Abstract

Objective: Rabbit ventricular myocardium is characterized by a biphasic response to stretch with an initial, rapid increase in force followed by a delayed, slow increase in force (slow force response, SFR). The initial phase is attributed to increased myofilament Ca2+ sensitivity, but the mechanisms of the delayed phase are only incompletely understood. We tested whether stretch-dependent stimulation of Na+/H+ exchange (NHE1) and consecutive changes in pHi and/or [Na+]i may underlie the SFR.

Methods: Isometric contractions of rabbit ventricular muscles were recorded in bicarbonate-containing Tyrode's (Tyrode) or bicarbonate-free HEPES-buffered solution (HEPES). Muscles were loaded with the Ca2+ indicator aequorin, the pH indicator BCECF, or the Na+ indicator SBFI and rapidly stretched from 88% (L88) to 98% (L98) of optimal length. The resulting immediate and slow increases in twitch force (1st phase and SFR) as well as changes in [Ca2+]i, [Na+]i, or pHi were quantified before and after inhibition of NHE1 by HOE 642 (3 μM) or reverse-mode Na+/Ca2+ exchange (NCX) by KB-R 7943 (5 μM).

Results: In both Tyrode (n=21) and HEPES (n=22), developed force increased to ∼160% during the 1st phase followed by a further increase to ∼205% during the SFR. The SFR was accompanied by a 21% increase of the aequorin light transient (n=4; normalized to the 1st phase) and a ∼3 mM increase in [Na+]i (n=4–7). The SFR was also associated with an increase in pHi. However, this increase was delayed and was significant only after the SFR had reached its maximum. The delayed pHi increase was larger in HEPES than in Tyrode. HOE 642 and/or KB-R 7943 reduced the SFR by ∼30–40%. In addition, HOE 642 diminished the stretch-mediated elevation of [Na+]i by 72% and the delayed alkalinization.

Conclusions: The data are consistent with the hypothesis that SFR results from increases in [Ca2+]i secondary to altered flux via NCX in part resulting from increases in [Na+]i mediated by NHE1.

1. Introduction

In 1973 Parmely and Chuck demonstrated that stretch of cardiac muscle causes a biphasic increase in developed force: an initial rapid rise, attributed to an increase in myofilament Ca2+ responsiveness, followed by a slow rise, taking 15–20 min to develop and thus termed the slow force response (SFR) [1]. Though of potential functional relevance, the cellular mechanisms of the SFR remain a matter of debate.

Experiments in papillary muscles or trabecular preparations from rat, cat and rabbit provided evidence that increases in [Ca2+]i transients underlie the SFR [2–6]. In rat and cat muscle, a stretch-dependent NHE1 stimulation was demonstrated and associated with the SFR. NHE1 is an electroneutral antiporter that extrudes 1 H+ coupled to the influx of 1 Na+. Therefore, both intracellular alkalinization with increased Ca2+ sensitivity of the myofilaments or elevated [Na+]i, leading to enhanced Ca2+ influx via reverse-mode NCX, were implicated in the SFR [7,8]. Alvarez et al. demonstrated that the rise in pHi could be easily detected in the absence of bicarbonate. It is minimized or even absent when there is CO2–bicarbonate buffer in the medium [2]. The absence of pH changes in bicarbonate buffer seems to result from a simultaneous stimulation of NHE1 and a bicarbonate-dependent acid loading mechanism after the release of endogenous angiotensin II/endothelin-1 (AngII/ET-1) by stretch [2,5–7,9,10].

However, autocrine–paracrine AngII/ET-1 stimulation does not contribute to the SFR in rabbit or human myocardium [6,8]. Therefore, the goal of the present study was to determine the effects of stretch on isometric contractions, pHi and [Na+]i in rabbit myocardium under physiological (bicarbonate-containing) and unphysiological (bicarbonate-free) buffering conditions. The main findings were that in physiological buffer [Na+]i–but not pHi–increased during the development of the SFR. pHi started to increase only after the SFR had reached its maximum. The stretch-induced rise in pHi was much more pronounced in unphysiological HEPES buffer. The data support that NHE1 stimulation, followed by increases in [Na+]i and [Ca2+]i, but not alkalinization, underlies the SFR under physiological conditions in the rabbit heart.

Materials and methods

Rabbit myocardium

Experiments were performed in isolated left ventricular muscle strips from Chinchilla bastard rabbits. The study protocol was approved by the local ethics committee and was performed in accordance with the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23, revised 1996).

Muscle preparation and fluorescence measurements

Muscles were dissected and prepared as described previously [6]. Trabeculae were mounted in a cylindrical glass cuvette, connected to an isometric force transducer, and superfused with modified Tyrode's solution (in mM: Na+ 152, K+ 3.6, Cl 135, HCO3 25, HEPES 5, Ca2+ 2.5, Mg2+ 0.6, H2PO4 1.3, SO42− 0.6, pH 7.4), 30 °C, 0.2 Hz stimulation, gassed with 95% O2 and 5% CO2 (bicarbonate-containing buffer) or 100% O2 (bicarbonate-free HEPES buffer) [11]. Isometric twitches were evoked through stimulation with rectangular voltage pulses (5 ms; 20% above threshold). Twitch force (mN) was measured by means of a force transducer (Scientific Instruments). Twitch force (or developed force) was defined as the difference between peak systolic force and diastolic force. For comparison between muscle strips, twitch force was normalized to muscle cross-sectional area (mm2) which was assessed by the help of a stereo microscope at the beginning of an experiment at slack length. In connection with the assessment of the slow force response, baseline values represent developed force parameters at the end of the first phase.

We chose to conduct the present study at 30 °C, 0.2 Hz stimulation, where cellular mechanisms of the SFR appeared to be unaffected and conditions for fluorescence measurements were optimized compared with experiments at 37 °C and 1 Hz stimulation [own unpublished results].

Measurements of pHi or [Na+]i were performed as previously reported [12,13]. Muscles were loaded with BCECF-AM (15 μM, 45 min) or SBFI-AM (35 μM, 180 min) in Tyrode's solution by incubation at room temperature. Excitation light from a 100 W mercury lamp was passed alternately through two bandpass filters (450 nm and 495 nm for BCECF; 340 nm and 380 nm for SBFI) at 125 Hz and focussed on the muscle strip. Fluorescence emission was collected by a photomultiplier (Scientific Instruments) after passage through a bandpass filter (535 ± 5 nm for BCECF; 505 ± 5 nm for SBFI). To limit photobleaching of the dyes, a neutral density filter (1% transmittance) was placed in the excitation light path. Furthermore, recording of fluorescence was limited to short intervals (∼20 s every minute during the first 10 min after the stretch, then ∼20 s every 2–5 min). In contrast to [Ca2+]i, pHi and [Na+]i did not change during the twitch. Reported values for pHi and [Na+]i were obtained as the mean value measured during the first 50 ms of the twitch.

At the start of the pH experiments fluorescence emission of BCECF was 10–15 times larger than the autofluorescence. Values of pHi were estimated from the ratio of the BCECF fluorescence signals (F495/F450) after subtraction of background fluorescence. At the end of each experiment, the BCECF fluorescence ratio was calibrated in vivo by means of the high K+–nigericin method as previously reported [12]. The calibration solution contained (in mM): 140 KCl, 1.2 MgCl2, 5 HEPES, 0.01 nigericin, and 30 2,3-butanedione monoxime (BDM). The pH of the calibration solution was adjusted with KOH to 5 different values ranging from pH 6.8 to 7.6.

Fluorescence emission of SBFI-loaded muscles was 4–5 times larger than the autofluorescence. Values of [Na+]i were estimated from the SBFI fluorescence ratio (F340/F380) after correction for autofluorescence. Similar to BCECF, SBFI ratios were calibrated by an in vivo approach. Two calibration solutions were prepared. The first one was composed of (in mM): 140 NaCl, 10 HEPES, 1 EGTA, pH 7.4 (NaOH). The second calibration solution was identical to the first one except that it contained K+ instead of Na+. The two solutions were mixed to yield [Na+] of 0, 5, 10, 15, 20, and 30 mM, respectively. Finally, 100 μM strophantidin, 40 μM monensin, 0.02 μg/ml gramicidin, and 30 mM BDM were added to the respective calibration solutions [13].

Aequorin-measurements: at steady state contractile function, electrical stimulation was switched off and the Ca2+-regulated bioluminescent photoprotein aequorin was macroinjected into the quiescent muscle as described previously [14]. Aequorin light emission was detected using a photomultiplier, which was vertically mounted with its cathode just above the glass cuvette containing the muscle. The signal was analyzed as the difference between peak systolic light emission and diastolic baseline values (mV amplifier output) and is referred to as the aequorin light transient or, alternatively, the [Ca2+]i transient.

Experimental protocol

A complete length–tension relationship was obtained in each muscle strip by stepwise increases in the pre-stretch level of the muscle. Lmax was defined as the muscle length at which maximum developed (systolic–diastolic) force was observed. Afterwards muscles were relaxed to 88% of Lmax for 30 min, then stretched to 98% of Lmax until complete stabilisation of mechanical and fluorescence signals (20 min). In Tyrode's solution this protocol could be repeated in an individual muscle strip preparation. However, in HEPES pHi remained elevated after a first stretch protocol for a prolonged period of time. Drugs were applied 20 min before the stretch and remained present throughout the experimental protocol.

Reported changes in twitch force refer to changes in developed (systolic–diastolic) force.

All experiments were performed at 30 °C and 0.2 Hz stimulation rate, since fluorescent dye leakage became a problem at higher temperature and stimulation rate.

Drugs

The NHE1 inhibitor HOE 642 (Cariporide; Aventis Pharma, Frankfurt, Germany) was diluted from an aqueous stock solution (10 mM). The reverse-mode NCX inhibitor KB-R 7943 (Tocris, Ballwin, USA) was added from a 10 mM stock (50% DMSO, 50% water).

Statistical analysis

Average values are given as mean ± S.E.M. Differences were compared by one-way ANOVA analysis and unpaired Student's t-test followed by Bonferroni–Holmes equation when appropriate. Differences were considered significant at p<0.05.

Results

Immediate and slow force response

Rabbit myocardium is characterized by a biphasic response to stretch as shown in the original recording of a typical experiment in Tyrode's solution (Fig. 1A). Stretching the preparation from 88% (L88) to 98% of its optimal length (L98) resulted in an immediate increase in twitch force, followed by a slowly (∼10 min) developing further rise in twitch force (slow force response, SFR). Both phases were reproducible: after releasing the muscle back to L88, a 2nd stretch protocol revealed almost identical increases in twitch force during both phases (not shown). On average, in Tyrode's solution (n=21) the stretch protocol resulted in an immediate increase in twitch force to 161.7 ± 3.8% (p<0.01 vs. baseline L88 values) and a further increase to 208.5 ± 6.9% (p<0.05 vs. the 1st phase) during the SFR. In HEPES (n=22) twitch force increased to 163.7 ± 2.8% (p<0.01) during the immediate phase and further to 205.5 ± 5.2% (p<0.05 vs. 1st phase) during the SFR. Thus, the stretch-induced increases in twitch force were essentially identical under the two different buffer conditions.

Fig. 1

(A) Original tracing of effects of stretch from 88% to 98% of Lmax in a muscle strip in Tyrode's solution. Top: pHi tracing. Bottom: Force tracing. At the end of the experiment, an in vivo calibration was performed (right). pHi increased during the late phase of stretch. (B) Influence of stretch on SFR (bottom) and pHi (top) in bicarbonate-containing buffer (Tyrode; n=21) and in bicarbonate-free buffer (HEPES; n=22). “Max” represents the maximal developed twitch force in percent of first phase. “Last” represents measurements at 20 min after stretch.

Fig. 1

(A) Original tracing of effects of stretch from 88% to 98% of Lmax in a muscle strip in Tyrode's solution. Top: pHi tracing. Bottom: Force tracing. At the end of the experiment, an in vivo calibration was performed (right). pHi increased during the late phase of stretch. (B) Influence of stretch on SFR (bottom) and pHi (top) in bicarbonate-containing buffer (Tyrode; n=21) and in bicarbonate-free buffer (HEPES; n=22). “Max” represents the maximal developed twitch force in percent of first phase. “Last” represents measurements at 20 min after stretch.

pHi changes during stretch

As also shown in Fig. 1A, stretch induced a delayed increase in BCECF emission ratio. On the right-hand part of the figure, the calibration procedure for the BCECF signal is illustrated. Following the stretch protocol, the muscle was exposed to five calibration solutions of pH 6.8 to 7.6. According to this in vivo calibration, pHi of the muscle strip was 6.96 at L88. It remained unchanged for the first ∼10 min after stretching to L98. Then pHi started to rise gradually to 7.01 until the end of the stretch protocol. Importantly, the observed increase of pHi occurred only after the SFR had reached its maximum. This was a consistent finding. Thus, the stretch-induced changes in twitch force and pHi were analysed (1) at the peak of the SFR, when twitch force was maximal (i.e. typically after ∼6–7 min=“max” in Figs. 1B and 2 at the end of the stretch protocol (i.e. after 20 min=“last” in Fig. 1B), when the pHi increase was maximal, and compared to the values obtained immediately after the stretch (i.e. the immediate increase or 1st phase=“1st” in Fig. 1B).

Fig. 2

Influence of stretch on [Na+]i and twitch force in Tyrode solution. Changes in [Na+]i (A) and corresponding twitch force (B) during a typical stretch experiment in Tyrode. SBFI fluorescence ratio was converted to [Na+]i by the calibration curve depicted in the inset. (C) Average values of the maximal increases in twitch force during SFR (left) and [Na+]i (right) without (grey bars; n=7) and in the presence of HOE 642 (n=7). *p<0.05 vs. baseline value, #p<0.05 vs. ctrl. Bar graphs represent the average of maximal increase in developed tension and [Na+]i.

Fig. 2

Influence of stretch on [Na+]i and twitch force in Tyrode solution. Changes in [Na+]i (A) and corresponding twitch force (B) during a typical stretch experiment in Tyrode. SBFI fluorescence ratio was converted to [Na+]i by the calibration curve depicted in the inset. (C) Average values of the maximal increases in twitch force during SFR (left) and [Na+]i (right) without (grey bars; n=7) and in the presence of HOE 642 (n=7). *p<0.05 vs. baseline value, #p<0.05 vs. ctrl. Bar graphs represent the average of maximal increase in developed tension and [Na+]i.

Fig. 1B compares the stretch-induced changes of twitch force and pHi obtained from 21 muscles in Tyrode's solution and 22 muscles in HEPES buffer. In Tyrode's solution, stretch increased maximum SFR to 129.6 ± 2.1% of the baseline value (i.e. at the end of the immediate increase) after 5.8 ± 1.6 min, with non-significant increases in pHi. The SFR slightly decreased during the next minutes, but pHi further increased by 0.07 ± 0.06 pHi units (p<0.05 vs. baseline).

In HEPES buffer maximum twitch force increased during the SFR to 129.1 ± 3.4% after 7.3 ± 1.7 min with non-significant increases in pHi. However, while twitch force slightly decreased, pHi further increased by 0.20 ± 0.03 pHi units after 20 min (p<0.05 vs. baseline). These data indicate that stretch results in a delayed increase in pHi, which is more prominent in bicarbonate-free buffer and which does not directly contribute to the SFR.

[Na+]i changes during stretch

Since previous studies indicated that the SFR may be Na+-dependent, we used SBFI epifluorescence to measure directly whether stretch alters [Na+]i in rabbit ventricular muscle. Fig. 2 shows an original recording of stretch-induced changes in [Na+]i (A) and twitch force (B) from a muscle strip superfused with Tyrode's solution. During the SFR [Na+]i rose from 19.4 mM to a maximum of 22.2 mM and the increase in [Na+]i started almost immediately (∼60 s) after the stretch. Average basal [Na+]i amounted to 14.4 ± 1.6 mM (n=14). This is somewhat higher than reported by others, but within the range of our previously published data [15]. At the peak of the SFR approximately 8–10 min after the stretch, [Na+]i had risen by ∼2 mM to 21.0–21.5 mM. Thus, the time course of the stretch-induced [Na+]i increase closely followed the development of the SFR. In a total of seven experiments, the SFR amounted to 127.3 ± 3.8% and [Na+]i increased by 3.16 ± 0.51 mM. Similar results were obtained in HEPES-buffered solution. Under these conditions, SFR and [Na+]i increased to 132.9 ± 11.0% and by 2.65 ± 1.34 mM of basal value, respectively (n=4; p=n.s. vs. Tyrode; data not shown). Thus, in rabbit ventricular myocardium the SFR is accompanied by a parallel increase in [Na+]i that is similar in bicarbonate- and HEPES-buffered solutions.

Contribution of Na+/H+ exchange and Na+/Ca+ exchange to the SFR

In the next set of experiments, we assessed the contribution of NHE1 to the SFR and the stretch-induced changes of pHi and [Na+]i. For this purpose, muscle strips were preincubated with the selective NHE1 inhibitor HOE 642 (3 μM) for 20 min. HOE 642 did not affect the immediate increase in twitch force during phase 1 in either of the two buffers used.

Fig. 2C compares the magnitude of the SFR and the accompanying rise in [Na+]i in muscle strips bathed in Tyrode's solution in the absence (control, n=7) and in the presence of 3 μM HOE 642 (n=7). With NHE1 inhibition, the SFR was reduced from 127.3 ± 3.8% to 115.6 ± 3.3% (p<0.05). In addition, the [Na+]i increase was reduced by HOE 642: [Na+]i increased by 0.90 ± 0.24 mM with HOE 642 as compared to 3.16 ± 0.51 mM without HOE 642 (p<0.01). In other words, HOE 642 (3 μM) reduced the SFR and the concomitant rise in [Na+]i by 43% and 72%, respectively, suggesting a major contribution of NHE1 activity to both the stretch-induced slow increases in twitch force and [Na+]i.

We then tested the effects of NHE1 inhibition on the delayed increase in pHi (Fig. 3).

Fig. 3

Influence of NHE1 inhibition by HOE 642 (3 μM) on twitch force (bottom) and pHi (top) in bicarbonate-containing (Tyrode; n=7) and bicarbonate-free (HEPES; n=8) buffers. *p<0.05 vs. baseline values in unstretched (L88) muscle strips. pHi was measured 20 min after stretch.

Fig. 3

Influence of NHE1 inhibition by HOE 642 (3 μM) on twitch force (bottom) and pHi (top) in bicarbonate-containing (Tyrode; n=7) and bicarbonate-free (HEPES; n=8) buffers. *p<0.05 vs. baseline values in unstretched (L88) muscle strips. pHi was measured 20 min after stretch.

In Tyrode's solution, HOE reduced the SFR (from 128.6 ± 4.5% to 119.1 ± 3.5%, p<0.05) and largely prevented the increase in pHi (by 0.07 ± 0.06 units without and by 0.03 ± 0.03 units with HOE). The effects of NHE1 inhibition were similar in HEPES buffer: HOE significantly reduced the SFR (from 129.1 ± 3.4% to 119.7 ± 2.5%, p<0.05) and the increase in pHi (from 0.20 ± 0.03 to 0.11 ± 0.01 units, respectively; p<0.05) (Fig. 3).

These results indicate that NHE1 activity contributes to the SFR and that the stretch-dependent increases in both [Na+]i and pHi are mediated in large part by stimulation of NHE1 activity.

To test the hypothesis that reverse-mode NCX activity underlies the SFR, we conducted experiments in the presence of 5 μM KB-R 7943. KB-R 7943 did not affect the immediate increase in twitch force elicited by stretch, but significantly attenuated the SFR both in Tyrode's solution (from 126.8 ± 2.9 to 119.2 ± 3.8; n=7) and in HEPES buffer (from 127.7 ± 3.6 to 118.3 ± 2.8; n=8; Fig. 4). However, in contrast to HOE 642, KB-R 7943 did not affect the stretch-induced elevation of pHi neither in Tyrode's solution nor in HEPES buffer (Fig. 4). Furthermore, additional experiments were performed in the presence of 3 μM HOE 642 plus 5 μM KB-R 7943 to characterize the effect of dual inhibition of NHE1 and NCX function on the SFR. In the presence of dual inhibition of NHE1 plus NCX, the SFR was largely reduced to 114 ± 2.1% (n=10, Fig. 5).

Fig. 5

Influence of combined reverse-mode NCX inhibition by KB-R 7943 (5 μM) and NHE1 inhibition by HOE642 (3 μM) on SFR in bicarbonate-containing Tyrode (n=10).

Fig. 5

Influence of combined reverse-mode NCX inhibition by KB-R 7943 (5 μM) and NHE1 inhibition by HOE642 (3 μM) on SFR in bicarbonate-containing Tyrode (n=10).

Fig. 4

Influence of reverse-mode NCX inhibition by KB-R 7943 (5 μM) on twitch force (bottom) and pHi (top) in bicarbonate-containing (Tyrode; n=7) and bicarbonate-free buffer (HEPES; n=8) after stretch. *p<0.05 vs. baseline values in unstretched (L88) muscle strips. pHi was measured 20 min after stretch.

Fig. 4

Influence of reverse-mode NCX inhibition by KB-R 7943 (5 μM) on twitch force (bottom) and pHi (top) in bicarbonate-containing (Tyrode; n=7) and bicarbonate-free buffer (HEPES; n=8) after stretch. *p<0.05 vs. baseline values in unstretched (L88) muscle strips. pHi was measured 20 min after stretch.

Increased reverse-mode NCX should cause a net gain in intracellular Ca2+.

Therefore, we directly tested the effects of stretch on [Ca2+]i transients in aequorin-loaded muscle strips (Tyrode, n=4). Stretching aequorin-loaded muscle strips from L88 to L98 resulted in an SFR of 117 ± 4% (p<0.05), which was accompanied by a parallel increase in aequorin light transients to 121 ± 3% (p<0.05). These data suggest that increases in [Ca2+]i transients ultimately underlie the SFR in the rabbit heart (Fig. 6).

Fig. 6

Influence of stretch on SFR and aequorin light emission. Bar graphs indicate percent changes of twitch force during the SFR (left) and of the corresponding aequorin light transient (right) normalized to the end of the first phase (100%). Average data from 4 preparations from 4 hearts (Tyrode).

Fig. 6

Influence of stretch on SFR and aequorin light emission. Bar graphs indicate percent changes of twitch force during the SFR (left) and of the corresponding aequorin light transient (right) normalized to the end of the first phase (100%). Average data from 4 preparations from 4 hearts (Tyrode).

Discussion

The results of the present study demonstrate that (1) stretching isolated rabbit myocardium results in an immediate, followed by a slow force response; (2) the SFR is associated with an increase in [Na+]i and [Ca2+]i transients; (3) both NHE1 and reverse-mode NCX partly contribute to the SFR; (4) delayed pHi changes may occur, but do not contribute to the SFR neither in bicarbonate-containing nor bicarbonate-free buffer.

A biphasic response to stretch was initially demonstrated by Parmely and Chuck in isolated mammalian cardiac muscle [1]. The initial response to stretch (Frank–Starling mechanism) was shown to be independent of increases in [Ca2+]i and was related to enhanced responsiveness of the myofilaments to Ca2+ (for reviews see Fuchs and Smith [16]). In contrast, Allen and Kurihara demonstrated slowly rising [Ca2+]i transients as the underlying mechanism for the SFR, but the source of [Ca2+]i remained obscure [3]. Recently the subcellular mechanism of the SFR has been studied in more detail. Using rat and feline myocardium, Cingolani and coworkers reported a stretch-dependent autocrine/paracrine stimulation of angiotensin-II and endothelin-1 receptors with subsequent increase of NHE1 activity [4]. The latter resulted in elevated [Na+]i leading to Ca2+ gain by increasing reverse-mode NCX. The stretch-dependent release of angiotensin-II and endothelin-1 from isolated myocardium was in line with previous work of Sadoshima et al. in isolated rat myocytes [17]. However, it remains unknown at present whether the SFR is a general phenomenon in mammalian myocardium and whether the signal transduction mechanisms are identical among species. For example, we recently reported that stimulation of Ang II or ET-1 receptors plays no role in the generation of the SFR in rabbit and human ventricular myocardium, suggesting species-dependent differences in signalling pathways [6,8]. In addition, the contribution of intracellular alkalinization to the SFR remains a matter of debate.

Influence of stretch on force and Ca2+, Na+, and H+ homeostasis in rabbit myocardium

Excitation–contraction (e–c) coupling in rabbit myocardium resembles e–c coupling in human myocardium [18]. Therefore, we further investigated the mechanism(s) of the SFR in rabbit myocardium. Rabbit myocardium was characterized by a slowly developing, 25–30% increase in twitch force during the late phase after stretch. This is in line with previous results in rat, cat, and human cardiac preparations and underlines that the SFR is a general phenomenon in mammalian cardiac muscle [2,4,6,8,9,19].

The mechanism(s) underlying the SFR are under debate. Increases in [Ca2+]i transients and SR Ca2+ content have been associated with the SFR, and may even be underestimated given the simultaneous stretch-dependent increase in troponin C Ca2+ affinity, acting as a potent Ca2+ buffer [8,20]. In line with previous data from Allen and Kurihara, we demonstrate increases in [Ca2+]i transients as the final mechanism for the SFR in rabbit myocardium in the present work [3].

However, mechanoreceptors and transducers involved in the increase in [Ca2+]i transients remain to be characterized. In addition, it is unclear whether intracellular alkalinization, followed by increased Ca2+ sensitivity of the myofilaments, might contribute to the SFR. Using HOE 642 for inhibition of the NHE1 in Tyrode's solution and HEPES buffer, we and others have shown that NHE1 stimulation is involved in the stretch-induced slow increase in force and this notion is further strengthened by the ∼50% reduction of SFR after NHE 1 inhibition in the present work [4,6,9].

Despite NHE1 stimulation, however, we did not observe an increase in pHi during the rising phase of the SFR. Only after the SFR had fully developed, pHi slightly increased. There are at least two possible explanations for the lack of pHi changes during the SFR: (1) other pHi-dependent transporters, such as the Cl/HCO3 exchanger, compensate for the extrusion of protons by the NHE1; (2) intracellular H+ buffering may prevent a change in pHi despite H+ efflux early after stimulation of the NHE1. Our results provide clear experimental evidence to support a role for both mechanisms: (a) the pHi increase following stretch is larger in HEPES- than in bicarbonate-buffered solutions, as demonstrated in this and a previous study [2]; (b) inhibition of Cl/HCO3 exchange by means of an antibody alters the SFR [21]; (c) intracellular H+ buffering is potent (>1000:1, i.e. for each free proton there are more than 1000 protons bound to intracellular buffers; [18]); (d) the intracellular buffering power for protons in ventricular myocytes is more than doubled in physiological, bicarbonate-containing (48.3 mM) as compared to unphysiological, bicarbonate-free solution (21.2 mM) [22].

In contrast to H+, intracellular Na+ buffering is small. Despa and Bers determined a value of 1.39:1 in rabbit ventricular myocytes, almost three orders of magnitude lower than H+ buffering [23]. Thus, it is likely that stretch-dependent stimulation of NHE1 could induce increases of [Na+]i in the absence of detectable changes in pHi. Consistent with this notion, [Na+]i clearly increased after stretch in the present study. The [Na+]i increase was ∼3 mM, in fair agreement with the 6 mM reported by Alvarez et al. in rat myocardium [2]. Importantly, the increase in [Na+]i was substantially inhibited by HOE 642 indicating that it was mediated in large part through stimulation of NHE1. Moreover, it started almost immediately after the stretch and had a time course similar to the SFR suggesting that the two phenomena are causally related. Elevated [Na+]i, in turn, is expected to increase intracellular Ca2+ by attenuating forward mode and facilitating reverse-mode Na+/Ca2+ exchange [5,8,24]. In line with this hypothesis we found that KB-R 7943, an inhibitor of reverse-mode Na+/Ca2+ exchange, suppressed the SFR by ∼50%.

Taken together, the results of the present as well as previous studies suggest the following sequence of events: (1) Stretch activates the NHE1 through an as yet unknown mechanosensor; (2) NHE1 stimulation increases [Na+]i; (3) a simultaneous increase in pHi is prevented by efficient intracellular H+ buffering and, possibly, stimulation of compensatory transporters, particularly in physiological, bicarbonate-containing solutions; (4) the elevation of [Na+]i elicits a secondary gain in [Ca2+]i mediated largely by reverse-mode NCX; (5) the increase in [Ca2+]i (and [Ca2+]i transients) causes augmented contractions and thus underlies the SFR.

It is noteworthy that inhibition of NHE1 and reverse-mode NCX, either alone or in combination, did not completely suppress the SFR. In the presence of both HOE 642 and KB-R 7943 (at concentrations assumed to be as specific and effective as possible in inhibiting NHE1 and NCX), a slow force response to ∼114% remained, implying that an as yet unidentified mechanism is responsible for ∼50% of the SFR. The lack of effect of blockers of stretch-activated cation channels (SACs) observed in previous studies makes a contribution of these Na+- and Ca2+-conducting ion channels to the SFR unlikely [6,9]. Likewise, the involvement of the positive inotropes angiotensin II and endothelin-1 can be ruled out from previous experiments [6,9]. Thus, further studies are necessary to identify this novel (presumably Na+-independent) slowly developing, stretch-activated positive inotropic mechanism.

Limitations of the study

Our conclusion that the stretch-induced [Na+]i and [Ca2+]i increases are mediated by NHE1 stimulation followed by decreased [Ca2+] efflux and/or increased influx via NCX relies on the assumption that HOE642 and KB-R 7943, as used in our study, selectively inhibit NHE1 and NCX, respectively. Reported IC50 values for the inhibition of NHE1 and reverse-mode NCX activities amount to 30 nmol/l and 0.3–2.4 μmol/l, respectively, indicating potent inhibition of the two Na+-dependent transporters by HOE642 and KB-R 7943 [25–27]. Presumably, however, neither drug may be entirely selective. Therefore, results obtained with HOE642 and KB-R 7943 have to be interpreted with caution. For example, HOE642 has been shown to reduce the slowly inactivating component of the voltage-activated Na+ current (INa(slow)) [28]. This current is activated by hypoxia or the ischaemic metabolite lysophosphatidylcholine, but almost absent under normal physiological conditions [28–30]. Thus, a possible contribution of INa(slow) to the stretch-induced (and HOE642-sensitive) [Na+]i increase is conceivable under the assumption that stretch significantly activates INa(slow). At present, however, there is no evidence suggesting a stretch-dependent stimulation of INa(slow) in cardiac myocytes [31–34]. Furthermore, if INa(slow) was activated by stretch, this would have to result in an alteration of the action potential. Direct measurements of the action potential, however, clearly showed that during our experimental protocol stretch does not change the action potential [6].

Similarly, KB-R 7943 was reported to inhibit Na+, Ca2+, and K+ currents with IC50 values of 14, 8, and 7 μM [26]. The authors of that study, however, admitted that the IC50 values could be underestimated due to the voltage protocols used. By contrast, in another study 10 μM KB-R 7943 did not affect NHE activity, DHP-sensitive Ca2+ uptake, passive Na+ uptake as well as Ca2+-ATPases and the Na+/K+-ATPase [25]. According to these results, 5 μM KB-R 7943 is expected to block NCX activity almost completely with minimal impact on other ion transporters and channels. Most importantly, however, direct measurements of action potential parameters in guinea-pig papillary muscle, a preparation very similar to ours, revealed that KB-R 7943, at 10 μM, did not affect resting membrane potential, action potential amplitude, the maximum rate of rise, and the action potential duration at 90% repolarization [25]. This is compelling evidence that KB-R 7943, at 5 μM, is unlikely to inhibit Na+, Ca2+, and K+ currents significantly in multicellular papillary muscle preparations. Collectively, these data suggest that HOE642 and KB-R 7943, as used in our study, acted mainly through inhibition of NHE1 and NCX activities and support our interpretation of the data that both NHE1 and NCX are involved in the stretch-induced SFR in rabbit myocardium.

Conclusions

In conclusion, stretch activates a [Na+]i-dependent SFR in rabbit myocardium. NHE1 stimulation underlies [Na+]i increases. Changes in pHi are negligible. They occur late after stretch and do not directly contribute to the SFR. The SFR may play a role in adapting cardiac contractility to conditions of sustained hemodynamic load.

Acknowledgements

This work was supported by a grant from the Deutsche Forschungsgemeinschaft (PI-414/3) and the German Federal Ministry of Education and Research (Competence Network Heart Failure, TP 8, Basic Mechanisms).

References

[1]
Parmely
W.W.
Chuck
L.
Length-dependent changes in myocardial contractile state
Am J Physiol
 
1973
224
1195
1199
[2]
Alvarez
B.V.
Perez
N.G.
Ennis
I.L.
Camilion de Hurtado
M.C.
Cingolani
H.E.
Mechanisms underlying the increase in force and Ca(2+) transient that follow stretch of cardiac muscle: a possible explanation of the Anrep effect
Circ Res
 
1999
85
716
722
[3]
Allen
D.G.
Kurihara
S.
The effects of muscle length on intracellular calcium transients in mammalian cardiac muscle
J Physiol
 
1982
327
79
94
[4]
Cingolani
H.E.
Alvarez
B.V.
Ennis
I.L.
Camilion de Hurtado
M.C.
Stretch-induced alkalinization of feline papillary muscle: an autocrine–paracrine system
Circ Res
 
1998
83
775
780
[5]
Perez
N.G.
de Hurtado
M.C.
Cingolani
H.E.
Reverse mode of the Na+–Ca2+ exchange after myocardial stretch: underlying mechanism of the slow force response
Circ Res
 
2001
88
376
382
[6]
von Lewinski
D.
Stumme
B.
Maier
L.S.
Luers
C.
Bers
D.M.
Pieske
B.
Stretch-dependent slow force response in isolated rabbit myocardium is Na+ dependent
Cardiovasc Res
 
2003
57
1052
1061
[7]
Orchard
C.H.
Kentish
J.C.
Effects of changes of pH on the contractile function of cardiac muscle
Am J Physiol
 
1990
258
C967
C981
[8]
Cingolani
H.E.
Perez
N.G.
Pieske
B.
von Lewinski
D.
Camilion de Hurtado
M.C.
Stretch-elicited Na+/H+ exchanger activation: the autocrine/paracrine loop and its mechanical counterpart
Cardiovasc Res
 
2003
57
953
960
[9]
von Lewinski
D.
Stumme
B.
Fialka
F.
Luers
C.
Pieske
B.
Functional relevance of the stretch-dependent slow force response in failing human myocardium
Circ Res
 
2004
94
1392
1398
[10]
Calaghan
S.C.
White
E.
Contribution of angiotensin II, endothelin 1 and the endothelium to the slow force response to stretch in ferret papillary muscle
Pflugers Arch
 
2001
441
514
520
[11]
Pieske
B.
Kretschmann
B.
Meyer
M.
Holubarsch
C.
Weirich
J.
Posival
H.
et al
Alterations in intracellular calcium handling associated with the inverse force–frequency relation in human dilated cardiomyopathy
Circulation
 
1995
92
1169
1178
[12]
Hasenfuss
G.
Maier
L.S.
Hermann
H.P.
Luers
C.
Hunlich
M.
Zeitz
O.
et al
Influence of pyruvate on contractile performance and Ca(2+) cycling in isolated failing human myocardium
Circulation
 
2002
105
194
199
[13]
Pieske
B.
Maier
L.S.
Piacentino
V.
III
Weisser
J.
Hasenfuss
G.
Houser
S.
Rate dependence of [Na+]i and contractility in nonfailing and failing human myocardium
Circulation
 
2002
106
447
453
[14]
Pieske
B.
Sutterlin
M.
Schmidt-Schweda
S.
Minami
K.
Meyer
M.
Olschewski
M.
et al
Diminished post-rest potentiation of contractile force in human dilated cardiomyopathy. Functional evidence for alterations in intracellular Ca2+ handling
J Clin Invest
 
1996
98
764
776
[15]
Maier
L.S.
Pieske
B.
Allen
D.G.
Influence of stimulation frequency on [Na+]i and contractile function in Langendorff-perfused rat heart
Am J Physiol
 
1997
273
H1246
H1254
[16]
Fuchs
F.
Smith
S.H.
Calcium, cross-bridges, and the Frank–Starling relationship
News Physiol Sci
 
2001
16
5
10
[17]
Sadoshima
J.
Xu
Y.
Slayter
H.S.
Izumo
S.
Autocrine release of angiotensin II mediates stretch-induced hypertrophy of cardiac myocytes in vitro
Cell
 
1993
75
977
984
[18]
Bers
D.M.
Exitation–contraction coupling and cardiac contractile force, 2nd ed.
2001
Dordrecht
Kluwer Academic Publishers
[19]
Calaghan
S.
White
E.
Activation of Na+–H+ exchange and stretch-activated channels underlies the slow inotropic response to stretch in myocytes and muscle from the rat heart
J Physiol
 
2004
559
205
214
[20]
Stuyvers
B.D.
McCulloch
A.D.
Guo
J.
Duff
H.J.
ter Keurs
H.E.
Effect of stimulation rate, sarcomere length and Ca(2+) on force generation by mouse cardiac muscle
J Physiol
 
2002
544
817
830
[21]
Cingolani
H.E.
Chiappe
G.E.
Ennis
I.L.
Morgan
P.G.
Alvarez
B.V.
Casey
J.R.
et al
Influence of Na+-independent Cl–HCO3 exchange on the slow force response to myocardial stretch
Circ Res
 
2003
93
1082
1088
[22]
Leem
C.H.
Lagadic-Gossmann
D.
Vaughan-Jones
R.D.
Characterization of intracellular pH regulation in the guinea-pig ventricular myocyte
J Physiol
 
1999
517
Pt 1
159
180
[23]
Despa
S.
Bers
D.M.
Na/K pump current and [Na](I) in rabbit ventricular myocytes: local [Na](I) depletion and na buffering
Biophys J
 
2003
84
4157
4166
[24]
Despa
S.
Islam
M.A.
Weber
C.R.
Pogwizd
S.M.
Bers
D.M.
Intracellular Na(+) concentration is elevated in heart failure but Na/K pump function is unchanged
Circulation
 
2002
105
2543
2548
[25]
Iwamoto
T.
Watano
T.
Shigekawa
M.
A novel isothiourea derivative selectively inhibits the reverse mode of Na+/Ca2+ exchange in cells expressing NCX1
J Biol Chem
 
1996
271
22391
22397
[26]
Watano
T.
Kimura
J.
Morita
T.
Nakanishi
H.
A novel antagonist, No. 7943, of the Na+/Ca2+ exchange current in guinea-pig cardiac ventricular cells
Br J Pharmacol
 
1996
119
555
563
[27]
Kawamoto
T.
Kimura
H.
Kusumoto
K.
Fukumoto
S.
Shiraishi
M.
Watanabe
T.
et al
Potent and selective inhibition of the human Na+/H+ exchanger isoform NHE1 by a novel aminoguanidine derivative T-162559
Eur J Pharmacol
 
2001
420
1
8
[28]
Chattou
S.
Coulombe
A.
Diacono
J.
Le Grand
B.
John
G.
Feuvray
D.
Slowly inactivating component of sodium current in ventricular myocytes is decreased by diabetes and partially inhibited by known Na(+)–H(+) Exchange blockers
J Mol Cell Cardiol
 
2000
32
1181
1192
[29]
Ju
Y.K.
Saint
D.A.
Gage
P.W.
Hypoxia increases persistent sodium current in rat ventricular myocytes
J Physiol
 
1996
497
Pt 2
337
347
[30]
Undrovinas
A.I.
Fleidervish
I.A.
Makielski
J.C.
Inward sodium current at resting potentials in single cardiac myocytes induced by the ischemic metabolite lysophosphatidylcholine
Circ Res
 
1992
71
1231
1241
[31]
Sasaki
S.
Sawada
Y.
Morita
M.
Hikita
Y.
Nakagaki
I.
Hori
S.
et al
Effects of a benzothiazepine calcium blocker on electrolyte alteration in human ischemic and reperfused myocardium
J Mol Cell Cardiol
 
1994
26
1037
1044
[32]
Hu
H.
Sachs
F.
Stretch-activated ion channels in the heart
J Mol Cell Cardiol
 
1997
29
1511
1523
[33]
Kamkin
A.G.
Kiseleva
I.S.
Iarygin
V.N.
Ion mechanisms of the mechanoelectrical feedback in myocardial cells
Usp Fiziol Nauk
 
2001
32
58
87
[34]
Zeng
T.
Bett
G.C.
Sachs
F.
Stretch-activated whole cell currents in adult rat cardiac myocytes
Am J Physiol Heart Circ Physiol
 
2000
278
H548
H557

Author notes

Time for primary review 18 days