INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

CENTRAL CHEMORECEPTORS MEDIATE changes in breathing and blood pressure in response to changes in CO₂ and H⁺ within the brain (14, 16-18, 23). Whereas early work located central chemoreceptors at sites just beneath the ventral medullary surface (14, 16), recent experiments in vitro and in vivo have indicated that they have a widespread distribution within the brain stem (4, 5, 13, 17-19). These locations include the regions superficially just beneath the surface of the ventral lateral medulla, the nucleus tractus solitarius, the locus ceruleus, the midline raphe, and the ventral respiratory group. Why are central chemoreceptors located at so many sites? We hypothesize that some chemoreceptor sites vary in their effectiveness, depending on the state of arousal.

In this study, we focused on one part of the chemosensitive region lying just beneath the rostral ventral medullary surface, the retrotrapezoid nucleus (RTN). We asked whether the response to focal acidification of the RTN differs during wakefulness vs. sleep. The RTN is one of several brain stem sites that communicate with the respiratory control network (1, 6, 7, 17, 18, 30). Destruction of the RTN in anesthetized and decerebrate animals reduces resting ventilatory output, often to apnea, and substantially reduces the ventilatory response to systemic CO₂ stimulation (17, 20). Focal CO₂ stimulation of the RTN in anesthetized animals increases ventilatory output by 25-40% of the change observed with stimulation of all central chemoreceptor sites by systemic CO₂ stimulation (13). The RTN appears to provide two sources of input to the respiratory control network: tonic and chemosensory. However, lesions of the RTN in unanesthetized rats (1) and cooling of the rostral ventrolateral medulla (RVLM), including the RTN in unanesthetized goats (9, 21), result in much less dramatic effects. Ventilation (E) at rest during wakefulness is unchanged (1) or decreased slightly (9, 21), and the response to CO₂ is decreased much less in comparison to responses to RTN disruption in anesthetized animals (1, 9, 21). These data suggest that the role of the RTN may vary considerably in different arousal states.

In this study, we used a microdialysis probe (3) (CMA/Microdialysis, Acton, MA) to produce a focal acidosis in the RTN of unanesthetized, unrestrained rats. The probe tip with semipermeable membrane is 1 mm in length and 240 µm in diameter, with a volume of 45 nl. The guide tube and dialysis probe are made of a rigid, sturdy material, allowing their use in a chronic animal. In a separate group of anesthetized rats, we measured the distribution of tissue pH changes during dialysis with this probe. Although we have produced a very focal tissue acidosis with rapid and reversible pH changes in the chemoreceptor regions of anesthetized animals by a CO₂ diffusion pipette (13), this glass pipette has limitations in unanesthetized and unrestrained animals.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

General Preparation

Anesthetized group. Seventeen male Sprague-Dawley rats (300-450 g) were anesthetized with -chloralose (60 mg/kg) and urethan (550 mg/kg). The trachea, femoral artery, and vein were cannulated. We paralyzed the rats with Gallamine triethiode (3 mg/kg) and artificially ventilated them with 100% O₂. The arterial blood pressure and end-tidal PCO₂ were monitored, and rectal temperature was maintained at 37.6°C. Bilateral thoracotomies were performed, and a positive end-expiratory pressure of 3-5 cmH₂O was maintained during the experiment. Both vagi were cut, and the ventral medullary surface was surgically exposed. The phrenic nerve was isolated, cut, and placed on a bipolar electrode for recording. Phrenic nerve activity was amplified (BMA831 amplifiers, Charles Ward), rectified, and integrated. Integrated phrenic nerve amplitude (PNA), arterial blood pressure, and end-tidal PCO₂ were recorded on a strip-chart recorder (MFE 1400, MFE).

CO₂ microdialysis and pH electrodes. A microdialysis probe (CMA/11, CMA/Microdialysis) with 1-mm length of cuprophane membrane and 0.24-mm outside diameter was placed directly into the RTN region from the ventral surface and was connected to a syringe pump for dialysis. A pH electrode coupled to a potentiometer (model EA920, Orion Research) was placed at various distances from the dialysis probe to monitor the tissue pH change. Each rat received one dialysis probe and one pH electrode. The details of the construction of the pH electrode were described previously (5, 13). To allow comparison of tissue pH among animals, the changes in each animal were normalized to the response of the electrode in vivo to the change in end-tidal CO₂ from 4 to 9%. All pH electrodes were calibrated in vitro by using standard buffer solutions at pH values of 4, 6, 7, and 10 before and after the experiment.

Conscious group. Fifteen male Sprague-Dawley rats (300-450 g) were anesthetized with ketamine (100 mg/kg im) and xylazine (20 mg/kg ip). The crown of the skull was shaved, and the skin sterilized with betadine and alcohol. The head was placed into a Kopf stereotaxic holder, and a dialysis guide cannula (0.38 mm outer diameter) with a dummy was implanted into the medulla. Each rat received one guide tube. The coordinates for probe placement were 2.2 mm caudal and 1.8 mm lateral from lambda and 10.6-10.8 mm below the dorsal surface. The guide cannula was secured with cranioplastic cement, and the wound sutured. The abdominal surface was shaved, the skin was sterilized, an incision was made through the linea alba, and a sterile telemetry temperature probe (TA-F20, Data Sciences, St. Paul, MN) was placed in the abdominal cavity. The incision was closed, and the animal was allowed to recover for 3-4 days. Of the 15 rats, 9 resulted in successful experiments, 7 with guide tubes in the RTN and 2 with guide tubes placed outside of the RTN region.

CO₂ dialysis solution. The artificial cerebrospinal fluid (aCSF) was equilibrated with 1) 5%, 2) 25%, 3) 50%, or 4) 100% CO₂. The composition of the aCSF was (in mM) 152 sodium, 3.0 potassium, 2.1 magnesium, 2.2 calcium, 131 chloride, and 26 bicarbonate. The calcium was added after the aCSF was warmed to 37°C and equilibrated with CO₂. The pH of each solution was monitored to ensure that the equilibration was reliable.

E measurement. The plethysmograph chamber used in these experiments is similar to the setup described by Jacky (12) and Pappenheimer (22). The analog output of the pressure transducer was recorded on a strip-chart recorder (model 1200, Honeywell) and on tape by using a Vetter Digital system (model 3000A). The animal chamber operates at atmospheric pressure, with the inflow and outflow of inspired gases balanced to prevent hyper- or hypobaric conditions in the box. The inflow gas was humidified, and the flow rate was controlled by a flowmeter (model 7491T, Matheson). The outflow port was connected to the in-house vacuum system via a flowmeter. A high-resistance "bleed" of the outflow line provided ~100 ml/min of outflow gas to the O₂ analyzer (model SA-3, Applied Electrochemistry). The flow rate through the plethysmograph was maintained at or above 1.4 l/min to prevent CO₂ rebreathing. The plethysmograph was calibrated with 0.3-ml injections.

O₂ consumption (O₂) and temperature measurement. O₂ was measured by calculating the difference in O₂ content between inspired and expired gas. O₂ = (_in × FI_O2)  (_out × FI_O2), normalized to ml · g body wt¹ · h¹, where _in is inflow, FI_O2 is fractional inspired O₂, and _out is outflow. The inflow O₂ content was calibrated at the beginning of each experiment, and the flow rate of gas was set at 1.4-1.5 l/min. The outflow content of O₂ was read constantly from the O₂ sensor during the experiment. The chamber temperature was measured by a thermometer inside the chamber. Rat body temperature was measured by using the analog output via telemetry from the temperature probe in the peritoneal cavity.

Anatomic analysis. At the end of the experiment, the rats were killed, and the medulla was quickly removed and fixed in 4% paraformaldehyde. Brain stems were frozen and then sectioned at 50-µm thickness with a Reichert-Jung cryostat. The sections were counterstained with cresyl violet. We identified anatomic landmarks and the site of dialysis probe placement by using a rat brain atlas (24) for reference. The guide tubes were removed postmortem but before brain stem removal and sectioning. To remove the guide tubes required manipulation and produced tissue disruption. This facilitated the anatomic verification of guide-tube and probe-tip location but also increased the volume of tissue disruption compared with that produced by simple insertion.

Data analysis. Respiratory data were transferred from the Vetter Digital system to the DataPac III software system. With the use of the DataPac III system, a breath-by-breath analysis was performed with the pressure deflections and the respiratory cycle time for each breath being determined over a 20- to 30-s time period. The data were exported to Sigmaplot 4.0 (Jandel Scientific software), and tidal volume per 100 g body wt, frequency, and E per 100 g body wt were calculated for each breath.

Experimental Protocol

Experiments with anesthetized rats. After finishing the surgeries, we placed a CO₂ dialysis probe directly into the RTN region of each rat and a pH electrode at a single distance from the probe. The precise locations of the dialysis probe and pH electrode were ascertained during postmortem anatomic examination. All animals were first tested for their responsiveness to inspired CO₂ after at least 30 min of recovery from probe and pH electrode placement. The baseline end-tidal CO₂ was set just above the apneic threshold, usually at 28 Torr. A CO₂ response was determined by increasing the inspired CO₂ and monitoring PNA, frequency, and tissue pH at end-tidal CO₂ values of 28, 42, 56, and 63 Torr. We allowed at least 30 min for the rat to recover from the CO₂ stimulation. Dialysis was then performed over a 15- to 30-min period at a speed of 45 l/min by using aCSF equilibrated with 25, 50, or 100% CO₂. The effect on PNA, phrenic discharge frequency, tissue pH, and blood pressure was observed and recorded. When dialysis was completed, another systemic CO₂ response test was performed. The pH and phrenic amplitude value at 63 Torr end-tidal PCO₂ was used as the maximum value for normalizing the pH and phrenic amplitude CO₂ dialysis response data. The phrenic response to CO₂ dialysis was normalized to percent baseline.

Chronic experiments. 25% CO₂ dialysis group. We dialyzed each rat during both wakefulness and sleep. The dialysis tube and cannula dead space are taken into account along with the dialysis fluid flow rate so that t = 0 is the estimated time at which the CO₂-equilibrated solution reaches the exchange membrane. The rat was judged to be asleep by behavioral criteria: it was curled up, motionless, with eyes closed. We monitored the rat's behavior carefully during the period of dialysis. Most of the sleep experiments were conducted between 9:00 AM and 3:00 PM, and most of the awake experiments after 3:00 PM when rats were more alert. The rats were initially weighed and then gently held while the dummy cannula was removed and the dialysis probe inserted into the guide cannula. The animals were placed into a plethysmograph chamber (12, 22) and allowed 30-40 min to acclimate. Measurements were taken in room air and with 7% CO₂ inhalation during periods of wakefulness and sleep. After the systemic CO₂ response, the animals were exposed to room air and allowed to recover for at least 30 min. All dialysis experiments were performed in the room-air-breathing condition. Baseline E, O₂, and body temperature measurements were taken. The dialysis pump was run for 20 min at a speed of 45 l/min. The measurements were taken over a 20- to 30-s period at 0, 5, 10, 15, and 20 min during dialysis and at 5- to 20-min intervals after dialysis until E returned to, or near to, the control level. The seven rats with correct guide-tube placement received 13 dialysis trials during wakefulness and 10 during sleep.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Measurement of Tissue Spread of pH During Dialysis in the RTN

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Change () in brain stem tissue pH (expressed as %maximum) is shown as function of distance of recording pH electrode from dialysis probe (in µm). There are 3 sets of data: artificial cerebrospinal fluid (aCSF) dialysates equilibrated with high (100%; ), medium (50%; ), and low CO2 (25%; ). Each symbol represents 1 experiment in 1 anesthetized rat. Lines are drawn by hand through data points. Change in tissue pH produced by dialysis is normalized in each animal to change in tissue pH observed in that animal when end-tidal CO2 is increased from 4 to 9%. The 100% value then represents tissue pH change like that observed when end-tidal PCO2 is increased from 28 to 63 Torr.

Tissue pH was measured by pH electrodes at varying distances from the dialysis probe, which were measured anatomically postmortem. Results are shown for three different dialysis conditions, high (100% CO₂ equilibrated with the dialysis solution), medium (50%), and low (25%). With high-CO₂ dialysis, the tissue pH change at the probe was approximately that which would occur with an end-tidal PCO₂ of 110 Torr, and it decreased with distance such that at ~750 µm from the probe no change was detected. With 50% CO₂ in the dialysate, the pH change at the probe was approximately that which would occur with an end-tidal PCO₂ of 81 Torr. At 600 µm from the probe, no pH change was detected. With 25% CO₂ in the dialysate, the pH change at the probe was approximately that which would occur with an end-tidal PCO₂ of 63 Torr, and there was no detectable pH change at ~550 µm from the probe.

In these anesthetized rats, we also measured the amplitude of the integrated phrenic nerve signal and the frequency of phrenic bursts during the systemic CO₂ responses and during dialysis in the RTN region. Focal acidification had no effect on frequency and increased phrenic amplitude by ~20% of baseline with 25, 50, or 100% CO₂ in the dialysate. This increase was ~25% of the response, with all chemoreceptors stimulated by the change in end-tidal PCO₂ from 28 to 63 Torr (data not shown).

Responses to Dialysis With 100 and 50% CO₂ in Unanesthetized Rats

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Two representative responses are shown in unanesthetized rats with dialysis into retrotrapezoid nucleus (RTN) region of an aCSF solution equilibrated with 100% CO2. Ventilation (E) measured by whole body plethysmography is expressed as %baseline value. A: rat was initially asleep when dialysis was begun and E decreased. After 2 min of dialysis, rat awoke and E increased by 70% of baseline. When dialysis was turned off, E decreased to values of 15% of baseline. B: slightly different response pattern. While asleep, rat did increase E by 20% of baseline during first 5 min of dialysis. Rat then awoke, and E increased abruptly to 95% of baseline during next 3 min of dialysis. With cessation of dialysis, E decreased, reaching value of 5% of baseline 7 min later. Insets: computer-modified images of partial cross section of medulla at level corresponding to location of dialysis probe. Rectangle in each image shows region of tissue disruption produced by probe tip. VII, facial nucleus; 7N, facial nerve.

These preliminary data suggested that the ventilatory response to focal acidification of the RTN depended on the state of arousal, but the stimulus intensity and spread was large and seemed to arouse the rats on some occasions. Our next step was to lower the amount of CO₂ in the dialysate to see if a ventilatory response could still be detected. We used 50% CO₂ in the dialysate in three trials of 20-min dialysis in two rats and found that E increased in all three by an average of 20% (data not shown). These studies were performed only during wakefulness. The anatomic locations of the dialysis probes for the three rats that received 100% CO₂ dialysis and the two rats that received 50% CO₂ dialysis were in the RTN (Fig. 3). The spread of tissue pH changes measured in the anesthetized rats indicates that regions adjacent or contiguous to the RTN might have been acidified during dialysis with 50 or 100% CO₂.

View larger version (27K): [in this window] [in a new window]   Fig. 3.   Schematized (24) anatomic cross sections show location of dialysis probe tips (solid rectangles) for the 3 animals receiving 100% CO2 equilibrated dialysate (A) and the 2 animals receiving 50% CO2 equilibrated dialysate (B). Nos. below cross section refer to millimeters caudal to bregma. LC, locus ceruleus; NTS, nucleus tractus solitarius; 7, facial nerve. Bar = 1 mm.

In the major series of experiments of this report, we decreased the stimulus intensity to 25% CO₂ in the dialysate, thus producing a stimulus at the center that was similar in intensity to that observed with 9% endtidal CO₂ but focal in nature with the stimulus intensity rapidly decreasing with radial distance from the dialysis probe. We conducted a series of experiments measuring body temperature by telemetry, O₂, and E using the whole body plethysmograph during exposure to 7% inspired CO₂, as well as during focal dialysis of the RTN with 25% CO₂ equilibration of the dialysate.

Responses to Dialysis of the RTN Region With 25% CO₂ in Unanesthetized Rats

View larger version (27K): [in this window] [in a new window]   Fig. 4.   Schematized (24) anatomic cross sections show location of dialysis probe tips (solid rectangles) for animals receiving 25% CO2 equilibrated dialysate. Solid rectangles show, in 7 animals, locations that were judged to be within RTN region. Open rectangles show, in 2 animals, locations that were judged to be outside RTN region. These 2 animals had no response to dialysis. Nos. below cross section refer to millimeters caudal to bregma. Bar = 1 mm.

Five rats with probes located in the RTN were tested for their response to inhalation of 7% CO₂ both during sleep and during wakefulness (two rats did not undergo the systemic 7% CO₂ stimulation). Body temperature data, obtained via telemetry from the rat during the time it was in the plethysmograph, are shown in Fig. 5A. Body temperature was significantly lower during sleep (P < 0.05, two-way ANOVA), and, with 7% CO₂ inhalation, body temperature decreased significantly (P < 0.05, one-way ANOVA) in both the sleep and the awake tests. There was no significant effect of sleep vs. wakefulness on O₂ (Fig. 5B), although O₂ did tend to decrease during the exposure to 7% CO₂ (P < 0.05, one-way ANOVA). E expressed in absolute values (Fig. 6A) was slightly but not significantly lower in room-air breathing during sleep, but with 7% CO₂ inhalation it increased less during sleep than during wakefulness (Fig. 6A), a difference that was significant. However, when E was expressed as a percentage of baseline (Fig. 6B), there was no difference between the response to increased CO₂ during sleep vs. wakefulness. When the E was normalized for O₂ during CO₂ inhalation, again there was no significant difference between the responses during sleep and wakefulness, although the values during sleep were lower than during wakefulness (data not shown). The increase in E with 7% CO₂ inhalation was made up of a roughly 50% increase in tidal volume and an 80% increase in frequency. During sleep, tidal volume increased slightly more in absolute terms, but the value before CO₂ stimulation was lower, as was E. Frequency showed similar changes in absolute and relative-to-baseline terms.

View larger version (18K): [in this window] [in a new window]   Fig. 5.   Body temperature (A) and whole body O2 consumption (O2; B) in unanesthetized rats (n = 5) exposed to 7% CO2 during wakefulness () and behaviorally defined sleep (). Mean ± SE values are shown. Four control room air values were obtained before and after 25-min period of exposure to 7% CO2. Body temperature was significantly lower during sleep than during wakefulness, and both body temperature and O2 decreased during CO2 exposure.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   E in absolute terms (A) and as %baseline (B) in unanesthetized rats (n =5) exposed to 7% CO2 during wakefulness () and behaviorally defined sleep (). Mean ± SE values are shown. Four control room air values were obtained before and after 25-min period of exposure to 7% CO2. Absolute E was significantly lower during sleep than during wakefulness, but, when expressed as %baseline, there was no difference.

For dialysis of 25% CO₂ into the RTN region, there were, in the seven rats, 13 trials during wakefulness and 10 trials during sleep. As a control, there were 7 trials in five rats with RTN dialysis with the use of a 5% CO₂ equilibration. With 5% CO₂ dialysis, there was no significant change in body temperature, O₂, or E. These data are not shown.

The effects of focal acidification of the RTN on body temperature are shown in Fig. 7A. During sleep, body temperature was significantly decreased, but there was no effect of RTN dialysis with 25% CO₂ on body temperature. O₂ (Fig. 7B) was unaffected by sleep or dialysis.

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Body temperature (A) and whole body O2 (B) in unanesthetized rats (n = 7) dialyzed with 25% CO2 in RTN during wakefulness () and behaviorally defined sleep (). Mean ± SE values are shown. Four control room air values were obtained before and after 20-min period of exposure to 25% CO2. Body temperature was significantly lower during sleep than during wakefulness, and neither body temperature nor O2 decreased during CO2 exposure localized to RTN.

The major findings of the study are shown in Fig. 8. During dialysis, E expressed in absolute terms (Fig. 8A) or in relative terms (Fig. 8B) increased, on average, only during wakefulness. One-way ANOVA showed a significant increase (P < 0.001; 17, 19, and 24% increase at 5, 10, and 15 min) in E in absolute or relative terms only during wakefulness. Two-way ANOVA showed a significant response of E, expressed in absolute terms or as percent baseline (P < 0.001), to CO₂ during wakefulness compared with sleep. This average ventilatory response to focal RTN acidification during wakefulness was entirely due to an increase in tidal volume (P < 0.01; Fig. 9). Frequency did not change. Normalizing the E data for O₂ did not change the nature of these results. Of the 13 dialysis trials during wakefulness, 9 could be classified as a "response" by using an arbitrary definition of an increase in E of >10%. With this definition, there was a response in only 1 of 10 trials during behavioral sleep and in 1 of 7 trials with the 5% CO₂ dialysis control.

View larger version (15K): [in this window] [in a new window]   Fig. 8.   E in absolute terms (A) and expressed as %baseline (B) in unanesthetized rats (n = 7) dialyzed with 25% CO2 in RTN region during wakefulness (; n = 13 trials) and behaviorally defined sleep (; n = 10 trials). Mean ± SE values are shown. Control room air values were obtained before and after 20-min period of dialysis. The 4 preexposure control values were combined into single value; 25-min postexposure value was deleted, as many rats showed an artifactual increase in E because of manipulation of dialysis apparatus when it was shut off. E during focal RTN acidification was significantly greater during wakefulness when expressed in absolute terms or as %baseline. Note that E increased to 24% of baseline. There was no response during sleep.

View larger version (14K): [in this window] [in a new window]   Fig. 9.   Tidal volume (VT; A) and respiratory frequency (B) in unanesthetized rats (n = 7) dialyzed with 25% CO2 in RTN region during wakefulness (; n = 13 trials) and behaviorally defined sleep (; n = 10 trials). Mean ± SE values are shown. Control room air values were obtained before and after 20-min period of dialysis. The 4 preexposure control values were combined into single value; 25-min postexposure value was deleted, as many rats showed an artifactual increase in E because of manipulation of dialysis apparatus when it was shut off. VT during focal RTN acidification was significantly greater during wakefulness when expressed in absolute terms or as %baseline. There was no response during sleep.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Technical Issues

There are two concerns with this model: 1) the definition of sleep vs. wakefulness (15), and 2) the intensity and degree of spread of the tissue acidosis produced by the dialyzed CO₂. For these initial studies, we defined sleep by the use of straightforward behavioral criteria. Rats were judged to be asleep when curled up motionless with their eyes closed. All other states were included during wakefulness. We are aware that active sleep (AS) and quiet sleep (QS) are very different states (15) and plan to add electroencephalogram and electromyogram measures to this model to describe more accurately wakefulness and sleep and AS vs. QS. Our behavioral definition is a reasonable beginning for these studies. Rats generally have frequent AS episodes of brief duration such that one would expect that our behaviorally defined sleep periods would consist predominantly of QS (15).

With respect to the tissue acidosis, we measured the brain stem tissue pH at varying distances from the dialysis probe using pH electrodes in anesthetized rats (Fig. 1) (5, 13, 19). Ideally we would do this in the unanesthetized model during sleep and wakefulness, but this is a technically difficult problem. For now, we rely on these data obtained under anesthesia. With dialysis by using aCSF equilibrated with 100% CO₂, the observed pH change is constrained to within 750 µm and, at the probe, is approximately equivalent to that observed with an end-tidal PCO₂ of 110 Torr. The affected tissue volume is 1.4 µl (assuming a spherical volume of pH distribution with a radius of 750); the volume of the RTN region in the rat is ~1.12 µl (assuming a width of 1.4 mm from the pyramidal tract to the lateral aspect of the facial nucleus, a depth of 0.5 mm from the ventral medullary surface to the ventral aspect of the facial nucleus, and a length of 1.6 mm from the rostral to the caudal poles of the facial nucleus) (24, 30). The tissue volume with an acid pH would include structures in the RVLM contiguous to the RTN, e.g., the facial nucleus, parapyramidal and juxtafacial nucleus, paragigantocellularis lateralis, and perhaps dendrites from subretrofacial and retrofacial neurons (1, 17). Even with this large stimulus, both in terms of intensity and spread, the ventilatory response to this focal acidosis is present only during wakefulness or is substantially greater during wakefulness than during sleep.

With dialysis by using aCSF equilibrated with 25% CO₂, the pH change at the probe is approximately equivalent to that observed with an end-tidal CO₂ of 63 Torr, and it decreases with distance such that, at 550 µm from the probe, there is no detectable change. The volume of affected tissue here is 700 nl. Most, if not all, of the focal acidosis with 25% CO₂ in the dialysate is limited to the RTN region and the facial nucleus.

These estimates of the stimulus intensity and spread made in the anesthetized animal are probably greater than that which occurred in the unanesthetized preparation. In the absence of anesthesia, it is likely that the response of cerebral blood flow to the focal acidosis is greater (31), which would decrease the intensity and spread of the acidosis.

Location of Dialysis Probes

Body Temperature and Metabolic Rate Responses to Sleep, 7% CO₂ Inhalation, and Focal RTN Acidification

Blood Flow During Sleep and With CO₂

The rostral pressor region (8), located near the RTN, when stimulated can increase blood pressure via sympathetic efferent stimulation. We did not measure blood pressure in these unanesthetized rat experiments, and it is possible that in some cases it may have been stimulated by the focal acidosis. Blood pressure did increase in some animals with focal RTN CO₂ application in anesthesia (13).

Chemoreception in the RTN During Wakefulness, Sleep, and Anesthesia

These findings at first glance seem to be a paradox. During sleep, focal acidification of the RTN has no effect, in anesthesia the effect is a large fraction of that observed with all chemoreceptors exposed to CO₂, and during wakefulness the effect is significant but is a relatively small percentage of the effect observed with all central chemoreceptors stimulated. Anesthesia (with chloralose-urethan) and sleep clearly differ in their effects on chemoreception. Our interpretation is that anesthesia affects other central chemoreceptor locations more than it does the RTN region; the RTN becomes an important source of input to the respiratory control system related to CO₂, perhaps the most important. This would explain why, in anesthetized animals, lesions of the RTN region virtually abolish central chemosensitivity and can result in apnea.

In the unanesthetized and awake state, the ventilatory response to focal acidification of the RTN region is but a small fraction of the response to acidification of all central chemoreceptors. All chemoreceptor sites are operational in this state, and the overall response is likely tempered by hypocapnia at other sites. This interpretation is consistent with recent results showing that lesions or cooling of the RTN region in the unanesthetized, awake rat or goat does not have the same dramatic effects on baseline E or chemosensitivity as it does when performed with the animal under anesthesia. In the unanesthetized animal with RTN disruption, baseline E is unaffected, or is minimally affected, and the CO₂ response is reduced but not abolished (1, 2, 9, 20, 21). In the unanesthetized, awake state, the RTN chemoreceptor site seems to act as one of many sites, each providing a reasonable but not exceptionally large fraction of the total central chemoreceptor input.

Surprisingly, focal acidification of the RTN region during sleep has no effect on E. Although we raised the possibility above that this lack of response could possibly be due to a sleep-related, enhanced cerebral blood flow response to focal acidosis of the RTN region during sleep, on close examination of existing data this seems to be unlikely. Instead, we are left with the intriguing explanations that during sleep either the RTN chemoreceptors are no longer functioning or the brain stem respiratory control system is not listening to the input from this source, or, if it is, the gain of the response is curtailed. From the data obtained with systemic 7% CO₂ stimulation in these same rats during behaviorally defined wakefulness and sleep, we know that there is a ventilatory response when all chemoreceptors are exposed to the stimulus. Thus the absence of any response to focal acidification of the RTN region must represent a specific effect in the RTN. We suggest that different central chemoreceptor sites may play different roles in the control of breathing, depending on the state of arousal of brain stem neurons. Recent in vitro studies show the presence of neurons responsive to acidification in the midline raphe (27) and in the locus ceruleus (26), sites that, when acidified focally in the anesthetized rat, stimulate ventilatory output (4, 5). These regions have not been traditionally identified as important in the control of breathing but are important in arousal-related functions of the brain stem. We hypothesize that their role in the control of breathing may be enhanced during sleep.