Myosin Heavy Chain Expression in Contracting Myocytes Isolated During Embryonic Stem Cell Cardiogenesis
Abstract Mouse embryonic stem (ES) cells are totipotent cells derived from the inner cell mass of the preimplantation blastocyst and are capable of differentiating in vitro into cardiac myocytes. Attached cultures of differentiating ES cells were established to document the timing of contractile development by microscopic observation and to permit the microdissection of cardiac myocytes from culture. The onset of spontaneous contraction varied markedly in differentiation culture, with contraction being maintained on average for 9 days (range, 1 to 75 days). Indirect immunofluorescence microscopy showed that myosin expression was localized to the contracting cardiac myocytes in culture. Myosin heavy chain (MHC) isoform expression in microdissected ES cell–derived cardiac myocytes was determined by means of sodium dodecyl sulfate–polyacrylamide gel electrophoresis. The distribution of MHC isoform expression in isolated ES cell cardiac myocytes was as follows: 27% expressed the β-MHC isoform, 33% expressed both the α- and β-MHC isoforms, and 40% expressed the α-MHC isoform. MHC phenotype was correlated to the duration of continuous contractile activity of the myocytes. Myocytes that had just initiated spontaneous contractile activity predominantly expressed the β-MHC (average days of contraction before isolation, 2.5±0.7). The α-MHC isoform was detected after more prolonged contractile activity in vitro (1 to 5 weeks). A strong correlation was obtained between MHC phenotype and days of contraction of the cardiac myocyte preparations isolated from ES cell cultures (r=.93). The apparent transition in MHC isoform expression during ES cell differentiation parallels the β- to α-MHC isoform transition characteristic of murine cardiac development in vivo. These findings are evidence that ES cell cardiac myocyte differentiation follows the normal developmental program of murine cardiogenesis.
Mouse embryonic stem (ES) cells have recently been established as an important tool for analysis of murine development. ES cells are derived from the inner cell mass of the preimplantation blastocyst and represent a continuous cell line in that these cells can proliferate in an undifferentiated, pluripotential state when grown on a feeder layer of mitotically quiescent mouse embryonic fibroblasts (MEF) and in the presence of leukemia inhibitory factor (LIF).1 2 3 The pluripotential nature of ES cells is demonstrated in studies in which ES cells are genetically manipulated, microinjected into a host blastocyst, and implanted into pseudopregnant mice.2 4 5 The foreign ES cells can contribute extensively to the developing embryo and have been demonstrated to colonize somatic cells as well as the germ line, thus enabling the establishment of a transgenic line of mice.2 4 5
A second notable feature of ES cells is their capacity to differentiate in vitro upon removal from MEF. Various differentiated cell types have been observed during the in vitro differentiation of ES cells, including cardiac myocytes, neuronal tissue, embryonic erythrocytes, melanocytes, and others.1 2 3 There is growing evidence that the in vitro differentiation of ES cells may provide a useful model system to determine cellular and molecular mechanisms of cardiac development. Doetschman and colleagues1 showed that embryoid bodies, which are aggregates derived from suspension cultures of differentiating ES cells, develop arch-shaped rhythmically contracting ridges in vitro, a phenotype characteristic of developing myocardium. Furthermore, electron micrographs of the contracting ridges revealed myofilaments assembled into sarcomeres and intercalated discs, a structural marker of the cardiac phenotype.
Recent cellular and molecular biological studies have demonstrated cardiac gene expression in differentiating embryoid bodies. For example, phospholamban, a protein specific to cardiac and slow skeletal muscle sarcoplasmic reticulum, was detected in embryoid bodies at the mRNA and protein level.6 The appearance of phospholamban coincided with the initiation of spontaneous contractile activity in the embryoid bodies in vitro. Cardiac troponin C and cardiac troponin T gene expression have also been detected in embryoid bodies7 as well as the cardiac markers myosin light chain (MLC) 2V and atrial natriuretic factor (ANF).8 Both MLC 2V and ANF are expressed in ventricular myocytes in vivo, with the MLC 2V gene thought to be a specific marker of the ventricular cardiac myocyte. These findings provide evidence that the ES cell differentiation culture system may provide an in vitro model system to determine the functional significance of altered cardiac gene expression during cardiogenesis.
Cardiac myosin heavy chain (MHC) gene expression has been detected in differentiation cultures of ES cells. Myosin is a large hexameric protein (≈500 000 D) consisting of two intertwining polypeptides, each with a molecular weight of ≈200 000 D, termed heavy chains, and four smaller peptides with molecular weights varying from ≈16 000 to 26 000 D, termed light chains.9 The globular head (S1) of the MHC is notable in that it contains the actin binding site and is the site of ATP hydrolysis. Myosin S1 propels actin cables in in vitro motility assays, thus identifying myosin S1 as a molecular motor.10 Three MHC isoforms have been identified in the mammalian heart: V1, a homodimer of the α-MHC isoform; V2, a heterodimer of the α- and β-MHCs; and V3, a homodimer of the β-MHC.11 12 The phenotype of the cardiac myosin isoform expressed in vivo is controlled by numerous factors including age, stress, and hormonal status of the animal.11 12 13 14
In ventricular myocytes of small mammals there is a transition in MHC phenotype during development. Prenatally the β isoform predominates; however, after birth there is a progressive downregulation of the β-MHC gene and upregulation of the α-MHC gene, so that in the adult ventricular myocyte the α-MHC isoform predominates.11 12 13 14 During the in vitro differentiation of ES cells, cardiac β-MHC mRNA was detected early in whole embryoid body development, a time point before the appearance of rhythmic contractions of myocytes within the embryoid bodies.15 16 At the onset of spontaneous contraction in embryoid bodies both β-MHC and α-MHC mRNA transcripts were observed.16 The apparent developmental transition in MHC transcripts from β to α in embryoid bodies paralleled the MHC mRNA transitions found during murine cardiogenesis in vivo.16 While this study clearly showed expression of the cardiac β- and α-MHC genes in ES cell differentiation cultures, it is less clear whether these results provide direct evidence of a developmental transition of MHC genes in vitro. It is possible, for example, that the detection of α-MHC gene expression is due to induction of atrial myocytes in embryoid bodies that constitutively express the α-MHC gene.
The uncertainty in the interpretation of the earlier results relates to difficulties associated with the inherent cellular and developmental complexity of embryoid bodies. For example, cardiac myocytes were not isolated from embryoid bodies, and therefore it is not known precisely how long specific cardiac myocyte foci were contracting in the embryoid bodies before the determination of MHC gene expression. This is important because embryoid bodies may have multiple foci of contracting myocytes, each with potentially different onset times of contraction and possibly each expressing a different subset of developmentally regulated genes. Analysis of whole embryoid bodies would therefore reflect an admixture of gene expression in all the myocytes.
In the present study we have overcome this experimental obstacle and now provide characterization of the MHC protein isoforms expressed in this differentiation culture system using attached cultures established by plating ES cell aggregates onto gelatin-coated glass coverslips. This format permitted the daily observation of the contracting cardiac myocytes during development in vitro. In addition, the attached differentiation culture format permitted microdissection of cardiac myocytes as they developed in vitro, so that MHC expression could then be determined in these isolated cardiac myocyte preparations by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE) and silver staining. Importantly, in this experimental approach both the phenotype of the MHC isoform expressed and the duration of contraction were determined in the same cardiac myocyte preparation. This made it possible to test whether a developmental transition in MHC protein expression is present in cardiac myocytes derived from the in vitro differentiation of mouse ES cells. We observed a correlation between MHC isoform expression and duration of contraction in the isolated ES cell–derived cardiac myocytes, a result that is consistent with the developmental pattern of MHC expression during murine cardiogenesis in vivo.
Materials and Methods
Mouse ES Cells
Mouse ES cell lines D3,1 CCE,5 and E1417 were compared for their potential to produce spontaneously beating cardiac myocytes. D3 cells were originally obtained from the inner cell mass of a mouse 129/Sv+/+ day 4 blastocyst. CCE cells were originally derived from a single XY blastocyst from the 129/Sv/Ev strain,5 and E14 cells were derived from day 4 blastocysts obtained from 129/Ola mice.17
MEF feeders were prepared by pretreatment with gamma irradiation (45 Gy) and plating at a density of 5×104 cells per square centimeter onto gelatin-coated tissue culture plates. ES cells were cultured on top of the monolayer of MEF feeders at a density of 1×105 cells per square centimeter. The culture medium was renewed daily, and the ES cells were subcultured onto new feeders every 2 to 3 days.
Culture medium for the undifferentiated ES cells consisted of DMEM with high glucose, supplemented with 15% fetal calf serum, 0.1 mmol/L β-mercaptoethanol, and 2% LIF-containing medium. LIF has been demonstrated to inhibit the differentiation of cultured ES cells.18 The LIF-containing medium was collected from cultures of Chinese hamster ovary cells that were transformed with an LIF expression plasmid (kindly provided by Genetics Institute, Cambridge, Mass).
Attached Cultures of Differentiating ES Cells
Cultures of differentiating ES cells were established by the formation of ES cell aggregates in hanging drop cultures, designated as day 1 of differentiation.19 20 The ES cells were dissociated from the MEF feeder layer by trypsinization and resuspended at a concentration of 1×104 cells per milliliter in differentiation media consisting of 20% fetal calf serum, 1% penicillin-streptomycin, and 0.1 mmol/L β-mercaptoethanol in DMEM. Drops (30 μL) of 300 cells were placed on the underside surface of a 100-mm tissue culture dish lid, and the lid was then carefully placed over PBS. After 2 days in differentiation culture, the embryoid bodies, having formed from aggregates of the ES cells in each hanging drop, were transferred to suspension culture in 100-mm bacterial dishes (Fisher Scientific) and were cultured an additional 3 days. Attached cultures of differentiating ES cells were initiated by plating the embryoid bodies onto gelatin-coated glass coverslips in six-well tissue culture dishes. Gelatin-coated coverslips (22 mm, Fisher) were prepared in advance and stored at 4°C until needed. Before the embryoid bodies were plated, the treated coverslips were placed into each well of the six-well tissue culture dishes and UV-irradiated for 5 minutes to ensure sterility. A single embryoid body was transferred onto a gelatin-coated coverslip with a small volume of culture medium and incubated at 37°C for 4 to 5 hours to allow attachment to the center of the coverslip before additional differentiation medium was added to the well.
Cultures were observed daily with the use of an inverted light microscope (magnification, ×25 to ×100) to document the contraction initiation day and total duration of continuous contractile activity for each myocyte focus that developed in the different regions of the coverslip. The growth medium for the attached differentiation cultures was changed three times per week.
Microdissection of Cardiac Myocytes From Differentiation Culture
Glass micropipettes (OD, 1.0 mm; ID, 0.58 mm) were pulled to yield tip diameters of ≈1 μm and used as a dissecting tool. Differentiation cultures were observed via light microscopy (×50 to ×100) to aid in the dissection of the spontaneously contracting cardiac myocytes from differentiation culture. Cardiac myocytes that were electrically coupled and therefore contracting synchronously in differentiation culture (eg, Fig 1D⇓ and 1E⇓) were dissected from culture by means of the micropipettes. It was essential that synchronously contracting cardiac myocytes were isolated separately from other neighboring contracting myocytes because the differentiation cultures frequently contained multiple, distinct, contracting cardiac myocytes that had different days of onset and durations of spontaneous contractile activity (Fig 2⇓).
Adult and Fetal Ventricular Cardiac Myocyte Preparations
Whole hearts were removed from female mice and rats, from which ventricles were dissected from the atria and minced in relaxing solution (see below). The minced ventricular tissue was placed in a Waring blender and homogenized for ≈6 seconds at low speed.21 The resulting cell suspension was then centrifuged at 120g for 1 minute, and the resulting pellet was resuspended in 10 mL relaxing solution. The centrifugation and resuspension steps were completed one more time.
Determination of MHC Phenotype by SDS-PAGE and Silver Staining
Isolated cardiac myocyte preparations were placed in a 0.5-mL microfuge tube containing SDS sample buffer (10 μL/mm of preparation length) and stored at −80°C for subsequent analysis of the MHC isoform expressed by SDS-PAGE and silver staining as described previously.22 The average dimensions of the isolated ES cell–derived cardiac preparations were ≈500 μm in length and 75 μm in width. Each isolated preparation contained multiple, electrically coupled cardiac myocytes (the arrows and arrowheads in Fig 1D⇑ and 1E⇑, respectively, each denote foci of synchronously contracting myocytes before their isolation from culture).
The gel electrophoresis procedure used a multiphasic buffer system that incorporated the following features: (1) acrylamide–piperazine diacrylamide ratio of 200:1, (2) pH of 9.3 in the running gel buffer, and (3) molarity of the running gel buffer of 0.75 mol/L. The acrylamide in the stacking gel was 3.5% and was 9% to 12% in the running gel. The gel electrophoresis conditions were similar to those used previously,22 although piperazine diacrylamide replaced Bis acrylamide as the cross-linker in the present study, which appeared to improve resolution of the MHC bands. Gels were fixed with glutaraldehyde overnight, washed several days in distilled water, silver stained, and then dried between mylar and cellophane sheets. Gels were scanned with the use of an LKB densitometer, and the areas under the peaks were integrated. The relative expression of each cardiac MHC isoform was calculated as the ratio of the area corresponding to the α-MHC or β-MHC peak to the sum of the areas corresponding to the α- and β-MHC peaks.
Localization of myosin during the in vitro differentiation of ES cells was determined by indirect immunofluorescent microscopy. Sketches indicating the specific location of spontaneously contracting cardiac myocytes in differentiation culture were made before removal of the glass coverslips from the six-well plates. This permitted the direct matching of the location of spontaneously contracting myocytes and the location of myosin fluorescence in differentiation culture. The glass coverslips containing the attached differentiation cultures were washed several times with PBS and placed into histology jars containing 2% paraformaldehyde in PBS for 2 hours at room temperature. After several washes with PBS and treatment with ammonium chloride (50 mmol/L, 30 minutes), blocking was accomplished with the use of 20% normal goat serum in PBS containing 0.5% Triton X-100. Myosin was detected by incubating with an anti-myosin antibody (Sigma M 7648) diluted 1:100 in PBS containing 2% normal goat serum with 0.5% Triton X-100 in a humidity chamber saturated with PBS at 4°C overnight. The next day the coverslips were washed several times with PBS and blocked with PBS containing 2% normal goat serum plus 0.5% Triton X-100 for 0.5 hours, followed by application of the secondary antibody (fluorescein isothiocyanate isomer I–conjugated goat anti-rabbit antibody) diluted 1:200 for 1.5 hours at room temperature. After several washes with PBS, the coverslips were mounted on microscope slides and stored at −80°C. Fluorescence was detected with the use of a Leitz Aristoplan fluorescence microscope. Similar methods were used for immunofluorescence localization of myosin in isolated adult cardiac myocytes and in slow and fast single skeletal muscle fibers (data not shown). In this case, the isolated cells were attached to coverslips by applying a thin film of silicone adhesive to the coverslip.
Establishment of Attached Cultures of Differentiating ES Cells
Differentiation culture procedures were established to permit the controlled in vitro differentiation of ES cells (described in “Materials and Methods”). Our procedure results in attached cultures of differentiating ES cells. This format permits visual inspection to determine the time of onset and duration of spontaneous contractile activity in these cultures. Photomicrographs obtained at various time points during the establishment of attached cultures of differentiating ES cells are shown in Fig 1⇑. A representative embryoid body at 5 days after initiation of differentiation, and just before plating onto the gelatin-coated coverslip, is shown in Fig 1A⇑. The attached culture was established by plating a single, day 5 embryoid body onto a gelatin-coated coverslip (Fig 1B⇑; 10 hours after plating) and allowing continued cellular proliferation and differentiation (Fig 1C⇑; day 8). In this differentiation culture, the initiation of spontaneous contractile activity occurred on day 11, and Fig 1D⇑ indicates (demarcated by arrows) the formation of a functional syncytium of linearly aligned, spontaneously contracting cardiac myocytes (picture taken on day 14). The size and shape of the contracting foci of myocytes in each culture were variable. In some differentiation cultures, foci of linearly aligned, spontaneously contracting cardiac myocytes were observed spanning the ridges of cysticlike structures (Fig 1E⇑; synchronously contracting cardiac myocytes demarcated by arrowheads). In ≈50% of the differentiation cultures studied, complex structures of branching/anastomosing spontaneously contracting cardiac myocytes were observed (Fig 1F⇑).
Characterization of Spontaneous Contractile Activity During In Vitro Differentiation of ES Cells
During the time course of this study, 490 attached differentiation cultures of ES D3 cells were initiated, of which 81% resulted in the development of spontaneously contracting cardiac myocytes. When the spontaneous contractile activity in differentiation cultures was plotted as a function of the time (days) after initiation of differentiation, an approximate bell-shaped curve was obtained (Fig 2A⇑). Day 1 refers to the day of dissociation of ES cells from MEF and the initiation of differentiation by the formation of ES cell aggregates. The percentage of differentiation cultures containing contracting cardiac myocytes reached a maximal value of 45% to 50% on approximately day 25. The difference between the 81% and 45% to 50% values is explained by noting (1) the significant variation in the time of onset of contractile activity for individual cardiac myocytes and (2) differences in the duration of continuous contractile activity among the cardiac myocytes (Fig 2B⇑ and 2C⇑). In ≈7% of the differentiation cultures studied, the earliest spontaneous contractile activity occurred on day 9 to day 10 after removal of ES cells from MEF (Fig 2B⇑). This result is in general agreement with earlier reports in embryoid bodies.1 3 6 On average, contraction was initiated on day 18, with values ranging from day 9 to day 40.
The distribution of the duration (days) of continuous contractile activity in ES cell–derived cardiac myocytes is shown in Fig 2C⇑. After the initiation of spontaneous contractile activity, the cardiac myocytes contracted for a mean duration of 9 days, with values ranging from 1 day to the longest observed of 75 days.
There was also significant variation in the number of distinct foci of synchronously contracting cardiac myocytes observed in each differentiation culture (Fig 2D⇑). The mean number of synchronously contracting cardiac myocyte foci observed over the lifetime of the differentiation cultures was 2.4 (range, 1 to 9). It was relatively straightforward to visually identify these distinct contracting foci of cardiac myocytes because they frequently were well separated and characteristically had different rates of spontaneous contraction. Foci of cardiac myocytes were defined by myocytes that were electrically coupled; that is, they contracted synchronously (examples are denoted by arrows in Fig 1D⇑ and arrowheads in Fig 1E⇑).
In summary, attached cultures of differentiating ES cells demonstrated marked developmental heterogeneity in the time of onset and duration of contractile activity among the nascent cardiac myocytes developing in vitro.
Influence of Passage Number and Various ES Cell Lines on Cardiac Muscle Formation In Vitro
The passage number of the undifferentiated ES D3 cultures appeared to affect the subsequent development of cardiac myocytes in culture. Optimal results were obtained from ES D3 cultures at approximate passage numbers 20 to 40. Furthermore, it was observed that ES D3 sublines that are not capable of germline chimera formation after microinjection into blastocysts have a greater capacity to produce contracting myocytes than ES D3 sublines that are capable of forming germline chimeras (data not shown). These observations suggest that ES cells that are somewhat restricted in their differentiation potential are best suited for spontaneous formation of cardiac myocytes during differentiation in vitro.
The mouse ES cell lines CCE and E14 (see “Materials and Methods”) were compared with ES D3 for their relative potential to produce spontaneously contracting cardiac myocytes in differentiation cultures in vitro (Table 1⇓). Spontaneously contracting cardiac myocytes derived from differentiation cultures of ES D3 cells resulted in a greater duration of continuous contractile activity and greater numbers of contracting cardiac myocytes compared with differentiation cultures that were established with the use of either CCE or E14. The percentages of differentiation cultures that developed spontaneously contracting cardiac myocytes varied among the three ES cell lines studied: D3, 81%; CCE, 71%; and E14, 58%. The day of onset of spontaneous contractile activity was not different among the three ES cell lines studied (Table 1⇓).
In general, differentiation cultures established with D3 cells routinely developed better defined, linearly aligned foci of contracting cardiac myocytes than the other ES cell lines. This is an important difference between the ES cell lines in terms of using the cardiac myocytes derived from culture for biophysical studies of contractility.23 Although caution must be used in drawing conclusions from these comparisons, in part because of the wide variation in the total number of differentiation cultures studied among the three ES cell lines, these results suggest that the ES D3 cells may be better suited for studies involving cardiac embryogenesis in vitro. Thus, for the remainder of this study ES D3 cells were used exclusively for the establishment of differentiation cultures.
Localization of Myosin Expression in ES Cell Differentiation Cultures
The distribution and localization of myosin during the in vitro differentiation of ES cells was determined by means of a polyclonal myosin antibody and indirect immunofluorescence microscopy. Fig 3⇓ displays fluorescent and Nomarski images obtained from a differentiation culture at day 14 of differentiation. In this differentiation culture all regions that demonstrated marked fluorescence were spontaneously contracting before fixation.
This differentiation culture also demonstrates the variation in morphology between different foci of synchronously contracting cardiac myocytes forming in vitro. In this culture there were three distinct foci of contracting myocytes: each was contracting separately from the others, indicating that the foci were not electrically coupled to each other (Fig 3⇑). The three contracting foci were as follows: (1) a large circular mass of synchronously contracting cardiac myocytes (marked “a”); (2) elongated, crescent-shaped, syncytium of synchronously contracting myocytes extending across the center of the differentiation culture (“b”), which upon isolation from culture has been shown to be suitable for biophysical studies of cardiac contractility23 ; and (3) two smaller, circular, syncytia of synchronously contracting myocytes (“c, d”). Each of these distinct contracting foci of myocytes had a different day of contraction onset and a different duration of continuous contractile activity, which were chronicled by daily microscopic observation of the culture. By microdissecting each of these foci from culture, it was possible to directly determine the relationship between MHC isoform expression and days of contraction in ES cell cardiac myocytes.
For comparison, Fig 3⇑ also shows a fluorescent image obtained from adult cardiac myocytes isolated from the ventricular chamber of a rat. The myosin antibody used in this study cross-reacted with myosin in skeletal fast and slow muscle fibers (data not shown) but did not cross-react with nonmuscle cells in differentiation culture (Fig 3⇑). Because the myosin antibody used was not cardiac specific, these immunofluorescence studies could only be used to show localization of myosin expression in differentiation culture without regard to the specific cardiac isoform expressed in a particular foci. For identification of the myosin isoform expressed in culture, subsequent studies were performed in which MHC isoform was determined by means of SDS-PAGE and silver staining (shown below). The results of these analysis identified the contracting myocytes in culture as cardiac myocytes.
Determination of MHC Isoform in Cardiac Myocytes Isolated From ES Cell Differentiation Cultures
The phenotype of the MHC isoform expressed in the ES cell–derived cardiac myocytes was determined by means of SDS-PAGE and silver staining. Cardiac myocytes that were electrically coupled and therefore were contracting synchronously were isolated from culture by microdissection (see “Materials and Methods”). This made it possible to determine the MHC phenotype of each of the distinct syncytia of contracting myocytes that were developing in different regions of the differentiation culture. It was important to isolate and study separately each preparation because of the marked variability in the time of onset and the duration of spontaneous contractility among the numerous, distinct, contracting cardiac myocytes formed in culture (Fig 2⇑).
The MHC isoform expressed in an ES cell–derived cardiac myocyte preparation was compared with that of an isolated adult cardiac myocyte expressing α-MHC, a fast skeletal fiber expressing fast MHC, and a slow skeletal muscle fiber expressing β-MHC (Fig 4⇓). The MHC isoform expressed in the cardiac myocyte preparation isolated from ES differentiation culture comigrated with the slow MHC isoform obtained from the soleus fiber. Because both the slow skeletal and cardiac β-MHC isoforms are derived from the same cardiac/slow β-MHC gene,13 this is evidence of expression of the cardiac β-MHC isoform in this ES cell–derived cardiac myocyte preparation. Results in Fig 5⇓ demonstrate the coexpression of the α- and β-MHC isoforms in another synchronously contracting cardiac myocyte preparation that was isolated from differentiation culture. Densitometric scans of the MHC bands indicated an approximate 1:1 stoichiometry of α/β-MHC protein expression in this ES cell–derived cardiac preparation. In some ES cell–derived cardiac myocyte preparations the α-MHC isoform was expressed exclusively (Fig 6⇓). This finding demonstrates cardiac-specific MHC isoform expression (eg, α-MHC isoform) in a spontaneously contracting cardiac myocyte preparation isolated from differentiation culture.
Overall, the relative distribution of MHC isoforms expressed in these ES cell–derived cardiac myocytes was as follows: 27% exclusively expressed the β-MHC isoform, 33% expressed both the β- and α-MHC isoforms, and 40% exclusively expressed the α-MHC isoform.
The MHC phenotype in the ES cell–derived cardiac preparations shown in Figs 4 through 6⇑⇑⇑, eg, β,β homodimer, α/β heterodimer, and α,α homodimer, suggests that the ES cell system may recapitulate the transition in MHC isoform expression obtained during murine cardiac development.15 16 24 To directly test this possibility, differentiation cultures were observed daily to document both the onset and duration of spontaneous contractile activity of each cardiac myocyte foci before microdissection. Thirty ES cell cardiac myocyte preparations with known contractile history were microdissected from culture, and MHC isoform expression was determined by SDS-PAGE and silver staining. The relationship between MHC isoform expression and days of contractile activity of individual ES cardiac myocyte preparations is summarized in Table 2⇓. Cardiac myocytes that had been contracting for a mean of 2.5 days in vitro predominantly expressed the β-MHC isoform. In ES cell–derived cardiac myocyte preparations that had been contracting for ≈1 week or more before isolation from culture, there was an increase in expression of the α-MHC isoform and a corresponding decrease in expression of the β-MHC isoform. In summary, a strong correlation was obtained between MHC phenotype and days of contraction of the cardiac myocyte preparations isolated from ES cell cultures (r=.93; Table 2⇓).
MHC Isoform Expression in Murine Ventricular Myocytes
For comparison of the results obtained from cardiac myocytes isolated from differentiation cultures of mouse ES cells, the MHC phenotype was determined in adult and fetal ventricular myocytes isolated from mice. In agreement with earlier findings obtained from mouse cardiac muscle at the transcript16 24 and protein25 26 levels, ventricular myocytes isolated from adult mice expressed predominantly the α-MHC isoform (Figs 5⇑ and 7⇓). In ventricular myocytes isolated from mice at day 18 to day 19 postcoitum, the β-MHC isoform was expressed predominantly, although in myocytes obtained at day 19 postcoitum expression of the α-MHC isoform was also evident, albeit at reduced levels compared with the β-MHC isoform (Fig 7⇓).
The ES cell differentiation culture system is noted for its marked developmental and cellular complexity. Previous studies on cardiac gene expression in this system have used embryoid bodies, which are differentiating ES cell aggregates growing in suspension culture. A significant limitation of using embryoid bodies to study cardiac gene expression and function relates to difficulties in establishing the time course of the onset and the duration of contractile activity of the cardiac myocytes during development in vitro. This is important because there are multiple foci of contracting cardiac myocytes in this system, each with potentially widely varying times of onset and duration of spontaneous contractile activity (Fig 2⇑), which cannot be monitored in embryoid bodies. This poses a significant experimental obstacle in using the embryoid body suspension culture format to study developmental regulation of cardiac gene expression.
To overcome the marked cellular and developmental complexity evident in the embryoid body system, the present study established attached cultures of differentiating ES cells. The attached differentiation culture format offers several important advantages over the embryoid body suspension culture format. First, the onset and duration of contractile activity of each cardiac myocyte can be determined directly by daily visual inspection of the differentiation cultures by means of conventional microscopy. Thus, the precise number of days of contractile activity (eg, day of onset, duration; Fig 2⇑) can be determined directly for each of the synchronously contracting cardiac myocyte foci rather than using the days in differentiation culture as an index of development. Second, the attached culture format makes it possible to microdissect the contracting myocytes from culture for determination of both cardiac contractile structure (Figs 4 through 6⇑⇑⇑) and function.23 The ability to isolate cardiac myocytes with known contractile history from differentiation cultures overcomes both the cellular and developmental difficulties associated with the ES cell differentiation culture system. Finally, the attached culture format significantly increases the viability of the contracting cardiac myocytes in differentiation culture. An individual embryoid body will beat for only ≈7 days.3 In contrast, in the attached culture format the average number of days of continuous contractile activity was >9 days, with the longest being 75 days in culture. In summary, the attached ES cell culture format overcomes previous limitations of this culture system by permitting (1) the direct determination of the time of onset and duration of contractile activity for each distinct synchronously contracting cardiac myocyte foci; (2) the microdissection of individual foci with documented contractile histories from differentiation culture; and (3) increased duration of spontaneous contractile activity.
In the present study attached cultures of differentiating ES cells were established to permit determination of the MHC isoform expressed in cardiac myocyte preparations that had been contracting for known time periods before their microdissection from culture. The MHC isoform expressed in the isolated cardiac myocytes was variable, ranging from the sole expression of the β-MHC isoform, to coexpression of the α and β isoforms, to exclusive expression of the α-MHC isoform. Results showed a strong correlation between the MHC phenotype and the days of contractile activity of the ES cell–derived cardiac myocytes before their isolation from culture (r=.93; Table 2⇑). The β-MHC isoform was expressed predominantly in cardiac preparations isolated after a mean of 2.5 days of spontaneous contractile activity in vitro. In preparations that were permitted to continue contractile activity for a week or more before isolation from culture, there was an increase in expression of the α-MHC isoform and a corresponding decrease in expression of the β-MHC isoform. The correlation observed between the days of contractile activity in vitro and MHC phenotype provides evidence that ES cell–derived cardiac myocytes recapitulate the programmed transition in MHC isoforms observed during murine cardiac development in vivo.16 24 25 26 27
The apparent β- to α-MHC isoform transition obtained during the in vitro differentiation of ES cells is qualitatively similar to that obtained during murine cardiac development. Alterations in MHC isoform expression occur at several different times during heart development. In embryonic heart development there is greater expression of the β-MHC gene compared with the α-MHC gene.27 Sanchez et al16 used a polymerase chain reaction assay to determine cardiac MHC gene expression during murine cardiogenesis at time points just before and after formation of the spontaneously contracting heart tube. At day 7.5 postcoitum, there was evidence of greater expression of the β-MHC gene than the α-MHC gene, whereas on day 9.5 postcoitum the α- and β-MHC transcripts were roughly equivalent. This finding is similar to that of Lyons et al,24 who used in situ hybridization to study the pattern of murine cardiac MHC gene expression. They found that MHC gene expression alters markedly upon cardiac chamber formation, which occurs at approximately day 9 postcoitum in the mouse.28 In general, from approximately day 9.5 postcoitum to approximately day 17.5 postcoitum, expression of the α-MHC transcript decreased in the ventricular chambers and expression of the β-MHC transcript decreased in the atrial chambers. After birth there was significant reexpression of the α-MHC gene and inactivation of the β-MHC gene in the ventricular chambers. Thus, the α-MHC isoform is expressed at high levels in both the atrial and ventricular chambers in adult mice, and high levels of β-MHC gene expression in the heart are normally seen only before birth.12 25 26
Robbins and colleagues3 15 16 have shown that during the early time points of ES cell differentiation in vitro, on days 3 to 4 after aggregation into embryoid bodies, there is expression of the β-MHC gene. In comparison, expression of the α-MHC gene had a delayed onset of activation, with expression apparent on approximately day 6 of differentiation.16 From day 6 through day 15 in differentiation culture there was approximately equal expression of cardiac α- and β-MHC genes in the embryoid bodies. Because gene expression was not determined in isolated myocytes from culture, it is not clear whether these results directly show a developmental transition in MHC gene expression in vitro. For example, an alterative interpretation of their findings is that detection of the α-MHC gene represents formation of atrial myocytes that constitutively express the α-MHC gene. However, our results provide evidence against this possibility. If atrial cardiac myocytes are forming in ES cell differentiation cultures, then it should be possible to detect expression of the α-MHC gene at the very onset of spontaneous contraction in these cells. This is apparent because the α-MHC gene is constitutively expressed in atrial myocytes.24 However, the earliest we detected α-MHC expression was 6 days after the initiation of contraction; at the onset of contraction only the β-MHC isoform was detected (Table 2⇑). This is evidence against the idea that atrial myocytes are forming in these cultures and alternatively supports the hypothesis of a developmental transition in MHC expression in the ES cell–derived cardiac myocytes.
We found that preparations that expressed the α-MHC isoform could be isolated from culture over a wide range of days of contraction, whereas preparations that predominantly expressed the β-MHC isoform could be obtained only if the preparation was isolated relatively soon after initiation of spontaneous contractile activity (Table 2⇑). The basis of the greater variation in the days of contraction for isolated cardiac myocytes that expressed the α-MHC isoform compared with myocytes that expressed exclusively the β-MHC isoform is not known. The comparatively small variation in days of contraction for those preparations that exclusively expressed the β-MHC isoform can be explained, however, if as described above, the activation of the β-MHC gene is initiated early during the in vitro differentiation of ES cells. If activation of the α-MHC gene is delayed with respect to activation of the β-MHC gene in vitro, and furthermore, if once activated the α-MHC gene remains activated for a prolonged period (eg, weeks), this could account for the range of days at which the α-MHC isoform was detected (Table 2⇑).
The functional significance of transitions in MHC gene expression during the very early stages of cardiogenesis in vivo and in vitro is not known. Because of the emerging complexity of the factors involved in determining cardiac mechanical properties, it cannot be assumed that transitions in cardiac MHC isoform expression will be solely responsible for alterations in myocardial function. A direct determination of the MHC structure-function relation during cardiac embryogenesis will require obtaining information on cardiac myocytes at the very early stages of cardiac myogenesis, where it has been demonstrated that there are complex alterations in cardiac gene expression.16 24 Progress toward elucidating the MHC structure-function relation during cardiac myogenesis has been hindered by the significant technical difficulties associated with determining contractile function at these early time points of cardiac development in vivo. The ES cell differentiation culture system may permit the determination of the MHC structure-function relation during the early stages of cardiac myogenesis. The apparent transition in cardiac MHC protein isoform expression during the in vitro differentiation of ES cells (Table 2⇑) is consistent with the hypothesis that this in vitro system recapitulates a developmental transition in MHC gene expression characteristic of murine cardiogenesis. In addition, methods are available to determine the contractile function of cardiac myocytes upon microdissection from ES cell differentiation culture.23 The ability to determine contractile gene structure and function in this in vitro differentiation system should permit new insight into the functional significance of alterations in cardiac gene expression during cardiac myogenesis.
In summary, attached cultures of differentiating ES cells were established to determine the phenotype of the MHC isoform expressed in synchronously contracting cardiac myocytes after microdissection from culture. A correlation was obtained between MHC phenotype and the days of continuous contractile activity of the isolated cardiac preparation. In preparations isolated from culture soon after initiation of spontaneous contractile activity the β-MHC isoform was expressed; however, with increased days of contraction detection of the β-MHC isoform decreased, with a corresponding increase in the α-MHC isoform. These results are in qualitative agreement with the transition in the MHC phenotype observed during murine cardiogenesis in vivo.16 24 This is taken as evidence that differentiating ES cells recapitulate cardiac myogenesis in vitro. The apparent transition in cardiac MHC isoform expression during the in vitro differentiation of ES cells, together with the ability to determine contractile properties of ES cell–derived cardiac myocytes,23 will make it possible to determine the functional significance of altered cardiac gene expression during cardiogenesis in vitro.
This study was supported by a Grant-in-Aid from the American Heart Association of Michigan, by the University of Michigan Office of Vice President for Research, and by a Rackham faculty grant. Dr Metzger is an Established Investigator of the American Heart Association. We thank Heather Burrows, Lisa Rankin, Dr Steve Ernst, and Noel Badyna for technical assistance, Dr Thomas Doetschman for the ES D3 cells, and Genetics Institute (Cambridge, Mass) for the gift of the Chinese hamster ovary/LIF cells.
- Received September 29, 1994.
- Accepted January 23, 1995.
- © 1995 American Heart Association, Inc.
Capecchi MR. Altering the genome by homologous recombination. Science. 1989;244:1288-1292.
Ganim JR, Luo W, Ponniah S, Grupp I, Kim HW, Ferguson DG, Kadambi V, Neumann JC, Doetschman T, Kranias EG. Mouse phospholamban gene expression during development in vivo and in vitro. Circ Res. 1992;71:1021-1030.
Pajek L, Muthuchamy M, Howles P, Doetschman T, Wieczorek DF. Analysis of contractile protein gene expression in developing embryoid bodies. J Cell Biochem. 1993;17D(suppl):202. Abstract.
Miller-Hance WC, LaCorbiere M, Fuller SJ, Evans SM, Lyons G, Schmidt C, Robbins J, Chien KR. In vitro chamber specification during embryonic stem cell cardiogenesis. J Biol Chem. 1993;268:25244-25252.
Hoh JFY, McGrath PA, Hale PT. Electrophoretic analysis of multiple forms of rat cardiac myosin: effects of hypophysectomy and thyroxine replacement. J Mol Cell Cardiol. 1977;10:1053-1076.
Mahdavi V, Izumo S, Nadal-Ginard B. Developmental and hormonal regulation of sarcomeric myosin heavy chain gene family. Circ Res. 1987;60:804-814.
Nadal-Ginard B, Mahdavi V. Molecular basis of cardiac performance. J Clin Invest. 1989;84:1693-1700.
Swynghedauw B. Developmental and functional adaptation of contractile proteins in cardiac and skeletal muscles. Physiol Rev. 1986;66:710-771.
Robbins J, Gulick J, Sanchez A, Howles P, Doetschman T. Mouse embryonic stem cells express the cardiac myosin heavy chain genes during development in vitro. J Biol Chem. 1990;265:11905-11909.
Sanchez A, Jones WK, Gulick J, Doetschman T, Robbins J. Myosin heavy chain gene expression in mouse embryoid bodies. J Biol Chem. 1991;266:22419-22426.
Handyside AH, O’Neill GT, Jones M, Hooper ML. Use of BRL-conditioned medium in combination with feeder layers to isolate a diploid embryonal stem cell line. Roux Arch Devel Biol. 1989;198:48-55.
Sweitzer NK, Moss RL. The effect of altered temperature on Ca2+-sensitive force in permeabilized myocardium and skeletal muscle. J Gen Physiol. 1990;96:1221-1245.
Metzger JM, Lin W, Samuelson LC. Transition in cardiac contractile sensitivity to calcium during the in vitro differentiation of mouse embryonic stem cells. J Cell Biology. 1994;126:701-711.
Lyons GE, Schiaffino S, Sassoon D, Barton P, Buckingham M. Developmental regulation of myosin gene expression in mouse cardiac muscle. J Cell Biol. 1990;111:2427-2436.
Ng WA, Grupp I, Subramaniam A, Robbins J. Cardiac myosin heavy chain mRNA expression and myocardial function in the mouse heart. Circ Res. 1991;69:1742-1750.
Sissman NJ. Developmental landmarks in cardiac morphogenesis: comparative chronology. Am J Cardiol. 1970;25:141-148.