OUP user menu

The human adult cardiomyocyte phenotype

S.D. Bird , P.A. Doevendans , M.A. van Rooijen , A. Brutel de la Riviere , R.J. Hassink , R. Passier , C.L. Mummery
DOI: http://dx.doi.org/10.1016/S0008-6363(03)00253-0 423-434 First published online: 1 May 2003


Aim: Determination of the phenotype of adult human atrial and ventricular myocytes based on gene expression and morphology. Methods: Atrial and ventricular cardiomyocytes were obtained from patients undergoing cardiac surgery using a modified isolation procedure. Myocytes were isolated and cultured with or without serum. The relative cell attachment promoting efficiency of several reagents was evaluated and compared. Morphological changes during long-term culture were assessed with phase contrast microscopy, morphometric analysis and immunocytochemistry or RT-PCR of sarcomeric markers including α-actinin, myosin light chain-2 (MLC-2) and the adhesion molecule, cadherin. Results: The isolation method produced viable rod-shaped atrial (16.6±6.0%, mean±S.E.; n = 5) and ventricular cells (22.4±8.0%, mean±S.E.; n = 5) in addition to significant numbers of apoptotic and necrotic cells. Cell dedifferentiation was characterized by the loss of sarcomeric structure, condensation and extrusion of sarcomeric proteins. Cells cultured with low serum recovered and assumed a flattened, spread form with two distinct morphologies apparent. Type I cells were large, had extensive sarcolemmal spreading, with stress fibers and nascent myofibrils, whilst type II cells appeared smaller, with more mature myofibril organisation and focal adhesions. Conclusion: Characterization of the redifferentiation capabilities of cultured adult cardiac myocytes in culture, provides an important system for comparing cardiomyocytes differentiating from human stem cells and provides the basis for an in vitro transplantation model to study interaction and communication between primary adult and stem cell-derived cardiomyocytes.

  • Adult human cardiomyocyte
  • Cell isolation
  • Cell attachment
  • Laminin
  • Myofibrillogenesis
  • Sarcomere
  • Cadherin

Time for primary review 22 days.

1 Introduction

The myocardium consists of a three-dimensional arrangement of rod-shaped cardiomyocytes attached to adjacent myocytes to form myofibers. Myofibers are adjacent to interstitial fibroblasts, blood vessels and the extracellular matrix, interspersed within a proteoglycan gel matrix. Each cardiomyocyte has a bundle of myofibrils divided into contractile units, or sarcomeres, which consist of several contractile proteins including actin and myosin [1,2]. Our understanding of the developmental processes underlying this tissue complexity and those that maintain it, has advanced in recent years. In particular, several studies have demonstrated the sequence of events in myofibrillogenesis in embryonic heart with respect to the recruitment and assembly of sarcomeric proteins [3,4]. An important point of convergence was the discovery that the pattern of myofibrillar formation in redifferentiating adult cardiomyocytes resembles their embryonic counterparts during development [5,6]. Both systems share a common program of recruitment and there is general agreement with respect to the order of assembly of myofibrillar components, although the exact processes in each system are unknown. In differentiating adult cardiomyocytes, titin and α-actinin polymerization preceded that of myosin and α-actin [5,6].

At the cellular level, myofibrillar orientation is dependent on a guidance system that builds individual myofibrils and aligns developing myofibers into an ordered and coordinated array. The exact identity of the mechanical and molecular signals which orchestrate these events is not known. However, recent studies have indicated that the formation of focal adhesions and cell–cell adhesions play a significant role during this process [7,8].

Damage to the myocardium following ischaemia and reperfusion injury leads to the irreversible loss of cardiac function and is the leading cause of death in western society [9]. The rapidly developing area of stem cell biology may eventually provide an effective treatment to replace damaged tissue associated with ischaemic heart disease and other cardiomyopathies, placing an emphasis on myocardial regeneration or repair [10].

Currently it is unclear at which stage during the differentiation program, stem cell derived myocytes might be used for transplantation therapy. Consequently, detailed characterization of each stage during the progression from stem cell through the progenitor, precursor (cardiomyoblast) fetal and to the adult phenotype is necessary, with profiles of primary human adult and fetal cells in culture providing the essential reference tissue. Various parameters have been used to describe the changes that occur during cardiogenesis and these have involved both contractile and electrical function studies. However, these approaches have limited value at the early stages of stem cell differentiation, since the development and maturation of the excitation–contraction components including ion channel formation, occur after the assembly of the early contractile apparatus. The present study used a combination of morphological studies and molecular analysis of a restricted panel of cardiac genes, to characterize adult human cardiac myocytes isolated from myocardial biopsies. Cultured cardiac myocytes were examined during sarcomeric breakdown and reorganization, thereby providing a complementary model system for comparing myofibrillogenesis in other models such as cardiomyocyte differentiation from stem cells.

2 Methods

2.1 Cardiomyocyte isolation and culture

Cardiac tissue specimens were obtained with informed consent from patients undergoing cardiac surgery. This study was approved by the University Medical Center of Utrecht Ethics Committee. All patients were scheduled for aortic valve replacement because of a transvalvular gradient of >80 mmHg and symptoms. In total we used tissue from seven patients, five males and two females. Mean age was 67±8 years. Atrial cells were obtained from the atrial appendage whilst ventricular cells were obtained from endocardial biopsies obtained during valve surgery. Individual specimens were immediately transferred to ice-cold Ca2+ free Krebs-Ringer saline solution consisting of: 10 mmol/l HEPES, 129 mmol/l NaCl, 4.7 mmol/l KCl, 1.2 mmol/l KH2PO4, 1.2 mmol/l MgSO4·7H2O, 5.0 mmol/l NaHCO3, 5.5 mmol/l glucose, 2.0 mg/ml bovine serum albumin (BSA), further supplemented with 20 mmol/l taurine, 2.0 mmol/l l-carnitine, 5.0 mmol/l creatine, and 30.0 mmol/l 2,3-butanedione monoxime, buffered at pH 7.4 [11]. Specimens were transported to the cell culture laboratory within 15 min. Connective and adipose tissue was trimmed and the myocardial tissue was washed with Krebs-Ringer solution and minced with sterile scissors.

Tissues were subjected to a five step isolation procedure [12]. Briefly, cardiac tissue was incubated for 15 min in Krebs-Ringer solution supplemented with protease XXIV. The partially digested tissue was transferred to Krebs-Ringer saline containing a combination of collagenase A (Roche, Molecular Biochemicals, Mannheim, Germany) and hyaluronidase. The tissue was incubated with this solution for 20 min at 37°C with gentle agitation. Next, the tissue was incubated with 1.0 mg/ml collagenase A solution, three times for 20 min each at 37°C. Rod-shaped cardiomyocytes were visible by phase contrast light microscopy after the second incubation step. After each incubation step, the supernatants were transferred to a tube and centrifuged at 600 rpm for 4 min. The pellets were resuspended in 1 ml of Ca2+ free Krebs-Ringer solution and calcium was reintroduced to the pooled suspension by incremental additions of sterile CaCl2 solution (1.0 mol/l) at 4-min intervals until the final concentration of 1.79 mmol/l was achieved [13]. The cell suspension was decanted into a 10-ml tube, centrifuged at 600 rpm for 4 min and resuspended with medium 199 (M199) supplemented with 20 mmol/l taurine, 2.0 mmol/l l-carnitine, 5.0 mmol/l creatine, 2.0 mg/ml BSA, 100 μU/ml penicillin and 100 μg/ml streptomycin [14]. Culture medium was further supplemented with either 2% heat inactivated foetal bovine serum (FBS) or with 10−7 mmol/l insulin (Roche, Mannheim, Germany; serum free medium). Viability of calcium tolerant cells was assessed immediately after isolation by the appearance of a rod-shaped morphology and the exclusion of trypan blue. Cardiomyocytes were plated and random fields of cells were photographed between 2 and 4 h after isolation, using an Olympus IMT-2 inverted phase contrast microscope (Olympus Optical, Tokyo, Japan), and Kodak Tmax-100 print film (ASA 100). Cells were counted directly from photomicrographs using morphometric analysis.

2.2 Cardiomyocyte attachment

Cardiomyocytes were dispensed onto glass coverslips, 24-well plates (Costar, Corning, USA) or Primaria 35-mm tissue culture dishes (Becton Dickinson, Franklin Lakes, NJ, USA). Substrates without attachment factors were compared with substrates coated with a range of attachment factors including 4% FBS, 0.1% gelatin, matrigel, extracellular matrix, poly-l-lysine, Cell Tak™ and laminin. In addition, experiments were conducted to evaluate whether cell attachment could be stimulated in a dose dependent manner. Cardiomyocytes were dispensed directly onto substrates and cultured for 2–4 h. Cell attachment was assessed by gently moving the culture dish and inspecting adherent cells by phase contrast light microscopy. Cells were cultured overnight and dishes were scored on the basis of the number of cells remaining after the medium was exchanged.

2.3 Immunocytochemistry

Cell morphology was assessed with phase contrast light microscopy and indirect immunocytochemistry. Cells were fixed with 3.0% paraformaldehyde in PBS with Ca2+ for 30 min or alternatively, ice-cold methanol. Cells were washed and permeablized with 0.1% Triton X-100 in PBS for 4 min and washed three times with PBS. Cell monolayers were blocked for several hours with 4% normal goat serum-PBS. Next, primary antibodies diluted with 4% normal goat serum-PBS were applied to coverslips or wells and incubated for 1 h at room temperature. Cryosectioned cardiac tissue and cardiac fibroblasts were used as positive and negative controls, respectively, for sarcomeric markers. Atrial and ventricular tissues were snap frozen and 5.0-μm cryosections were prepared using conventional methods. Monoclonal antibodies included mouse anti α-actinin (1:800), rabbit anti pan-cadherin (1:800), and polyclonal rabbit antibodies directed against atrial (1:500) and ventricular (1:50) isoforms of myosin light chain-2 [1]. For triple labelling experiments, cells were incubated with Cy-3 conjugated goat anti-mouse IgG for 1 h followed by FITC conjugated goat anti rabbit IgG for 1 h. All cell preparations were stained with Hoechst (1:500) for 5 min at room temperature to distinguish between viable and necrotic/apoptotic cells. Coverslip or wells were mounted with Mowiol, on standard glass slides or with coverslips, respectively.

All reagents were of the purest quality obtained from Sigma (St. Louis, MO, USA) unless otherwise stated. Solutions were prepared with ultrapure water obtained from a Milli-Q water system (Millipore, Bedford, MA, USA).

2.4 RT-PCR

RNA was isolated from human atrial and ventricular tissues using Ultraspec (Biotecx Laboratories) and reverse transcribed (1 μg total RNA) as described previously [15]. Primers were designed using primer designer software and three stage PCR for MLC-2a, MLC-2v and β-actin was performed at 56°C and 40 cycles and for atrial natriuretic factor (ANF) and sarcoendoplasmic reticulum Ca2+ ATPase (SERCA), at 55°C and 40 cycles, using a Biometra UNOII cycler (Table 1).

View this table:
Table 1

Primers used for RT-PCR

GenePrimerProduct size (bp)

2.5 Data analysis

All experiments (immunostaining and PCR) were repeated at least three times and the images provided represent typical results.

3 Results

3.1 Attachment and survival of cardiomyocytes

The isolation procedure provided a sub-population of calcium tolerant rod-shaped atrial and ventricular myocytes (Table 2) that had sharply defined edges and distinctive sarcomeric bands (Fig. 1A). Rod-shaped cells stained positive for sarcomeric proteins including α-actinin and atrial or ventricular isoforms of MLC-2, as depicted in Fig. 2B. Cells had well formed I and A band structures with a z-line periodicity of 2.8 μm, similar to myocytes from cryosections (Fig. 1C). The anti-MLC-2 antibodies distinguished human atrial and ventricular cardiomyocytes as shown in Fig. 2B. However, control experiments using cryosectioned myocardium showed that ventricular tissue expressed MLC-2v and weakly expressed MLC-2a (data not shown, but see Fig. 3). Non-viable cells were abundant after the isolation procedure and included elongated cells with ruffled membranes, disorganized sarcomeres, extensive blebbing and dispersed DNA, indicative of advanced apoptosis (Fig. 2A and B). In addition, necrotic (round/square) cells appeared amorphous or had large blebs present on the surface membrane as shown in Fig. 2C and D.

Fig. 3

Expression of regulatory myosin light chain (MLC-2) in human atrial and ventricular myocardial tissue. Isoforms of MLC-2 were detected in adult human myocardial tissue. Atrial tissue expressed MLC-2a whilst ventricular tissue expressed both MLC-2v and MLC-2a. In addition, atrial natriuretic factor (ANF) was present in both tissues. Sarcoendoplasmic reticulum Ca ATPase (SERCA) is expressed at significantly higher levels in ventricular tissue and not detected in atrial tissue. β-Actin was used for control purposes (40 cycles were used in every PCR).

Fig. 2

Cell populations present after isolation (magnification 40×). Phase contrast photomicrograph of an apoptotic cell (A). Nucleic acid stained with Hoechst (blue) was sequestered and extruded into vesicles, which is indicative of advanced apoptosis. MLC-2v stain (red) shows the intact myosin filaments and sequestered protein (B). Necrotic cells appeared rounded or square (C, arrows) or elongated with large membrane blebs (D).

Fig. 1

(A) Isolation of atrial and ventricular cardiomyocytes from fresh human biopsies (magnification 40×). (B) Cells immediately after isolation included quiescent rod-shaped atrial and ventricular cells. Upper panel: staining with polyclonal atrial myosin light chain (MLC-2a, FITC) and α-actinin (Cy-3). Lower panel: ventricular MLC-2(v) and α-actinin. Third panel: merging of both images. (C) Cryosection of ventricular tissue sample stained for sarcomeric proteins; here the MLC-2v antibody was used (40×).

View this table:
Table 2

Cell populations observed after isolation

Rod-shaped cells (%)16±6.022±8.0
Apoptotic cells (%)21±5.813±3.0
Necrotic cells (%)63±1265±11
  • Cells harvested using a three-step isolation procedure produced rod-shaped, apoptotic and necrotic cells observed by phase contrast light microscopy, 4 h after isolation. Cell counts were determined from photomicrographs using morphometric analysis. Mean±S.E; n, duplicate counts of five individual cell isolations.

Calcium tolerant cardiomyocytes were plated with several reagents in order to assess their attachment requirements. Atrial cells attached to uncoated substrates including glass coverslips survived although ventricular cells could survive in culture without attachment and remained viable for several days. Laminin provided the best overall cell attachment and was superior to other reagents examined, as indicated in Table 3. Laminin improved the plating efficiency in a dose dependent manner and was effective on plastic and glass substrates within 1 h of cell plating. The concentration of 10.0 μg/cm2 was used for the remaining experiments in this study.

View this table:
Table 3

Estimated cell attachment efficiency

Attachment factorsAtrialVentricular
Uncoated substrate+
Foetal bovine serum (4.0%)+
Gelatin (0.1%)+
Poly-l-lysine (50 μg/ml)++
Extracellular matrix (10 μg/ml)+
Matrigel (10 μg/cm2)
Laminin (10 μg/cm2)++++++
Cell Tak™ (5.0 μg/cm2)++++
  • Symbols represent numbers of cells attached to each substrate as seen using phase contrast microscopy. Microscope magnification 150×. −, no cells attached; +, <5; ++, 5–20; +++, >20 cells.

3.2 Expression of MLC-2 isoforms

RT-PCR on RNA from freshly isolated myocytes revealed the expression of the ventricular myosin light chain-2 (MLC-2v) isoform in ventricular tissue only. In contrast, atrial MLC-2 (MLC-2a) was detected in both atrial and ventricular tissues (Fig. 3). The expression was similar to tissue from human foetal ventricular myocardium (data not shown). The signal for ANF was detected in both atrial and ventricular tissue, which underscores the hypertrophy of the ventricular tissue. Potentially this also contributes to the difference in the SERCA expression level, which was much higher in the ventricular sample compared with the atrium.

3.3 Myofibrillar organization

Viable adult cardiac myocytes cultured on plastic substrates exhibited a diminished or complete loss of sarcomeric structure by day 1, although widespread amorphous staining of sarcomeric markers including α-actinin, MLC-2 and tropomyosin, was evident. Cell edges became rounded and remaining sarcomeres appeared disorganised. All cells cultured with low serum assumed a flattened, spread morphology with large aggregates and intracellular vacuoles in the peri-nuclear region as shown in Fig. 4A. Aggregates appeared to be extruded from the cell and stained positive for α-actinin, whilst the majority of intracellular vacuoles were not stained (data not shown). In contrast, the majority of cells cultured without serum did not produce large extracellular aggregates and retained a rod-like morphology with few cells appearing flat after 7 days in culture (Fig. 4B). Sarcomeric structure was disorganized or absent, although z-line structures (α-actinin) persisted during the culture period. Intracellular vacuoles did not stain positive for sarcomeric proteins (Fig. 5A). Serum deprived cells had punctate areas of protein that stained for α-actinin and MLC-2 at the cell edges of the rod-shaped cells and cells at an early stage of transition to a flatter morphology (Fig. 5B). Populations of spherical cells were identified in cultures without serum. These remained attached to the substrate and stained positive for amorphous α-actinin and MCL-2v and had some sarcomeric structure (Fig. 5C).

Fig. 5

Myofibrillar breakdown in cells without serum. (A) Cells cultured without serum retained a rod-shaped appearance with intracellular vacuoles and discrete positive staining for α-actinin and ventricular regulatory myosin light chain (MLC-2v), within the cytosol at day 5. Nuclei stained blue with the Hoechst nucleic acid stain (magnification 40×). (B) Cells with a spread appearance had punctate localization of α-actinin and MLC-2v within the cytosol and at the cell edges (magnification 40×). (C) Spherical myocytes with intact nucleus cultured without serum, stained positive for α-actinin and MLC-2v with a diffuse, but specific, staining pattern (magnification 40×).

Fig. 4

Morphology of cultured cardiac myocytes with or without serum. (A) Cells cultured with 2.0% serum formed flattened, spread cells with large protein aggregates on the cell surface (arrows) and perinuclear vacuoles (magnification 40×). (B) Cells cultured without serum retained an elongated or rod-like appearance with disorganized sarcomeric structure at day 7 (magnification 40×). Nuclei stained blue with Hoechst nucleic acid stain.

3.4 Myofibrillogenesis in adult cardiac myocytes

At day 16, ventricular myocytes cultured with low serum adopted two broad but slightly distinct flat cell morphologies. The first type (type I) was characterized by a flattened spread morphology and extended lamellapodia (Fig. 6A). These cells stained positive for α-actinin and consisted of beaded stress-like fibers or nascent myofibrils organized in parallel arrays. The second cell type (type II, Fig. 6B) appeared refractive with phase contrast microscopy, had well formed sarcomeres radiating in a uniform direction and converging at the cell edges.

Fig. 6

Differentiation of ventricular cardiac myocytes cultured with low serum. Cardiac myocytes cultured with low serum had two distinct morphologies at day 16. (A) Type I cells appear flat and spread with straight cell edges (magnification 40×). Type I cells stained positive for α-actinin (upper middle panel) and consisted of beaded stress-like fibres (right upper panel, higher magnification of indicated area). (B) Type II cells have sharp edges and visible cytoskeletal fibres (magnification 40×). Type II cells were polygonal shaped, with vacuoles and distinct myofibrils with broad z-disks, radiating to focal points at the cell edges (lower middle panel). Myofibrils stained positive for α-actinin at developing z-lines (right lower panel, higher magnification of indicated area).

To evaluate whether intercellular contact promoted type II morphology, populations of cultured cardiac myocytes (day 16) were examined with immunocytochemistry and phase contrast light microscopy. A total of 14 cell clusters consisting of two to four cells and 18 individual cells were examined and the morphological type determined. Some 50% of individual cells and 70% of cell clusters exhibited type II morphology. Fig. 7A shows an individual cell, well separated from its neighbors, exhibiting type II morphology. Prior to fixation, this cell spontaneously contracted at the perinuclear region and produced large contractile waves that propagated to the cell edges. As expected, individual separated cells had no cadherin staining at the cell edges (Fig. 7A). Fig. 7B shows a cluster of four, tightly associated myocytes; these clusters do express cadherin indicating that the cells have formed tight junctions. In addition, mixed clusters with type I and II cells were observed.

Fig. 7

Morphology of individual cells and cell clusters. (A) Individual type II cell with well defined myofibrils stained positive for α-actinin and negative for pan-cadherin at cell edges. (B) Myocyte positive for pan-cadherin (magnification 63×), and co-staining with phalloidin Cy3.

4 Discussion

4.1 Isolation of adult human cardiac myocytes

The enzymatic isolation of adult human cardiac biopsies produced viable rod-shaped atrial and ventricular cardiomyocytes in addition to apoptotic and necrotic cells as shown in Fig. 1 and Table 2 [16]. Sequestration of DNA and myosin distinguished cells undergoing programmed cell death from necrotic cells which appeared fragmented, as shown in Fig. 2. Cells were cultured with low serum (2%) or without serum in order to prevent proliferation of cardiac fibroblasts and other non-myocyte cells, eliminating the need for anti-mitotic agents such as cytosine-β-d-arabinofuranoside. The yield of ∼20% viable cells was an improvement over values of 1–5% reported by Peeters et al. [12]. Due to the kind of biopsies we received, chunk collagenase treatment was our only option, although perfusion of isolated vessels is the preferred method. However, studies using retrograde perfusion to isolate myocytes from rabbit and rat heart reported greater than 95% recoveries of viable cells [14]. A different approach has been followed by Forini et al. using small chunks of tissue in culture [17]. An alternative approach is to alter the cell dissociation concoction. Other enzymes were used successfully to isolate myocytes from human tissues, including protease and elastase [18]. No additional measures were tested to remove cations from the dissociation solution, although the inhibiting effect of calcium is well known [19]. All these issues highlight the need for further refinements to improve myocyte isolation from surgical biopsies.

4.2 Cell attachment

The extended exposure to the proteolytic enzymes during the isolation procedure disrupted the ability of cells to attach by damaging non-specific adhesion molecules including selectins or removing residual cell associated ECM that may adhere to the coverslip. In contrast, rat myocytes rapidly isolated using the Langendorf method attached to uncoated glass coverslips within seconds or minutes after plating [20]. Laminin provided the best overall attachment of human cardiomyocytes compared with other agents tested (Table 3), suggesting that β-integrins were either resistant to the effects of the protease or were rapidly re-expressed on the cell surface during the initial culture period. Studies in rats indicated that a substrate laminin concentration between 0.5 and 12.5 μg/ml was effective for myocyte attachment [21,22]. In the present study, a laminin concentration of 10 μg/cm2 effectively promoted cell attachment [23], whilst collagen containing products including commercial gelatin, extracellular matrix, and matrigel, failed to promote cell attachment. Lundgren et al. [21] also observed that rat adult cardiomyocyte attachment was markedly reduced in collagen type I treated dishes while in contrast, collagen type IV binding was identical to laminin. This study also demonstrated that laminin remained associated with isolated myocytes and that collagens were not detected by immunocytochemistry following isolation. Taken together, these studies indicate that laminin is critical for adult myocyte attachment and raises the possibility that outside-in signal transduction through the laminin–β-integrin complex may have a role in determining cardiac cell function in vivo. Cell Tak™, used as a tissue adhesive agent, anchored myocytes to the glass coverslip. This facilitated immunocytochemical studies on non-adherent and apoptotic cells. However, this treatment appeared to interfere with active cell attachment since cells detached after several days in culture.

4.3 Cardiomyocyte transformation in culture

The transformation of isolated adult cardiomyocytes from a rod-shaped morphology to a flattened, spread morphology occurred as they adapted to the cell culture substrate (Fig. 4A). Low serum promoted rapid cell spreading, loss of sarcomeric structure and the extrusion of sarcomeric proteins within 1 week, described previously in other species, and generally regarded as dedifferentiation [13,14,24,25]. Human cells cultured without serum retained a rod-like morphology, rounded edges, large vacuoles and discrete areas occupied by sarcomeric structures as shown in Fig. 4B and demonstrated in rabbit cardiomyocytes [14]. In addition, we identified spherical cells that resembled chick embryonic cardiac myocytes. Spherical cells were attached to the substrate and exhibited a smooth sarcolemma and stained positive for amorphous sarcomeric proteins. These cells retained some sarcomeric structure with an intact nucleus, but never recovered to become functional rod-shaped myocytes (Fig. 5C). Whether this state of the cells resembles cardiomyocyte ‘hibernation’ [26,27] requires further investigation.

4.4 Myofibrillogenesis in adult cardiac myocytes

Given the diverse growth promoting properties of serum, we used an intermediate concentration of 2% rather than the 4% used previously [13,25]. This treatment resulted in cardiomyocytes with two distinct cell morphologies. Type I and type II cells differed from one another in overall appearance and the extent of myofibrillar development.

This difference in morphology may reflect the cell response to the proliferative effects of serum (hypertrophy), balanced by the propensity to form stable focal adhesions with the substrate. The α-actinin stress-like fibres shown in Fig. 6B of type I cells were similar to those described in differentiated adult rat cardiomyocytes. α-Actinin provided a scaffold for sarcomeric proteins and preceded the recruitment of both α-actin and myosin during the formation of nascent myofibrils [5,6].

Given that type II cells were either physically attached to adjacent myocytes or existed as single cells with no intercellular attachments (Fig. 7A and B), it was concluded that cell–cell attachment promoted, but was not essential for, myofibrillar development. In the present study, only type II myocytes regained the contractile phenotype after 9 days in culture. This was characterized by large contractile waves originating at the peri-nuclear region, similar to those reported for mouse and rat adult cardiomyocytes [23]. Intercellular adhesions and focal adhesion involving integrin receptors play a critical role during the formation and polarity of the developing myofibrils, by providing a bipolar structure upon which myofibrils may align. The relative importance of each was not determined here, however, previous studies have shown that adult feline cardiomyocytes plated at high density changed from a randomly dispersed culture of beating myocytes into a parallel array of myofibrils as cells formed intercellular junctions [28]. In addition, cell adhesion facilitated and perpetuated differentiation of skeletal myoblasts by promoting the upregulation of dominant skeletal muscle genes including MyoD and myogenin in cultured skeletal myoblasts [29,30].

The expression of MLC-2a RNA in myocytes from both the atrial and ventricular tissues, shown in Fig. 3, was expected on the basis of the underlying pathology including left ventricular hypertrophy [31]. To what extent this altered gene expression influenced cell survival or exacerbated the changes we observed during the cell culture experiments is not known, however, the unavailability of specimens from normal hearts has so far precluded further quantification. We believe that the human model of cultured cardiac myocytes developed in this study may provide important information about the differentiation and maintenance of cardiac myocytes.

5 Conclusion

Cultured adult cardiomyocytes provide a convenient and complementary in vitro system with which cardiac pathology and human cardiogenesis can be investigated, particularly myofibrillar development, excitation–contraction coupling and apoptosis. Isolated human cardiomyocytes could help to gain insight into cardiomyopathies, particularly those associated with inter-related disorders of ion channel function, contractility and myofibrillogenesis. In addition, this system will enhance our understanding of the mechanisms underlying cardiac myocyte differentiation, benefiting efforts to direct and promote stem cell differentiation into cardiac myocytes suitable for transplantation into an adult heart. Here, the ability to communicate with host cardiac cells and respond appropriately in the working myocardium will be of the utmost importance. Co-culture of the adult myocytes, described here, with human embryonic stem cell derived cardiomyocytes [32] could be used as an in vitro transplantation model to understand how these cells might interact.


This research was funded by Embryonic Stem Cell International (SDB). The antibodies for MLC-2a and MLC-2v were a kind gift from Dr K. Chien, University of California School of Medicine at San Diego, La Jolla, CA, USA. The authors wish to thank J. Korving for his assistance with the preparation of cryosections, Rene Spijker for technical assistance and Dr B. Defize for discussion and help with microscopy.


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
  24. [24]
  25. [25]
  26. [26]
  27. [27]
  28. [28]
  29. [29]
  30. [30]
  31. [31]
  32. [32]
View Abstract