Abstract

Embryonic stem (ES) cells offer unprecedented opportunities for in vitro drug discovery and safety assessment of compounds. Cardiomyocytes derived from ES cells enable development of predictive cardiotoxicity models to increase the safety of novel drugs. Heterogeneity of differentiated ES cells limits the development of reliable in vitro models for compound screening. We report an innovative and robust approach to isolate ES-derived cardiomyocytes using laser microdissection and pressure catapulting (LMPC). LMPC cells were readily applied onto 96-well format in vitro pharmacology assays. The expression of developmental and functional cardiac markers, Nkx 2.5, MLC2V, GATA-4, Connexin 43, Connexin 45, Serca-2a, cardiac alpha actin, and phospholamban, among others, was confirmed in LMPC ES-derived cardiomyocytes. Functional assays exhibited cardiac-like response to increased extracellular calcium (5.4 mM extracellular Ca2+) and L-type calcium channel antagonist (1 μM nifedipine). In conclusion, laser microdissection and pressure catapulting is a robust technology to isolate homogeneous ES-derived cell types from heterogeneous populations applicable to assay development.

In vitro assessment of cardiac safety has become one of the leading causes of compound attrition in the pharmaceutical industry. In a process that can cost upwards of 900 million dollars, and require over a decade of development, there remains an unmet need for physiologically relevant and clinically predictive in vitro assays (DiMasi et al., 2003). Cardiotoxicity is a major culprit for retraction of several commercial therapeutics and remains a critical concern in the regulatory agencies. Currently, recombinant proteins such as sarcolemmal ion channels, which encode for various currents associated with the cardiac action potential, are stably transfected into heterologous expression systems in an effort to isolate the ionic current of choice. Although this method is commonly used for the electrophysiological determination of compound potency, such as in the case of L-type calcium and the rapid inward rectifier potassium channel (hERG), measurement of these currents assumes the importance of a single gene target and can be confounded by the fact that native channels are often encoded by several genes. As compounds are pushed through developmental stages, toxicity in more complex systems such as the isolated cardiac myocyte or canine Purkinje fiber is invariably assessed to predict in vivo efficacy and safety in the native background. An alternative to in vitro ionic current assays as well as the labor-intensive isolation of adult primary tissue is the differentiation of embryonic stem cells into cardiomyocytes.

In vitro differentiation of embryonic stem cells into cardiomyocytes has shown that cardiogenesis evolves in a temporal manner from embryoid bodies (Klug et al., 1996). Transcription factors that induce cardiac development in vivo, such as Nkx 2.5 and GATA-4, as well as structural proteins that are critical for cardiac function were detected following in vitro development of embryoid bodies (Hescheler et al., 1997; Klug et al., 1996). Additionally, mouse and human ES-derived cardiomyocytes exhibit electric coupling and stimulus propagation that reflect in vivo action potentials of atrial, pacemaker, and ventricular cardiomyocytes (Boheler, 2003; Maltsev et al., 1994). This system therefore has widespread applications in cardiac development studies, as well as in vitro drug screening and potential therapeutic intervention in heart failure (Davila et al., 2004). Nonetheless, generation of enriched populations of ES-derived cardiomyocytes remains a challenge (Kehat et al., 2002; Kumar et al., 2005), and investigators have resorted to time-consuming manual dissection (Maltsev et al., 1994; Yuasa et al., 2005) or complex genetic manipulation of ES cells (Fijnvandraat et al., 2003; van Kempen et al., 2003). Overall success rates of directed differentiation of cardiomyocytes from wild type ES cells range from 3.9 to 30% (Fijnvandraat et al., 2003, Zandstra et al., 2003). Genetic modification of ES cell-derived cardiomyocytes increases homogeneity of differentiated cells, but may compromise their therapeutic application as opposed to wild-type cells.

One of the hallmarks of cardiac function is the process of excitation-contraction (EC) coupling, the mechanism by which an electrical impulse is converted into mechanical force. EC-coupling is generated by a depolarizing stimulus, which initiates the cardiac action potential. The summation of various membrane potential–sensitive ionic currents produces the action potential, allowing for the influx of calcium into the myocyte, leading to the release of intracellular calcium stores. The calcium transient provides the necessary fuel for cross-bridge cycling and precedes mechanical contraction. Multiple subpopulations of contractile foci have been characterized in previous studies (Zandstra et al., 2003) as well as nodal, atrial, and ventricular-like action potentials in ES-derived cardiomyocytes (Kehat et al., 2002). The ability to generate homogeneous populations of these cardiac cell types would provide a robust and replenishable substrate for in vitro assay development and cardiac safety screening.

Laser microdissection and pressure catapulting (LMPC) is a high-precision laser micromanipulation system that yields pure cellular preparations from live or fixed tissue (Schütze, 2003). This technology is operated via a pulsed UV-A laser that excises targeted cells, followed by propulsion and capture of excised material, a phenomenon also known as “cold ablation” (Vogel and Venugopalan, 2003). Moreover, extensive analysis at mRNA and protein levels have shown that LMPC offers high resolution of sample content and lack of contamination with adjacent cells (Espina et al., 2005). The majority of LMPC applications have been focused on mRNA isolation and gene expression analysis in target cells from fixed tissue preparations (Chung et al., 2005). A single report was issued on the opportunity to use laser-mediated microdissection to subculture live cell lines (Stich et al., 2003). There are currently no published studies on the use of laser microdissection as a means to address one of the largest limitations in stem cell technology, i.e., the heterogeneity of differentiated cell types that arise in both embryonic and adult stem cells (Beltrami et al., 2003; Kehat et al., 2002; Thomson et al., 1998).

In the present study we validated LMPC as a novel approach to isolate differentiated cells from ES cells and obtained a homogenous cardiomyocyte substrate that was applied to in vitro cardiotoxicity assays. ES-derived cardiomyocytes enriched by LMPC expressed critical cardiac-specific markers both at the mRNA and protein levels. Most importantly, LMPC ES-derived cardiomyocytes were subject to functional assays for intracellular calcium release and L-type calcium channel activity in the presence of chemical modulators, i.e., functional pharmacology. This is the first report, to our knowledge, where LMPC was established as a means to purify ES-derived differentiated cell types. In summary, this approach yielded a novel in vitro screening tool based on the application of ES-derived cardiomyocytes for predictive cardiac safety.

MATERIALS AND METHODS

ES cell differentiation and laser microdissection and pressure catapulting (LMPC).

Mouse (DBA/1LacJ, Roach et al., 1995) embryonic stem cells were seeded onto 150-mm bacterial-grade plastic culture dishes at 9 × 106 cells in 30 ml of Knockout™ D-MEM (Invitrogen), supplemented with 15% heat-inactivated fetal calf serum 200 mM L-glutamine, 10 mM MEM Non-Essential Amino Acids Solution (Invitrogen), 1000 units/ml LIF (Chemicon #ESG1107) 55 μM 2-mercaptoethanol (Sigma #M7522), and 10 mg/ml gentamicin (Invitrogen). Embryoid bodies were generated following suspension culture of mouse DBA ES cells. On days two to seven of in vitro culture, embryoid bodies were exposed to 1 mM ascorbic acid in the culture media described above in the absence of LIF. Differentiation was subject to dynamic culture (shaker 50–70 rpm).

On day seven, cardiomyocyte-specific media containing 100 mM of norepinephrine (Claycombs Complete, JRH Biosciences # 51800) was added to suspension cultures. Embryoid bodies were maintained in dynamic suspension culture until day ten, when they were seeded onto PALM® DuplexDish double-membrane culture plates for LMPC, performed at day 12. Prior to microdissection, 90% of the media was removed, and the uncovered dish was placed in the appropriate holder of the RoboStage® (P.A.L.M. Microlaser Technologies AG). Actively contracting foci (Fig. 1) were circumscribed using the freehand tool of the P.A.L.M. (Positioning and Ablation with the Laser Microbeam) software. The Class III-A UV-A laser was set to the RoboLPC (robotic cutting and laser pressure catapulting) mode and activated. The circumscribed focus was cut along with the membrane layer of the DuplexDish and catapulted into the media-filled cap of a microcentrifuge tube positioned directly over the target area. Alternatively, when beating foci were located on the vicinity of PALM® DuplexDish borders, microdissected foci were pipetted manually in 40 μl of media. Beating foci were transferred into 96-well format for pharmacology assays or pooled (60–80 foci) for RNA extractions and gene expression analysis. Beating foci in 96-well format were dissociated into single cells for L-type calcium channel immunohistochemistry by incubation with 50 μl of trypsin EDTA 0.05% (Invitrogen) for 15 min at 37°C. Enzymatic dissociation was neutralized with 150 μl of serum-containing media.

FIG. 1.

Fluorescence immunostaining of ES-derived cardiomyocytes prior to LMPC and C2C12 skeletal myocytes: α-actinin, myosin, and Nkx 2.5 were detected in beating foci. S-myosin (skeletal myosin) was specific to C2C12 cells.

FIG. 1.

Fluorescence immunostaining of ES-derived cardiomyocytes prior to LMPC and C2C12 skeletal myocytes: α-actinin, myosin, and Nkx 2.5 were detected in beating foci. S-myosin (skeletal myosin) was specific to C2C12 cells.

Calcium fluorescence detection assay.

Individual laser microdissected beating foci were plated on gelatin-coated 96-well assay plates for intracellular calcium transient measurement. Stock solution of Fluo-3AM was prepared using Pluronic F-127 (20% w/v in dimethyl sulfoxide) (Molecular Probes, Eugene, OR). Assay wells were incubated (37°C) with 1-ml aliquot of Hank's Balanced Salt Solution containing Fluo-3AM at a final concentration of 4.4 μM for 30 min. Beating foci were studied on the stage of an inverted microscope (Nikon Diaphot); superfused at 1 to 2 ml/min with a modified Tyrode solution (150 mM NaCl, 5.4 mM KCl, 1.8 mM CaCl2, 1.2 mM MgCl, 10 mM glucose, 2 mM pyruvate, 5 mM HEPES, pH 7.4, 37°C). Fluo-3AM was excited at 480 nm with a xenon lamp, and the emitted light at 530 nm was recorded to represent the cytosolic Ca2+ transient ([Ca2+]). Intracellular calcium measurement was represented as a pseudo-ratio as previously described (Takahashi et al., 1999). Spontaneous calcium transients were acquired and stored on computer for analysis using Ionwizard software (Ionoptix, Milton, MA). Indicated parameters were measured on three consecutive transients from each cell per treatment.

RT-PCR.

After washing cells with PBS, total RNA was immediately extracted using RNeasy minicolumns (Qiagen, CA) according to the manufacturer's recommendations. Residual amounts of contaminating DNA were removed by on-column RNase-free DNase treatment during the RNeasy procedure. RNA concentration was determined using a NanoDrop ND-1000 spectrophotometer (NanoDrop Technologies, DE) and RNA purity and integrity was assessed using an Agilent 2100 Bioanalyzer (Agilent Technologies, CA). No DNA contamination was detected in any of our total RNA preparations after qualitative assessment.

All RT-PCR assays were carried out using primer and probe sets from Applied Biosystems (ABI Assays on Demand, http://www.appliedbiosystems.com/). Each assay was designed using ABI's primer/probe selection algorithm and bioinformatics pipeline, which includes access to both public and Celera DNA sequence databases. All ABI Assays on Demand are designed to generate amplicons of 50–150 bp and are carried out using identical cycling conditions. Aliquots of total RNA were used for RT-PCR experiments according to the manufacturer's protocols (GeneAmp® RNA PCR kit and 9600 thermocycler, Applied Biosystems, CA). Specifically, 1–2 μg total RNA was used for each RT reaction, where total RNA was first converted to cDNA. Following removal of the reverse transcription mix, 20 μl/well of appropriate PCR mixes containing 10 μl of Taqman 2× Universal Master Mix, 1 μl of 20× Assay On Demand, or 20× GAPDH endogenous control primer/probe sets, and 9 μl H2O was added per well. PCR reactions were carried out according to the thermal profile: 10 min at 95°C, followed by 40 cycles of 15 s at 95°C and 1 min at 60°C. All RT-PCR assays for a particular gene were undertaken at the same time under identical conditions and carried out in duplicate. After 40 cycles, all RT-PCR products were resolved on 2% agarose gels.

TaqMan® low density array assay.

LMPC isolated foci (60 foci of 400–700 μm diameter each) were lysed with ABI Nucleic Acid Purification Solution. RNA was extracted from LMPC lysates using the ABI PRISM™ 6700 Automated Nucleic Acid Workstation. The GeneAmp® PCR System 9700 was used to generate cDNA. Quantitative RT-PCR (qRT-PCR) was performed, and levels of specific transcripts were obtained with combined ABI PRISM® 7900HT Sequence Detection System and ABI® Micro Fluidic cards technology. Data generated from qRT-PCR was quantified using ABI PRISM® SDS software, version 2.2. Gene expression levels were normalized to an endogenous control gene (HPRT), and cDNA from undifferentiated embryonic stem cells served as the calibrator. Two custom-designed ABI® Micro Fluidic cards were used in the study, with corresponding genes (Table 1).

TABLE 1

List of Genes Analyzed per ABI® Micro Fluidic Custom Designed Card


Osf2-pendi 

osteoblast specific factor 2 (fasciclin I-like) 
Gpc3 glypican 3 
Nodal nodal 
Myhca myosin heavy chain, cardiac muscle, adult 
Prnp prion protein 
Cpe carboxypeptidase E 
Hba-x hemoglobin X, alpha-like embryonic chain in Hba complex 
Pou5f1 POU domain, class 5, transcription factor 1 
Actc1 actin, alpha, cardiac 
Afp alpha fetoprotein 
Hprt hypoxanthine guanine phosphoribosyl transferase 
Kcnh2 potassium voltage-gated channel, subfamily H (eag-related), member 2 
Gata4 GATA binding protein 4 
App amyloid beta (A4) precursor protein 
Cacna1g calcium channel, voltage-dependent, T type, alpha 1G subunit 
Tnc tenascin C 
Trap1a tumor rejection antigen P1A 
Ryr2 ryanodine receptor 2, cardiac 
Atp2a2 ATPase, Ca++ transporting, cardiac muscle, slow twitch 2 
Calm1 calmodulin 1 
Col1a2 procollagen, type I, alpha 2 
Hprt
 
hypoxanthine guanine phosphoribosyl transferase
 

Osf2-pendi 

osteoblast specific factor 2 (fasciclin I-like) 
Gpc3 glypican 3 
Nodal nodal 
Myhca myosin heavy chain, cardiac muscle, adult 
Prnp prion protein 
Cpe carboxypeptidase E 
Hba-x hemoglobin X, alpha-like embryonic chain in Hba complex 
Pou5f1 POU domain, class 5, transcription factor 1 
Actc1 actin, alpha, cardiac 
Afp alpha fetoprotein 
Hprt hypoxanthine guanine phosphoribosyl transferase 
Kcnh2 potassium voltage-gated channel, subfamily H (eag-related), member 2 
Gata4 GATA binding protein 4 
App amyloid beta (A4) precursor protein 
Cacna1g calcium channel, voltage-dependent, T type, alpha 1G subunit 
Tnc tenascin C 
Trap1a tumor rejection antigen P1A 
Ryr2 ryanodine receptor 2, cardiac 
Atp2a2 ATPase, Ca++ transporting, cardiac muscle, slow twitch 2 
Calm1 calmodulin 1 
Col1a2 procollagen, type I, alpha 2 
Hprt
 
hypoxanthine guanine phosphoribosyl transferase
 
Immunohistochemistry of heterogeneous adherent ES-derived cultures.

Differentiating ES cells were plated in gelatin-coated 24-well plates. Each well was rinsed twice with 1 ml Ca2+/Mg2+-free PBS. Cells were fixed with 4% paraformaldehyde solution, 500 μl/well, and incubated at room temperature for 20 min, followed by two rinses with Ca2+/Mg2+-free PBS, 750 μl/well. To block and permeabilize cells, 0.4% Triton X-100 in PBS + 10% appropriate serum was added to each well. The serum was the same as the host species of secondary fluor-labeled antibody (for example, AlexaFluor goat anti-mouse IgG1, was blocked with goat serum). Samples were incubated for 1 h at room temperature on shaker. Primary antibody and appropriate controls diluted in BP were added to respective wells, 400 μl/well. For each primary antibody, an isotype control was included (i.e., for mouse anti-α-actinin, so MOPC 21 IgG1 was used) at the same concentration, and a blank well receiving BP only (for the secondary antibody control). Samples were incubated for 2 h at room temperature on shaker.

Plates were rinsed four times with 0.4% Triton X-100/PBS (TPBS), 750 μl/well, and the appropriate fluor-labeled secondary antibody (Molecular Probes AlexaFluors 488 or 594), was added, generally 1:500, in PB, 400 μl/well. Incubation was done for 1.5 h at room temperature in the dark (covered with foil) on the shaker. Following this procedure, plates were rinsed three times, 10 min each on shaker, with TPBS, 750 μl/well. The following antibodies were used to analyze beating adherent foci in heterogeneous cultures: α-anti-α-actinin (EA-53), mouse IgG1, recognizes sarcomeric α-actinin (Sigma Chemical Co.); anti-Nkx-2.5 (N-19), goat polyclonal IgG, recognizes amino terminus of Nkx-2.5 (Santa Cruz Biotechnology, Inc.); anti-myosin (MF20), mouse IgG2b, recognizes MHC, skeletal, and cardiac myosin (developed by Donald A. Fischman, M.D. and obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by The University of Iowa, Iowa City, IA); and anti-myosin (My-32), mouse IgG1, recognizes fast skeletal myosin, not cardiac.

Immunohistochemistry of LMPC beating foci.

LMPC beating foci were placed into 8-chambered slides and cultured overnight. Cells collected in the 8-chambered slides (Nunc International LabTech #177445) were fixed with 4% paraformaldehyde (EMS #15710) and 0.1% Tween-20 (Sigma #P-5927) for 30 min prior to quenching in 5% normal donkey serum (Jackson #017–000–121) and 0.1% Tween-20 in 1% ovalbumin (Sigma #A-5378). Two chambers per slide were exposed to primary antibody (anti-myosin heavy chain, Abcam #ab15; anti-cardiac actin, RDI #RDIPRO61075) while the remaining two chambers were exposed to the corresponding control IgG (either mouse IgG1, Caltag #MG100, or mouse IgG2a, Pharmagin #03151D) for 60 min diluted to 10 μg/ml in 0.1× quenching solution. After several quick rinses with 0.1× quenching solution, all chambers were incubated for 30 min with 1 μg/ml biotinylated donkey anti-mouse (Jackson #712–065–153). This was followed by several quick rinses with 0.1× quenching solution, and binding was detected by incubating all chambers for 30 min in a fluorescent cocktail containing 1 μg/ml streptavidin Texas Red (Jackson #016–070–084) and 1 μg/ml Hoechst 33342 (Molecular Probes #H3570) in 0.1× quench. The chambers were then removed, and the slide washed in PBS prior to coverslipping with 80% glycerol in PBS. Images were captured digitally with a SPOT camera attached to a Nikon E-800 fluorescent microscope.

RESULTS

LMPC Beating Foci Express Markers of Cardiac Differentiation and Function

The expression of mature and developmental cardiac genes was investigated in LMPC beating foci. Customized micro fluidics cards also contained noncardiac transcripts associated with early development. cDNA from adult mouse heart was used as control, and transcript levels were normalized against those of undifferentiated ES cells. Expression of definitive cardiac tissue markers (adult cardiac muscle myosin heavy chain, cardiac alpha actin, cardiac ryanodine receptor) was confirmed in LMPC ES-derived cardiomyocytes that composed beating foci and adult heart (Fig. 2). This finding was also true for other critical cardiac physiology genes (GATA-4, potassium voltage-gated channel, voltage dependent T-type calcium channel). Transcripts for cardiac-specific and cardiac-related genes were absent in undifferentiated ES cells. Interestingly, early developmental genes were upregulated in LMPC ES-derived cardiomyocytes in comparison to adult heart. The expression of early developmental markers in day 13 ES-derived cardiomyocytes, such as alpha fetoprotein, glypican 3, alpha-like embryonic chain Hba complex, nodal, POU domain class 5, and tumor rejection antigen, suggests that these cells were not fully mature at the molecular level, in comparison to adult heart. A second panel of cardiac-specific genes (Fig. 3) was assessed in LMPC beating foci by RT-PCR in comparison to the cardiac atrial myocyte cell line HL-5 and skeletal myocytes (C2C12). This panel focused on transcripts that are essential for cardiac function, as opposed to development and differentiation, namely electrophysiology (calcium, potassium, and sodium ion channels), calcium handling/contraction (SERCA2a, phospholamban), and conduction (Connexin 45, Connexin 43). All transcripts were expressed in LMPC beating foci, which is indicative of adequate programming at the molecular level to support heart physiology.

FIG. 2.

Quantitative gene expression analysis of ES-derived cardiomyocytes (dark bars) isolated with LMPC by ABI® Micro Fluidic cards in comparison to adult heart (light bar). Transcript levels were normalized against undifferentiated ES cells (x axis). Cardiac-specific markers (Cardiac actin, actc-1, GATA-4, myosin heavy chain, myh6, and ryanodine receptor, ryr2) were expressed in LMPC ES-derived cardiomyocytes as well as early developmental genes (alpha-fetoprotein, afp, glypican 3, gpc3, Homeobox X, Hba-X, and tenascin-C, Tnc). See Table 1 for complete list of genes.

FIG. 2.

Quantitative gene expression analysis of ES-derived cardiomyocytes (dark bars) isolated with LMPC by ABI® Micro Fluidic cards in comparison to adult heart (light bar). Transcript levels were normalized against undifferentiated ES cells (x axis). Cardiac-specific markers (Cardiac actin, actc-1, GATA-4, myosin heavy chain, myh6, and ryanodine receptor, ryr2) were expressed in LMPC ES-derived cardiomyocytes as well as early developmental genes (alpha-fetoprotein, afp, glypican 3, gpc3, Homeobox X, Hba-X, and tenascin-C, Tnc). See Table 1 for complete list of genes.

FIG. 3.

RT-PCR gene expression of functional cardiac markers in LMPC ES-derived cardiomyocytes, murine atrial tumor cell line (HL-5), and skeletal myocyte-derived cell line (C2C12): expression for all critical markers was confirmed in LMPC ES-derived cardiomyocytes. MLC2V, which is limited to heart ventricles was absent in atrial HL-5 cells as well as C2C12. Cardiac ion-gated channels (Cav1.2a, Kv 1.5, and mERG) were specific to ES-derived cardiomyocytes and atrial HL-5 cells.

FIG. 3.

RT-PCR gene expression of functional cardiac markers in LMPC ES-derived cardiomyocytes, murine atrial tumor cell line (HL-5), and skeletal myocyte-derived cell line (C2C12): expression for all critical markers was confirmed in LMPC ES-derived cardiomyocytes. MLC2V, which is limited to heart ventricles was absent in atrial HL-5 cells as well as C2C12. Cardiac ion-gated channels (Cav1.2a, Kv 1.5, and mERG) were specific to ES-derived cardiomyocytes and atrial HL-5 cells.

LMPC Beating Foci Are Composed of ES-Derived Cardiomyocytes with Structural Proteins for Cardiac L-Type Calcium Channel, Cardiac Actin, and Cardiac Myosin

Immunohistochemistry was conducted on heterogeneous adherent cultures containing beating foci prior to LMPC to investigate if contracting cellular aggregates were positive for muscle and cardiac muscle-specific markers. Immortalized mouse skeletal myocytes (C2C12) were used as controls for skeletal muscle–specific proteins (Fig. 1). At day 12, when LMPC was conducted, beating foci, but not adjacent cells, displayed robust muscle components (alpha-actinin, sarcomeric myosin) with cardiac specificity (Nkx2.5). Skeletal myosin, found in skeletal muscle fibers, was absent in ES-derived cardiomyocytes. Additionally, epitopes in C2C12 skeletal myocytes were not recognized by antibodies directed against cardiac-specific Nkx2.5. These findings suggest that beating foci are indeed composed of cardiac cells that should be enriched upon laser microdissection. Following LMPC, single ES-derived cardiomyocytes were analyzed by immunohistochemistry following enzymatic treatment of beating foci. Nuclear proteins were stained with DAPI, while the sarcolemmal calcium channel was labeled using an anti-Cav 1.2a antibody. This antibody, specific for the pore-forming alpha subunit of the cardiac L-type calcium channel, detected the presence of molecular mediators of excitation–contraction coupling in ES-derived cardiomyocytes (Fig. 4). The lack of a punctate staining pattern normally seen in adult cardiac myocytes containing the specialized dyadic cleft and T-tubular apparatus is suggestive of an early stage of cardiac development. Although diffuse and unorganized, this pattern of staining parallels similar findings in neonatal mammalian myocytes. Further, the isolated cells appear to be mononucleated as opposed to binucleated adult cardiac myocytes. Immunohistochemistry for cardiac actin and cardiac myosin was conducted on whole LMPC beating foci (not digested into single cells). Both cardiac markers were identified homogeneously throughout the isolated beating foci compared to negative IgG controls (Fig. 5). The intensity of antibody fluorescence, however, varied amongst cells within foci, suggesting that cardiomyocytes within aggregates may be at different developmental stages. Nonetheless, immunohistochemistry characterization was consistent in all LMPC-isolated foci and demonstrated the enrichment of cell types by LMPC applicable to in vitro functional studies for predictive cardiotoxicity.

FIG. 4.

Immunohistochemistry of ES-derived cardiomyocytes isolated with LMPC: alpha subunit 1 of the cardiac L-type calcium channel protein was detected in cells of beating foci.

FIG. 4.

Immunohistochemistry of ES-derived cardiomyocytes isolated with LMPC: alpha subunit 1 of the cardiac L-type calcium channel protein was detected in cells of beating foci.

FIG. 5.

Immunohistochemistry of ES-derived cardiomyocytes isolated with LMPC: structural cardiac proteins were detected in cells of beating foci. (a) Control IgG. (b) Anti-cardiac actin. (c) Anti-cardiac myosin.

FIG. 5.

Immunohistochemistry of ES-derived cardiomyocytes isolated with LMPC: structural cardiac proteins were detected in cells of beating foci. (a) Control IgG. (b) Anti-cardiac actin. (c) Anti-cardiac myosin.

ES-Derived Cardiomyocytes Exhibit Intracellular Calcium Transients and Functional Pharmacology in Response to Compounds

Spontaneously beating LMPC ES-derived cardiomyocytes were incubated with FLUO-3AM for measurement of intracellular calcium transients. Ventricular cardiac myocytes from adult mammalian species express a characteristic calcium transient, with a rapid upstroke caused by the influx of extracellular calcium, followed by intracellular calcium release from the sarcoplasmic reticulum. As displayed in Figure 6, the cultured ES-derived cardiomyocytes express heterogeneous calcium transients. Both pacemaker (Fig. 6a), and ventricular-like (Fig. 6b) calcium transients were present in separate beating foci from the same population of cells. In some instances, spontaneous pacemaker activity was overpowered by a depolarizing stimulus (generated by an external electrical pacer and platinum electrodes). In Figure 7, a beating focus was paced from 0.5–2.0 Hz in stepwise increments to mimic a frequency-dependent challenge.

FIG. 6.

Heterogeneity of calcium transients recorded from ES-derived cardiomyocytes. Spontaneously beating foci from ES-derived cardiomyocytes exhibit differing morphologies in the waveforms of their respective calcium transients, indicative of possible differences in calcium handling or sources (extracellular versus intracellular) contributing to cytosolic calcium. Both pacemaker (a), and ventricular-like (b) calcium transients were present in separate beating foci from the same population of cells. F = measured fluorescence.

FIG. 6.

Heterogeneity of calcium transients recorded from ES-derived cardiomyocytes. Spontaneously beating foci from ES-derived cardiomyocytes exhibit differing morphologies in the waveforms of their respective calcium transients, indicative of possible differences in calcium handling or sources (extracellular versus intracellular) contributing to cytosolic calcium. Both pacemaker (a), and ventricular-like (b) calcium transients were present in separate beating foci from the same population of cells. F = measured fluorescence.

FIG. 7.

Frequency dependence of calcium transients. Representative trace of an electrically induced calcium transient, following rhythmic pacing at multiple frequencies by field stimulation.

FIG. 7.

Frequency dependence of calcium transients. Representative trace of an electrically induced calcium transient, following rhythmic pacing at multiple frequencies by field stimulation.

Extracellular calcium levels in perfusion buffer were increased from 1.8 to 5.4 mM to modulate the chemical gradient of extracellular calcium. By increasing extracellular calcium, a marked increase in L-type calcium current was expected to promote larger intracellular calcium transient. As seen in Figure 8, increasing extracellular calcium three-fold caused a significant increase in the calcium transient (3.95% ± 1.1%) (n = 4, p ≤ 0.01). Further, blockade of the L-type calcium channel with 1 μM nifedipine, a dihydropyridine and anti-hypertension drug, significantly decreased the magnitude of the calcium transient (Fig. 9). In most cases, the potent inhibition of L-type calcium current abolished the contraction and calcium transient altogether.

FIG. 8.

Effect of increasing extracellular calcium on cytosolic calcium transient. Increasing extracellular calcium (black trace = 1.8 mM, red trace = 5.4 mM) increases the chemical gradient for calcium entry into the cell upon deploarization, allowing for a larger contraction and an increase in the calcium transient amplitude.

FIG. 8.

Effect of increasing extracellular calcium on cytosolic calcium transient. Increasing extracellular calcium (black trace = 1.8 mM, red trace = 5.4 mM) increases the chemical gradient for calcium entry into the cell upon deploarization, allowing for a larger contraction and an increase in the calcium transient amplitude.

FIG. 9.

L-type calcium channel blockade. Dihydropyridines such as nifedipine are specific blockers for the L-type calcium channel. Inhibition of the channel with 1 μM extracellular nifedipine caused a decrease in the calcium transient amplitude (black trace = control, red trace = nifedipine) and abolished the calcium transient and contraction in most cases. F = measured fluorescence. F/Fo = relative fluorescence or pseudo-ratio. This is defined as (F-Funst)/(Funst-Fbgrd) where F is the measured fluorescence, Funst is the unstimulated fluorescence of the cells, and Fbgrd is the background fluorescence.

FIG. 9.

L-type calcium channel blockade. Dihydropyridines such as nifedipine are specific blockers for the L-type calcium channel. Inhibition of the channel with 1 μM extracellular nifedipine caused a decrease in the calcium transient amplitude (black trace = control, red trace = nifedipine) and abolished the calcium transient and contraction in most cases. F = measured fluorescence. F/Fo = relative fluorescence or pseudo-ratio. This is defined as (F-Funst)/(Funst-Fbgrd) where F is the measured fluorescence, Funst is the unstimulated fluorescence of the cells, and Fbgrd is the background fluorescence.

DISCUSSION

We validated a novel approach to obtain enriched ES cell–derived cell types that were readily applicable to in vitro safety pharmacology assays. Heterogeneity is a major obstacle to in vitro differentiation of embryonic and adult stem cells (Iida et al., 2005). The ability to manipulate outcomes from ES cell differentiation is a significant challenge that needs to be addressed prior to drug discovery and therapeutic applications of this novel technology. Laser microdissection and pressure catapulting (LMPC) of beating foci from a population of ES cells yielded functional cardiomyocyte-like cells with regards to pharmacological response, protein immunohistochemistry, contractile properties, calcium homeostasis, and gene expression patterns.

Enriched populations of ES cell–derived cell types were previously reported following genetic selection and/or flow cytometry (Geijsen et al., 2004; Zhan et al., 2004). Genetic selection, however, requires timely and technically complex procedures to enable enrichment upon cell-type-specific gene function and/or promoter activities (Hidaka et al., 2003). LMPC offers an innovative and faster alternative and can potentially be used to isolate any ES cell–derived cell type. Routinely, LMPC has been successful as a means of cell enrichment in tumor samples (Haqq et al., 2005; Rubin and De Marzo, 2004). This is the first study, to our knowledge, to report the application of LMPC as a means for the enrichment of differentiated embryonic stem cells.

LMPC beating foci from ES cells expressed the structural markers cardiac alpha-actin and cardiac myosin heavy chain as well as cardiac-specific transcription factors, such as GATA-4. In vivo, cardiogenesis is dependent upon expression of these transcription factors, particularly GATA-4, Nkx 2.5, and Mef2c (reviewed in Harvey, 2002). These transcription factors promote the embryonic development of the heart, as well as the various cell types associated with the terminal differentiation of the cardiac conduction system and myocardium. We investigated expression of genetic markers and the presence of proteins that infer the cardiac phenotype (Figs. 2 and 3). Heart rate is determined in part by cardiac ryanodine receptor-mediated calcium release (Yang et al., 2002), whose expression was confirmed by RT-PCR. Specifically, the presence of a transcript for cardiac actin, troponin-c, alpha-myosin heavy chain, and MLC2v provide evidence for the presence of the contractile apparatus (Fig. 3). Although the focus of our study was to evaluate the calcium homeostasis associated with myocyte contraction, previous studies have demonstrated the ability of ES-derived cardiomyocytes to produce contractile force (Metzger et al., 1995). Myosin heavy chain is a critical protein for cardiac-specific motor function and is directly correlated to contractile velocity (Miyata et al., 2000). Indeed, our data suggest the transition from beta to alpha myosin heavy chain after several days of in vitro differentiation of ES-derived cardiomyocytes, a transition that mirrors the stoichiometry of myosin heavy chain isoforms from fetal to adult cardiac maturation. At the protein level, cardiomyocyte aggregates isolated by laser microdissection were homogeneous for cardiac-specific alpha-actin and myosin heavy chain (Franke, 1996).

ES-derived cardiomyocytes expressed mRNA for early development genes (alpha-fetoprotein, glypican 3, homeobox X, tenascin-C) that were absent or downregulated in the adult heart. This gene expression pattern (Fig. 2) suggests that ES-derived cardiomyocytes are immature in direct molecular comparison to adult heart. This finding is consistent with previous studies (Klug et al., 1996), where in vitro differentiation of ES cells generated cardiomyocytes with functional properties of perinatal versus adult cells. The same evidence exists in studies on human embryonic stem cells (hES), where ES-derived cardiomyocytes differed from adult counterparts despite functional cardiac properties of conduction and electrophysiology (Maltsev et al., 1994). Additionally, other findings in hES cells, such as variability in size and shape of contracting foci were shared in our study. LMPC beating foci varied between 400 and 700 microns in diameter. The challenge to drive ES-derived cardiomyocytes into an adult phenotype remains and may be resolved upon integrated approaches that also target nongenetic cardiac inducers, such as mechanical forces and neighboring cell signaling (Hove et al., 2003). Nonetheless, isolated foci were readily applicable to in vitro pharmacology assays in 96-well format, offering significant advantage to currently available in vitro cardiotoxicity substrates, i.e., immortalized cell lines that often lack native cardiac channels. Despite cardiac origin, immortalized cell lines usually lack typical features of cardiomyocytes, such as contraction and generation of action potential (Caviedes et al., 1993). It is also noteworthy to mention another benefit of this system, i.e., a predictive cardiotoxicity assay that does not require primary cardiomyocytes or the use of experimental animals.

In many instances, beating foci retained contractility during LMPC, whereas in others, foci recovered beating activity following a 6 to 12 h of in vitro culture. LMPC was not detrimental to survival of target cells. Functional and molecular characterization was performed exclusively on beating LMPC cells. Therefore, LMPC preserved cellular viability as well as functional cell–cell interactions that enable pacemaker-driven contractility of cardiomyocyte aggregates. As a matter of fact, studies are in progress to evaluate cardiac conduction and field potential recording in LMPC ES-derived cardiomyocytes with multielectrode arrays.

The most widely studied process of myocardial function is the mechanism of excitation–contraction coupling (EC-coupling), which begins with the propagation of a regenerative action potential, allowing for the influx of extracellular calcium to induce the release of intracellular calcium stores to produce mechanical force (Bers, 2002). The characteristic waveform of the cardiac action potential is generated by the sequential activation and inactivation of sarcolemmal voltage-gated channels such as NAV 1.5 sodium, L-type calcium, and various repolarizing potassium channels (Bers, 2001). Transcripts for these channels were detected in the ES-derived cardiomyocytes, in comparison to a spontaneously contracting murine atrial tumor cell line (HL-5). Transcription detection, however, cannot infer the presence of functionally competent channels. Ionic channels were expressed in c-kit+ stem cells but did not support the coupling between excitation and contraction typical of cardiac myocytes (Lagostena et al., 2005). In our study, the presence of protein for cardiac alpha subunit of the L-type calcium channel on the sarcolemmal membrane (Fig. 4) also corroborated the cardiac lineage of these cells (Acosta et al., 2004).

To provide functional evidence for the presence of cardiac excitation contraction coupling and generate a safety pharmacology assay, we chose to modulate the activity of the L-type calcium channel while recording intracellular calcium transient dynamics. This is the first study of its kind, utilizing the calcium transient to infer a cardiac lineage while also describing the heterogeneity of calcium transients found in the same population of cultured ES-derived cells. Previous studies have described the heterogeneity of the action potentials found within the same population of differentiating ES-derived cardiomyocytes, with direct comparisons to action potentials found in the various regions of the heart (He et al., 2003). In these studies, action potentials resembled those found in nodal, conducting fibers and those from atrial versus ventricular myocytes (He et al., 2003). The calcium transients observed in our studies showed similar variability (Fig. 6). In other words, spontaneously beating foci generated transients with slowly rising diastolic calcium followed by a rapid upstroke of the transient. Conversely, other foci showed rapidly rising calcium transients, followed by a rapid relaxation/restitution phase. The genesis of the ES-derived calcium transient remains unclear, although studies in neonatal and ES-derived cardiomyocytes suggest the trans-sarcolemmal dependence of cells in these stages of development as compared to the SR dependent transient in adult myocytes (Chin et al., 1990).

In order to establish a safety pharmacology assay based upon modulation of L-type calcium channel and the calcium transient, extracellular calcium was increased three-fold. The calcium dependence of intracellular calcium release is based upon an increase in the trans-sarcolemmal flux of calcium. In the presence of increased extracellular calcium, ES-derived cardiomyocytes responded similarly to adult murine ventricular myocytes by increasing magnitude of the calcium transient (Gao et al., 1998). Although we have not isolated the contribution of the SR versus the L-type calcium channel influx, blockade of the calcium channel with nifedipine reduced or abolished the calcium transient in each beating foci tested. The robust response of ES-derived cardiomyocytes to chemical modulators in 96-well format is the basis for a stem cell-based safety screen. This system is therefore applicable to evaluate cardiotoxic effects of multiple compounds.

The absence of differentiation methods that generate large percentages (>50%) of cardiomyocytes from ES cells has substantially hindered progress of in vitro screening assays and clinical application studies (Kehat et al., 2002; Kumar et al., 2005). The successful outcomes of LMPC to obtain an enriched population of differentiated cells from ES cells may support its application in cell therapy studies. ES-derived cardiomyocytes were viable and retained critical cardiac functions that suggest that intact, enriched cell types could be generated via laser microdissection for therapeutic intervention in the future. Further investigation is required to determine if other ES cell–derived cell types, such as hepatocytes, pancreatic islets, and neuronal precursors can be manipulated by LMPC for robust toxicity screening and clinical applications.

1
Current address: Lexicon Genetics Inc., 8800 Technology Forest Place, The Woodlands, Texas 77381.
2
Current address: University of Wisconsin-Madison, Department of Animal Sciences, 1675 Observatory Drive, Madison, Wisconsin, 53706.

The authors thank Jessica Quam (University of Wisconsin-Madison, Department of Animal Sciences) for her support on the preparation of this manuscript.

References

Acosta, L. H., Haase, H., Morano, I., Moorman, A. F., and Franco, D. (
2004
). Regional expression of L-type calcium channel subunits during cardiac development.
Dev. Dyn.
 
230
,
131
–136.
Beltrami, A., Barlucchi, L., Torella, D., Baker, M., Limana, F., Chimenti, S., Kasahara, H., Rota, M., Musso, E., Urbanek, K., et al. (
2003
). Adult cardiac stem cells are multipotent and support myocardial regeneration.
Cell
 
114
,
763
–776.
Bers, D. M. (
2001
). Excitation-Contraction Coupling and Cardiac Contractile Force. Kluwer Academic Publishers, Boston, MA.
Bers, D. M. (
2002
). Cardiac excitation-contraction coupling.
Nature
 
415
,
198
–205.
Boheler, K. R. (
2003
). ES cell differentiation to the cardiac lineage.
Methods Enzymol.
 
365
,
228
–241.
Caviedes, P., Olivares, E., Salas, K., Caviedes, R., and Jaimovich, E. (
1993
). Calcium fluxes, ion currents and dihydropyridine receptors in a new immortal cell line from rat heart muscle.
J. Mol. Cell. Cardiol.
 
25
,
829
–845.
Chin, T. K., Friedman, W. F., and Klitzner, T. S. (
1990
). Developmental changes in cardiac myocyte calcium regulation.
Circ. Res.
 
67
,
574
–579.
Chung, C. Y., Seo, H., Sonntag, K. C., Brooks, A., Lin, L., and Isacson, O. (
2005
). Cell type-specific gene expression of midbrain dopaminergic neurons reveals molecules involved in their vulnerability and protection.
Hum. Mol. Genet.
 
14
,
1709
–1725.
Davila, J. C., Cezar, G. G., Thiede, M., Strom, S., Miki, T., and Trosko, J. (
2004
). Use and applications of stem cells in toxicology.
Toxicol. Sci.
 
79
,
214
–223.
DiMasi, J. A. R., Hansen, R. W., and Grabowski, H. G. (
2003
). The price of innovation: New estimates of drug development costs.
J. Health Econ.
 
22
,
151
–185.
Espina, V., Geho, D., Mehta, A. I., Petricoin, E. F., 3rd, Liotta, L. A., and Rosenblatt, K. P. (
2005
). Pathology of the future: Molecular profiling for targeted therapy.
Cancer Invest.
 
23
,
36
–46.
Fijnvandraat, A. C., van Ginneken, A. C., Schumacher, C. A., Boheler, K. R., Lekanne Deprez, R. H., Christoffels, V. M., and Moorman, A. F. (
2003
). Cardiomyocytes purified from differentiated embryonic stem cells exhibit characteristics of early chamber myocardium.
J. Mol. Cell. Cardiol.
 
35
,
1461
–1472.
Franke, W. (
1996
). Specific immunohistochemical detection of cardiac fetal αctin in human cardiomyocytes and regenerating skeletal muscle cells.
Differentiation
 
60
,
245
–250.
Gao, W. D., Perez, N. G., and Marban, E. (
1998
). Calcium cycling and contractile activation in intact mouse cardiac muscle.
J. Physiol.
 
507
,
175
–184.
Geijsen, N., Horoschak, M., Kim, K., Gribnau, J., Eggan, K., and Daley, G. Q. (
2004
). Derivation of embryonic germ cells and male gametes from embryonic stem cells.
Nature
 
427
,
148
–154.
Haqq, C., Nosrati, M., Sudilovsky, D., Crothers, J., Khodabakhsh, D., Pulliam, B. L., Federman, S., Miller, J. R., 3rd, Allen, R. E., Singer, M. I., et al. (
2005
). The gene expression signatures of melanoma progression.
Proc. Natl. Acad. Sci. U.S.A.
 
102
,
6092
–6097.
Harvey, R. P. (
2002
). Patterning the vertebrate heart.
Nat. Rev. Genet.
 
3
,
544
–556.
He, J., Ma, Y., Lee, Y., Thomson, J. A., and Kamp, T. J. (
2003
). Human embryonic stem cells develop into multiple types of cardiac myocytes: Action potential characterization
Circ. Res.
 
93
,
32
–39.
Hescheler, J., Fleischmann, B. K., Lintini, S., Maltsev, V. A., Rohwedel, J., Wobus, A. M., and Addicks, K. (
1997
). Embryonic stem cells: A model to study structural and functional properties in cardiomyogenesis.
Cardiovasc. Res.
 
36
,
149
–162.
Hidaka, K., Lee, J. K., Kim, H. S., Ihm, C. H., Iio, A., Ogawa, M., Nishikawa, S., Kodama, I., and Morisaki, T. (
2003
). Chamber-specific differentiation of Nkx2.5-positive cardiac precursor cells from murine embryonic stem cells.
FASEB J.
 
17
,
740
–742.
Hove, J. R., Koster, R. W., Forouhar, A. S., Acevedo-Bolton, G., Fraser, S. E., and Gharib, M. (
2003
). Intracardiac fluid forces are an essential epigenetic factor for embryonic cardiogenesis
Nature
 
421
,
172
–177.
Iida, M., Heike, T., Yoshimoto, M., Baba, S., Doi, H., and Nakahata, T. (
2005
). Identification of cardiac stem cells with FLK1, CD31, and VE-cadherin expression during embryonic stem cell differentiation.
FASEB J.
 
19
,
371
–378.
Kehat, I., Gepstein, A., Spira, A., Itskovitz-Eldor, J., and Gepstein, L. (
2002
). High-resolution electrophysiological assessment of human embryonic stem cell-derived cardiomyocytes: A novel in vitro model for the study of conduction.
Circ. Res.
 
91
,
659
–661.
Klug, M. G., Soonpaa, M. H., Koh, G. Y., and Field, L. J. (
1996
). Genetically selected cardiomyocytes from differentiating embryonic stem cells form stable intracardiac grafts.
J. Clin. Invest.
 
98
,
216
–224.
Kumar, D., Kamp, T. J., and LeWinter, M. M. (
2005
). Embryonic stem cells: Differentiation into cardiomyocytes and potential for heart repair and regeneration.
Coron. Artery Dis.
 
16
,
111
–116.
Lagostena, L. D., Avitabile, D., De Falco, E., Orlandi, A., Grassi, F., Iachininoto, M. G., Ragone, G., Fucile, S., Pompilio, G., Eusebi, F., et al. (
2005
). Electrophysiological properties of mouse bone marrow c-kit(+) cells co-cultured onto neonatal cardiac myocytes.
Cardiovasc. Res.
 
66
,
482
–492.
Maltsev, V. A., Wobus, A. M., Rohwedel, J., Bader, M., and Hescheler, J. (
1994
). Cardiomyocytes differentiated in vitro from embryonic stem cells developmentally express cardiac-specific genes and ionic currents.
Circ. Res.
 
75
,
233
–244.
Metzger, J. M., Lin, W. I., Johnston, R. A., Westfall, M. V., and Samuelson, L. C. (
1995
). Myosin heavy chain expression in contracting myocytes isolated during embryonic stem cell cardiogenesis.
Circ. Res.
 
76
,
710
–719.
Miyata, S., Minobe, W., Bristow, M. R., and Leinwand, L. A. (
2000
). Myosin heavy chain isoform expression in the failing and nonfailing human heart.
Circ. Res.
 
86
,
386
–390.
Roach, M. L., Stock, J. L., Byrum, L., Kohler, B. H., and McNeish, J. D. (
1995
). A new embryonic stem cell line from DBA/1lacJ mice allows genetic modification in a murine model of human inflammation.
Exp. Cell Res.
 
2
,
520
–525.
Rubin, M. A., and De Marzo, A. M. (
2004
). Molecular genetics of human prostate cancer
Mod. Pathol.
 
17
,
380
–388.
Schütze, K. (
2003
). Tissue microdissection, laser pressure catapulting. In DNA Microarrays: A Molecular Cloning Manual. (D. Bowtell, and J. Sambrook, Eds.), pp. 331–356. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY.
Stich, M., Thalhammer, S., Bugemeister, R., Friedmann, G., Ehnle, S., Luthy, C., and Shütze, K. (
2003
). Live cell catapulting and recultivation.
Pathol. Res. Pract.
 
199
,
405
–409.
Takahashi, A., Camacho, P., Lechleiter, J. D., and Herman, B. (
1999
). Measurement of intracellular calcium.
Physiol. Rev.
 
79
,
1089
–1125.
Thomson, J. A, Itskovitz-Eldor, J., Shapiro, S. S., Waknitz, M. A., Swiergiel, J. J., Marshall, V. S., and Jones, J. M. (
1998
). Embryonic stem cell lines derived from human blastocysts.
Science
 
282
,
1145
–1147.
van Kempen, M. J. A., van Ginneken, A., de Grijs, I., Mutsaers, N., Opthof, T., Jongsma, H., and van der Heyden, M. (
2003
). Expression of the Electrophysiological system during murine embryonic stem cell cardiac differentiation.
Cell Physiol. Biochem.
 
13
,
263
–270.
Vogel, A., and Venugopalan, V. (
2003
). Mechanisms of pulsed laser ablation of biological tissues.
Chem. Rev.
 
102
,
577
–644.
Yang, H. T., Tweedie, D., Wang, S., Guia, A., Vinogradova, T., Bogdanov, K., Allen, P. D., Stern, M. D., Lakatta, E. G., and Boheler, K. R. (
2002
). The ryanodine receptor modulates the spontaneous beating rate of cardiomyocytes during development.
Proc. Natl. Acad. Sci. U.S.A.
 
99
,
9225
–9230.
Yuasa, S., Itabashi, Y., Koshimizu, U., Tanaka, T., Sugimura, K., Kinoshita, M., Hattori, F., Fukami, S., Shimazaki, T., Okano, H., et al. (
2005
). Transient inhibition of BMP signaling by Noggin induces cardiomyocyte differentiation of mouse embryonic stem cells.
Nat. Biotechnol.
 
23
,
607
–611.
Zandstra, P. W., Bauwens, C., Yin, T., Liu, Q., and Schiller, H. (
2003
). Scalable production of embryonic stem cell-derived cardiomyocytes.
Tissue Eng.
 
9
,
767
–778.
Zhan, X., Dravid, G., Ye, Z., Hammond, H., Shamblott, M., Gearhart, J., and Cheng, L. (
2004
). Functional antigen-presenting leucocytes derived from human embryonic stem cells in vitro.
Lancet
 
364
,
163
–171.

Author notes

*Pfizer Global Research and Development, Chesterfield, Missouri 63017; and †Pfizer Global Research and Development, Groton, Connecticut 06340