Pontine μ-opioid receptors mediate bradypnea caused by intravenous remifentanil infusions at clinically relevant concentrations in dogs

Surgical Procedures

Anesthesia was induced by mask with isoflurane, and the dogs were intubated with a cuffed endotracheal tube and their lungs mechanically ventilated with an air-O₂-isoflurane mixture. The femoral artery was cannulated for blood pressure recording and blood gas sampling and the femoral vein for continuous infusion of maintenance fluids and administration of drugs. The animals were positioned in a stereotaxic device (model 1530; David Kopf Instruments, Tujunga, CA) with the head ventrally flexed (30°). A bilateral pneumothorax was performed to minimize brain stem movement and phasic inputs from chest wall mechanoreceptors. The animal was then decerebrated by midcollicular transection (Tonkovic-Capin et al. 1998), and isoflurane was discontinued. In a subset of dogs in protocol 2 (see below), studies were performed during isoflurane anesthesia without decerebration. The animals were ventilated with an air-O₂ mixture and maintained in hyperoxic isocapnia (Fi_O2 >0.6, end-tidal CO₂ range 40–50 mmHg). Complete access of the dorsal surface of the brain stem was obtained by an occipito-parietal craniotomy followed by cerebellectomy and external sagittal and nuchal bone crest removal. After bilateral neck dissections, phrenic nerve activity was recorded from the desheathed right C₅ rootlet. Bilateral vagotomy was performed to achieve peripheral deafferentation to avoid interference of the mechanical ventilation with the underlying central respiratory rhythm and respiratory neuronal activity. The PNG was obtained from the moving-time average (100 ms) of the amplified phrenic nerve activity and was used to produce timing pulses corresponding to the beginning and end of the inspiratory phase for the measurement of inspiratory duration (TI) and expiratory duration (TE). Peak phrenic activity (PPA) was also obtained from the PNG. The breath-by-breath fictive breathing rate [phrenic activity burst rate; f_B, breaths/min (BPM)] was calculated as 60/(TI + TE). Continuous neuromuscular block was achieved with pancuronium (0.1 mg·kg⁻¹·h⁻¹) to reduce motion artifacts during neuronal recordings. Esophageal temperature was maintained at 38.5 ± 1°C with a heating pad. If required, mean arterial pressure was maintained above 80 mmHg with adjustment of intravenous fluids and, when needed, phenylephrine infusion (0.5–5 μg·kg⁻¹·min⁻¹).

Microinjection Technique

A minimum of 1 h was allowed for preparation stabilization before data collection. A Hamilton microliter syringe (5 μl) with a calibrated scale (resolution: 50 nl) was used for microinjections of the μOR agonist DAMGO (100 μM) and the μOR antagonist NAL (100 μM) to locate μORs in the rostral pons that alter f_B. The relatively high concentrations of DAMGO and NAL were used to saturate the μORs in a small region (∼1-mm diameter) of the pons to map out the extent of the opioid-sensitive region; remi was not used for injection at high concentrations since the only available formulation contains glycine, which is in itself inhibitory to neurons and would have thus confounded the results.

Remifentanil Infusion

The short-acting μOR agonist remi was continuously infused intravenously at a rate of ∼0.5 μg·kg⁻¹·min⁻¹, which is the analgesic dose that reduces the requirements of a volatile anesthetic to blunt surgical stimuli in humans and dogs by 50% (Michelsen and Hug 1996; Michelsen et al. 1996). This infusion rate was adjusted to produce >60% reduction in f_B during hyperoxic isocapnea, which was maintained by mechanical ventilation.

Immunohistochemistry Methods Used for Preliminary Data

For immunohistochemical mapping of μOR, a carotid artery was cannulated rostrally in an anesthetized dog (n = 1) and perfused with phosphate-buffered saline (PBS) followed by 4% paraformaldehyde fixative with 0.2% picric acid in 0.1 M phosphate buffer. The brain was removed and postfixed in 4% paraformaldehyde fixative. Frozen transverse brain sections (25 μm) were cut from 10–17 mm rostral to the obex, and every fourth section was washed in PBS and incubated for 1 h in 5% normal goat serum in 0.3% Triton X-100 in PBS (PTX). Sections were then incubated in primary antibody, guinea pig anti-μOR-1 (1:3000 in 0.3% PTX; Millipore), for 24 h prior to Alexa Fluor 488-conjugated goat anti-guinea pig secondary (1:200; Invitrogen) for 1 h to yield a fluorescent signal for μOR. Sections were washed in PBS, floated onto slides, air-dried, dehydrated, and coverslipped. Control sections were processed simultaneously by incubation with serum instead of primary antibody and showed no staining for μOR. Sections were examined with a Nikon TE2000 microscope equipped for epifluorescence and photographed with an Optronics Quantifier 4 megapixel thermoelectrically cooled digital monochrome camera. Images were acquired at a resolution of 2,048 × 2,048 pixels with Optronics picture frame software and stored in TIFF format.

Preliminary Study Protocol: Dorsolateral Pons as a Potential Site of Opioid-Induced Bradypnea

Studies in four dogs, with experimental procedures similar to those described above, were used to examine the possible involvement of the dorsolateral pons in opioid-induced bradypnea. Pledgets saturated with 100 μM DAMGO or 500 μM and 10 mM NAL were bilaterally placed on the dorsal surface of the pons between the IC and the superior cerebellar peduncle. The pledgets consisted of strands, three per side, of ∼2-mm-diameter cotton string ∼10 mm in length. They were submerged in a beaker containing either DAMGO or NAL so as to be completely saturated. During the protocol, they were exchanged about every 6 min during a ∼30-min period to provide a continuous supply of the drug to the surface. After a maximum bradypneic response determined from the PNG, the dorsal pontine surface was rinsed with saline and NAL pledgets were then applied and exchanged as above to reverse the DAMGO-induced effect.

Protocols

Protocol 1: respiratory effects of direct pontine μOR activation via pontine microinjection of DAMGO.

In a first series of animals (n = 10), pontine microinjections of DAMGO were used to define the regions that mediate opioid-induced bradypnea. A 5-μl Hamilton syringe was stereotaxically positioned over the dorsal surface of the pons, between the IC and superior cerebellar peduncles. A series of injections of DAMGO (100 μM, 300–500 nl each), ∼10 min apart, were given covering 1 mm × 1 mm grids over a region from 4 to 8 mm lateral of midline and extending from 1 mm rostral to the caudal pole of the IC to 3 mm caudal to that pole at a depth of 3–4 mm from the dorsal surface (see Fig. 3, shaded areas). The 3- to 4-mm depth was suggested by results of preliminary studies. TI, TE, and PNG were continuously monitored online with a PowerLab system (ADInstruments, Castle Hill, Australia; 16SP and Chart). Average values of these variables and f_B were obtained for off-line analysis from data starting 2 min before and for the 10-min period after each DAMGO microinjection. At the end of each experiment, bilateral microinjections (∼1 μl) of fluorescent latex microspheres (Lumafluor.com; Red Retrobeads), diluted to 5% of the supplied concentration, were used to mark the most sensitive areas for histology. The brain stems were removed postmortem and submersed in paraformaldehyde fixative (4%) for 1 wk before being frozen and sectioned. Sequential sections (25 μm) were cut from 10 to 17 mm rostral to obex. Alternate sections were stained with neutral red for histological identification of nuclear groups and compared with unstained adjacent sections showing fluorescent beads.

Open in a separate window

Fig. 3.

Schematic of the dorsal view of the canine pons/medulla with cerebellum removed showing the coordinate system used in this study. Shaded areas indicate the pontine region explored for opioid-induced bradypnea. The caudal pole of IC and the midline were used as the R-C and M-L initial (0/0) reference points, respectively.

Data Analysis

Protocol 1: respiratory effects of direct pontine μOR activation via pontine microinjection of DAMGO.

The DAMGO microinjections resulted in a cumulative slowing of f_B in sensitive areas. Therefore we calculated the incremental effect of each additional microinjection off-line. A new local control was obtained from the data ∼1 min prior to each microinjection. A second measurement was obtained 5–10 min after each microinjection but prior to the subsequent microinjection. These two values were used to calculate the percent change in f_B. To aid in this analysis, the breath-by-breath data were fitted via a regression procedure known as LOESS or LOWESS (Cleveland and Devlin 1988), in which a series of subsets of the data is used to provide the best fit that follows the trend features of the data. SigmaPlot 11 (Systat Software, San Jose, CA) was used for this purpose and for plots used in the figures. Examples of the fits are indicated with red lines in the f_B data (BPM) in Fig. 1, A and B. Figure 1A shows an example of the effects of a series of DAMGO microinjections (vertical arrows) at the three locations indicated on breath-by-breath f_B (BPM) with the data fit superimposed (red line). Note the cumulative effect of the microinjections on the slowing of phrenic respiratory rate. Time-expanded plots of the same data for each microinjection site are shown with time shifted relative to the time of microinjection in Fig. 1B (black arrowheads at microinjection, t = 0). From these plots, the incremental changes in f_B were calculated as a percent change from the preinjection local control values as indicated by arrows in Fig. 1B. Contour plots were constructed to display the percent change in f_B as a function of the location on the rostral-caudal (R-C) by medial-lateral (M-L) grid (Fig. 1A, bottom).

Open in a separate window

Fig. 1.

Example of data illustrating the methods used to quantify the effects of [d-Ala²,N-Me-Phe⁴,Gly-ol⁵]-enkephalin (DAMGO) microinjections during the pontine mapping protocol (protocol 1). Microinjections of DAMGO (100 μM, 500 nl) were sequentially injected beginning at 0 rostral-caudal (R-C) and 4 lateral and progressed laterally until 8 mm lateral. The medial to lateral sequence of microinjections was subsequently repeated for −1, −2, −3, +1 R-C. Typical depth of injections was 3 mm from dorsal surface (see 3rd coordinate). A, top: effects of a series of DAMGO microinjections [vertical arrows with coordinates as indicated: R-C/medial-lateral (M-L)/depth (D)] on the phrenic neurogram (PNG) and breath-by breath values of TI (blue symbols), TE (red symbols) and fictive breathing rate [f_B, breaths/min (BPM), black symbols]. The corresponding locations are shown by the 3 black circles at 0 R-C on the contour plot (A, bottom). Thick red line in rate data is the local regression curve (LOWESS function) indicating the best-fit trend. Note that there is a progressive increase in TI and TE, a decrease in rate, and no change in peak PNG. B: time-expanded plots of f_B for the corresponding microinjections indicated on the contour plot (black circles) with time of injection set to 0 (vertical dashed lines). Red lines: best-fit regression curves. Because of cumulative effects of DAMGO, incremental % changes relative to the preinjection values (dashed horizontal lines) were calculated 6–10 min after injection (arrows with values). Arrowheads: time of microinjections. R-C, lateral, and depth coordinates are given at right of each plot. Contour plot (A, bottom) shows % reduction in breathing rate caused with DAMGO microinjections at the various R-C and M-L locations. In this example, the maximum change in f_B of ∼25% occurred at −2 R-C and 6 lateral. Corresponding data are shown in B, bottom, with changed scale.

Protocol 1: Respiratory Effects of Direct Pontine μOR Activation via Pontine Microinjection of DAMGO

In 10 dogs, 286 DAMGO microinjections (390 ± 13 nl/microinjection; 28.6 ± 3.3 microinjections/dog) in the dorsal lateral pons extending from 1 mm rostral to the caudal pole of the IC to 3 mm caudal and from 4 to 8 mm lateral of the midline (shaded areas in Fig. 3) were performed to delineate the region that mediates an opioid induced decrease in f_B. Figure 4 shows an example of a series of unilateral DAMGO (100 μM, 500 nl each) microinjections in the lateral direction, which resulted in significant slowing from ∼11 BPM to apnea (Fig. 4, bottom, f_B). The ∼10-min apnea was subsequently reversed by intravenous NAL (∼10 μg/kg; Fig. 4, right). PPA increased with the increase in TI. Time-expanded records of PNG (Fig. 4, top) are shown in Fig. 5A and superimposed traces of the first cycle in each record in Fig. 5B. Note that the ramp slope of the PNG is not altered and the peak height is dependent on magnitude of TI, suggesting that respiratory drive has not changed. A nonlinear relationship between TI and the following TE of each cycle is also seen after vagotomy (Fig. 5C). While DAMGO increases both TI and TE, the major decrease in f_B is primarily due to the increase in TE.

Open in a separate window

Fig. 4.

Example of the effects of a series of unilateral DAMGO microinjections (diamond symbols with pontine R-C/lateral coordinates). Note the progressive increases in TI and TE and decrease in f_B and subsequent apnea. An IV bolus of NAL (10 μg/kg) partially antagonized the apnea. Larger subsequent doses were required to fully antagonize the DAMGO effects. A moderate increase in peak PNG is entirely due to the increase in TI without changes in drive (see Fig. 5 and text for further explanation).

Open in a separate window

Fig. 5.

Pontine DAMGO microinjections produce changes in timing but not drive. A: time-expanded segments of PNG record in Fig. 4, top, at the times indicated (a–e) below the PNG trace. The 1st PNG cycle in each segment has been time aligned at 0 s. B: superimposed traces of the 1st breath cycle in each record. Note that the slope of the PNG ramp is not altered and the peak height is dependent on magnitude of TI, suggesting that respiratory drive has not changed. C: plot of TI vs. the subsequent TE of each cycle during vagotomy appears to be nonlinear. While DAMGO caused increases in both TI and TE, the large decreases in f_B are mainly due to the large increases in TE.

The data of the contour plots from 10 dogs were averaged at each of the grid coordinates to produce the contour plot of the average responses (Fig. 6, top). The average magnitude of the incremental responses in f_B for the most sensitive locations, indicated by superimposed open circles in Fig. 6, top, are shown in Fig. 7A. The largest incremental responses were located at 5- to 6 mm lateral to midline and 0 to −1 caudal to the IC and were in the range of 15–20%. The cumulative maximum effect of a series of DAMGO microinjections per experiment was on average a 74 ± 6% reduction in baseline f_B from 18.6 ± 4.2 to 5.0 ± 1.5 BPM with 8.2 ± 1.1 microinjections. PPA was not significantly changed (0.78 ± 1.52%, n = 74 microinjections) by the DAMGO microinjections. The prominent and most consistent response over the lateral range of 4 to 8 mm from the midline was a >5% decrease in f_B as shown by the pooled response data from all R-C sites corresponding to each lateral site (Fig. 7B, Bradypnea). At some of the sites in some animals, DAMGO microinjections caused no response or an increase in f_B. However, increases in f_B >5% were seldom seen at 5–7 mm lateral (Fig. 7B, Tachypnea) and were observed less often than little or no responses (Fig. 7B, No Response).

Open in a separate window

Fig. 6.

Top: contour plot for DAMGO-induced decreases in breathing frequency based on mean % change in f_B values obtained in 10 dogs. The most opioid-responsive region was confined to an area of <2 × 2 mm. The largest decreases in f_B at any injection site were ∼20% and were located at −1 R-C and 5 lateral. Superimposed open circles indicate sites from which data were tested for statistical significance (see text for details) (Fig. 7A). Bottom: contour plot for NAL microinjection reversal of IV remifentanil (remi)-induced bradypnea. Values are the average % reversal of the maximum depression of f_B caused by remi from 21 dogs (see methods for more details regarding calculations). The largest mean values were ∼60% (red). Open circles indicate sites where data were statistically analyzed as shown in Fig. 7C.

Open in a separate window

Fig. 7.

A: % change in f_B for each microinjection of 300–500 nl of 100 μM DAMGO at the site indicated above each bar (also in Fig. 6A, open circles; n = 10 dogs). ^#P < 0.05 difference from zero change (1-way ANOVA). B: types of responses to DAMGO as % of the number (N) of DAMGO microinjections per lateral coordinate (all R-C locations pooled). Tachypnea was rarely observed throughout the mapped region (darkest bars). C: % reversal of the remi-induced depression of f_B per NAL microinjection (540 nl of 100 μM NAL). *P < 0.05, **P < 0.01 changes relative to no effect (1-way ANOVA).

Protocol 2: Pontine Microinjections of Naloxone to Reverse Bradypnea Induced by Systemic Remifentanil

The reversal of remi-induced bradypnea by NAL microinjections in the region delineated by the DAMGO microinjections (protocol 1) was studied in 21 dogs. An example of a protocol is given in Fig. 8, where remi (0.8 μg·kg⁻¹·min⁻¹) markedly decreased the peak PNG and PNG ramp slope (not shown) and altered f_B. As illustrated in this example, after initiation of the remi infusion a transient increase in respiratory rate (tachypnea) was frequently observed (e.g., f_B above the dashed horizontal line, Fig. 8). This was followed by bradypnea and finally apnea as the remi infusion continued. Throughout this time, PPA gradually decreased until it reached a plateau well above zero (see PNG record before start of NAL microinjections and after and time-expanded records c–f, Fig. 8). This suggests that the apnea resulted from an interruption of respiratory rhythm rather than a completely suppressed peak PNG amplitude. The intermittent large tidal breaths illustrated in the peak PNGs are sighlike cycles that are also frequently observed during remi infusions. During the apneic period, short periods of phrenic activity reappeared, suggesting that the selected remi infusion rate did not depress respiratory activity much below the phrenic apneic threshold. After ∼10 min of apnea, bilateral pontine microinjections of NAL (1 μl each) were made at 7 mm lateral to the midline and 1 mm caudal to the caudal pole of the IC at a depth of ∼3 mm, which completely reversed the apnea and resulted in a transient respiratory rate above the pre-remi control frequency (transient tachypnea). In this example, the first NAL microinjection was sufficient to initiate the reversal of the apnea. Despite these profound effects of pontine NAL microinjections on f_B, there was very little effect on peak PNG (Fig. 8, middle section between the 2 arrows). As the NAL effect gradually declined during continued remi infusion, bradypnea and apnea reoccurred. After termination of the remi infusion, peak PNG and f_B gradually returned to pre-remi levels. The length of the entire protocol was almost 2 h.

Open in a separate window

Fig. 8.

Example of the reversal of IV remi-induced bradypnea/apnea by NAL (100 μM) microinjections in the dorsolateral pons (−1 mm R-C, 7 mm lateral to midline, 3 mm ventral to dorsal surface). The IV remi infusion at a rate of 0.8 μg·kg⁻¹·min⁻¹ decreased peak PNG, PNG slope, and breathing rate f_B until transient apneas ensued (PNG record, arrows). Intermittently, large sighlike respiratory cycles were also observed. A transient increase in f_B above control baseline (RATE record, horizontal dashed line) was observed after initiation of remi infusion prior to bradypnea. Throughout the infusion, peak PNG gradually decreased until a steady-state plateau at ∼20% peak PNG was reached. Peak PNG remained depressed after the two bilaterally symmetrical 1-μl NAL microinjections (compare records at times c, d, e, and f), while respiratory rate transiently recovered despite ongoing remi infusion. During the initial apneic period, short periods of phrenic activity reappeared (between times c and e), suggesting that the opioid plasma concentration was close to the phrenic apneic threshold. After 10 min of apnea, the first unilateral NAL microinjection reversed the apnea (the 2nd contralateral microinjection given as part of the protocol), and a transient tachypnea similar to the initial one resulted without any notable change in peak PNG. As the effect of the bilaterally microinjected NAL gradually declined during sustained IV remi infusion, bradypnea and apnea eventually reoccurred. After termination of the remi infusion, both peak PNG and f_B gradually returned to pre-remi levels. The entire protocol lasted almost 2 h. Time-expanded records of PNG are shown for times indicated by the labeled arrows (a–g). e: Onset of reversal of the apnea by the 1st NAL microinjection, where the 1st PNG cycle shows a sigh in this case.

During 39 NAL microinjection protocols, the pre-remi f_B of 14.8 ± 1.6 BPM was reduced by 72 ± 4% by 0.41 ± 0.05 μg·kg⁻¹·min⁻¹ of intravenous remi. The level of remi-induced depression of f_B was similar in decerebrate (n = 18 runs) versus isoflurane-anesthetized (n = 21 runs) dogs (−64 ± 6% vs. −78. ± 3%); however, to obtain this same level of depression, decerebrate dogs required a ∼35% higher remi infusion rate than anesthetized dogs (0.46 ± 0.04 vs. 0.34 ± 0.04 μg·min⁻¹·kg⁻¹; P = 0.029). Remi also reduced the peak PNG by 25.3 ± 3.6%. There was no significant correlation between the degree of remi-induced depression in peak PNG and f_B (r = −0.24; P = 0.14, Spearman rank order).

A total of 84 NAL microinjections (540 ± 35 nl each) in 21 dogs (decerebrate and anesthetized) were used to map the region where reversal of remi-induced bradypnea was observed (Fig. 6, bottom). The average magnitude of the pooled responses at the most sensitive NAL locations of Fig. 6, bottom (open circles) are given in Fig. 7C. The magnitude of bradypnea reversal was ∼50% per NAL microinjection, and typically two NAL microinjections in these locations were able to completely reverse remi-induced bradypnea or apnea (101.8 ± 11.3%). There was no difference in the amount of NAL-induced reversal of the remi effect between the two types of preparations (107 ± 9% vs. 82 ± 14%, respectively). A total of 59 of the 84 NAL microinjections were located within the sensitive region extending from 0 to −1 mm rostrally and from 5 to 7 mm laterally (Fig. 7C). Unilateral microinjections (28/59) were just as effective as the bilateral microinjections (31/59), 46.7 ± 9.5% vs. 49.8 ± 8.2% (P = 0.60) reversal/microinjection. For 12 protocols in 9 of the dogs, only a single microinjection at the first site selected within the sensitive region was able to produce complete reversal. The NAL microinjections that reversed remi-induced bradypnea had no effect on peak PNG (0.75 ± 4.1% change; P = 0.86) and only a small increase (7.1 ± 2.5% change; P = 0.02) in PNG/TI (average slope of the PNG).

Since the remi-induced bradypnea protocols take 1.5–2 h to complete and only a small number of NAL microinjections were required to totally reverse the bradypnea, the number of NAL microinjections per experiment was less than the number of DAMGO injections in protocol 1.

Role of Pontine Parabrachial/Kölliker-Fuse Complex in Control of Breathing Frequency

The results of this study emphasize the powerful role that the pontine PB-KF complex plays in the control of phase timing and breathing frequency (e.g., Fig. 4). The range of f_B control is remarkable, since microinjected DAMGO was capable of producing significant bradypnea up to complete respiratory arrest through increases in TE. On the other hand, remi-induced bradypnea of similar magnitude (∼72% reduction in baseline f_B) could be fully reversed by the pontine NAL microinjections.

The importance of the pontine PB-KF complex in the control of breathing has been known for some time. This area has been referred to as the “pneumotaxic center” or, more recently, the pontine respiratory group (PRG) (Bianchi et al. 1995; Feldman 1986). PRG neurons project to the ventral respiratory group (Chamberlin et al. 1999). Electrical stimulation or chemical stimulation with glutamate in the lateral PB complex in cats or rats predominantly evoked respiratory facilitation, while stimulation of the more ventrally located Kölliker-Fuse nucleus or medial PBN decreased respiratory rate (Chamberlin and Saper 1994; Cohen 1971; Lara et al. 1994). Song and Poon (2009) also showed that bilateral lesions of the lateral PBN resulted in significant bradypnea (increase in TE) under hyperoxia and in marked attenuation of the tachypneic response (decrease in TE) to hypoxia. In vitro voltage-clamp experiments in rat lateral PB neurons show that μOR agonists increase the potassium conductance of the membrane and inhibit the neurons through hyperpolarization (Christie and North 1988). It is thus reasonable to deduce that the inhibition of neurons whose excitation results in tachypnea would cause bradypnea.

The maximum responses to DAMGO and NAL microinjections were located 5–7 mm lateral from midline, 0–1 mm caudal of the inferior colliculi, and 3–4 mm deep and correspond histologically with the structures of the PBN. There were more tachypneic responses medial of this area, suggesting that DAMGO microinjections into the medial PBN may have inhibited neurons that produce bradypnea (Lara et al. 1994). Still, in a few animals DAMGO microinjections into the presumed lateral PBN had no effect or resulted in tachypnea (Fig. 7B), possibly due to anatomical differences among animals. The fact that complete reversals of bradypnea by single NAL microinjections were seen in the most lateral locations (i.e., 7 mm lateral, Fig. 6, bottom; also example in Fig. 8) suggests that the lateral PB region may play a significant role in the bradypneic effect of clinically used μOR agonists. However, significant effects were also found at 6 mm lateral to midline, which appears to correspond with the medial PBN. Because of the size of the microinjections and the subsequent diffusion, our ability to resolve the location within the PB complex is limited.

Other Brain Stem Sites Possibly Involved with Opioid-Induced Respiratory Depression

In adult rats, small numbers of weakly μOR-ir neurons and more abundant μOR-ir fibers and terminals were found throughout the rostrocaudal extent of the VRG (Lonergan et al. 2003). Interestingly, there appeared to be no difference in the number of μOR-ir neurons in the preBC compared with the Bötzinger complex. Other respiratory-related areas with intense μOR-like immunoreactivity were the pontine lateral PBN, the locus coeruleus, the rostral ambiguus nucleus, and the medial and commissural subnuclei of the NTS (Chamberlin et al. 1999; Ding et al. 1996). The mere presence of μORs, however, does not allow any conclusions as to their clinical significance. Similarly, the presence of μ- and δ-opioid receptors on canine respiratory premotor neurons in the VRC, demonstrated by picoejection of the respective agonist onto single neurons, did not translate into a depression of these neurons by clinical concentrations of a μOR agonist (Stucke et al. 2008). The apparent reason was that the concentrations required for neuronal depression were in the 100–1,000 μM range with a threshold of 10 μM for morphine (Stucke et al. 2008). Perhaps the concentration of synaptically released endogenous opioids is high enough to be effective on these neurons. In contrast, the plasma concentration of remi in dogs at the infusion rate used in the present study (∼0.5 μg·kg⁻¹·min⁻¹) is estimated to be ∼14 nM (Michelsen et al. 1996), and the effect site concentration (extracellular fluid) may be ∼50% lower (Kabbaj et al. 2005). In whole brain tissue, the binding affinity for remi was found to be 21.1 nM (Poisnel et al. 2006).

NAL microinjections did not always achieve a full reversal of the remi effect. Although this could be due to the spread of the drug within the PBN, it is possible that other areas of the brain stem may also contribute, to a lesser degree, to the opioid-induced respiratory depression. With microdialysis of fentanyl (100 μM) into the preBC region of anesthetized spontaneously breathing adult rats, it was reported that the preBC was solely responsible for the respiratory rate suppression by systemic opioids (Montandon et al. 2011). However, this study used relatively large, stereotaxically guided (relative to bregma) microdialysis probes (0.24-mm diameter, 1-mm length of permeable membrane) located at an average distance of 760 μm from the preBC (n = 8; based on corrected vector differences from Supplemental Table 1 in Montandon et al. 2011). Since the respiratory effect was measured after dialyzing 100 μM fentanyl over 30–60 min, it is very difficult to judge the amount of fentanyl released or its concentration distribution within the brain stem. For example, genioglossus muscle activity was significantly depressed, even though the presumed location of hypoglossal premotor neurons was on average twice as far from the dialysis probes as the preBC (Supplemental Table 1 in Montandon et al. 2011). The effective concentration of fentanyl in the brain stem most likely exceeded by far the plasma concentration of ∼24 nM, which has been shown to depress minute ventilation by 57% in awake adult rats (Yassen et al. 2005). Montandon et al. (2011) used an indirect method based on probe locations and response latencies to estimate the site of respiratory depression. However, because of the heterogeneity of brain tissue, the rate of diffusion will vary with path direction, as shown in their data by comparing diffusion-front velocities (distance/latency; Supplemental Table 1 in Montandon et al. 2011), and result in dubious estimates of effect-site locations.

Other in vivo studies suggest that the preBC is not the location of opioid-induced respiratory depression. Unilateral microinjection of the potent endogenous μOR agonist endomorphin-1 (60 nl, 10 mM, binding affinity K_i: 0.36 nM; Zadina et al. 1997) into both the preBC and Bötzinger complex in adult rats produced an increase in phrenic burst rate, rather than a bradypneic response that is observed clinically (Lonergan et al. 2003). These findings are consistent with those in decerebrate dogs, where localized microinjections of DAMGO (∼140 nl, 100 μM) within the functionally identified preBC region produced tachypnea (44% increase in f_B; n = 16), while intravenous infusion of remi counteracted the DAMGO-induced tachypnea and produced marked bradypnea (55% decrease in f_B relative to baseline control; Mustapic et al. 2010). More importantly, multiple bilateral microinjections of NAL (500 μM, 120 nl each, 3 per side) into the preBC region during steady-state, remi-induced bradypnea had no effect on f_B and PPA (Mustapic et al. 2010). Finally, in awake goats, relatively large microinjections (10 μl) of DAMGO (100 nM) into the preBC had no effect on the eupneic breathing pattern (Krause et al. 2009). Altogether, these studies suggest that the preBC is not directly involved in decrease in respiratory rate by clinical concentrations of opioids.

Because of the invasive nature of our studies that require anesthesia, it is not possible to assess the opioid-induced effects on breathing in the cortex and other supra-brain stem structures, which may play an important role in the awake state. For example, Pattinson et al. (2009), using MRI studies in awake volunteers, showed that remi reduced the urge to breathe and this was associated with a decrease in activity in the left anterior insula and operculum cortical areas. However, our studies were focused on the effects of opioids on the automatic control of breathing, where opioids cause dose-dependent bradypnea.

The authors thank Jack Tomlinson, Biological Laboratory Technician, Zablocki VA Medical Center, for excellent technical assistance.