Carotid chemoafferent activity is not necessary for all phrenic long-term facilitation following acute intermittent hypoxia

Abstract

Phrenic long-term facilitation (pLTF) is a form of respiratory plasticity induced by acute intermittent hypoxia (AIH) or episodic carotid chemoafferent neuron activation. Surprisingly, residual pLTF is expressed in carotid denervated rats. However, since carotid denervation eliminates baroreceptor feedback and causes profound hypotension during hypoxia in anesthetized rats, potential contributions of these uncontrolled factors or residual chemoafferent neuron activity to residual pLTF cannot be ruled out. Since ATP is necessary for hypoxic carotid chemotransduction, we tested the hypothesis that functional peripheral chemoreceptor denervation (with intact baroreceptors) via systemic P2X receptor antagonism blocks hypoxic phrenic responses and AIH-induced pLTF in anesthetized rats. Pyridoxal-phosphate-6-azophenyl-2',4'-disulfonic acid (PPADS; 100 mg/kg i.v.), a non-selective P2X receptor antagonist, was administered to anesthetized, vagotomized, paralyzed and ventilated male Sprague-Dawley rats prior to AIH (3, 5 min episodes of 10% O(2); 5 min intervals). Although PPADS strongly attenuated the short-term hypoxic phrenic response (20 ± 4% vs. 113 ± 15% baseline; P < 0.001), pLTF was reduced but not eliminated 60 min post-AIH (25 ± 4% vs. 51 ± 11% baseline; n = 8 and 7, respectively; P < 0.002). Thus, AIH initiates residual pLTF out of proportion to the diminished hypoxic phrenic response and chemoafferent neuron activation. Although the mechanism of residual pLTF following functional chemo-denervation remains unclear, possible mechanisms involving direct effects of hypoxia on the CNS are discussed.

Fig. 1

Representative tracings of compressed phrenic nerve recordings during experimental protocols. (A) AIH following vehicle injections, demonstrating normal pLTF. (B) AIH with PPADS, demonstrating residual pLTF. (C) Time control experiment following PPADS administration without AIH. (D) Time control experiment following vehicle injection without AIH. Vertical bars to the left of each experimental trace represent 1V. Dotted horizontal lines represent the initial, baseline value of phrenic burst amplitude. The time represented after the final hypoxic episode is 1 h.

Fig. 2

Changes in phrenic burst amplutude (A), burst frequency (B) and minute phrenic activity (C) during each of three hypoxic episodes (i.e. AIH) in vehicle-treated (filled bars, n = 7) and PPADS-treated (open bars, n = 8) rats. Values are means±1 SEM. *Different from baseline, P < 0.005; ^#different from corresponding vehicle control (i.e. same protocol, AIH or hyperoxia), P < 0.05.

Fig. 3

Changes in phrenic activity from baseline after episodic hypoxia in vehicle-treated (●, n = 7) and PPADS-treated (■, n = 8) rats. Time control rats without AIH are also shown (vehicle-treated, ○, n = 6; PPADS-treated, □, n = 6): (A) phrenic burst amplitude; (B) phrenic burst frequency; (C) minute phrenic activity. Values are means±1 SEM. *Different from baseline, P < 0.005; ⁺different from corresponding time control group (i.e. same drug), P < 0.005; ^#different from corresponding vehicle control group (i.e. same protocol, AIH or hyperoxia) at same time point, P < 0.005.

Fig. 4

Regression analysis of pLTF magnitude vs. the change in mean arterial pressure during hypoxia (A), or the change in mean arterial pressure during an experimental protocol (B; 60 min post-AIH vs. baseline). In (A), there was no significant correlation between the change in MAP during hypoxia and pLTF (P = 0.24, r² = 0.124) in vehicle-treated (●, n = 7) and/or PPADS-treated (○, n = 8) rats. Similarly (B) MAP at 60 min post-AIH and pLTF are not significantly correlated (P = 0.72, r² = 0.01) in PPADS-treated rats. However, a significant correlation was apparent between pLTF and MAP at 60min post-AIH in vehicle-treated rats (P = 0.01, r² = 0.717). This latter observation is not consistent with numerous previous observations from our laboratory and is regarded as a spurious result here (see discussion; Baker-Herman and Mitchell, 2008).