5-HT4(a) Receptors Avert Opioid-Induced Breathing Depression Without Loss of Analgesia

See allHide authors and affiliations

Science  11 Jul 2003:
Vol. 301, Issue 5630, pp. 226-229
DOI: 10.1126/science.1084674


Opiates are widely used analgesics in anesthesiology, but they have serious adverse effects such as depression of breathing. This is caused by direct inhibition of rhythm-generating respiratory neurons in the Pre-Boetzinger complex (PBC) of the brainstem. We report that serotonin 4(a) [5-HT4(a)] receptors are strongly expressed in respiratory PBCneurons and that their selective activation protects spontaneous respiratory activity. Treatment of rats with a 5-HT4 receptor–specific agonist overcame fentanyl-induced respiratory depression and reestablished stable respiratory rhythm without loss of fentanyl's analgesic effect. These findings imply the prospect of a fine-tuned recovery from opioid-induced respiratory depression, through adjustment of intracellular adenosine 3′,5′-monophosphate levels through the convergent signaling pathways in neurons.

Serotonin (5-hydroxytryptamine, or 5-HT) is an important neurotransmitter that is involved in a wide range of neuromodulatory processes in the central nervous system, by acting on a number of different 5-HT receptor isoforms (1, 2). The 5-HT4 receptor is a recently identified subtype that is widely and abundantly expressed as alternatively spliced variants in various brain regions (35). The receptor exerts excitatory effects through its positive coupling to heterotrimeric Gs proteins to activate adenylyl cyclases and to induce robust increases of intracellular adenosine 3′,5′-monophosphate (cAMP) levels (6, 7). The receptor also couples to G13 proteins to activate small guanosine triphosphates of the Rho family (8) (Fig. 1A).

Fig. 1.

(A) Schematic illustration of the signal transduction pathways mediated by 5-HT4 and μ-opioid receptors. Whereas 5-HT4 receptors stimulate adenylyl cyclases (ACs) throughbothGs and G13 proteins, μ-opioid receptors inhibit AC activities through a Gi/o-mediated inhibitory pathway. μOR, μ-opioid receptor; Gαi, Gαs, and Gα13, the αi, αs, and α13 subunits of heterotrimeric G protein; 5-HT 4R, 5-HT4 receptor; AKAP, A-kinase anchor protein; PKA, protein kinase A. (B) Western blot analysis of the brainstem lysate witha polyclonal antibody raised against the C-terminal part of the 5-HT4(a) receptor isoform. Left: Model of the 5-HT4(a) receptor withthe amino acids CHSGHHQELEKLPIHNDP (red) (27) used for the production of an antibody (ab). Right: Membrane (m) and cytosolic (c) fractions prepared from the lysate of the rat brainstem were separated by SDS–polyacrylamide gel electrophoresis and then subjected to Western blotting withan antibody to the 5-HT4(a) receptor.

Recent cloning of the 5-HT4 receptor (9) initiated the development of 5-HT receptor subtype–specific immunocytochemistry and pharmacology. We produced a specific antibody against a synthetic peptide that corresponds to the C-terminal sequence of the 5-HT4(a) receptor's isoform (amino acids His364 to Pro380) (10) (Fig. 1B), which allowed us to specifically identify the spatial expression of the 5-HT4 receptors in the central nervous system, including the brainstem (11). We found that 5-HT4 receptors are abundantly expressed in the Pre-Boetzinger complex (PBC), a region in the lower brainstem that is known to generate and control spontaneous breathing movements (12).

This specific antibody was also used to analyze the coexpression of the 5-HT4(a) receptor with μ-opioid receptors and Substance P–reactive neurokinin-1 (NK-1) receptors, which have been suggested as potential immunocytochemical markers for respiratory neurons (1315). Medullary motoneurons were visualized by choline acetyl transferase (ChAT) staining and excluded from the analysis (16). In multiple-labeling experiments (with NK-1 receptors, 5-HT4(a) receptors, and ChAT), we identified three different types of immunoreactive medullary interneurons: 35.5% of immunoreactive interneurons displayed intense NK-1 and 5-HT4(a) receptor co-immunoreactivities and 34.1% of interneurons revealed 5-HT4(a) receptor immunoreactivity alone, whereas 30.4% of interneurons revealed only NK-1 receptor immunoreactivity (Fig. 2, A and B) (supporting online text). In a similar study, the brainstem was analyzed for co-immunoreactivities of μ-opioid and 5-HT4(a) receptors within the PBC region, a region essential for respiratory rhythm generation (12). We found positive staining in 46.8% of interneurons for both μ-opioid and 5-HT4(a) receptors, whereas a population of 53.2% of interneurons exhibited only μ-opioid receptor immunoreactivity (Fig. 2C) (supporting online text). These data suggest that approximately one-half of all 5-HT4(a) receptor–positive interneurons in the PBC region coexpress both NK-1 receptors and μ-opioid receptors. We confirmed this by multiple staining of 5-HT4(a), NK-1, and μ-opioid receptors in the same slice (fig. S1) as well as by reverse transcription–polymerase chain reaction (RT-PCR) analysis.

Fig. 2.

Distribution of 5-HT4(a), NK-1, and μ-opioid receptor and ChAT immunoreactivities within the ventrolateral region of the brainstem that contains the PBC. (A) Triple-labeling withappropriate antibodies of a 40-μm-thick transversal slice cut at the level of the PBC, showing 5-HT4(a) receptors (Alexa 488, green), NK-1 receptors (Alexa 546, red), and ChAT (Alexa 647, blue). NA, Nucleus ambiguus. Scale bar, 50 μm. (B) A single neuron from the PBC region, stained for 5-HT4(a) (green) and NK-1 (red) receptors. 5-HT4(a) receptors are predominantly expressed on the soma and proximal dendrites. Scale bar, 10 μm. (C) Triple-labeling of 5-HT4(a) receptors (Alexa 488, green), μ-opioid receptors (Alexa 546, red), and ChAT (Alexa 647, blue) at the level of the PBC. Scale bar, 50 μm. All images were obtained by confocal laser scanning microscopy.

To verify that the 5-HT4(a) receptor–immunoreactive interneurons indeed represent respiratory neurons, single-cell RT-PCR analysis was performed on the cytosol of identified inspiratory neurons in the rhythmically active slice preparation (Fig. 3) (17). We found that 95.2% of the neurons analyzed expressed 5-HT4 receptors (Fig. 3B) (supporting online text). In view of the large number of 5-HT4 receptor isoforms, it was of particular interest to analyze the expression of the different 5-HT4 receptor splice variants. RT-PCR analysis of the PBC region and of individual respiratory neurons proved the expression of 5-HT4(a), 5-HT4(b), and 5-HT4(f) receptor mRNA, but not of 5-HT4(e) receptor mRNA, in the PBC region (Fig. 3B) (16), whereas defined inspiratory neurons expressed only the 5-HT4(a) mRNA (Fig. 3B). Immunocytochemical studies confirmed these observations, demonstrating a strong 5-HT4(a) receptor immunoreactivity in biocytin-labeled inspiratory neurons (Fig. 3C).

Fig. 3.

Expression of 5-HT4 receptors in functionally identified inspiratory neurons. (A) Top: The site of electrophysiological recordings (left) with an inspiratory neuron on the tip of the patch pipette (right). Traces: Integrated hypoglossal nerve (XIIa) activity (middle) corresponds to rhythmic inward currents in a single inspiratory neuron recorded in the whole-cell configuration (bottom). Im, membrane current. (B) Top: Single-cell RT-PCR analysis of inspiratory neurons. Gel electrophoresis was carried out for RT-PCR products amplified with5-HT4 primers. The control reaction without reverse transcription is shown in the first line. Ins, inspiratory neuron; bp, base pairs. Bottom: RT-PCR analysis of 5-HT4 receptor splice variants in the PBC region (left) and in an individual inspiratory neuron (right). Lane 1 shows primers amplifying the (a), (e), and (f) isoforms. Lane 2 shows primers amplifying the (b) isoform. All RT-PCR products were evaluated by direct DNA sequencing. (C) Example of an inspiratory PBC neuron labeled intracellularly withbiocytin (arrows) and exhibiting strong 5-HT4(a) receptor immunoreactivity (Alexa 546, red). The neuron is surrounded by 5-HT4(a) receptor immunoreactivities on somatic profiles (arrowheads) within the PBC. Scale bar, 50 μm.

We also determined, by single-cell RT-PCR, the expression profile for NK-1 receptor mRNA as well as for the μ-opioid receptor mRNA (16). Expression of the NK-1 receptor mRNA was found in only 30.7% of defined inspiratory neurons, whereas immunohistochemical data from the PBC region indicated NK-1 receptor expression in 65.9% of PBC neurons (Fig. 2). This difference revealed the presence of nonrespiratory neurons that express NK-1 (18). In contrast, μ-opioid receptor mRNA was detected in all inspiratory neurons analyzed, which is in line with immunocytochemical data reported by other laboratories (13). Our data demonstrate that 5-HT4(a) and μ-opioid receptor–mediated signaling pathways are coexistent in inspiratory neurons (Fig. 1A) and therefore are capable of interacting in an antagonistic manner; μ-opioid receptors operate through Gi/o proteins to decrease the cAMP levels (19), and 5-HT4(a) receptors counteract by activating Gs proteins to raise cAMP concentration (7).

The physiological significance of such potential interaction between 5-HT4(a) and μ-opioid receptor–mediated signaling in the regulation of respiration was further explored in the in vivo–like perfused rat brainstem–spinal cord preparation (16, 20), which contains the fully intact respiratory network, and finally verified in the live rat. First, we tested the effects of the 5-HT4(a) receptor–specific agonistic drug BIMU8 (10) on ongoing respiratory activity and found that vascular application of this drug significantly increased phrenic nerve activity at all doses tested (concentration range: 0.3 to 10 μM) (Fig. 4A). The whole-animal experiments verified that application of BIMU8 (1 to 2 mg/kg) significantly increased respiratory minute volume (RMV) in vivo (supporting online text). This stimulatory effect was 5-HT4 receptor–specific, because it was blocked by the specific antagonist GR 113808 in both experimental approaches (supporting online text). Involvement of 5-HT4/Gs signaling in the regulation of respiratory activity was confirmed by the findings that application of dibutyryl-cAMP increased phrenic nerve activity, whereas application of the adenylyl cyclase blocker SQ 22,536 decreased phrenic nerve activity. Coactivation of 5-HT3 receptors by BIMU8 could be excluded, because these receptors are not functionally expressed in respiratory neurons (21, 22)

Fig. 4.

Stimulation of 5-HT4 receptors by the selective agonist BIMU8 removes opioid-induced respiratory depression without loss of the antinociceptive effect of opioids in [(A) and (B)] the perfused brainstem preparation, as well as in (C) the intact in vivo rat. (A) The dose-dependent effect of BIMU8 on phrenic nerve minute activity (PNAmin) (16). Statistically significant changes from untreated controls are indicated by asterisks (P < 0.05). (B) Left: Application of fentanyl (blue) caused a marked depression of PNAmin by 91.2 ± 4.2% (n = 8, P < 0.05) as compared witha control (black). In three cases, fentanyl even led to apnea. Respiratory activity was reestablished by a subsequent application of BIMU8 (red) in a dose-dependent manner (16). Right: Responses of the spinal CFR. The black trace represents the CFR of the untreated control. Application of fentanyl (blue) suppressed the CFR by 60.9 ± 6.5% (n = 8, P < 0.01). Subsequent administrations of BIMU8 (red) did not affect opioid-induced depression of the CFR (16). (C) Left: Respiratory airflow in an anesthetized, spontaneous-breathing, in vivo rat. Application of 10 to 15 μg/kg fentanyl (blue) induced a marked reduction in RMV to 3.9 ± 8.5% of control RMV (n = 5, P < 0.001). Consecutive application of 1 to 2 mg/kg BIMU8 (red) prevailed over this effect of fentanyl and restored stable breathing, with RMV recovering to 70.6 ± 18.1% (n = 5, P < 0.001) of control RMV (16). Application of naloxone reestablished respiratory activity to the control level. Asterisks indicate transient periods of artificial ventilation, which were necessary to rescue the animal during genuine fentanyl treatment. Exp, expiration; Insp, inspiration. Right: Analysis of nociception based on the TFR. A quick TFR was obtained in control conditions (black), which was completely abolished after application of fentanyl (blue). This absence of analgetic response remained unchanged after subsequent administration of BIMU8 (red). Application of naloxone (1 mg/kg) immediately reestablished the TFR (supporting online text).

The physiological consequences of μ-opioid receptor activation were tested with the specific agonist fentanyl. Fentanyl is a synthetic opioid widely used for anesthesia and for the relief of acute and chronic pain, although it produces serious adverse reactions, such as hypoventilation (19, 23, 24). Application of fentanyl to the perfused rat brainstem–spinal cord preparation induced the expected antinociceptive effects, as seen by a 60.9 ± 6.5% (n = 8, P < 0.01) reduction of the C-fiber reflexes (CFRs) (Fig. 4B). At the same time, however, respiratory activity was almost completely suppressed (Fig. 4B). In three cases, exposure to fentanyl led to apnea that would have been lethal under normal conditions. The effects obtained in in vivo animals were even more pronounced. Here, fentanyl produced strong antinociceptive effects that resulted in a complete abolishment of the tail flick response (TFR) (Fig. 4C). However, as in the brainstem preparation, spontaneous respiratory movements were completely blocked (Fig. 4C) (supporting online text).

Therefore, we tested whether activation of the 5-HT4 receptor–mediated signaling pathway is effective in overcoming fentanyl-induced respiratory depression and apnea (Fig. 4B) (19, 25). To verify the power of 5-HT4 receptors in restoration of respiratory activity, we performed successive applications of fentanyl and of BIMU8. The crucial result was that consecutive applications of BIMU8 indeed reestablished stable respiratory activity within 3 min in the perfused brainstem preparation (Fig. 4B) (supporting online text). This effect was fully reproduced in vivo. In the latter cases, subsequent application of BIMU8 (1 to 2 mg/kg) overcame the fentanyl-induced apnea and restored stable breathing, with RMV recovered to 70.6 ± 18.1% within 3 min (Fig. 4C) (supporting online text).

Lastly, we investigated whether 5-HT4 receptor stimulation obliterates the nociceptive function of opioids. We tested CFRs in the brainstem–spinal cord preparation and the TFR in vivo (16). Application of BIMU8 after fentanyl treatment was sufficient to reestablish stable respiration in both test systems without any significant effects on the CFRs (Fig. 4B) or the TFR (Fig. 4C). Additional application of naloxone (1 mg/kg) immediately reestablished the TFR (Fig. 4C). The absence of BIMU8-induced effects on nociception can be explained by the finding that dorsal horn spinal interneurons reveal abundant μ-opioid but not 5-HT4(a) receptor immunoreactivity (fig. S2).

This study provides evidence that activation of 5-HT4 receptors in neurons of the medullary respiratory center represents a method for the treatment of respiratory depression induced by opioids. Stimulation of 5-HT4 receptors effectively counteracts fentanyl-induced respiratory depression without compromising its antinociceptive potency. An inspiring possibility is that application of 5-HT4 receptor agonists could be used for the treatment of critical respiratory events caused by fentanyl in postoperative situations and for the treatment of pain patients against overdose of opioids (23). In essence, a straightforward therapy that targets convergent intracellular signal pathways by means of a receptorspecific pharmacology (26) might open strategies for effective treatment in a wide spectrum of critical clinical situations.

Supporting Online Material

Materials and Methods

SOM Text

Figs. S1 and S2

References and Notes

References and Notes

View Abstract

Navigate This Article