Heart and Circulatory Physiology

Validation of formamide as a detubulation agent in isolated rat cardiac cells

Fabien Brette, Kimiaki Komukai, Clive H. Orchard


Kawai M, Hussain M, and Orchard CH.Am J Heart Circ Physiol 277: H603–H609, 1999 developed a technique to detubulate rat ventricular myocytes using formamide and showed that detubulation results in a decrease in cell capacitance, Ca2+ current density, and Ca2+ transient amplitude. We have investigated the mechanism of this detubulation and possible direct effects of formamide. Staining ventricular cells with di-8-ANEPPS showed that the t tubule membranes remain inside the cell after detubulation; trapping of FITC-labeled dextran within the t tubules showed that detubulation occurs during formamide washout and that the t tubules appear to reseal within the cell. Detubulation had no effect on the microtubule network but resulted in loss of synchronous Ca2+ release on electrical stimulation. In contrast, formamide treatment of atrial cells did not significantly change cell capacitance, Ca2+ current amplitude, action potential configuration, the Ca2+ transient or the response of the Ca2+ transient to isoprenaline. We conclude that formamide washout induces detubulation of single rat ventricular myocytes, leaving the t tubules within the cell, but without direct effects on cell proteins that might alter cell function.

  • transverse tubules
  • rat ventricular cell
  • di-8-ANEPPS
  • rat atrial cell
  • Ca2+ current
  • Ca2+ transient

the sarcolemma of mammalian cardiac ventricular myocytes has invaginations, called t tubules, which occur perpendicular to the longitudinal axis of the cell at intervals of ≈2 μm (19). Interest in the physiological role of the t tubules has grown in the last few years, although evidence for their function is still sparse. One approach to the investigation of their function has been to use confocal microscopy which, used in conjunction with Ca2+ indicators, has shown that Ca2+ sparks arise more frequently around the t tubules than elsewhere in the cell (21), and immunocytochemistry has shown that several key proteins involved in excitation-contraction coupling are concentrated at the t tubules (4), suggesting an important role for this structure in excitation-contraction coupling (20). Another approach has been to compare data from ventricular myocytes with data from preparations that lack t tubules, e.g., atrial cells (15), Purkinje cells (6), or cultured ventricular cells, which lose their t tubules after 1–3 days (18). However, differences in protein expression in these systems may also contribute to observed differences.

In 1999, Kawai et al. (14) developed a new technique in their laboratory to investigate the role of the t tubules by inducing acute detubulation of freshly isolated rat ventricular myocytes. They showed that incubation of these cells in 1.5 mol/l formamide (a membrane permeant agent) for 15 min, followed by washout, induces loss of the t tubules, confirmed by staining the cell membrane with the lipophilic dye di-8-ANEPPS. They also showed that this procedure decreased Ca2+ current density by ≈75%, decreased the amplitude of the Ca2+ transient, and decreased cell capacitance by ∼26%, a value close to the proportion of the cell membrane in the t tubules determined with electron microscopy (19).

However, several questions remain. First, how does detubulation occur? Second, what are the functional consequences of detubulation on the spatial and temporal distribution of the Ca2+ transient? Third, does formamide have direct effects that might account for the observed changes? The present study was designed to address these questions.


Isolation of single rat myocytes.

Myocytes were isolated from rat hearts according to the method of Boyett et al. (3), which enabled atrial and ventricular myocytes to be isolated from the same heart. Male adult Wistar rats (250–300 g) were euthanized by stunning, followed by cervical dislocation, in accordance with the Home Office Guidance on the Operation of the Animals (Scientific Procedures) Act of 1986. The heart was quickly removed and mounted on a Langendorff apparatus and retrogradely perfused with isolation solution containing (in mmol/l) 130 NaCl, 5.4 KCl, 0.4 NaH2PO4, 1.4 MgCl2 · 6 H2O, 0.5 CaCl2, 10 HEPES, 10 glucose, 20 taurine, and 10 creatine, pH set to 7.3 using NaOH. When the coronary circulation had cleared of blood, perfusion was continued with Ca2+-free isolation solution (in which CaCl2 had been replaced with 0.1 mM EGTA) for 4 min, followed by perfusion for a further 10 min with Ca2+-free isolation solution containing 0.8 mg/ml collagenase (type I; Worthington Biochemical; Lakewood, NJ) and 0.1 mg/ml protease (type XIV; Sigma; St. Louis, MO). The atria and ventricles were then excised from the heart, minced, and gently shaken at 37°C in collagenase-containing solution supplemented with 1% bovine serum albumin. Atrial cells were filtered from this solution at 15-min intervals and resuspended in isolation solution containing 0.5 mM Ca2+. Ventricular cells were filtered from this solution at 5-min intervals and resuspended in isolation solution containing 0.5 mM Ca2+. All experiments were performed at room temperature (22–25°C).

Visualization of the cell membrane and incubation in FITC-dextran.

To stain the cell membrane, cells were incubated with the lipophilic fluorescent indicator di-8-ANEPPS (5 μM; Molecular Probes) for 2 min (14). The cells were then resuspended in control solution and imaged using confocal laser scanning microscopy (Zeiss LSM 510) using 488-nm excitation light with detection at >505 nm. Representative confocal images are shown in thex-y plane at the level of the nucleus or, in the absence of clear staining, midway through the cell. The periodicity of staining within the cell was investigated using Zeiss LSM 5 Image Examiner software version 2.81 (Zeiss, Germany) to obtain the fluorescence profile along a line drawn on the long axis of the cell (see Fig. 1); the cross-striation (t tubule) interval was determined by applying a fast Fourier transformation algorithm (Origin) to these profiles. Images in thex-z plane were constructed from a series ofx-y images taken through the depth of the cell (az-stack) by using the Zeiss LSM 5 Image Examiner Software V2.81 (Zeiss, Germany).

Fig. 1.

The effect of detubulation on t-tubule structure of ventricular myocytes. Top, representative x-y and x-zconfocal images of rat ventricular cells stained with the lipophilic dye di-8-ANEPPS. Middle, fluorescence intensity (FI) profile along dotted line indicated on the longitudinal axis of each cell at top in arbitrary units (AU). Bottom, the results of a fast Fourier transformation of the fluorescence profile.A: control ventricular cell is shown; the t tubule network is clearly visible. Similar staining was obtained in 31 cells. The fluorescence profile shows peaks at ≈2-μm intervals; for this cell, the fast Fourier transform gives a peak at 0.5183 μm−1i.e., 1.93 μm. B: rat ventricular cell stained after formamide treatment. Staining is visible predominantly on the surface membrane (top and middle; n = 28/34), and the fast Fourier transformation does not show a peak.C: rat ventricular cell that was first stained and then detubulated. The confocal image shows staining in the interior of the cell (n = 36/47) but with no clear periodicity (middle and bottom).

In some experiments, FITC-labeled dextran (10,000 mol wt, 50–100 μg/ml, Molecular Probes) was present in the solutions used for detubulation. Protocols are given in results.

Immunolabeling of microtubules.

The microtubules were labeled with monoclonal antibody to β-tubulin (1:200; Sigma) and FITC-conjugated donkey anti-mouse IgG (1:50; Jackson ImmunoResearch), as described by Howarth et al. (9). Confocal microscopy was used to obtain 2-μm optical sections at the level of the nucleus. The same settings of the confocal microscope were used for imaging control and formamide-treated cells. Analysis was performed using Scion Image for Windows software (Scion) by selecting an area of the cell and measuring mean fluorescence.

Ca2+ imaging.

Ventricular myocytes were loaded with 10 μM fluo 3-AM for 20 min at room temperature, followed by 30 min for deesterification, and then transferred to a bath mounted on the stage of a laser scanning confocal microscope (LSM 510, Zeiss, Germany). The cells were superfused with control solution (see below) and stimulated at 0.5 Hz by 2- to 4-ms suprathreshold rectangular voltage pulses via a pair of extracellular platinum electrodes. Fluo 3 was excited at 488 nm and emitted fluorescence collected at wavelengths >505 nm. The confocal slit aperture was set so that the confocal plane was <1 μm with a ×63 oil immersion objective lens (Plan-Neofluar, numerical aperture 1.2, Zeiss). Line scans were performed repetitively (1,000 lines of 512 pixels) at 5-ms intervals across the width of the cell. Image analysis was performed off-line using Zeiss LSM 5 Image Examiner Software V2.81 (Zeiss, Germany). Linescan images (8-bit) are presented as the original signal. Traces showing the time course of fluorescence are presented as a ratio of fluorescence/background fluorescence (F/F0). Analysis of Ca2+ transients recorded from the subsarcolemmal space (SS) and cell center (CC) was performed using Origin software. The initial rate of rise of the fluo 3 transient was calculated by fitting a linear function to the upstroke and is expressed as change in fluorescence intensity per ms (dF/dt).

Electrophysiological recording.

Currents and voltages were recorded from rat atrial cells using the whole cell configuration of the patch-clamp technique. Pipettes were made from borosilicate glass tubing (GC150TF-15, Clark Electromedical Instrument) with a vertical puller (PP-83, Narashige) and were slightly fire polished (MF-83, Narashige). When filled with the pipette solution (see below), pipette resistance was 1–2.5 MΩ. Junction potentials between the pipette solution and the reference electrode were cancelled before obtaining a tight gigaseal (>1 GΩ). For measurement of membrane current, cell capacitance and series resistance were compensated (60–80%). Cell membrane capacitance was measured by integrating the capacitance current recorded during a 10-mV hyperpolarizing pulse from a holding potential of −80 mV.

The voltage-clamp amplifier was an Axopatch 1D (Axon Instruments), which was controlled by a Pentium personal computer connected through a CED 1401plus A/D interface (Cambridge Electronic Design), which was also used for data acquisition and analysis with Signal software (Cambridge Electronic Design). Signals were filtered at 2 kHz using an eight-pole Bessel low-pass filter before digitization at 10 kHz and storage.

L-type Ca2+ current (I Ca,L)-voltage (V) curves were obtained using 300-ms depolarizing pulses from a holding potential of −80 mV, to voltages between −40 and +50 mV, in 5-mV increments, at a frequency of 0.125 Hz. A 50- or 100-ms prepulse to −40 mV was used to inactivate sodium current and T-type Ca2+ current. I Ca,L was measured as the difference between the peak inward current and the current at the end of the depolarizing pulse currents are expressed as current density (pA/pF). The pipette solution used for these experiments was the same as that used by Kawai et al. (14) and contained (in mmol/l) 120.0 CsCl, 20.0 KCl, 10.0 NaCl, 5.0 Mg-ATP, 1.0 BAPTA, and 10.0 HEPES, adjusted to pH 7.2 with CsOH. Cs-based pipette solutions and 5 mmol/l CsCl in the external solution (instead of KCl) were used to avoid contamination ofI Ca,L by K+ currents.

Action potentials were elicited by 2-ms current steps just above threshold and were acquired using the current clamp facility of the Axopatch 1D patch-clamp amplifier and Signal software. The pipette solution for these experiments contained (in mmol/l) 130.0 KCl, 3 Mg-ATP, 0.4 Na-GTP, 10 EGTA, and 25 HEPES, adjusted to pH 7.2 with KOH. Stimulation frequency was 0.33 Hz. Analysis was performed using Origin software (Microcal).

Recording Ca2+ transients in atrial cells.

Rat atrial cells were loaded with the Ca2+-sensitive fluorescent indicator fura 2-AM (3 μM; Molecular Probes) for 10 min at room temperature. Cells were electrically field stimulated at 0.4 Hz. The ratio of fluorescence emitted at 510 nm in response to alternate excitation with 340- and 380-nm light was used as an index of intracellular Ca2+. Analysis was performed using Signal software (Cambridge Electronic Design) by averaging 20 signals at steady state.


The control bathing solution used in these experiments contained (in mmol/l) 113.0 NaCl, 5.0 KCl, 1.0 MgSO4, 1.0 CaCl2, 1.0 Na2HPO4, 20.0 sodium acetate, 10.0 glucose, and 10.0 HEPES and 5.0 U/l insulin, pH adjusted to 7.4 with NaOH. Acetate was included in the bathing solution because it is an important metabolic substrate for cardiac muscle (1). To induce detubulation by osmotic shock, cells were exposed to this solution plus formamide (1.5 mol/l) for 15–20 min, before returning the cells to control solution, as described by Kawai et al. (14). All solutions were made using ultrapure water supplied by a Milli-Q system (Millipore). All solution constituents were reagent grade and were purchased from Sigma unless stated otherwise. Di-8-ANEPPS was kept as a stock solution of 1.0 mmol/l in DMSO plus 20% pluronic acid, which was added to the bathing solution. Isoprenaline was added directly to the bathing solution from an ampoule (Guy's Hospital; London, UK) to give a final concentration of 0.5 μmol/l.


Data are presented as means ± SE. Unpaired t-tests were used to compare data from control and formamide-treated cells; paired t-tests were used to test the effect of isoprenaline. For the microtubule confocal data, which are not normally distributed (10), the Mann-Whitney rank sum test was used.P < 0.05 was taken as significant.


Formamide treatment induces detubulation of rat ventricular myocytes.

Figure 1 A, top, shows a confocal image of a representative control cell stained with di-8-ANEPPS. Thex-y image clearly shows the t tubule network and is representative of 31/31 cells. The x-zimage, reconstructed from a z-stack ofx-y images (see materials and methods), shows that the t tubule network is present throughout the depth of the cell. A plot of the fluorescence intensity along the longitudinal axis of the cell (Fig. 1 A, middle) shows peaks of fluorescence at ≈2-μm intervals; a fast Fourier transformation of this plot (Fig. 1 A, bottom) in 17 cells gave an interpeak interval of 0.5386 ± 0.0091 μm−1, i.e., 1.86 ± 0.03 μm.

Figure 1 B, top, shows that cells stained with di-8-ANEPPS after treatment with formamide (14) show staining of the surface membrane but little staining within the cell. The fluorescence versus distance plot (Fig. 1 B,middle) clearly shows that only surface membrane staining is present; little staining is present in the interior of the cell, and the fast Fourier transformation (Fig. 1 B, bottom) shows no periodicity in the fluorescence profile. This result is representative of 28/34 cells. Similar results were obtained in cells stained for 5 h (n = 6/6) or 6 h (n = 4/4) after detubulation, indicating that the detubulation is not reversed during this time.

T tubules remain inside cell after detubulation.

To investigate what happens to the t tubules during detubulation, we exposed the cells to formamide after the membrane had been stained with di-8-ANEPPS. Figure 1 C, top, shows a representative image from a cell after this procedure, showing staining on the cell surface and within the cell. However, the intracellular staining is not as well organized as in control cells (compare Fig.1 C, which is representative of 36/47 cells, with Fig.1 A): the profile of fluorescence versus distance shows that fluorescence is present within the cell (Fig. 1 C,middle) but shows no periodicity (Fig. 1 C,bottom; n = 11).

Thus it appears that the t tubule membrane remains inside the cell after detubulation. To investigate whether the t tubules form vacuoles inside the cell, as described after detubulation in skeletal muscle (16), 50–100 μg/ml FITC-dextran was added to the bathing solution for different parts of the detubulation procedure. Figure 2 A shows a confocal image of a control cell bathed in control solution containing FITC-dextran for 2 h, showing the this dextran does not cross the cell membrane: fluorescence can only be seen in the extracellular solution, not within the cell (n = 5/5). Figure2 B shows a confocal image of a representative cell that had undergone the detubulation procedure with FITC-dextran (seematerials and methods) present in all the solutions used, followed by resuspension in dextran-free solution 30 min after the removal of formamide. Fluorescence is clearly visible within the cell (n = 12/15) indicating FITC-dextran trapped within the cell after detubulation; the fluorescence appears to be localized within the cell, but showing a longitudinal pattern, rather than the transverse pattern shown by the t tubules in control cells (Fig.1 A). When this experiment was repeated, except that formamide was washed out with a FITC-dextran-free solution, no fluorescence was visible within the cell (Fig. 2 C,n = 10/10). Thus it appears that the FITC dextran becomes trapped within the cell on washout of formamide and thus that detubulation occurs at this point in the protocol.

Fig. 2.

The accessibility of FITC-dextran to the interior of ventricular myocytes. A: representative image of a ventricular myocyte that was bathed for 2 h in FITC-dextran. Fluorescence can be seen in the solution but not within the cell; similar results were obtained in 5 cells. B: representative image of a ventricular myocyte in which the formamide-containing and washout solutions contained FITC-dextran, showing that fluorescence is visible within the cell (n = 12/15). C: representative image of a ventricular myocyte in which the formamide-containing solution contained FITC-dextran, but the washout solution was dextran free, showing that no fluorescence is present within the cell (or the solution); similar results were obtained in 10 cells. Scale bar is 10 μm for AC.

Microtubule network remains unchanged after detubulation.

The cytoskeleton plays an important role in cell structure and in the localization of proteins (23). Thus it seemed possible that the observed changes could be secondary to changes in the cytoskeleton. We therefore stained the microtubule network in control and formamide-treated cells using a specific antibody to β-tubulin and a FITC-conjugated secondary antibody (see materials and methods). Figure 3 A,left, shows a confocal image of the microtubule network in a representative control cell (n = 116). The pattern of staining appeared the same after detubulation (n = 107; Fig. 3 A, right), and the mean fluorescence signal was not significantly different between control and detubulated cells [17.27 ± 1.37 arbitrary units (AU) in control, n= 116, vs. 17.56 ± 1.35 AU in detubulated cells,n = 107; not significant (NS); Fig. 3 B]. In addition, cell area, which is measured on confocal images through the center of the cells, was not significantly different after detubulation (2,681.9 ± 106.8 μm2 in control cells, n= 53, vs. 2,632.6 ± 71.1 μm2 in detubulated cells, n = 63; NS).

Fig. 3.

The effect of detubulation on the organization and density of microtubules in ventricular myocytes. A: immunofluorescence confocal images of sections taken at the level of the nuclei in representative control (left) and detubulated (right) ventricular myocytes labeled with an antibody to β-tubulin, showing no apparent difference in the organization of the microtubule network. Scale bar is 20 μm. B: fluorescence intensity (normalized to cell section area) of β-tubulin staining in control (open bar) and detubulated (hatched bar) ventricular myocytes. Means ± SE, n = 116 and 107 observations, respectively [not significant (NS)]. AU, arbitrary units.

Effect of detubulation on intracellular Ca2+ distribution in ventricular myocytes.

Because I Ca is concentrated in the t tubules (see Introduction), detubulation would be expected to alter Ca2+ distribution after electrical stimulation. Figure4 A shows transverse line scan images from a representative control cell (left) and detubulated cell (right) during electrical stimulation. In the control cell, electrical stimulation resulted in a rapid and synchronous rise in intracellular Ca2+ concentration ([Ca2+]i) across the width of the myocyte; this is clearly shown by the Ca2+ transients recorded from the SS and CC (Fig. 4 B, left), which have almost identical amplitudes and time courses (F/F0: SS, 2.7 ± 0.2; CC, 2.7 ± 0.6; NS and dF/dt: SS, 3.4 ± 1.0; CC, 3.3 ± 1.0 F U/ms; NS, n = 7). In contrast, in the detubulated cell, Ca2+ initially increases at the edge of the cell and then propagates to the center of the cell at 46 ± 6 μm/s (n = 10). Figure 4 B,right, shows that Ca2+ recorded from the SS rises earlier and more rapidly than that recorded from the CC (dF/dt: SS, 3.5 ± 0.3; CC, 0.6 ± 0.1 F U/ms;P < 0.05; n = 10), although peak systolic [Ca2+] was the same in SS and CC (F/F0: SS, 2.2 ± 0.2; CC, 1.9 ± 0.1; NS;n = 10).

Fig. 4.

Transverse line scans recorded from representative control (left) and detubulated (right) myocytes.A: 8-bit line scan images of fluo-3 fluorescence before, during, and after electrical stimulation. B: time course of the ratio of fluorescence to background fluorescence (F/F0) at the cell center (CC) and subsarcolemmal space (SS) of the control and detubulated ventricular cells shown in A. C: whole cell Ca2+ transient obtained from the line scans shown in A.

The amplitude of the Ca2+ transient was significantly (P < 0.05) smaller in detubulated cells than in control cells (Fig. 4, B and C), as reported previously (14). However, the rate of rise of Ca2+ in SS was not statistically different in control and detubulated cells (3.4 ± 1.0 and 3.5 ± 0.3 F U/ms, respectively; NS) but in CC was significantly faster in control cells than in detubulated cells (3.3 ± 1.0 and 0.6 ± 0.1 F U/ms, respectively; P < 0.05).

Figure 4 C shows global Ca2+ transients obtained from the whole scan. In contrast to the Ca2+ transient recorded from the control cell, in 7 of 10 detubulated cells, the Ca2+ transient showed a biphasic rising phase, as reported by Kawai et al. (14), as a consequence of the asynchronous rise of Ca2+ across the width of the cell.

Atrial cells lack t tubules.

To investigate the possible direct effects of formamide, the detubulation procedure was performed on atrial cells, which lack t tubules (24): this is confirmed in Fig.5 A, which shows a confocal image of a representative rat atrial myocyte stained with di-8-ANNEPS, showing the absence of a t tubule network (n = 22). A similar result was obtained after incubation of atrial cells with formamide (15–20 min), followed by washout (Fig. 5 B,n = 20), although measurements of cell length and width during addition and removal of formamide showed that atrial cells undergo the same changes in size as described for ventricular cells by Kawai et al. (14) (n = 3, data not shown).

Fig. 5.

The effect of formamide treatment on membrane structure of atrial myocytes. Representative confocal X-Y images of rat atrial myocytes stained with the lipophilic dye di-8-ANEPPS. A: control atrial cell; no t tubule network is visible. Similar staining was obtained in 22 cells. B: formamide-treated (FT) atrial myocyte. The same staining as control is observed (n = 20). Scale bar is 20 μm.

Effects of formamide treatment on electrophysiological parameters of atrial myocytes.

We repeated in atrial myocytes the experiments that Kawai et al. (14) carried out in ventricular myocytes. The results are shown in Fig. 6. Cell capacitance measured with the whole cell configuration of the patch-clamp technique did not change significantly after formamide treatment (Fig.6 A; 48.6 ± 2.2 pF in 38 control cells vs. 46.1 ± 2.4 pF in 26 formamide-treated cells, NS). TheI Ca-V curve, measured as described by Kawai et al. (14), did not change significantly after formamide treatment of atrial cells (Fig.6 B): I Ca density during a test pulse to −5 mV was −2.34 ± 0.39 pA/pF in control cells (n = 10) versus −2.50 ± 0.47 pA/pF in formamide-treated cells (n = 8; NS). These data are different from those obtained in rat ventricular cells in which formamide treatment significantly decreased cell capacitance andI Ca density (see Introduction).

Fig. 6.

The effect of formamide treatment on the electrophysiological properties of atrial myocytes. Control and FT atrial myocytes were studied using the patch-clamp technique. A: mean data for cell capacitance: open bar represents control atrial myocytes (n = 38), the hatched bar represents FT atrial myocytes (n = 26). B: means ± SE Ca2+ current (I Ca)-voltage relationship for control (n = 10) and FT (n = 8) cells, obtained as described in materials and methods. C: individual traces of action potentials from a representative control cell (left) and FT (right) cell. See Table 1 for mean values.

To investigate whether formamide treatment affects other physiologically important currents, we recorded action potentials from single atrial rat cells before and after exposure to formamide. Figure6 C shows action potentials recorded from a representative control atrial cell (left) and a representative formamide-treated atrial cell (right). Mean data from 10 cells in each group show that the resting membrane potential, action potential amplitude, and times for repolarization of the action potential (APD25, APD50, and APD90) were not significantly different in control and formamide-treated cells (Table 1).

View this table:
Table 1.

Effect of formamide treatment on action potential parameters in atrial myocytes

Effects of formamide treatment on Ca2+ cycling in atrial myocytes.

To investigate further the possible direct effects of formamide on proteins and cell function, we recorded Ca2+ transients from control and formamide-treated atrial cells, in the absence and presence of the β-adrenergic agonist isoprenaline (0.5 μmol/l). Figure 7 A shows representative Ca2+ transients recorded with the fluorescent dye fura 2 from control (left) and formamide-treated (right) atrial cells in the absence and presence of isoprenaline. The amplitude of the Ca2+ transient was not significantly different in control and formamide-treated cells [0.0607 ± 0.0093 ratio units (RU) for control cells,n = 11, vs. 0.0622 ± 0.0151 in formamide-treated cells, n = 8, NS; Fig. 7 B]. Similarly, treatment with formamide had no significant effect on the half-time of decay of the Ca2+ transient (T ½), which was 371 ± 34 ms in control cells, n = 11, versus 379 ± 34 ms in formamide-treated cells, n = 8; NS (Fig.7 C). This is different from the result obtained in rat ventricular cells by Kawai et al. (14) and reported above, in which formamide treatment significantly decreased Ca2+transient amplitude (see Introduction).

Fig. 7.

Effect of formamide on Ca2+ cycling in rat atrial cells. A: representative records of Ca2+transients from control (left) and FT (right) cells under in the absence (○) and presence (●) of isoprenaline (Iso). B: mean data for Ca2+ transient amplitude. C: mean data for the half time of decay of the Ca2+ transient (T ½). Open bars represent control data, and hatched bars represent data from FT cells. All bars are means ± SE of 11 control cells and 8 FT cells. *P < 0.01, Iso vs. control. Formamide treatment had no significant effect.

Isoprenaline-induced a significant inotropic and lusitropic effect in both control and formamide-treated cells (P < 0.01, Fig. 7, B and C, respectively), but there was no significant difference in the amplitude orT ½ of the Ca2+ transient between the two populations of cells in the presence of isoprenaline (amplitude was 0.0986 ± 0.0199 RU in control cells,n = 11, vs. 0.0962 ± 0.0205 RU in formamide-treated cells, n = 8; NS.T ½ was 306 ± 30 ms in control cells, n = 11, vs. 312 ± 42 ms in formamide-treated cells, n = 8; NS).


Formamide-induced detubulation of rat ventricular myocytes.

The present study shows that treatment of rat ventricular myocytes with formamide produces disruption of the t tubules, in agreement with the report of Kawai et al. (14). This detubulation was confirmed with the lipophilic dye di-8-ANNEPS. This dye is retained in the outer leaflet of the plasma membrane and does not cross the cell membrane (8); the observation that t-tubular staining was not observed in formamide-treated cells, despite the presence of t-tubular membranes within these cells (see below) supports this idea. Thus the membranes stained by the dye are those accessible from the extracellular space. The loss of t-tubular staining after treatment with formamide suggests, therefore, that the t tubules are no longer open to the extracellular space. This disruption was maintained for at least 6 h.

The present experiments help elucidate the mechanism and nature of this disruption. The use of FITC-labeled dextran in the extracellular solution showed that this dextran became trapped within the cell only when it was present in the solution used to wash out formamide (Fig.2). This agrees with previous suggestions (14) that it is the rapid expansion of the cell that occurs on formamide-washout that causes the t tubules to detach from the cell surface. It also agrees with the observation that incubation of cells in formamide for up to an hour, without washout, does not disrupt the t tubule system (not shown), and is compatible with previous data (14) showing that the major functional changes occur on washout of formamide.

The pattern of fluorescence from the trapped FITC-dextran suggests that this indicator is trapped in discrete compartments within the cell, which have a different spatial organisation from the original t tubules. This is consistent with the t tubules resealing within the cell, thus forming vacuoles, as suggested after detubulation of skeletal muscle (16). The idea that the t tubules reseal within the cell is also supported by the observation that the cells maintain their shape after detubulation (see below), whereas the high [Ca2+] in the t tubules would, if released into the cell interior, be expected to cause Ca2+ overload and cell death.

The present data suggest, therefore, that detubulation occurs on washout of formamide and that the t tubules reseal within the cell. It appears likely that they are physically uncoupled from the surface membrane because of the change in the pattern of staining and lack of access from the extracellular space after detubulation and electrically uncoupled because of the measured decrease in cell capacitance: the decrease in cell capacitance of ventricular cells after formamide treatment is ∼26% (Ref. 14 and F. Brette, data not shown). However, only ∼80% of cells are detubulated after this procedure (14). Correcting for the 20% of nondetubulated cells that will be included in the capacitance measurement, the true decrease in cell capacitance will be ∼33%, the same as the percentage of the cell membrane estimated from ultrastructural studies to be in the t tubules (19). It is unclear, however, whether the t tubules that have resealed within the cell remain capable of ion transport.

Detubulation does not affect cytoskeleton.

The cytoskeleton plays an important role in determining cell structure and shape and in anchoring proteins (2, 23). It seemed possible, therefore, that the changes produced by formamide could be secondary to effects on the cytoskeleton. However, the observation that there is no apparent disruption of the cytoskeleton (Fig. 3) and no change in cell size after formamide treatment makes this unlikely.

Detubulation changes spatial distribution of Ca2+ after electrical stimulation.

In control cells, Ca2+ increased homogeneously across the cell on electrical stimulation. However, in formamide-treated cells, Ca2+ initially increased close to the cell membrane and then propagated into the cell. This is similar to the pattern observed in atrial and Purkinje cells, which lack t tubules (6,11), and is consistent with the formamide-treated cells being detubulated and with an important role for the t tubules in ensuring rapid and synchronous Ca2+ release throughout the cell. Interestingly, the whole cell Ca2+ transients calculated from these line-scan images showed a biphasic rising phase as a consequence of this asynchronous release (Fig. 4); similar Ca2+ transients have been reported previously in detubulated cells by Kawai et al. (14), who, in agreement with the present data, suggested that it might be caused by initial Ca2+ release at the cell surface, followed by propagation into the cell.

It seems likely that the inward propagation of Ca2+ is due to propagated Ca-induced Ca release (CICR) because the rate of propagation is within the range reported previously for CICR (12), the amplitude of the Ca2+ transient is the same at SS and CC (Fig. 4), and propagation is inhibited by the sarcoplasmic reticulum inhibitors ryanodine and thapsigargin (not shown). This propagation is, therefore, similar to that in atrial cells (15) but different from that in Purkinje cells, in which buffered diffusion of Ca2+ appears to be responsible for the inward movement of Ca2+ from the cell surface (6) so that the amplitude of the Ca2+transient is smaller in the center of the cell than at the periphery (6).

Formamide has no direct effect on proteins.

As previously described for the atrial cells of other species, e.g., cat (15), rabbit (4), and guinea pig (7), the rat atrial cells used in the present study lacked t tubules (Fig. 5). This preparation therefore enabled us to investigate the possible direct effects of formamide on protein and cell function in the absence of t tubules. Formamide treatment of atrial cells did not induce a significant decrease in cell capacitance, or in the amplitude or density of the Ca2+ current, or in the configuration of the action potential or Ca2+transient. Thus formamide appears to have no direct effect onI Ca,L, or on the other currents, and hence presumably the channels that underlie the atrial action potential or Ca2+ transient. Similarly, the response of the β-adrenergic pathway to isoprenaline was unaltered after formamide treatment. These data are consistent with the idea that the effects seen in rat ventricular myocytes are due to detubulation rather than direct effects of formamide on proteins.

Significance of this technique.

Several different techniques have been used previously to investigate the role of the t tubules in cardiac muscle. Immunocytochemistry has been widely used to investigate protein distribution and has shown, for example, that the I Ca,L channel is concentrated in the t tubules (4). However, the observed distribution may depend on the antibody used and the accessibility of the epitope (17). Although this technique can yield useful information about protein distribution, it cannot give information about localized protein function, which may depend on other local factors, such as accessory proteins and modulators.

The patch-clamp technique has been used to study localization of functional channels. Jurevicius and Fischmeister (13) used a double patch-clamp technique and a double-barreled microperfusion system to show that Ca+ and Na+channels are uniformly distributed on the sarcolemmal membrane of frog ventricular myocytes. However, this technique does not have access to the t tubules. An alternative approach has been to use the diffusion delay between the bulk extracellular solution and the t tubules. This approach has been used to investigate the distribution of Na+ and Ca2+ channels in guinea pig ventricular myocytes (22) and has shown that 64% ofI Ca and I Na change slowly after a rapid change of extracellular [Ca+] and [Na+], indicating that ∼64% of functioning Na+ and Ca2+ channels are in the t tubules. This approach has also been used by Christé (5) to investigate the distribution of K+ inward rectifier current in rabbit ventricular myocytes. Comparison of preparations lacking t tubules with those having t tubules has also been used, although it is difficult to exclude differences in protein expression (see Introduction). Acute detubulation provides a complimentary technique to investigate the localization of membrane functions.

In summary, it appears that washout of formamide causes the t tubules to become uncoupled from the surface membrane and reseal within the cell. It seems most likely that the effects of formamide on the function of rat ventricular myocytes are due to this detubulation, rather than secondary to direct effects of formamide on the cytoskeleton or proteins. This technique therefore provides an additional tool to study the functional role of the t tubules in ventricular myocytes.


The authors thank Dr. H. Dobrzynski for helpful advice on confocal microscopy and Dr. S. C. Calaghan for valuable discussion.


  • This work was funded by the Wellcome Trust and the British Heart Foundation.

  • Address for reprint requests and other correspondence: C. H. Orchard, School of Biomedical Sciences, Univ. of Leeds, Leeds LS2 9NQ, UK (E-mail:c.h.orchard{at}leeds.ac.uk).

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

  • June 20, 2002;10.1152/ajpheart.00347.2002


View Abstract