Identification of cardiac stem cells within mature cardiac myocytes

Abstract

Cardiac stem cells are described in a number of mammalian species including humans. Cardiac stem cell clusters consisting of both lineage-negative and partially committed cells are generally identified between contracting cardiac myocytes. In the present study, c-kit⁺, Sca⁺, and Isl1⁺ stem cells were revealed to be located inside the sarcoplasm of cardiac myocytes in myocardial cell cultures derived from newborn, 20-, and 40-day-old rats. Intracellularly localized cardiac stem cells had a coating or capsule with a few pores that opened into the host cell sarcoplasm. The similar structures were also identified in the suspension of freshly isolated myocardial cells (ex vivo) of 20- and 40-day-old rats. The results from this study provide direct evidence for the replicative division of encapsulated stem cells, followed by their partial cardiomyogenic differentiation. The latter is substantiated by the release of multiple transient amplifying cells following the capsule rupture. In conclusion, functional cardiac stem cells can reside not only exterior to but also within cardiomyocytes.

Abbreviations

CSCs

cardiac stem cells

DIV

day in vitro

CMs

cardiac myocytes

CICS

cell-in-cell structure

TACs

transitory amplifying cells.

Introduction

Stem cells represent a promising material for therapies, disease modeling, and drug testing; these include cardiac stem cells (CSCs) of various mammals, including humans. To date, the following major subtypes of CSCs have been described: c-kit⁺,¹ Isl1⁺,² Sca1⁺,³ and cardiosphere-generating progenitors.⁴ Among these, the c-kit⁺ subtype of CSCs has been characterized in detail, demonstrating the abilities of self-renewal, clonogenicity, and multipotency.⁵ Furthermore, the apparent functional benefits of CSC intramyocardial transplantation after myocardial infarction have been shown and verified in a number of animal studies.⁶ However, many significant aspects of CSC biology still require further investigation, such as the distribution of intracardiac CSCs and the role of the microenvironment in their transformation from a quiescent to a proliferating phenotype. To date, clusters of lineage-negative and partially committed CSCs have only been identified within the microdomains between contracting cardiac myocytes (CMs); these sites are referred to as “cardiac niches.”^7,8 In the present study, on the other hand, CSCs were found to also reside within the CM sarcoplasm in cardiac cell cultures derived from newborn, 20-, and 40-day-old rats and in the suspension of freshly isolated myocardial cells (ex vivo) of 20- and 40-day-old rats. Herein we present the first description of this phenomenon, which has not been previously reported.

Results

Cardiac cell suspensions were obtained according to standard technique,⁹ as described in detail in the Material and methods. Briefly, after plating and incubation, cultured cells were fixed, permeabilized, and stained using a set of antibodies against CSC antigens (Fig. 1A). Careful examination showed that some of the CSCs were localized inside the host cell sarcoplasm (Fig. 1B–D) thereby forming membrane-bound cell-in-cell structures (CICSs). These were frequently seen either in the vicinity of or between the nuclei of the mature CMs (Fig. 1B–D). Serial optical tomography provided additional evidence in favor of the CSCs intrasarcoplasmic residence (Fig. 2A–C and also Material and Methods). Nested in their intracellular niche, we observed that the CSC-containing CICSs were surrounded with densely packed filaments of cytoskeletal actin (Fig. 2B, C). The CICSs enlarged in parallel with the duration of cell culture, most likely due to proliferation of the CSCs within the CICSs as documented by Ki-67 labeling (Fig. 1C). This marker was also positive in host myocytes obtained from newborn rats (Fig. 1C) but not in those from 20- and 40-day-old animals, which is consistent with the notion that cardiomyocytes exit the cell cycle 5 to 6 days after birth. Furthermore, as long as the CICS dimensions increased and irrespective of the age of the animal, the CSC coating thickened, which we termed as a “capsule.” The capsule maintained several openings (“pores”) between the capsule content and the host cell sarcoplasm (Fig. 1D–G). The CICSs with similar characteristics were identified in freshly isolated suspension of myocardial cells derived from 20- and 40-day-old rats (Fig. 3A, C, D). Mature cardiomyocyte with small Isl1⁺ CSC-containing capsule is shown in Figure 3B and Video S1. The dimensions of the mature CICSs found both in vitro and ex vivo were: Dmax = 28.8 ± 1.2 μm, Dmin = 25.3 ± 1.0 μm, vertical dimension = 17–36 μm. Some of the capsules were so large that they could displace both the neighboring nucleus and the myofilaments of the host cell to its periphery (Fig. 1E, Video S2). Serial, time-lapse and confocal microscopy showed the eventual rupture of the capsule following continued CICS expansion and, of note, the release into the medium of an abundance of transient amplifying cells (TACs) expressing not only CSC markers but also cardiac proteins (Fig. 4A-E, G, Video S3), with some of the released cells spreading over the substrate (Fig. 4F, H, I). We suggest that the latter constitutes evidence for the division of the encapsulated CSCs followed by their partial cardiomyogenic differentiation prior to their release. The proposed concept of intracellular CSC development is summarized in Fig. 5

Figure 1.

The CSCs inside CMs and the formation of CSC-containing CICSs in the cultures obtained from newborn and 20- and 40-day-old rats. (A) Experimental design. The cells were plated and cultured for up to 30 days, followed by immunostaining or time-lapse microscopy. (B–G) Immunocytochemistry. The nuclei of the cells have been stained with Hoechst. Transmitted light and fluorescent images are merged. (B) c-kit⁺ CSC inside a CM obtained from a newborn rat (day in vitro 6). (C) Isl1⁺ CSC inside a CM obtained from a newborn rat (day in vitro 4). As documented by the expression of Ki67, both the CSC and the host cell exhibit proliferative ability. (D) Sca⁺ CSC encapsulated between the nuclei of the host cell (20-day-old rat, day in vitro 11). (E) A mature c-kit⁺ CSC-containing CICS with a prominent coating (“capsule”) with 3 pores (white arrows, 40-day-old rat, day in vitro 6). Optical sectioning shows the host cell nucleus (blue) just above the CICS (see sections 13–15 in Video S1). (F-G) The CICS capsule in detail. (F) Erosion of the Isl1⁺ CSC-containing CICS capsule (black arrow) obtained from a 40-day-old rat, day in vitro 8. The pores are also visualized (white arrows). The capsule interior is positive for sarcomeric α-actinin, also observed in (G). (G) Erosion of the c-kit⁺ CSC-containing CICS capsule (black arrow) obtained from a newborn rat, day in vitro 20. The pores are seen (white arrows).

Figure 2.

Optical tomography of the CSC-containing CICSs. The optical sections were spaced 1.01 μm (A, B) and 2.01 μm along the z-axis (C). Images are placed in consecutive order from the bottom to the top of structures. (A) C-kit⁺ CSC-containing CICS in the culture obtained from a 40 day old rat (FITC, green, day in vitro 6) After counterstaining for α-sarcomeric actin (Alexa 543 nm, red), the nuclei (Hoechst, blue) of both the CSC and the host cardiomyocytes are visualized; the vertical dimension of the CICS is 20 μm (slices 3 to 20). Transmitted light and fluorescent images are merged. (B) The Isl1⁺ CSC-containing CICS (FITC, green, left column) inside a given host CM obtained from a newborn rat (day in vitro 11), with a vertical dimension of 7 μm (slices 2 to 6). Cytoskeletal actin (rhodamine phalloidine, red, central column) can be seen. Green and red fluorescent images are merged and presented in the right column. (C) The Isl1⁺ CSC-containing CICS in the culture of newborn rat (day in vitro 4). Isl1⁺ CSCs (FITC, green), cytoskeletal actin (rhodamine phalloidine, red), the nuclei (Hoechst, blue) are merged with transmitted light images.

Figure 3.

Cell-in-cell structures (CICSs) identified in the suspension of freshly isolated myocardial cells (ex vivo) of 20- and 40-day-old rats. Transmitted light and fluorescent images are merged. (A and B) Isl1⁺ CSCs inside cardiomyocytes of 20-day-old rats (Isl1, green), α-Sarcomeric actin, red). (C) c-kit⁺ CICS. (40-day-old rat, c-kit, green; Ki67, red). (D) c-kit⁺ CICS. (40-day-old rat, c-kit, green; α-Sarcomeric actin, red).

Figure 4.

The release of transient amplifying cells from the CICSs. (A) The ruptured CICS capsule in the live cell culture (40-day-old rat, day in vitro 21). (B) The ruptured capsule of the Isl⁺ CSC-containing CICS still within the sarcoplasm (newborn rat, day in vitro 25). (C) The exit of Isl⁺ cardiac progenitors (newborn rat, day in vitro 14). See also Video S3. (D) The cluster of progenitors released to the outside of the host cell, now in the live cell culture (newborn rat, day in vitro 19). The fragments of the capsule can be seen in the center of the image. (E) Clusters of c-kit⁺/α-sarcomeric actin⁺ transient amplifying cells in aggregate (arrow) and locally dissociated (newborn rat, day in vitro 10). (F) Dozens of Isl⁺ transient amplifying cells among mature CMs (newborn rat, day in vitro 16). (G) An aggregation of c-kit⁺/α-actinin⁺ transient amplifying cells embedded into the mucous media (newborn rat, day in vitro 20). (H) A cluster of c-kit⁺/α-actinin⁺ transient amplifying cells spreading over the substrate; dividing cells are indicated by an arrow (newborn rat, day in vitro 20). (I) C-kit⁺/α-actinin⁺ cells separated from each other (newborn rat, day in vitro 20). A binucleated cell is shown (arrow). In (B, C and E-I), transmitted light and fluorescent images are merged.

Figure 5.

Schematic representation of the sequence of events during the intracellular development of cardiac stem cells inside mature cardiomyocytes.

Discussion

The present study represents the first description of the residence, survival, division, and partial differentiation of CSCs within mature CMs. The occurrence of viable cells internalized within different kinds of host cells has been recognized for more than 100 years (for review, see).¹⁰ To date, 3 major types of cell-in-cell interactions have been described: cannibalism, emperipolesis, and entosis, which differ in both effector and host cell identity, mechanism of penetration, and function. Cannibalism is frequently observed in neoplasms, in which the tumor cells engulf either malignant or immune cells followed by intracellular degradation of the effector cell.¹¹ It has been shown that cannibalism favors the survival and proliferation of the host cells: in the case of lymphocyte ingestion, this is most likely due to the escape of the host cells from immunologic attack; preferential host cell survival might also be due to the increased nutrient supply of the host cell in the process of cytophagocytosis.¹² In contrast, emperipolesis is thought to be an active invasion of the host cell by lymphocytes, neutrophils, or natural killer cells.¹³ The apparent lack of both host and target cell damage is an important feature of this phenomenon at its initial stage, although eventually the effector cell either kills the host, is killed by it, or simply exits it. Additionally, in emperipolesis the host cells can even nurture their effector cells, as has been documented for thymic nurse cells containing T-lymphocytes,¹⁴ and for fetal liver Kupffer cells harboring erythroblasts.¹⁵ The third form of cell-in-cell interaction, entosis, was first noted for epithelial cells that entered the cytoplasm of their neighbors and detached from each other and/or the extracellular matrix.¹⁶ The effector cell thus internalized was shown to either die, exit its host, or even proliferate inside it. Both emperipolesis and entosis appear to share similar features with the development of CSCs inside cardiomyocytes, as considered thoroughly below, although the distinct features of the latter also deserve special attention.

To our knowledge, CICSs have never previously been described for CMs. Although hybrid binucleated cells produced by the fusion between CMs and bone marrow-derived haematopoietic cells,¹⁷ lymphocytes,¹⁸ or adipose tissue-derived stem cells,¹⁹ have been previously implicated in myocardial regeneration, these hybrids differ significantly from CSC-containing CICSs described herein. The results of the present study demonstrated both intracellular residence of CSCs and their ability to proliferate inside the intrasarcoplasmic capsule as evidenced by Ki67 expression (Fig. 1C, Fig. 3C). Capsule rupture resulted in the release of Hoechst-positive cells (Fig. 4C, D, G, Video S3), followed by their attachment to the substrate (Fig. 4E, F, H, I), which is indicative of cell viability. The following stages of CSCs release could be recognized on Fig. 4: partial dissociation of c-kit⁺/α-sarcomeric actin⁺ TAC clusters (Fig. 4E), attachment of TACs to the substrate (Fig. 4F), and their flattening (Fig. 4H, I).

As noted above, prominent changes in the CICSs were observed during cell culture, including their significant enlargement as well as the thickening of their coating and the formation of a few pores therein. The outer rhodamine-phalloidin-positive layer of the capsule (Fig. 1D) suggests that it might consist of compacted actin cytoskeleton, presumably of host cell origin. Surprisingly, the presence of a prominent capsule surrounding the intracellularly localized CSCs resembles some of the features of Apicomplexa parasite invasion into intestinal epithelial cells with its subsequent encapsulation.²⁰ Electron microscopic analysis will be needed to provide a more detailed description of the capsule in particular and the organization of the CICS as a whole. Two inter-related events are of special interest: capsule rupture (Fig. 4A, B) and the exit of its obviously cellular content (Fig. 4C–E). Both large and thick, the capsule is poorly permeable for immunocytochemical agents, thus resulting in the lack of reliable CSC labeling in our analyses. In contrast, upon capsule rupture the progenitor cells stained intensely for both CSC-specific antigens and nuclear DNA (Fig. 4C), and were shown to be released into the host cell cytoplasm and to migrate out of the maternal cell. The number of uniform progenitor cells scattered around the capsule fragments was estimated to be approximating 200 (Fig, 4F), concurrent with the in vivo findings showing that genetically tagged CSCs each generated 230 CMs.²¹

The process of intracellular CSC development in the culture of myocardial cells obtained from neonatal rats coincides in time with CM hypertrophy and formation of contracting CM colonies from single CSCs.²² The first contractions of CSC-derived CM colonies heralding cardiomyogenic differentiation of CSCs within the colony were recorded starting from the 8th day of culturing, as opposed to the process of CSC-containing CICS maturation, which required 20–25 days in culture prior to capsule rupture. In addition, contracting CM colonies were not found in cell cultures obtained from 20 and 40 day old rats, whereas CSC-containing CICSs were identified in cultured cardiomyocytes irrespective of the age of the source animal.

Under physiological conditions, the stem cells of different organs are localized in specialized microdomains that have been termed “niches” by Schofield.²³ The microenvironment of the niche is considered to protect the stem cells from a variety of noxious agents and to contribute to the renewal and differentiation of the stem cells by symmetric and asymmetric division, respectively. In the heart, clusters of CSCs and also their progenies were found between contracting CMs.²⁴ c-kit-expressing CSCs were shown to have direct contacts through gap junctions with surrounding CMs, thereby implying that in the cardiac niche the CMs serve as supporting cells for the CSCs.²⁴ Relatively abundant in the atria and apex of the heart, the CSC niches are thought to be protected from hemodynamic stress,^7,25 this description is not inconsistent with the findings of the present study on intracellular residence of CSCs. There are at least 2 arguments in favor of this similarity. First, the intracellular milieu itself provides the highest possible protection for CSCs and, second, the bidirectional transport of key regulatory factors involved in the cross-talk between encapsulated CSCs and the host/supporting CMs most likely proceeds through the pores identified in the capsule (Fig. 1E–G, Fig. 3A, C, D).

There are, however, still many issues which remain unresolved. Since the specificity of CSC membrane markers has been questioned, the additional rigorous experimental evidence should be gathered in support of the fact that the cells identified belong to different types of CSCs. On the other hand, it is impossible to document that these cells fulfill such criteria of CSC as self-renewal, clonogenicity and multipotency,⁸ because the cells inside cardiac myocytes undergo several rounds of division, which is associated with their partial cardiomyogenic differentiation. Therefore, the encapsulated cells could be defined as TACs. Partially differentiated TACs lose the ability to produce clones meaning that their clonogenicity and multipotency cannot be demonstrated in principle. Proof awaits the demonstration that TACs can undergo final differentiation into contracting myocytes. Notwithstandingly, Omatsu-Kanbe et al.^26,27,28 have described in mice a population of small (˜10 µm) self-beating cardiomyocytes of “a novel type,” which start to contract soon after plating without prior division. It might be speculated that these authors described the final cardiomyogenic differentiation of CICS-derived TACs.

Moreover, the question might arise regarding the rate at which the CICS-bearing CMs might be found with the methods utilized in the present study. In fact, 2–3 CICS-containing CMs were found per 2 × 10⁵ plated cells, which corresponds well with the published occurrence of contracting CM colonies in cardiac cell culture.²² The outstanding question remaining regards the mechanism of CSC internalization into the host CMs. A plausible option would be the penetration of CSCs into the CM through the T-tubule canals, although alternative explanations cannot currently be ruled out. It is also crucial to define whether CSC-containing CICSs exist in vivo in the other species. We feel that intracellular development of CSCs represents a new biological principle of cardiomyocyte renewal, which, along with clone formation, may be responsible for myocardial self-renewal and regeneration in mammals.

In conclusion, further investigation of the development of intracellular CSCs is required for better understanding of the mechanisms that govern cardiac regeneration in healthy and diseased states. These studies might substantially advance our knowledge regarding the biology of cardiac stem cells and their potential in myocardial repair.

Materials and Methods

Animals

Newborn, 20-, and 40-day-old Wistar rats were used throughout the study. The experiments were performed in accordance with the Guide for the Care and Use of Laboratory Animals and were approved by the local ethics committee.

Isolation of cardiac cells

Dissected hearts were enzymatically dissociated into a single cell suspension as previously described (Lam et al., 2002). Briefly, the hearts were excised and rinsed in Ringer's solution (pH 7.4) consisting of 146 mM NaCl, 5 mM KCl, 2 mM CaCl₂, 1 mM MgCl₂, 11 mM glucose, and 10 mM HEPES. After mincing and incubation in the same solution with the addition of 1 mg/mL collagenase IA (Sigma-Aldrich, USA) and 0.12% trypsin (FLUKA, Sigma-Aldrich) at 37°C for 30 min, the suspensions thus obtained were left to rest without further stirring for 2–3 min to precipitate the undissociated tissue fragments. The supernatant was centrifuged at 400 × g for 10 min for the enrichment of viable cells.

Cell culture

The cells were transferred to DMEM supplemented with 10% fetal calf serum (Biolot, Russia), 50 U/mL penicillin, and 50 μg/mL streptomycin (Biolot) followed by a one-hour preincubation in Petri dishes for at least partial purification from the non-myocytic cells. Then the cells were plated at an initial density of 1 Χ 10⁵ cells/mL on 12 × 24 mm glass strips pre-coated with 0.1 mg/mL poly-D-lysine (Sigma-Aldrich) to be further incubated in 40 mm Petri dishes (Medpolimer, Russia) in a CO₂ incubator (Jouan, France) at 37°C in humid air containing 5% CO₂. The medium was changed twice a week.

Immunocytochemistry

After rinsing with PBS and fixation for 20 min in 2.5% paraformaldehyde at room temperature, the cells were permeabilized with 0.25% Triton-X100 for 10 min. For immunostaining of the CSCs, 3 protocols were used: 1) immunostaining was performed using 5 μg/mL mouse anti-c-kit monoclonal antibodies (Invitrogen, USA), 1:100 mouse anti-Sca1 polyclonal antibodies (Abcam, USA), and 1:100 rabbit anti-Isl1 monoclonal antibodies (Abcam). The secondary antibodies utilized, respectively, were: 1:100 goat anti-mouse FITC-conjugated antibodies (AbD Serotec, United Kingdom), and 1:100 donkey anti-rabbit FITC-conjugated antibodies (AbD Serotec); 2) primary mouse anti-Isl1 and anti-Sca1 (Abcam) antibodies pre-conjugated with Alexa 532, 546, 568, 594, or 647 according to Zenon technology (Invitrogen) were used at a 1:100 dilution; or 3) commercially available FITC-conjugated anti-c-kit antibodies (Abcam) were used at 1:100 dilution for the detection of c-kit⁺ CSCs.

To analyze the cardiomyogenic differentiation of the CSCs, mouse antibodies to sarcomeric α-actinin (Abcam), α-sarcomeric actin (Sigma-Aldrich), cardiac troponin T (Abcam) and cardiac myosin type II (Abcam) were used. In addition to immunolabeling, rhodamine-phalloidin (10 μg/mL, Sigma-Aldrich) and Hoechst (10 μg/mL, Molecular Probes, USA) staining were used for detection of filamentous actin and cell nuclei, respectively.

Ex vivo experiments

The myocardium of 20- and 40-day-old rats was dissociated enzymatically as described above. The enzyme-free cell suspension was fixed and stained using antibodies, followed by suspending the cells between the slide and cover slip.

Visualization

A confocal microscope (Leica TCS SP5, Germany) with 40×, 63× oil, and 63× glycerol objectives was used to visualize CSCs. For optical tomography, the sections were spaced 1.01 or 2.01 μm along the z-axis.

Time-lapse microscopy

The images of the living cardiomyocytes were recorded at a magnification of 40× (glycerol) for 7 days at a rate of 1 frame per min (Zeiss Cell Observer SD, Carl Zeiss, Germany).

Statistical analysis

All of the data are expressed as the mean ± standard deviation. The statistical analyses were performed using the SPSS 13.0 software package (SPSS Inc.. Software, USA).

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Author Contributions

GB performed all experiments, analyzed and interpreted the data, performed the statistical analysis, and wrote the manuscript; AN and MG analyzed and interpreted the data, helped to draft the final manuscript and added important comments to the paper. All authors read and approved the final manuscript.

Supplemental Material

Supplemental data for this article can be accessed on the publisher's website.

Funding

This work was supported by the grants from the Russian Foundation for Basic Research (No 12-04-00941), Program of Presidium of Russian Academy of Sciences “Fundamental Sciences for Medicine” (2012–2014) and by Government of Russian Federation, Grant 074-U01. Time-lapse and confocal microscopy were performed at the Research Resource Center “Molecular and Cell Technologies”of St-Petersburg State University.