INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

AVIAN INTRAPULMONARY chemoreceptors (IPC) are CO₂-sensitive vagal afferents that innervate the lungs of birds. The rate of action potential generation by IPC is inversely proportional to intrapulmonary PCO₂, which varies during the breathing cycle (4, 9). IPC provide a sensory feedback path that has been implicated in the regulation of breathing pattern and arterial PCO₂ (2, 8, 21) and the adjustment of ventilation to metabolic demands (2, 10). Despite the clear phasic discharge of IPC with tidal breathing and its strong inverse dependence on lung PCO₂, the CO₂ transduction mechanisms of IPC are not well understood.

CO₂ transduction in IPC is dependent on the enzyme carbonic anhydrase (CA), which catalyzes the hydration-dehydration reaction between CO₂, H₂O, and H₂CO₃. Systemic CA blockade with the cell-permeable inhibitor acetazolamide increases IPC discharge to near maximal levels and makes IPC insensitive to increasing PCO₂ (16, 19). Although the complete mechanism of CO₂ transduction in IPC remains unclear, the action of acetazolamide on IPC suggests that H⁺ from the dissociation of hydrated CO₂, rather than molecular CO₂, is the cellular stimulus for chemotransduction. However, acetazolamide is freely permeable to cell membranes and its actions do not distinguish between intracellular and extracellular sites for catalysis by CA and detection of H⁺.

Other lines of evidence support the idea of an intracellular site for chemotransduction. IPC are much more sensitive to pH changes in pulmonary arterial blood when it is acidified with CO₂ than when it is acidified with HCl (5). CO₂ gas should traverse cell membranes of sensory endings more easily than HCl (20), and CO₂ may be hydrated intracellularly to yield H⁺ for chemotransduction. Therefore, these results are consistent with transduction of intracellular H⁺ levels altered by CO₂ hydration or with transduction of molecular CO₂ directly. However, transduction of molecular CO₂ by IPC appears unlikely because of the nearly complete loss of CO₂ sensitivity with acetazolamide (16, 19). Transduction of extracellular H⁺ also seems unlikely because HCl infusion has little consistent effect on IPC (5). Transduction of intracellular H⁺ remains the most viable hypothesis and is further supported by the observation that insufflation of the soluble, hydratable acidic gas SO₂ mimics the effects of CO₂ on IPC discharge (6).

If H⁺ controls IPC excitability, interventions that alter pH buffering of PCO₂ changes should affect chemotransduction. Specifically, metabolic acidosis should inhibit IPC discharge at any given PCO₂ level by depleting bicarbonate and increasing H⁺ concentration ([H⁺]), and metabolic alkalosis should excite IPC discharge by increasing bicarbonate and decreasing [H⁺]. In fact, chronic metabolic acidosis induced by ammonium chloride ingestion inhibits IPC discharge and reduces IPC sensitivity to PCO₂ as would be expected from the H⁺ transduction model (3). Conversely, acute metabolic alkalosis produced by intravenous infusion of 1.2 M NaHCO₃ modestly stimulates IPC discharge (1, 17), as would be expected from the H⁺ model, but also reduces IPC responsiveness to PCO₂ (1). It has been suggested that interpretation of the effects of acute 1.2 M NaHCO₃ infusion on IPC may be complicated by transient ionic disequilibrium between extracellular and intracellular compartments and the large sodium load administered with the bicarbonate (3).

Metabolic alkalosis induced by chronic hypercapnia offers an effective physiological alternative for increasing bicarbonate levels and testing the H⁺ transduction model without the transient side effects of acute 1.2 M NaHCO₃ infusion. CO₂ is freely permeable to cells (20) and chronic hypercapnia should produce a stable compensatory intracellular and extracellular metabolic alkalosis that is proportional to the magnitude of hypercapnia (7). Therefore, to further test the hypothesis that IPC transduce H⁺, we acclimatized birds to 7.5% inspired CO₂ for several weeks and compared their IPC responses to unacclimatized controls. We reasoned that bicarbonate production and retention in acclimatized birds should increase pH at all PCO₂ values relative to controls and allow a test for independent effects of H⁺ and CO₂ on chemotransduction. If average IPC discharge at a given PCO₂ were unchanged by acclimatization, even though pH were more alkaline, this would suggest that IPC transduce CO₂ directly. If average IPC discharge at a given PCO₂ was increased after acclimatization, this would suggest that IPC transduce H⁺ and are stimulated by the increased alkalinity. If average IPC discharge at a given PCO₂ were lower after acclimatization, this would not fit predictions of either the H⁺ or CO₂ chemotransduction models and would indicate operation of other mechanisms.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Animals and acclimatization. Fifteen adult Pekin ducks, Anas platyrhynchos, 4-8 mo old, of either sex were studied. Nine ducks [body mass 2.5 ± 0.1 kg (means ± SE)] were kept under normal vivarium conditions breathing room air [0.02% CO₂, 0.1 Torr inspiratory PCO₂ (PI_CO2)] with food and water ad libitum. Six other ducks (body mass 2.3 ± 0.1 kg) were acclimatized for 12 ± 1.2 days in a 700-liter Plexiglas environmental chamber ventilated with 20 l/min of 7.5% CO₂ in air. Chamber PI_CO2 was monitored with a Beckman LB-2 CO₂ analyzer and ranged from 45 to 55 Torr. Ducks were removed from the acclimatization chamber for 20 min daily to clean the bedding material and replenish food and water and were maintained in otherwise normal vivarium conditions. Animal husbandry and experimental procedures were reviewed and approved by the local Institutional Animal Care and Use Committee.

Reflex ventilatory measurements. Awake ventilatory responses to inspired CO₂ were studied before and after chronic CO₂ exposure (n = 6 ducks) by whole body plethysmography. Animals were placed in a 14-liter sealed chamber that was ventilated through high-resistance inlet and outlet valves. Air with or without added CO₂ was admitted at 20 l/min through the inlet and removed at the same rate through the outlet connection to a vacuum source. PI_CO2 was measured with a Perkin Elmer MGA-1100 mass spectrometer. Mean chamber pressure was measured with a water manometer and set equal to atmospheric pressure by balancing the high-resistance valves. Tidal fluctuations in chamber pressure associated with heating and humidification of the inspired gas were measured with a Validyne MP-45 transducer and demodulator. Body and chamber temperatures were monitored with Yellow Springs Instruments thermistor probes, relative chamber humidity was monitored with an analog hygrometer, and temperature and pressure signals were recorded on a Gould pen writer using universal couplers.

Animals were allowed to relax in the darkened plethysmographic chamber until ventilation became regular and exploratory movements ceased. Inspired CO₂ levels of 0, 2, 4, and 6% were then given in random order for ~5-10 min each, and ventilatory pressure traces, temperatures, and humidity were recorded. The chamber was calibrated after each measurement by injection of a known volume of air with a syringe. Tidal pressure traces were converted to tidal volume (12, 13), and minute ventilation was calculated from respiratory frequency and tidal volume.

Minute ventilatory response was analyzed using two-way ANOVA. Main effects were inspired CO₂ test levels (0, 2, 4, and 6%) and acclimatization state (normal vs. acclimatized). Post hoc paired comparisons were made using the Scheffé method and least-squares means (SAS System GLM procedure), and P levels 0.05 were accepted as significant.

Neural recording. Animals were anesthetized by intravenous infusion of 35-40 mg/kg pentobarbital sodium. Supplemental doses were administered through an implanted brachial vein cannula to maintain deep surgical anesthesia (absent or minimal toe withdrawal response to strong pinch). Body temperature was monitored with a Yellow Springs Instruments colonic thermistor probe and regulated at 41 ± 1°C with a servocontrolled heat lamp and circulating hot water pad.

The thorax was opened with a sternal incision, and each lung was independently unidirectionally ventilated with 0.5 l/min of air mixed with 3% CO₂. Ventilation was delivered through Foley catheters inserted into each primary bronchus to a point just caudal to the ostia of the medioventral secondary bronchi. Gases for the right lung were mixed with a Matheson Rotameter. Gases for the left lung were mixed with a Cameron Instruments GF-1 mass flow controller. A loose umbilical tape snare was placed around the left pulmonary artery to allow reversible occlusion of blood flow.

The left vagus nerve was exposed in the neck and bathed in a mineral oil pool bounded by skin flaps. Single-cell recordings were made by dissecting the vagus and placing fine filaments on a 35-gauge platinum-iridium electrode. IPC electrical activity was referenced to an indifferent Ag-AgCl electrode on the nerve sheath through a high-impedance differential probe (Grass). Action potentials were amplified with a Grass P511K AC preamplifier and AM-5 audio amplifier, digitized with a Haer window discriminator, timed with a microcomputer, visualized on HP 130C and Tektronix 561A oscilloscopes, and recorded on a Vetter D FM tape system. A 60-Hz notch filter was engaged, and filters were set for bandpass between 100 and 3,000 Hz. IPCs were identified by their nearly immediate response to step changes in ventilatory gas PCO₂. Single cell IPC were identified by constancy of amplitude and shape of their action potentials.

When a single IPC was identified in the left vagus, right lung ventilation was adjusted to 2.5 l/min of 5% CO₂ in air to maintain gas exchange in the animal. The snare was tightened around the left pulmonary artery to redirect cardiac output to the right lung, thereby stabilizing the left lung's intrapulmonary PCO₂ at inspired levels (1, 2, 3, 5, 11, 15). Left lung ventilation was adjusted to 0.5 l/min of 50% O₂ balanced with N₂, and CO₂ levels were set at steady levels ranging from 2 to 11% using the mass flow controller. IPC discharge rate was allowed to stabilize at each PCO₂ level tested (usually 1 min) and was then recorded by computer for 10-30 s. After measurments of IPC discharge at each PCO₂ value, duplicate measurements were made at several PCO₂ stimulus levels to verify repeatability. The pulmonary artery ligature was then loosened, ventilation was returned to premeasurement values, and the search for IPCs was resumed.

Stimulus response relationships for each IPC were quantified by simple linear regression of discharge frequency vs. the natural logarithm of intrapulmonary PCO₂ (1, 3, 11, 15, 19). Mean values and SEs of the resultant slope, intercept, and regression r values were calculated for control and acclimatized groups. Comparisons of slopes and intercepts were made by unpaired t-test and P levels 0.05 were accepted as significant.

Changes in blood pH buffering in acclimatized vs. normal birds were quantified from anaerobic blood samples drawn into heparinized 1-ml syringes from a cannula in the right carotid artery. PCO₂ and plasma pH were analyzed on an Instrumentation Laboratories IL 813 blood gas system, and plasma bicarbonate was calculated using the Henderson-Hasselbalch equation (7). Data were pooled separately for normal and acclimatized birds, and blood buffer curves were determined using simple linear regressions of pH vs.log₁₀PCO₂. Regression coefficients for normal and acclimatized groups were compared with t-tests, and P levels 0.05 were considered significant.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Chronic CO₂ acclimatization caused a right shift of the ventilatory response curve to inspired CO₂ (Fig. 1). Two-way ANOVA on minute ventilation revealed effects of both PI_CO2 (P < 0.05) and acclimatization (P < 0.05) and an interaction between PI_CO2 and acclimatization (P < 0.05). Post hoc analysis showed that mean ventilations at 0 and 14 Torr PI_CO2 were unaffected by acclimatization (P > 0.3), but mean ventilations at 28 and 42 Torr PI_CO2 were reduced by acclimatization (P < 0.05, n = 6 animals; Fig. 1).

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Reflex CO2 sensitivity of awake ducks. Ducks acclimatized to 7.5% inspired CO2 (PICO2) for 12 days had a higher stimulation threshold (the PICO2 that increases ventilation above baseline) compared with normal ducks acclimatized to room air. Graphs show means ± SE. * Acclimatized different from normal (P < 0.05), n = 6.

Blood pH-log₁₀PCO₂ relationships of CO₂-acclimatized birds were elevated and right shifted compared with normal birds, indicating compensatory metabolic alkalosis and bicarbonate retention (Fig. 2). Linear regressions of the form pH = (slope) · log₁₀PCO₂ + (intercept) were correlated with r values of 0.93 for normal birds (n = 21 blood samples) and 0.97 with acclimatization (n = 25 blood samples). Intercept values (±SE) were increased in acclimatized (8.760 ± 0.068 pH) compared with normal animals (8.413 ± 0.098 pH, P < 0.05). Slope values (±SE) in acclimatized animals (0.751 ± 0.039 pH/log₁₀PCO₂) were not different from normal animals (0.645 ± 0.059 pH/log₁₀PCO₂, P > 0.05). Mean plasma pH at 40 Torr PCO₂ increased from 7.38 ± 0.03 (normal) to 7.56 ± 0.02 (acclimatized, P < 0.05), and mean plasma bicarbonate increased from 22.9 ± 1.6 (normal) to 34.6 ± 1.5 mM (acclimatized, P < 0.05).

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Blood buffer curves showing the relationship between plasma pH and PCO2 in normal ducks acclimatized to room air and acclimatized ducks after 12 days breathing 7.5% PCO2 in air. Semilogarithmic regression lines and 95% confidence intervals show extent of metabolic compensation for chronic respiratory acidosis.

Individual stimulus-response relationships for 22 normal and 34 CO₂-acclimatized IPC are shown in Fig. 3. As a group, discharge of CO₂-acclimatized IPC was elevated relative to normals. Note that the lowest PCO₂ administered to acclimatized IPC was 21 Torr, while the lowest administered to normal IPC was 14 Torr. Earlier studies have shown that normal IPC sometimes discharge erratically and appear impaired with sustained PI_CO2 values <7-14 Torr in nonperfused lungs (1, 11, 15). Similarly, we found that IPC acclimatized to high PCO₂ sometimes discharged erratically <14-21 Torr PI_CO2, perhaps due to increased receptive site pH accompanying metabolic compensation. We therefore did not expose acclimatized animals to PI_CO2 values <21 Torr.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   A: Individual stimulus-response curves showing action potential discharge frequency vs. intrapulmonary PCO2 for 22 intrapulmonary chemoreceptors (IPC) recorded in normal ducks acclimatized to room air. B: response curves for 34 IPC recorded in chronic CO2-acclimatized ducks exposed to 7.5% inspired CO2 for 12 days. IPC discharge is right shifted after acclimatization.

Coefficients from the individual regressions of IPC discharge frequency vs. the natural logarithm of PCO₂, f_IPC = (slope) · lnPCO₂ + intercept, were averaged for the 22 normal and 34 acclimatized IPC to estimate population responses. Mean slope values (±SE) were greater for acclimatized IPC [19.0 ± 1.0 impulses · s¹ · lnPCO₂¹] than for normal IPC [12.0 ± 1.1 impulses · s¹ · lnPCO₂¹, P < 0.05]. Mean intercept values (±SE) were also greater for acclimatized IPC (81.1 ± 4.0 impulses/s) than for normal IPC (48.4 ± 3.6 impulses/s, P < 0.05). Average correlation coefficients indicated excellent fit to the logarithmic stimulus-response model for both normal (r = 0.994 ± 0.001) and acclimatized (r = 0.992 ± 0.001) IPC.

Regression relationships for each individual IPC were used to interpolate discharge frequency at four standard PCO₂ levels: 14, 21, 35, and 56 Torr. Interpolated discharge rates were then averaged at each PCO₂ level for normal and acclimatized groups, yielding the mean stimulus response curves shown in Fig. 4B. Mean IPC discharge at 14 Torr for acclimatized animals is shown with a different symbol and a dotted line to indicate it was extrapolated rather than interpolated. Two-way ANOVA revealed significant effects of intrapulmonary PCO₂ (P < 0.05), acclimatization state (P < 0.05), and PCO₂-acclimatization interaction (P < 0.05). Significant interaction is consistent with the increased slope of IPC responses to PCO₂ after acclimatization as shown in regression analysis above. Post hoc analysis using methods of Scheffé and least-squares means showed that IPC discharge was higher in acclimatized vs. normal animals at all PCO₂ levels (P < 0.05).

View larger version (18K): [in this window] [in a new window]   Fig. 4.   A: mean discharge rates (means ± SE) for normal IPC and CO2-acclimatized IPC as a function of plasma pH predicted from PCO2 using the relationship in Fig. 2. Discharge rate vs. pH curves for normal and acclimatized IPC overlap each other between pH 7.55 and 7.67 (P < 0.05). Data suggest that IPC discharge rate may be a unique function of pH, not of PCO2. B: mean IPC discharge rates (means ± SE) as a function of intrapulmonary PCO2 in normal ducks () and CO2-acclimatized ducks (). The 14 Torr point for CO2-acclimatized IPC was extrapolated from regression equations (see RESULTS). CO2-acclimatized IPC (n = 34) had right-shifted and steeper responses to CO2 compared with normal IPC (n = 22). * Differences between acclimatized and normal values, P < 0.05.

Figure 4A shows mean IPC discharge frequencies as a function of plasma pH associated with 14, 21, 35, and 56 Torr PCO₂. Plasma pH was calculated from the regression relationships between pH and log₁₀PCO₂ given above. Figure 4A reveals that normal and acclimatized IPC responses to pH were not different over the pH range of 7.55 to 7.67 (P > 0.05), in contrast to normal and acclimatized IPC responses to PCO₂, which were different at all PCO₂ studied (Fig. 4B, P < 0.05). Despite the overlap of IPC responses to pH, linear regression analysis of IPC discharge frequency vs. plasma pH, f_IPC = (slope) · (pH) + intercept, revealed larger slope (57.7 ± 0.3 vs. 39.3 ± 2.1 impulses · s¹ · pH¹, P < 0.05) and larger intercept (425 ± 2 vs. 285 ± 15 impulses/s, P < 0.05) values in acclimatized IPC. Correlation between f_IPC and plasma pH was high in both normal and acclimatized groups (0.997 vs. 0.999).

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

Acclimatization model. Birds breathing 7.5% inspired CO₂ for 12 days showed clear acclimatization. Their ventilatory responses had an increased threshold to CO₂ stimulation, as noted in other chronically hypercapnic species (22). Their blood PCO₂-pH relationships were right shifted, suggesting metabolic compensation for respiratory acidosis (7) and mean IPC discharge rates were increased at all intrapulmonary PCO₂ levels studied. Because the changes were chronic, the shifted pH-log₁₀PCO₂ relationship in acclimatized animals provided a stable method for estimating the separate effects of pH and PCO₂ on IPC chemotransduction.

H⁺ vs. CO₂ signal transduction. The different IPC responses to PCO₂ in normal and acclimatized animals showed that IPC discharge is not a unique function of intrapulmonary PCO₂ (Fig. 4B). After acclimatization, both the slope and the intercept of the IPC response to PCO₂ were increased, increasing IPC discharge rate relative to normal at all PCO₂ values tested. This suggests that IPC do not transduce CO₂ changes directly or that CO₂ transduction is strongly modulated by other stimuli. Conversely, the overlapping curves of IPC discharge vs. pH for normal and acclimatized animals (Fig. 4A) suggest that IPC actually do transduce H⁺ changes caused by CO₂ hydration. Although IPC response curves to pH overlapped between pH 7.55 and 7.67, the slope of the IPC response was steeper after acclimatization. This may reflect a fundamentally curvilinear relationship between IPC membrane excitability and pH or perhaps a change in buffering, chemical environment, or gene expression that modifies chemotransduction after acclimatization.

Intracellular vs. extracellular pH. We could not measure intracellular pH directly because IPC endings are inaccessible in the lung parenchyma. However, changes in PCO₂ are extremely effective in changing both intracellular and extracellular pH due to high permeability of CO₂ and generation of H⁺ and HCO₃ from CO₂ hydration (7, 20). Furthermore, most cells in steady-state acid base balance have a 0.3-0.4 pH unit differential between the intracellular and extracellular spaces, with the intracellular space more acid (20). Because 2 wk of hypercapnia produced a steady-state compensatory metabolic alkalosis, we reasoned that changes in intracellular pH should track changes in plasma pH but be ~0.3-0.4 pH units lower. Although our results indicate that IPC transduce mainly H⁺ rather than CO₂, we cannot differentiate between extracellular or intracellular sites of H⁺ transduction using our data alone.

Comparisons with other studies. Chronic metabolic acidosis in chickens depresses IPC discharge at low PCO₂ and reduces the slope and intercept of IPC discharge vs. PCO₂ (3). Our results showing increased IPC discharge and increased slope and intercept of IPC discharge vs. PCO₂ are consistent with this earlier study. Because metabolic acidosis and metabolic alkalosis produce opposite changes in bicarbonate and pH levels, they should have opposite effects on IPC chemotransduction if pH rather than PCO₂ is transduced by IPC. The mechanism for changes in slope and intercept of IPC response to PCO₂ during metabolic acid-base disturbances remains to be determined. However, increased slope may indicate weaker intracellular pH buffering of PCO₂ changes in metabolic alkalosis compared with acidosis, and increased intercept may reflect chronically elevated intracellular bicarbonate and increased pH stimulation at low PCO₂ in metabolic alkalosis compared with acidosis.

Similar to the chronic metabolic alkalosis studied here, acute metabolic alkalosis induced by sodium bicarbonate infusion in ducks also increases IPC discharge (1, 17). However, acute bicarbonate infusion causes smaller increases in IPC discharge rate, and unlike the effect of chronic metabolic alkalosis reported here, acute bicarbonate infusion causes a small decrease in slope of IPC responses to PCO₂ (1). These somewhat different effects of acute and chronic metabolic alkalosis on IPC may reflect transient ionic imbalances related to infusion of 1.2 M sodium bicarbonate (3) or relatively smaller bicarbonate changes in intracellular vs. extracellular compartments with acute bicarbonate infusion. Although chronic metabolic alkalosis used here takes longer to produce than acute metabolic alkalosis, it has the advantage of minimizing transient ionic imbalances and producing stable acid-base changes in intracellular and extracellular compartments (3, 7).

Acclimatization vs. evolutionary adaptation to hypercapnia. An earlier study investigated ventilatory responses and IPC discharge in burrowing owls, animals that are genetically adapted to subterranean life and high inspired CO₂ levels. Ventilatory CO₂ responses of burrowing owls are blunted (14) like those of CO₂-acclimatized ducks in this study. However, burrowing owls also have blunted IPC responses to PCO₂ (14) relative to most other birds studied (mean slope 6.9 ± 0.4 impulses · s¹ · lnPCO₂¹), unlike the increased mean IPC-PCO₂ response slope seen in acclimatized ducks in this study. These contrasting observations probably reflect fundamental differences between genetic adaptation and physiological acclimatization to hypercapnia. Whereas acclimatization invokes the physiological plasticity available with an animal's existing genetic makeup, adaptation results from generations of selection for traits particularly suited for a given lifestyle. The end results may be quite different.

Perspectives

Avian IPC are exquisitely sensitive CO₂ chemoreceptors that are inhibited rather than excited by increased PCO₂. Although IPC sensory endings have yet to be studied using intracellular techniques, their afferent activity is easily recorded from vagal filaments, and their in situ microenvironment is easily controlled with pulmonary ventilation and perfusion. Continued study of avian IPC may reveal fundamental aspects of CO₂ chemoreception and should provide a useful comparative contrast to mammalian central and peripheral CO₂ chemoreceptors.