CD117-Positive Cells in Adult Human Heart Are Localized in the Subepicardium, and Their Activation Is Associated with Laminin-1 and α6 Integrin Expression

Authors


Abstract

CD117-positive cells contributing to cardiac cell turnover in normal and pathological conditions have recently been described in adult human heart. Since the precise spatial and temporal expression of extracellular matrix proteins and their receptors is critical for organ formation, we compared the distribution of cardiac primitive CD117-positive cells in the human adult normal and pathological hearts with ischemic cardiomyopathy, with respect to localization and expression of laminin and integrin isoforms. In the pathological hearts, CD117-positive cells were significantly more numerous than in the normal hearts. They were localized mainly in the atria and were up to 38-fold more numerous in the subepicardium than in the myocardium. Compared with normal hearts, most CD117-positive cells in the subepicardium of pathological hearts were α6 integrin-positive. Laminin-1, typical of developing heart, was found predominantly in the subepicardium of adult heart. Immunoblotting revealed its highest expression in the normal atrium and pathological left ventricle. Both laminin isoforms reduced apoptosis and increased proliferation and migration of CD117-positive cells in vitro with respect to control, but the effects of laminin-1 significantly outweighed those of laminin-2. Signaling mediated by α6 integrin was implicated in the migration and protection from apoptosis, as documented by transfection with specific small interfering RNA. These data reveal that the increase in the number of cardiac CD117-positive cells and the expression of laminin-1 are observed in ischemic cardiomyopathy. Subepicardial localization of CD117-positive cells and expression of laminin-1 and α6 integrin subunits may all correspond to the activation of regeneration involving an epithelial-mesenchymal transition recently described in adult heart.

Disclosure of potential conflicts of interest is found at the end of this article.

Introduction

Author contributions: C.C. and F.D.M.: conception and design, collection and assembly of data, data analysis and interpretation, final approval of manuscript; D.N.: conception and design, collection and assembly of data, data analysis and interpretation, manuscript writing; G.R., C.M., C.B., P.M., and M.B.: provision of study material; M.C.: provision of study material, administrative support; S.M.: administrative support.

Stem cells and progenitor cells with the capacity to differentiate into three major cardiac cell types—cardiomyocytes, smooth muscle cells, and endothelial cells—have been described in both embryonic [1, [2]–3] and adult [4, [5]–6] heart tissue. These cells have been characterized by an array of membrane, cytoplasmic, and nuclear antigens, yet the expression of different markers could be associated with the degree of stem cell differentiation. Although the population of resident cardiac stem cells remains to be determined, many studies have reported the presence of hematopoietic lineage-negative, stem cell factor receptor (CD117)-positive primitive cells in the adult myocardium [4, 5, 7].

It is known that the precise spatial and temporal expression of extracellular matrix proteins and their receptors is critical for proper organ formation during organogenesis [8]. Laminin-1 (α1β1γ1) is the first extracellular matrix protein to be expressed during embryonic development, and it has been observed that heart organogenesis does not proceed in the absence of this protein [9]. The absence of laminin-2 (α2β1γ1), an isoform typical of the muscle, causes congenital muscular dystrophy with cardiac involvement [10]. Similarly, the presence of specific integrins influences the fate and biological properties of cells [11]. Receptors for laminin-1 and -2 present on adult cardiomyocytes include integrins α1β1, α3β1, and α7β1, whereas α6β4 is typically found on epithelial cells [12]. Although a primary function of α6β4 integrin is to maintain the integrity of epithelia, because of its ability to mediate the formation of hemidesmosomes on the basal cell surface [13], some studies provide definitive evidence to implicate α6β4 in migration, documenting its localization in membrane protrusions associated with migration [14]. Importantly, the acquisition of motile properties by epithelial cells can be a consequence of an epithelial-mesenchymal transition [15].

Epithelial-mesenchymal transition is one of the major events in heart organogenesis. Studies of developing avian embryos have shed light on the mechanisms by which cardiac cells are formed from the proepicardial organ [16]. The epithelial cells of epicardium migrate over the surface of the developing myocardium and subsequently invade the subepicardial space and acquire a mesenchymal phenotype. The epithelial-mesenchymal transition is triggered by an interplay of extracellular matrix proteins and soluble growth factors and involves dissociation of tight junctions, acquisition of motile properties, and molecular changes involving epithelial and mesenchymal markers [15]. Recent studies have observed the preservation of vasculogenic potential of adult epicardial cells in vitro [17, 18], but the possibility that these cells are really cardiac stem cells of the human heart awaits investigation.

We have analyzed and compared the spatial distribution of CD117-positive cells in the adult human normal and pathological hearts with chronic ischemic cardiomyopathy, with respect to the localization and expression of laminin-1 and laminin-2 and their receptor, integrin α6. Furthermore, the role of laminin-1 and laminin-2 in the primitive adult cardiac cell proliferation apoptosis and migration was examined in vitro.

Materials and Methods

Samples of normal adult hearts were derived from patients who died for reasons other than cardiovascular disease (n = 11; mean age, 41 ± 12 years; seven males, four females). Pathological hearts were explanted because of end-stage heart failure associated with ischemic cardiomyopathy (n = 20; mean age, 55 ± 5.5 years; 14 males, 6 females; mean ejection fraction, 25% ± 1%). In each case, fragments of right ventricle, left atrium, left atrioventricular junction, left ventricle, and apex were excised across the full thickness of heart wall, including epicardium and endocardium. The investigation conformed with the principles outlined in the Declaration of Helsinki.

Immunofluorescence

Heart tissue was embedded in Killik cryostat embedding medium (Bio-Optica, Milan, Italy, http://www.bio-optica.it), frozen, and stored at −80°C or fixed in formaldehyde, embedded in paraffin, and then sliced into serial 4-μm-thick sections. Double labeling was performed using primary antibodies against CD117 (monoclonal IgG1; Dako, Glostrup, Denmark, http://www.dako.com), α-sarcomeric actin (monoclonal IgG2a) and fibronectin (rabbit polyclonal; both from Sigma-Aldrich, St. Louis, http://www.sigmaaldrich.com), laminin α1 and α2 chain (both rabbit polyclonal), α4 (goat polyclonal) and α6 (rabbit polyclonal) integrin (all from Santa Cruz Biotechnology Inc., Santa Cruz, CA, http://www.scbt.com), and secondary antibodies conjugated with fluorescein or rhodamine (Jackson Immunoresearch Laboratories, West Grove, PA, http://www.jacksonimmuno.com). Nuclei were counterstained with 4,6-diamidino-2-phenylindole (DAPI). Negative controls were included for each staining using an isotype-matched nonspecific antibody. Circulating blood cells were excluded from the analysis by the use of antibodies against the following hematopoietic lineage markers (supplemental online Fig. 1): CD34, CD45, CD133 (Miltenyi Biotec, Bergisch Gladbach, Germany, http://www.miltenyibiotec.com), CD8, and CD20 (BD Biosciences, San Jose, CA, http://www.bdbiosciences.com). Microscopic analysis was performed with a Leica DMLB microscope equipped for epifluorescence (Leica Microsystems, Wetzlar, Germany, http://www.leica.com). For every field, images corresponding to different immunofluorescence filters were taken with a digital camera connected to a microscope (Leica DC200; Leica Microsystems) and then merged (Leica QFluoro; Leica Microsystems). The number of CD117-positive cells per 100 mm2 was established by counting all positive cells in the section and measuring the area of the entire section (SigmaScan Pro5 software; SPSS, Chicago, http://www.spss.com). Counting was performed by three independent investigators blinded to the type of sample. Each slide was analyzed three times, and the results were averaged.

For the characterization of cells in vitro, CD117-positive cells were fixed with 4% paraformaldehyde. Nuclear and cytoplasmic markers of different cardiac cell lineages were labeled with primary antibodies against Nkx 2.5 (rabbit polyclonal), α-sarcomeric actin (monoclonal IgG2a), Ets-1 (rabbit polyclonal), GATA-6 (goat polyclonal; all from Santa Cruz Biotechnology), von Willebrand factor (monoclonal IgG1; Chemicon, Temecula, CA, http://www.chemicon.com), smooth muscle actin (monoclonal IgG2a; Sigma-Aldrich), and secondary antibodies conjugated with fluorescein or rhodamine (Jackson Immunoresearch Laboratories). Nuclei were counterstained with DAPI.

Immunoprecipitation and Immunoblotting

Protein extracts were prepared from fragments of normal and pathological hearts. Lysates containing 300 μg of proteins were incubated with antibody against laminin α1 or α2 chain (both rabbit polyclonal; Santa Cruz Biotechnology) and then incubated with protein G-Agarose beads (Invitrogen, Carlsbad, CA, http://www.invitrogen.com). Nonspecific IgG was used as a negative immunoprecipitation control. The immunoprecipitated proteins were size fractionated by electrophoresis on 8% SDS-polyacrylamide gel and transferred onto a nitrocellulose membrane. Molecular weight markers (Bio-Rad, Hercules, CA, http://www.bio-rad.com) and denaturated samples of purified laminin (Santa Cruz Biotechnology) were loaded onto each gel as a weight control and a positive control, respectively. The membranes were blocked and then incubated with anti-β1 laminin antibody (monoclonal IgG2a; Santa Cruz Biotechnology) followed by horseradish peroxidase-labeled secondary IgG. Antibody binding was visualized by chemiluminescence (Amersham Biosciences, Piscataway, NJ, http://www.amersham.com) and autoradiography. The intensity of individual bands was determined using ImageJ software (NIH).

RNA Interference

CD117-positive cells from adult human hearts were transfected with pooled specific human α6 integrin small interfering RNA (siRNA) duplex oligoribonucleotides at a concentration 20–50 nM in the presence of Lipofectamine RNAiMAX. For assessment of double stranded RNA delivery, red fluorescent oligo not homologous to any known gene was used on control cells. Negative control duplexes served as a control against the nonspecific RNA interference. All reagents for siRNA experiments were purchased from Invitrogen. The reduction of α6 integrin expression was controlled at the protein and mRNA levels by Western blotting and reverse transcription-polymerase chain reaction, respectively.

Proliferation and Apoptosis in the Presence of Laminin-1 and Laminin-2 In Vitro

For in vitro assays, CD117-positive cells were isolated from the fragments of left ventricular myocardium of pathological hearts. The primary culture was obtained according to the protocol described previously [19], with some modifications. Briefly, tissue fragments were minced and digested in the presence of 0.25% trypsin-EDTA and collagenase II (0.1% wt/vol). Cardiomyocytes were removed by sequential centrifugation, and the supernatant was filtered with a 40-μm nylon cell strainer (Becton, Dickinson and Company, Franklin Lakes, NJ, http://www.bd.com). The cells were plated at a density of 2 × 104 cells per cm2 in Dulbecco's modified Eagle's medium-Ham's F-12 medium (Sigma-Aldrich) supplemented with 5% fetal calf serum (Invitrogen), basic fibroblast growth factor (Peprotech, Rocky Hill, NJ, http://www.peprotech.com), glutathione (Sigma-Aldrich), penicillin, and streptomycin (Invitrogen) and allowed to proliferate. Once adherent cells were more than 75% confluent, they were detached with 0.25% trypsin-EDTA, and CD117-positive cells were purified using positive selection with anti-human-CD117 MicroBeads (Miltenyi Biotec) as recommended by the manufacturer. Purity of sorted cells was determined by immunofluorescence and reached 97%.

CD117-positive cells from adult human heart were cultured on bovine serum albumin (BSA; Sigma-Aldrich), laminin-1 (purified natural protein; R&D Systems Inc., Minneapolis, http://www.rndsystems.com), and laminin-2 (purified natural protein; Sigma-Aldrich)-coated (5 μg/cm2) chamber slides (Becton Dickinson). For evaluation of proliferation, quiescent cells were incubated with complete medium for 24 hours, and 5-bromo-2′-deoxyuridine (BrdU) was added (1:1,000) for 1 hour before cell fixation. Incorporation of BrdU was evaluated using the BrdU Labeling and Detection Kit (Roche Diagnostics, Basel, Switzerland, http://www.roche-applied-science.com) according to the manufacturer's protocol. For evaluation of apoptosis, the fragmentation of DNA after 24 hours of culture was detected using the ApopTag Plus Fluorescein In Situ Apoptosis Detection Kit (Chemicon) according to the manufacturer's protocol. All in vitro experiments were repeated a minimum of three times in triplicate.

Migration in the Presence of Laminin-1 and Laminin-2 In Vitro

The speed of migration and the number of migrating cells were evaluated with a scratch wound assay and a Boyden chamber-based migration assay, respectively. For the assessment of the speed of transverse displacement, a scratch wound was inflicted across the confluent monolayer of cells, and then the movement of cells across the scratch area was observed. The distance between the cells at opposite scratch edges was measured in five distinct points at the regular time intervals using a computer-assisted contrast-phase microscope (CKX41; Olympus, Tokyo, http://www.olympus-global.com).

To assess invasive displacement, a modified Boyden chamber assay was performed. Transwell membranes (8 μm; Chemicon) were coated on both sides with BSA, laminin-1, or laminin-2 (5 μg/cm2) overnight at +4°C. Control and α6 integrin-siRNA-transfected cells resuspended in serum-free medium containing 5% BSA were added to coated Boyden chambers (5 × 105 cells per well) placed over the triplicate wells with complete medium. After 24 hours of incubation at 37°C, cells remaining on the upper surface of the filters were mechanically removed with a flattened swab, and cells that had migrated to the lower surface were fixed with 4% formaldehyde, stained with DAPI, and counted, by two independent investigators blinded to samples, in five fields at 40× using a fluorescence microscope (Leica DC200; Leica Microsystems). The results were expressed as percentage of migrated control cells.

Statistics

Results are expressed as mean ± SEM. Statistical differences were evaluated using Student's two-tailed t test for comparison between pairs of groups, and analysis of variance with post hoc Bonferroni's t test was used when multiple groups were compared. p values <.05 were considered statistically significant.

Results

Localization of CD117-Positive Cells in the Adult Human Heart

The presence of CD117-positive, hematopoietic lineage-negative cells was analyzed by immunofluorescence (Fig. 1; supplemental online Fig. 2). In the normal hearts, they were most numerous in left atrium, followed by right ventricle, left ventricle, atrioventricular junction, and apex (Fig. 2A). In the hearts with chronic ischemic cardiomyopathy, CD117-positive cells were significantly more numerous than in the normal hearts. Interestingly, their distribution in the various heart regions did not differ significantly between damaged and normal hearts, and in the pathological hearts, also, it was the left atrium that contained more CD117-positive cells, followed by the left ventricle and the right ventricle, apex, and atrioventricular junction (Fig. 2B).

Figure Figure 1..

CD117-positive cells in the normal adult human heart. The presence of CD117-positive cells was detected by immunofluorescent labeling of heart sections. Top row: two CD117-positive ([A], red) cells (arrow) within the subepicardium expressed α6 integrin subunit ([B], green). Middle row: CD-117-positive ([E], red) cell (arrow) within the myocardium is α6 integrin-negative (F). Bottom row: fibronectin ([I], red) lined epithelial cells within the epicardium; all cells forming the epicardium were α4 integrin-positive ([J], green) and colocalized with fibronectin. Nuclei of cells were counterstained with 4,6-diamidino-2-phenylindole ([C, G, K], blue). (D, H, L): Overlays of the other three images from each row.

Figure Figure 2..

Quantification of CD117-positive cells in the adult human normal and pathological heart. CD117-positive α6 integrin-positive (gray bars) and α6 integrin-negative (white bars) cells in the E and M of RV, LA, AVJ, LV, and Apx were visualized by immunofluorescence and quantified. The bars correspond to the mean ± SEM number of CD117-positive cells in 100 mm2. (A): In the normal hearts CD117-positive cells were localized mainly in the epicardium of the atrium and RV; the fraction of α6 integrin-expressing cells was identical in the subepicardium E and M. (B): In the pathological hearts CD117-positive cells were significantly more numerous (note the difference in the scale), with their highest number in the E of LA, LV, and RV; the fraction of α6 integrin-positive cells was two- to fourfold higher than in the normal hearts and twofold higher in the E than in the M. *, p < .05 between CD117-positive α6 integrin-positive cells in the E of LA and all other regions. Abbreviations: Apx, apex; AVJ, atrioventricular junction; E, epicardium/subepicardium; LA, left atrium; LV, left ventricle; M, myocardium; RV, right ventricle.

In both the normal and the pathological hearts, CD117-positive cells were strikingly more numerous in the epicardium and subepicardium than in the main myocardium. In the normal hearts (Fig. 2A), this difference reached a maximum of 27.5-fold in the left atrium (p < .0001) and a minimum of 6.7-fold in the apex (p < .05). In the hearts with ischemic cardiomyopathy (Fig. 2B), the subepicardium within the left atrium contained 38-fold more CD117-positive cells than the main myocardium (p < .0001). Similarly, there were 10.5-, 42-, 8-, and 37.5-fold more CD117-positive cells in the subepicardium compared with the main myocardium of the atrioventricular junction (p < .0001), left ventricle (p < .005), apex (p < .0001), and right ventricle (p < .0001), respectively.

It emerges that the increase in the number of CD117-positive cells in chronic pathological conditions is more pronounced in the subepicardium than in the myocardium of all heart regions. Compared with normal hearts, the number of CD117-positive cells in the hearts with ischemic cardiomyopathy rises 10-fold in the subepicardium and only 5-fold in the myocardium of the left ventricle.

Integrin α6 Expression of CD117-Positive Cardiac Cells

A fraction of CD117-positive cells expressed the α6 integrin subunit (Fig. 1B), which together with β4 constitutes a typical and specific receptor for laminin. Examination of heart sections by immunofluorescence showed that this fraction was almost identical in the subepicardium (19.51% ± 5.6%) and myocardium (21.67% ± 3.48%; p value not significant) of the normal hearts (Fig. 2A). In the pathological hearts (Fig. 2B), by contrast, α6 integrin-expressing cells constituted 82.73% ± 9.1% of CD117-positive cells in the subepicardium and only 40.7% ± 17% of CD117-positive cells in the myocardium (p < .05) of the left atrium. All the numbers referring to the CD117-positive cells with or without α6 integrin coexpression in different heart regions and cardiac wall sections are reported in supplemental online Table 1.

Laminin Isoforms Expression in the Adult Human Heart

Laminin α1 chain makes part of laminin-1 and -3, whereas α2 laminin is present in laminin-2, -4, and -12 isoforms [20]. Of these, only laminin-2 and -4 (merosins) are present typically in the normal adult human heart, whereas laminin-1 is expressed in the developing myocardium [9]. Fluorescent labeling of heart sections with antibodies specific for laminin α1 and α2 chains revealed the presence of laminin-1 and merosins in the human adult heart. In the left ventricle of normal hearts, the presence of laminin-1 was restricted to the epicardium and subepicardium, whereas laminin α2 chain bordered cardiomyocytes but did not entirely surround them. In the hearts with ischemic cardiomyopathy, laminin-1 formed a meshed network within the subepicardium, with frequent inward-reaching branches (Fig. 3A), and filled the interstitial space in the myocardium, revealing a granular-like pattern of fluorescence around the cardiomyocytes (Fig. 3B, 3C). Merosins lined the epicardium (Fig. 3D) and formed a conspicuous network lining the basement membrane of the cells in the myocardium (Fig. 3E, 3F).

Figure Figure 3..

Laminin-1 and merosin expression in the adult human heart. The presence of laminin-1 and merosins (laminin-2 and -4) in the adult human heart with ischemic cardiomyopathy was detected by immunofluorescent labeling of heart sections with specific antibodies against α1 or α2 laminin (green). Antibody against α-sarcomeric actin was used to stain cardiomyocytes (red); nuclei of cells were counterstained with 4,6-diamidino-2-phenylindole (blue). (A): Laminin-1 filled the subepicardium with a meshed network; a few branches spread between epicardium and myocardium were present. (B, C): A granular-like pattern of laminin-1 immunofluorescence were observed within the myocardium, where laminin-1 filled interstitial spaces and clustered around the cardiac cells. (D–F): Merosins lined the epicardium and formed a conspicuous network corresponding to the basement membrane of the cells in the myocardium.

Protein analysis by immunoprecipitation of α1 or α2 followed by electrophoresis and immunoblotting of β1 chain showed differences in the expression of laminin-1 and laminin-2 between different regions, as well as between the same regions of the normal and pathological hearts. Whereas in the atria the expression of laminin α1β1 chains was only slightly higher in the normal than in pathological hearts, in the left ventricle it was up to threefold higher (p < .001) in the hearts with ischemic cardiomyopathy (Fig. 4). Moreover, the expression of laminin-1 in the normal atria was fourfold higher than in the normal left ventricle (p < .001), whereas in the pathological conditions it was the left ventricle that contained more laminin-1 than any other heart region (Fig. 5). Laminin-2 expression between the atria from normal and pathological hearts did not differ significantly, whereas in the left ventricle α2β1 chains were up to 2.5-fold more abundant (p < .001) in the pathological hearts.

Figure Figure 4..

Laminin-1 and laminin-2 expression in the adult human normal and pathological heart. The expression of laminin-1 (left) and laminin-2 (right) isoforms in the adult human heart was detected by IP (α1 or α2) followed by WB (β1) of the proteins from N and P. Representative results with bands of 220 kDa, corresponding to laminin β1 chain, are shown above the bars, indicating their mean ± SEM O.D. Immunoblotting of α-actinin was used as control of IP from an equal amount of proteins in the lysate. *, p < .05, pathological versus normal within the same heart region; #, p < .05, normal left ventricle versus normal atrium. Abbreviations: CTR, positive control for Western blot; IP, immunoprecipitation; N, normal hearts; O.D., optical density; P, pathological hearts; WB, Western blotting.

Figure Figure 5..

Laminin-1 and laminin-2 expression in different regions of adult human pathological heart. The expression of laminin-1 (top) and laminin-2 (bottom) in the adult human hearts with ischemic cardiomyopathy was detected by IP (α1 or α2) followed by WB (β1) of the proteins from RV, LA, AVJ, LV, and Apx. A representative result with bands of 220 kDa, corresponding to laminin β1 chain, is shown above the bars, indicating mean ± SEM O.D. Immunoblotting of α-actinin was used as control of IP from an equal amount of proteins in the lysate. *, p < .05 versus LV. Abbreviations: Apx, apex; AVJ, atrioventricular junction; IP, immunoprecipitation; LA, left atrium; LV, left ventricle; O.D., optical density; RV, right ventricle; WB, Western blotting.

The Effects of Laminin-1 and Laminin-2 on the Proliferation, Apoptosis, and Migration of Cardiac Primitive Cells In Vitro

To investigate the role of different laminin isoforms in the cardiac primitive cell proliferation, survival, and migration, CD117-positive cells were isolated from the adult human heart and cultured. These cells expressed markers of different cardiac cell lineages (Fig. 6), constituting the population of CD117-positive cardiac primitive cells [21]. Specific human α6 integrin siRNA duplex oligoribonucleotides were used to examine the contribution of α6 integrin and laminin interactions to the cell proliferation, apoptosis, and migration in vitro. The efficiency of α6 integrin gene silencing reached maximum of 81% at 24 hours after transfection with 30 nM siRNA, as evaluated at the mRNA level (Fig. 7A).

Figure Figure 6..

CD117-positive cells in vitro. CD117-positive cells were isolated from the fragments of left ventricular myocardium of adult human hearts by enzymatic dissociation and immunomagnetic separation, as described in Materials and Methods. The presence of cells expressing nuclear and cytoplasmic markers of endothelial (Ets-1 and vWF), smooth muscle (GATA6 and smooth muscle actin), and cardiomyocyte (Nkx2.5 and α-sarcomeric actin) cell lineages was shown by immunofluorescence. Magnification, ×400. Abbreviations: DAPI, 4,6-diamidino-2-phenylindole; vWF, von Willebrand factor.

Figure Figure 7..

Proliferation, apoptosis, and migration of cardiac CD117-positive cells in the presence of laminin-1 and laminin-2 in vitro. Specific human α6 integrin RNA interference was used to evaluate the role of α6 integrin. (A): The efficiency of α6 integrin silencing was evaluated at the protein (left) and mRNA (middle) levels, and the silencing reached a maximum of 81%. (B): Proliferation (left) and apoptosis (right) of CD117-positive CTR and α6 integrin silenced cells. (C): The speed of transverse displacement was calculated in the scratch wound assay (left), whereas the invasive migration was evaluated in the Boyden chamber. The bars correspond to the mean value ± SEM. *, p < .05 laminin-1 and laminin-2 CTR versus BSA CTR; #, p < .05 small interfering RNA versus the respective CTR. Abbreviations: BrdU, 5-bromo-2′-deoxyuridine; BSA, bovine serum albumin; CTR, control cells; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; h, hours; siRNA α6-int, small interfering RNA α6 integrin silencing; siRNA neg, small interfering RNA negative transfection control.

Both laminin-1 and laminin-2 increased the proliferation of cells, evaluated by the incorporation of BrdU in vitro, with respect to BSA (3.7% ± 0.1%; n = 3). In the presence of laminin-2 the proliferation rate (Fig. 7B) reached 7.9% ± 0.4% (n = 4), whereas during the culture of cardiac primitive cells on laminin-1-coated dishes, the incorporation of BrdU was more than fourfold higher and reached 32.7% ± 1.9% (n = 4; p < .001). Incorporation of α6 integrin siRNA did not significantly influence proliferation rate in the presence of either laminin-1 or laminin-2. Moreover, apoptosis in the presence of laminin-1 was twofold lower (2.6% ± 0.3%; n = 5) than in the culture of cells on laminin-2 (4.3% ± 0.5%; n = 5; p < .001) and sixfold lower than in the presence of BSA alone (12.2% ± 1.8%; n = 3; p < .001). These effects of both laminin isoforms were inhibited after α6 integrin RNA interference, when the apoptosis rates increased sevenfold (19.1% ± 2.2%; n = 5; p < .001) and fivefold (21.17% ± 2.7%; n = 5; p < .001) in the presence of laminin-1 and -2, respectively, implicating α6 integrin and laminin interaction in the regulation of CD117-positive cell survival in vitro.

The presence of laminin-1 or laminin-2 increased both transverse and invasive displacement of cardiac primitive cells in vitro with respect to control (Fig. 7C). In the scratch wound assay the speed of migration was 36.8 ± 2.4 μm/hour (n = 4) in the presence of albumin alone, whereas it reached 51.5 ± 5.4 μm/hour (n = 3) on laminin-2-coated dishes and 94.7 ± 7.1 μm/hour (n = 3) on laminin-1-coated dishes. In a similar manner, the number of cells migrating across laminin-1 and laminin-2 was significantly higher with respect to control. Because integrins are essential for cell motility, we next studied the role of α6 integrin in the migration of cardiac primitive cells across laminin-1 and laminin-2. When CD117-positive cells from adult human hearts were transfected with the pooled specific human α6 integrin siRNA duplex oligoribonucleotides, the invasive displacement in the laminins presence diminished significantly. In all cases, migration on laminin-1 significantly outweighed that on laminin-2.

Discussion

Both different laminin isoforms and integrin subunits have different functions and convey specific signals within the cell. Moreover, integrins are associated with growth factor and cytokine receptors, coordinating the response of cells to the changes of the fibrillar and soluble components of extracellular matrix [11]. The subepicardial space is rich in both extracellular matrix proteins and growth factors, and in this exact localization, numerous CD117-positive cells were found in the adult human heart in our study. To the best of our knowledge, this is the first study showing that the increase in the number of hematopoietic lineage-negative, CD117-positive cells between human normal and pathological hearts with ischemic cardiomyopathy predominantly involves the cells localized in the subepicardium and to a lesser, although still significant, extent those within the myocardium. The discussion concerning the contribution of bone marrow-derived cells to the population of CD117-positive cells in the adult human myocardium is beyond the scope of the present study. The subepicardial localization of CD117-positive cells seems to correspond to their likely origin, that is, an epithelial-mesenchymal transition of epicardial cells. This is typically the case during heart morphogenesis, when a subset of epicardial cells detaches from epicardial layer and gives rise to different cardiac cell types (as discussed extensively in [22]). The process in which epicardially derived cells acquire mesenchymal phenotype and invade myocardium, giving rise to cells of cardiac lineages, has recently been suggested as the source of stem cells also in the adult heart, and in vitro studies document the multipotentiality of cells derived from adult epicardium [17, 18]. The precise mechanisms controlling the frequency and efficacy of epithelial-mesenchymal transition and subsequent cardiac stem cells differentiation and maturation in the adult normal and pathological heart are currently under intensive investigation.

A numeric increase of cardiac primitive cells in the adult human heart has already been described in the patients with chronic aortic stenosis and ischemic heart failure [23]. In these studies, more c-kit-positive cells were found within the myocardium. It is possible that the epicardial and subepicardial localization of CD117-positive cells has not been taken into consideration, or else different types of pathological stress evoke different types of response, with the prevalent activation of differentiation and maturation of primitive cells resident within the myocardium in the pressure overload and the generation, proliferation, and migration of CD117-positive cells within the subepicardial space in chronic ischemia.

The presence of CD117-positive cells in the subepicardium was also not described in other studies looking for cardiac stem cells niches in the adult heart. Urbanek et al. [24] described clusters of cardiac lineage-negative, c-kit-positive, α4 integrin-positive cells surrounded by fibronectin and laminin α2 chain within the myocardium of normal adult mice. A far-reaching analogy with bone marrow made them conclude that the interaction of α4 integrin with laminin-8/9, -10/11, and fibronectin is implicated in cardiac stem cell renewal and the preservation of their undifferentiated state [24]. This is a very plausible theory; however, to the best of our knowledge, the data showing the interaction of laminin with α4 integrin (the map representing current knowledge on extracellular matrix-receptor interactions is available online from GenomeNet) or the presence of α2 chain in the laminin-8/9 or -10/11 isoforms are missing from the literature. In our study of the localization of CD117-positive primitive cells in the adult human heart, we found only sparse CD117-positive cells in the normal myocardium and numerous CD117-positive cells that expressed α4 integrin within the epicardium lined with fibronectin. This is consistent with the results of Dettman et al. [25], who found that the receptor of fibronectin, α4 integrin, normally restrains epicardial-mesenchymal transition, as well as invasion and migration of epicardially derived mesenchymal cells during organogenesis. Integrin α4 has also been implicated in the epicardium development and integrity during embryogenesis [26, 27]. In the adult, this integrin subunit is typically present on differentiated cells of mesenchymal origin [28], and its interaction with interstitial fibronectin may enable functional integration of newly formed cardiac cells in the myocardial syncytium in vivo.

CD117-positive cells in the subepicardial space colocalized with laminin-1, lacked α4 integrin, and expressed the α6 integrin subunit. Moreover, the branches of laminin-1 could be observed spread between the subepicardium and the main myocardium. Apart from the classic role of integrin α6 in the maintenance of cell adhesion, it has been described for several different cell types that, rather than being downregulated, this subunit becomes widely distributed in the cell membrane and participates in cell migration on laminin-1 [14]. Our results indicate that migration of CD117-positive cells increases in the presence of laminin-1 and -2 in vitro. As demonstrated by the specific α6 integrin RNA interference, the displacement of cells is significantly reduced in the absence of this integrin subunit. Importantly, migration in the presence of laminin-1 exceeded migration in the presence of laminin-2. Moreover, laminin-1 and laminin-2 protect from apoptosis and stimulate proliferation of cardiac primitive cells in vitro. Human α6 integrin expression was not implicated in the proliferation of cells, but its abolishment significantly increased the apoptosis rate in the presence of laminin isoforms. This effect is similar to observations reported by Maitra et al. [28], in which myoblasts transfected with α6 did not proliferate but were able to differentiate. Our findings of the increased fraction of α6 integrin-expressing CD117-positive cells in the pathological human hearts with respect to the normal hearts support the role of α6 integrin-laminin interaction in the survival, migration, and differentiation of cardiac primitive cells activated in chronic pathological conditions.

Laminin-1 predominates among the laminin forms during early embryogenesis and further organogenesis. Its unique role is underlined by the fact that embryogenesis will not proceed in the absence of this form of laminin [9]. To the best of our knowledge, this is the first study documenting the presence of laminin-1 in the adult human heart. In the immunochemical study of the laminin content in embryonic and adult mouse tissues, Sasaki et al. [29] found laminin-2 expression in the normal adult murine heart and skeletal muscle to be the highest compared with all other organs, whereas the content of α1 chain was much lower. Whereas that study analyzed whole-organ tissue extracts, we focused on the differences in laminin expression between different regions (most importantly atrium and left ventricle) and different layers (subepicardium and myocardium) of normal and pathological adult hearts. We found that in the adult human heart, the presence of laminin α1 chain, as detected by immunofluorescence, was restricted mostly to the subepicardial space, with the highest expression in the normal atria and pathological left ventricle. Laminin-1 is essential in epithelial tissue polarization, as well as in epithelial-mesenchymal contact and interactions [30]. These two functions may be essential also in the heart, particularly in the subepicardial space, which hosts CD117-positive cardiac cells. In the liver, an organ with high regenerative capacity in the adult life, laminin α1, which disappears from the space of Disse by 6–8 weeks of postnatal life, reappears during hepatic regeneration [31]. The expression of laminin-1 in the adult human heart, involving molecular reprogramming and revoking the mechanisms operative during organogenesis, may be directly associated with and aimed at the regeneration of cardiac tissue in chronic pathological conditions. The presence of laminin-1 in the subepicardial space of ischemic heart may be essential for creating an environment permissive for epithelial-mesenchymal transition in the adult heart.

It is known that the absence of or alteration in the laminin α2 chain weakens the muscle cell basement membrane, which leads to muscle fiber damage under the stress of contractions [32]. Whereas α2 chain is expressed by differentiated mature cells of mesodermal origin, laminin α1 chain is highly expressed by developing epithelial cells [29]. It follows that although the expression of laminin-1 isoform in the diseased heart may have an important role in the formation of new functionally competent cells, laminin-2 expression would be essential for the maintenance of the pre-existing cardiac cells. Importantly, the differences in the laminin expression in the pathological hearts found and described in our study mostly regarded the laminin α1 chain with the preserved and even increased laminin α2 expression in the ischemic myocardium compared with the normal heart. Moreover, the atria of the normal adult human hearts contained more laminin-1 than the left ventricles, whereas the increase of laminin-1 content in the pathological conditions mostly involved the left ventricular walls. The atrium, a region of low mechanical stress, with the highest CD117-positive cell and laminin content in the normal heart, may constitute the privileged site for the primitive cells, which survive during pathological conditions involving the ventricles and migrate across the subepicardial space or myocardium to the regions of damage and regeneration.

Conclusion

In organs with regenerative capacity, the physiological cell turnover enables the preservation of tissue structure and function, inasmuch as old or dead cells are replaced with the new and better-functioning ones. In pathological conditions, the greater the damage, the greater the resources that must be involved in the healing process, and it has already been suggested that organ regeneration may require molecular reprogramming and reactivation of the mechanisms operative during organogenesis [33]. The epithelial-mesenchymal transition and the generation of CD117-positive cells, as well as the expression of laminin α1 chain in the adult pathological heart, followed by different expression of integrin subunits on cardiac cells, may represent such a process, inasmuch as an extensive damage and chronic pathological conditions activate a regenerative response involving all cardiac stem cells, extracellular matrix, and its receptors.

Disclosure of Potential Conflicts of Interest

The authors indicate no potential conflicts of interest.

Acknowledgements

We thank Drs. Daniela Cesselli and Antonio P. Beltrami for fruitful discussions and careful reading of the manuscript. This work was supported by a grant from Ministero Italiano dell'Università e della Ricerca-Progetto di Ricerca di Interesse Nazionale prot. 2006060854_004 (University “Federico II,” Naples, Italy). C.C., F.D.M., and D.N. contributed equally to this work.

Ancillary