1. Introduction

Cardiomyocytes are the muscle cells that provide for the contractility of bulk heart tissue. They exhibit a fascinating and complex range of dynamic behavior that forms the building blocks of the physiology of the heart [1], and have been the subject of a vast number of inter-species studies in biomedical fields [2]. Isolated from their original tissue, cardiomyocytes have shown great potential as a basis for the artificial engineering of heart tissue [3, 4]. Although the therapeutic applications of cardiomyocyte-based structures are not yet at hand, the outlook is promising. There are also a number of interesting applications where the self-contracting or electrically controlled, chemically-powered actuator properties of the cardiomyocytes may be utilized for structured actuator sheets [5] or as single actuators [6]. There is therefore considerable interest in understanding the mechanisms of heart muscle cell contraction, extending the existing knowledge [1], and in determining what methods may be used to regulate the contraction. The work discussed here outlines a novel method with which to regulate or modify the contraction of cardiomyocytes.

We have previously demonstrated cellular responses to light, and showed that cells without added photosensitive compounds can respond to a brief exposure to focused near-infrared ultrashort-pulsed laser irradiation by an elevation of the intracellular calcium concentration [7], and that different cell types can exhibit a similar response to laser irradiation [8]. Since many cellular responses are regulated by calcium, this laser irradiation can be used to provoke calcium-directed cellular responses without necessarily causing adverse affects on the target cell viability [9]. In cardiomyocytes in particular, the calcium ion dynamics exhibit a wide range of subcellular [10] and cell-level activity [1], and facilitates the binding of actin and myosin, which in turn directly controls the cell contraction, and opens up the possibility of using light to control the muscle cell contraction. While photolysis of caged calcium can be achieved in a cell by flashlamp or laser photolysis [11–13] and can be applied to perturbation of muscle cell contraction, it requires the addition of caged compounds to the target cell before irradiation. Our previous work, as well as reports by other groups [14–17], indicates that the femtosecond laser light can independently provide sufficient stimulus to perturb and possibly control the contraction dynamics of a cardiomyocyte.

The claim under test here is then simply: whether periodic exposure of femtosecond laser irradiation focused into a cardiomyocyte can generate trigger amounts of cytosolic calcium that can be used to cause the cell to contract and whether such contraction can occur multiple times in synchronization with the exposure periodicity.

To test this, we use periodic exposure to focused femtosecond laser light to trigger, periodically control, and/or eliminate contraction cycles in cardiomyocytes by repeatedly causing a small amount of calcium release in accordance with previous work [9]. To avoid overloading the cell with calcium ions, the laser power was set to the minimum power level that was sufficient to cause repeatable calcium release at a single exposure. This power level was determined from our previous work to be between approximately 15 to 25 mW of average power from a 780 nm, 80 fs, 82 MHz pulsed laser source, focused by a 0.9 NA objective lens. This technique is applicable to the study of perturbations on a single-cell level in heart muscle contraction. It may additionally be used as a novel technique to control the contraction of microactuators based on cultured cardiomyocytes.

2. Experimental methods

2.1 Cardiomyocyte isolation and usage

All animal experiments were carried out in accordance with university and federal regulations. Cardiomyocytes were isolated from neonatal rats by Percoll gradient density centrifugation. The hearts of 1 day old neonatal Wister rats (Hamaguchi Doubutsu and Nihon Doubutsu) were extracted by opening the chest cavity under diethylether anesthesia. The hearts were washed with PBS (NaCl: 136.9 mM, KCl: 2.68 mM, Na₂HPO₄: 8.10 mM, KH₂PO₄: 1.47 mM, CaCl₂: 0.90 mM, and MgCl₂ • 6H₂O: 0.49 mM, [pH 7.4]) at 4°C, and then washed twice with PBS(-) (NaCl: 136.9 mM, KCl: 2.68 mM, Na₂HPO₄: 8.10 mM, KH₂PO₄: 1.47 mM) at 4 °C. The hearts were minced into cubes of 1–2 mm² approximate volume, and rinsed with PBS(-). Afterward, the tissues were processed with 0.2 weight % collagenase dissolved in 20 ml PBS(-) for 30 minutes at 37 °C, while stirring. The supernatant solution was rapidly transferred into a tube with 20mL cell culture medium (pH 7.4), containing 9.5 mg/1 mL Dulbecco’s Modified Eagle’s Medium (DMEM, [Nissui]), 10% fetal bovine serum (FBS), 1% penicillin-streptomycin-glutamine (Gibco), 1 µg/mL cytosine-1-β-D(+)-arabinofuranoside, and 1.932 mg/mL NaHCO₃, and centrifuged at 180 G for 5 minutes. The cell pellet was resuspended in the 10mL cell culture medium, and filtered through a 40 µm pore nylon mesh, and the filtered cell suspension was centrifuged again at 180 G for 5 minutes. The cell pellet was then resuspended in HEPES buffer (pH 7.30–7.40), containing HEPES: 20 mM, NaCl: 116 mM, NaH₂PO₄: 1.0 mM, glucose: 5.5 mM, KCl: 5.4 mM, and MgSO₄: 0.8 mM. The cardiomyocytes present in the suspension were separated from other cells by way of Percoll density gradient centrifugation. 40.5% Percoll diluted in HEPES buffer was layered on 58.5% Percoll diluted in HEPES buffer in a tube. The cell suspension was layered on the top of the bilayer, and centrifuged at 2200 G for 30 minutes. Cardiomyocytes were retrieved from the interface between 40.5% and 58.5% Percoll layers. The retrieved cells were then suspended in DMEM, and centrifuged at 180 G for 5 minutes to remove the remaining Percoll solution. The cardiomyocyte pellet was then resuspended in DMEM and the cardiomyocyte suspension was seeded on culture dishes (non-coated glass bottom dish, Matsunami), which in some experiments were treated with a thin collagen layer, so as to achieve 40–60% confluence. The cardiomyocytes were incubated for one day at 37 °C in a 5% CO₂ incubator. After initial seeding, the cell culture medium was changed every 2 or 3 days. Experiments were performed between 3–6 days following seeding.

Prior to observation, cell culture dishes were bathed in a solution containing CaCl₂: 1 mM, NaCl: 145 mM, KCl: 4 mM, MgCl₂: 1 mM, glucose: 10 mM, and HEPES buffer (pH=7.4): 10 mM and Fluo-4AM: 0.018mM. The cells were loaded with Fluo-4 by incubating in this solution for 30 minutes at 23 °C. Laser irradiation and observation experiments were also carried out at room temperature in a controlled environment at 23 °C.

2.2 Optical setup and image processing

The laser source was a Ti:Sapphire laser (Tsunami, Spectra Physics) of 780 nm radiation, modelocked to produce 80 fs pulses at an 82 MHz repetition rate. The laser beam entered an inverted microscope (IX70, Olympus) via a dichroic mirror, to allow simultaneous laser irradiation and fluorescence imaging. A beam expander was used to fill the aperture of a 60x/0.9 NA water-immersion lens, and the light source for fluorescence microscopy was a mercury lamp. Phase contrast optics were also used to confirm the relationship between calcium imaging and physical contraction. Cells can be imaged in fluorescence or phase contrast mode microscopy. The two modes show similar information since intracellular calcium (visible by fluorescence microscopy) causes the mechanical contraction that can be visualized by phase contrast microscopy. Since the laser illumination was usually not visible in phase contrast microscopy modes, experiments were done in fluorescence mode imaging (unless otherwise noted). The laser power was measured at the focus by a photodiode meter (PD-300, Ophir), and the laser irradiation timing was controlled by a mechanical shutter (F77-4&F77-6, Suruga Seiki) set to 8 ms for each individual exposure. Each 8 ms exposure was intended to trigger a single response (i.e. contraction) in a single cell, and the exposure was periodically repeated by a trigger signal on the TTL input of the shutter controller. The trigger signal to control the shutter was provided by software (Cubase, Steinberg) running on a pc with soundcard, and amplified to TTL voltage levels, allowing the periodic exposure to be continuously varied at rates of between 0.5 and 3 Hz in increments of 0.01 Hz. Unless otherwise noted, all experiments were carried out with a laser irradiation periodicity of 1.0 Hz. Note that the exposure periodicity is distinct from and unrelated to the pulse repetition rate of the laser source. The “exposure periodicity” refers to the period between shutter opening times, is variable, and is on the order of a few Hz. The “repetition rate” refers to the pulsed nature of the laser source and is fixed at 82 MHz.

Images were taken in fluorescence and phase contrast microscopy, and fluorescence video sequences were recorded using Fluo-4 loaded cells. The video sequences were analysed by software (ImageJ, NIH and Igor Pro, WaveMetrics) and selected cell regions were highlighted for further time profile, correlation, or Fourier analysis. For analysing frequencies, Gabor sonograms were used with 0.3 Hz resolution. There were 400 time points per spectrum, and spectra were generated every 2 samples, using 1601 frequency bins.

3. Experimental results

3.1 Laser pacemaking in single cardiomyocytes

We found that the periodic irradiation could produce a periodic response seen as a calcium elevation in target cells. Fig. 1(a) shows a cardiomyocyte with two regions labeled. Region “i” contains the fluorescence from the calcium indicator in the cell, as well as fluorescence generated by the laser focal spot, positioned approximately 2 µm below the top membrane. Region “ii” encapsulates fluorescence from the cell as a whole, and has a higher signal to noise ratio due to the larger integrated area. Fluorescence signals were integrated over these regions and plotted versus time, as shown in Fig. 1(b). The fluorescent signal from region ‘i’ is dominated by transient laser-induced multiphoton absorption causing fluorescence from the Fluo-4 calcium indicator, as well as the fluorescence evoked from endogenous fluorophores.

 

Fig. 1. (a). Cardiomyocyte loaded with Fluo-4 calcium indicator. The region ‘i’ shows the laser focal spot allowing visualization of the irradiation timing (8 ms exposures at 1 Hz, 25 mW). Region ‘ii’ encapsulates the surrounding area of the cell. (b) Shows the induced contraction due to the calcium response over time. The laser is visible as a series of spikes in the fluorescence signal in trace ‘i’, while the overall cell response is averaged over time and plotted as trace ‘ii’. The first 3 laser exposures are indicated by vertical arrows.

Download Full Size | PDF

The contraction of the cell is most clearly observed by the change in calcium concentration in the cell, and from Fig. 1(b), we can see that the target cell was not initially contracting (evident by the horizontal line ii, from 0 to 9 seconds). Following laser irradiation at 25 mW average power for 8 ms with sequential exposures at a rate of 1 Hz, the cardiomyocyte fluorescence begins to follow the periodicity of the laser exposures. The cardiomyocyte shows no visible response to the first and second laser exposures, and the first whole-cell response occurs after the third exposure (at ~9 seconds), with subsequent responses appearing after each additional laser exposure. In the data in Fig. 1, it can be seen that the calcium level rises and falls following the laser exposures, and subsequent laser exposures cause a build-up of calcium in the cell. This means that, for longer-term laser pacemaking, either the period between exposures is too short, or the amount of calcium released by the laser exposure is too high, and the cardiomyocyte eventually becomes incapable of periodic contraction (from 18 seconds onwards).

 

Fig. 2. Cardiomyocyte synchronized at 120 beats per minute periodicity during laser irradiation of 8 ms exposures at 2 Hz with 25 mW average power, and then spontaneously contracting at 40 beats per minute following irradiation. (a) The region ‘i’ shows the laser focal spot, while region ‘ii’ encapsulates the surrounding area of the cell. (b) Shows the induced contraction due to the calcium response over time. The first 3 laser exposures are indicated by vertical arrows.

Download Full Size | PDF

Similarly, Fig. 2(a) shows the fluorescence traces in a separate cell from the laser irradiation region (region i) and cell fluorescence (region ii), indicating a cell contracting in synchronization with the laser irradiation at 2 Hz, corresponding to 120 beats per minute (bpm). Similar to the data shown in Fig. 1, the cell contraction is controlled by the whole-cell calcium concentration (Fig. 2(b), trace ii), and synchronization occurs with the laser periodicity (Fig. 2(b), trace i) after the fourth laser exposure. The first response of the cell is observed between the second and third exposures, in a similar manner to the measurements shown in Fig. 1, and after the laser periodic exposure is stopped, the cell returns to a contraction periodicity of approximately 0.66 Hz, or 40 bpm. The delay between the first laser exposure and the first observable response in the cell was a common occurrence. For all experiments where any evidence of synchronization was observed, approximately 1 in 3 showed the first observable cell response occurring after the second or third laser exposure (i.e. as shown in Figs. 1 and 2)

3.2 Laser pacemaking in cardiomyocyte groups

Cultured cardiomyocytes undergo spontaneous contractions that depend on substrate properties [18], proximity of surrounding cells [19], and other conditions [20]. Spontaneous contractions were observed in our experiments, and varied between 0.2 Hz and 1.5 Hz, depending predominantly on the number and location of surrounding cells. For target cells with such a pre-existing contraction cycle, periodic laser exposure did not immediately cause the cells to begin contracting with the same periodicity as the irradiation, and the laser pacemaking effect was complicated by the contraction periodicities of the surrounding cells. In some cases, however, the laser irradiation could be seen to dominate the contraction periodicities of all cells and act as a pacemaker not only for the target cell but also for the surrounding cells. An example of this is shown in Fig. 3. Here, a small group of cells contract in phase with each other, at a rate of approximately 0.2 Hz before laser irradiation.

 

Fig. 3. Laser pacemaking in a group of cardiomyocytes. The cardiomyocyte group consists of more than 6 cells, spontaneously contracting in synchronization. The laser power was set to 20 mW with 8 ms exposures at 1 Hz and targeted in the region marked ‘i’ in Fig. 3(a), with surrounding cells labeled by regions ii–v. The laser irradiation timing can be seen most clearly in Fig. 3(b) by the spikes in trace ‘i’. Arrows indicate the onset of laser irradiation. No observable effect in surrounding cells (traces ii–v, Fig. 3(b)) is seen for the first 20 seconds of periodic laser exposures, after which, all cells can be seen to contract at the periodicity of the laser exposures. Of particular interest is that all cells respond identically even though only one cell is targeted by laser irradiation (evident by the negligible differences in traces ii–v). Following irradiation, all cells enter a 5 s duration rest phase during which the intracellular calcium level is lowered before spontaneous contractions restart. (See supplemental video material for the data shown in Fig. 3(a)). [Media 1][Media 2]

Download Full Size | PDF

Following periodic laser exposures of 8 ms each at a rate of 1 Hz (shown by trace i), the contractions in all cells begin to synchronize with the exposure periodicity (traces ii–v). The synchronization occurs after the 22^nd irradiation, and continues until the irradiation is ceased. The calcium level then decreases without cell contraction, until the cell contraction occurs again, approximately 5 seconds after irradiation is ceased. The final contraction periodicity is unrelated to the periodicities before, and during irradiation. This is evidence that the periodic laser stimulation can be used to push the existing dynamics of a group of cardiomyocytes into a new regime. The fluorescence intensity in all cells decreases over several seconds following the final laser exposure before contraction spontaneously restarts, showing that the cells can respond to laser exposure periodicities that are faster than spontaneous contraction rates for the culturing conditions used in these experiments.

3.3 Pacemaking periodicity degradation following laser irradiation

The fluorescence signal in time can be used to confirm the synchronization of the laser irradiation periodicity with the whole-cell calcium cycle, which in turn regulates the contraction. Since the laser exposure is only 8 ms, and the periodicity is on the order of 1 Hz, the duty cycle of the laser exposure is approximately 0.01. In the frequency domain, this can be observed as a high number of harmonics above the fundamental periodicity of the laser (which was controlled by the shutter opening and closing). The fluorescence data was analysed by windowed Fourier transform with a Gaussian window and a frequency resolution of 0.3 Hz. In Fig. 4(a), the fluorescence from the laser region of a target cell (trace ‘i’) is plotted against time and compared with the fluorescence from the cell region excluding the laser focal spot (in trace ‘ii’), as denoted in Fig. 4(b). The harmonics visible for the laser irradiation signal (Fig. 4(c)) exhibit obvious periodicity, and the corresponding harmonics from the cell calcium signal are shown in Fig. 4(d).

 

Fig. 4. Laser synchronized cardiomyocyte contraction cycles that continue after the irradiation of 25mW at 1 Hz is stopped. (a) shows fluorescent signals from the cell regions, with region ‘i’ corresponding to the laser focal spot, and region ‘ii’ representing the surrounding cell area, as outlined in (b). Panels (c) and (d) show the time-dependent Fourier transforms (TDFT) of the signals in the laser region ‘i’ and surrounding cell region ‘ii’. The color lookup table shows the amplitude of the frequency components in log scale. The cardiomyocyte retains the laser periodicity for approximately 20 contraction cycles after irradiation is stopped. The cell contraction cycles stabilize during irradiation and continue in the absence of irradiation with no significant changes in the time or frequency domain behavior. Final application of periodic laser exposure at a higher rate of 1.5 Hz brings the contractions to a halt.

Download Full Size | PDF

The fluorescence signals in Fig. 4(a) combined with the frequency analysis in Fig. 4(d) show that the changes in cell contraction periodicity occur rapidly; the cell contractions do not slowly approach the periodicity of the laser. Rather, the question of synchronization with laser exposures seems to occur on a case-by-case basis, even though the target cell, laser power, and exposure conditions remain the same from 10 to 90 s. The peak at 1 Hz can be seen as a horizontal line at the bottom of Fig. 4(d) and represents the fundamental contraction periodicity in the cell. There are minimal differences between the harmonics of the cell contraction when synchronized to the laser (panel (d), 40 to 65 s) and when the cell contracts without laser irradiation (panel (d), 65 to 80 s). This supports the hypothesis that the laser interaction with the cardiomyocyte contraction cycle is exerted as a triggering-type effect rather than an overriding control of normal cardiomyocyte function. It further indicates the strong harmonics in the fluorescent signal from the laser focal region (region ‘i’) do not significantly influence the measurements in the surrounding area of the cell (region ‘ii’).

Remarkably, the target cell retains the contraction periodicity of the laser exposure for approximately 20 cycles after the laser is switched off (at 64 seconds), and is relatively unchanged by a further single exposure (at 83 seconds). When the laser exposure is again started periodically at a rate of 1.5 Hz (at 92 seconds) the cell movement is halted. This occurs apparently due to the inability of the cell to sufficiently reduce the intracellular calcium level during the refractory (relaxation) stage of the contraction cycle, and is evident from the constant high fluorescent signal from the cell. Whether this results simply from intracellular damage due to calcium overload or is a more complex mechanism resulting from the fact that the laser exposure periodicity of 1.5 Hz is substantially different from the cellular contraction periodicity of 1 Hz will require further study. Finally, a significant difference between the dataset shown in Fig. 4 and those in the previous figures is that in Fig. 4, over the time range of 20 to 90 s, the calcium level does not build up between laser exposures. This is obviously a pre-requisite for longer-term synchronization between the laser and target cell contraction. The fact that 25 mW at 1 Hz exposure periodicity did not induce a build-up of calcium in Fig. 4 while 20 mW irradiation at the same exposure conditions was sufficient to diminish cell contractions in Fig. 2 serves to illustrate the variations in individual cell dynamics.

3.4 Optical parameters for laser synchronization and feasibility of the technique

In total, over 200 laser-cardiomyocyte synchronization trials were performed in the experiments reported here with different laser powers and periodicities. The synchronization of the cell contraction periodicity with the laser exposure periodicity occurred in approximately 25% of all trials for average laser powers of between 15 to 30 mW (measured at the sample) at a laser periodicity of 1 Hz. To quantify the relationship, we restricted the definition of synchronization so that around 5 or more consecutive beats were required to follow the laser periodicity in frequency and phase. With this definition of synchronization, we recorded the results shown in Fig. 5(a), over a total of 181 cells using 1 Hz irradiation periodicity. We investigated the number of cells that exhibited synchronization and also the number that exhibited some calcium response to the irradiation and/or showed indications of synchronization that did not fit our definition of synchronization as given above. We compared the probabilities of these outcomes while varying the average laser power. The data necessarily includes experiments from different days, which means that there is a degree of variability in the cell culture conditions. This, as well as the difficulty in achieving large amounts of data where only one parameter is changing, probably accounts for the variation in probabilities across adjacent laser powers. The data shows that the synchronization generally occurs at the onset of an observable response in the cardiomyocytes and shows that while synchronization can be repeatedly induced, more work needs to be done to clarify the precise optical and/or biological parameters that govern the interaction of the cell’s spontaneous contraction rate and the external periodicity of the laser influence on the cell dynamics. The data are then further divided into two constituent groups: experiments where a single isolated cell was irradiated (a), and experiments where one cell within a group of 2–14 connected cells (b) was targeted and irradiated (similar to that shown in Fig. 3). While the data is insufficient to conclude the distinction between isolated cells and those within a group, it does point to an intriguing possibility; the cell synchronization of an entire group of 2–14 cardiomyocytes may be more probable than the synchronization of a single cardiomyocyte. More research will be required to determine if this is the case, and this may have implications for the onset of new and often undesirable contraction periodicities in groups of cardiomyocytes.

 

Fig. 5. The probability of generating a synchronized cell response or a calcium response without synchronization. Both are functions of laser power. The graph in (a) shows all data sets for a laser irradiation periodicity of 1 Hz. In (b), the graph shows the cases where the target cell was an isolated cardiomyocyte, and (c) shows the cases where the target cell was within a group of 2–14 cardiomyocytes. For synchronization to occur in cases shown in (c), all cells in the group were seen to synchronize with the laser periodicity. Conditions necessary to merit the term “synchronization” are described in the main text.

Download Full Size | PDF

To further quantify the precise parameters of the synchronization phenomenon requires control of the laser power, periodicity, and timing, which can be readily achieved. It is complicated, however, by the additional variations in individual cells. The spontaneous contractions that occur in cultured cardiomyocytes depend strongly on proximity to surrounding cells [19], providing for substantial variation during each trial in the contractions existing before the laser irradiation. The laser periodicity and timing (relative to spontaneous contractions in the cells) that is best suited to synchronizing individual cardiomyocytes is not yet clear but appears to be between 0.5 and 2 Hz. The laser power window where cardiomyocyte contraction can be synchronized with the laser periodicity of 15 to 30 mW is quantitative and shown in Fig. 5, with lower laser powers producing no short-term observable effects, and higher laser powers producing constant calcium elevations and no cyclic contractions. The effect was dependent on the mode-locked operation of the laser and was not observed in continuous-wave mode irradiation, even for average laser powers of up to 100 mW. This corresponds well with previous work in non-excitable cells which shows that laser-induced calcium responses can occur at a threshold average power of approximately 30 mW and only when the laser is operated in pulsed mode [9]. In that report, laser-induced calcium responses were only observed above 20 mW average power using single exposures of 13 ms. For cardiomyocytes described in the current work, we used periodic exposures and observed a lower limit of 15 mW for generating a visible effect in target cells. This lower limit may decrease further in future experiments, particularly if the beam is scanned, the exposures used are longer than 8 ms, or if the exposure is periodically repeated over periods longer than the seconds to minutes time frame used in the current work. A large number of femtosecond laser pulses has also been shown to create reactive oxygen species that can mediate cellular damage [7] and can sequentially degrade the sample at the laser focus via low density plasma induced chemical decomposition [21]. The effects of scanning the beam are not treated here but have been shown to further decrease the lower limit for the onset of laser-induced effects in living cells [22].

It is interesting to note that the laser irradiation effects demonstrated in this paper (Figs. 1–4, for example) do not initially cause observable calcium responses, but instead after repetitive stimulation cause whole-cell calcium elevations which control cell contraction. This shows that the laser technique described here acts as a trigger for perturbing and synchronizing cardiomyocyte contraction but does not directly generate the majority of the calcium which dominates the contraction cycle. The possibility of generating trigger amounts of calcium which perturb the existing cell contraction cycle offers the opportunity to use this technique to study or simulate the onset of abnormal calcium fluctuations in cardiomyocytes. It is often claimed that such cell-level calcium dynamics are a significant factor in the onset of fibrillation and finally heart-failure [23].

While more research must be carried out to isolate the parameters of the periodic laser interaction with the cardiomyocyte contraction, the use of femtosecond laser irradiation allows several unique possibilities; the irradiation can be used to remotely trigger, synchronize or impede contractions in heart muscle cells, examples of which are shown in this work. The near-infrared wavelengths additionally offer the possibility of deep penetration into actual cardiac tissue with minimal interaction with cells apart from those at the focal spot.

With regard to the use of the calcium fluorescence signal in the cardiomyocyte to measure contraction, the calcium directly regulates the contraction in normally functioning cardiomyocytes. The laser irradiation is not part of the natural biological and mechanical driving and damping forces in a cell, and more work needs to be done to quantify what departure, if any, occurs from the normal calcium played by calcium in the excitation-contraction coupling of the cardiomyocyte. Without using the calcium signal, visualizing and quantifying the simultaneous laser signal and cell contraction is more difficult, but can still be achieved by analyzing the video sequence, as shown in Fig. 6, where phase contrast is used to highlight cell contraction movements over time. Fig. 6 shows 3 adjacent cells that are initially contracting without exposure to laser irradiation. Since the cells are contracting but not in one preferred direction, the contractions are not easily plotted. The video data was therefore processed by subtracting the n^th frame from the (n-1)^th frame in the phase contrast video signal, to differentiate the data in time. The resulting frame sequence then contained positive and negative coefficients corresponding to the movements in time of image features. These were converted to purely positive coefficients by squaring the data, and then each frame was summed over space coordinates to produce a final 1-d signal that shows total movement of all cells in the viewing area as a function of time. We then correlated the cell movement during the laser irradiation (Fig. 6(c)), regions ‘v’ and ‘vi’) with the laser signal (regions ‘vii’ and ‘viii’), and obtained correlation coefficients of .33 and .27 (see Fig. 6(b)), when the laser was on. To estimate the statistical significance of these correlations, we carried out the analogous calculation, but using the movement of cells during a 10 second interval before the laser was firing (region ‘ix’). For each such calculation, we shifted the phase of the cell movement signal in region ‘ix’ forward by 0–30 time-steps and correlated these with 10 second data segments ‘v’ and ‘vi’ from the laser signal when the laser was on, thus obtaining 31 reference data sets with which to compare the significance of the observed correlations. The reference correlations spanned a range from 0 to 0.2, with two peaks at 0.02 and 0.06, as shown in Fig. 6(b). The correlation between cell contraction and laser pulse periodicity is clearly significant, with the observed correlations for “Laser on” time periods lying far outside the distribution of reference correlations. The observance of some positive correlation of the cell movement before the irradiation starts is due to the cells spontaneous contraction, at a rate that is close to the 1 Hz laser irradiation periodicity. The 15 mW laser irradiation at 1 Hz then acts to synchronize the contractions of the 3-cell group. The cell movement signal can also be seen to increase in amplitude at the locations of closest overlap with the laser fluorescence signal (Fig. 6(c) 38–49 s, and 112–121 s).

 

Fig. 6. Time analysis of a phase contrast video showing the effect of 1 Hz, 15 mW periodic irradiation. Frame (a) shows an image from the video where the laser region is outlined and marked ‘i’, while the region for cell movement analysis includes all pixels outlined by the box marked ‘ii’. (b) shows correlation coefficients between the laser signature and the cell movement with the laser on, and for multiple time-shifted locations where the laser was off. (c) shows the time dependent signal representing cell movement plotted under the laser fluorescence signal. See main text for details.

Download Full Size | PDF

3.5 Laser-driven cardiomyocytes as remotely triggered micro-actuators

The ability to remotely trigger cardiomyocyte contraction also allows its use as a means to target and cause contraction in cultured cardiomyocytes that can then be used as micro-actuators. As long as the cell remains viable, we can increase the power and exposure time and force the cell to contract by intentionally overloading the cell’s intracellular calcium levels. Although this can cause irreparable damage to the cell, we can elicit a significant contraction and force the cell to act as a single-shot actuator. Compared to experiments where synchronization was achieved, this type of response in a cardiomyocyte is much simpler to elicit and should require only that the cell is viable and in a relaxed state. An example of this is shown in Fig. 7. To force the contraction to occur, sufficient average laser power is necessary, but as shown in the image sequence, was only 30 mW for the cell shown in Fig. 7. The laser is focused in the cytoplasm of the target cell and periodically exposed at 1 Hz (and is not visible due to the background light of the phase contrast imaging mode). The contraction occurs over a total time of approximately 16 seconds. This type of contraction was different from the type of contraction that occurred in experiments where the contraction synchronized with the laser exposure periodicity (see for example, Fig. 3). The contraction occurs less rapidly than the laser-synchronized beat periods (e.g. Fig. 4(a) 40–60 s) or spontaneous contraction periods (e.g. Fig. 4(a) 0–10 s), and unlike the synchronization results shown in Figs. 1–4, the cardiomyocyte in Fig. 7 remained in a contracted state for several minutes. To test the repeatability of the type of contraction demonstrated in Fig. 7, we set the average laser power at the sample to 35 mW and chose 15 cells which exhibited morphology consistent with good cell health, and irradiated each of them at 1 Hz periodicity for up to 20 seconds, while waiting for a contractile response. Out of these, 11 exhibited behavior similar to that shown in Fig. 7, showing the potential for using irradiation to control cardiomyocyte-based actuators.

 

Fig. 7. Laser actuation of a single cultured cardiomyocyte by irradiation, causing it to contract and stay contracted. The image sequence was taken by phase contrast microscopy and the outline of the cell is highlighted. The laser was focused inside the cell using 30 mW of average power, and 8 ms exposures occurred at 1 Hz intervals. The time sequence is (a) 1 sec, (b) 8 sec, (c) 14.7 sec, and (d) 16.7 sec.

Download Full Size | PDF

4. Conclusion

In conclusion, we have shown that focused femtosecond laser irradiation of 15 to 30 mW power can be used to trigger contraction in individual heart muscle cells, and the contractions can be synchronized to the periodic application of laser light, allowing the laser to be used as an optical pacemaker. The pacemaking effect was also observed in groups of cells, even where only one cell in the group was targeted by the laser, showing that the periodic laser exposure can play an integral role in the dynamics of coupled and synchronized groups of cells. By measuring the power dependence of laser synchronization in isolated cardiomyocytes vs single target cells within cardiomyocyte groups, initial results indicate that target cells within cardiomyocyte groups of 2–14 in number are as susceptible, or may be more susceptible to synchronization than isolated cells. Depending on the interactions with surrounding cells, laser-generated contraction rhythms were observed in some cases to continue in the target cell even after the laser irradiation was stopped. Although there are other simpler methods by which to synchronize cardiomyocytes (e.g. electrical current-based regulation of contraction or even photolysis of loaded caged compounds), the femtosecond laser interaction may be a useful tool with which to apply a driving stimulus that can synchronize contractions in heart muscle cells and may possibly be able to penetrate through substantial depths of heart muscle tissue due to the multiphoton absorption and may facilitate the investigation of such synchronization in-vivo, where other methods cannot be used or cannot achieve the same degree of subcellular localization. Apart from synchronization, it may also be used to study how perturbations affect existing excitation-contraction coupling in cardiomyocytes. We also propose that the laser interaction can be put to use as a method of controlling cardiomyocyte-based actuators, which can be driven periodically or forced to undergo contraction for longer periods of time, as shown here.