inward rectifier potassium channel
- T1 and T2 APs
type 1 and type 2 action potentials
- • Two types of extracellular action potentials, bi- and tri-phasic, were recorded within the same fibre of skeletal muscle of frog.
- • Unlike bi-phasic, tri-phasic action potentials were very sensitive to the potassium channel blocker barium, and to a frequency of stimulation of muscle preparation.
- • It is suggested that currents generated in a specialized t-tubular compartment of a muscle fibre form the third phase of these action potentials and studies of tri-phasic action potentials may provide a new tool for evaluating the role of t-tubules in skeletal muscle contractility.
Abstract Two types of muscle fibre action potentials (APs) were recorded using narrow-tipped extracellular pipettes in isolated sartorius muscles of frog, Rana temporaria. The waveform of type 1 responses (T1 AP, 75% of recordings) was biphasic, ‘positive–negative.’ The type 2 signals were tri-phasic, ‘positive–negative–positive’ (T2 AP, 18%). The type of AP was preserved for up to 1 h of collecting data from the given site on the muscle fibre. However, a re-positioning of the recording pipette by a few micrometres along the axis of the studied fibre resulted in a change of the type of AP in 73% of such attempts. In experiments with detubulated muscles, only T1 APs were observed. In experiments using 10 μmol l−1 Ba2+ filled pipettes, the characteristics of the T1 waveform did not change, but the amplitude of the third phase of T2 APs increased progressively during continuous recording. In experiments with 100 μmol l−1 Ba2+ filled pipettes, a decline and eventual inversion of the third phase of the T2 signal was observed. Amplitude and duration of the inverted third phase of T2 APs increased progressively with frequency of muscle stimulation. These results can be explained by suggesting that currents generated within the t-tubular system of muscle fibre form the third phase of extracellularly recorded APs. Measurement and analysis of T2 APs present a unique approach in determining the mechanisms controlling accumulation of t-tubular K+ and the role of these mechanisms in regulation of muscle contractility.
Transport of K+ across the plasma membrane of skeletal muscle fibres is carried out by several different ion-transporting mechanisms. Due to the development of molecular and cell biology techniques, the molecular structure and cellular localization of these mechanisms is known reasonably well. However, our understanding of the relative functional significance of K+ transporting systems is still limited. Existing data on physiological specialization and quantitative estimates of the role of these systems in the maintenance of ionic homeostasis, muscle fibre excitability and contractility of adult skeletal muscles are incomplete and controversial. This is specifically true with regard to the functional role of Kir2.1 subfamily potassium channels of inward rectification located within the t-tubular system (Ashcroft et al. 1985; Kristensen et al. 2006; Kristensen & Juel, 2010). Muscle activity is intrinsically associated with the accumulation of extracellular K+. Timely removal of the excess K+ is critical for maintaining muscle excitability and contractility. It is generally accepted that clearance of extracellular potassium is accomplished mostly by the re-uptake of K+ into the muscle fibres by Na+,K+-ATPase and by free diffusion of K+ out of the interstitial muscle compartment (Clausen, 2003). Moreover, the experimental data and the results of computer simulations suggest that the re-uptake of K+ via inward rectifier potassium, Kir, channels may constitute a third route of extracellular K+ clearance. Unlike classical voltage-gated channels, Kir channels are opened at potentials negative to K+ equilibrium potential, which makes them potentially very efficient route for inward K+ current to move excess of extracellular K+ back into the muscle fibre. This route would be especially critical for control of the concentration of K+ in the narrow t-tubular muscle space where accumulation of K+ is especially high, while free diffusion of K+ is greatly limited (Wallinga et al. 1999; Kristensen et al. 2006). However, the t-tubular system is not accessible for direct electrophysiological studies. Therefore, the functional data needed for validation of the hypothesis above are lacking. Considering this problem, our study was designed to test if some useful information regarding the t-tubular electrogenesis may be obtained via extracellular recordings of the muscle fibre APs with narrow-tipped pipettes. The reasoning behind this approach was based on data showing great quantitative and qualitative differences in ion channels and transporters expressed in the surface and the t-tubular plasma membranes of skeletal muscle fibres (Kristensen et al. 2006) and on the fact that the signal recorded by an extracellular electrode mostly represents the currents running across the membrane patch located immediately under the orifice of the pipette (Wolters et al. 1994). Thus, it appeared reasonable to speculate that waveforms of locally recorded muscle fibre APs may be different depending on the position of the mouth of the extracellular pipette with respect to the opening of the t-tubular system on the surface of the studied fibre. The purpose of this work was to test this suggestion experimentally.
Studies were conducted in accordance with regulations of the National Committee on Bioethics of the Russian Academy of Sciences. The protocol was approved by the local ethics Committee of the Sechenov Institute of Physiology and Evolutionary Biochemistry.
Muscle preparation and experimental conditions
All experiments were performed at room temperature, 20–22 °C, during a spring (March–May) period on isolated sartorius muscle preparations of male common frog, Rana temporaria. A total of 35 animals were used. Sartorius muscles excised from pithed animals were kept in Ringer solution (in mmol l−1: NaCl, 112; KCl, 2.0; CaCl2, 1.8; NaHCO3, 2.0; pH 7.2–7.4) and used in experiments within 2–3 h after isolation. To evoke propagating action potentials the preparation was stimulated directly by 1 ms-long super-threshold pulses of current delivered at 0.1 Hz frequency via a pair of Ag–AgCl wire electrodes positioned above and below the muscle close to its proximal tendon (electro stimulator ES-50–1, EFSE, Chernogolovka, Russia). Muscle contractility was suppressed either by adding Dantrolene (10 μmol l−1) to the bath solution or by 20–30 min-long pre-incubation of the preparation in 300 mmol l−1 glycerol solution (glycerination; Krolenko, 1975; Sheikh et al. 2001). In experiments with Ba2+, a potassium channel blocker, BaCl2 (10–100 μmol l−1) was added to the pipette's Ringer solution. Bath Ringer solution was refreshed during the experiment between sequential recordings from the fibres of a given muscle preparation.
Muscle APs were recorded extracellularly in the non-synaptic muscle region about 15–20 mm distal to the stimulating electrodes, using the S52 glass recording pipettes with 3–5 μm tip (outer diameter). The pipettes were pulled in two steps using the ME-4 microelectrode puller (Biolink, Sankt-Petersburg, Russia) and fire-polished. When filled with Ringer solution, the pipettes had a resistance of 0.5–1.0 MΩ. The pipette was positioned on the muscle fibre under visual control at 200× magnification with a Carl Zeiss microscope (Jena, Germany) using a P-54-011 mechanical manipulator (ESIB, Moscow, Russia). Negative pressure was applied to the pipette via suction to achieve pipette tip–fibre contact resistance of 3–5 MΩ. Recorded signals were amplified using a ROC-3M single channel amplifier (Biophysequipment, Saint-Petersburg, Russia), filtered between 0.03 and 10 kHz, digitized at 10 μs intervals by a 16 bit analog–digital converter (NI USB-6211, National Instruments, Austin, TX, USA) and stored on the hard-drive of a Pentium-4 computer. Signals were analysed off-line (Clampfit-6; Axon Instruments, Union City, CA, USA) for the wave amplitude (from baseline to the peak), and rise and half-decay times. The signals recorded with this technique represent a potential drop across the sealing (pipette tip–tissue) resistance, which is directly proportional to a local current leaving or entering the muscle fibre under the electrode opening. Therefore, the technique is sometimes referred to as the ‘focal current recording’ technique with signals reported in either voltage (as in this report) or current units (see Brigant & Mallart, 1982; Wolters et al. 1994).
Statistical analysis and curve fitting were carried out using Origin 7.0 (OriginLab Corp., Northampton, MA, USA), Prism 5.0 (GraphPad Software Inc., La Jolla, CA, USA), and a Levenberg–Marguardt algorithm of minimization. Average values are expressed as mean (±SEM). Significant differences were defined as having a P value less than 0.05. Individual traces shown in figures represent point-by-point averages of 10 consecutive responses.
Two distinct types, bi-phasic and tri-phasic (designated here as type 1 (T1) and type 2 (T2) APs, respectively), of extracellularly recorded AP waveforms were observed in experiments in intact muscle preparations (Fig. 1). The amplitude and temporal characteristics of the first two phases of these signals were nearly identical (Fig. 1C), but the presence or absence of the positive phase at the end of the waveform allowed unquestionable classification of recorded APs in most cases (93% of a total of 546 recordings from 297 fibres from 32 intact muscles). In the recordings in which the fibres were approached at random with the electrode, observations of T1 APs were by far most prevalent, 75%, with T2 APs constituting 18% of recorded waveforms only.
Interconversion between types of APs was not observed in any of the control experiments with up to 60 min of continuous recording at a given site of the muscle fibre (5 muscles, 23 fibres). Furthermore, the type of recorded AP did not change with an increase in frequency of stimulation of the preparation (Fig. 1A and B). However, a change in the type of recorded AP was observed in 12 of 16 experiments in which the tip of the recording pipette was repositioned by a few micrometres along the studied fibre with respect to the original site (Fig. 1E).
Another important observation was that in the studies of glycerol treated muscles (12 muscles, 43 fibres; to disrupt the t-tubule system) no identifiable T2 waveforms were found: 92% of the recorded APs were T1 signals and the other 8% constituted unclassified forms. Furthermore, unlike that in studies of intact muscles, in none of the 39 fibres of glycerinated muscles generating T1 responses did screening of the surface of these fibres by a repositioning of the pipette result in a change of the type of recorded signal (Fig. 1D). Thus, the specific position of the electrode on the muscle fibre and the presence of an intact t-tubular system constitute two critical determinants of the waveform of extracellular AP to be recorded.
Analysis of the peak amplitude distribution of the third phase of identified T2 APs has shown that this group of signals could be separated further into at least four subgroups composed of T2 APs having a progressively stronger third phase (marked as T21, T22, T23 and T24 in Fig. 2A). Overall prevalence of observations of T1 or of one of the subtypes of T2 responses declined exponentially, being highest for T1 and lowest for strong third phase T2 responses, respectively (Fig. 2B, continuous curve). Interestingly, the decline in prevalence of the types and the subtypes of AP could also be approximated using a spatial Poisson function (Fig. 2B, columns).
Further support for the suggestion that bi-phasic and tri-phasic APs constitute mechanistically different groups of signals was provided by experiments utilizing pipettes filled with a solution containing a potassium channel blocker, Ba2+. In experiments with 10 μmol l−1 Ba2+ filled pipettes there was no change over time in either amplitude or temporal characteristics of T1 APs of either intact or glycerinated muscle fibres at 0.1 Hz stimulation frequency (Fig. 3A). For example, average amplitude of the second phase of T1 AP for the 10th–20th minutes of recording was 104 ± 11% (4 intact muscles, 18 fibres, 18 recordings) and 106 ± 17% (4 glycerinated muscles, 22 fibres, 27 recordings) relative to its amplitude immediately after the beginning of the record. In contrast, T2 APs recorded under similar conditions demonstrated progressive decline of the second phase and an increase in amplitude and duration of the third phase (Fig. 3B). On average, after 10 min of recording, the amplitude of the third phase of T2 AP increased to 232 ± 23% relative its initial value (5 muscles, 32 fibres, 43 recordings). Interestingly, the time course of relative changes of the second and third phases of T2 APs was nearly identical (Fig. 3C), implying that the same mechanism governs both changes. Changes in the voltage on the pipette tip seal resistance obtained by subtraction of the AP traces recorded at the beginning and at the end of the experiment are consistent with the idea that this putative mechanism consists of the development of some outward current (Fig. 3D). As this current had no or little effect on the first phase of the AP, it is probably generated immediately under the rim of the pipette.
In experiments with higher concentrations of Ba2+, 0.1–1 mmol l−1, T1 APs remained insensitive to the presence of the potassium channel blocker in the pipette except for a gradual decrease in the amplitude of the second phase of the signal at high, 10 Hz, rates of stimulation (Fig. 4A; the peak of the second phase declined under these conditions to 78 ± 6% of its initial value by the 10th minute of recording; 18 fibres/recordings, 4 intact muscles). A similar moderate decline in the second phase was observed also for T2 APs recorded with 100 μmol l−1 Ba2+-filled pipettes at 10 Hz (Fig. 4B). With respect to the third phase of T2 APs, however, in about half of those experiments it did not change at low stimulation rates but did disappear completely at high stimulation rates (17 of 36 records; Fig. 4B). In the remaining of these experiments, a gradual replacement of the third positive phase of T2 AP with a phase of late negativity was observed even at low frequency of muscle stimulation (Fig. 4C). At the 10th minute of continuous recording the average amplitude and half-decay time of this newly developed negative phase were –0.22 ± 0.06 mV and 24 ± 6 ms (n= 19). Both the amplitude and duration of the Ba2+ induced late negative phase increased with the rate of stimulation. (Fig. 4D).
The major finding of this work is a description of the existence of two distinct, bi- and tri-phasic types of waveforms of extracellularly recorded APs in frog sartorius muscle, designated here as T1 and T2 APs.
The paucity of transient forms (7% of non-identifiable responses only), the absence of interconversion between types of AP during the prolonged recordings and the sharply different sensitivity of T1 and T2 waveforms to intra-pipette Ba2+ all strongly suggest qualitative differences in mechanisms of bi- and tri-phasic APs. Considering the putative nature of this difference, it is important to note that extracellular micropipettes collect information about local electrical events, the currents originating in the membrane located immediately under the orifice of the pipette (Brigant & Mallart, 1982; Mallart, 1985; Wolters et al. 1994). Therefore, our observation that the type of recorded AP waveform can be changed as a result of a small change in the electrode position on the studied fibre suggests that some specifics of the surface of the fibre, not the type of muscle fibre, are responsible for recording either a T1 or a T2 AP. Following this idea further, it appears most reasonable to suggest that the third phase of T2 AP is formed primarily by currents generated within the t-tubular compartment of the muscle fibre (Fig. 5A) and the type of recorded AP is determined by the location of the tip of the pipette with respect to the opening of the t-tubules on the surface of the fibre. In accord with this suggestion, recordings of T1 APs or T2 APs with incrementally increasing amplitude of their third phase could correspond with situations when the tip of the pipette happens to be positioned over the area of the muscle fibre containing respectively none or one, two, or three t-tubular openings (Fig. 5B and C; sites ‘a’, ‘b’, ‘c’ and ‘d’).
Limits of the resolution of the light microscope do not allow direct evaluation of the suggestion above. However, it is consistent with the data showing that the first and second phases of extracellularly recorded AP are determined mostly by the activity of voltage-dependent Na+ channels (located just outside and within the rim of the micropipette, respectively), while the third phase of these signals is governed mostly by potassium currents (Wolters et al. 1994). While present at high density in sarcolemma, Na+ channels are relatively scarce within the t-tubular membrane of the frog skeletal muscle fibres (Huxley & Taylor, 1958; Sheikh et al. 2001), which explains the relative stability of characteristics of the first two phases of T1 and T2 APs recorded under different experimental conditions in these experiments. It also explains the apparent resistance of the T1 AP waveform to the detubulation of the muscle fibres by glycerination. At the same time, the t-tubular compartment of a muscle fibre appears to be the major source of outward potassium currents (Wolters et al. 1994). Therefore, it should not be surprising that glycerination eliminates the third positive wave of T2 APs, as was observed in our experiments.
Another line of support for the hypothesis above relates to the observation of the peculiar ‘quantal’ distribution of the peak values of the third phase of T2 signals and the apparent exponential decay in the prevalence of signals with T1 APs and T2 APs, with strong third phase being most and least frequently observed, respectively (Fig. 2). Such a distribution of the AP type/subtype prevalence appears well matched to one predicted by the two-dimensional spatial Poisson process describing the probability of encountering exactly κ events (t-tubular openings) within the arbitrarily selected area (determined by the pipette position and inner diameter) of the plane containing events positioned stochastically with average intensity of λ (events/t-tubular openings per unit of the plane/muscle fibre area; Fig. 5B and C). Considering the goodness of fit shown in Fig. 2C (columns) we need to emphasize that some conflict between the predicted and experimental data must be expected. The spatial Poisson process requires events to be randomly distributed over the plain, while t-tubular openings on the surface of muscle fibres tend to be arranged into a rectangular mesh (Huxley & Taylor, 1958; Kostin et al. 1998; Sheikh et al. 2001). Specifically for amphibian skeletal muscles, t-tubules open at z-lines, spaced at 2–7 μm (depending on the muscle stretch) along the fibre and at about 5 μm intervals across the fibre along the z-line (Huxley & Taylor, 1958; Sheikh et al. 2001). Although such an arrangement is not perfect (Huxley & Taylor, 1958; Soeller & Cannell, 1999), it certainly does not satisfy the Poisson process requirement above. It is interesting nonetheless that the best fit parameters of the Poisson equation in our study were well within reasonable ranges. Thus, predicted by the fit, the rectangular area of the muscle fibre per one t-tubular opening was 27 μm2. The equivalent area calculated assuming 5 μm distances between t-tubular lumens along a z-line and 2–7 μm between z-lines (Huxley & Taylor, 1958) is between 10 and 35 μm2. Used in the fit electrode lumen diameter, 2.7 μm, is a quite reasonable estimate of the inner diameter of fire-polished pipettes, used by us (2/3 of 3–5 μm, mean 4 μm outer tip pipette diameter). It is also important that our hypothesis and estimates as given above predict that no T1 AP recordings could be possible with large pipettes, exceeding in internal diameter 6.5 μm (area of opening ≥35 μm2); such pipettes will always cover the fibre area containing at least one t-tubular opening. In accord with this, no T1 AP waveforms have been reported in the study of extracellular APs in rat skeletal muscle by Wolters et al. (1994), in which large, 10 μm diameter, pipettes were used.
The final line of support for the t-tubular origin of the third phase of T2 APs comes from experiments showing distinct behaviour of T1 and T2 waveforms in experiments with Ba2+ loaded pipettes. At low concentration Ba2+ selectively blocks inward-rectifier potassium (Kir) channels (Kristensen et al. 2006). At both high and low concentrations, local application of Ba2+ had little effect on T1 AP or the first two phases of T2 AP, confirming that potassium currents are of minor significance for generation of these segments of the AP waveform (see also Wolters et al. 1994). Although further studies are needed, this result could indicate a leading role for Na+ and Cl− conductances in shaping electrical events generated in sarcolemma of the skeletal muscle fibre. Probably because of a relatively strong presence of K+ channels, the situation in the t-tubular compartment appears to be very different in this respect. Indeed, the third phase of T2 APs increased during the recordings with micromolar Ba2+ loaded pipettes and was suppressed and replaced with late negative and frequency of stimulation-dependent wave in about half of the experiments using high (0.1–1 mmol l−1) Ba2+ concentrations in the pipette solution. We believe the simplest explanation for these observations may be based on the following postulates. First, amplitude and direction of the third phase of T2 AP is determined by a balance of several currents generated within the t-tubular compartment, with the potassium voltage-dependent, leak and inwardly rectified currents and the sodium pump current playing the major role (Fig. 6A). Second, because of the progressive increase in concentration and a decline in K+ reversal potential toward the depth of the t-tubule, potassium currents generated deep and at the mouth of the t-tubule have mostly inward and mostly outward direction, respectively (Fig. 6A). With these postulates, in experiments with low concentrations of Ba2+, selective block of inward Kir current by Ba2+ diffusing from the pipette into the depth of the t-tubule appears as the most likely explanation for the gradual increase of the integral t-tubular current in the outward direction resulting in an increase of the third positive phase of T2 APs (Fig. 6B). Experiments with high concentrations of Ba2+ that would be non-selective to different K+ conductances should result in a complete block of inward potassium current via Kir channels leaving free diffusion and Na+,K+-ATPase as the only mechanisms controlling K+ concentration in the t-tubular system. The failure of these mechanisms to prevent use-dependent accumulation of t-tubular K+ will eventually result in a severe deterioration of the chemical gradient of K+ on the t-tubular membrane leading to a decrease in outward potassium current and disappearance or inversion of the third positive phase of the T2 AP waveform (Fig. 6C). The amplitude and temporal characteristics of this late wave will be determined by the rate of accumulation of potassium in the t-tubule (dependence on stimulation frequency) and by the rate of the K+ removal from the t-tubule by Na+,K+-ATPase and by diffusion into the intra-pipette solution.
Overall, our data are consistent with the idea that the third phase of T2 APs recorded with relatively narrow extracellular pipettes is formed by currents generated within the t-tubular system of the skeletal muscle fibre. Measurement and analysis of these signals therefore presents a useful tool in functional studies of the role of the t-system in normal physiology and pathophysiology of skeletal muscle. In particular, such studies open an avenue for further dissection of the relative roles of Kir channels and Na+,K+-ATPase in clearance of intra-tubular potassium and regulation of muscle contractility and fatigue. In this respect, further studies using this technique in combination with a battery of various ion channels and transporter blockers are of great interest and importance.
I.V.K. and M.D. equally contributed to the conception and design of the experiments, the collection, analysis and interpretation of the data, and the drafting and revision of the article.
We thank Ms Gemma DiTommaso and Dr Joseph R. Stimers for their help with editing this manuscript. This work was supported by grant 02.740.11.5135 from the Federal Program of the Ministry of Science and Education of the Russian federation (to M.D. and I.V.K.) and in part by the COM UAMS pilot grant program (to M.D.).