Sensitizing effect of antimigraine drug sumatriptan on capsaicin-sensitive lung vagal neurons via 5-HT 1B/1D receptors and PKA: Relevance to adverse chest effects

Animals

Male Brown-Norway rats were purchased from the National Laboratory Animal Center (Taipei, Taiwan). Animals were handled in accordance with standards established by the Guide for the Care and Use of Laboratory Animals published by the National Institutes of Health, and approved by the IACUC (permits LAC-2018-0005 and LAC-2021-0538) of Taipei Medical University. Animals were housed two per cage under a 12:12 hour light/dark photocycle (lights on at 07.00 hours) with ad libitum water and food in the Laboratory Animal Center of Taipei Medical University. All efforts were made to minimize the number of animals used and their suffering. Anesthetics were administered before any surgical procedures, and supplemental doses were provided as needed to sustain the elimination of pain reflexes produced by pinching the rat's tail throughout the experiments. Rats were randomly divided into control and experimental groups. At the end of the experiment, the animals were euthanized using an intravenous injection of KCl during deep anesthesia. All experiments were carried out in accordance with ARRIVE (Animal Research: Reporting of In Vivo Experiments) guidelines (30).

In vivo study

Electrophysiological recording of CSLV afferents

The measurement of fiber activity (FA) of CSLV afferents was described in detail in our previous study (31). A schematic diagram of the experimental setup is shown in Figure 1(a). Briefly, the anesthetized rats were artificially ventilated with a respirator (model 683; Harvard Apparatus, Holliston, MA, USA); tidal volume and respiratory frequency were set at 8 ml/kg and 50 breaths/minute, respectively, to mimic those of anesthetized, spontaneously breathing rats. In our previous study, the CO₂ levels can be constantly maintained in a physiological condition under this ventilation setting (32). Ptr was measured by a differential pressure transducer (model P300D; Validyne, Northridge, CA, USA) via a side-port of the trachea cannula, and arterial blood pressure was measured by a pressure transducer (Statham P23XL; Spectramed, Columbus, OH, USA) via a femoral catheter. A midline thoracotomy was then performed to expose the thoracic cavity. Both vagus nerves were ligated just above the diaphragm to eliminate afferent signals originating from visceral organs below the diaphragm. To maintain a near-normal functional residual capacity, the expiratory outlet of the respirator was placed under 3 cmH₂O pressure. The neck was opened along the midline, and a segment (approximately 1 cm) of the right cervical vagus nerve was carefully separated from the adjacent common carotid artery. The nerve was then sectioned as far rostrally as possible, and the distal (caudal) end of the cut vagus nerve was gently placed on a small dissecting platform immersed in a pool of mineral oil. A fine filament was teased from the desheathed nerve trunk and positioned on a platinum–iridium hook electrode for recording. Action potentials were amplified by an AC/DC differential amplifier (model 3000; A-M Systems, Sequim, WA, USA), monitored by an audio monitor (AM8RS; Grass Technologies, West Warwick, RI, USA) and displayed on an oscilloscope (model 2211; Tektronix, Beaverton, OR, USA). The filament was progressively teased until a single-unit action potential was isolated. Signals were recorded and analyzed by a polygraph recorder (MP36; Biopac, Goleta, CA, USA). CSLV afferents were identified by 1) intense response to capsaicin (2 μg/kg) within two seconds after the right-atrial bolus injection via jugular catheter and 2) the locations of the fibers could be identified by their responses to pressing of the lungs with a cotton swab at the end of the experiment. FA responses were calculated as the difference between peak and baseline FA values (ΔFA). Peak responses to capsaicin, lung inflation, and adenosine were averaged over 3, 5 and 10 seconds, respectively, following stimulation.

OPEN IN VIEWER

Figure 1.

Sumatriptan (Suma) increased the excitability of capsaicin-sensitive lung vagal (CSLV) afferents to a capsaicin injection in rats. (a) Schematic diagram illustrating the experimental setup of single-fiber recording of CSLV afferents in an anesthetized, artificially ventilated rat. The activity of capsaicin-sensitive afferents from pulmonary branches of the vagus nerve was measured from a thin filament of the caudal end of the right sectioned cervical vagus nerve placed on a platinum–iridium electrode. (b) Experimental timeline illustrating CSLV-afferent responses to right-atrial capsaicin injections (arrows) obtained from 10 minutes before and 3–80 minutes after termination of the infusion of Suma or its vehicle. (c) Experimental records illustrating CSLV-afferent responses to capsaicin injections (arrows) before (control), 3 minutes and 80 minutes after termination of the Suma (1.2 mg/kg, left panel) or vehicle infusion (isotonic saline, right panel) in two anesthetized rats. Notice that the fiber response was potentiated 3 minutes after termination of the Suma infusion and recovered by 80 minutes later. (d) Group data of the increase in fiber activities (ΔFA values) elicited by capsaicin injections collected at three time points: before the infusion (control), at peak response times of 3 or 20 minutes (sensitization), and at 80 minutes (recovery) after infusion termination. The open bars and two sets of green bars represent infusion of the vehicle, and low-dose (0.6 mg/kg) and high-dose Suma (1.2 mg/kg), respectively. ΔFA is the difference between the peak FA (averaged over a 3-second interval) and the baseline FA for each CSLV afferent. Data are the mean ± SEM (n = 8). *p < 0.05 and **p < 0.01 using a mixed-effects model followed by Tukey's HSD post-hoc test. AP, action potential; Ptr, tracheal pressure; ABP, arterial blood pressure; imp, impulses.

In vitro study

Chemical agents

Statistical analysis

The minimal sample size was based on the following criteria: 1) the sample size required to achieve statistical significance in comparable experiments conducted in our laboratory (33,34,38) and 2) the minimal sample sizes in each group are required to achieve a power of 0.8 and alpha = 0.05 calculated using G*Power software (http://www.gpower.hhu.de) based on preliminary data. The experimental protocol for each study is illustrated and described where appropriate. To ensure objectivity, data analysis and calculation were carried out in a blinded manner by two individual investigators. Statistical analyses were performed using Prism, version 10 (GraphPad Software Inc., San Diego, CA, USA). Prior to statistical analysis, Shapiro–Wilk tests were conducted to evaluate whether the data followed a normal distribution. For datasets that met the assumption of normality, two-way analysis of variance (ANOVA) followed by Tukey's honestly significant difference (HSD) post-hoc test was applied. Non-normally distributed variables were analyzed using a Mann–Whitney U-test, a Kruskal–Wallis test followed by Dunn's post-hoc test or a mixed-effects model (restricted maximum likelihood, REML) followed by Tukey's HSD post-hoc test. p < 0.05 was considered statistically significant. All data are reported as the mean ± SEM. Statistical analyses, numbers of animals used (n) and p-values are reported where appropriate. Detailed statistical results are provided in the supplementary material (Table S1).

Sumatriptan enhanced the excitability of CSLV afferents to capsaicin in anesthetized, ventilated rats

To evaluate the potentiating effect of sumatriptan on capsaicin-evoked discharges (via activation of TRPV1), CSLV-afferent responses to a right-atrial bolus injection of capsaicin (0.5 μg/kg) were measured at control (10 minutes before infusion) and at 3, 20 and 80 minutes after the infusion of sumatriptan or its vehicle (Figure 1(b)). Under control conditions (before sumatriptan infusion), capsaicin injection elicited a mild discharge in CSLV afferents (Figure 1(c), upper left). This response was markedly enhanced 3 minutes after termination of the sumatriptan infusion (1.2 mg/kg; 0.1 ml/minute for 20 minutes) (Figure 1(c), middle left) and returned to baseline by 80 minutes post-infusion (Figure 1(c), lower left). In contrast, infusion of the sumatriptan vehicle (isotonic saline) did not induce such potentiation (Figure 1(c), right). Among CSLV afferents, peak capsaicin-induced activity typically occurred at 3 or 20 minutes after sumatriptan infusion. Sumatriptan potentiated these responses in a dose-dependent manner; relative to pre-infusion levels, the higher dose (1.2 mg/kg) increased the response by approximately 3.1-fold, whereas the lower dose (0.6 mg/kg) produced an approximately 1.4-fold increase (Figure 1(d)). Notably, Ptr and FA remained unchanged throughout the experimental period in both sumatriptan- and vehicle-treated groups (Table 1). These findings demonstrate that sumatriptan enhances the excitability of CSLV afferents to capsaicin without altering baseline Ptr and FA.

OPEN IN VIEWER

Table 1.

Effect of an infusion of sumatriptan or its vehicle on the baseline tracheal pressure (Ptr) and fiber activity (FA) in anesthetized, artificially ventilated rats.

	Vehicle			Sumatriptan
	Before infusion	3 minutes	20 minutes after infusion	Before infusion	3 minutes	20 minutes after infusion
		after infusion			after infusion
Baseline Ptr (cmH2O)	8.23 ± 0.13	8.36 ± 0.05	8.34 ± 0.14	8.49 ± 0.16	8.45 ± 0.19	8.62 ± 0.14
Baseline FA (imp/second)	0.00 ± 0.00	0.00 ± 0.00	0.00 ± 0.00	0.03 ± 0.02	0.00 ± 0.00	0.03 ± 0.03

Vehicle or sumatriptan (1.2 mg/kg) were slowly infused into the femoral vein at a constant rate of 0.1 ml/min for 20 minutes. Baseline values were averaged over a 10-second period immediately before, and 3 minutes and 20 minutes after termination of the infusion. There was no significant difference in either the baseline Ptr or FA between before and after infusion of the vehicle or sumatriptan. Data are the mean ± SEM (n = 8). Statistical analysis was performed using a mixed-effects model followed by Tukey's HSD post-hoc test. imp, impulses.

Sumatriptan-induced sensitization of CSLV afferents was not restricted to capsaicin stimulation in anesthetized, ventilated rats

To determine whether the excitability increase induced by sumatriptan was specific to capsaicin stimulation, we evaluated CSLV afferent responses to two additional stimuli: adenosine injection (0.6 mg/kg), a chemical stimulant of CSLV fibers via adenosine A1 receptors (39), and constant-pressure lung inflation (Ptr = 30 cmH₂O for 10 seconds), a mechanical stimulus. Under control conditions, both adenosine injection and lung inflation elicited only mild activation of CSLV afferents. In contrast, responses to both stimuli were significantly potentiated following sumatriptan infusion (Figure 2(a) and (b), left), whereas no such potentiation was observed in vehicle-treated animals (Figure 2(a) and (b), right). Group data confirmed that sumatriptan increased the adenosine-induced (Figure 2(c)) and lung inflation-induced (Figure 2(d)) changes in FA, suggesting that sumatriptan induces a non-selective sensitizing effect on CSLV afferents.

OPEN IN VIEWER

Figure 2.

Sumatriptan (Suma) increased the excitability of rat capsaicin-sensitive lung vagal (CSLV) afferents to an adenosine injection and to lung inflation. (a, b) Experimental records illustrating CSLV-afferent responses to right-atrial adenosine injections (arrows, a) and lung inflation (b) before and after termination of a Suma (1.2 mg/kg) or vehicle infusion in four anesthetized, artificially ventilated breathing rats. (c, d) Group data of fiber activities (ΔFA values) elicited by adenosine injections (c) or by lung inflation (d) in rats treated with Suma or its vehicle. For further explanation of the ΔFA calculation, see Figure 1. Data are the mean ± SEM (n = 8). *p < 0.05 using a mixed-effects model followed by Tukey's HSD post-hoc tests. AP, action potential; Ptr, tracheal pressure; ABP, arterial blood pressure.

Sensitizing effect of sumatriptan was observed in isolated rat CSLV neurons

Because sumatriptan was administrated by systemic infusion route in the in vivo study, whether the sensitization resulted from a direct action on CSLV neurons or an indirect effect via non-CSLV pathways remained unclear. To clarify this, we examined the effects of sumatriptan on isolated CSLV neurons. Neuronal responses were assessed by monitoring changes in intracellular calcium concentration ([Ca²⁺]_i), using the F340/F380 ratio. Under control conditions (before sumatriptan perfusion), capsaicin (0.1 μM for 30 seconds) elicited a mild, transient increase in [Ca²⁺]_i. However, 10 minutes after perfusion of high-concentration sumatriptan (1 nM for 60 seconds, n = 23), reapplication of capsaicin evoked an enhanced Ca²⁺ transient (Figure 3(c) and (d)). This enhancement was not observed in neurons perfused with the vehicle (ECS, n = 21) or with low-concentration sumatriptan (0.5 nM, n = 26) (Figure 3(d)). These findings indicate that sumatriptan can directly sensitize CSLV neurons.

OPEN IN VIEWER

Figure 3.

Sumatriptan (Suma) enhanced capsaicin (Cap)-evoked Ca²⁺ transients in isolated capsaicin-sensitive lung vagal (CSLV) neurons. (a–c) Experimental records illustrating Ca²⁺ transients evoked by Cap perfusion (arrows) before and 10 minutes after termination of perfusion. Treatment with the vehicle (Veh, extracellular solution) (a), low-concentration Suma (0.5 nM) (b) or high-concentration Suma (1.0 nM) (c) is indicated by white, and light- and dark-green bars, respectively. Cell viability was tested with a KCl solution (arrows, 60 mM). Notice that only high-concentration Suma treatment induced a significantly enhanced response. (d) Concentration effects on Suma-enhanced Ca²⁺ transients. The sensitizing effect was determined by the ratio of Ca²⁺ transient evoked by Cap after perfusion (2^nd Cap) to those before perfusion (1^st Cap). The horizontal dashed line depicts a ratio of 1 (indicating no sensitizing effect). Respective n values for the vehicle, and low- and high-Suma groups were 21 (from two rats), 26 (from two rats), and 23 (from two rats). Data are the mean ± SEM. *p < 0.05 and ***p < 0.005 using a Kruskal–Wallis test followed by Dunn's post-hoc test.

5-HT_1B and 5-HT_1D receptors were expressed in isolated CSLV neurons

Immunofluorescence staining was performed in cultured vagal neurons isolated from rat nodose and jugular (N/J) ganglia. Bronchopulmonary neurons were identified by retrograde labeling with DiI, which marked 25.7% of the total N/J neuronal population. Among the DiI-labeled neurons, 76.3% were immunopositive for TRPV1, indicating that these cells represent CSLV neurons. Notably, both 5-HT_1B and 5-HT_1D receptors were expressed in TRPV1-positive CSLV neurons (Figure 4). Specifically, 84.3% and 88.9% of these neurons expressed 5-HT_1B and 5-HT_1D receptors, respectively. The selectivity of the antibodies was confirmed through two standard control conditions in which either the primary or secondary antibody was omitted (see supplementary material, Figure S1). The expression of 5-HT_1B and 5-HT_1D receptors in CSLV neurons provides a rationale for further pharmacological investigation into their functional roles.

OPEN IN VIEWER

Figure 4.

5-Hydroxytryptamine 1B and 1D (5-HT_1B/1D) receptors expressed in rat capsaicin-sensitive lung vagal (CSLV) neurons. Double-labeling immunofluorescence staining was performed on CSLV neurons following intratracheal injection of 1,1′-dioctadecyl-33,3′,3′-tetrameth-ylindocarbocyanine perchlorate (DiI) (a retrograde tracer), showing (a) DiI (lung labeling, red)/transient receptor potential vanilloid 1 (TRPV1) (a marker of capsaicin-sensitive neuron, blue)/5-HT_1B receptors (green). (b) As in (a), green staining targeted 5-HT_1D receptors instead of 5-HT_1B receptors. Arrows indicate neurons with triple labeling. Scale bar = 150 μm.

Sumatriptan-induced sensitization of CSLV neurons required 5-HT_1B , 5-HT_1D , and PKA activation, but was independent of transient receptor potential ankyrin 1 (TRPA1)

To determine whether 5-HT_1B and 5-HT_1D receptors are critical for sumatriptan-induced sensitization, selective antagonists targeting the 5-HT_1B receptor (SB224289), the 5-HT_1D receptor (BRL15572), or both 5-HT_1B/1D receptors (GR127935) were administered 15 minutes before sumatriptan perfusion (1 nM for 60 seconds). Pretreatment with any of these antagonists (1 nM each) did not affect baseline F340/F380 (see supplementary material, Table S2) or the initial capsaicin-evoked Ca²⁺ transients (see supplementary material,, Table S3), but completely abolished the sumatriptan-induced enhancement of these responses (Figure 5(b) to (d)). In contrast, vehicle pretreatment (ECS) did not block the enhancing effect (Figure 5(a)), and group analysis confirmed that the sensitization was fully suppressed by receptor antagonism but remained intact in the vehicle group (Figure 5(e)).

OPEN IN VIEWER

Figure 5.

Pretreatment with an antagonist of 5-hydroxytryptamine 1B and 1D (5-HT_1B/1D) receptors abolished the enhancing effect of sumatriptan (Suma) on Ca²⁺ transients evoked by capsaicin (Cap) in isolated capsaicin-sensitive lung vagal (CSLV) neurons. (a-d) Experimental records illustrating the effect of a Suma perfusion (1 nM, black bar) on Ca²⁺ transients evoked by Cap (arrows) after pretreatment with the vehicle (Veh) (a), SB224289 (SB, a 5-HT_1B receptor antagonist) (b), BRL15572 (BRL, a 5-HT_1D receptor antagonist) (c) or GR127935 (GR, a 5-HT_1B/1D receptors antagonist) (d). (e) Group data showing enhancement of Ca²⁺ transients after Suma was pretreated with Veh (n = 22 from two rats), SB (n = 26 from three rats), BRL (n = 20 from two rats) or GR (n = 25 from two rats). Data are the mean ± SEM. ***p < 0.005 using a Kruskal–Wallis test followed by Dunn's post-hoc test. (f, g) Experimental records illustrating the effect of a Suma perfusion (1 nM) on Ca²⁺ transients evoked by Cap after pretreatment with the Veh (f) or HC-030031 (HC, a TRPA1 antagonist) (g). (h) Group data showing enhancement of Ca²⁺ transients after Suma was pretreated with the Veh (n = 25 from two rats) or HC (n = 24 from two rats). Data are the mean ± SEM. Statistical analysis was performed using a two tailed unpaired Mann–Whitney test. For further explanations of the horizontal dashed line and calculation of the sensitized ratio, see Figure 3.

Given that sumatriptan has been shown to activate TRPA1 in HEK293 cells (40), we next investigated whether TRPA1 contributes to this sensitization. However, pretreatment with an TRPA1 antagonist (HC-030031; 10 μM) or its vehicle (0.01% DMSO) did not alter the sumatriptan-induced enhancement of capsaicin-evoked Ca²⁺ transients (Figure 5(f) to (h)). The efficacy and selectivity of HC-030031 were validated by use of cinnamaldehyde (a TRPA1 agonist, 0.3 mM for 40 seconds) and capsaicin (0.1 μM for 30 seconds): HC-030031 suppressed the cinnamaldehyde-evoked Ca²⁺ transients (ΔF340/F380: 0.51 ± 0.12 vs. 0.04 ± 0.02; p < 0.005 using two tailed unpaired Mann–Whitney test), but had no effect on capsaicin responses (ΔF340/F380: 0.60 ± 0.14 vs. 0.69 ± 0.13; p = 0.2406 using two tailed unpaired Mann–Whitney test) (see supplementary material, Figure S2). These findings indicate that TRPA1 activation is not involved in the sensitizing effect of sumatriptan on CSLV neurons.

Finally, to assess whether PKA activation is essential for the sensitizing effect of sumatriptan, we pretreated cells with H89, a selective PKA inhibitor, before sumatriptan perfusion. Pretreatment with H89 (10 μM), but not its vehicle (0.02% DMSO), completely abolished the sumatriptan-induced enhancement of the capsaicin-evoked Ca²⁺ transient (Figure 6), indicating that PKA activation is required for this sensitizing effect.

OPEN IN VIEWER

Figure 6.

Pretreatment with a protein kinase A (PKA) inhibitor abolished the enhancing effect of sumatriptan (Suma) on Ca²⁺ transients evoked by capsaicin (Cap) in isolated capsaicin-sensitive lung vagal (CSLV) neurons. (a, b) Experimental records illustrating the effect of Suma perfusion (1 nM, black bar) on Ca²⁺ transients evoked by Cap (arrows) after pretreatment with vehicle (Veh) (a) or H89 (a PKA inhibitor) (b). (c) Group data showing enhancement of Ca²⁺ transients after Suma pretreated with the Veh (n = 22 from two rats) or H89 (n = 20 from two rats). For further explanations of the horizontal dashed line and calculation of the sensitized ratio, see Figure 3. Data are the mean ± SEM. ***p < 0.005 using a two tailed unpaired Mann–Whitney test.

Sumatriptan enhanced CSLV afferent–mediated airway reflexes in anesthetized, spontaneously breathing rats

Given that sumatriptan increases the excitability of CSLV afferents, we next investigated whether this peripheral sensitization also potentiates central processing, leading to enhanced airway reflexes. To address this, we returned to the in vivo model to assess CSLV afferent–mediated airway reflexes after sumatriptan treatment. Under control conditions (before sumatriptan infusion), a right-atrial bolus injection of capsaicin (1.0 μg/kg) elicited a mild, transient apneic response, comprising a hallmark reflex mediated by activation of CSLV afferents (Figure 7(a), left). Following sumatriptan infusion (1.2 mg/kg; 0.1 ml/minute for 20 minutes), this apneic response was potentiated 3 minutes post-infusion (Figure 7(a), middle) and returned to baseline by 80 minutes later (Figure 7(a), right), paralleling the time course of afferent sensitization observed in Figure 1 and Figure 2. In contrast, no potentiation of the apneic response was observed in animals receiving vehicle infusion (isotonic saline) (Figure 7(b)).

OPEN IN VIEWER

Figure 7.

Infusion of sumatriptan (Suma) potentiated apneic responses to a capsaicin injection in rats. (a, b) Experimental records illustrating apneic responses to right-atrial capsaicin injections (arrows) before and after termination of a Suma (1.2 mg/kg) (a) or vehicle (Veh) (b) infusion in two anesthetized, spontaneously breathing rats. Notice that the apneic response was potentiated 3 minutes after termination of the Suma infusion and recovered by 80 minutes later. (c) Temporal response (left) and peak response (right) showing the dose effect on Suma-potentiated apneic responses to capsaicin. In the histogram, apneic responses were collected before, 3–120 minutes after termination of the infusion (gray vertical bar) of the Veh, or low-dose (0.6 mg/kg) or high-dose (1.2 mg/kg) Suma. In the bar diagram, apneic responses are compared between three time points: before infusion (control), at peak response times of 3 or 20 minutes (sensitization), and at 80 minutes (recovery) after infusion termination. Data are the mean ± SEM (n = 6). The horizontal dashed line depicts an apneic ratio of 1 (indicating no apnea). *p < 0.05 and **p < 0.01 using a mixed-effects model followed by Tukey's HSD post-hoc tests. (d) Role of 5-hydroxytryptamine 1B and 1D (5-HT_1B/1D) receptors in Suma (1.2 mg/kg)-induced potentiation of the apneic response to capsaicin. The arrow on the left represents pretreatment with GR (300 μg/kg) or its Veh (0.75% DMSO). Collections of apneic responses for comparison were the same as those in (c). The horizontal dashed line depicts an apneic ratio of 1 (indicating no apnea). Data are the mean ± SEM (n = 6). *p < 0.05 using a two-way ANOVA followed by Tukey's HSD post-hoc test. V ˙ airflow rate; V_T, tidal volume; ABP, arterial blood pressure.

As summarized in Figure 7(c), apneic responses were measured before and after infusion of either vehicle or sumatriptan at low (0.6 mg/kg) or high (1.2 mg/kg) doses. Only the high-dose sumatriptan group showed significant potentiation at 3 and 20 minutes post-infusion, with responses returning to baseline by 80 minutes. Importantly, baseline cardiovascular parameters recorded at the corresponding time points did not change across groups (Table 2), indicating that potentiation of the airway reflex was independent of hemodynamic changes. These findings suggest that sumatriptan augments CSLV afferent-mediated signaling to the central nervous system, thereby potentiating airway reflex responses in a dose-dependent and reversible manner.

OPEN IN VIEWER

Table 2.

Effect of infusion of sumatriptan or its vehicle on baseline heart rate (HR) and mean arterial blood pressure (MABP) in anesthetized, spontaneously breathing rats.

	Vehicle			Sumatriptan
	Before infusion	3 minutes	20 minutes	Before infusion	3 minutes	20 minutes
		after infusion	after infusion		after infusion	after infusion
Baseline HR (beats/minute)	386 ± 8	388 ± 11	388 ± 9	368 ± 9	350 ± 13	364 ± 8
Baseline MABP (mmHg)	102 ± 6	103 ± 6	105 ± 8	112 ± 3	102 ± 4	107 ± 3

The vehicle or sumatriptan (1.2 mg/kg) was slowly infused into the femoral vein. There was no significant difference in either baseline HR or MABP between before and after infusion of the vehicle or sumatriptan. Data are the mean ± SEM (n = 6). Statistical analysis was performed using a two-way analysis of variance followed by Tukey's HSD post-hoc test.

Sumatriptan potentiated CSLV afferent–mediated airway reflexes through 5-HT_1B/_1D receptors in anesthetized rats

To determine whether the sumatriptan-induced potentiation of CSLV-mediated airway reflexes depends on activation of 5-HT_1B/1D receptors, GR127935 (300 μg/kg) or its vehicle (0.75% DMSO) was administered 10 minutes prior to sumatriptan infusion (1.2 mg/kg) in anesthetized, spontaneously breathing rats. GR127935, but not the vehicle, markedly suppressed the sumatriptan-induced potentiation of apneic responses to capsaicin (Figure 7(d)). To determine whether GR127935 alone affected the apneic response to capsaicin, capsaicin was injected before and after sumatriptan vehicle infusion in rats pretreated with GR127935. No significant difference was observed in the apneic responses (2.61 ± 0.37 vs. 2.40 ± 0.31; n = 6; p = 0.24 using two-tailed paired t-test), indicating that GR127935 did not attenuate the capsaicin-evoked reflex per se but specifically inhibited the potentiating effect of sumatriptan.

Consistent with this, electrophysiological recordings in anesthetized, artificially ventilated rats further demonstrated that GR127935 pretreatment abolished the sumatriptan-induced potentiation of CSLV fiber responses to both capsaicin injection (see supplementary material, Figure 3(a)) and lung inflation (see supplementary material, Figure 3(b)). These findings demonstrate that the potentiation of both reflex and afferent responses induced by sumatriptan requires activation of 5-HT_1B and 5-HT_1D receptors.

A limited role for myocardial ischemia in sumatriptan-induced chest discomfort

Sumatriptan-induced chest discomfort was reported shortly after its marketing began in 1991 and was systematically documented in The Lancet in 1993 (10). Initially, researchers speculated that the mechanism was of cardiac origin since the chest discomfort resembled angina symptoms and sumatriptan had vasopressor properties. Indeed, a vasoconstrictive effect was observed in ex-vivo experiments which showed that sumatriptan caused a significant contraction in isolated human coronary arteries (41,42). However, the maximum concentration/median effective concentration (C_max/EC₅₀) ratio of sumatriptan was only about 0.4, which meant that the maximum plasma concentration in humans after an effective dose of sumatriptan reached just 40% of the level needed for a half-maximal effect in isolated coronary arteries, suggesting that sumatriptan at a therapeutic plasma concentration could not cause significant coronary vasoconstriction (41). Numerous human studies supported this finding, such as that by angiographic examinations, where only 12–16% decreases in coronary artery diameter occurred after intravenous or subcutaneous sumatriptan in migraineurs without cardiovascular diseases (14,15). Notably, even when these patients experienced chest-related adverse effects, sumatriptan did not alter myocardial perfusion (11). Throughout those human studies, sumatriptan did not affect the heart rate or electrocardiographic readings (11,14,15). Current investigations indicated that therapeutic doses are unlikely to cause myocardial ischemia in healthy migraineurs.

Involvement of CSLV afferents in sumatriptan-induced chest discomfort

For many decades, CSLV afferents were postulated to be the source of chest unpleasant sensations (e.g. dyspnea, chest tightness, breathlessness, etc.) (23). Studies in humans showed that projections of CSLV afferent signals to the central nervous system play a vital role in triggering chest discomfort symptoms, such as shortness of breath, chest tightness and breathlessness (23,24). In healthy human subjects, dyspneic sensations can be induced by inhalations or injections of various CSLV-afferent stimulants (e.g. adenosine (28,29), histamine (43) and nicotine (44)) or sensitizers (e.g. prostaglandin E2 (45)). Furthermore, when intravenous adenosine (known to stimulate CSLV afferents (27,39)) is used to treat tachycardia, the dyspneic adverse reaction can be reduced by inhaling lidocaine, a local anesthetic. This indicates a primary role of lung sensory nerves in the sensation of dyspnea (28,29). In addition, studies showed that patients with certain lung diseases experience significant relief from chest discomfort after the vagal nerve pathway is blocked, either through local anesthetic injections or surgical vagal nerve cutting (a vagotomy) (46–49). Our experimental results showed that intravenous infusion of sumatriptan sensitized rat CSLV afferents, thus enhancing airway reflexes through projections of sensory signals to the central nervous system. Furthermore, we observed that the timing of sumatriptan-induced airway reflex enhancement in rats (3–20 minutes) matched the period in which patients reported chest discomfort after taking sumatriptan, implying that enhancement of signal inputs of CSLV afferents may contribute to the development of chest discomfort after sumatriptan treatment in migraineurs.

Determination of the dose and administrated route of sumatriptan in this study

Although sumatriptan is mostly administered orally (50), we gave it by systemic infusion route in our study for two main reasons. First, oral administration exhibits higher variability in absorption and lower bioavailability patterns than systemic infusion routes among experimental subjects (51). Second, oral users experience chest discomfort less frequently than subcutaneous and intravenous users (6,7), which is highly associated with bioavailability patterns. Based upon these observations, oral administration may have constrained our capacity to observe and analyze the sensitizing effect of sumatriptan on CSLV afferents. In humans, intravenous or subcutaneous sumatriptan doses range 0.05–0.2 mg/kg (52,53). Applying the human-to-rat conversion factor (K_m ≈ 6.2) (54), the translation from a human dose (0.2 mg/kg) to a rat dose is approximately 1.2 mg/kg. In addition, we administered sumatriptan through a slow infusion over 20 minutes (2 ml administered over 20 minutes) to prevent a rapid surge in the plasma concentration, simulating the time required to reach peak concentration following subcutaneous administration. Overall, the pharmacokinetic profile of sumatriptan in our rat model may have closely resembled that observed with clinical use in humans.

Sumatriptan-induced sensitization of CSLV neurons: Role of 5-HT_1B/1D receptors

This study revealed that pretreatment with either an antagonist of 5-HT_1B receptor or 5-HT_1D receptor fully abolished sumatriptan-induced sensitization of CSLV neurons, suggesting that both receptors play essential roles in this process. However, if these receptors act through independent pathways, partial residual sensitization would typically be expected when only one is inhibited. A similar all-or-none blocking pattern was observed in other models, such as in sumatriptan-inhibited hyperalgesia induced by formalin in rats, where antagonism of either receptor alone eliminated the effect (55). Although the underlying mechanism remains to be fully elucidated, it seems to be unlike the involvement of insufficient antagonist selectivity in our study. The high-selectivity antagonists we used (BRL15572 and SB224289) were reported in other studies examining the effects of sumatriptan on the sensitivity of sensory neuron (56,57) and the contractility of vascular smooth muscle (58,59). Moreover, concentrations of antagonists we used were lower than those reported in the referenced studies, minimizing potential non-selective binding effects. Given prior reports of functional interaction between 5-HT_1B and 5-HT_1D receptors (60), we speculate that receptor crosstalk, such as heterodimerization or shared intracellular signaling, may allow inhibition of one receptor to impair the activity of the other. Confirming this hypothesis requires further experimental investigations.

Sumatriptan-induced sensitization of CSLV neurons: The role of PKA

Pretreatment with GR127935 or H89 effectively prevented sumatriptan-induced sensitization of CSLV neurons, demonstrating activation of 5-HT_1B/1D receptors and PKA plays a key role. 5-HT_1B/1D receptors known to be coupled with G_i/G_o proteins (61) suggests that activation of 5-HT_1B/1D receptors triggers G_i/G_o protein-mediated signaling, further leading to elevation of PKA levels. This postulation was supported by a study which demonstrated that hyperalgesia elicited by an intradermal injection of sumatriptan in rat hind paws was reversed by inhibiting activation of either 5-HT_1B/1D receptors, G_i/G_o protein, or PKA (26). However, this pathway of PKA activation mediated through the G_i/G_o protein contrasts with the classical signaling pathway. Typically, G_i/G_o activation decreases PKA levels through inhibition of cAMP activity (61). Consistent with this classical pathway, sumatriptan administration was found to suppress forskolin-induced cAMP activation in isolated dog saphenous veins (62) and rabbit carotid arteries (63). Interestingly, some studies using other species of cells showed inconsistent results with the above findings. For example, sumatriptan did not alter intracellular cAMP levels in rat trigeminal sensory neurons (64,65), astrocyte cells (66), or anococcygeus muscle (67). We assumed that sumatriptan-mediated PKA activation may be independent of cAMP and only exhibited in specific species and cell types. However, further studies are required to investigate this mechanism.

Dual effects of sumatriptan on hyperalgesia and analgesia

Sumatriptan demonstrates efficacy in migraine treatment; nevertheless, our findings revealed its capacity to increase the sensitivity of CSLV neurons, presenting an apparent paradox in the modulation of pain perception. This dual effect can be attributed to distinct mechanisms of sumatriptan's action on sensory neurons. The therapeutic benefits primarily arise from its ability to suppress the releases of inflammatory neuropeptides, particularly calcitonin gene-related peptides, from sensitized trigeminal nerve terminals to the dura mater (64,68). These neuropeptides typically trigger a cascade of pro-inflammatory mediator release, further resulting in the self-sustaining inflammatory cycle and hypersensitivity of sensory nerves (69,70). Sumatriptan effectively disrupts this vicious cycle by inhibiting neuropeptide synthesis and release (61,64). After excessive inflammatory mediators in the dura matter are degraded, neural sensitivity can return to the baseline. However, substantial studies revealed that before the onset of analgesic effects, sumatriptan increases the sensitivity of peripheral sensory nerves (5,26,71,72). For example, transient hyperalgesia, a common adverse effect of sumatriptan, results from sensitization of somatic nociceptors (26); this comparable effect of transient sensitization was documented in both cultured rat sensory trigeminal neurons (71) and CSLV neurons examined in our investigation. Taken together, these findings indicate that sumatriptan administration elicits a biphasic response in sensory neurons, characterized by an initial sensitization phase followed by therapeutic analgesic effects.

Limitations of this study

A major limitation of this study is the exclusive use of male rats. Clinically, both the prevalence of migraine (73) and the incidence of chest symptoms after sumatriptan use (6,7) are higher in women. In our preliminary experiments using female rats, sumatriptan did not enhance airway sensitivity, with responses showing considerable individual variability. We speculate that this sex difference may be due to fluctuations in estrogen levels during the estrous cycle. Estrogen modulates airway sensitivity in a concentration-dependent, biphasic manner. Huang et al. (74) showed that intermittent hypoxia-induced airway hypersensitivity was markedly potentiated in ovariectomized female rats, and attenuated by low-dose estrogen replacement. In contrast, replacement of high-dose estrogen, which raised plasma estrogen to levels comparable to the proestrus phase, re-enhanced airway sensitivity. In the same study, a similar biphasic effect of estrogen on the excitability of CSLV afferents was also observed (74). Therefore, a comprehensive assessment of the sensitizing effect of sumatriptan in female rats would require the use of ovariectomized rats with controlled estrogen replacement to mimic hormonal cycling. However, conducting such investigations would require a significantly larger number of animals, which is beyond the scope of the present study. Despite these limitations, whether sumatriptan alters the sensitivity of CSLV afferents in female rats remains an important and unanswered question that requires further investigation.

Clinical implications of this study

In this study, we have proposed a novel perspective based on sensory neural mechanisms to elucidate the longstanding phenomenon of sumatriptan-induced chest discomfort. However, because dyspnea is a subjective sensation that can only be verbally expressed and evaluated in conscious humans, further clinical investigations are required to validate the translational relevance of our findings from rodents to humans. Nevertheless, our findings indicated that sumatriptan sensitized CSLV afferents, vagal afferents primarily responsible for dyspnea, through its direct action on nerve endings. Furthermore, 5-HT_1B/1D receptor-mediated signaling is critical for sensitizing CSLV afferents. We suggest that inhibiting activation of these receptors in the lungs could alleviate chest discomfort caused by sumatriptan. Given that the lungs can receive targeted drug delivery, airway exposure to an antagonist of 5-HT_1B/1D receptors might serve as a therapeutic strategy. Localized pulmonary administration minimally interferes with oral sumatriptan for headache relief and maintains a therapeutic effect on chest side effects.