCs frequently formed homocellular networks, establishing both adherens (AJ) and gap junctions (GJ) at the site of contact (Fig. 2) and this was consistent with previous observations of TCs in tissue [36]. Gap junctions were easily observed connecting cultured TCs to other TCs (Fig. 2C), however, not TCs to CSC. Interestingly, no GJs could be identified between TCs in the cardiac tissue.
Telocytes formed junctions with adjacent CSCs ( The Puncta adhaerentia type junction was infrequently found connecting TCs and CSCs (Fig. 4C). Telocytes and CSCs also maintained contact by means of paracrine signalling, via different types of extracellular vesicles (Fig. 5A), however, these are described elsewhere [38].
Intercellular contacts between TCs and CSCs were also analysed in tissue samples for comparison. The stromal synapse associated with electron-dense nanostructures (Figs 5B and 6) was the most common type of junction found in tissue. The intercellular distances in tissue were between 20 and 30 nm and the minimum distance between the cellular membranes was approximately 15 nm. Contours drawn over the opposing cellular membranes emphasize the similarities of heterocellular junctions between the TCs and CSCs in culture and in tissue (Fig. 7).

Discussion
Here, we confirm that TCs display a similar morphology in culture as they do in tissue and form the same types of cell-to-cell junctions between themselves, as well as with CSCs, supporting the idea of an active involvement in the homoeostasis of cardiac tissue. This study    suggests that replicating the architecture of CSC niches in vitro may be possible.
In tissue, TCs may be identified as a result of their peculiar morphology, with Tp forming vast cellular networks, and these character-istics were exhaustively discussed in a previous review from our group [31]. In cardiac tissue especially, TCs were suggested to function as 'nurse-cells' facilitating the maturation of cardiac stem/progenitor cells [35,36,40]. However, although recent cell-culture data  indicated that TCs communicate over long distances with stem cells through extracellular vesicles [39], the direct contacts between TCs and other cell types in culture remained largely unexplored. Understanding cardiac heterocellular communication (Fig. 8), both direct [36] and through paracrine signalling [38,39] may prove essential for successful cell-based therapies.
Telocytes formed homocellular networks by means of GJs, however, no such junctions could be seen between TCs and CSCs in culture. This may have been the result of the relatively short culture times, but more likely represents a characteristic of the partnership between the two types of cells. Similarly, embryonic fibroblast feeder cells were not found to form GJs with human ESCs and mouse ESCs [44]. In fact, heterocellular GJs are not normally seen between different cell types. Contrary to a previous report indicating that fibroblasts establish GJs with cardiomyocytes in tissue and in culture [45], our group has not been able to visualize such junctions between TCs and cardiomyocytes in tissue [36], reinforcing the idea that GJs are restricted to homocellular communication.
Telocytes formed both homocellular, as well as heterocellular AJs with CSCs. Classic AJs formed between stem and support cells in epithelial and stromal niches were previously shown to control the geometry of division, by facilitating the proper positioning of centrosomes [46]. As we have seen an increase in the number of CSCs in the presence (but not the absence) of cardiac TCs, it could be speculated that similar mechanisms take place in CSC-TC co-cultures, however, this hypothesis warrants further study. Puncta adhaerentia, consisting of cadherin-catenin clusters, are found at cell-cell contacts often during early stages of AJ assembly [47] and such junctions were also seen in the case of TC-CSC cultures.
Additionally, previous studies from our group [36,48] showed that heterocellular contacts in tissue can occur by means of stromal  synapses and these were also observed in TC-CSC co-cultures. Stromal synaptic regions, akin to immune synapses [49,50], are defined by intercellular distances within the molecular interaction range (15-100 nm), allowing the presence of a gap (synaptic cleft) in which receptor-ligand interactions occur. Consistent with the observations of TCs and CSCs or cardiomyocytes in tissue [36,48,51], electrondense nanostructures connecting the two apposing membranes over distances shorter than 25 nm were frequently associated with these types of junctions. The molecular composition of these nanostructures is unknown.
The exact role of TCs in cardiac tissue regeneration remains an elusive, but desirable target, with potential applications in stem cell therapy. Future in-depth studies of the molecular components of stromal synapse and the dynamics of these components in normal and diseased tissues will shed a new light on the signalling mechanisms between stem and support cells (Fig. 8). This, in turn, may help over-come the current limitations of stem cell engraftment in myocardial tissue following ischaemic events.