A myosin activator improves actin assembly and sarcomere function of human-induced pluripotent stem cell-derived cardiomyocytes with a troponin T point mutation

Cardiac myocytes were derived from induced pluripotent stem cells from normal and family members expressing a mutant cardiac troponin T linked to dilated cardiomyopathy. Shortening, actin content and assembly dynamics were depressed in the severely affected mutant but reversed by a myosin activation reagent. Sarcomeric isoform composition was fetal/neonatal.

Abstract

We have investigated cardiac myocytes derived from human-induced pluripotent stem cells (iPSC-CMs) from two normal control and two family members expressing a mutant cardiac troponin T (cTnT-R173W) linked to dilated cardiomyopathy (DCM). cTnT is a regulatory protein of the sarcomeric thin filament. The loss of this basic charge, which is strategically located to control tension, has consequences leading to progressive DCM. iPSC-CMs serve as a valuable platform for understanding clinically relevant mutations in sarcomeric proteins; however, there are important questions to be addressed with regard to myocyte adaptation that we model here by plating iPSC-CMs on softer substrates (100 kPa) to create a more physiologic environment during recovery and maturation of iPSC-CMs after thawing from cryopreservation. During the first week of culture of the iPSC-CMs, we have determined structural and functional characteristics as well as actin assembly dynamics. Shortening, actin content, and actin assembly dynamics were depressed in CMs from the severely affected mutant at 1 wk of culture, but by 2 wk differences were less apparent. Sarcomeric troponin and myosin isoform composition were fetal/neonatal. Furthermore, the troponin complex, reconstituted with wild-type cTnT or recombinant cTnT-R173W, depressed the entry of cross-bridges into the force-generating state, which can be reversed by the myosin activator omecamtiv mecarbil. Therapeutic doses of this drug increased both contractility and the content of F-actin in the mutant iPSC-CMs. Collectively, our data suggest the use of a myosin activation reagent to restore function within patient-specific iPSC-CMs may aid in understanding and treating this familial DCM.

NEW & NOTEWORTHY

Cardiac myocytes were derived from induced pluripotent stem cells from normal and family members expressing a mutant cardiac troponin T linked to dilated cardiomyopathy. Shortening, actin content and assembly dynamics were depressed in the severely affected mutant but reversed by a myosin activation reagent. Sarcomeric isoform composition was fetal/neonatal.

in experiments reported here, we have investigated cardiac myocytes derived from human-induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) from two normal and two family members expressing a mutant cardiac troponin T (cTnT-R173W) linked to dilated cardiomyopathy (DCM) (7, 44, 49). Family members expressing cTnT-R173W demonstrate varying degrees of DCM, with a most severe case occurring in a 14-year-old male, who required cardiac transplantation (44). cTnT is a scaffolding protein of the sarcomeric thin filament with regions of interactions with cardiac troponin I (cTnI), the inhibitory unit of the troponin complex, with cardiac troponin C (cTnC), the Ca²⁺-binding unit, and with tropomyosin (Tm). By positioning Tm on the actin strand, cTnT has a major role in establishing the diastolic state at Ca²⁺ concentrations below the threshold for activation (45). cTnT is also critical for transduction of the Ca²⁺-cTnC signal to cTnI and to Tm in triggering the systolic state (23). R173 of cTnT is located at a hinge region between the tail of cTnT (T1), which interacts with Tm, and the head (T2), which in turn interacts with cTnI-cTnT R173 (32). Thus cTnT-R173 is strategically located to participate in these control mechanisms, and it is not surprising that loss of this basic charge has consequences leading to progressive DCM.

Despite the promise of iPSC-CMs as a valuable platform for understanding the connections between mutations in sarcomeric proteins, DCM, and sudden death, there are important questions to be addressed. First, understanding sarcomeric assembly and maturation is of critical importance for these iPSC-CMs, and we monitored the cells as they built their myofibrillar architecture in culture. One way to create a condition known to influence the maturation of myocytes is to use softer polymers as the substrate for the culture dish (12, 14). Our lab has used polydimethylsiloxane (PDMS) for primary cardiomyocyte culture experiments (34), and here we use 100 kPa PDMS (5) to create a more physiologic environment during recovery and maturation of iPSC-CMs after thawing from cryopreservation. Using these substrates, we have determined sarcomeric structural and functional characteristics as well as actin assembly dynamics in iPSC-CMs from family members who do or do not possess a cTnT-R173W point mutation. Moreover, our data show that there was a depression in shortening and actin assembly in the severely affected mutant CM at 1 wk of maturation in culture, but by 2 wk the morphology and shortening were more normal. This leads to a follow-up question of how can the deficiency in maturation and assembly of the cTnT-R173W iPSC-CMs be rectified?

On the basis of effects of other thin filament mutations leading to DCM, we expected that sarcomeres controlled by cTnT-R173W would demonstrate a reduced response to Ca²⁺ (13, 33, 37). To test this expectation, we employed myofilament preparations controlled by a troponin (Tn) complex reconstituted with recombinant wild-type cTnT or cTnT-R173W. The data showed that cTnT-R173W induced a depression in activation of cross-bridge kinetics. In view of evidence that the small molecule omecamtiv mecarbil (OM) is able to promote entry of cross-bridges into the force-generating state, we treated the iPSC-CMs with this sarcomere activator, which has been shown to be effective in improving acquired heart failure by directly interacting with myosin (18, 31). OM treatment increased both contractility and the content of F-actin expressed in the iPSC-CMs. Thus our data suggest the use of a myosin activation reagent for treatment of point mutation cTnT-R173W to restore function within patient-specific iPSC-CMs. These results demonstrate a potential therapeutic intervention in understanding and treating this familial DCM.

METHODS

Substrate fabrication and preparation for culture.

The process of fabricating flat polydimethylsiloxane (PDMS) substrates has previously been described by our lab (5, 34). Briefly, a 50:1 elastomer base-to-curing agent (Dow Corning, Midland, MI) ratio was used to create softer PDMS, mixed for 5 min and vacuum degassed for ∼30 min to remove air bubbles. The PDMS was then spun onto cell culture–compatible plates (MatTek, Ashland, MA) to create an approximately 50-micron thickness. PDMS substrates were polymerized in a 60°C oven for 24 h. When cooled, stiffness was measured by atomic force microscopy to yield a Young's modulus of 98.4 ± 7.36 kPa, which was rounded off to 100 kPa (5). Substrates were rinsed with 200 proof isopropyl, washed three times with PBS, and layered with a coat of a 1:60 solution of Matrigel Basement Membrane Matrix Growth Factor Reduced (BD Biosciences, San Diego, CA)/DMEM/F12 1:1 (Life Technologies, Grand Rapids, NY) and kept at least 30 min in an incubator for before seeding the cells. The stiffness of glass is 61.9 GPa (47).

iPSC-CM culture.

The procedure of converting iPSCs from fibroblasts taken from skin biopsies of patients is well documented (6, 42). Cells used in the present study were from two normal cell lines, and two mutant cell lines verified by DNA sequencing and PCR to be affected with the cTnT R173W point mutation as well as showing spontaneous beating and cardiac electrophysiological properties (42). Certification of human research for derivation of human-induced pluripotent stem cells was approved by the Stanford University IRB Protocol ID 17576, IRB 350 (Pane:3) to Joseph Wu, MD, PhD. The nomenclature of the pedigree is given in Fig. 1 of Sun et al. (44), and here we use N1 for IIIb, N2 for Ib, M2 for IIb, and M1 for IIIa. iPSC-CMs were aliquoted and stored in liquid nitrogen at Stanford University, coded then shipped on dry ice in vials containing 1–2 million cells, and stored at −80°C at UIC until used. In total, UIC analyzed 20 vials of normal iPSC-CMs and 16 vials of mutant iPSC-CMs. The experiments were completed and then decoded for statistical analysis. Results were reproducible from lot to lot over a 2-yr time period.

Fig. 1.

iPSC-CM maturation of normal and cTnT R173W mutant cells with time after thawing and plating on 100-kPa substrate. At 1 and 2 wk after cryopreservation and plating, both normal and mutant iPSC-CMs were fixed and stained for sarcomeric α-actinin (antibody, green), F-actin (phalloidin, red), and nucleus (DAPI, blue). A: at 2 wk, normal (N1 line) iPSC-CMs have formed well-organized myofibrils with sarcomeric striations. The severely affected mutant (M1 line) iPSC-CMs at 1 wk are less organized and have few striations, whereas the less affected mutant (M2 line) has some sarcomeres. B: at week 2, normal (N1) iPSC-CMs are well striated, whereas the mutant (M1) iPSC-CMs are still less well organized. Scale bar, 20 μm.

Cells were thawed in warm bath water, resuspended in RPMI (Life Technologies) + 10% fetal bovine serum (FBS) (Life Technologies) and centrifuged at 500 RPM for 5 min into pellets. The supernatant was aspirated, and cells were resuspended in RPMI with 2% B-27 supplement with insulin (Life Technologies), 1% Pen-Strep antibiotic (Life Technologies), and 2% FBS to improve cell attachment. Excess Matrigel was aspirated, and cells were seeded at low density (250,000 to 500,000 cells/35-mm dish) on the substrates and maintained in a standard cell culture incubator (37°C, 5% CO₂) until used for experiments. The iPSC-CM media (RPMI with 2% B-27 supplement with insulin, 2% FBS and 1% Pen-Strep) was aspirated and replenished with fresh media every 2 days.

Neonatal rat ventricular myocyte culture.

Primary heart cultures were obtained from neonatal rats according to Institutional Animal Care and Use Committee and National Institutes of Health guidelines for the care and use of laboratory animals. Hearts were removed and cells isolated from 1- to 2-days-old neonatal Sprague-Dawley rats with collagenase (Worthington), as previously described (29).

Immunohistochemistry and actin quantification.

iPSC-CMs from both normal and mutant cell lines were fixed in 4.0% paraformaldehyde, rinsed in 0.3 M glycine/PBS (Media Tech, Manassas, VA), and incubated for 10 min to permeabilize membranes with 0.1% Triton X-100 (Sigma, St. Louis, MO). The primary antibody for sarcomeric α-actinin (mouse monoclonal ab9465, Abcam, Cambridge, MA) was diluted 1/200 in 1% BSA (Sigma)/PBS and incubated overnight at 4°C. iPSC-CMs were washed multiple times with PBS and incubated for 45 min at room temperature with rabbit antimouse Alexa Fluor IgG 488 antibody (Invitrogen) diluted 1/1,000. All cells were mounted in Vecashield with DAPI (Vector Labs, Burlingame, CA) and stored at 4°C until imaged. At least 10 iPSC-CM images were randomly taken from each culture condition with a Zeiss LSM 710 confocal microscope (Zeiss, Peabody, MA) with oil on an α Plan-Apochromat ×63/1.46 Imm Corr DIC (WD = 0.24 mm) objective lens. These images were used for the analysis of iPSC-CM myofibril development using fluorescent wavelengths of 358, 488, or 561 nm for the nucleus, α-actinin, and F-actin, respectively.

For the assessment of sarcomeric F-actin, the cell membranes were permeabilized, and Rhodamine Phalloidin (R415, Invitrogen) was diluted 1/100 in 1% TBST (Sigma) with 1% BSA/PBS mix at room temperature for 45 min and washed with PBS. Confocal microscopy had established that almost all the myofibrils were in the bottom few microns of the cell; therefore, fluorescent images taken with the pin hole wide open are equivalent to the 3D volume of a cell (34). Ten randomly selected images were recorded (pixel dimensions of x: 0.09 μm; y: 0.09 μm) for the two normal and two mutant cell lines, untreated or treated with 500 nM OM yielding 10–30 cells per image. Cells with striations were used to quantify the actin content per CM. The background threshold intensity of nonspecific staining was first subtracted and then the image was converted to a binary notation, meaning that the pixels were either red (F-actin) or black, making it easy to distinguish between assembled actin from the rest of the cell. The perimeter of the cell was drawn by hand, and ImageJ software was used to determine the fraction of red pixels per cell, which is a metric for the fraction of the CM volume filled by phalloidin fluorescence (a semiquantitative measure of F-actin content in the striated cells).

Live cell video recording.

iPSC-CMs were examined with a Zeiss LSM Observer Z1 microscope (Zeiss). Videos of 10 randomly selected fields from each dish were recorded for 15 s with phase microscopy (×20 objective). The percent of cells beating per dish and beats per minute were calculated each day for normal and mutant cells from 2 to 14 days after plating. Detailed studies were conducted at week 1 (days 5–8) and week 2 (days 12–14).

Kymograph analysis of shortening and shortening velocity.

Myofibrillar contractility was measured by line scans taken along the major axis of the cell as reported by our lab (5, 11). This method is analogous to echocardiography where displacement of the heart wall over time is used to assess contractile function. The cells were selected based on an easily defined myofibril axis for the line scan. A Zeiss 710 Confocal Microscope (Zeiss) was used through a water objective on a Universal ×25/0.80 Imm Corr DIC (WD = 0.21 mm) lens and recorded with Zen software (Zeiss). Transmitted light from a differential interference contrast channel was captured from each line every 7.56 ms over the course of 1,000 to 5,000 ms, thereby recording multiple complete and regular iPSC-CM contraction cycles. Peaks of contraction are rendered and used to mark states of maximum shortening and relate such distance to time, from which to calculate shortening velocity. Image processing software (LSM Browser, Zeiss) was used to scale the pixel to micron distance as well as the time to obtain the shortening and the associated velocity in single iPSC-CM contractions. Measurements of shortening and shortening velocity were taken of beating cells at 1 and 2 wk after plating. The maximum shortening and associated velocity for iPSC-CMs along the major myofibrillar axis was measured for 3 contractions per iPSC-CMs and a total of at least 12 iPSC-CMs over the course of at least three independent experiments per condition were measured. (n = 24 to 42).

SDS-PAGE and Western blot analysis.

Protein concentrations were determined by a RCDC assay (BioRad), following the manufacturer's instructions. Western blot samples were loaded at 8 μg of total protein per lane, and samples for myosin heavy chain (MHC) separation were loaded at 6 μg of total protein per lane. Samples were separated in 12% SDS-PAGE (17) and transferred onto Immun-blot PVDF membrane (BioRad) for Western blot analysis (25). A 6% SDS-PAGE gel was run to separate MHC isoforms (48). The primary antibodies used were JLT-12 cTnT 1:1,000 from Sigma Chemical and C5 TnI 1:5,000 from Fitzgerald. The secondary antibody was an antimouse from Sigma and used at 1:80,000. The signal was detected with Thermo Scientific SuperSignal West Femto Maximum substrate, following manufacturer's recommendations. The gels and Western blots were imaged with a Chemidoc XRS+ from BioRad. Band intensities were determined by densitometry with Image Lab software version 5.0 from BioRad. We only ratioed two isoform bands for cTnI, cTnT, and MHC within each lane to determine percent of the isoform vs. the total, thus no loading controls were necessary.

Actin dynamics by analysis of fluorescence recovery after photobleaching.

To examine sarcomere assembly in iPSC-CMs, actin-GFP expression was induced by CellLight Reagents BacMam 2.0 actin-GFP in a 2-day infection process. Two days before experiments, an appropriate volume of CellLight Reagent (35 μl/500,000 cells) was used as modified from the manufacturer's instructions. One day before experiments, a volume of 70 μl/500,000 cells of CellLight Reagent was used. Infected iPSC-CMs were returned to the culture incubator for at least 16 h. Fluorescence recovery after photobleaching (FRAP) analysis was conducted as previously described by our lab (29). Striated iPSC-CMs were selected for FRAP with a Zeiss LSM 710 confocal microscope. The intensity (I_FRAP) of the region of interest (ROI) (area = 3.75 × 3.75 μm) was observed both before and after bleaching at full power. For FRAP assays, the sample number was defined as individual cells, of which one to three cells were analyzed per culture and at least three separate cultures analyzed per experimental condition.

Protein purification and ATPase measurements with reconstituted myofilaments.

We expressed and purified recombinant cTnI employing pET3d, cTnT-Wt, cTnT-R173W containing an N-terminal myc tag in pSBET, and cTnC in pET3d (24). Tn complex was reconstituted and verified. Actin was isolated from rabbit fast skeletal muscle acetone powder, and tropomyosin was prepared from bovine cardiac ether powder. Myosin-S1 was made by chymotryptic digestion of rabbit psoas muscle myosin. Myofilaments were reconstituted with varying concentrations of Tn complex containing either cTnT-Wt or cTnT-R173W, 0.2 μM myosin S1, 7 μM actin, and 1 μM tropomyosin. Final assay conditions were 35 mM NaCl, 5 mM MgCl₂, 20 mM MOPS, 1 mM ATP, pH 7.0. The progress of ATP splitting was determined at 27°C in 0.1 mM Ca²⁺ or in 2 mM EGTA. At 30 min, a time when the reaction rate remained linear, the reaction was quenched in 0.2 M perchloric acid at 10°C. Phosphate production was measured with a malachite green assay by measuring absorbance at 655 nm.

Omecamtiv mecarbil therapy.

OM (Selleckchem, Houston, TX), formerly called CK-1827452 (Cytokinetics, San Francisco, CA), was prepared at 200 or 500 nM (42). Efficacy of the drug in culture conditions was first confirmed with neonatal rat ventricular myocytes (NRVM) cultured by our published methods (29). For OM dosing, iPSC-CMs were treated with OM at the initial time of cell seeding, and new drug included with the media exchanged every other day until experimentation before actin content, FRAP, or contractility assessments were made.

Statistical analysis.

Data was organized with Excel software (Microsoft, Redmond, CA), and statistical analysis was performed with GraphPad Prism (GraphPad Software, La Jolla, CA). Sample size varied for the different techniques as described above and is clearly indicated in the figures. The data shown for hIPSC-CM are many observations on a single population from a cell line, not multiple replicates. Therefore all data from the hIPSC-CM cell lines are expressed as means ± SD according to AJP guidelines (9, 10). SE is used for rodent studies where multiple replicates are available. The number of cells or samples is indicated on the bar and/or in the legend. Significant differences in variables were determined by using either two-tailed Student's t-test, one-way ANOVA, or two-way ANOVA.

RESULTS

iPSC-CM structure and actin content as a function of time after plating on 100 kPa substrates.

Representative images show myofibril formation after cryopreservation, replating, and 1 or 2 wk of culture when stained for actin, α-actinin, and the nucleus (Fig. 1). Cells fixed at the times stated and did not have any signs of cell death. At week 1, normal iPSC-CMs had well-developed striations, and the M2 mutant had striations but not as well delineated by α-actinin. However, the mutant M1 cells from the severely affected patient were much less organized with fewer striations (Fig. 1A). At week 2, the normal cells remained striated and the mutant M1 cells were improved (Fig. 1B). There was a larger range of myofibrillar architecture from very low to high in the iPSC-CM population than one would find, for example, in cultured NRVMs where all cells are beating and filled with striated myofibrils.

Contractility of iPSC-CMs.

The maximum shortening of iPSC-CMs was calculated with a kymograph (Fig. 2, A and B). At week 1, the average shortening and standard deviation for normal N1 was 0.89 ± 0.28 μm, and N2 was 1.09 ± 0.48, which were similar to each other. In the mutant M1 cells from the severely affected patient it was 0.47 ± 0.17 μm, which was significantly less than normal N1 (P < 0.01). However, the M2 mutant at week 1 was not different from the normal CMs and these cells were, therefore, not subjected to analysis for longer time periods. At week 2, the average shortening and standard deviation in the N1 normal cells was 0.81 ± 0.18 μm and in M1 mutant cells was 0.70 ± 0.23 μm, which was significantly higher than week 1 (P < 0.01) (Fig. 2C). Substrate stiffness (100 kPa) and surface area adhering to the substrate can be used to express shortening in terms of surface tension (5), where normal cells with greater shortening generate more force than mutant cells at week 1, but have the same force by week 2.

Fig. 2.

Measuring shortening in iPSC-CMs by line scans and kymographs. A: contractile motion was measured in a single iPSC-CM, via a line scan (black arrow). B: kymograph of cell displacements over 3 s with several contractions to illustrate how the maximum shortening was measured. Vertical white lines indicate the beginning and end of a contraction used to calculate beats per minute. Shortening time and distance were used to calculate shortening and velocity. C: normal (N) and mutant (M) iPSC-CMs were cultured for 2 wk. The shortening of normal iPSC-CMs (N1) on the softer (100 kPa) PDMS substrate was significantly greater than on the very stiff glass (61.9 GPa) (N1-G) at week 1 (**P < 0.01). At week 1 after plating on PDMS-100 kPa, normal iPSC-CMs (N1) have a greater maximum shortening than M1 mutant iPSC-CMs (**P < 0.01), but N2 normal and M2 mutant are similar. At 2 wk after plating on PDMS-100 kPa, the maximum shortening of the M1 mutant cells has increased significantly (**P < 0.01) and reached near normal levels. Normal and mutant values remain significantly different (*P < 0.05) at week 2 although much improved over week 1 (**P < 0.01). The number of cells (n) is given for each condition.

Substrate stiffness is known to influence cardiomyocyte shortening (5, 11, 30) and was verified with the iPSC-CMs by comparing normal iPSC-CMs plated on glass (61.9 GPa) and PDMS (100 kPa) (Fig. 2). In Tables 1 and 2, measurements of beats per minute, percentage of cells beating per dish, and shortening velocity are shown. There was wide variation from cell to cell in many parameters for all lines. Some significant differences are found between N1 and M1 cells at both week 1 and week 2, but cells from the more normal M2 patient were indistinguishable from N2 in all parameters.

Table 1.

iPSC-CM beating rate, percentage of cells beating, and shortening velocity at 1 and 2 wk after plating for N1 and M1

	Week 1		Week 2
	N1	M1	N1	M1
BPM	15.80 ± 3.63	10.77 ± 2.92a	14.87 ± 5.28	13.80 ± 4.59
N	15	13	15	15
% Cells beating	89.67 ± 13.77	10.35 ± 8.84b	77.87 ± 20.14	16.46 ± 7.98c
N	8	n = 8	4	4
Shortening velocity, μm/s	2.06 ± 0.65	0.93 ± 0.39d	1.77 ± 0.47	1.80 ± 0.56
N	42	36	42	40

Values are means ± SD.

M1 or M2, mutant cell line number; N1 or N2, normal cell line number; BPM, beats per minute.

^aP < 0.001 vs. N1 (week 1), % cells beating:

^bP < 0.001 vs. N1 (week 1);

^cP < 0.001 vs. N1 (week 2); shortening velocity:

^dP < 0.001 vs. N1 (week 1).

Table 2.

iPSC-CM beating rate, percentage of cells beating, and shortening velocity at 1 wk after plating for N2 and M2

	Week 1
	N2	M2
BPM	19.00 ± 9.39	19.00 ± 7.39
N	22	25
% Cells beating	35.25 ± 4.92	36.00 ± 5.66
N	4	6
Shortening velocity, μm/s	2.51 ± 1.13a	2.57 ± 0.73
N	42	42

Values are means ± SD.

^aP < 0.05 vs. N1 (week 1).

Profiles of sarcomeric proteins in iPSC-CMs.

To understand the structure/function relations in the iPSC-CMs of the DCM patient more clearly, we performed gel electrophoresis to determine the protein isoform populations (Fig. 3). No detectable cTnI isoform was identified by Western blot analysis of TnI (Fig. 3A) in iPSC-CMs, suggesting exclusive expression of the slow skeletal TnI (ssTnI) variant. ssTnI is predominant in the embryonic/early neonatal stage of heart development, as verified in the positive controls of developing mouse heart TnI (Fig. 3A). Substantial evidence indicates that sarcomeres controlled by ssTnI rather than cTnI are resistant to deactivation by acidic pH and highly sensitive to Ca²⁺ (49). In contrast to the case with TnI, the TnT isoform population (Fig. 3B) has a relatively high cTnT3, which is the adult human isoform (1). Owing to a limited number of cells, to perform further analysis we have relied on relative mobility to identify TnT isoforms, which we have employed extensively in past studies (20, 21). Moreover, previous studies (20) from our lab investigating relative expression thin filament isoforms in the developing rabbit heart indicated a lack of coordination of the developmental transition of TnI and TnT isoforms, which may be influenced by adrenergic stimulation and stress. After 1 wk in culture, the normal and the cTnT-R173W iPSC-CMs both expressed a relatively high percentage of β-MHC isoform (Fig. 3C). However, by 2 wk, whereas the mutant percentage remained about the same as at 1 wk, both the normal lines were mostly β-MHC, which is characteristic of the adult human heart (38). Compared with α-MHC, hearts regulated by β-MHC demonstrate slower cross-bridge kinetics and increased economy of contraction (unit ATPase rate/unit tension development) (39). Taken together, isoform expression reflects the fetal/neonatal stage of development of these hIPSC-CMs in culture.

Fig. 3.

iPSC-CM protein analysis. SDS-PAGE and Western blot analysis of iPSC-CMs. A: Western blot analysis of troponin I isoform expression. B: Western blot analysis of cardiac troponin T isoform expression. C: 6% SDS-PAGE of myosin heavy chain (MHC) isoforms. The predominant myosin heavy chain isoform in human adult cardiac tissue is the β isoform. Data represented are the means of two normal cell lines with standard deviation, and the M1 and M2 cell lines are separated. The 2-wk mutant M1 cell line for cTnT is not available in B. N, normal both cell lines combined; M1 or M2, mutant cell line number; N1 or N2, normal cell line number; MHC, myosin heavy chain; cTnT, cardiac troponin T; ssTnI, slow skeletal troponin I; cTnI, cardiac troponin I; D, day; wk, week.

Verification of the depressant effect of cTnT-R173W on myofilament activity.

Functional changes in the iPSC-CMs may be due in part to the altered sarcomeric isoform population (Fig. 3) as well as myosin ATPase activity (Fig. 4). We therefore thought it important to determine more clearly how the presence of cTnT-R173W per se affects sarcomere activity. It is generally found that DCM-inducing mutants in thin filament protein induce a depression in myofilament activity. To determine whether this is the case with cTnT-R173W, we expressed wild-type (wt) and mutant cTnT, and fully reconstituted a Ca²⁺-sensitive myofilament preparation. Inasmuch as the other sarcomeric proteins were the same in these experiments, we were able to compare the specific effects of control cTnT with mutant cTnT. We titrated the thin filaments with varying concentrations of Tn complex, achieving full relaxation at the expected ratio of 7actins:1Tm:1Tn (Fig. 4). There was a significantly depressed ATPase rate in the myofilaments reconstituted with the mutant cTnT-R173W compared with the controls.

Fig. 4.

In vitro ATPase from reconstituted myofilament preparations. Thin filaments were reconstituted with increasing concentrations of Tn complex containing either wild-type (cTnT-WT) or mutant TnT-R173W with a 7:1 ratio of actin:tropomyosin in the presence or absence of Ca²⁺. Following incubation with S1 myosin and 1 mM ATP, the rate of ATPase activity was determined via malachite green assay and was f ound to be decreased in TnT-R173W mutants compared with TnT-WT at the expected stoichiometric concentration of actin:tropomyosin:Tn complex (7:1:1), with a similar effect at higher concentrations (7:1:2). *P < 0.05 compared with TnT-WT.

Omecamtiv mecarbil control studies on NRVMs.

NRVMs were used as a control to study the effect of OM on shortening (Fig. 5). When untreated, the average shortening was 0.95 ± 0.10 μm (means ± SE), whereas it increased to 1.65 ± 0.32 μm with 200 nM, and to 2.04 ± 0.15 μm with 500 nM OM. The maximum shortening of NRVM was significantly higher at the 500 nM dose compared with untreated NRVM (P < 0.01), but not different with the 200 nM dose. Actin FRAP dynamics in NRVM were also measured at the different OM concentrations (Fig. 5). The K_FRAP for the untreated group was 2.69 ± 0.02 × 10⁻⁴ s⁻¹ and for the 200 nM treatment was 3.65 ± 1.10 × 10⁻⁴ s⁻¹. With the 500 nM OM treatment, the actin dynamics were significantly higher (7.10 ± 1.3 × 10⁻⁴ s⁻¹) than the untreated group (P < 0.01).

Fig. 5.

Shortening and actin FRAP for neonatal rat ventricular myocytes treated with omecamtiv mecarbil (OM). NRVMs served as a control cells to test for OM activity and efficacy in culture. The 200 nm OM dose had no effect on NRVM. A: shortening of NRVMs is significantly increased when treated at a 500 nM dose of OM (***P < 0.001). B: actin dynamics are significantly increased in NRVMs with a 500 nM dose of OM (***P < 0.001).

Actin content in iPSC-CM with omecamtiv mecarbil treatment.

Representative images are shown for week 1 and week 2 (Fig. 1) in normal and mutant cells and when treated with 500 nM OM (Fig. 6). Given the heterogeneity of iPSC-CMs, an unbiased measure of actin content was necessary. The content of assembled F-actin was measured semiquantitatively per cell by fraction the fluorescent pixels at week 1. There were 45 and 4% striated CMs for N1 and M1 (Fig. 6B) and for 47 and 23% striated CMs for N2 and M2 (Fig. 6C), respectively. The fraction of striated CM volume filled by F-actin (Fig. 6D) was significantly higher in N1 (0.36 ± 0.14) than in M1 mutant cells (0.18 ± 0.04) (means ± SD, P < 0.01). However, the F-actin fraction per cell with 1 wk of exposure to 500 nM OM the M1 mutant cells increased significantly to 0.25 ± 0.04 compared with the untreated M1 0.18 ± 0.04 (P < 0.01). The M2 mutant CMs did not change their actin content with OM treatment (Fig. 6E).

Fig. 6.

F-actin content of normal and mutant cells when treated with omecamtiv mecarbil. A: both normal (N1) and mutant (M1) iPSC-CMs were fixed and stained for F-actin (phalloidin, red), nucleus (DAPI, blue), and sarcomeric α-actinin (antibody, green) to show striations. Scale bar = 50 μm at lower magnification and enlarged in inset. After 1 wk, normal iPSC-CMs grown on the 100-kPa PDMS exhibited striated myofibrils, compared with the more weakly striated and meshlike architecture from the M1 mutant cells. OM treatment with either 200 or 500 mM for a week did not alter the cytoarchtecture. B and C: percentage of striated cells for normal N1 and N2 vs. mutant M1 and M2, with or without 500-nM treatment. Ten randomly selected images were analyzed with 10–30 cells per image, n = 10 images for all conditions. D and E: the F-actin content per striated CM was assessed semiquantitatively by using the phalloidin stain. Normal (N1) cells have a higher fraction of assembled F-actin than mutant (M1) cells at week 1 (**P < 0.01), and with OM treatment (500 nM) there was a significant increase (**P < 0.01) with treatment compared with untreated M1 cells. No differences are noted for the fraction of assembled F-actin in the M2 line in E. The number of cells (n) is given for each condition.

Effects of omecamtiv mecarbil on actin FRAP dynamics and shortening.

To address potential therapeutic effects of OM on actin FRAP dynamics and shortening, iPSC-CMs were treated with 200 or 500 nM OM, which had no effect on cell viability over a 2-wk period. The dynamics of sarcomeric actin were measured at week 1 via FRAP from cells infected with actin-GFP fusion virus to visualize the striated sarcomeres and quantify the recovery profile of the fluorescence (Fig. 7A). The kinetic constant (K_FRAP) obtained from curve fitting the raw data was 5.40 ± 3.00 × 10⁻⁴ s⁻¹ for the N1 normal cells and 2.42 ± 1.06 × 10⁻⁴ s⁻¹ for M1 mutant cells at week 1 (**P < 0.01) (Fig. 7B). However, no significant differences resulted from treatment of these M1 cells with 500 nM OM. K_FRAP for normal N2 and mutant M2 cells were not different.

Fig. 7.

Effect of omecamtiv mecarbil on actin FRAP dynamics and shortening in normal and mutant iPSC-CMs. A: sarcomeres, visualized by confocal microscopy with actin-GFP (white), are found throughout the cell. White box denotes region of interest (ROI) (area = 3.75 × 3.75 μm) of thin filaments in which FRAP was analyzed; scale bar = 20 μm. B: at week 1, actin dynamics were significantly higher between normal and mutant M1 CMs (**P < 0.01), but similar for the M2 mutants CMs. However, treatment with the 500-nM dose of OM showed no gain in function for the N1 or M1 cell lines. C: the effect of 200- and 500-nM doses is shown on shortening for normal (N1) and mutant (M1) iPSC-CMs after 1 wk of treatment. Maximum shortening did not alter with either 200 or 500 nM in N1 cells compared with untreated cells. However, there was significantly greater shortening in M1 mutant cells when treated with 200 nM OM (*P < 0.05) and ∼50% higher with the 500 nM dose compared with untreated M1 cells (**P < 0.01). The number of cells (n) is given for each condition.

Maximum shortening with standard deviation was assessed on N1 and M1 CMs to test the effects of OM (Fig. 7C). OM did not affect the normal CMs, but differences were found with a 1-wk treatment of 200 nM OM between N1 normal cells (0.96 ± 0.25 μm) and M1 mutant cells (0.59 ± 0.25 μm) (means ± SD, P < 0.05). At the 500 nM dose of OM, maximum shortening did not significantly increase in the N1 CMs (0.96 ± 0.37 μm) compared with untreated N1 CMs (0.89 ± 0.28 μm). However, maximum shortening increased very significantly (P < 0.01) for M1 mutant cells (0.74 ± 0.21 μm), which is ∼50% higher than the untreated M1 cell (0.47 ± 0.17 μm).

DISCUSSION

Data reported here are the first to demonstrate a linkage between expression of a DCM-inducing thin filament mutation and altered thin filament assembly in iPSC-CMs from a severely affected patient. Thus the novel concept is presented for pathogenicity of cTnT mutations on sarcomere assembly as well as contractility, rather than solely by affecting contractile parameters. Electron micrographs of iPSC-CMs with the cTnT R173W mutation showed myofibrillar disruption (44), but a mechanistic association with the modification in TnT remained unknown. In studies reported here, the structure, actin content, and function of cTnT R173W mutation iPSC-CMs from the severely afflicted family member matured at a slower rate than the normal iPSC-CMs, with altered actin FRAP dynamics and reduced sarcomeric assembly. These findings indicate sarcomeric assembly is slower in the iPSC-CMs from a severely afflicted patient and serve as evidence to support the hypothesis that the link between a TnT mutation and sarcomere assembly is a maladaptive alteration in mechanical properties in the cell via mechanosignal transduction pathways (40, 41). The epigenetic effects of the severity of the disease in this family were noted recently (50) and were also seen here with respect to contractility and actin content.

Previous studies (15) reported a failure in sarcomere assembly when TnT activity was disrupted during zebra fish development. Those investigators concluded that TnT was important for both regulation of the actin-myosin interaction and myofibrillogenesis (15). However, to the best of our knowledge, this concept has not been identified in the case of DCM in human hearts or iPSC-CMs. Thus our finding suggests a new problem that may contribute into disease progression in familial DCM. The presence of both depressed sarcomeric organization and muscle performance has been found in other models, such as DCM induced by loss of MLP (2) or by loss of nexillin (21).

Our data also are the first to show that a sarcomere activator, OM, acting via thick filament myosin improves sarcomere shortening that had been depressed because of the cTnT-R173W mutation. An important aspect of our choice of this particular agent is that it has gone through safety trials in humans and has been reported to improve function in acquired heart failure. In addition, there is no evidence for off-target effects of OM as an inhibitor of phosphodiesterase III (PDE III) or on alterations in Ca²⁺ fluxes (31, 42). Recent studies from our labs have reported proof of principal that OM is able to increase tension and ATPase rate of sarcomeres from a DCM model with a mutation in the thin filament protein Tm (46). Indeed, with 1 wk of culture, the chronic treatment with the 500-nM dose of OM increased about 50% iPSC-CM shortening with additional gains in actin accumulation over untreated mutant iPSC-CMs, but we found no detectable changes on actin dynamics with OM. These results suggest that OM is effective in improving function by promoting the actin-myosin interaction in normal cells, but not effective in overcoming the depressive affects the mutant TnT has on actin dynamics. TnT can reportedly contact and generate molecular swivel for the successive tropomyosin to interact equivalently with the actin filament (35). Also, tropomyosin is well-known for providing stability during actin assembly (8). Thus other structural dysfunctions of TnT might lead to delayed abnormal actin dynamics and assembly rate, which was also seen in this study.

An important novel aspect of our findings is data analyzing sarcomeric protein composition in the iPSC-CMs. Comparisons of sarcomeric protein isoform composition have not been generally assessed, and our studies were limited by the amount of protein available in the cell preparations. Yet, an important finding is the persistence of expression of ssTnI, the fetal/neonatal form of TnI. We have previously reported that variations in N-terminal TnT isoforms produce different effects on sarcomere Ca-responsiveness in myofilaments controlled by cTnI and ssTnI (20). It is interesting that in contrast to the case with TnI, the adult isoform cTnT3 is the most abundant isoform population. Myosin isoforms also more closely resembled the adult human heart (predominantly β-MHC) in normal cells and at 2 wk were ∼95% β-MHC, but the β-MHC isoform was about half of what was present in the normal cells at 2 wk (36, 43). Moreover, the lack of effect of OM on actin dynamics in the cells expressing the mutant TnT might be partly explained by a difference in MHC isoform population.

Moreover, In contrast to cases with normal NRVMs, there as an intriguing lack of effect of OM on iPSC shortening. This difference is not likely to be due to differences in ssTnI, which is nearly identical in rat and human hearts. Figure 3 shows that there is a preponderance of the adult isoform of cTnT in the iPSCs, which would not be the case in the NRVM. However, we do not have a clear mechanism for why OM did not induce an increase in shortening in normal IPSC.

The sarcomere protein composition directly corresponded with the function of human iPSC-CMs, which differed after plating on physiologically soft substrates (100 kPa PDMS) between normal human iPSC-CMs and those derived from a DCM patient. The freeze-thaw process itself delivers a major stress to iPSC-CMs (28), but this was common to all cells. At 1 wk after plating, the M1 mutant cells had significantly less actin, poorly organized myofibrillar structure, and less ability to shorten than the normal cells. By 2 wk after plating, the mutant cells were almost indistinguishable from the normal cells. Interestingly, these findings correlate with the reduction in actin assembly dynamics known to regulate myofibril growth (26, 27, 29). However, the structural and functional maturation of the CMs from the severely afflicted cTnT R173W was delayed by several days, demonstrating a reduced ability of adaptive response to an altered mechanical environment. Interestingly, the less affected family member, M2, were similar to normal cells in most regards.

The variation in functional performance of the different cell lines was readily detected by plating cells on material with a more physiologic stiffness. Isolated embryonic cardiomyocytes, for example, cultured over time on very hard surfaces lack striated myofibrils and stop beating; very soft surfaces (under 10 kPa) did not last long in culture. But at the physiologic stiffness (10–30 kPa), myofibrils were robust and contracted well over time (14, 22). Plating the iPSC-CMs on PDMS substrates with a stiffness of 100 kPa was physiologically similar to heart tissue (3, 5, 12), which is in the higher end of the normal physiologic range, but was sufficiently soft to permit excellent beating and sarcomere formation. Stiffness varies with age, species, disease state, and the method of measurement. Muscle stiffness is in the 10-kPa range for embryonic/neonatal myocytes (4, 12), but is over 70 kPa in the normal adult rat (51). Infarct stiffness and collagen content increase with time so that by 6 wk postinfarct it may rise to 400 kPa (16). Furthermore, stiffness of muscle is dynamic with an increase when muscles contract (3, 4, 30). Therefore, 100 kPa is a useful stiffness to mimic normal adult heart that may aid in the maturation of hIPSC-CMs. Note that in previous studies these normal and cTnT R173W iPSC-CMs were grown on very stiff glass substrates (61.9 GPa) (44), which may have masked the subtleties of sarcomere assembly characteristics as found in these studies. Moreover, previous studies did not measure actual shortening properties of the myofibrils, but used the indirect method of atomic force microscopy to indent the upper surface of the cell and deduce myofibrillar force using certain assumptions (44).

In summary, cardiac myocytes were derived from induced pluripotent stem cells from normal and family members expressing a mutant cardiac troponin T linked to dilated cardiomyopathy. Our studies provide new evidence relating to the structure/function relations of patient-specific iPSC-CMs and most likely applicable to other forms of acquired and familial DCM. Shortening, actin content, and assembly dynamics were depressed in the severely affected mutant but reversed by a myosin activation reagent. Our data indicate that the alterations noted in these cells are due to a combination of modifications triggered by the cTnT-R173W in a patient with DCM, which as we demonstrated depresses the actin-myosin interaction. Moreover, there are multiplex alterations, including the stress of cell processing, altered sarcomere isoform populations, altered expression of sarcomeric proteins, and altered thin filament assembly. Our data also emphasize the important role of depressed mechanics in inducing progression to DCM, and provide evidence of the usefulness of iPSC-CMs in testing the effects of realistic therapies enhancing these mechanics via direct interactions with the sarcomeres. As iPSC-CMs serve as an important means to model disease and pharmacological intervention, research to characterize and fully derive iPSCs into adult myocytes remains ongoing. The use of bioengineering approaches, such as modified substrate stiffness, provides a means to continue studies in a physiologic manner allowing for the detection of subtle differences in cellular development and functional performance. The future of iPSCs remains a valuable means to develop individual healthcare studies both for mimicking disease and determining therapeutic intervention.

GRANTS

Support to conduct this research was provided by National Heart, Lung, and Blood Institute Grant PO HL-62426 (to R. Solaro and B. Russell), Proteomics and Analytical Biochemistry Core (to C. Warren), National Heart, Lung, and Blood Institute Grant T32 HL-07692 (to K. Broughton and M. Henze), and American Heart Association Predoctoral Fellowship 12PRE12050371 (to Y.-H. Lin).

DISCLOSURES

R. J. Solaro serves on the Scientific Advisory Board of Cytokinetics, Inc.

AUTHOR CONTRIBUTIONS

K.M.B., Y.-H.L., R.J.S., and B.R. conception and design of research; K.M.B., J.L., E.S., C.M.W., Y.-H.L., M.P.H., and V.S.-F. performed experiments; K.M.B., J.L., E.S., C.M.W., Y.-H.L., and M.P.H. analyzed data; K.M.B., J.L., E.S., C.M.W., Y.-H.L., R.J.S., and B.R. interpreted results of experiments; K.M.B, J.L., E.S., and Y.-H.L. prepared figures K.M.B., R.J.S., and B.R. drafted manuscript; K.M.B., J.L., E.S., C.M.W., Y.-H.L., M.P.H., V.S.-F., R.J.S., and B.R. edited and revised manuscript; B.R. approved final version of manuscript.