- Email: [email protected]
Spiral reentry waves in confluent layer of HL-1 cardiomyocyte cell lines
Biochemical and Biophysical Research Communications 377 (2008) 1269–1273
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Spiral reentry waves in confluent layer of HL-1 cardiomyocyte cell lines Jin Hee Hong, Joon Ho Choi, Tae Yun Kim, Kyoung J. Lee * Center for Cell Dynamics and Department of Physics, Korea University, Anam-Dong 5-1, Sungbuk-Gu, Seoul 136-713, Republic of Korea
a r t i c l e
i n f o
Article history: Received 20 October 2008 Available online 8 November 2008
Keywords: Arrhythmia Cardiac reentries Spiral waves HL-1 cardiomyocytes Mechanical contraction Mouse atrial cells Multi-electrode array (MEA) Inter-beat-interval (IBI) Myolysis Optical mapping Myosin
a b s t r a c t Cardiac excitation waves that arise in heart tissues have long been an important research topic because they are related to various cardiac arrhythmia. Investigating their properties based on intact animal whole hearts is important but quite demanding and expensive. Subsequently, dissociated cardiac cell cultures have been used as an alternative. Here, we access the usefulness of cardiomyocyte cell line HL-1 in studying generic properties of cardiac waves. Spontaneous wave activities in confluent populations of HL1 cells are monitored using a phase-contrast optical mapping system and a microelectrode array recording device. We find that high-density cultures of HL-1 cells can support well-defined reentries. Their conduction velocity and rotation period both increase over few days. The increasing trend of rotation period is opposite to the case of control experiments using primary cultures of mouse atrial cells. The progressive myolysis of HL-1 seems responsible for this difference. Ó 2008 Elsevier Inc. All rights reserved.
Many cardiac arrhythmias are known to be driven by spiral reentries (rotating spiral wave), and different spatiotemporal dynamics of them underlie a different class of cardiac arrhythmia [1–2]. For example, the steady rotation of a single reentry has been associated with monomorphic tachycardia [3] and a non-stationary meandering reentry is believed to cause polymorphic tachycardia [4]. Currently, there is a strong consensus in the medical as well as scientific community that the chaotic wave states following a break-up of regular reentries underlie cardiac fibrillations [5]. Subsequently, there has been a great deal of effort for understanding various aspects of cardiac reentries and their instabilities. The related experimental investigations are, however, a very demanding task, particularly so, when they involve whole heart preparation of a large animal. First of all, it is difficult to keep the whole heart under investigation viable and controlled for more than a day. In other words, long-term investigations lasting several days are not practical if not accessible. Second, the anatomical complexity of animal whole heart often prevents one to make any easy interpretation on the observed phenomena. With these difficulties, various techniques have been developed for primary cultures of dissociated cardiac cells and they have been used very successfully as an alternative to the whole heart, especially when the fundamental properties of cardiac reentries were to be examined [6–9]. One can easily and actively manipulate the density, * Corresponding author. Fax: +82 2 3290 3534. E-mail address: [email protected] (K.J. Lee). 0006-291X/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2008.10.168
composition, and the structure of cultured tissues. Thus, the precise experimental quantification of cardiac waves following a systematic parameter change is possible and the result can be compared with those of analytical and/or computational studies. Several research groups have recently demonstrated the existence of reentries in primary cultures of dissociated ventricular cells of chicks [10] and rats [11–14], and studied their various properties. A natural extension of these efforts is to use cardiac cell lines. Several years ago, Claycomb et al. developed a cardiomyocyte cell line termed HL-1 [15], and it is still the only cardiomyocyte cell line currently available. The HL-1 cells are derived from the mouse atrial cardiomyocyte tumor lineage AT-1. They continuously divide and spontaneously contract under some suitable conditions. Many different aspects of HL-1 cells have been studied based on genetic, immunohistochemical, pharmacological, and electrophysiological techniques [16]. However, these studies were designed mainly to answer questions pertaining to cardiac biology at cellular or subcellular levels. We view confluent populations of HL-1 cells as a potentially useful model cardiac tissue for investigating generic features of cardiac spiral reentries. Here, we demonstrate that they can exhibit cardiac reentries just like primary cultures of cardiac cells, and we quantify their long-term wave properties, namely, conduction velocity and rotation period, based on our home-built phase-contrast macroscope and micro multi-electrode array (MEA) plates. These wave properties are compared with those of primary cultures of mouse atrial cells. Interestingly, the rotation period increases in the HL-1
1270
J.H. Hong et al. / Biochemical and Biophysical Research Communications 377 (2008) 1269–1273
cultures while it decreases in the primary cultures. The amount of motor protein, myosin, in Hl-1 cell cultures is quantified at several different stages of the HL-1 culture as a first step for elucidating the observed discrepancy. Materials and methods HL-1 cell line cultures. The HL-1 cell line (passage no. 38) was obtained from Dr. William Claycomb (Louisiana State University, USA). The cells were cultured in a specially formulated growth medium (Claycomb medium, JRH Biosciences, USA) supplemented with 100 l M norepinephrine, 2 mM L-glutamine (Gibco, USA) and 10% FBS (Life Technologies, USA). They were grown in T-flasks coated with 12.5 lg/ml fibronectin (Sigma) and 0.02% gelatin (Sigma) and kept in a 5% CO2 atmosphere at 37 °C. When a HL-1 cell culture flask became confluent, it was split into three batches, and each of which was then considered as the next passage. The culture medium was changed at every 24–48 h. The initial cell plating density was in the range of 0.5–1.4 103 cells/mm2. Culture samples with an initial plating density lower than 0.5 103 cells/mm2 exhibit only small noisy wave segments that do not develop into a stable reentry. Those with an initial plating density higher than 1.4 103 cells/mm2 quickly overpopulate the dish and become unhealthy. Primary cultures of dissociated mouse atrial cells. The atria of 3- to 4-day-old C57BL/6J mice were dissociated enzymatically at 37 °C using 0.05% trypsin (Gibco) and Pancreatin (Sigma). The cells were resuspended in Medium199 (Gibco) supplemented with 1 lM vitamin B12, 1 mM L-glutamine, 100 U/ml penicillin/streptomycin antibiotic mixtures, 5% HS (Gibco) and 10% FBS. The cells were either plated onto MEA plates and (or) dishes that were coated with 12.5 lg/ml fibronectin and 0.02% gelatin. The plating density was kept at 2.0 103 cells/mm2 for all experiments. The culture medium was changed once every 48 h. All the culture procedures abide
by the Guide for the Care and Use of Laboratory Animals published Korea University, Collage of Medicine, Animal Research Policy Committee. Optical mapping of mechanical contractile waves and image analysis. A home-built phase-contrast macroscope along with a custom-built incubation chamber was used to optically map the mechanical contractile activities of cultured cardiac tissues. The construction of the phase-contrast macroscope and the chamber was described in our earlier publication [11]. The imaging (30 frames per second, 512 512 pixels) is based on a ‘‘propagationinduced phase contrast optics” detecting minute shape changes of activated cardiac cells [11–14]. This imaging method does not rely on any extrinsic fluorescent dyes, thus, non-invasive to the samples under investigation and uninterrupted long-term observations over several days are possible. Multi-electrode array (MEA) recordings. The MEA plates were either purchased from Panasonic (MED-P210A) or fabricated at the micro fabrication facilities at Korea University. The construction of home-fabricated MEA plates was described in our earlier publication [17]. The electrical signals picked up by 32 micro-electrodes were amplified 10,000 times (by 10 pre- and 1000 filteramplifier) and band-pass filtered (100–10 kHz). The field potential peaks that had a maximum 4.5–6.0 times (depending on the signal to noise ratio) larger than the standard deviation for 10 min duration of recording were identified as an action potential driven excitation. Based on these activation events, the sequence of interbeat-interval (IBI) was built. The MEA recording system provides a far better temporal resolution (10 kHz) than the phase imaging (30 Hz) but at the expanse of much lower spatial resolution. Immunostaining and quantification for cardiac myosin. HL-1 cardiomyocytes were fixed for 15 min at room temperature in 4% paraformaldehyde in PBS, permeabilized in PBS containing 0.1% Triton X-100, and blocked in 5% BSA for 1 h at room temperature. Then, the permeabilized cells were incubated overnight at 4 °C
Fig. 1. Spiral reentries in population of HL-1 cells (A) and dissociated atrial cells (C). The corresponding space–time plots shown below are obtained along the segment a–b as marked. The long-term evolutions of the rotation period (IBI) and conduction velocity (CV) of spiral reentries for HL-1 cells and dissociated atrial cells are given in (B) and (D), respectively. Images shown in (A) are taken 120 h after the initial plating (projected cell density at t = 120 h is 2.0 103 cells/mm2). The initial cell plating density is 1.0 103 cells/mm2 for the HL-1 cell culture (B) and 2.0 103 cells/mm2 for the atrial primary culture (C,D). In (B), the discontinuity around t = 90 h is due to the fact that the spiral reentry has disappeared briefly and reemerged there. In (D), the error bars are large during the early stage since several competing reentries coexist and their optical signals are rather small. The scale bars are 3 mm. The error bars in (B) and (D) represent spatial fluctuation.
J.H. Hong et al. / Biochemical and Biophysical Research Communications 377 (2008) 1269–1273
1271
Fig. 2. Long-term IBI trends of HL-1 reentries. (A) a schematic diagram of MEA recording system, (B) a typical time trace acquired through a MEA electrode, and (C–E) increasing mean IBI curves of HL-1 reentry for three different initial plating densities [0.7 103 cells/mm2 (C),1.0 103 cells/mm2 (D), and1.4 103 cells/mm2 (E)]. The arrows in (C–E) mark the point of propagation failure. Standard error (66.3 ms) in space domain is smaller than the size of the data symbols in Fig. 2C–E. (F) HL-1 cell population growth assay. HL-1 cells were stained with 4% tryphan blue for counting (n = 5). The error bars represent a standard error.
with a mouse monoclonal antibody against cardiac myosin heavy chain (ab15, abcam, UK) diluted 1:400 in blocking buffer. Fluorescein-labeled isothiocyanate (FITC) anti-mouse (abcam) was used as secondary antibody. The nuclei were stained by 2 lg /ml propidium iodide (PI). Images of FITC and PI fluorescence were obtained by using a fluorescence microscope (IX71, Olympus, Tokyo, Japan) and a confocal laser-scanning microscope (BX51WI, FV 500, Olympus). Results Reentries supported by confluent cultures of HL-1 cell lines It has been known that HL-1 cells can contract spontaneously [15]. However, we find that the contractibility of HL-1 cells depends very much on the developmental stage of the culture, not to mention the cell population density. Especially in the early stage after the initial cell plating, the majority of HL-1 cells do not contract spontaneously. When the culture becomes 1–2 days old, a few cells that are scattered randomly in the populations start to
beat, but rather weakly at 1–3 Hz. Their minute contractile motion is discernable only when inspected under a microscope. As the population density increases and culture becomes more matured in time, the localized activation sites recruit more neighboring cells and expand their territories progressively. The majority of HL-1 cells contracts regularly by the time the culture medium reaches around 3 days in vitro, at this stage they support well-defined rotating spiral waves as shown in Fig. 1A (S1). As far as we understand, this is the first spiral reentry in populations of HL-1 cells ever produced and documented. In other words, HL-1 cells can constitute an excitable tissue that allows wave conduction and reentries. The reentry shown in Fig. 1A rotates quite stably in a periodic manner having a frequency of 3 Hz approximately, as well reflected in the accompanying spacetime plot. The uniformity of the culture medium can be inferred from the regularity of the pitch of the reentry. The reentries in HL-1 cell cultures can last for quite some time under the culture condition described in Methods. Fig. 1B illustrates a typical long-term behavior in terms of their rotation period and conduction velocity. The overall trend is that both increase
1272
J.H. Hong et al. / Biochemical and Biophysical Research Communications 377 (2008) 1269–1273
gradually as the culture gets older. The fact that the rotation period gets longer in time is rather unusual and opposite to our earlier observations made with primary cultures of rat ventricle cells, in which the period of reentries gets shorter progressively over several days [11]. The conduction velocity, on the other hand, shows a similar increasing trend in both systems. Reentries supported by confluent primary cultures of mouse atria As for a comparison, cardiac waves in primary cultures of dissociated mouse atrial cells are also investigated, because HL-1 cell line is derived from mouse atria. Neonatal mouse atria are rather small that many (14–16) of them are needed to form a confluent tissue in a 35 mm dish. The confluent atrial cell cultures also support spiral reentries, and one typical example is given in Fig. 1C (S2). The particular reentry conducts about six times faster and has 2–3 times shorter IBI than the one shown in Fig. 1A. Since the pitch of a reentry is the product of IBI and CV, it is conceivable that the pitch of the reentry wave in the primary cell culture is about 2–3 times larger than that of the HL-1 cell line tissue. The snapshot images in Fig. 1A and 1C are consistent in this regard. One dramatic difference exists between the primary culture of mouse atrial cells and the culture of HL-1 cells; just as in the primary culture of ventricle cells [11], the IBI of the reentries observed in the cultures of atrial cells (Fig. 1D) decreases in time. Thus, the increasing trend of IBI associated with the HL-1 reentries is rather an unusual property, and our subsequent investigation is aimed for explaining this long-term property of HL-1 cell populations.
Long-term trend of HL-1 spiral reentries: initial plating density dependence The MEA electrode recording system (Fig. 2A) is highly sensitive to the local activity and provides a high temporal resolution (10 kHz), thus, it suits better for monitoring a long-term trend of IBI than the optical imaging, including the very early stage of HL1 culture when the contraction of cells is very minute. Several different IBI temporal sequences of HL-1 reentries are obtained, and a typical time series acquired through an electrode is given in Fig. 2B. The sensitivity of the micro-electrodes is indeed far better than that of the optical measurement; the action potential events can be detected as early as 2 days in vitro (about 24 h in advance to the point when the optical imaging is able to detect the motion signals; compare Fig. 2D with Fig. 1B) under the same culture condition. It is surprising that, even during this early period, the HL-1 cell culture supports a very regular sequence of beats. Pacemaker cells usually beat with a much lower frequency and non-periodically that the observed periodic sequences of IBIs must have its origin at self-rotating reentries. Fig. 2C reconfirms the long-term trend of IBI shown in Fig. 1B. Moreover, the long-term trend does not depend on the initial cell density (Fig. 2C–E); in all cases, the increase is gradual at the initial stage but sharply rising at the later stage. Although not exhaustively tested, we find that when the initial cell density is low, the onset of wave initiation and the lengthening of IBI come at a delayed time. The time constants associated with the rising IBI curves corresponding to different initial cell densities, however, do not vary significantly [s = 6.9 ± 1.4 h (n = 3) for 1.0 103 cells/mm2, s = 6.3 h for 1.4 103 cells/mm2, and s = 6.6 h for 0.7 103 cells/ mm2]. When the initial plating density is 1.0 103 cells/mm2 (Fig. 2D), the IBI lengthening begins around t = 96 h. Based on our HL-1 cell counting assay (Fig. 2F), the estimated cell density at this point is about 2.8 103 cells/mm2. Myolysis in HL-1 cardiomyocytes
Fig. 3. Myolysis in proliferating HL-1 cells. (A) Superimposed images of myosin immunostaining (green) and PI stained nuclei (red). Some cells that have undergone myolysis or membrane-blebbing are marked by arrows and arrowhead, respectively. The scale bar is 25 lm. (B) Myosin-expressing (green) area fraction vs. time. The initial cell density is 1.0 103 cells/mm2. Between 48 and 72 h (*** mark) after the initial plating, the area fraction reduces more than 50% of its peak value (p < 0.001; n = 5). Five fluorescent images (512 512 pixels) were obtained and analyzed for each data points. A green scale value of 80 (out of total 255 dynamic range) was chosen as a threshold for defining the myosin rich area. All error bars represent a standard error.
As the Hl-1 culture gets older, two obvious changes take place: first of all, the cell density increases (approximately, doubles at every 48 h, see Fig. 2F), and second of all, the connectivity mediated by gap junctions increases (not shown). These two factors, however, should work to increase the wave conduction velocity and to shorten the IBI. Primary cultures of atrial as well as ventricular cells of rats and mice both support this conclusion [11]. Therefore, the consistently observed long-term trend of lengthening IBI of HL-1 reentries points to a possibility that the HL-1 cells might have changed their cellular properties during the observation periods. Subsequently, we have used cardiac myosin heavy chain antibodies and PI to access the possible occurrence of myolysis and cell degeneration. Beyond 48 h after the plating, the immunostaining clearly reveals progressive myolysis and membrane blebbing in the HL-1 cells (Fig. 3A). The number of nuclei increases steadily, while the number of functional myocytes having a well-developed network of cardiac myosins clearly decreases in time except for the very early stage of the culture when there is no significant wave activity (see Fig. 3B); here, we note that the reentry wave activities shown in Fig. 2D also begin around 48 h in vitro.
Discussion The present study demonstrates the existence of reentry waves in confluent two-dimensional culture of HL-1 cells, and this observation suggests a potential use of HL-1 cells for studying spiral reentry wave dynamics and its instability to spatiotemporal chaos
J.H. Hong et al. / Biochemical and Biophysical Research Communications 377 (2008) 1269–1273
akin to cardiac fibrillation. Unfortunately, however, our study finds that the HL-1 cells go through the process of myolysis in the long run. Thus, HL-1 cell populations may be useful only for studying some short-term (<3 days) properties of cardiac reentries or their instabilities. The degradation process appears similar to that of the myocytes in the atria that has experienced an atrial fibrillation [18]. Incidentally, the recent study of Brundel et al. shows that HL-1 cardiomyocytes can undergo a cell remodeling within 24 h when they are subjected to high-frequency electrical field stimulations [19]. So, the self-rotating high-frequency reentries supported by the HL-1 culture may have elicited the HL-1 cell myolysis. From a clinical point of view, spiral reentry waves are a pathological dynamic state, and as such, the cellular remodeling caused by reentries seems to be a plausible hypothesis. On the other hand, primary cultures of dissociated atrial as well as ventricular cells, which had supported spiral reentries, show no indication of myolysis during the same term of observation (not shown). Therefore, an alternative explanation for the HL-1 degradation can simply be that they get deficient in their motor system because the cells have been regenerated many times over a short duration. Acknowledgments This work was supported by Acceleration Project (R17-2007017-01000-0) of the Korea Ministry of Science and Technology. We thank E. Entcheva and F. Fenton for their suggestions and comments. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bbrc.2008.10.168. References [1] J.M. Davidenko, A.M. Pertsov, R. Salomonsz, W.T. Baxter, J. Jalife, Stationary and drifting spiral waves of excitation in isolated cardiac muscle, Nature 355 (1992) 349–351.
1273
[2] A.T. Winfree, Electrical instability in cardiac muscle: phase singularities and rotors, J. Theor. Biol. 138 (1989) 353–405. [3] A. Garfinkel, Z. Qu, Nonlinear dynamics of excitation and propagation in cardiac tissue, in: D.P. Zipes, J. Jalife (Eds.), Cardiac Electrophysiology. From Cell to Bedside, W.B. Saunders Company, Philadelphia, 1999, pp. 315–320. [4] R.A. Gray, J. Jalife, A.V. Panfilov, W.T. Baxter, C. Cabo, A.M. Pertsov, Nonstationary vortex-like reentrant activity as a mechanism of polymorphic ventricular tachycardia in the isolated rabbit heart, Circulation 91 (1995) 2454–2469. [5] J.A. Abildskov, R.L. Lux, The mechanism of simulated Torsade de Pointes in a computer model of propagated excitation, J. Cardiovasc. Electr. 2 (1991) 224– 237. [6] L. Tung, Y. Zhang, Optical imaging of arrhythmias in tissue culture, J. Electrocardiol. 39 (2006) S2–S6. [7] G. Bub, N. El-Sherif, Monolayer cell cultures as model systems for studying paroxysmal atrial fibrillation, J. Electrocardiol. 37 (2004) 44–46. [8] N. Bursac, F. Aguel, L. Tung, Multiarm spirals in a two-dimensional cardiac substrate, P. Natl. Acad. Sci. USA 101 (2004) 15530–15534. [9] X.L. Tian, Y. Cheng, T. Zhang, M.L. Liao, S.L. Yong, Q.K. Wang, Optical mapping of ventricular arrhythmias in LQTS mice with SCN5A mutation N1325S, Biochem. Biophys. Res. Commun. 352 (2007) 879–883. [10] H. Gonzalez, Y. Nagai, G. Bub, L. Glass, A. Shrier, Reentrant waves in a ring of embryonic chick ventricular cells imaged with a Ca2+ sensitive dye, Biosystems 71 (2003) 71–80. [11] S.M. Hwang, K. Yea, K.J. Lee, Regular and alternant spiral waves of contractile motion on rat ventricle cell cultures, Phys. Rev. Lett. 92 (2004) 198103. [12] S.M. Hwang, T.Y. Kim, K.J. Lee, Complex-periodic spiral waves in confluent cardiac cell cultures induced by localized inhomogeneities, Proc. Natl. Acad. Sci. USA 102 (2005) 10363–10368. [13] T.Y. Kim, S.J. Woo, S.M. Hwang, J.H. Hong, K.J. Lee, Cardiac beat-to-beat alternations driven by unusual spiral waves, Proc. Natl. Acad. Sci. USA 104 (2007) 11639–11642. [14] S.J. Woo, J.H. Hong, T.Y. Kim, B.W. Bae, K.J. Lee, Spiral wave drift and complexoscillatory spiral waves caused by heterogeneities in two-dimensional in vitro cardiac tissues, New J. Phys. 10 (2008) 015005. [15] W.C. Claycomb, N.A. Lanson, B.S. Stallworth Jr., D.B. Egeland, J.B. Delcarpio, A. Bahinski, N.J. Izzo Jr., HL-1 cells: a cardiac muscle cell line that contracts and retains phenotypic characteristics of the adult cardiomyocyte, Proc. Natl. Acad. Sci. USA 5 (1998) 2979–2984. [16] S.M. White, P.E. Constantin, W.C. Claycomb, Cardiac physiology at cellular level: use of cultured HL-1 cardiomyocytes for studies of cardiac muscle cell structure and function, Am. J. Physiol. Heart Circ. Physiol. 286 (2004) H823– H829. [17] N. Seong, J.J. Pak, J.H. Choi, D.J. Ahn, S.M. Hwang, K.J. Lee, Fabrication of microstamps and patterned cell network, J. Kor. Phys. Soc. 40 (2002) 605–609. [18] M. Allessie, J. Ausma, U. Schotten, Electrical, contractile and structural remodeling during atrial fibrillation, Cardiovasc. Res. 54 (2002) 230–246. [19] B.J. Brundel, H.H. Kampinga, R.H. Henning, Calpain inhibition prevents pacinginduced cellular remodeling in a HL-1 myocyte model for atrial fibrillation, Cardiovasc. Res. 62 (2004) 521–528.
Contents lists available at ScienceDirect
Biochemical and Biophysical Research Communications journal homepage: www.elsevier.com/locate/ybbrc
Spiral reentry waves in confluent layer of HL-1 cardiomyocyte cell lines Jin Hee Hong, Joon Ho Choi, Tae Yun Kim, Kyoung J. Lee * Center for Cell Dynamics and Department of Physics, Korea University, Anam-Dong 5-1, Sungbuk-Gu, Seoul 136-713, Republic of Korea
a r t i c l e
i n f o
Article history: Received 20 October 2008 Available online 8 November 2008
Keywords: Arrhythmia Cardiac reentries Spiral waves HL-1 cardiomyocytes Mechanical contraction Mouse atrial cells Multi-electrode array (MEA) Inter-beat-interval (IBI) Myolysis Optical mapping Myosin
a b s t r a c t Cardiac excitation waves that arise in heart tissues have long been an important research topic because they are related to various cardiac arrhythmia. Investigating their properties based on intact animal whole hearts is important but quite demanding and expensive. Subsequently, dissociated cardiac cell cultures have been used as an alternative. Here, we access the usefulness of cardiomyocyte cell line HL-1 in studying generic properties of cardiac waves. Spontaneous wave activities in confluent populations of HL1 cells are monitored using a phase-contrast optical mapping system and a microelectrode array recording device. We find that high-density cultures of HL-1 cells can support well-defined reentries. Their conduction velocity and rotation period both increase over few days. The increasing trend of rotation period is opposite to the case of control experiments using primary cultures of mouse atrial cells. The progressive myolysis of HL-1 seems responsible for this difference. Ó 2008 Elsevier Inc. All rights reserved.
Many cardiac arrhythmias are known to be driven by spiral reentries (rotating spiral wave), and different spatiotemporal dynamics of them underlie a different class of cardiac arrhythmia [1–2]. For example, the steady rotation of a single reentry has been associated with monomorphic tachycardia [3] and a non-stationary meandering reentry is believed to cause polymorphic tachycardia [4]. Currently, there is a strong consensus in the medical as well as scientific community that the chaotic wave states following a break-up of regular reentries underlie cardiac fibrillations [5]. Subsequently, there has been a great deal of effort for understanding various aspects of cardiac reentries and their instabilities. The related experimental investigations are, however, a very demanding task, particularly so, when they involve whole heart preparation of a large animal. First of all, it is difficult to keep the whole heart under investigation viable and controlled for more than a day. In other words, long-term investigations lasting several days are not practical if not accessible. Second, the anatomical complexity of animal whole heart often prevents one to make any easy interpretation on the observed phenomena. With these difficulties, various techniques have been developed for primary cultures of dissociated cardiac cells and they have been used very successfully as an alternative to the whole heart, especially when the fundamental properties of cardiac reentries were to be examined [6–9]. One can easily and actively manipulate the density, * Corresponding author. Fax: +82 2 3290 3534. E-mail address: [email protected] (K.J. Lee). 0006-291X/$ - see front matter Ó 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2008.10.168
composition, and the structure of cultured tissues. Thus, the precise experimental quantification of cardiac waves following a systematic parameter change is possible and the result can be compared with those of analytical and/or computational studies. Several research groups have recently demonstrated the existence of reentries in primary cultures of dissociated ventricular cells of chicks [10] and rats [11–14], and studied their various properties. A natural extension of these efforts is to use cardiac cell lines. Several years ago, Claycomb et al. developed a cardiomyocyte cell line termed HL-1 [15], and it is still the only cardiomyocyte cell line currently available. The HL-1 cells are derived from the mouse atrial cardiomyocyte tumor lineage AT-1. They continuously divide and spontaneously contract under some suitable conditions. Many different aspects of HL-1 cells have been studied based on genetic, immunohistochemical, pharmacological, and electrophysiological techniques [16]. However, these studies were designed mainly to answer questions pertaining to cardiac biology at cellular or subcellular levels. We view confluent populations of HL-1 cells as a potentially useful model cardiac tissue for investigating generic features of cardiac spiral reentries. Here, we demonstrate that they can exhibit cardiac reentries just like primary cultures of cardiac cells, and we quantify their long-term wave properties, namely, conduction velocity and rotation period, based on our home-built phase-contrast macroscope and micro multi-electrode array (MEA) plates. These wave properties are compared with those of primary cultures of mouse atrial cells. Interestingly, the rotation period increases in the HL-1
1270
J.H. Hong et al. / Biochemical and Biophysical Research Communications 377 (2008) 1269–1273
cultures while it decreases in the primary cultures. The amount of motor protein, myosin, in Hl-1 cell cultures is quantified at several different stages of the HL-1 culture as a first step for elucidating the observed discrepancy. Materials and methods HL-1 cell line cultures. The HL-1 cell line (passage no. 38) was obtained from Dr. William Claycomb (Louisiana State University, USA). The cells were cultured in a specially formulated growth medium (Claycomb medium, JRH Biosciences, USA) supplemented with 100 l M norepinephrine, 2 mM L-glutamine (Gibco, USA) and 10% FBS (Life Technologies, USA). They were grown in T-flasks coated with 12.5 lg/ml fibronectin (Sigma) and 0.02% gelatin (Sigma) and kept in a 5% CO2 atmosphere at 37 °C. When a HL-1 cell culture flask became confluent, it was split into three batches, and each of which was then considered as the next passage. The culture medium was changed at every 24–48 h. The initial cell plating density was in the range of 0.5–1.4 103 cells/mm2. Culture samples with an initial plating density lower than 0.5 103 cells/mm2 exhibit only small noisy wave segments that do not develop into a stable reentry. Those with an initial plating density higher than 1.4 103 cells/mm2 quickly overpopulate the dish and become unhealthy. Primary cultures of dissociated mouse atrial cells. The atria of 3- to 4-day-old C57BL/6J mice were dissociated enzymatically at 37 °C using 0.05% trypsin (Gibco) and Pancreatin (Sigma). The cells were resuspended in Medium199 (Gibco) supplemented with 1 lM vitamin B12, 1 mM L-glutamine, 100 U/ml penicillin/streptomycin antibiotic mixtures, 5% HS (Gibco) and 10% FBS. The cells were either plated onto MEA plates and (or) dishes that were coated with 12.5 lg/ml fibronectin and 0.02% gelatin. The plating density was kept at 2.0 103 cells/mm2 for all experiments. The culture medium was changed once every 48 h. All the culture procedures abide
by the Guide for the Care and Use of Laboratory Animals published Korea University, Collage of Medicine, Animal Research Policy Committee. Optical mapping of mechanical contractile waves and image analysis. A home-built phase-contrast macroscope along with a custom-built incubation chamber was used to optically map the mechanical contractile activities of cultured cardiac tissues. The construction of the phase-contrast macroscope and the chamber was described in our earlier publication [11]. The imaging (30 frames per second, 512 512 pixels) is based on a ‘‘propagationinduced phase contrast optics” detecting minute shape changes of activated cardiac cells [11–14]. This imaging method does not rely on any extrinsic fluorescent dyes, thus, non-invasive to the samples under investigation and uninterrupted long-term observations over several days are possible. Multi-electrode array (MEA) recordings. The MEA plates were either purchased from Panasonic (MED-P210A) or fabricated at the micro fabrication facilities at Korea University. The construction of home-fabricated MEA plates was described in our earlier publication [17]. The electrical signals picked up by 32 micro-electrodes were amplified 10,000 times (by 10 pre- and 1000 filteramplifier) and band-pass filtered (100–10 kHz). The field potential peaks that had a maximum 4.5–6.0 times (depending on the signal to noise ratio) larger than the standard deviation for 10 min duration of recording were identified as an action potential driven excitation. Based on these activation events, the sequence of interbeat-interval (IBI) was built. The MEA recording system provides a far better temporal resolution (10 kHz) than the phase imaging (30 Hz) but at the expanse of much lower spatial resolution. Immunostaining and quantification for cardiac myosin. HL-1 cardiomyocytes were fixed for 15 min at room temperature in 4% paraformaldehyde in PBS, permeabilized in PBS containing 0.1% Triton X-100, and blocked in 5% BSA for 1 h at room temperature. Then, the permeabilized cells were incubated overnight at 4 °C
Fig. 1. Spiral reentries in population of HL-1 cells (A) and dissociated atrial cells (C). The corresponding space–time plots shown below are obtained along the segment a–b as marked. The long-term evolutions of the rotation period (IBI) and conduction velocity (CV) of spiral reentries for HL-1 cells and dissociated atrial cells are given in (B) and (D), respectively. Images shown in (A) are taken 120 h after the initial plating (projected cell density at t = 120 h is 2.0 103 cells/mm2). The initial cell plating density is 1.0 103 cells/mm2 for the HL-1 cell culture (B) and 2.0 103 cells/mm2 for the atrial primary culture (C,D). In (B), the discontinuity around t = 90 h is due to the fact that the spiral reentry has disappeared briefly and reemerged there. In (D), the error bars are large during the early stage since several competing reentries coexist and their optical signals are rather small. The scale bars are 3 mm. The error bars in (B) and (D) represent spatial fluctuation.
J.H. Hong et al. / Biochemical and Biophysical Research Communications 377 (2008) 1269–1273
1271
Fig. 2. Long-term IBI trends of HL-1 reentries. (A) a schematic diagram of MEA recording system, (B) a typical time trace acquired through a MEA electrode, and (C–E) increasing mean IBI curves of HL-1 reentry for three different initial plating densities [0.7 103 cells/mm2 (C),1.0 103 cells/mm2 (D), and1.4 103 cells/mm2 (E)]. The arrows in (C–E) mark the point of propagation failure. Standard error (66.3 ms) in space domain is smaller than the size of the data symbols in Fig. 2C–E. (F) HL-1 cell population growth assay. HL-1 cells were stained with 4% tryphan blue for counting (n = 5). The error bars represent a standard error.
with a mouse monoclonal antibody against cardiac myosin heavy chain (ab15, abcam, UK) diluted 1:400 in blocking buffer. Fluorescein-labeled isothiocyanate (FITC) anti-mouse (abcam) was used as secondary antibody. The nuclei were stained by 2 lg /ml propidium iodide (PI). Images of FITC and PI fluorescence were obtained by using a fluorescence microscope (IX71, Olympus, Tokyo, Japan) and a confocal laser-scanning microscope (BX51WI, FV 500, Olympus). Results Reentries supported by confluent cultures of HL-1 cell lines It has been known that HL-1 cells can contract spontaneously [15]. However, we find that the contractibility of HL-1 cells depends very much on the developmental stage of the culture, not to mention the cell population density. Especially in the early stage after the initial cell plating, the majority of HL-1 cells do not contract spontaneously. When the culture becomes 1–2 days old, a few cells that are scattered randomly in the populations start to
beat, but rather weakly at 1–3 Hz. Their minute contractile motion is discernable only when inspected under a microscope. As the population density increases and culture becomes more matured in time, the localized activation sites recruit more neighboring cells and expand their territories progressively. The majority of HL-1 cells contracts regularly by the time the culture medium reaches around 3 days in vitro, at this stage they support well-defined rotating spiral waves as shown in Fig. 1A (S1). As far as we understand, this is the first spiral reentry in populations of HL-1 cells ever produced and documented. In other words, HL-1 cells can constitute an excitable tissue that allows wave conduction and reentries. The reentry shown in Fig. 1A rotates quite stably in a periodic manner having a frequency of 3 Hz approximately, as well reflected in the accompanying spacetime plot. The uniformity of the culture medium can be inferred from the regularity of the pitch of the reentry. The reentries in HL-1 cell cultures can last for quite some time under the culture condition described in Methods. Fig. 1B illustrates a typical long-term behavior in terms of their rotation period and conduction velocity. The overall trend is that both increase
1272
J.H. Hong et al. / Biochemical and Biophysical Research Communications 377 (2008) 1269–1273
gradually as the culture gets older. The fact that the rotation period gets longer in time is rather unusual and opposite to our earlier observations made with primary cultures of rat ventricle cells, in which the period of reentries gets shorter progressively over several days [11]. The conduction velocity, on the other hand, shows a similar increasing trend in both systems. Reentries supported by confluent primary cultures of mouse atria As for a comparison, cardiac waves in primary cultures of dissociated mouse atrial cells are also investigated, because HL-1 cell line is derived from mouse atria. Neonatal mouse atria are rather small that many (14–16) of them are needed to form a confluent tissue in a 35 mm dish. The confluent atrial cell cultures also support spiral reentries, and one typical example is given in Fig. 1C (S2). The particular reentry conducts about six times faster and has 2–3 times shorter IBI than the one shown in Fig. 1A. Since the pitch of a reentry is the product of IBI and CV, it is conceivable that the pitch of the reentry wave in the primary cell culture is about 2–3 times larger than that of the HL-1 cell line tissue. The snapshot images in Fig. 1A and 1C are consistent in this regard. One dramatic difference exists between the primary culture of mouse atrial cells and the culture of HL-1 cells; just as in the primary culture of ventricle cells [11], the IBI of the reentries observed in the cultures of atrial cells (Fig. 1D) decreases in time. Thus, the increasing trend of IBI associated with the HL-1 reentries is rather an unusual property, and our subsequent investigation is aimed for explaining this long-term property of HL-1 cell populations.
Long-term trend of HL-1 spiral reentries: initial plating density dependence The MEA electrode recording system (Fig. 2A) is highly sensitive to the local activity and provides a high temporal resolution (10 kHz), thus, it suits better for monitoring a long-term trend of IBI than the optical imaging, including the very early stage of HL1 culture when the contraction of cells is very minute. Several different IBI temporal sequences of HL-1 reentries are obtained, and a typical time series acquired through an electrode is given in Fig. 2B. The sensitivity of the micro-electrodes is indeed far better than that of the optical measurement; the action potential events can be detected as early as 2 days in vitro (about 24 h in advance to the point when the optical imaging is able to detect the motion signals; compare Fig. 2D with Fig. 1B) under the same culture condition. It is surprising that, even during this early period, the HL-1 cell culture supports a very regular sequence of beats. Pacemaker cells usually beat with a much lower frequency and non-periodically that the observed periodic sequences of IBIs must have its origin at self-rotating reentries. Fig. 2C reconfirms the long-term trend of IBI shown in Fig. 1B. Moreover, the long-term trend does not depend on the initial cell density (Fig. 2C–E); in all cases, the increase is gradual at the initial stage but sharply rising at the later stage. Although not exhaustively tested, we find that when the initial cell density is low, the onset of wave initiation and the lengthening of IBI come at a delayed time. The time constants associated with the rising IBI curves corresponding to different initial cell densities, however, do not vary significantly [s = 6.9 ± 1.4 h (n = 3) for 1.0 103 cells/mm2, s = 6.3 h for 1.4 103 cells/mm2, and s = 6.6 h for 0.7 103 cells/ mm2]. When the initial plating density is 1.0 103 cells/mm2 (Fig. 2D), the IBI lengthening begins around t = 96 h. Based on our HL-1 cell counting assay (Fig. 2F), the estimated cell density at this point is about 2.8 103 cells/mm2. Myolysis in HL-1 cardiomyocytes
Fig. 3. Myolysis in proliferating HL-1 cells. (A) Superimposed images of myosin immunostaining (green) and PI stained nuclei (red). Some cells that have undergone myolysis or membrane-blebbing are marked by arrows and arrowhead, respectively. The scale bar is 25 lm. (B) Myosin-expressing (green) area fraction vs. time. The initial cell density is 1.0 103 cells/mm2. Between 48 and 72 h (*** mark) after the initial plating, the area fraction reduces more than 50% of its peak value (p < 0.001; n = 5). Five fluorescent images (512 512 pixels) were obtained and analyzed for each data points. A green scale value of 80 (out of total 255 dynamic range) was chosen as a threshold for defining the myosin rich area. All error bars represent a standard error.
As the Hl-1 culture gets older, two obvious changes take place: first of all, the cell density increases (approximately, doubles at every 48 h, see Fig. 2F), and second of all, the connectivity mediated by gap junctions increases (not shown). These two factors, however, should work to increase the wave conduction velocity and to shorten the IBI. Primary cultures of atrial as well as ventricular cells of rats and mice both support this conclusion [11]. Therefore, the consistently observed long-term trend of lengthening IBI of HL-1 reentries points to a possibility that the HL-1 cells might have changed their cellular properties during the observation periods. Subsequently, we have used cardiac myosin heavy chain antibodies and PI to access the possible occurrence of myolysis and cell degeneration. Beyond 48 h after the plating, the immunostaining clearly reveals progressive myolysis and membrane blebbing in the HL-1 cells (Fig. 3A). The number of nuclei increases steadily, while the number of functional myocytes having a well-developed network of cardiac myosins clearly decreases in time except for the very early stage of the culture when there is no significant wave activity (see Fig. 3B); here, we note that the reentry wave activities shown in Fig. 2D also begin around 48 h in vitro.
Discussion The present study demonstrates the existence of reentry waves in confluent two-dimensional culture of HL-1 cells, and this observation suggests a potential use of HL-1 cells for studying spiral reentry wave dynamics and its instability to spatiotemporal chaos
J.H. Hong et al. / Biochemical and Biophysical Research Communications 377 (2008) 1269–1273
akin to cardiac fibrillation. Unfortunately, however, our study finds that the HL-1 cells go through the process of myolysis in the long run. Thus, HL-1 cell populations may be useful only for studying some short-term (<3 days) properties of cardiac reentries or their instabilities. The degradation process appears similar to that of the myocytes in the atria that has experienced an atrial fibrillation [18]. Incidentally, the recent study of Brundel et al. shows that HL-1 cardiomyocytes can undergo a cell remodeling within 24 h when they are subjected to high-frequency electrical field stimulations [19]. So, the self-rotating high-frequency reentries supported by the HL-1 culture may have elicited the HL-1 cell myolysis. From a clinical point of view, spiral reentry waves are a pathological dynamic state, and as such, the cellular remodeling caused by reentries seems to be a plausible hypothesis. On the other hand, primary cultures of dissociated atrial as well as ventricular cells, which had supported spiral reentries, show no indication of myolysis during the same term of observation (not shown). Therefore, an alternative explanation for the HL-1 degradation can simply be that they get deficient in their motor system because the cells have been regenerated many times over a short duration. Acknowledgments This work was supported by Acceleration Project (R17-2007017-01000-0) of the Korea Ministry of Science and Technology. We thank E. Entcheva and F. Fenton for their suggestions and comments. Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bbrc.2008.10.168. References [1] J.M. Davidenko, A.M. Pertsov, R. Salomonsz, W.T. Baxter, J. Jalife, Stationary and drifting spiral waves of excitation in isolated cardiac muscle, Nature 355 (1992) 349–351.
1273
[2] A.T. Winfree, Electrical instability in cardiac muscle: phase singularities and rotors, J. Theor. Biol. 138 (1989) 353–405. [3] A. Garfinkel, Z. Qu, Nonlinear dynamics of excitation and propagation in cardiac tissue, in: D.P. Zipes, J. Jalife (Eds.), Cardiac Electrophysiology. From Cell to Bedside, W.B. Saunders Company, Philadelphia, 1999, pp. 315–320. [4] R.A. Gray, J. Jalife, A.V. Panfilov, W.T. Baxter, C. Cabo, A.M. Pertsov, Nonstationary vortex-like reentrant activity as a mechanism of polymorphic ventricular tachycardia in the isolated rabbit heart, Circulation 91 (1995) 2454–2469. [5] J.A. Abildskov, R.L. Lux, The mechanism of simulated Torsade de Pointes in a computer model of propagated excitation, J. Cardiovasc. Electr. 2 (1991) 224– 237. [6] L. Tung, Y. Zhang, Optical imaging of arrhythmias in tissue culture, J. Electrocardiol. 39 (2006) S2–S6. [7] G. Bub, N. El-Sherif, Monolayer cell cultures as model systems for studying paroxysmal atrial fibrillation, J. Electrocardiol. 37 (2004) 44–46. [8] N. Bursac, F. Aguel, L. Tung, Multiarm spirals in a two-dimensional cardiac substrate, P. Natl. Acad. Sci. USA 101 (2004) 15530–15534. [9] X.L. Tian, Y. Cheng, T. Zhang, M.L. Liao, S.L. Yong, Q.K. Wang, Optical mapping of ventricular arrhythmias in LQTS mice with SCN5A mutation N1325S, Biochem. Biophys. Res. Commun. 352 (2007) 879–883. [10] H. Gonzalez, Y. Nagai, G. Bub, L. Glass, A. Shrier, Reentrant waves in a ring of embryonic chick ventricular cells imaged with a Ca2+ sensitive dye, Biosystems 71 (2003) 71–80. [11] S.M. Hwang, K. Yea, K.J. Lee, Regular and alternant spiral waves of contractile motion on rat ventricle cell cultures, Phys. Rev. Lett. 92 (2004) 198103. [12] S.M. Hwang, T.Y. Kim, K.J. Lee, Complex-periodic spiral waves in confluent cardiac cell cultures induced by localized inhomogeneities, Proc. Natl. Acad. Sci. USA 102 (2005) 10363–10368. [13] T.Y. Kim, S.J. Woo, S.M. Hwang, J.H. Hong, K.J. Lee, Cardiac beat-to-beat alternations driven by unusual spiral waves, Proc. Natl. Acad. Sci. USA 104 (2007) 11639–11642. [14] S.J. Woo, J.H. Hong, T.Y. Kim, B.W. Bae, K.J. Lee, Spiral wave drift and complexoscillatory spiral waves caused by heterogeneities in two-dimensional in vitro cardiac tissues, New J. Phys. 10 (2008) 015005. [15] W.C. Claycomb, N.A. Lanson, B.S. Stallworth Jr., D.B. Egeland, J.B. Delcarpio, A. Bahinski, N.J. Izzo Jr., HL-1 cells: a cardiac muscle cell line that contracts and retains phenotypic characteristics of the adult cardiomyocyte, Proc. Natl. Acad. Sci. USA 5 (1998) 2979–2984. [16] S.M. White, P.E. Constantin, W.C. Claycomb, Cardiac physiology at cellular level: use of cultured HL-1 cardiomyocytes for studies of cardiac muscle cell structure and function, Am. J. Physiol. Heart Circ. Physiol. 286 (2004) H823– H829. [17] N. Seong, J.J. Pak, J.H. Choi, D.J. Ahn, S.M. Hwang, K.J. Lee, Fabrication of microstamps and patterned cell network, J. Kor. Phys. Soc. 40 (2002) 605–609. [18] M. Allessie, J. Ausma, U. Schotten, Electrical, contractile and structural remodeling during atrial fibrillation, Cardiovasc. Res. 54 (2002) 230–246. [19] B.J. Brundel, H.H. Kampinga, R.H. Henning, Calpain inhibition prevents pacinginduced cellular remodeling in a HL-1 myocyte model for atrial fibrillation, Cardiovasc. Res. 62 (2004) 521–528.