Studying electrophysiological properties of cardiac ion channels with the patch-clamp technique and the exploration of cellular Ca²⁺ handling abnormalities require isolated cardiomyocytes. These are usually obtained following the in vitro exposure of cardiac tissue samples to digestive enzymes (collagenase, hyaluronidase, peptidase etc.). Since the first report of isolation of viable cardiac myocytes in 1955² a large quantity of protocols has been developed in order to harvest single atrial and ventricular cardiomyocytes from different species including mouse, rat, rabbit, dog, guinea pig and human. In this review we focus on isolation of human atrial myocytes. Regarding procedures for isolation of myocytes from other species we refer to the "Worthington Tissue Dissociation guide" provided by Worthington Biochemical Corp., USA (www.tissuedissociation.com).

Human atrial myocyte isolation protocols are generally derived from the method described by Bustamante, et al.³ Here we provide a step-by-step description of a technique, which is adapted from a previously published method, in order to obtain atrial myocytes suitable not only for patch-clamp experiments but also for simultaneous intracellular Ca²⁺ measurements.^4-11

Experimental protocols need to be approved by the local ethics committee and all patients need to give written informed consent. Our research was approved by the ethics committee of the Medical Faculty Mannheim, University of Heidelberg (#2011-216N-MA) and was performed in compliance with all institutional, national and international guidelines for human welfare. All patients gave written informed consent.

0. Obtaining Human Atrial Tissue

During routine cannulation procedures in patients undergoing open-heart surgery for cardiopulmonary bypass grafting, the tip of the right atrial appendage is usually removed and can be used for isolation of atrial cardiomyocytes. After excision the tissue sample is transferred immediately into a 50 ml Falcon tube with sterile Ca²⁺-free transport solution containing 2,3-butanedione monoxime (Table I; BDM, contractile inhibitor, preventing myocyte contracture). With transport times between 30 and 45 min, we did not recognize a clear advantage of cooling or oxygenation with respect to both number and quality of isolated atrial cardiomyocytes. In general transport to the lab should be as quick as possible. However, if longer transportation time cannot be avoided, transport at 4 °C in oxygenated solution might be advantageous.

1. Prearrangements

1. Switch on the thermocirculator, which controls the temperature of the jacketed beaker (Table II). Make sure the temperature within the beaker equals 37 °C. Cover the beaker with a glass lid to maintain a constant temperature within the beaker throughout the whole isolation process.
2. Weigh the Collagenase I and Protease XXIV into three glass beakers and store it at room temperature. The enzymes will be diluted just before use.
3. Store 20 ml of Ca²⁺ free solution in a Petri dish at 4 °C.
4. Store 250 ml of Ca²⁺ free solution in a water bath at 37 °C.

2. Cleaning of the Tissue

1. Transfer the tissue sample together with the transport solution into a Petri dish (~10 cm diameter) and remove fatty tissue roughly using scissors.
2. Weigh the tissue sample. Between 200 and 600 mg are used for cell isolation. The remaining tissue can be frozen in liquid nitrogen for later biochemical analysis.
3. Transfer the tissue sample into the Petri dish containing 20 ml Ca²⁺ free solution at 4 °C (see step 1.3) and chop the tissue sample into small chunks of approximately 1 mm³ in size.
4. The following steps should be executed at 37 °C and under continuous gassing with 100% O₂. A pipette tip can be placed onto the oxygen tube to avoid strong bubbling and generate continuous air flow.
5. Transfer the tissue chunks together with the Ca²⁺-free solution into the jacketed beaker and stir carefully for about 3 min with a magnetic stirring bar.
6. Stop stirring and allow the tissue chunks to settle down for a few seconds. Carefully strain the supernatant through a nylon mesh (200 μm, Table II). Return the restrained tissue chunks from the mesh into the beaker using forceps.
7. Refill the beaker with Ca²⁺ free solution and repeat the washing step 2.4 and 2.5 two times.

3. First Enzymatic Digestion

1. Re-suspend the tissue chunks with 20 ml Enzyme solution E1 (Table I) containing collagenase and protease and stir carefully for 10 min.
2. Add 40 μl of 10 mM CaCl₂-solution to obtain a final concentration of 20 μM Ca²⁺.
3. After 35 min strain the supernatant carefully through a nylon mesh (200 μm, Table II) in a way that most of the tissue chunks remain in the beaker. Return restrained tissue chunks from the mesh into the beaker.

4. Second Enzymatic Digestion

1. Resuspend the tissue chunks again with 20 ml enzyme solution E2 containing collagenase I only (Table I). Add 40 μl of 10 mM CaCl₂-solution immediately to obtain a final concentration of 20 μM Ca²⁺.
2. During the second digestion use scissors to further chop tissue clots occasionally.
3. After 5 min take a sample using a Pasteur pipette to check the dissociation of cells. Repeat this every 2-3 min until rod-shaped, striated cardiomyocytes appear.
4. Stop stirring and allow the tissue chunks to settle down for about 20-30 sec.
5. Strain the supernatant carefully through a nylon mesh (200 μm, Table II) into a 50 ml Falcon tube (Tube A). Return the restrained tissue chunks from the mesh into the beaker.
6. Re-suspend the tissue chunks in the beaker with 20 ml storage solution (Table I)¹⁰ and further dissociate the cells by gentle mechanical trituration using a 20 ml serological pipette with dispenser. Try to avoid bubble formation since bubbles are detrimental to cell survival and quality.
7. Strain the supernatant carefully through a nylon mesh (200 μm, Table II) into a 50 ml Falcon Tube (Tube B).

5. Final Preparation and Adjustment of Final Ca²⁺ Concentration

1. Centrifuge both Falcon tubes A and B (see step 4.5 and 4.7) at 95 x g for 10 min.
2. Remove the supernatant from both tubes carefully by aspiration. Make sure not to disturb the pellet. Discard the supernatant.
3. Re-suspend both pellets in 1.5 ml storage solution (Table I) each (room temperature).
4. Add two times 7.5 μl of 10 mM CaCl₂ solution to each Falcon and stir carefully. Incubate for 10 min after each step.
5. Add 15 μl of 10 mM CaCl₂ solution to obtain a final Ca²⁺ concentration of 0.2 mM.

6. Loading of Myocytes with the Fluorescent Ca²⁺-indicator Fluo-3 AM (Figure 1)

1. Transfer 1.5 ml of cell suspension (tube A and/or tube B) into a 2 ml microcentrifuge tube (Eppendorf tube).
2. The following steps should be executed under consideration of the light sensitivity of the fluorescent Ca²⁺ Indicator Fluo-3.
3. Dissolve 50 μg of the membrane permeable acetoxymethyl ester derivative of Fluo-3 (Fluo-3 AM, Table III) in 44 μl of the Pluronic F-127 (Table III) stock solution (20% w/v in anhydrous DMSO) to get a 1 mM Fluo-3 AM stock solution, which can be stored at -20 °C for at most 1 week.
4. Add 15 μl of the Fluo-3 AM stock solution to the microcentrifuge tube containing 1.5 ml of cell suspension (see step 6.1) and agitate carefully.
5. Incubate the cell suspension for 10 min in an optically opaque box.
6. Briefly centrifuge at about 6,000 rpm.
7. Discard the supernatant and re-suspend the pellet in 1.5 ml bath solution (Table IV).
8. Leave the cell suspension for about 30 min for de-esterification before beginning with experiments.

7. Simultaneous Patch-clamp and Epifluorescent Ca²⁺ Measurements

Since patch-clamp measurements are not the major topic of this review, we refer the interested reader to other publications providing a more in depth description of this technique.^11-14 For the sake of completeness we provide a brief summary of a protocol to measure action potentials or L-type Ca²⁺ currents, both together with simultaneous Ca²⁺-transient recordings.

During experiments myocytes are superfused at 37 °C with bath solution (Table IV) using a rapid perfusion system (Octaflow IITM, ALA Scientific Instruments, NY). For voltage-clamp experiments, K⁺ currents are blocked by adding 4-aminopyridine (5 mmol/L) and BaCl₂ (0.1 mmol/L) to the bath solution. Borosilicate glass microelectrodes are used and should have tip resistances of 2-5 MΩ when filled with pipette solution (Table V). In addition to the Fluo-3 AM loading of the myocytes (see step 6), Fluo-3 is also included in the pipette solution (Table V). Fluorescence is excited at 488 nm and emitted light (<520 nm) converted to [Ca²⁺]_i assuming

where k_d=dissociation constant of Fluo-3 (864 nM), F=Fluo-3 fluorescence; F_max=Ca²⁺-saturated fluorescence obtained at the end of each experiment.¹² Both electrical signals and epifluorescent Ca²⁺ signals are recorded simultaneously. Action potentials are stimulated at 0.5 Hz in current-clamp mode using 1 msec current pulses of 1.2x threshold strength. L-type Ca²⁺-currents are measured in voltage-clamp mode using a holding potential of -80 mV and a 100-msec ramp-pulse to -40 mV to inactivate the fast Na⁺-current, followed by a 100-msec test-pulse to +10 mV at 0.5 Hz.

Figure 2A shows three representative examples from isolated human right atrial myocytes. To quantify the cell yield we pipette 10 μl of cell suspension (step 5.5) on a CellFinder microscope slide (http://www.antenna.nl/microlab/index-uk.html). Averaged cell yields in Figure 2B clearly indicate that there is a tendency to lower cell yields in chronic AF (cAF) patient samples (tube A: 16.5±3.1 cells/10 μl (n=29) vs. 5.1±2.3 cells/10 μl (n=10) in SR and cAF, respectively, p<0.05; tube B: 17.9±3.9 cells/10 μl (n=29) vs. 5.9±2.0 cells/10 μl (n=9) in SR and cAF, respectively, p=0.107).

Representative examples of action-potential measurements and simultaneous recordings of cytosolic Ca²⁺ transients are given in Figure 3. In about 90% of the investigated cells, the action-potential-triggered Ca²⁺ release causes clear and regular cell contractions. As reported previously, the resting membrane potential, which is an accepted indicator for cell integrity, averaged about -73.9±2.7 mV (n = 23/10 myocytes/patients) and -77.7±1.8 mV (n = 19/8 myocytes/patients) in SR and cAF respectively (p>0.05).¹⁵ Figure 4 shows representative simultaneous recordings of voltage-gated L-type Ca²⁺ currents and cytosolic Ca²⁺ transients. Application of the non-selective β-adrenoceptor agonist isoprenaline (1 μM) increases amplitudes of both I_Ca,L and cytosolic Ca²⁺ transients, suggesting intact β-adrenergic signal transduction cascade.

Figure 1. Flow chart of the myocytes Fluo-3 AM loading protocol (see step 6.1-6.5). m/v, mass/volume.

Figure 2. A, Isolated human right atrial myocytes after one hour in storage solution. B, Mean±SEM of the cell yield counted in 10 μl of cells in storage solution (see step 5.5). n refers to the number of preparations within each group. *p<0.05.

Figure 3. Representative recordings of action-potential-triggered Ca²⁺-transients (CaT) in an atrial myocyte from a sinus rhythm and a chronic atrial fibrillation patient. Top: Injected membrane current (I_M) used for stimulation (0.5 Hz). Below: Simultaneous recording of membrane potential (V_M), and triggered CaT (bottom). (Replotted with permission from Voigt et al. 2012)¹⁵

Figure 4. Representative recordings of the isoprenaline (1 μM) effect on L-type Ca²⁺ current-triggered Ca²⁺-transients (CaT) in an atrial myocyte from a sinus rhythm and a chronic atrial fibrillation patient. Top: Voltage-clamp protocol (0.5 Hz). Below: Simultaneous recording of total net membrane current (I_M), predominantly reflecting L-type Ca²⁺ current (middle) and triggered CaT (bottom). (Replotted with permission from Voigt, et al. 2012)¹⁵

Table I. Solutions.

Table II. Specific equipment.

Table III. Substances for loading of myocytes with Fluo-3 AM.

Table IV. Bath solution for patch-clamp.

Table V. Pipette solution for patch-clamp*.

Here we describe a method for isolation of human atrial myocytes from right atrial appendages obtained from patients undergoing open heart surgery. In order to use these myocytes for measurements of cytosolic Ca²⁺ we adapted a previously described method^4-11 by omitting EGTA from the storage solution.

Already in 1970 it was observed that although myocytes dissociate in the presence of Ca²⁺ during digestion, all of them were in contracture and non-viable.^16,17 Therefore, cell isolation is performed in Ca²⁺-free solution. However, the re-introduction of physiological concentrations of Ca²⁺ resulted in rapid Ca²⁺ influx and cell death. This has been described as the Ca²⁺ paradox phenomenon which was originally observed in perfused hearts by Zimmerman and Hulsman.¹⁸ Modifications of the isolation media including reduction of the pH to 7.0,¹⁹ addition of taurin²⁰ or of small amounts of Ca²⁺ (see step 3.2 and 4.1),²¹ as well as storage of isolated myocytes in EGTA containing storage-solution²² have been suggested to prevent the Ca²⁺ paradox.¹⁷ However, it is well known that Ca²⁺ buffering through EGTA reduces the amplitude of L-type Ca²⁺ current-induced Ca²⁺ transient amplitudes and results in a biphasic decay of the Ca²⁺ transients.²³ Therefore, we omitted EGTA throughout the whole isolation process in order to obtain Ca²⁺ transients with typical properties and monophasic decays. To protect the cells from the Ca²⁺ paradox we increased the final Ca²⁺ concentration of the storage solution in a stepwise manner until 0.2 mM.

The choice of collagenase is probably the most critical step for successful myocyte isolation. Conventional collagenases are crude preparations obtained from Clostridium histolyticum and contain collagenase in addition to a number of other proteinases, polysaccharidases and lipases. Based on their general composition collagenases are subdivided into different types.²⁴ Worthington collagenase Types I and II have been successfully used for isolation of human atrial myocytes.^{4-10,15,25-30} In our presently described protocol we recommend the use of collagenase Type I, although we were also able to obtain acceptable amounts of viable cells using collagenase Type II. However, even within a single collagenase type there is a significant batch-to-batch variation regarding the enzyme activities. These variations require careful batch selection and testing of various batches to optimize isolation procedure. The online available batch-selection tool from Worthington Biochemical Corp. (http://www.worthington-biochem.com/cls/match.php) may be used to find available batches with a composition that has been shown to be suitable for the isolation of human atrial myocytes. Currently we use collagenase type I with 250 U/mg collagenase activity, 345 U/mg caseinase activity, 2.16 U/mg clostripain activity and 0.48 U/mg tryptic activity (lot# 49H11338).

The cells obtained using the procedure described in this manuscript may be used within 8 hr for patch-clamp studies, Ca²⁺ transient measurements and a combination of both.¹⁵ In addition, these cells allow measurements of cellular contraction in response to electric field stimulation or electric stimulation using the patch-clamp pipette (unpublished observations).