Dynamic Epicardial Contribution to Cardiac Interstitial c-Kit and Sca1 Cellular Fractions

Background: The cardiac interstitial cellular fraction is composed of multiple cell types. Some of these cells are known to express some well-known stem cell markers such as c-Kit and Sca1, but they are no longer accepted to be true cardiac stem cells. Although their existence in the cardiac interstitium has not been disputed, their dynamic throughout development, specific embryonic origin, and potential heterogeneity remain unknown. In this study, we hypothesized that both c-KitPOS and Sca1POS cardiac interstitial cell (CIC) subpopulations are related to the Wilms’ tumor 1 (Wt1) epicardial lineage. Methods: In this study, we have used genetic cell lineage tracing methods, immunohistochemistry, and FACS techniques to characterize cardiac c-KitPOS and Sca1POS cells. Results: Our data show that approximately 50% of cardiac c-KitPOS cells are derived from the Wt1-lineage at E15.5. This subpopulation decreased along with embryonic development, disappearing from P7 onwards. We found that a large proportion of cardiac c-KitPOS cells express specific markers strongly suggesting they are blood-borne cells. On the contrary, the percentage of Sca1POS cells within the Wt1-lineage increases postnatally. In accordance with these findings, 90% of adult epicardial-derived endothelial cells and 60% of mEFSK4POS cardiac fibroblasts expressed Sca1. Conclusion: Our study revealed a minor contribution of the Wt1-epicardial lineage to c-KitPOS CIC from embryonic stages to adulthood. Remarkably, a major part of the adult epicardial-derived cell fraction is enriched in Sca1, suggesting that this subpopulation of CICs is heterogeneous from their embryonic origin. The study of this heterogeneity can be instrumental to the development of diagnostic and prognostic tests for the evaluation of cardiac homeostasis and cardiac interstitium response to pathologic stimuli.


INTRODUCTION
The vertebrate heart is composed of a plethora of different cell types. Around 70% of these cells reside in the cardiac interstitium (CI), the extracellular space between cardiomyocytes (Bergmann et al., 2015;Pinto et al., 2016). All these cells are jointly referred to as cardiac interstitial cells (CICs) and are known to play an essential role in myocardial embryonic development, and adult homeostasis (Krenning et al., 2010;Pérez-Pomares and de La Pompa, 2011;Pogontke et al., 2019), and in the adaptive responses of the heart to pathological conditions (Takeda et al., 2011;Ruiz-Villalba et al., 2015).
The CIC population is highly heterogeneous. It comprises multiple cell types such as endothelial cells (ECs), vascular smooth muscle cells (vSMCs), cardiac fibroblasts (CFs), pericytes, and circulating cells among others (Ruiz-Villalba et al., 2015;Pinto et al., 2016;Sampaio-Pinto et al., 2020). These different cellular pools are not, however, equally represented. Indeed, ECs, which approximately represent 58% of all CICs, are very abundant. This finding is in accordance with extensive myocardial vascularization by coronary vessels, including arteries, veins, and a massive capillary bed (Tomanek et al., 2010). Furthermore, the coronary endothelium has a key scaffolding role during embryonic interstitium formation (Pogontke et al., 2019) since coronary EC interaction with vSMCs, pericytes, and CFs is necessary for the building of periendothelial coronary domains. Moreover, due to the cellular complexity, a niche role for the periendothelial coronary milieu has been proposed (Fioret et al., 2014;Pogontke et al., 2019). Unlike EC, other cell types are far less frequent in the CI. For example, the small numbers of blood-borne cells have been described to permanently or transiently reside in the adult cardiac interstitium in the absence of evident pathological stimuli (Epelman et al., 2014;Molawi et al., 2014). The functions played by these cells in the healthy heart are poorly known, and data on the specific diversity and the spatiotemporal distribution of these cells are scarce (Pinto et al., 2016;Farbehi et al., 2019).
Other sparse cells holding an intrinsic multipotent differentiation potential were reported to be present in the cardiac interstitium. These cells, frequently dubbed cardiac stem cells (CSCs), were thought to be the origin of de novo cardiomyocyte differentiation in the adult heart. A bone marrow origin for these cells was discarded by some authors based on the absence of CD45 expression (Limana et al., 2005;Bearzi et al., 2007) but not by other researchers (Zhou et al., 2010). CSCs were originally identified based on their characteristic expression of molecules like Bmi1, Abcg2, Isl-1, Sca1, or c-Kit (Martin-Puig et al., 2012;Anversa et al., 2013), all of which are known to be expressed by other organresident, well-characterized stem or progenitor cells (Bradfute et al., 2005;Thoren et al., 2008). Of these, c-Kit-and Sca1expressing CSCs have been the most widely studied. Both molecules are associated with stem/progenitor cell properties and are often considered markers for these cell types. The membrane-bound stem cell factor (SCF) receptor c-Kit (a.k.a. CD117) is expressed by hematopoietic cells but also by other cell types, many of which do not display stem/ progenitor cell properties (Lennartsson and Rönnstrand, 2012;Liang et al., 2013). Similarly, Sca1 (Stem Cell Antigen 1) is expressed in different cell populations, including blood and cancer cells, and only a fraction of Sca1-expressing cells has been unambiguously shown to represent a stem/progenitor cell fraction (Holmes and Stanford, 2007).
As indicated, since 2003, cardiac c-kit POS cells have been thought to be a self-renewing, clonogenic, and multipotent population of CSC. These cells were described to differentiate into a minimum of three different cardiogenic cell lineages (myocytes, smooth muscle cells, and endothelial cells) in vitro (Beltrami et al., 2003) and, when grown in non-adherent in vitro assays, c-Kit POS CSCs were shown to form cardiospheres (Vicinanza et al., 2017). Most importantly, these cells were characterized, isolated, expanded, and used in cell-based experimental therapies to treat the diseased heart (Gude and Sussman, 2018;Keith and Bolli, 2015). Intensive recent research, however, has revealed that the myocardiogenic potential of these c-kit POS CSCs is negligible (Van Berlo et al., 2014;Sultana et al., 2015), so their actual multipotency has been refuted. As for cardiac Sca1 POS cells, different studies have shown that they robustly differentiate into endothelial cells in vitro after treatment with VEGF or PDGF BB (CD31 POS , vWF POS , caveolin POS ), and form endothelial tubules when cultured in 3D matrices (e.g., Matrigel) (Takamiya et al., 2011;Wang et al., 2006). More recently, these cells have also been associated with cardiac endothelial cells and fibroblasts in vivo using a transgenic model for lineage tracing (Tang et al., 2018).
Regardless of their multipotent potential, the existence of cardiac cells expressing molecular markers classically associated with stem cells such as c-Kit and Sca1 has not been disputed. However, their dynamics along with embryonic development and adulthood, specific origin, nature, and function remain obscure. In this work, we aimed to progress in the characterization of these elusive cells by the evaluation of a possible epicardial origin for, at least, part of them, also considering the bone marrow/circulation as a plausible source for cardiac c-Kit POS and Sca1 POS cells. This working hypothesis is supported by the well-known contribution of the embryonic epicardium to the adult heart resident CICs pool (Ruiz-Villalba et al., 2015;Volz et al., 2015;Cano et al., 2016). In order to tackle this objective, permanent genetic tracing of epicardial cell derivatives has been carried out taking advantage of strong expression of the Wilms' tumor suppressor gene (Wt1) in embryonic epicardial cells and by carefully analyzing tissues using both immunohistochemistry and FACS.

Animal Models
All animals used in this study were handled in compliance with institutional and European Union guidelines for animal care and welfare under a specific experimental procedure approved by the Committee on Ethics of Animal Experiments of the University of Málaga/BIONAND.
For lineage-tracing studies, homozygous Tg(WT1-cre) #Jbeb line (Wt1Cre from now on), in which an IRES/GFP-Cre cassette was inserted 17 bp downstream of the translation stop site of the Wilms' Tumor Gene 1 (Wt1) gene was used as driver line (Del Monte et al., 2011). The B6.129X1-Gt(ROSA) Representative dot blots and cytograms obtained were obtained from the FACS analysis showing the sequential gating used for the characterization of CD31 NEG / CD45 NEG cardiac interstitial cells. (C) Representative dot blot was used for the characterization of c-Kit and YFP expression within the subpopulation of CD31 NEG / CD45 NEG cells. The same gating was applied for all the different stages included in this study (above). The bar plot representing the percentages of CD31 NEG /CD45 NEG / c-Kit POS cells present in all stages considered; YFP NEG (black bars), YFPPOS (green bars). (D) Representative dot blot for the characterization of CD34 POS cells within the CD31 NEG /CD45 NEG /c-Kit POS subpopulation (left). The bar plots indicated the percentage of CD34 POS cells found in the CD31 NEG /CD45 NEG /c-Kit POS /YFP POS (green (Continued ) Frontiers in Cell and Developmental Biology | www.frontiersin.org May 2022 | Volume 10 | Article 864765 3 26Sor tm1(EYFO)Cas/J line (Rosa26R-eYFP from now on), that contains the enhanced yellow fluorescent protein gene downstream of a loxP-flanked stop sequence, was used as reporter line (Srinivas et al., 2001). Both lines were crossed in homozygosis, obtaining Wt1Cre-YFP POS mice in which Wt1driven Cre activity mediates the excision of the LoxP-flanked STOP sequence in R26R mice, activating permanent reporter enhanced yellow fluorescent protein (eYFP) expression in the Wt1 POS cell linage.
For conditional tracing studies, homozygous Wt1 tm2(cre/ERT2)Wtp/J mice (Wt1Cre/ERT2 from now on), in which exon 1 of the Wt1 locus was replaced by a CreERT2 fusion gene, were crossed with homozygote Rosa26R-eYFP, obtaining Wt1Cre/ERT2-YFP POS , where eYFP expression is permanently activated in the Wt1 POS cell lineage from the moment of the chemical induction using tamoxifen (Sigma) (Zhou et al., 2008). Tamoxifen was dissolved in corn oil (Sigma) at 1 mg/ml and 50 µl were injected into newborns by intragastric injection at P1, P2, and P3 (Lizen et al., 2015). An equivalent volume of corn oil was administered to control animals. The processing of the pup's hearts was performed at P7. The embryos were staged considering the moment of vaginal plug observation, which was designated as E0.5.

Flow Cytometry
Wt1Cre-YFP POS and Wt1Cre/ERT2-YFP POS hearts were dissociated through two sequential steps (10 min) in 37°C pre-warmed liberase TH (0.125 mg/ml, Roche). After complete digestion, the samples were filtered to eliminate cardiomyocytes (Cell Strainer 40 μm filters, BD) and red blood cells were lysed (Roche). The cell samples were then incubated in the proper primary antibodies (see Supplementary  Table S1) diluted in 2% BSA, 1 mM EDTA in PBS for 15 min, using non-immune IgGs or PBS as negative controls (Supplementary Figure S1). For the exclusion of non-viable cells in flow cytometry analysis, 7AAD (EBioscience) was employed. The results were recorded in a Gallios flow cytometer (Beckman Coulter) and analyzed using the Kaluza software.

Statistical Analysis
For FACS studies, the statistical significance was analyzed by the Student's t-test, and shown as mean plus standard deviation (p < 0.05). Three biological replicates were used in embryonic samples (5 littermates were pooled for each replicate for E15.5, and 3 littermates for E18.5). Three individual samples were used for each neonatal time point and eight for adults, without significant differences derived from the gender.  Figure 1A). In order to characterize Wt1 lineage-derived cardiac interstitial cells (CICs) population, we first FACS screened the colocalization of c-Kit and YFP in the CD31 NEG /CD45 NEG interstitial cell fraction. Our analysis showed that the CD31 NEG /CD45 NEG CIC pool, which excludes endothelial and circulating cells, contains a low c-Kit POS CICs number (between 1.6 and 22%, see Table 1). Notwithstanding, the proportion of CD31 NEG /CD45 NEG /c-Kit POS significantly varies through time, peaking at P1 (around 22% of the total CD31 NEG /CD45 NEG CICs) ( Table 1). At E15.5 almost 14% of CD31 NEG /CD45 NEG CICs coexpressed c-Kit. Of them, almost half were YFP POS (43%, see Figure 1C; Table 1), suggesting a significant contribution of Wt1-lineage to the CD31 NEG /CD45 NEG CIC population (6.33 ± 2.06% of CD31 NEG /CD45 NEG were c-Kit POS /YFP POS and 8.31 ± 5.13% c-Kit POS /YFP NEG ) ( Table 1). From E15.5 until P1, the proportion of CD31 NEG /CD45 NEG /c-Kit POS /YFP POS cells was reduced to 10% of the total c-Kit POS CICs present in the heart. Finally, no CD31 NEG /CD45 NEG /c-Kit POS /YFP POS CICs were detected from P7 onwards ( Figure 1C; Table 1).
In order to characterize both subpopulations of c-Kit POS CICs (YFP POS vs. YFP NEG ), we analyzed the presence of CIC surface markers often associated with multipotency in literature, such as CD34 and Sca1. Both the YFP POS and YFP NEG cell subpopulations showed a similar proportion of CD34 expressing cells along with embryonic development ( Figure 1D). The percentage of CD31 NEG /CD45 NEG /c-Kit POS / YFP POS /CD34 POS CICs is however reduced at P7, disappearing in the adult ( Figure 1D). The percentage of Sca1 POS cells in the c-Kit POS /YFP POS fraction was negligible, and in the c-Kit POS / YFP NEG one was highly variable (11.48 ± 11.51%, Figure 1E), indicating that both subpopulations are different cell types.
After the preliminary evaluation of c-Kit POS CICs relationship with the Wt1 cell lineage, we decided to explore a potential alternative origin for these cells. Based on the published evidence of blood-borne cell contribution to the cardiac interstitium Pinto et al., 2016;Sampaio-Pinto et al., 2020), we investigated whether CD45, a panleukocyte marker for circulating cells, was expressed (Supplementary Figure S2A). Our analysis showed that the number of CD45 POS /c-Kit POS cells decreased from 65% to 13% between E15.5 and adulthood (Table 1). Surprisingly, a significant number of CD45 NEG /c-Kit POS /YFP NEG CICs were enriched in cell surface markers classically associated with different subpopulations of circulating cells, such as myeloid cells (CD11b), monocyte/ macrophages lineage (CCR2), macrophages (F4/80), and some lymphoid cell types (CXCR3), strongly suggesting an extracardiac origin for these cells ( Figure 1F). Some of these circulating cell markers were also present in CD45 NEG /c-Kit POS /YFP POS CICs in perinatal stages, in accordance with the described expression of Wt1 in a restricted population of adult blood cells (Alberta et al., 2003;King-Underwood et al., 2005). Finally, we found a reduced contribution of c-Kit POS cells into CD31 POS endothelium from E15.5 (around 11%) to adulthood (around 1.6%) (Supplementary Figure S2B). In both the CD45 POS and the CD31 POS /c-Kit POS subpopulations, the co-expression with YFP was restricted.
To validate these FACS data as well as to identify the specific anatomical distribution of c-Kit POS cells, we analyzed the localization of c-Kit in the heart using immunohistochemistry. We could not find any c-Kit POS cells in any embryonic hearts subpopulations, respectively. (F) Immunolocalization of SCA1 staining in the representative sections from E13.5, 17.5, P1, and adult Wt1Cre-YFP hearts. SCA1 POS / YFP NEG (red, arrowheads) and SCA1 POS /YFP POS cells (yellow, arrows) home at the cardiac interstitium. Abbreviations: CD31/PECAM1, platelet, and endothelial cell adhesion molecule 1; CD34, cluster of differentiation 34; CD45, protein tyrosine phosphatase receptor type C; c-Kit, KIT proto-oncogene receptor tyrosine kinase; DAPI, 4′,6-diamidino-2-fenilindol; FS, forward scatter; Sca1, stem cell antigen-1; SS, side scatter; V, lumen of coronary vessels; YFP, yellow fluorescent protein. Each experimental group includes the following biological replicates: 3 litters for embryonic stages (5 or 3 embryos peer biological replicate for E15.5 and E18.5, respectively); 3 individual, biological replicates for P1 and P7; 8 animals, biological replicates for adults). Scale bars: 25 μm *p value ≤0.05; **p value ≤0.01; ***p value ≤0.005. before 13.5 days of development (data not shown). At E13.5, most c-Kit POS cell co-expressed CD45 (arrows in Supplementary  Figure S2C), and a few of them CD31 (arrows in Supplementary Figure S1D) by immunohistochemistry. From E13.5 onwards both CD45 POS /c-Kit POS and CD45 NEG /c-Kit POS CICs were always located in the subepicardial of embryonic ventricles (Figure 2A). No c-Kit POS /YFP POS cells were found in any of the analyzed hearts (n = 2 per stage), except at P1 (arrows in Figure 2). It is important to note that most postnatal c-Kit POS cells were located either close to coronary vessels or incorporated into the surface of the coronary adventitial layer ( Figure 2B).

Wt1 Cardiac Lineage Significantly Contributes to the Interstitial Sca1 POS Cell Fraction
To evaluate the contribution of Wt1-derived cells to the cardiac Sca1 POS cell population, we performed FACS at the same stages selected for the study of c-Kit POS CICs ( Figure 1A) using the same gating strategy ( Figure 3A). Sca1 was detected in a small percentage of CD31 NEG /CD45 NEG CICs only, although the amount of Sca1 POS cells significantly varied over time ( Table 2). During developmental and perinatal stages, Sca1 POS CICs represent between the 0%-2% of CD31 NEG /CD45 NEG CICs but importantly increased between P1 and P7. The majority of Sca1 POS CICs derived from the epicardial lineage at P7 (7.65 ± 0.35% of CD31 NEG /CD45 NEG were Sca1 POS / YFP POS and 3.37 ± 0.86% were Sca1 POS /YFP NEG ), although this contribution was reduced in the adult ( Figure 3B; Table 2). In order to characterize the Sca1 POS CICs subpopulation in detail, we first checked for the expression of CD34. The percentage of CD31 NEG / CD45 NEG /Sca1 POS /YFP POS CICs expressing CD34 remained relatively stable from E18.5 onward. In contrast, the percentage of CD31 NEG /CD45 NEG /Sca1 POS /YFP NEG /CD34 POS CICs was highly dynamic along the different stages ( Figure 3C). Regarding the coexpression of c-Kit and Sca1, CD31 NEG /CD45 NEG /Sca1 POS / c-Kit POS /YFP POS CICs were detected at P7 only ( Figure 3D). CD31 NEG /CD45 NEG /Sca1 POS /c-Kit POS /YFP NEG CICs were abundant at E15.5, but their number sharply decreased later on.
Further evaluation of the samples revealed that the percentage of CD45 POS /Sca1 POS CICs increased from E15.5 to P7 and then decreased in the adult (Supplementary Figure S3A; Table 2). Remarkably, more than 70% of Sca1 POS CICs coexpressed CD31 after birth (up to 90% in the adult); most of these cells were YFP NEG (Supplementary Figure S3B; Table 2).
To validate the FACS analysis and map the cardiac location of Sca1 POS cells, we performed immunohistochemical analysis. Quite differently from our c-Kit immunohistochemical screening, Sca1 POS CICs were found to be distributed throughout all the cardiac ventricular interstitium. Furthermore, Sca1 POS /YFP POS CICs were identified in different stages (arrows in Figure 3E). In the adult, the majority of Sca1 colocalizes with Griffonia simplicifolia lectin (IB4), strongly suggesting an endothelial nature for these cells (arrowheads in Supplementary Figure S3C). Some of these Sca1 POS endothelial cells were also found to be YFP POS (arrows in Supplementary Figure S3C).
In summary, our data demonstrated that Sca1 POS cells represent a low percentage of CICs with a heterogeneous cellular origin, including Wt1 lineage. Interestingly, these cells are mainly found in postnatal and adult stages, and are closely related to the cardiac endothelial lineage, suggesting a potential difference in this cell type in terms of cell proliferation, migration, and differentiation.

Majority of Cardiac Wt1 Lineage-Derived Endothelial Cells and Fibroblasts Express Sca1 in the Adult Heart
Since a significant percentage of Sca1 POS CICs co-expressed YFP in the adult, we aimed at characterizing these cells in detail at this stage. We, therefore, searched for Sca1 POS cells within the epicardial lineage-derived endothelial and cardiac fibroblast populations (Acharya et al., 2012;Katz et al., 2012;Ruiz-Villalba et al., 2015;Cano et al., 2016). Considering the total YFP POS cardiac cells ( Figure 4A), our analysis revealed that more than 90% of Wt1 lineage-derived CD31 POS ventricular endothelial cells co-express Sca1 ( Figure 4B; Table 3). Then, to determine the percentage of Sca1 POS cells in the cardiac fibroblasts (CFs) population we used the mEFSK4 antibody, a known CF panmarker (Pinto et al., 2016;Ruiz-Villalba et al., 2020). Our analysis showed that around 60% of YFP POS /CD31 NEG /mEFSK4 POS CFs also co-expressed Sca1 ( Figure 4C; Table 3). These data clearly indicate that both the putative epicardial-derived endothelial and cardiac fibroblasts populations express Sca1.
These results are in accordance with the recently described contribution of Sca1 POS cells into the cardiac interstitium (Tang  , 2018, 2020). However, our results point out the heterogeneous nature of this subpopulation of adult Sca1 POS CICs and a close relationship with specific subpopulations of epicardial-derived lineages, such as cardiac endothelium and fibroblasts. This is again relevant in terms of understanding the cardiac interstitium in both homeostasis and response to pathologic stimuli.

Postnatal Epicardial Cell Tracing Discards De Novo Contribution of This Cell Lineage Into the Adult Sca1 POS Interstitial Cell Subpopulation
In a final series of experiments, we aimed at discarding a potential de novo postnatal appearance of Sca1 POS /YFP POS CICs (between P1 and P3). To tackle this task, we crossed the conditional Wt1Cre ERT2 transgenic line with the Rosa26R-eYFP reporter one ( Figure 5A). Our analysis revealed that YFP POS cells were absent from the CD31 NEG /CD45 NEG cardiac cells fraction at P7 (Figures 5B,C). These results supported an embryonic epicardial origin for the adult CD31 NEG /CD45 NEG /Sca1 POS /YFP POS CICs identified and characterized in this work.

DISCUSSION
Cardiac interstitial cells (CICs) are a dynamic and heterogeneous cell population. This poorly studied group of cells has been historically overshadowed by the size and functional relevance of cardiomyocytes. CICs, however, largely exceed cardiac muscle cells in number and their enormous importance in the diseased heart has recently attracted much attention. Indeed, CICs contain progenitors and quiescent forms of cardiac fibroblasts (CFs), which are central to the reparative fibrotic response that characterizes the pathophysiology of multiple cardiac diseases, from ischemic, dilated, and arrhythmogenic cardiomyopathies to the aging and diabetic heart. Increasing our understanding of this unique group of cells will open new avenues for the early diagnosis of and effective therapy for these ailments.
The ontogenetic origin of CICs has been shown to be very diverse, but two of the main sources for these cells have been identified, namely, the epicardial-derived mesenchymal cell progeny often referred to as epicardial-derived cells or EPDCs and the blood cell lineages (for a comprehensive review see Pogontke et al., 2019). In both cases, the cellular colonization of the interstitial space takes place progressively during the perinatal stages (Epelman et al., 2014;Moore-Morris et al., 2015;Ruiz-Villalba et al., 2015). The main aim of this study was to study two specific CICs fractions (c-Kit POS and Sca1 POS ) in relation to epicardial-derived and hematopoietic/blood cell lineages. Since the beginning of this century, both c-Kit POS and Sca1 POS have been considered to represent multipotent resident cardiac stem cell (CSC) populations (Beltrami et al., 2003;Vicinanza et al., 2017). This concept was first intensely debated and then recently refuted, so that the existence of this CSCs as they were originally described is no longer accepted (Van Berlo et al., 2014;Sultana et al., 2015). The existence of CICs expressing c-Kit and Sca1 remains, however, an undisputed fact, but their real nature is far from having been deciphered. Our FACS analysis first revealed that very few c-Kit POS cells are related to cardiac Wt1-lineage-derived cells, which can be genetically tagged by their original embryonic expression of the Wilms' tumor suppressor gene 1 (Wt1). This analysis was specifically carried out in the CD31 NEG /CD45 NEG CICs fraction to exclude endothelial and blood-borne cells from the screening. We nonetheless identified a small proportion of c-Kit POS /YFP POS CICs whose identity is not still clear. The presence of c-Kit POS , EPDCs homologous mesothelial-derived Cajal cells in the intestinal tract (Carmona et al., 2013) strongly suggested that cardiac c-Kit POS /YFP POS cells are not an artifact of the genetic cell tracing technology but a real biological entity. In any case, immunohistochemistry unambiguously shows that c-Kit POS /YFP POS is normally located in the subepicardial space or in contact with the wall of coronary blood vessels, two domains massively contributed by epicardial cells (Cano et al., 2016). The sharp postnatal decrease in c-Kit POS /YFP POS cell numbers within the CD31 NEG /CD45 NEG CICs fraction observed by FACS may reflect a progressive dilution of a poorly proliferating cell subset with respect to other expanding cardiac cell types and also reflects the difficulties of finding significant numbers of these cells in adult cardiac tissues using immunohistochemical approaches. Alternatively, it could be argued that the decrease in the number of c-Kit POS /YFP POS cells is due to the loss of this cell surface marker during cell differentiation/maturation. More research is needed to clarify this point.
As the majority of cardiac c-Kit POS cells are also YFP NEG in Wt1Cre-YFP mice, we then decided to evaluate a possible origin for these cells from hematopoietic/blood lineages. In order to do so, we searched for the specific pan-leukocyte marker CD45 in CICs. The results from this inspection show that CD45 POS /c-Kit POS cells are abundant among CICs (up to 65% at E15.5). As expected, virtually none of these cells was YFP POS , excluding a possible relation between blood-borne cells and Wt1 cell lineages. Of note, nearly all c-Kit POS CICs co-expressed CD34, an immature endothelium and stem cell marker (Carmona et al., 2020). Furthermore, c-Kit POS cells expressing the endothelial and hematopoietic cell marker CD31 POS were also identified in our immunohistochemical analysis. Other authors have already identified c-Kit POS / CD45 NEG CICs as endothelial progenitors in the adult (Sandstedt et al., 2010). However, as described above, we found a relevant contribution of c-Kit POS cardiac cells in the CD45 POS CIC fraction, suggesting that these cells are therefore more likely to be blood-borne cells than endothelial cells. In accordance with this conclusion, CD45 NEG /c-Kit POS /YFP NEG CICs were found to encompass cells expressing markers for macrophages (F4/80), lymphoid (CXCR3), and myeloid (CD11b) cell types, a result that endorses an extracardiac origin for these CIC. Finally, a marked decrease in the number of CD45 POS /c-Kit POS CICs was observed between embryonic and postnatal stages, a finding that can be interpreted either as a cell dilution similar to that previously described for c-Kit POS /YFP POS cells or as an actual decrease of blood circulating cell recruitment to the adult heart.
The second part of our study focused on the study of Sca1 POS CICs. FACS characterization of Wt1Cre-YFP CICs clearly indicates that the majority of CD31 NEG /CD45 NEG /Sca1 POS CICs were YFP POS at P7, a finding that relates these cells with the Wt1-derived cardiac mesenchymal lineage. The progressive, sustained increase of Sca1 POS CICs from developmental to postnatal stages (rising from 0% to 2% to almost 70%) is evidence of the fast expansion of these cells after birth. Further FACS characterization indicates that an important number of these cells are CD31 POS endothelial ones (Tang et al., 2018(Tang et al., , 2020. Indeed, we show that 30% of CD31 POS ventricular endothelial cells also expressed Sca1 at E18.5, reaching 80% in the adult, indicating that the contribution of Sca1 POS /CD31 POS cells to the cardiac interstitium is highly dynamic over time. Relevant to this discussion, 90% of YFP POS /CD31 POS cells and 60% of YFP POS /CD31 NEG / mEFSK4 POS CFs expressed Sca1, suggesting that the expression of this molecule is related to the epicardial cell lineages. This finding could be relevant due to the importance of epicardial cells in the context of cardiac repair (Ruiz-Villalba et al., 2015;Chen et al., 2018). However, since Sca1 POS /YFP POS and Sca1 POS /YFP NEG cells included similar percentages of mesenchymal stem-like cells, such as CD73, CD90, and CD105, more research is needed to determine potential differences between both cell subpopulations.
We have shown that postnatal CD31 NEG /CD45 NEG /Sca1 POS / YFP POS CICs are likely to arise through the proliferation of prenatal cell populations, although the possibility of this increase results from the de novo expression of Wt1 in the Sca1 POS population cannot be directly discarded. To ensure this was not the case, we used an inducible epicardial driver line (Wt1Cre ERT2 ) to activate the promoter in the postnatal stages. At P7, we did not identify YFP POS CICs, confirming that CD31 NEG /CD45 NEG /Sca1 POS /YFP POS CICs derive from embryonic founder cells.
Our study provided novel evidence on the relationship existing between cardiac interstitial c-Kit POS and Sca1 POS cell populations and the Wt1 cell lineage, also contributing relevant information on the spatiotemporal distribution of these cells in the embryonic and postnatal heart. Since the majority of cardiac Wt1 lineage cells (E11-E17) are known to derive from the embryonic epicardium (Wessels et al., 2012;Cano et al., 2016), we generally assumed that YFP cells are epicardial derivatives. In accordance with this view, recent publications based on singlecell RNA-Seq revealed the existence of an epicardial transcriptomic signature related to Wt1 lineage-derived cells in the adult heart (Quijada et al., 2020;Hesse et al., 2021). For this reason, we considered that adult YFP cardiac cells are mostly derived from the embryonic epicardium. Notwithstanding, this conclusion does not rule out the possibility of Wt1 lineage cells deriving from other alternative cellular sources (King-Underwood et al., 2005).
Frontiers in Cell and Developmental Biology | www.frontiersin.org May 2022 | Volume 10 | Article 864765 As it can be inferred from the previous discussion, Sca1 POS / YFP POS CICs are very likely to be epicardial derivatives. The enrichment of Sca1 in a major part of adult putative epicardialderived cells suggested that this molecule may be involved in the regulation of epicardial biology. The colonization of the cardiac interstitium by these cells starts during embryonic development and is highly dynamic. In this respect, since the turnover rate for CICs is unlikely to be steady throughout adult life, the specific cellular dynamics and composition of the cardiac interstitium have been proposed to be a sensor for the health status of the heart (Pogontke et al., 2019).
In summary, increasing our knowledge of the origin, diversity, and functions of CICs will be instrumental to the development of a diagnostic and prognostic test for the evaluation of cardiac homeostasis and the cardiac interstitium´s response to pathologic stimuli. However, further, systematic research on CICs biology is needed to bring about significant progress in our understanding of interstitial responses to pathologic conditions.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusion of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The animal study was reviewed and approved by the Committee on Ethics of Animal Experiments of the University of Málaga (procedure code 2018-0018).

AUTHOR CONTRIBUTIONS
CP and AS-T: performing the experiments and collecting the data. CP, JG, RM-C, and AR-V: data analysis and interpretation. CP, JG, AR-V, and JP-P: research design and manuscript writing. CP, JG, AS-T, RM-C, AR-V, and JP-P: final manuscript approval. All authors contributed to the article and approved the submitted version.

ACKNOWLEDGMENTS
We thank Maria Isabel Castillo and John Pearson (BIONAND), and David Navas (UMA) for technical help with animal care, confocal microscopy, and flow cytometry.