The mechano-gated channel inhibitor GsMTx4 reduces the exercise pressor reflex in rats with ligated femoral arteries

We found that the mechanically sensitive component of the exercise pressor reflex contributes to its exaggeration in rats with ligated femoral arteries (model of simulated peripheral arterial disease). Thin-fiber mechanoreceptors, therefore, may contribute to the exaggerated exercise pressor reflex found in patients with peripheral arterial disease.

Abstract

Mechanical and metabolic stimuli arising from contracting muscles evoke the exercise pressor reflex. This reflex is greater in a rat model of simulated peripheral arterial disease in which a femoral artery is chronically ligated than it is in rats with freely perfused femoral arteries. The role played by the mechanically sensitive component of the exaggerated exercise pressor reflex in ligated rats is unknown. We tested the hypothesis that the mechano-gated channel inhibitor GsMTx4, a relatively selective inhibitor of mechano-gated Piezo channels, reduces the exercise pressor reflex in decerebrate rats with ligated femoral arteries. Injection of 10 μg of GsMTx4 into the arterial supply of the hindlimb reduced the pressor response to Achilles tendon stretch (a purely mechanical stimulus) but had no effect on the pressor responses to intra-arterial injection of α,β-methylene ATP or lactic acid (purely metabolic stimuli). Moreover, injection of 10 μg of GsMTx4 into the arterial supply of the hindlimb reduced both the integrated pressor area (control 535 ± 21, GsMTx4 218 ± 24 mmHg·s; P < 0.01), peak pressor (control 29 ± 2, GsMTx4 14 ± 3 mmHg; P < 0.01), and renal sympathetic nerve responses to electrically induced intermittent hindlimb muscle contraction (a mixed mechanical and metabolic stimulus). The reduction of the integrated pressor area during contraction caused by GsMTx4 was greater in rats with ligated femoral arteries than it was in rats with freely perfused femoral arteries. We conclude that the mechanically sensitive component of the reflex contributes to the exaggerated exercise pressor reflex during intermittent hindlimb muscle contractions in rats with ligated femoral arteries.

NEW & NOTEWORTHY

We found that the mechanically sensitive component of the exercise pressor reflex contributes to its exaggeration in rats with ligated femoral arteries (model of simulated peripheral arterial disease). Thin-fiber mechanoreceptors, therefore, may contribute to the exaggerated exercise pressor reflex found in patients with peripheral arterial disease.

the exercise pressor reflex (25) is evoked by mechanical and metabolic signals arising within contracting skeletal muscles and is manifested in part by increases in sympathetic nerve activity, arterial blood pressure, and cardiac output. The exercise pressor reflex has been shown in both humans and animals to increase arterial perfusion of contracting muscles (1, 17, 30). In patients with peripheral arterial disease (PAD), the exercise pressor reflex is exaggerated and results in greater increases in sympathetic nerve activity and blood pressure than those found in healthy, age-matched control subjects (3, 7, 8, 27, 28). Most importantly, the exaggerated exercise pressor reflex has been shown to increase the risk of cardiovascular morbidity and mortality in patients with PAD (12).

The exercise pressor reflex is also exaggerated in a rat model of simulated PAD in which a femoral artery is ligated for 72 h (44). The majority of the studies investigating the mechanisms that underlie the exaggerated exercise pressor reflex in rats with ligated femoral arteries have focused on the role played by the metabolically sensitive component of the reflex. Our laboratory has reported, for example, that stimulation of acid-sensing ion channel 3 (ASIC3) (45), purinergic 2X (P_2X) (41), and endoperoxide4 (EP4) (47) receptors contribute to the exaggerated exercise pressor reflex in rats with ligated femoral arteries. In contrast to what is known about the metabolically sensitive component of the reflex, the evidence suggesting that the mechanically sensitive component of the reflex also plays a role in the exaggerated exercise pressor reflex in rats with ligated femoral arteries is at best indirect. Lu et al. (21, 22) reported, for example, that stretching the Achilles tendon of rats with ligated femoral arteries produced increases in renal sympathetic nerve activity (RSNA) and blood pressure that were larger than those produced by stretching the tendon of rats with freely perfused femoral arteries. Moreover, in PAD patients Muller et al. (27, 28) found that the pressor responses to low levels of rhythmic plantar flexion exercise were greater than they were in healthy age-matched control subjects, a finding that suggests that the mechanically sensitive component of the exercise pressor reflex is exaggerated in this patient population. Despite these indirect observations, no study has investigated specifically the role played by the mechanically sensitive component of the exaggerated exercise pressor reflex either in rats with ligated femoral arteries or in patients with PAD.

We found recently that injection of the relatively selective mechano-gated channel inhibitor GsMTx4, a tarantula toxin that is an inhibitor of mechano-gated Piezo channels (4, 5, 26), into the arterial supply of the hindlimb reduced the pressor and sympathetic nerve responses to intermittent hindlimb muscle contraction of rats with freely perfused femoral arteries (9). This finding prompted us to conduct the present study, in which we tested the hypothesis that injection of GsMTx4 into the arterial supply of the hindlimb reduces the exaggerated sympathetic nerve and pressor responses to intermittent hindlimb muscle contractions in rats with ligated femoral arteries. We also tested the hypothesis that the magnitude of the reduction of the pressor response to contraction caused by GsMTx4 in rats with ligated femoral arteries was greater than the magnitude of the reduction of the pressor response to contraction caused by GsMTx4 in rats with freely perfused femoral arteries.

MATERIALS AND METHODS

All experimental procedures were approved by the Institutional Animal Care and Use Committee of the Penn State College of Medicine. Experiments were performed on adult (∼12–15 wk old) male Sprague-Dawley rats (n = 29, body wt range 360–560 g; Charles River Laboratories, Malvern, PA). At the end of each experiment, the decerebrated rats (see below) were killed with an intravenous injection of saturated potassium chloride (>3 ml/kg) and the chest was opened bilaterally.

Surgical procedures.

Seventy-two hours before the experiments, 28 of the 29 rats in this study had their left femoral artery ligated (see below for more information on the 1 rat that was not subjected to a ligation procedure). Specifically, these 28 rats were anesthetized with 3–4% isoflurane (balance O₂), and the left femoral artery was surgically exposed and ligated (5-0 silk suture) ∼3 mm distal to the inguinal ligament. The incision was closed, and the rats were given 72 h to recover before any subsequent procedure. We did not perform any “sham” ligation procedures in this study because we have shown previously that a sham procedure does not result in an exaggerated pressor response to either static (11) or intermittent (10) hindlimb muscle contractions.

In vivo experiments were performed on 20 “ligated” rats. On the day of the experiment, rats were anesthetized with 4% isoflurane (balance O₂). Adequate depth of anesthesia was confirmed by the absence of toe-pinch and blink reflexes. The trachea was cannulated, and the lungs were mechanically ventilated (Harvard Apparatus, Holliston, MA) with the gaseous anesthetic until the decerebration was completed (see below). In all rats, the right jugular vein and both carotid arteries were cannulated with PE-50 catheters that were used for the injection of fluids and measurement of arterial blood pressure (P23 XL, Statham), respectively. Heart rate (HR) was calculated beat to beat from the arterial pressure pulse (Spike2 software). In 15 ligated rats, the left superficial epigastric artery was cannulated with a PE-8 catheter whose tip was placed near its junction with the femoral artery. A reversible snare (2-0 silk suture) was placed around the left iliac artery and vein. The superficial epigastric artery catheter was distal, and the snare was proximal, to the site of ligation in the rats whose femoral arteries were occluded. In 15 ligated rats, the left sciatic nerve and triceps surae (gastrocnemius, soleus, and plantaris) muscles were exposed and the left calcaneal bone was severed and linked by string to a force transducer (model FT10, Grass Technologies, Warwick, RI), which, in turn, was attached to a rack and pinion. In these 15 rats, bundles from the left renal sympathetic nerve were exposed with a retroperitoneal approach. The bundles were then glued (Kwik-Sil, World Precision Instruments, Sarasota, FL) onto a pair of thin stainless steel recording electrodes that, in turn, were connected to a high-impedance probe (Grass model HZP) and amplifier (Grass P511). Multiunit signals from the renal sympathetic nerve fibers were filtered at 100 Hz (low frequency) and 1 kHz (high frequency). At the end of each experiment, hexamethonium (10 mg) was injected into the jugular vein to abolish RSNA, thereby demonstrating that the activity was postganglionic.

After the initial surgical procedures were completed, the rats were placed in a Kopf stereotaxic frame and spinal unit with clamps placed around the pelvis and rostral lumbar vertebrae. Dexamethasone (0.2 mg iv) was injected to minimize brain stem edema. A precollicular decerebration was performed, and all neural tissue rostral to the section was aspirated. After the decerebration was completed, anesthesia was terminated and the lungs were mechanically ventilated with room air. The rats were given at least 1 h to recover and stabilize prior to the initiation of any experimental protocol. Experiments were performed in decerebrate, unanesthetized rats because anesthesia has been shown to depress the exercise pressor reflex in this species (34). Arterial blood gases and pH were measured periodically and maintained within normal limits (Pa_CO2: 35–45 mmHg, Pa_O2: ∼100 mmHg, pH: 7.35–7.45) by adjusting ventilation and/or administering intravenous sodium bicarbonate (8.5%). Core temperature was measured by a rectal probe and maintained at ∼37–38°C by a heating lamp.

Experimental protocols.

In nine ligated rats we compared the pressor, cardioaccelerator, and renal sympathetic nerve responses to both calcaneal (Achilles) tendon stretch and intermittent hindlimb muscle contractions before and after the injection of 10 μg of GsMTx4 into the arterial supply of the hindlimb. Control tendon stretch and intermittent contraction maneuvers were performed in random order and separated by ∼5–10 min. For tendon stretch, triceps surae muscle tension was set at ∼80–100 g and baseline data were collected for 30 s. We then stretched the calcaneal tendon for 30 s by rapidly turning the rack and pinion (38). For intermittent contractions, baseline tension was set and baseline data were collected as described for tendon stretch. We then stimulated the sciatic nerve (40 Hz, 0.01-ms pulse duration, 500-ms train duration, ≤2 × motor threshold) with shielded stimulating electrodes for 30 s. After the control stretch and contraction maneuvers, the snare on the left iliac artery and vein was tightened and 10 μg of GsMTx4 (dissolved in 0.1 ml of saline) was injected into the arterial supply of the hindlimb via the left superficial epigastric artery catheter. Ten minutes after injection the snare was released. Tightening the snare in this manner partially trapped GsMTx4 within the arterial supply of the hindlimb. Twenty minutes after the snare was released (i.e., 30 min after the injection of GsMTx4) the stretch and contraction maneuvers were repeated as described above.

In an additional group of six ligated rats we performed the same stretch and contraction maneuvers described above before and after GsMTx4 except that the toxin was injected into the jugular vein. We injected the toxin intravenously to determine whether circulation of GsMTx4 to central nervous system sites could explain the reduced pressor, cardioaccelerator, and RSNA responses that we observed after injection of GsMTx4 into the arterial supply of the hindlimb (see results).

In five ligated rats we compared the pressor responses to injection of α,β-methylene ATP and lactic acid into the arterial supply of the hindlimb before and after injection of 10 μg of GsMTx4 into the arterial supply of the hindlimb. Before initiating the protocol, we paralyzed the rats with pancuronium bromide (1 mg/kg iv). The snare around the left iliac artery and vein was tightened, and α,β-methylene ATP (20 μg in 0.2 ml of saline) and lactic acid (0.2 ml of a 24 mM concentration in saline) were injected, separately (at least 5 min apart) and in random order, as a bolus over ∼2 s. Ten micrograms of GsMTx4 was then injected into the arterial supply of the hindlimb via the left superficial epigastric artery catheter as described above, and the injections of α,β-methylene ATP and lactic acid were then repeated.

In each of the ligated rats in which we tested the effects of injection of GsMTx4 into the arterial supply of the hindlimb, we injected blue dye at the end of the experiment in the same manner as GsMTx4 was injected to determine that the toxin did in fact have access to the triceps surae muscles.

After all experiments in which the sciatic nerve was stimulated to induce muscle contractions, rats were paralyzed with pancuronium bromide (1 mg/kg iv) and the sciatic nerve was stimulated with the same parameters as those used to induce muscle contraction. This was done to ensure that the pressor response to contractions was not the result of electrical activation of the axons of thin-fiber afferents in the sciatic nerve. Stimulation of the sciatic nerve after pancuronium bromide injection did not increase blood pressure, HR, or RSNA in any of the experiments in this study.

Western blotting assays.

In eight ligated rats that were not used in any of the experiments described above, both the left (which innervated the ligated hindlimb) and right (which innervated the freely perfused hindlimb) L₄ and L₅ dorsal root ganglia (DRGs) were harvested. Total protein was isolated with the NucleoSpin Protein Kit (Macherey-Nagel, Bethlehem, PA) according to the manufacturer's instructions and as described previously (14). Western blot experiments were performed with the Wes System (Protein Simple, San Jose, CA), and tissue samples and reagents were prepared according to the manufacturer's instructions. The protein samples, primary and secondary antibodies, blocking reagent, wash buffer, and chemiluminescent substrate were loaded on to the provided microplate. For Piezo1, we loaded 0.2 μg/μl protein per lane, the anti-Piezo1 rabbit antibody (Alomone Labs, Jerusalem, Israel) was used at 1:50 dilution, and the anti-vinculin (loading control) rabbit antibody was used at a dilution of 1:1,200 (Abcam, Cambridge, UK). For Piezo2, we loaded 0.1 μg/μl protein per lane, the anti-Piezo2 rabbit antibody (a kind gift from Dr. Ardem Patapoutian of The Scripps Research Institute, La Jolla, CA) (46) was diluted at 1:600, and the anti-vinculin antibody was diluted at 1:600. For Western blots in which we assessed the ratio of vinculin to actin (n = 7), we loaded 0.2 μg/μl protein per lane. The anti-vinculin rabbit antibody was used at a dilution of 1:1,200 (Abcam) and the anti-actin antibody at a dilution of 1:100 (EMD Millipore, Billerica, MA). Data analysis was performed with Compass software (Protein Simple).

Comparison of effects of GsMTx4 in ligated vs. freely perfused rats.

The surgical procedures (except for the ligation procedure), the intermittent contraction protocol, and the 10-μg dose of GsMTx4 and timing of the subsequent bout of intermittent contractions used in this study in ligated rats were identical to those used in our previous study in freely perfused rats (9). We used the same procedures and protocols so that we could compare the effects of GsMTx4 on the pressor response to intermittent contractions between “ligated” and “freely perfused” rats. Moreover, we performed an additional experiment in one freely perfused rat in which we tested the effects of GsMTx4 on the pressor response to contraction [using the same protocol as described above for ligated rats and previously for freely perfused rats (9)]. We added the results of the additional experiment to the freely perfused rat sample (n = 8) from our previous study (9) to increase the statistical power of the between-group comparisons (n = 9 in each group). Specifically, we compared the absolute and relative (normalized to the control condition) reductions in the peak pressor response and the integrated area of the pressor response to intermittent contractions caused by GsMTx4 between groups. We did not compare the effects of GsMTx4 on the pressor response to tendon stretch between ligated and freely perfused rats because there were differences between the tendon stretch protocol used in this study in ligated rats and the tendon stretch protocol used in our previous study in freely perfused rats (9). For example, in this study the rats were not paralyzed with pancuronium bromide during the tendon stretch protocol, whereas in the previous study (9) they were paralyzed with pancuronium bromide, a compound that markedly increases baseline HR.

Data analysis.

Data were collected with a Spike2 data acquisition system (Cambridge Electronic Design) and saved for future off-line analysis. The original RSNA recordings were integrated and corrected for background noise recorded after hexamethonium administration. Baseline mean arterial pressure (MAP), HR, and RSNA values were determined from the 30-s baseline periods that preceded the stretch or contraction. The peak pressor and cardioaccelerator responses to stretch and intermittent contraction were calculated as the difference between the peak MAP and HR values wherever they occurred during the maneuvers and their corresponding baseline MAP and HR. Second-by-second time courses of the pressor and RSNA responses were plotted as their change (absolute increase for MAP and relative increase for RSNA) from baseline during the 30-s stretch or contraction periods. The areas of the second-by-second time course of the pressor response to contraction were integrated (mmHg·s). Second-by-second integrated RSNA values were analyzed and expressed as a percent change from baseline. The tension-time indexes (TTIs, in kg·s) for stretch and intermittent contraction were calculated by subtracting the area under the tension signal trace for the 30-s baseline period from the area under the tension signal trace for the 30-s stretch or contraction period.

All data are expressed as means ± SE. Within-group (control vs. GsMTx4) comparisons were performed with paired Student's t-tests (peak and TTI data) or two-way (condition and time) repeated-measures ANOVAs (time course data) with Holm-Sidak post hoc tests used as indicated from the ANOVA result. Between-group (ligated vs. freely perfused) comparisons of peak and integrated pressor area data were performed with unpaired Student's t-tests. The criterion for statistical significance was P < 0.05.

RESULTS

Stretching the Achilles tendon of ligated rats for 30 s produced sustained increases in blood pressure and transient (i.e., lasted for ∼5 s) increases in RSNA. Injection of 10 μg of GsMTx4 into the arterial supply of the hindlimb of ligated rats (n = 9) significantly reduced the peak pressor response to tendon stretch but did not have a statistically significant effect (P = 0.06) on the cardioaccelerator response (Table 1). Second-by-second analysis revealed that GsMTx4 significantly reduced the pressor response to tendon stretch at multiple time points throughout the duration of the stretch and significantly reduced RSNA during the fourth second of the stretch (Fig. 1). Intravenous injection of 10 μg of GsMTx4 in ligated rats (n = 6) had no effect on the peak pressor or cardioaccelerator response (Table 1) or the time course of the pressor response and RSNA (Fig. 1) to tendon stretch. There were no differences in the TTI (Table 1) or the tension time course (Fig. 1) during tendon stretch between the control and GsMTx4 conditions for either the intra-arterial or intravenous injection group.

Table 1.

Baseline and peak change in MAP and HR in control and GsMTx4 conditions for tendon stretch and intermittent contractions in ligated rats

	Intra-arterial Injection (n = 9)		Intravenous Injection (n = 6)
	Control	GsMTx4	Control	GsMTx4
Tendon stretch
Baseline MAP, mmHg	97 ± 6	101 ± 7	91 ± 4	96 ± 6
Peak ΔMAP, mmHg	18 ± 2	11 ± 2*	18 ± 2	20 ± 3
Baseline HR, bpm	367 ± 9	369 ± 18	354 ± 12	369 ± 19
Peak ΔHR, bpm	14 ± 3	9 ± 2	12 ± 3	15 ± 4
TTI, kg·s	12 ± 1	12 ± 1	11 ± 3	11 ± 1
Intermittent contractions
Baseline MAP, mmHg	97 ± 7	101 ± 8	91 ± 6	97 ± 7
Peak ΔMAP, mmHg	29 ± 2	14 ± 3*	25 ± 3	26 ± 4
Baseline HR, bpm	368 ± 11	369 ± 15	354 ± 12	369 ± 19
Peak ΔHR, bpm	18 ± 3	11 ± 3*	26 ± 5	22 ± 6
TTI, kg·s	11 ± 1	12 ± 1	12 ± 2	13 ± 2

Values are mean ± SE baseline and peak change in (Δ) mean arterial pressure (MAP) and heart rate (HR) in the control and GsMTx4 conditions (intra-arterial and intravenous injection groups) for both tendon stretch and intermittent contractions in ligated rats.

TTI, tension-time index; bpm, beats per minute.

^*P < 0.05.

Fig. 1.

Injection of 10 μg of GsMTx4 into the arterial supply of the hindlimb of rats with a ligated femoral artery (ia, n = 9; left) significantly reduced the pressor response and renal sympathetic nerve activity (RSNA) at specific time points during tendon stretch. Horizontal black line and/or asterisks indicate a statistically significant difference (P < 0.05) from control. Injection of GsMTx4 into the jugular vein (iv, n = 6; right) had no effect on the pressor response and RSNA during tendon stretch. The tension development was not different between control and GsMTx4 conditions at any time point in either group of rats. MAP, mean arterial pressure.

Stimulating the sciatic nerve of ligated rats for 30 s to intermittently contract the hindlimb muscles markedly increased arterial blood pressure and RSNA (Fig. 2). In the nine ligated rats in which we tested the effects of intra-arterial injection of GsMTx4 on the responses to tendon stretch, injection of 10 μg of GsMTx4 into the arterial supply of the hindlimb significantly reduced the peak pressor and cardioaccelerator responses to intermittent contraction (Table 1). Second-by-second analysis revealed that GsMTx4 significantly reduced the time course of the pressor response from the fifth second on and significantly reduced RSNA from the first through fifth seconds of contraction (Fig. 3). In the six ligated rats in which we tested the effects of intravenous injection of GsMTx4 on the responses to tendon stretch, injection of 10 μg of GsMTx4 into the jugular vein had no effect on either the peak pressor or cardioaccelerator responses (Table 1) or the time courses of the pressor response and RSNA (Fig. 3) to intermittent contractions. There were no differences in either the TTI (Table 1) or the tension time course (Fig. 1) during intermittent contractions between the control and GsMTx4 conditions for either the intra-arterial or intravenous injection group.

Fig. 2.

Example of original data from 1 ligated rat showing the effects of 10 μg of GsMTx4 on RSNA and blood pressure (BP) responses to the first ∼15 s of intermittent contractions. Note that in the control condition the large increase in RSNA during the first ∼5 s of contraction was followed by a synchronization of RSNA bursts to the rising phase of the tension trace. GsMTx4 markedly reduced the initial increase in RNSA as well as the subsequent synchronization between RSNA and tension development. a.u., Arbitrary units.

Fig. 3.

Injection of 10 μg of GsMTx4 into the arterial supply of the hindlimb of rats with a ligated femoral artery (n = 9; left) significantly reduced the pressor and RSNA responses at specific time points during intermittent contraction. The injection of GsMTx4 into the jugular vein (n = 6; right) had no effect on the pressor and RSNA responses to intermittent contractions. The tension values in both groups represent the average tension over each 1-s period, which is why the plots do not appear to represent intermittent contractions. The tension development was not different between control and GsMTx4 conditions at any time point in either group of rats. Horizontal black lines and/or asterisks indicate statistically significant (P < 0.05) differences from control.

In five rats the injection of 10 μg of GsMTx4 into the arterial supply of the hindlimb had no effect on the pressor response to intra-arterial injection of α,β-methylene ATP or lactic acid (Fig. 4), findings that are consistent with the fact that GsMTx4 did not block metabolically sensitive channels.

Fig. 4.

Injection of 10 μg of GsMTx4 into the arterial supply of the hindlimb (n = 5) had no effect on the peak pressor responses that resulted from the injection of α,β-methylene ATP (top) or lactic acid (bottom) into the arterial supply of the hindlimb.

In eight rats there were no differences in the Piezo1-to-vinculin ratio or the Piezo2-to-vinculin ratio between protein extracted from DRG tissue innervating the ligated hindlimb and protein extracted from DRG tissue innervating the freely perfused hindlimb (Fig. 5). Likewise, in the seven rats tested, there was no significant difference between the vinculin-to-actin ratio in the freely perfused rats (1.2 ± 0.3) and the vinculin-to-actin ratio in the ligated rats (1.4 ± 0.2; P = 0.28). This finding indicates that vinculin was an appropriate loading control for the Piezo channel comparisons between freely perfused and ligated rats.

Fig. 5.

Top: example of the Western blots for Piezo1 and Piezo2 channels in dorsal root ganglion (DRG) tissue innervating the freely perfused (FP) hindlimb and the ligated (Lig) hindlimb from 1 rat with vinculin used as a loading control. For Piezo1, we loaded 0.2 μg/μl protein per lane, the anti-Piezo1 antibody was used at 1:50 dilution, and the anti-vinculin antibody was used at a dilution of 1:1,200. For Piezo2, we loaded 0.1 μg/μl protein per lane, and both the anti-Piezo2 and anti-vinculin antibodies were diluted at 1:600. Bottom: group mean data (n = 8). Note that the exposure times for anti-Piezo1 and anti-vinculin were the same, whereas the exposure times for anti-Piezo2 and anti-vinculin were different. Consequently, the anti-vinculin blot was placed adjacent to the anti-Piezo2 blot when in fact it was in the same lane as the anti-Piezo2 blot.

The absolute, but not the relative, reduction of the integrated pressor area induced by the injection of GsMTx4 into the arterial supply of the hindlimb was significantly greater in ligated rats than in freely perfused rats. Further analysis revealed that the absolute reduction of the integrated pressor area caused by GsMTx4 was significantly greater in ligated rats than in freely perfused rats during the first 10 s of contraction but was not different between ligated and freely perfused rats during seconds 11–30 of contraction. The relative reduction of the integrated pressor area caused by GsMTx4 was not different between ligated and freely perfused rats during any time period of the contraction (Table 2).

Table 2.

Comparison between freely perfused and ligated rats of absolute and relative reductions of the pressor response to intermittent contractions caused by GsMTx4

	Freely Perfused (n = 9)	Ligated (n = 9)	P Value
ΔMAP total area, mmHg·s	−190 ± 33	−316 ± 21	0.046*
ΔMAP total area, %	−55 ± 8	−57 ± 3	0.462
ΔMAP s 1–10 area, mmHg·s	−44 ± 13	−105 ± 9	0.032*
ΔMAP s 1–10 area, %	−38 ± 12	−55 ± 6	0.221
ΔMAP s 11–20 area, mmHg·s	−71 ± 13	−110 ± 8	0.093
ΔMAP s 11–20 area, %	−59 ± 7	−53 ± 3	0.317
ΔMAP s 21–30 area, mmHg·s	−75 ± 13	−102 ± 7	0.151
ΔMAP s 21–30 area, %	−66 ± 8	−58 ± 3	0.248
ΔMAP peak, mmHg	−9 ± 2	−15 ± 1	0.098
ΔMAP peak, %	−43 ± 6	−52 ± 4	0.276

Values are means ± SE for freely perfused and ligated rats of the absolute (mmHg or mmHg·s) and relative (%) reductions of the pressor response to intermittent contractions caused by GsMTx4. In the control condition both the total pressor area (freely perfused 361 ± 61, ligated 535 ± 21 mmHg·s, P = 0.034) and the peak pressor (freely perfused 23 ± 2, ligated 29 ± 2 mmHg, P = 0.044) were greater in ligated rats than they were in freely perfused rats. TTIs for intermittent contraction were not different (P > 0.05) between conditions (control vs. GsMTx4) or across groups (freely perfused vs. ligated). Freely perfused data are from our recent publication (9), with 1 rat added to that sample to increase the statistical power of the comparison to ligated rats.

^*P < 0.05.

The absolute and relative reductions of the peak pressor response to intermittent contraction induced by the injection of GsMTx4 into the arterial supply of the hindlimb were not statistically different between ligated and freely perfused rats (Table 2).

DISCUSSION

We found that injection of the mechano-gated channel inhibitor GsMTx4 into the arterial supply of the hindlimb of rats with ligated femoral arteries reduced the pressor and sympathetic nerve responses to tendon stretch and intermittent hindlimb muscle contraction. Intravenous injection of GsMTx4 had no effect on these responses, a finding that indicates that the toxin did not circulate systemically to exert its effects within the central nervous system. Moreover, injection of GsMTx4 into the arterial supply of the hindlimb of ligated rats had no effect on the pressor responses to intra-arterial injection of α,β-methylene ATP or lactic acid, findings that indicate that GsMTx4 did not inhibit metabolically sensitive receptors. We also found that the absolute reduction of the integrated pressor area during contraction caused by GsMTx4 was greater in ligated rats than it was in freely perfused rats. Collectively, these findings indicate that the mechanically sensitive component of the exercise pressor reflex contributes to its exaggeration during intermittent contractions in ligated rats.

We used GsMTx4 to inhibit mechano-gated channels (42). GsMTx4 inhibits these channels by inserting into the lipid bilayer at the channel-membrane interface and “pushing” the channel to a closed state (43). Importantly, GsMTx4 inhibits both Piezo1 (4, 5, 26) and Piezo2 (33), with at least some selectivity for Piezo channels because it did not inhibit mechano-gated TREK-1 channels (5). Our finding that 10 μg of GsMTx4 did not inhibit metabolically sensitive receptors in the present study is important because GsMTx4 has been reported to exert off-target effects. For example, GsMTx4 has been shown to inhibit voltage-gated sodium (Na_V)1.7 channels (32), which are responsible, at least in part, for the afferent transmission of the exercise pressor reflex (40). If GsMTx4 had blocked Na_V1.7 channels in our study the pressor responses to injection of α,β-methylene ATP and lactic acid would likely have been reduced.

We could have used gadolinium to inhibit mechano-gated channels to investigate the role played by the mechanically sensitive component of the exercise pressor reflex in ligated rats. Gadolinium has been used, for example, to show that the mechanically sensitive component of the exercise pressor reflex is important in cats (15, 16) and that it underlies the exaggerated exercise pressor reflex found in rats with heart failure (35). We used GsMTx4 in this study because it is a more selective mechano-gated inhibitor than gadolinium (5) and because we have reported recently that GsMTx4 reduced the mechanically sensitive component of the exercise pressor reflex during intermittent contractions in rats with freely perfused femoral arteries, a finding that we attributed primarily to the GsMTx4-induced inhibition of mechano-gated Piezo channels (9).

Our finding that the absolute reduction of the integrated pressor area during contractions caused by GsMTx4 was greater in ligated rats than it was in freely perfused rats suggests to us that the mechanically sensitive component of the exercise pressor reflex is exaggerated in ligated rats compared with freely perfused rats. The exaggerated mechanically sensitive component of the reflex in ligated rats may be caused by an increase in the number of mechano-gated channels on the sensory endings of the thin-fiber muscle afferents, an increase in metabolite-induced sensitization of mechano-gated channels, or some combination of both. Our finding that Piezo1 and Piezo2 expression were not different between DRG tissue innervating freely perfused and ligated hindlimbs suggests that an increased metabolite-induced sensitization of mechano-gated channels underlies the exaggerated mechanically sensitive component of the reflex in ligated rats. Several lines of in vitro and in vivo evidence raise the possibility that bradykinin is a prime candidate for this metabolite-induced sensitization (13, 21, 39). Other metabolites, such as prostaglandin E₂ (PGE₂), have also been shown to sensitize sensory neurons (36, 37) and have been shown to contribute to the exaggerated exercise pressor reflex found in ligated rats (47). Nevertheless, specific interactions between PGE₂ and Piezo channels have not, to our knowledge, been investigated.

The absolute reduction of the integrated pressor area during contraction caused by GsMTx4 was greater in ligated rats than it was in freely perfused rats during the first 10 s of contraction, whereas the reduction of the integrated pressor area was the same between ligated and freely perfused rats during the remaining two-thirds of the contraction period. This finding suggests that the sensitization of mechano-gated channels in ligated rats was limited to the onset (i.e., first 10 s) of the contraction period. This may be explained by the fact that a low pH has been shown to markedly reduce Piezo1 channel activation (6). Ligating a femoral artery in a rat for 72 h reduces blood flow reserve capacity during exercise to ∼20–30% of normal (31). Such a reduced blood flow response to contraction in ligated rats compared with that in freely perfused rats would be expected to progressively reduce pH during the 30-s contraction period used in this study and may have resulted in the inactivation of some Piezo channels during the latter two-thirds of the contraction period. We should point out, however, that GsMTx4 inhibits both Piezo1 and Piezo2, and, consequently, we do not know whether one or the other, or both, contribute to evoke the exercise pressor reflex. Although low pH has been shown to inactivate Piezo1 (6), the effects of low pH on Piezo2 are unknown.

There are three important considerations of this study. The first is that we found that the absolute reduction of the integrated pressor area during contraction caused by GsMTx4 was greater in ligated rats than it was in freely perfused rats. Both the integrated pressor area and the peak pressor response to contraction provide important information. The peak pressor response, for example, provides an index of the peak workload placed on the heart, whereas the integrated pressor area provides an index of the workload placed on the heart throughout the contraction period. The second consideration is that we found that the relative reductions of the peak pressor response and integrated pressor area during contraction caused by GsMTx4 were not different between ligated and freely perfused rats. Some may argue that the relative reductions are the appropriate comparison between groups because the peak pressor response and integrated pressor area were both greater in ligated rats than they were in freely perfused rats in the control condition. The absolute reductions of these variables caused by GsMTx4, however, reflect the actual physiological contributions of the mechanically sensitive component of the exercise pressor reflex. Nevertheless, we have presented both the absolute and relative comparisons to be complete. The third consideration is that in humans PAD is an atherosclerotic process that develops slowly over time whereas in our experiments in rats simulated PAD was induced rapidly by ligating the femoral artery (29).

Our study has three limitations. First, our assessment of the effect of femoral artery ligation on Piezo channels was performed on the DRG instead of on the terminals of sensory nerve endings in the triceps surae muscles. Second, GsMTx4 does not distinguish between the role played by Piezo1 channels and that played by Piezo2 channels in evoking the mechanoreceptor component of the exercise pressor reflex. Third, GsMTx4 may block TRPC1 and TRPC6 channels, both of which are mechanosensitive; with regard to the last limitation, the threshold concentration of GsMTx4 needed to block TRPC1 (20) and TRPC6 (2) channels appears to be 8- to 10-fold higher than the calculated concentration of GsMTx4 in our experiments (9).

In conclusion, we found that the mechano-gated channel inhibitor GsMTx4, a relatively selective inhibitor of mechano-gated Piezo channels, reduced the exercise pressor reflex in rats with ligated femoral arteries. Moreover, the absolute reduction of the integrated pressor area during intermittent contraction caused by GsMTx4 was greater in ligated rats than it was in freely perfused rats. These findings indicate that the mechanically sensitive component of the exercise pressor reflex is exaggerated in rats with ligated femoral arteries. Our findings extend previous work from our laboratory that found that the metabolically sensitive component of the exercise pressor reflex was exaggerated in rats with ligated femoral arteries (18, 41, 45, 47). Moreover, our findings may have important implications for pathophysiological conditions such as heart failure (23, 24, 35) and hypertension (19) in which the mechanically sensitive component of the exercise pressor reflex contributes to the exaggerated pressor and sympathetic nerve responses to exercise.

GRANTS

This work was supported by National Institutes of Health Grants HL-096570 and AR-059397.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).

AUTHOR CONTRIBUTIONS

Author contributions: S.W.C. and M.P.K. conception and design of research; S.W.C., J.S.K., and V.R.-V. performed experiments; S.W.C. and V.R.-V. analyzed data; S.W.C., J.S.K., V.R.-V., and M.P.K. interpreted results of experiments; S.W.C. prepared figures; S.W.C. drafted manuscript; S.W.C., J.S.K., V.R.-V., and M.P.K. edited and revised manuscript; S.W.C., J.S.K., V.R.-V., and M.P.K. approved final version of manuscript.