Brainstem P_CO2 modulates phrenic responses to specific carotid body hypoxia in an in situ dual perfused rat preparation

Abstract

Inputs from central (brainstem) and peripheral (carotid body) respiratory chemoreceptors are coordinated to protect blood gases against potentially deleterious fluctuations. However, the mathematics of the steady-state interaction between chemoreceptors has been difficult to ascertain. Further, how this interaction affects time-dependent phenomena (in which chemoresponses depend upon previous experience) is largely unknown. To determine how central P_CO2 modulates the response to peripheral chemostimulation in the rat, we utilized an in situ arterially perfused, vagotomized, decerebrate preparation, in which central and peripheral chemoreceptors were perfused separately (i.e. dual perfused preparation (DPP)). We carried out two sets of experiments: in Experiment 1, we alternated steady-state brainstem P_CO2 between 25 and 50 Torr in each preparation, and applied specific carotid body hypoxia (60 Torr P_O2 and 40 Torr P_CO2) under both conditions; in Experiment 2, we applied four 5 min bouts (separated by 5 min) of specific carotid body hypoxia (60 Torr P_O2 and 40 Torr P_CO2) while holding the brainstem at either 30 Torr or 50 Torr P_CO2. We demonstrate that the level of brainstem P_CO2 modulates (a) the magnitude of the phrenic responses to a single step of specific carotid body hypoxia and (b) the magnitude of time-dependent phenomena. We report that the interaction between chemoreceptors is negative (i.e. hypo-additive), whereby a lower brainstem P_CO2 augments phrenic responses resulting from specific carotid body hypoxia. A negative interaction may underlie the pathophysiology of central sleep apnoea in populations that are chronically hypocapnic.

The chemical-dependent regulation of blood gasses is co-ordinated through central and peripheral respiratory chemoreceptor feedback loops (e.g. Cunningham et al. 1986; Topor et al. 2004). The central respiratory chemoreceptors are located within the brainstem and respond to changes in tissue pH/CO₂ (Nattie, 2000). The peripheral (carotid body) respiratory chemoreceptors, located at the bifurcation of the common carotid arteries, respond to changes in both arterial P_O2 and P_CO2 (Lahiri & DeLaney, 1975). Recently, considerable advances have been made toward locating and characterizing the phenotype of putative central chemoreceptors (Guyenet et al. 2005; Richerson et al. 2005) and determining the signal transduction mechanism involved in peripheral chemosensitivity (Lahiri et al. 2006; Prabhakar, 2006). Rodent preparations have been used extensively in these studies. However, the fundamental question as to how central and peripheral chemoreceptors interact to produce ventilatory responses remains contentious, and testing this interaction experimentally in small mammals, such as rodents, has not been attempted.

In non-rodent preparations, the mathematics of the interaction between central and peripheral respiratory chemoreceptors has been reported to be: (a) additive, where the contribution of the two receptor compartments do not interact in any significant way (van Beek et al. 1983; Daristotle & Bisgard, 1989; Clement et al. 1992, 1995; St. Croix et al. 1996;); (b) positive (i.e. hyper-additive; Adams et al. 1978; Robbins, 1988); and (c) negative (i.e. hypo-additive; Gesell et al. 1940; Giese et al. 1978; Berger et al. 1978; Eldridge et al. 1981; Adams & Severns, 1982; Smith et al. 1984).

The uncertainty regarding the nature of central and peripheral chemoreceptor interaction is further confounded by ventilatory time-dependent phenomena, which influence the dynamics of chemoresponses relative to previous experience. Numerous time-dependent responses to chemostimuli have been described (see Powell et al. 1998 for review), including hypoxic ventilatory decline (HVD) (Robbins, 1995), short-term potentiation or ventilatory after-discharge (STP or VAD, respectively) (Eldridge & Millhorn, 1986; Engwall et al. 1991), post-hypoxic frequency decline (PHxFD) (Hayashi et al. 1993; Coles & Dick, 1996; Dick & Coles, 2000; Day & Wilson, 2005), progressive augmentation (PA) (Millhorn et al. 1980; Fregosi & Mitchell, 1994; Turner & Mitchell, 1997; Day & Wilson, 2005) and long-term facilitation (LTF) (Baker & Mitchell, 2000; Olson et al. 2001). These time-dependent phenomena are likely to be important in the stability of breathing, and their magnitude may depend on chemoreceptor interaction.

To facilitate the study of central and peripheral chemoreceptor interaction we developed a novel in situ dual perfused rat preparation (DPP) based on the working heart–brainstem preparation (WHBP) (Paton, 1996). The WHBP has a well-oxygenated brainstem (>200 Torr P_O2), a pH that matches the perfusate, intact vascular responses to P_O2 and P_CO2, and a eupnoeic-like ramping phrenic discharge (Wilson et al. 2001). In the DPP, central (brainstem) and peripheral (carotid body) compartments are perfused separately, allowing compartment-specific chemochallenges with defined medium containing precisely controlled gas concentrations. We have previously demonstrated (a) the independence of peripheral and central perfusion circuits, and (b) that specific carotid body intermittent hypoxia is sufficient to elicit a number of time-dependent phenomena (Day & Wilson, 2005).

In the present study, we tested whether the level of steady-state brainstem P_CO2 modulates components of the phrenic response resulting from specific carotid body hypoxia. Specifically, we investigated phrenic responses to single and intermittent bouts of carotid body hypoxia under differing levels of steady-state brainstem P_CO2. Here we provide data to demonstrate that the level of brainstem P_CO2 modulates (a) the magnitude of the phrenic responses to a single bout of specific isocapnic carotid body hypoxia (60 Torr P_O2 and 40 Torr P_CO2), and (b) the magnitude of time-dependent phenomena resulting from intermittent isocapnic hypoxia.

Methods

In situ DPP

We modified the WHBP (Paton, 1996), making it appropriate for studies of respiratory chemoreceptor interaction. We will refer to our preparation as the in situ DPP (Day & Wilson, 2005). Our modifications include: (a) separate perfusion of central and peripheral chemoreceptor compartments, (b) clamping of descending aorta perfusion pressures at ∼90 mmHg via computer feedback control, (c) vagotomy, and (d) normokalaemic perfusate.

Dissection and perfusion system

Experiments were conducted using juvenile male Sprague-Dawley albino rats (Charles River, Quebec, Canada; 120–150 g, ∼4–6 weeks old) and 23 animals were included in the study. Experimental procedures were approved by the University of Calgary Animal Care Committee and were in accordance with national guidelines. Procedures were identical to a previous study (Day & Wilson, 2005). Briefly, rats were heparinized (1500 units i.p.) ∼15–30 min prior to being anaesthetized for dissection. Deep anaesthesia was induced using an overdose of halothane via inhalation. Adequate anaesthesia was assessed by testing for absence of response to noxious tail pinch. In rapid succession, rats were transected subdiaphragmatically and decerebrated at the midcollicular level. The decerebration was performed with a vertical cut using a new scalpel, removing all tissue rostral and dorsal to the colliculi. Transection and decerebration were performed in cold perfusate (∼5–8°C, 500 ml volume) containing (mm): 115 NaCl, 24 NaHCO₃, 4 KCl, 2 CaCl₂, 1 MgSO₄, 1.25 NaH₂PO₄ and 10 dextrose, and equilibrated with 95% O₂–5% CO₂. The skin was removed, and the preparation was placed into an experimental chamber and secured with ear bars. The descending aorta was canulated with a double-lumen catheter. One lumen of the catheter was connected to a peristaltic pump (Gilson Minipuls 3) and used to perfuse the descending aorta retrogradely with room-temperature perfusate containing 1.33% Ficoll Type-70 (Sigma-Aldrich F-2878) equilibrated with 40 Torr P_CO2 in O₂. The other lumen was attached to a pressure transducer to monitor perfusion pressure. Once canulated, the speed of the peristaltic pump was increased so as to increase perfusion pressure to ∼90 mmHg using a custom-built computer-controlled feedback system. A bilateral vagotomy was performed at the mid-cervical level. The common carotid arteries were tied off above the clavicles and canulated proximal to the carotid bodies. A separate peristaltic pump with two channels was used to perfuse the common carotid arteries at ∼15–20 ml min⁻¹ per carotid. Up to this stage in the dissection, central and peripheral perfusion was from the same tonometer (∼300 ml).

Independent perfusion of central (descending aorta) and peripheral (carotid arteries) circuits was then initiated by pulling fresh Ficoll-containing medium from two different reservoirs of a custom-built tonometer. This custom-built system was designed to accommodate a common return, while preventing mixing of perfusate once equilibrated in the two reservoirs. Between the reservoir and the preparation, the central perfusate passed through a bubble trap, heat exchanger and a 25 μm filter. The carotid perfusate passed through a bubble trap and a heat exchanger. Perfusate leaked from the many cut vessels in the preparation and returned to the experimental tonometer where it was recycled and re-equilibrated (total volume of perfusate, 750 ml).

Following the initiation of independent perfusion, the central perfusate was equilibrated with 40 Torr P_CO2 in O₂, and the peripheral perfusate was equilibrated with 40 Torr P_CO2 and 100 Torr P_O2 in N₂. Over the next 30 min, the temperature of the preparation and the central perfusion pressure was ramped to 32–33°C and ∼90 mmHg, respectively. The central P_CO2 was then adjusted slowly to the level utilized for the subsequent protocol (25, 30 or 50 Torr P_CO2; estimated pH for 33°C: 7.58, 7.5 and 7.27, respectively). Protocols were started at 70 ± 5 min after central canulation. Gasses used to equilibrate the perfusate were mixed by computer-controlled mass flow controllers (MKS Instruments) and sampled to ensure accuracy (SensorMedics, Medical Gas Analyser LB-2 and Sable Systems PA-1B O₂ analyser, calibrated daily). Gas concentrations reported here were accurate to within 0.1% (∼1.3 Torr in Calgary). A neuromuscular blocker was not used.

Electrophysiology

The left phrenic nerve was dissected out with a small piece of diaphragm attached to its distal end, and it was placed in a custom-made plexi-glass recording chamber. The nerve was bathed in perfusate and protected with petroleum jelly. Silver bipolar extracellular electrodes were used to record the phrenic neurogram, which was amplified (A-M Systems Differential AC Amplifier Model 1700), filtered (low cut-off, 300 Hz; high cut-off, 5 KHz), rectified and integrated (Amplitude Demodulator; Saga Tech), computer archived (Axon Instruments Digidata 1322A and Axoscope 9.0) at a sampling rate of 50 Hz, and analysed off-line.

Data analysis

Only preparations that exhibited a ramping phrenic discharge (the peak of phrenic activity occurs in the last 50% of the duty cycle) and an inspiratory time (T_I) of ≤1 s under control conditions were included in the study. Preparations with a T_I > 1 s were considered apneustic. Analysis of phrenic neurogram characteristics was performed using in-house software written by R.J.A.W. The following respiratory variables were quantified from the integrated phrenic neurogram: period (T_TOT), respiratory frequency (f_R, 60 times the inverse of the period), time to peak (T_p), inspiratory duration (T_I), expiratory duration (T_E), neural tidal volume (nV_T, the peak phrenic amplitude), neural minute ventilation (nV_E, the product of f_R and nV_T) and eupnoeic index (EI, T_p divided by T_I). A schematic of this quantification has been previously published (Day & Wilson, 2005).

Baseline conditions and protocols

Twenty-three preparations were used to quantify the phrenic responses to single or intermittent bouts of specific carotid body hypoxia at various levels of steady-state brainstem P_CO2. During the 70 min recovery period, baseline conditions were established: brainstem, 25, 30 or 50 Torr P_CO2 in O₂; and carotid bodies, 40 Torr P_CO2 and 100 Torr P_O2 in N₂. One of two experimental protocols was superimposed upon these baseline conditions.

Experiment 1: modulation of responses to carotid body hypoxia by brainstem P_CO2

This protocol consisted of three phases (15 min each), all identical other than the level of steady-state brainstem P_CO2. Protocol: 5 min baseline (brainstem 25 or 50 Torr P_CO2, carotid bodies 100 Torr O₂, 40 Torr P_CO2 in N₂), 5 min specific carotid body hypoxia (60 Torr P_O2 and 40 Torr P_CO2 in N₂), 5 min washout (one 15 min phase). A 15 min transition period, equilibrating the brainstem compartment with the alternate level of P_CO2 (50 or 25 Torr), preceded initiation of the next phase. The brainstem P_CO2 was alternated between 25 and 50 Torr three times, where four animals started and ended with 25 Torr, and four others started and ended with 50 Torr. The total protocol time was 75 min. This protocol design eliminated the possibility of time-dependent effects resulting from repeated carotid body stimulation from contaminating the results. Every phrenic burst over the protocol period was analysed, and respiratory variables were averaged in 30 s bins. Each 15 min phase was cut and grouped together with other matching brainstem CO₂ phases. Where an animal experienced a brainstem condition twice, the data were averaged together to contribute one sample. Thus, eight samples were taken from eight animals for each brainstem level. We chose this method of analysis so that (a) no one animal's responses disproportionately weighted the mean, and (b) we were able to utilize a single statistical test (two-factor repeated measures ANOVA) for each respiratory variable (f_R, nV_T and nV_E). To generate a brainstem P_CO2 response curve from this data (Fig. 6), baseline phrenic f_R, nV_T and nV_E at 50 Torr P_CO2 were normalized to the values obtained at 25 Torr P_CO2.

Figure 6.

Proposed theoretical model of the interaction between central and peripheral respiratory chemoreceptors f_R, nV_T and nV_E responses to brainstem P_CO2 are data from Experiment 1. Respiratory variables at 50 Torr P_CO2 are normalized to the baseline activity when the brainstem was held at 25 Torr P_CO2 for each animal. The shape of the central curve is extrapolated from unpublished data (filled symbols; ▪ and •). Open symbols (○ and □) represent the relative responses to specific carotid body hypoxia (60 Torr P_O2 and 40 Torr P_CO2). Each hypoxic response in A, B and C (○ or □) is from Fig. 3A, B and C, respectively.

Experiment 2: modulation of responses to intermittent carotid body hypoxia by brainstem P_CO2

Protocol: 5 min baseline, four 5 min specific carotid body perturbations (60 Torr P_O2 and 40 Torr P_CO2 in N₂), each separated by a 5 min washout, followed by a 30 min recovery period. This protocol was carried out under steady-state brainstem P_CO2 of either 30 (n = 9) or 50 Torr (n = 6). The total protocol time was 70 min. Every phrenic burst over the protocol period was analysed, and respiratory variables were averaged in 30 s bins.

Terminology

For statistical tests of respiratory variable dynamics, bins of raw data were selected from each animal as follows: Initial Baseline, minutes 4–5 (two bins) prior to the first perturbation (to which all data were normalized to for graphical presentation); Bout Baseline, the last bin prior to each perturbation (for the first bout, the Bout Baseline is the same as the Initial Baseline); Peak, the highest bin during each perturbation; Trough, the bin with the lowest value (nadir) during each washout. These points were selected for each of the four perturbations. The End Point was taken as an average of the last two bins of each protocol period. These values were selected for f_R, nV_T and nV_E.

Using this nomenclature, we define (a) PHxFD as when the frequency Trough differs from the Initial Baseline, and (b) Delta (Δ) Peak as the difference between Peak and the preceding Initial or Bout Baseline (for f_R, nV_T and nV_E).

Statistical tests

Firstly, we tested for within- and between-group differences in Initial Baseline and End Point variables, for both experiments, using two sets of tests. Two-factor repeated measures (2F RM) ANOVAs were used to test for differences between Initial Baseline and End Point in f_R, nV_T, nV_E, T_I, T_E and EI and brainstem P_CO2 (25 or 50 Torr P_CO2) for Experiment 1 data. One-factor (1F) RM ANOVAs were used to test for within-group differences between Initial Baseline and End Point in f_R, T_I, T_E and EI for Experiment 2. 1F ANOVAs were then used to test for between-group differences (brainstem 30 versus 50 Torr P_CO2). Data are summarized in Table 1.

Table 1.

Statistical notes

Note no.	Statistics
1	Table 2; two-factor (2F) RM ANOVA: Baseline versus End Point for nVT, F = 20.39, P = 0.003. NSD in interaction between brainstem PCO2 and time for nVT (F = 0.53, P = 0.49)
2	Table 2; 2F RM ANOVA: Brainstem 25 versus 50 Torr PCO2 for Baseline fR (F = 0.034, P = 0.86), nVT (F = 52.94, P < 0.001), nVE (F = 8.47, P = 0.023), TI (F = 10.7, P = 0.014) and EI (F = 11.68, P = 0.011)
3	Table 3; one-factor (1F) ANOVA: Baseline versus End Point for brainstem 30 Torr PCO2 EI (F = 11.77, P = 0.009) and brainstem 50 Torr PCO2TE (F = 8.12, P = 0.036). Brainstem 30 versus 50 Torr PCO2 for Baseline fR (F = 0.83, P = 0.38) and TI (F = 9.35, P = 0.009)
Experiment 1
4	Figs 1B and C, and 2A; 2F RM ANOVA: frequency response (F = 15.33, P = 0.006)
5	SNK post hoc test: Baseline versus Peak frequency for brainstem 25 Torr PCO2 (q = 7.09, P < 0.001) and 50 Torr PCO2 (q = 2.02, P = 0.18)
6	Figs 2A and 6A; 2F RM ANOVA: interaction between delta frequency response and brainstem PCO2 (F = 9.93, P = 0.016); SNK post hoc test: comparison between brainstem PCO2 and Peak frequency (q = 3.7, P = 0.025)
7	Figs 1C and 2A; 1F RM ANOVA: Baseline versus Peak fR under conditions of 50 Torr brainstem PCO2 (F = 14.44, P = 0.007)
8	Figs 1B and C, and 2B; 2F RM ANOVA: nVT response (F = 11.32, P = 0.012)
9	SNK post hoc test: Baseline versus Peak nVT for brainstem 25 Torr PCO2 (q = 5.52, P = 0.002) and 50 Torr PCO2 (q = 1.44, P = 0.33)
10	Figs 2B and 6B; 2F RM ANOVA: interaction between ΔnVT response and brainstem PCO2 (F = 4.46, P = 0.072)
11	Figs 1C and 2B; 1F RM ANOVA: Baseline versus Peak nVT under conditions of 50 Torr brainstem PCO2 (F = 12.47, P = 0.01)
12	Figs 1B and C, and 2C; 2F RM ANOVA: nVE response (F = 15.94, P = 0.005)
13	SNK post hoc test: Baseline versus Peak nVE for brainstem 25 Torr PCO2 (q = 7.47, P < 0.001) and 50 Torr PCO2 (q = 2.49, P = 0.11)
14	Figs 2C and 6C; 2F RM ANOVA: interaction between ΔnVE response and brainstem PCO2 (F = 13.88, P = 0.007)
15	Figs 1C and 2C; 1F RM ANOVA: Baseline versus Peak nVE under conditions of 50 Torr brainstem PCO2 (F = 10.234, P = 0.015)
Experiment 2
16	Fig. 4A; 1F RM ANOVA: Troughs compared with Initial Baseline and Washout (F = 5.048, P = 0.001)
17	SNK post hoc test: Troughs compared with Initial Baseline, Trough 1 (q = 3.34, P = 0.059), Trough 2 (q = 4.032, P = 0.033), Trough 3 (q = 4.76, P = 0.014), and Trough 4 (q = 4.923, P = 0.015)
18	Fig. 4B; 1F RM ANOVA: Troughs compared with Initial Baseline and Washout (F = 3.86, P = 0.01)
19	SNK post hoc test: Troughs compared with Initial Baseline (q < 1.44, P > 0.85), End point compared with Initial Baseline and Troughs (q > 4.036, P < 0.019)
20	Fig. 5A, left; 2F RM ANOVA:ΔfR responses for brainstem 30 Torr PCO2 (F = 26.06, P < 0.001), interaction between ΔfR and Bout (F = 6.93, P = 0.002)
21	Confirmed with 1F RM ANOVA on derived Δ values for fR: SNK post hoc test, pair-wise comparison of Bout 1 with Bouts 2, 3 and 4 (F > 4.58, P < 0.004)
22	Fig. 5B, left; 2F RM ANOVA: ΔfR responses for brainstem 50 Torr PCO2 (F = 14.79, P = 0.012), interaction between ΔfR and Bout (F = 2.53, P = 0.097)
23	Fig. 5A, middle; 2F RM ANOVA: ΔnVT responses for brainstem 30 Torr PCO2 (F = 9.63, P = 0.015), interaction between ΔnVT and Bout (F = 0.59, P = 0.63)
24	Fig. 5B, middle; 2F RM ANOVA: ΔnVT responses for brainstem 50 Torr PCO2 (F = 4.39, P = 0.09), interaction between ΔnVT and Bout (F = 0.53, P = 0.67)
25	Fig. 5A, right; 2F RM ANOVA: ΔnVE responses for brainstem 30 Torr PCO2 (F = 57.31, P < 0.001), interaction between ΔnVE and Bout (F = 5.91, P = 0.004)
26	Confirmed with 1F RM ANOVA on derived Δ values for nVE: SNK post hoc test, pair-wise comparison of Bout 1 with Bouts 2, 3 and 4 (F > 4.4, P < 0.008)
27	Fig. 5B, right; 2F RM ANOVA: ΔnVE responses for brainstem 50 Torr PCO2 (F = 23.94, P = 0.005), interaction between ΔnVE and Bout (F = 1.03, P = 0.41)

f_R, respiratory frequency; nV_T, neural tidal volume; nV_E, neural minute ventilation; RM, repeated measures; SNK, Student-Newman-Keuls. Note numbers refer to notes in Results and Tables 2 and 3 regarding specific statistical tests used and the respective results.

Secondly, for Experiment 1 data, 2F RM ANOVAs were performed to test for the significance of within-delta (Δ) responses in f_R, nV_T and nV_E (two levels: Initial Baseline and Peak variable) and between-group differences (two levels: brainstem 25 versus 50 Torr P_CO2).

Thirdly, for Experiment 2 data, 1F RM ANOVAS were used to test for differences in phrenic frequency between the Initial Baseline, End Point and each of the four Troughs for both brainstem conditions (30 and 50 Torr P_CO2). Response delta values for Peak f_R, nV_T and nV_E were tested using 2F RM ANOVAs. In these ANOVAs, one factor was the bout (four levels) and the other, the constituent of the delta values (two levels, Bout Baseline and Peak variable during carotid body hypoxic bout).

Statistical tests were performed on raw data. In all cases, once significant differences had been determined (assumed at P < 0.05), Student-Newman-Keuls (SNK) post hoc tests were used for pair-wise comparisons.

Results

Baseline variables and stability

The Initial Baseline (the last minute of the baseline period) and End Point (the last minute of the protocol, i.e. washout) for f_R, nV_T, nV_E, T_I, T_E and EI were compared to assess these respiratory variables under different brainstem conditions (summarized in Tables 2 and 3) for both experimental protocols.

Table 2.

Raw data from Initial Baseline and End Points under different steady-state brainstem P_CO2 conditions from Experiment 1, Single Bout CB Hypoxia Within Animal

Respiratory variable	Baseline	End Point
25 Torr PCO2 brainstem = 8
fR (min−1)	30.87 ± 4.4	32.04 ± 4.82
nVT (a.u.)	3.5 ± 0.32	3.29 ± 0.28*1
nVE (a.u.)	109.24 ± 17.47	103.91 ± 15.77
TI (s)	0.97 ± 0.066	0.99 ± 0.089
TE (s)	1.49 ± 0.52	1.81 ± 0.93
EI	0.67 ± 0.24	0.68 ± 0.24
50 Torr PCO2 brainstem n = 8
fR (min−1)	31.7 ± 2.012	32.39 ± 1.88
nVT (a.u.)	4.42 ± 0.24+,2	4.27 ± 0.24*1
nVE (a.u.)	140.81 ± 13.23+2	139.61 ± 13.54
TI (s)	0.73 ± 0.072+2	0.73 ± 0.067
TE (s)	1.23 ± 0.14	1.18 ± 0.12
EI	0.60 ± 0.036+2	0.58 ± 0.033

a.u., arbitrary units; T_I, inspiratory duration; T_E, expiratory duration; EI, eupnoeic index. These data illustrate the phrenic neurogram characteristics during the baseline period prior to and during a washout period following the experimental perturbation. Note that the baseline frequency is no different under any brainstem P_CO2 condition.

^*End Point variable is significantly different than the same variable for Baseline.

⁺Variable at 50 Torr P_CO2 brainstem is significantly different than the same variable under conditions of 25 Torr P_CO2 brainstem.

Table 3.

Raw data from Initial Baseline and End Points under different steady-state brainstem P_CO2 conditions from Experiment 2, Intermittent Hypoxia

Respiratory variable	Baseline	End Point
30 Torr PCO2 brainstem n = 9
fR (min−1)	33.42 ± 3.31	33.31 ± 4.52
TI (s)	0.81 ± 0.05	0.88 ± 0.06
TE (s)	1.19 ± 0.26	1.58 ± 0.6
EI	0.72 ± 0.027	0.76 ± 0.031*,3
50 Torr PCO2 brainstem n = 6
fR (min−1)	29.43 ± 1.73	31.53 ± 2.33
TI (s)	0.59 ± 0.05+,3	0.61 ± 0.034
TE (s)	1.51 ± 0.164	1.37 ± 0.17*,3
EI	0.72 ± 0.019	0.74 ± 0.012

These data illustrate the phrenic neurogram characteristics during the baseline period prior to and during a washout period following the experimental perturbation. Note that the baseline frequency is no different under any brainstem P_CO2 condition.

^*End Point variable is significantly different than the same variable for Baseline.

⁺Variable at 50 Torr P_CO2 brainstem is significantly different than the same variable under conditions of 30 Torr P_CO2 brainstem. nV_T and nV_E were not compared because these values contain arbitrary units (a.u.) and these two data sets are from different animals.

For Experiment 1, there were no differences between most Baseline and End Point respiratory variables for both brainstem conditions (25 and 50 Torr P_CO2). However, when comparing Baseline and End Point nV_T for both brainstem conditions, there was a decline over time (P = 0.003, Table 1¹). When comparing Baseline between brainstem conditions (a) baseline f_R was no different (P = 0.86, Table 1²), (b) nV_T was different (P < 0.001, Table 1²), and (c) nV_E was different (P = 0.023, Table 1²).

For Experiment 2, f_R, T_I, T_E and EI were similarly tested. There were no differences between most Baseline and End Point respiratory variables for both brainstem conditions (30 and 50 Torr P_CO2). Of note, when comparing Baseline between brainstem conditions, baseline f_R was no different (P = 0.38, Table 1³).

The variables of T_I, T_E and EI are reported in Tables 2 and 3 as indices of the shape of the neurogram and to ensure that the preparations used in the study fit the inclusion criteria. Preparations were included if they maintained a T_I≤ 1 s and EI between 0.5 and 1, indicating the peak of the neurogram occurred late in the duty cycle.

Experiment 1: modulation of responses to carotid body hypoxia by brainstem P_CO2

In Experiment 1, each preparation experienced each brainstem condition (25 and 50 Torr P_CO2) at least once (four preparations experienced 25 Torr twice and the other four experienced 50 Torr twice). Under each of these brainstem conditions, preparations experienced one bout of specific carotid body hypoxia (60 Torr P_O2 and 40 Torr P_CO2). Figures 1 and 2 illustrate f_R, nV_T and nV_E responses.

Figure 1.

Time course of respiratory variables in response to specific carotid body hypoxia (60 TorrP_O2and 40 TorrP_CO2) under differing steady-state brainstemP_CO2(Experiment 1) A, sample rectified, integrated neurogram of one entire experimental protocol whereby each animal experienced both brainstem P_CO2 levels shown. Total protocol time is 75 min. B and C, combined mean respiratory frequency (f_R), neural tidal volume (nV_T) and neural minute ventilation (nV_E) responses to carotid body hypoxia when the brainstem was held at steady-state 25 Torr P_CO2 and 50 Torr P_CO2, respectively. Data represent eight samples from eight animals. Bins represent 30 s averages of the respiratory variable shown. Black bars represent hypoxic carotid body stimulation (5 min). Total time is 15 min. Error bars represent s.e.m. Data are normalized to last minute of the baseline immediately preceding the hypoxic challenge.

Figure 2.

Respiratory responses (normalized to Initial Baseline) to specific carotid body hypoxia under conditions of 25 and 50 Torr steady-state brainstemP_CO2 A, f_R; B, nV_T; C, nV_E. *Response is significantly different from baseline. ⁺Response is significantly different from 25 Torr P_CO2 brainstem (i.e. interaction). Data are from eight samples (at each brainstem P_CO2 level) taken from eight animals. Error bars represent s.e.m.

f_R responses

Figure 2A illustrates the quantification of f_R responses to carotid body hypoxia under conditions of brainstem P_CO2 of 25 and 50 Torr P_CO2. Carotid body hypoxia elicited a frequency response (Peak f_R compared to Baseline; P = 0.006, Table 1⁴). Under conditions of 25 Torr brainstem, there was a frequency response, whereas under conditions of 50 Torr, there was not (P < 0.001 (Table 1⁵) and P = 0.18 (Table 1⁵), respectively). Thus, there was an interaction between brainstem P_CO2 and the delta response for frequency (P = 0.016, Table 1⁶). Although the 2F RM ANOVA revealed no f_R response under conditions of 50 Torr brainstem (Figs 1C and 2A), a 1F RM ANOVA on this data set revealed a statistically significant response (P = 0.007, Table 1⁷).

nV_T responses

Figure 2B illustrates the quantification of nV_T responses to carotid body hypoxia under conditions of brainstem P_CO2 of 25 and 50 Torr P_CO2. Carotid body hypoxia elicited an nV_T response (Peak nV_T compared with Baseline; P = 0.012, Table 1⁸). Under conditions of 25 Torr brainstem, there was an nV_T response, whereas under conditions of 50 Torr, there was not (P = 0.002 (Table 1⁹) and P = 0.33 (Table 1⁹), respectively). Thus, although an interaction between brainstem P_CO2 and the delta response for nV_T was not significant, the similarity was weak (P = 0.072, Table 1¹⁰). Although the 2F RM ANOVA revealed no nV_T response under conditions of 50 Torr brainstem (Figs 1C and 2B), a 1F RM ANOVA on this data set revealed a statistically significant response (P = 0.01, Table 1¹¹).

nV_E responses

Figure 2C illustrates the quantification of nV_E responses to carotid body hypoxia under conditions of brainstem P_CO2 of 25 and 50 Torr P_CO2. Carotid body hypoxia elicited an nV_E response (Peak nV_E compared to Initial Baseline; P = 0.005, Table 1¹²). Under conditions of 25 Torr brainstem, there was a frequency response, whereas under conditions of 50 Torr, there was not (P < 0.001 (Table 1¹³) and P = 0.11 (Table 1¹³), respectively). Thus, there was an interaction between brainstem P_CO2 and the delta response for nV_E (P = 0.007, Table 1¹⁴). Although the 2F RM ANOVA revealed no nV_E response under conditions of 50 Torr brainstem (Figs 1C and 2C), a 1F RM ANOVA on this data set revealed a statistically significant response (P = 0.015, Table 1¹⁵).

Experiment 2: modulation of responses to intermittent carotid body hypoxia by brainstem P_CO2

In Experiment 2, under steady-state brainstem conditions of either 30 or 50 Torr P_CO2, preparations experienced four bouts of 5 min specific carotid body hypoxia (60 Torr P_O2 and 40 Torr P_CO2), separated by 5 min and followed by a 30 min washout. Figures 3, 4 and 5 illustrate f_R, nV_T and nV_E responses.

Figure 3.

Time course of respiratory variables during intermittent specific carotid body hypoxia (60 TorrP_O2) under differing steady-state brainstemP_CO2(Experiment 2) A, f_R, nV_T and nV_E when the brainstem was held at 30 Torr P_CO2 (n = 9). B, f_R, nV_T and nV_E when the brainstem was held at 50 Torr P_CO2 (n = 6). Bins represent 30 s averages of variables shown. Black bars represent hypoxic carotid body stimulation (5 min). Total protocol time is 70 min. Error bars are s.e.m. Data are normalized to a 1 min baseline prior to the first perturbation.

Figure 4.

Post-hypoxia responses (normalized to Initial Baseline) in frequency (f_R) to specific carotid body hypoxia (60 TorrP_O2and 40 TorrP_CO2) under conditions of steady-state 30 (n = 9) or 50 Torr (n = 6)P_CO2brainstem A, post-hypoxia f_R Trough where the brainstem was held at 30 Torr P_CO2. B, post-hypoxia f_R Trough where the brainstem was held at 50 Torr P_CO2. The post-hypoxic Trough represents the lowest 30 s f_R bin (nadir) following each bout of hypoxia during each 5 min washout. Where the Trough is significantly different than the initial baseline, post-hypoxic frequency decline (PHxFD) is present. Data are normalized to the last minute of the initial baseline period. Error bars represent s.e.m.*Significantly different than the initial baseline.

Figure 5.

Delta (Δ) responses (relative to each bout baseline) to intermittent carotid body hypoxia under differing steady-state brainstemP_CO2conditions A, Δf_R, ΔnV_T and ΔnV_E when the brainstem was held at 30 Torr P_CO2. B, Δf_R, ΔnV_T and ΔnV_E when the brainstem was held at 50 Torr P_CO2. Data are normalized to the last minute of the initial baseline period. Error bars are s.e.m.*Peak is significantly different from each respective bout baseline. ⁺Delta response is significantly different from the first response, demonstrating progressive augmentation.

f_R responses

Figure 4A and B illustrates the quantification of posthypoxic responses to specific carotid body intermittent hypoxia under conditions of 30 and 50 Torr brainstem P_CO2. For steady-state brainstem 30 Torr P_CO2, following termination of hypoxic bouts the activity between bouts decreased below the initial baseline (i.e. PHxFD, Troughs compared with Initial Baseline and End Point; P = 0.001, Table 1¹⁶). PHxFD appeared to have some bout dependence, as Troughs 2, 3 and 4 were significantly different than baseline whereas Trough 1 was not (Trough for bouts 1, 2, 3 and 4: P = 0.059, P = 0.033, P = 0.014 and P = 0.015, respectively, Table 1¹⁷). Conversely, for steady-state brainstem 50 Torr P_CO2, following termination of hypoxic bouts, the activity between bouts did not decrease below baseline (Troughs compared to Initial Baseline and End Point; P > 0.85, Table 1^18,19).

The left-hand panels of Fig. 5A and B illustrate the quantification of delta (Δ) f_R responses to specific carotid body intermittent hypoxia under conditions of 30 and 50 Torr brainstem P_CO2. For steady-state brainstem 30 Torr P_CO2, each of the four hypoxic bouts elicited increases in f_R that were bout dependent (Δf_R responses, P < 0.001, Table 1²⁰; interaction between Δf_R and bout, P = 0.002, Table 1²⁰). Thus, the ΔPeak f_R from each bout baseline got progressively larger with each bout, where the derived delta values from bouts 2, 3 and 4 were larger than bout 1 (P < 0.004, Table 1²¹). For steady-state brainstem 50 Torr P_CO2, each of the four hypoxic bouts elicited increases in f_R that were not bout dependent (Δf_R responses, P = 0.012, Table 1²²; interaction between Δf_R and bout, P = 0.097, Table 1²²).

nV_T responses

The middle panels of Fig. 5A and B illustrate the quantification of ΔnV_T responses to specific carotid body intermittent hypoxia under conditions of 30 and 50 Torr brainstem P_CO2. For steady-state brainstem 30 Torr P_CO2, each of the four hypoxic bouts elicited increases in nV_T that were not bout dependent (ΔnV_T responses, P = 0.015, Table 1²³; interaction between ΔnV_T and bout, P = 0.63, Table 1²³). For steady-state brainstem 50 Torr P_CO2, none of the four hypoxic bouts elicited significant increases in nV_T (ΔnV_T responses, P = 0.090, Table 1²⁴).

nV_E responses

The right-hand panels of Fig. 5A and B illustrate the quantification of ΔnV_E responses to specific carotid body intermittent hypoxia under conditions of 30 and 50 Torr brainstem P_CO2. For steady-state brainstem 30 Torr P_CO2, each of the four hypoxic bouts elicited increases in nV_E that were bout dependent (ΔnV_E responses, P < 0.001, Table 1²⁵; interaction between ΔnV_E and bout, P = 0.004, Table 1²⁵). Thus, the ΔPeak nV_E from each bout baseline got progressively larger with each bout, where bouts 2, 3 and 4 were larger than bout 1 (P < 0.008, Table 1²⁶). For steady-state brainstem 50 Torr P_CO2, each of the four hypoxic bouts elicited increases in nV_E that were not bout dependent (ΔnV_E responses, P = 0.005, Table 1²⁷; interaction between ΔnV_E and bout, P = 0.41, Table 1²⁷).

Discussion

Using an in situ DPP, we report that the level of steady-state brainstem P_CO2 modulates the magnitude of (a) phrenic responses resulting from specific carotid body hypoxia (60 Torr P_O2 and 40 Torr P_CO2), and (b) time-dependent phenomena. Numerous studies have examined the interaction of central and peripheral chemoreceptors previously. However, ours is the first study to separate central and peripheral chemosensitivity in the rat and explore how the interaction of chemoreceptors without vagal or cortical descending inputs affects time-dependent ventilatory phenomena.

The DPP: a model for studying chemoreceptor interactions in the rat

The DPP provides a novel method for studying respiratory chemoreceptor interaction in the rat. Uniquely, the DPP allows the study of respiratory chemosensitivity in non-anaesthetized decerebrate rodents, independent of heart rate, systemic blood pressure, hormonal factors and vagal inputs. As such, the DPP provides an important bridge between studies of central and peripheral respiratory chemosensitivity at the system level and recent studies in rodents at the cellular level (Mulkey et al. 2004; Takakura et al. 2006).

The main limitations of the DPP are shared with other decerebrate and vagotomized preparations. Particularly, there is a pronounced leftward shift of the apnoeic threshold and the CO₂ responsiveness relative to intact animals (compare Boden et al. 1998 with Day & Wilson, 2005). Other studies have demonstrated that both decerebration (Hayashi & Sinclair, 1991; Mitchell, 1990; Nielsen et al. 1986; St-John & Paton, 2000) and vagotomy (Takakura et al. 2006) can shift chemosensitivity in the hypocapnic direction. This consistent observation is an important caveat to consider when interpreting data from studies of chemosensitivity using decerebrate and vagotomized preparations.

An additional caveat to consider is that the artificially perfused brainstem of the DPP is likely to be hyperoxic (Tissue P_O2, ∼200 Torr; Wilson et al. 2001). To date, we have been unable to demonstrate that hyperoxia affects the output of the DPP, although such a possibility cannot be excluded.

Like anaesthesia, decerebration eliminates cortical inputs influencing respiration, but without the chemical desensitizing effects of anaesthesia on chemosensitivity (Czapla & Zadina, 2005). The CO₂ chemosensitivity of this preparation is similar in magnitude to that reported in awake freely behaving rats of the same species (compare Czapla & Zadina, 2005 with Day & Wilson, 2005). Further, the removal of descending influences by decerebration provides a level of reduction that is likely to be essential toward understanding the complexity of chemoreceptor interactions without the confounding influences of wakefulness drive on ventilation (Fink, 1961).

Finally, an important strength of our model is that both central and peripheral chemoreceptor compartments are intact and independently perfused artificially. Unlike in vivo breathing preparations, we can hold the brainstem with levels of P_CO2 below eupnoeic levels. Thus, the role of one chemoreceptor can be explored across the entire physiological range while the other chemoreceptor is maintained at steady-state levels of chemostimulation.

Hypo-additive interaction between peripheral and central chemoreceptors

In the DPP, peripheral chemoreceptor stimulation had little effect on nV_T; nV_T was largely dictated by the level of brainstem P_CO2 (Fig. 6B). In contrast, increasing brainstem P_CO2 from 25 to 50 Torr, while maintaining normoxic and normocapnic carotid body perfusion, had little effect on frequency; changes in frequency were largely brought about by changes in carotid body P_O2 (Figs 1B and C, 2A and 6A).

As peripheral chemoreceptor stimulation had little effect on nV_T, the interaction between central and peripheral chemoreceptors in terms of nV_T is best described as additive. Remarkably, however, frequency responses were greatest when peripheral chemoreceptors were stimulated while the brainstem was hypocapnic (increased ∼50% above that without peripheral stimulation; Figs 2A and 6A). When both peripheral and central chemoreceptors were stimulated (carotid body hypoxia and brainstem 50 Torr P_CO2), mean frequency was elevated only slightly above that without peripheral stimulation (∼10%; Figs 2A and 6A). Thus, central chemoreceptor activation appears to clamp respiratory frequency, nullifying the effect of excitatory inputs from peripheral chemoreceptors. In terms of frequency then, the interaction between central and peripheral chemoreceptors is strongly hypo-additive.

The combined effect of additive nV_T and hypo-additive frequency interactions is a hypo-additive nV_E interaction. Specifically, carotid body hypoxia induced a large increase in nV_E when brainstem P_CO2 was low (∼70%), but only a minimal increase in nV_E when brainstem P_CO2 was high (∼10%; Figs 2C and 6C).

Other investigators have demonstrated hypo-additive V_E interactions in larger mammals, including awake goats (Smith et al. 1984), anaesthetized dogs (Gesell et al. 1940; Adams & Severns, 1982) and cats (Berger et al. 1978; Giese et al. 1978; Eldridge et al. 1981). For example, Adams & Severns (1982) used specific carotid body perfusion with either normoxic or hypoxic–hypercapnic blood in anaesthetized dogs breathing various levels of inspired CO₂ in order to maintain systemic (central) isocapnia. They found that specific carotid body stimulation produced significant increases in tidal volume when systemic P_aCO2 was normocapnic but not hypercapnic. Respiratory frequency continued to increase with carotid body stimulation at all levels of systemic P_aCO2. The model used by these investigators only allows changes in systemic P_CO2 in the hypercapnic direction. Smith et al. (1984) investigated the ventilatory responses to carotid body stimulation across a wider range of cerebral spinal fluid (CSF) pH in awake goats by perfusing the cisterna magna artificially with mock CSF containing varying levels of bicarbonate. Employing bolus intravenous injections of NaCN to stimulate the carotid bodies, they found that carotid body stimulation was more efficacious at eliciting increases in tidal volume and breathing frequency when the cisterna magna perfusion was alkaline than acidic (Smith et al. 1984).

However, other studies have found qualitatively different results. For example, simple addition in V_E has been reported in awake goats (Daristotle & Bisgard, 1989), pons-medulla artificially perfused anaesthetized cats (Heeringa et al. 1979; van Beek et al. 1983) and humans (Clement et al. 1992, 1995; St. Croix et al. 1996). Moreover, simple addition between inputs is assumed in models of the human respiratory system (Khoo et al. 1991; Painter et al. 1993; Ursino et al. 2001; Duffin & Mahamed, 2003; Topor et al. 2004).

An additive interaction in V_E necessitates a hypo-additive interaction in at least one respiratory variable due to the multiplicative relationship between tidal volume and rate in generating minute ventilation (Mitchell et al. 1990). Thus, a number of discrepant reports may be reconciled to some degree by considering volume and timing components separately. Unfortunately, several of these interaction studies suggesting simple addition report only V_E (Heeringa et al. 1979, Van Beek et al. 1983; Clement et al. 1992, 1995; St Croix et al. 1996) or could not discriminate underlying hypo-additive interactions in volume or frequency when data from multiple animals were combined (e.g. Daristotle & Bisgard, 1989).

Our findings are in contrast to other studies in intact humans (Robbins, 1988) and anaesthetized dogs (Adams et al. 1978) that suggest the interaction is hyper-additive. However, we note that studies in humans have reported both hyper-additive interaction and simple addition (compare Robbins, 1988 with Clement et al. 1992, 1995, and St Croix et al. 1996). Similarly, whereas Adams et al. (1978) report a hyper-additive interaction in dogs, other studies in dogs report hypo-additive interactions (Gesell et al. 1940; Adams & Severns, 1982). Further, we note that studies in humans assume that the contribution of central and peripheral chemoreceptors in stimulating ventilation can be temporally separated by exploiting the difference in time constants between the responses from stimulation of the two receptor sites (e.g. DeGoede et al. 1985; Pedersen et al. 1999). This assumption may be only partly correct given work in humans where the central chemoreceptor gain (determined via temporal separation technique) is decreased when the carotid body input is eliminated by resection (Bellville et al. 1979; Fatemian et al. 2003). Finally, in vivo studies of chemoreceptor interaction necessitate measurements in the presence of vagal and descending cortical inputs. These influences may themselves interact with central and peripheral chemoreceptors, causing an apparent change in the algebra of peripheral and central chemoresponses.

Our results also differ, at least superficially, from recent cellular work suggesting simple addition between the CO₂ response of putative central respiratory chemoreceptor cells in the retrotrapezoid nucleus (RTN) and carotid body stimulation (Takakura et al. 2006). These authors suggest that carotid body stimulation increases RTN cell firing rates across a range of inspired CO₂ (above eupnoeic levels) in an additive manor; however, in this study: (a) recordings were made from single cells from only one of several putative central chemoreceptor sites; and (b) in the protocol, both chemoreceptor compartments were stimulated when systemic CO₂ was increased. The known multiplicative interaction between O₂ and CO₂ at the level of the carotid body is likely to confound these results.

In summary, the various experimental models, techniques and designs may contribute to the discrepant literature on central–peripheral chemoreceptor interaction. However, we note that in all but one study (Adams et al. 1978) in which stimuli to the central and peripheral chemoreceptors were physically separated, hypo-additivity has either been explicitly documented or is necessarily present algebraically, albeit in different respiratory variables, depending on the species.

Brainstem P_CO2 modulates time-dependent effects

Importantly, our data also suggest that central chemoreceptor activation can nullify time-dependent ventilatory phenomena produced by episodic activation of the peripheral chemoreceptors.

In Experiment 2 we examined the effects of differing steady-state brainstem P_CO2 on the phrenic responses resulting from specific carotid body intermittent hypoxia. Previously, we showed that specific carotid body hypoxia (40 Torr P_O2, 40 Torr P_CO2 balanced N₂) is sufficient to elicit PHxFD under conditions of 30 Torr steady-state brainstem P_CO2. Repeated hypoxic bouts exacerbate PHxFD, leading to larger response deltas for each subsequent bout of hypoxia (progressive augmentation). The progressive augmentation in frequency translated into nV_E (Day & Wilson, 2005). Here we demonstrate that a milder hypoxic stimulus (60 Torr P_O2) was also sufficient to elicit post-hypoxic frequency decline and progressive augmentation (in both f_R and nV_E) under the same brainstem conditions as the previous study, although it took two bouts to get a significant post-hypoxic depression (Fig. 4A). However, when the brainstem was held at 50 Torr P_CO2, PHxFD and progressive augmentation were absent, even after four bouts of carotid body hypoxia (Figs 3B, 4B and 5B).

The magnitude of other time-dependent effects may also depend on brainstem P_CO2. For example, Eldridge (1974) demonstrated that ventilatory after-discharge, defined as a sustained increase in ventilation following termination of stimulus, appears to be masked by hypercapnia (∼10 Torr above the apnoeic threshold). However, another study reported that ventilatory after-discharge is prolonged under conditions of systemic hypercapnia and truncated under conditions of poikilocapnia (lowered systemic P_CO2; Engwall et al. 1994). Neither study specifically address post-hypoxic frequency decline.

Several studies have demonstrated the importance of the noradrenergic A5 pontine region in PHxFD. Lesions to this region abolishes PHxFD in both the anaesthetized, artificially ventilated and intact awake rat (Coles & Dick, 1996; Schlenker & Prestbo, 2003), while electrical stimulation of this region causes prolongation of T_E. Further, systemic application of α-adrenoreceptor antagonists blocks PHxFD, although whether this is through a central or peripheral mechanism is currently unclear (Coles et al. 1998; Bach et al. 1999). Recent work suggests that the A5 region may contain neurons that have an endogenous sensitivity to hypercapnia (Ilyinsky et al. 2003). The A5 region is also stimulated by carotid sinus nerve activation (Guyenet et al. 1993; Erickson & Millhorn, 1994). Thus, this region may be an integration site for PHxFD, responding to specific carotid body hypoxia and modulated by hypercapnia.

Interaction versus saturation

Given the shift in the chemosensitive range of decerebrate preparations like the DPP, the possibility exists that the increase in ventilation produced by activation of central chemoreceptors might mask the effects of activating the peripheral chemoreceptors through saturation of a common pathway within the CNS (Eldridge et al. 1981; Smith et al. 1984). However, this saturation hypothesis does not appear to hold in the case of frequency, because (a) baseline frequencies are the same at both levels of brainstem P_CO2 (Table 2 and Fig. 6A), and (b) carotid body stimulation often elicited larger absolute maximum values when the brainstem was 25 Torr P_CO2 than when it was 50 Torr P_CO2, in the same animal (Fig. 6A).

Significance

Our data demonstrate that in the rat, when central and peripheral chemoreceptor compartments are perfused independently, the contribution of the carotid bodies to ventilation is largely dictated by the chemical status of the central compartment. Thus, despite the caveats of our preparation, our data are in agreement with all but one animal study in which the two compartments have been stimulated independently, confirming a negative interaction in at least one component of the ventilatory response. Hence, a negative interaction has been reliably demonstrated across species, and calls into question the validity of the assumption of simple addition used in models of the human respiratory controller.

We hypothesize that the negative interaction between the chemoreceptor compartments may be important in the pathophysiology of periodic breathing. In vivo, the central chemoreceptors are likely to be protected from rapid changes in blood gases by the temporal delay in CO₂ transit and equilibration in the central compartment (Pedersen et al. 1999; Teppema & Dahan, 2005; Smith et al. 2006). Consequently, conditions may arise in which activation of peripheral chemoreceptors occurs during chronic central hypocapnia. Our results suggest that in situations where the brainstem is hypocapnic, an augmented peripheral chemoreflex may result, compared with when the brainstem is normocapnic. An increase in peripheral chemoreflex gain has the potential to exacerbate the likelihood of periodic breathing during sleep (Chapman et al. 1988; Topor et al. 2004), when the sole reliance is on the chemical control system (Skatrud & Dempsey, 1983). In this regard, we note clinical observations that congestive heart failure patients (Xie et al. 1995; Solin et al. 2000) and those sleeping at high altitude (Lahiri et al. 1983; Burgess et al. 2004) have increased peripheral chemoreflex gain and periodic breathing during sleep in the face of chronic hyperpnoea and hypocapnia.

Conclusions

In conclusion, our data suggest a negative (i.e. hypo-additive) interaction exists between central and peripheral respiratory chemoreceptors in the rat, supporting data from studies in larger animals, but in contrast to models used for the human respiratory control system. We also demonstrated that PHxFD and progressive augmentation resulting from intermittent carotid body hypoxia can be modulated by the level of brainstem P_CO2. These results suggest that in clinical populations experiencing sustained hyperpnoea, the resulting brainstem hypocapnia may increase the gain of the peripheral chemoreflex, potentially exacerbating the likelihood of periodic breathing during sleep.