Adaptation within embryonic and neonatal heart environment reveals alternative fates for adult c-kit+ cardiac interstitial cells.

Abstract Cardiac interstitial cells (CICs) perform essential roles in myocardial biology through preservation of homeostasis as well as response to injury or stress. Studies of murine CIC biology reveal remarkable plasticity in terms of transcriptional reprogramming and ploidy state with important implications for function. Despite over a decade of characterization and in vivo utilization of adult c‐Kit+ CIC (cCIC), adaptability and functional responses upon delivery to adult mammalian hearts remain poorly understood. Limitations of characterizing cCIC biology following in vitro expansion and adoptive transfer into the adult heart were circumvented by delivery of the donated cells into early cardiogenic environments of embryonic, fetal, and early postnatal developing hearts. These three developmental stages were permissive for retention and persistence, enabling phenotypic evaluation of in vitro expanded cCICs after delivery as well as tissue response following introduction to the host environment. Embryonic blastocyst environment prompted cCIC integration into trophectoderm as well as persistence in amniochorionic membrane. Delivery to fetal myocardium yielded cCIC perivascular localization with fibroblast‐like phenotype, similar to cCICs introduced to postnatal P3 heart with persistent cell cycle activity for up to 4 weeks. Fibroblast‐like phenotype of exogenously transferred cCICs in fetal and postnatal cardiogenic environments is consistent with inability to contribute directly toward cardiogenesis and lack of functional integration with host myocardium. In contrast, cCICs incorporation into extra‐embryonic membranes is consistent with fate of polyploid cells in blastocysts. These findings provide insight into cCIC biology, their inherent predisposition toward fibroblast fates in cardiogenic environments, and remarkable participation in extra‐embryonic tissue formation.


| INTRODUCTION
Myocardial homeostasis is maintained by dynamic interaction on multiple levels between cardiomyocytes and the cardiac interstitial cell (CIC) population. Decades of study reveals CICs as a heterogeneous collection of cell types that defy simple categorization, due in part to their fluid adaptability in response to development, aging, acute injury, and chronic stress. [1][2][3] Parsing out CIC subtypes with specific markers such as periostin or Tcf21 has merged with the more impartial and nuanced approach of transcriptomic profiling at the single cell level. [4][5][6] Appreciation for the complexity of CIC biological properties continues to grow, as does recognition that environmental influences exert profound control over CIC phenotypic characteristics and functional activities.
Studies of CIC biology often rely upon assessments performed using populations expanded by in vitro cell culture for various reasons of sample yield, manipulability, and of course simplification compared with challenges of the myocardial milieu in vivo. [7][8][9][10][11] Such studies provide tremendous insights but also are limited by inescapable aspects of cell culture adaptation, natural selection ex vivo for robust proliferative cell subsets, and multiple choices for conditions of experimental design. Collectively, these variables contribute to the wide range of interpretations and published literature for CIC biology that has been extensively reviewed. 4,[12][13][14] Moreover, a plethora of selected subpopulations of in vitro expanded CICs have been intensively studied for cardioprotective and reparative potential upon reintroduction into pathologically injured myocardium for over a decade, 10,15,16 but consequences of cell culture environment upon CIC properties in terms of reshaping population characteristics or individual cellular functional capabilities remain relatively unstudied and poorly understood. Typically, such cultures involve two-dimensional (2D) monolayer growth and serial passaging to obtain sufficient numbers of cells for treatments. [17][18][19][20] Such 2D culture conditions promote reprogramming toward a common shared transcriptional profile, even between CIC subpopulations enriched by selection for unrelated markers as well as comparisons between multiple donor sources. 5,21,22 Taken further, our group found that relatively short-term 2D cell culture for five serial passages results in loss of cell-specific identity markers and increased homogeneity in a CIC subpopulation enriched for tyrosineprotein kinase kit or CD117 (c-Kit + ) cardiac interstitial cell (cCIC) compared with correspondingly selected freshly isolated cells by singlecell RNA-Seq transcriptional profiling. 22 Findings such as these support the contention that CIC isolation and propagation conditions exert profound influences upon biological and functional properties, consistent with our recent reports of hypoxic culture conditions antagonizing mitochondrial dysfunction and senescence in human cCICs 19 as well as tetraploid conversion of murine cCICs. 23  A major impediment to assessing readaptation of cultured CICs following delivery to host adult myocardium is poor retention and persistence of the donated cell population. [24][25][26][27] Although using augmented approaches to embed CICs offers some improvement over direct injection to recipient myocardium, bioengineering solutions involving injectable gels or cultured patches severely limits direct interaction between exogenously introduced CICs and host myocardium. Furthermore, delivery to pathologically injured myocardium further stresses the CIC population already coping with dramatic changes in environmental conditions. For example, host immunemediated reaction to pathologic injury including CIC delivery prompts a powerful inflammatory response involving cytotoxic action. Indeed, developing myocardium exhibits stage-specific permissivity for incorporation of introduced or migrating cells. 1,28 Therefore, we reasoned that assessment of cultured cCIC adaptation following reintroduction to myocardial tissue in vivo would be facilitated by delivery to early developmental stages characterized by cardiogenic activity and negligible inflammation.
Permissive conditions present in embryonic tissue or an early stage developing heart allows for engraftment and persistence of injected cCICs, then followed in subsequent days to weeks for determination of phenotypic characteristics exhibited by both exogenously introduced cells as well as host reaction to their presence.   Robust expression of c-Myc, Gata3, Gata6, and Gata4 mRNAs relative to embryonic stem cells (ESCs) is evident by quantitative PCR (Figure S1b), and cCICs showed the lowest pro-oncogene expression profile relative to ESC or the whole heart ( Figure S1c). Spontaneous aggregation into 3D embryoid body spheres (EBs) in suspension culture is commonly used to study ESC differentiation potential, 11,29 and culture expanded F I G U R E 1 C-Kit + cardiac interstitial cells (cCICs) integrate into preimplantation blastocysts and adopted extra-embryonic fate. A, Schematic of blastocyst injection and ex vivo incubation for 24-48 hours. (b-d) At 24 hours postinjection (hpi), injected cCICs were retained in blastocoel (B, n = 6/11), inner cell mass (ICM; C, n = 2/11), and trophoblast (D, n = 8/11). See also Video S1. E, At 48 hpi, whole-mount immunostaining of injected blastocyst showing cCICs anchored with host cells and spread out as spindle morphology in a hatching blastocyst blastocoel. See also Video S2. F, Left, whole-mount immunostaining of injected blastocyst showing cCICs sharing tight junction (ZO1, white) with host trophectoderm (TE) layer (CDX2, green). Right, higher magnification of boxed area. Arrowheads: ZO1 junctions. G, Immunostaining of ICM marker Oct3/4 (white) showing cCICs do not integrate into ICM. H, A longitudinal optical section showing nuclei (arrowheads) of cCICs located at TE layer. I, Higher magnification of transverse optical section showing cCICs (arrowhead) integrated among nuclei (DAPI, blue) of trophoblasts (CDX2, white), sharing tight junctions (ZO1, green). J, After uterine transfer into pseudopregnant female, cCICs were detected in a mosaic pattern in extraembryonic membrane from a chimeric embryo from blastocyst injection at 10 dpi/E13.5. K, Fluorescent scanning of a frozen sectioned extraembryonic membrane showing mosaic cCICs integration. Nuclei, DAPI, blue. L, Immunostaining of Laminin showing integrated cCICs localized to the opposite side of epithelial layer of extra-embryonic tissue. Laminin, green. M, Immunostaining showing cCICs locate in proximity of trophoblast (CDX2, white) and express fibroblast marker (vim, green) in extraembryonic tissue (n = 5). Scale bar, 50 μm cCICs similarly aggregate into spheres ( Figure S1d). Mesoderm induction treatment of cCIC-spheres in adherent culture showed increased expression of SM22 alpha (SM22α), whereas endoderm (α-Fetoprotein, AFP) and ectoderm (βIII-Tubulin, TUJ1) markers remained undetectable before and after differentiation ( Figure S1e). cCICs uniquely express SM22α but not AFP shown by confocal microscopy immunolabeling ( Figure S1f), confirming that in vitro expanded cCICs are capable of expressing SM22α + . In addition to mesoderm potential, a majority of mesodermal induced cCICs express the fibroblast marker vimentin (Vim), consistent with fibroblast origin ( Figure S1g). Collectively, these findings portray cCIC in culture as mesodermal-lineage derived cells with characteristic fibroblast-associated marker expression.

| Extra-embryonic tissue integration of cCIC in preimplantation blastocysts
Chimeras blastocyst formation following cell injection is used as a stringent assessment for testing stem cell pluripotency. 30,31 Adult multipotent cells may harbor properties similar to ESCs allowing for chimera formation when injected into blastocysts. [32][33][34] Therefore, cCICs were delivered into murine blastocysts that were subsequently cultured ex vivo for 24 to 48 hours postinjection (hpi; Figure 1A). The presence of injected Following the anticipated extra-embryonic pattern, cCICs mosaically integrate predominantly in chorionic lamina of amniochorionic membrane (AM) opposite from squamous amniotic epithelium (Laminin + ) at 10 days postinjection (dpi; E13.5, Figure 1J-L). Engrafted cCICs locate adjacent to CDX2 + cells and express fibroblast marker vim in extraembryonic tissue ( Figure 1M). In contrast, the absence of cCICs from the ICM of developing embryonic tissue was exhaustively evaluated without a single positive finding (n = 253), whereas embryo chimerism was readily observed with a frequency of 19.2% using ESC as a control cell (n = 10/52; Table 1, Figure S2). Therefore, although cCICs possess sufficient functional capacity for extra-embryonic tissue integration, they are unable to participate in embryonic chimerism.

| Neonatal myocardium allows for extended persistence of cCICs
Empirical testing for intramyocardial injection of approximately 5000 cCICs in a time course ranging from P0 to P5 (data not shown) revealed the optimal postnatal stage for engraftment and persistence was P3 ( Figure 3A). Assessment of cCIC fate performed every 7 days T A B L E 1 Generation of chimeric mice Abbreviations: cCIC, c-Kit + cardiac interstitial cell; ESC, embryonic stem cell.   Cultured murine cCICs acquire tetraploid DNA content with serial passaging and override cellular senescence. 23 Indeed, the tetraploid nature of cCICs ( Figure 6) likely accounts for the mechanism behind engraftment into TE and amniotic membrane integration (Figure 1 and Persistence of injected cCIC in neonatal hearts for up to 4 weeks (28 dpi) is remarkable given long-standing issues of retention and engraftment in the adult heart. Donated cells are typically lost shortly after delivery with engraftment rates below 5% to 10% by 24 hpi and less than 2% by 48 hpi. [24][25][26][27] In comparison, initial cCIC engraftment of 36.2% ± 17.0% at 2 hpi remained high at 33.4% ± 6.2% by 48 hpi in neonatal injections ( Figure S4). Moreover, histological analyses at the were isolated from two male mice, and cCICs used in FUCCI experiments were isolated from four mice (2 males + 2 females).

| Histology and immunofluorescence staining
Mice were heparinized (Sigma-Aldrich H3393, 10 units/g) by intraperitoneal injection and euthanized at harvest time points. For animals younger than 14 days, euthanasia was carried out by anesthetization on ice followed by decapitation. For animals at 14 days and older, euthanasia was carried out by isoflurane overdose followed by cervical dislocation. Hearts were perfused with PBS and 1% PFA before removal from thoracic cavity, followed by fixation in 1% PFA immersion overnight at 4 C. Fixed hearts were dehydrated in 30% sucrose in PBS overnight at 4 C, then in OCT + 30% Sucrose mix at 1:1 ratio, before mounting in NEG50 and frozen on dry ice. Frozen sections were cut at 20 μm thickness and collected onto Superfrost glass slides. Sections were allowed to dry for 48 hours prior to storage at −20 C.

| Immunoblotting
At the time of harvesting, cells were washed twice in cold PBS and lysed in RIPA buffer (Thermo, 89901) with freshly added proteinase inhibitor and phosphatase inhibitors cocktails (Sigma P0044, P8340, P5726) for 30 minutes on ice with intermittent vortexing. Cell lysates were then centrifuged for 10 minutes at 11000g at 4 C to remove insoluble debris.
Supernatants were quantified with Bradford assay (ThermoFisher, 23236) and 20 μg lysates were run on 4% to 12% Bis-Tris protein gels (Invitrogen, NP0335BOX) and transferred onto a PVDF membrane (Millipore, IPFL00010), followed by blocking in 10% nonfat dry milk (LabScientific) for 1 hour at RT. Primary antibodies (see dilutions in Table S1) were incubated overnight at 4 C and secondary antibodies (1:1000) for 90 minutes at RT. Immunoblots were scanned with LI-COR Odyssey Clx system.   Table S1) were incubated overnight at 4 C, and secondary antibodies (1:100) were incubated for 1.5 hours at RT. DAPI was added to last PBST washes to stain nuclei. All washes and incubations were performed in liquid bubbles under mineral oil immersion. Following staining, blastocysts were gradually transferred from PBST to 20%, 50%, and 70% glycerol, and mounted in 80% glycerol. Z-stack series scanning was performed using Leica SP8 confocal microscopy (×63) at a 5-μm interval depth. Three-dimensional reconstruction videos were generated using Leica LAS X analysis software.

| In utero transplantation
Timed pregnant FVB/J female inbred mice were anesthetized with ketamine/xylazine according to body weight at 10 μL/g. Uterine horns were exteriorized through a short ventral midline incision at lower abdomen. Cells were delivered using a microcapillary needle with the appropriate volume of cell suspension at approximately 5000 cells per embryo into pericardial space. After injection, the uterine horns were gently placed back into the abdomen, and the maternal abdominal muscle and peritoneum were closed by surgical adhesive. Following recovery, two buprenorphine doses (0.2 μg/body weight g) were given every 12 hours as analgesia. At 2 dpi, dams were euthanized by isoflurane overdose followed by cervical dislocation. Embryos were dissected out of uteri in cold PBS and fixed in 1% PFA immersion at 4 C overnight.

| FUCCI constructs and expression
The FUCCI system consists of two chimeric proteins, mKO-Cdt1 and AzG-Geminin, which oscillate reciprocally during cell cycle, labeling the nuclei in G1 phase orange and those in S/G2/M phases green. 35 During G1/S transition, both probes are present, resulting in a yellow fluorescence (overlaid green and red); during the brief gap between M and G1 phases, neither probe is present and fluorescence is absent. Oscillation between red, yellow, and green signals tracks cell cycle status 35,36 ( Figure 4A). FUCCI lentiviral plasmids were generated as previously described. 36

| Myocardial infarction and intramyocardial injection
Myocardial infarction and intramyocardial injection were carried out as previously described 67 on FVB/J strain mice. Briefly, hearts were popped out through the fourth intercostal space and the left anterior descending artery (LAD) was permanently ligated at the second distal branching point using a 7-0 silk suture. Following LAD ligation, three injections were delivered (Harvard Apparatus, Hamilton infusion pump) at the border zone surrounding the blanching area at a tangential angle parallel to the myocardial wall, in order to ensure intramyocardial cell delivery. A total of 100 000 cells/10 μL were injected per heart at the three injection sites. Following injection, the heart was immediately placed back into the intrathoracic space and muscle and skin were closed by surgical adhesive.

| Cardiac cell disassembly and quantification
Postinjection hearts were enzymatically disassembled into single-cell suspension and subjected to flow cytometry for fluorescence-based cell count. For neonates, postop pups at 2 and 48 hpi were heparinized and anesthetized on ice. Anesthesia was maintained by hypothermia in a petri dish filled with ice during the surgical procedure.
Perfusion and digestion were performed following a modified protocol as previously described. 68 Briefly, the heart was digested (Collagenase

| Flow cytometry
Single-cell resuspension was analyzed using a BD FACSCanto instrument. Cells digested from sham (uninjected) hearts were used to exclude autofluorescence disturbance, and cultured cCICs expressing mCherry fluorescence were used as positive gating to establish fluorescence levels. All cells from neonatal hearts were analyzed. A recorded volume of 100 to 200 μL cell suspension from adult interstitial cells was analyzed, and the whole heart cell count was calculated based on volumetric ratio relative to 1 mL initial cell suspension. Flow cytometry data were analyzed by FlowJo software (BD Biosciences).

| Echocardiography
Echocardiography was performed using the Vevo2100 (Visual Sonics) system from LV parasternal long and short axes at a heart rate range of 500-550 beats/minute. EF and FS were determined by offline analysis. Age-matching unoperated mice were used as baseline controls.

| Masson's trichrome staining
Masson's trichrome staining was performed using trichrome stain kits Samples were then washed in PBS, mounted in VectaShield, and scanned using a Leica SP8 confocal microscope.

| Ploidy quantification
Following euthanization, mouse sperm was collected from vas deferens and maintained in PBS/0.5% BSA on ice. BMC were collected from femur flushed with PBS/0.5% BSA using a 27-gauge needle and filtered through a 30-μm cell strainer to remove debris.
Cultured cCICs were trypsinized and pelleted at 300g for 5 minutes.
Cells were then stained with Sytox Green (Invitrogen, S7020, 1 μM) for 15 minutes at RT before subjected to flow cytometry analysis.
Unstained cells of each cell type served as negative gating controls.
Ploidy comparison was established using sperm as haploid and BMC as diploid control using FlowJo software.
Alternatively, sperm, BMC, and cCIC suspensions were manually mixed and cytospun (Thermo, Cytospin 4) for 3 minutes at 800 rpm with low acceleration onto a poly-D-lysine-coated slide. Cells were then fixed in 1% PFA for 20 minutes at RT, stained with DAPI for 5 minutes at RT, following by three PBS washes to remove excess staining. cCIC nuclei were identified by mCherry fluorescence, BMC nuclei were identified by mCherry negativity, and sperm nuclei were identified by unique fishhook-like nuclear morphology. Nuclear DAPI signals were scanned by z-series spanning entire nucleus at 1 μm interval using Leica SP8 confocal microscopy. Z-projection was reconstructed with sum intensity by ImageJ. Nuclear intensity was quantified by nuclear volume tracing using ImageJ and presented as arbitrary units.

| Statistical analysis
All data were presented as mean ± SEM and analyzed by GraphPad Prism 5.0b with unpaired Student's t-test, two-tailed. A P-value <.05 was considered statistically significant.

SUPPORTING INFORMATION
Additional supporting information may be found online in the Supporting Information section at the end of this article.