During hypoxia, release of platelet-activating factor (PAF) and activation of its cognate receptor (PAFR) regulate neural transmission and are required for full expression of peak hypoxic ventilatory response (pHVR) but not hypercapnic ventilatory response. However, it is unclear whether PAFR underlie components of long-term ventilatory adaptations to hypoxia. To examine this issue, adult male PAFR(+/+) and PAFR(−/−) mice were exposed to intermittent hypoxia (IH) consisting of 90 s 21% O2 and 90 s 10% O2 for 30 days, and normoxic and hypoxic ventilatory patterns were assessed using whole body plethysmography. Starting at day 14 of IH, normoxic ventilation in PAFR(−/−) was reduced significantly compared with PAFR(+/+) mice (P < 0.001), the latter exhibiting a prominent long-term ventilatory facilitation (LTVF). However, IH-exposed PAFR(−/−) mice had markedly enhanced pHVR and hypoxic ventilatory decline that became similar to those of IH-exposed PAFR(+/+) mice. Thus we postulate that PAFR expression and/or function underlies critical components of IH-induced LTVF but does not play a role in the potentiation of the hypoxic ventilatory response after IH exposures.
- respiratory control
- synaptic plasticity
- central chemosensitivity
- ventilatory adaptation
- long-term facilitation
platelet-activating factor (1-O-alkyl-2-acetyl-sn-glycero-3-phosphocholine, PAF) is a biologically active phospholipid that is now well established in its multifaceted roles involving several biological activities and tissue locations, including the central nervous system (CNS; see Refs. 3 and 22). In the CNS, PAF is primarily synthesized through a de novo mechanism involving cytidine diphosphate choline-1-O-2-acetyl-sn-glycerolcholine phosphotransferase (CGCP), which mediates the transfer of phosphocholine to alkylacetyl-sn-glycerol (28). CGCP is abundant in the rodent brain and is responsive to both physiological and pathological stimulation such as tissue ischemia/hypoxia, high-frequency electrical stimuli, and neurotransmitters such as ACh and dopamine in a Ca2+-dependent manner (8, 9, 13, 28, 40, 47, 50). Thus PAF has emerged as an important mediator of both excitotoxic injury and of synaptic transmission and neural plasticity, such as long-term potentiation (LTP; see Refs. 8, 9, 13, 28, 40, 47, 50).
Expression of the PAF receptor (PAFR) in the CNS has been demonstrated in both neurons and microglia (35) and may also colocalize with N-methyl-d-aspartate glutamate receptors in hippocampal neurons (8). PAF application modulates neuronal differentiation, calcium fluxes, and LTP (4, 10, 23, 26, 27, 55), and the latter can be blocked by administration of the PAFR antagonist BN-52021, both in vivo and in vitro, thereby supporting a role for PAF as a retrograde messenger during neuronal activation (21, 23, 25, 32).
PAFR are constitutively expressed in the pontomedullary regions in relative abundance (6–8), and microinjection of a long-acting PAF analog in the dorsocaudal brain stem induces significant ventilatory enhancements (16). Conversely, local or systemic administration of BN-52021 was associated with substantial attenuation of the hypoxic ventilatory response (HVR), but not of the hypercapnic ventilatory response (16), and this effect was primarily mediated through changes in tidal volume (VT) without significant changes in the frequency responses (16, 49). We recently confirmed these findings using a transgenic mouse model deficient in the PAFR [PAFR(−/−); see Ref. 43]. In the same study, PAFR(−/−) animals displayed marked differences in their metabolic responses when exposed to an acute hypoxic environment (43). However, although PAFR activity is now clearly implicated in acute components of the HVR, it remains unclear whether PAFR play any role in long-term ventilatory adaptations to hypoxia.
To further examine this issue, we took advantage of the unique long-term ventilatory effects induced by exposures to intermittent hypoxia (IH; see Refs. 33, 38, 42, 44, 48, 54) and exposed wild-type adult male mice [PAFR(+/+)] and PAFR(−/−) mice to IH for a period of 30 days, during which ventilatory responses were assessed. We hypothesized that PAFR(−/−) mice would have maladaptive responses to long-term hypoxia exposures and therefore provide further insight in the role(s) of PAFR in the ventilatory adaptations to hypoxia.
MATERIALS AND METHODS
The experimental protocols were approved by the Institutional Animal Use and Care Committee and are in close agreement with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.
Breeding pairs of PAFR-deficient mice (hybrids between C57BL/6J and 129/Ola) and their wild-type littermates were developed by Dr. Takao Shimizu and provided by Dr. Nicholas Bazan (Louisiana State University, New Orleans, LA). Littermates were genotyped according to previously reported procedures (19). Briefly, Quick Tail DNA Prep for Genotyping (1 mm mouse tail) was obtained under halothane anesthesia, added to digestion buffer [5 mM EDTA, 200 mM NaCl, 100 mM Tris·HCl (pH 8.0; 0.2% SDS), 0.5 mg/ml proteinase K, and 12.5 mg/ml RNase A], and digested overnight in a hybridization oven at 55°C. The DNA was extracted with phenol-chloroform-isoamyl alcohol (25:24:1), washed two times in chloroform-isoamyl alcohol (26:4), and precipitated with 1:1 isopropanol. Pellets were resuspended in Tris-EDTA (pH 8) and heated for 2 h at 65°C. PCR was as follows: 0.5 ml DNA, 13 ml PCR buffer, 1.5 mM MgCl, 0.1 mM dNTPs, 200 ng primers (forward: 59-gcc tgc ttg ccg at atc atg gtg gaa aat-39, reverse: 39-gcg atg cgc tgc gaa tcg gga gcg gcg ata-59 for deficient allele and forward: 59-tat ggc tga cct gct ctt cct gat-39, reverse: 39-tat tgg gca cta ggt tgg tgg agt-59 for wild-type allele), 11.5 ml water, and 1 unit Taq DNA Polymerase in a total volume of 25 ml. PCR was carried out using a Perkin-Elmer GeneAmp PCR System 9600 cycler with the following conditions: 94°C for 0.25 min, 60°C for 0.25 min, and 72°C for 0.5 min for 30 cycles. The PCR product was then analyzed on a 1.8% agarose gel containing ethidium bromide.
Animals were placed in four identical commercially designed chambers (30 × 20 × 20 in.; Oxycycler model A44XO; Biospherix, Redfield, NY) that were operated under a 12:12-h light-dark cycle (6:00 AM-6:00 PM). Gas was circulated around each of the chambers, attached tubing, and other units at 60 l/min (i.e., 1 complete change/10 s). The O2 concentration was measured continuously by an O2 analyzer and was changed by a computerized system controlling the gas valve outlets such that the moment-to-moment desired O2 concentration of the chamber was programmed and adjusted automatically. Deviations from the desired concentration were met by addition of N2 or O2 through solenoid valves. Ambient CO2 in the chamber was monitored periodically and maintained at <0.01% by adjusting overall chamber basal ventilation. Humidity was measured and maintained at 40–50% by circulating the gas through a freezer and silica gel. Ambient temperature was kept at 22–24°C. The IH profile consisted of alternating room air and 10% O2 every 90 s for the 12 h of light period. Control animals were exposed to circulating normoxic gas in one of the chambers.
Respiratory measures were acquired continuously in the freely behaving animals using the barometric method (Buxco Electronics, Troy, NY; see Refs. 3 and 41). To minimize the long-term effect of signal drift resulting from temperature and pressure changes outside the chamber, a reference chamber of equal size in which temperature was measured using a T-type thermocouple was used. In addition, a correction factor was incorporated in the software routine to account for inspiratory and expiratory barometric asymmetries (12). At least 60 min before the start of each protocol, animals were allowed to acclimate to the chamber, in which humidified air (70–90% relative humidity) was passed through at a rate of 1–2 l/min as appropriate, using a precision-flow pump-reservoir system. Pressure changes in the chamber resulting from the inspiratory and expiratory temperature changes were measured using a high-gain differential pressure transducer (model MP45–1; Validyne). Analog signals were digitized continuously and analyzed on-line by a microcomputer software program (Buxco Electronics). A rejection algorithm was included in the breath-by-breath analysis routine and allowed for accurate rejection of motion-induced artifacts. It is important to point out that the time-decay characteristics of pressure signals may impact on the estimates of volume and that therefore the estimated volumes from our open plethysmographic technique represent qualitative estimates rather than true absolute values (51). V̇T, respiratory frequency, and minute ventilation (V̇E) were computed and then stored for subsequent off-line analysis. Peak HVR (pHVR) was defined as the maximal ventilatory output over three consecutive minutes during the hypoxic challenge, while hypoxic ventilatory decline (HVD) was defined as the nadir ventilation over three consecutive minutes. Hypoxic challenges consisted of 10% oxygen-balance nitrogen for 20 min.
All values are shown as means ± SE, unless indicated otherwise. Ventilatory measures were averaged at 1-min intervals and plotted. Analyses of variance were employed to compare differences in data. Significant comparisons were followed by Newman-Keuls post hoc tests. A P value <0.05 was considered to achieve statistical significance.
Twelve male PAFR(+/+) mice and 12 PAFR(−/−) mice derived from multiple litters were assessed in each experimental group for ventilatory recordings. Animals were studied longitudinally for the duration of the 30 days of IH beginning with day 0 at normoxia and subsequently at defined intervals (7, 14, and 30 days) during the course of the IH exposures. No significant differences between the groups emerged when ventilatory recordings were performed in normoxia at day 7; however, at days 14 and 30, PAFR(+/+) animals displayed a time-dependent increase in normoxic V̇E similar to the long-term ventilatory facilitation (LTVF) induced by IH and previously observed in the rat (48). LTVF was attenuated markedly in PAFR(−/−) animals such that, relative to their baseline ventilatory pattern, PAFR(−/−) animals displayed no significant increase in normoxic V̇E [Fig. 1A; P < 0.001 vs. PAFR(+/+) mice].
At day 0, significant differences in V̇E between the groups were noted during pHVR (P < 0.001, Fig. 2A) and HVD (P < 0.01, Fig. 2D), as previously reported (43). As previously reported, these differences were primarily accounted for by differences in VT in both pHVR and HVD (P < 0.02 and P < 0.03, respectively). However, these differences were no longer present at either 7, 14, or 30 days of IH exposures.
Closer examination of the data collected at 30 days of IH revealed that the significant differences between the PAFR(−/−) and PAFR(+/+) groups during normoxic V̇E (Fig. 3A) were attributable to significant differences in VT (Fig. 3B) without significant changes in ventilatory frequency (Fig. 3C). However, in contrast with day 0, such group differences were not further present during the 20-min hypoxic challenges after 30 days of IH.
The present study demonstrates that PAFR activity is necessary for emergence of the time-dependent increases in normoxic ventilation that occur after long-term exposures to IH in adult male mice (i.e., LTVF). In contrast, PAFR activity does not appear to play a role in other alterations of ventilatory patterning associated with long-term IH exposures such as the robust increases in the magnitude of pHVR and the attenuation of HVD (46, 48). Thus the absence of LTVF in PAFR(−/−) mice suggests that PAFR mediate intrinsic components of the neural plasticity of respiratory centers controlling normoxic ventilation.
The concept that PAF/PAFR activity is involved in neuronal plasticity is not novel and has been explored in several studies using tetanic stimulation-induced LTP as a model of neural plasticity. Pharmacological PAFR blockade attenuates LTP in the rat hippocampal CA1 and dentate gyrus, particularly in the medial perforant pathway (10, 23, 24) and in the medial vestibular nuclei (17). In addition, PAF administration has been shown to enhance excitatory postsynaptic responses (23, 24, 55) and to increase the frequency of spontaneous miniature excitatory postsynaptic potentials (11, 23) in hippocampal neurons. The functional relevance of these experiments was further demonstrated in behavioral studies conducted in a Morris water maze (39, 52). Rats received a unilateral intracaudate injection of mc-PAF (PAFR agonist) or BN-52021 (PAFR antagonist) immediately after training, and retention of the task was assessed 24 h later. Latencies to find the hidden platform in rats injected with mc-PAF were significantly lower than in vehicle controls, whereas the latencies of BN-52021-injected rats were significantly longer (52). Thus, in rostral brain structures, PAFR play a major role in synaptic plasticity phenomena that classically involve recruitment and activation of glutamatergic transmission.
The analogy between such rostral PAFR and glutamate receptor-dependent activity and plasticity and the brain stem is particularly striking in the nucleus of the solitary tract (nTS). Indeed, it is now well established that, within this brain stem region, glutamatergic pathways are required for normal neural transmission in HVR responses (34, 36, 37, 53) and are also activated in nTS plasticity (44). In addition, previous experiments from our laboratory confirmed the contributions of PAF-PAFR activation in HVR (16, 49). This study shows for the first time that long-term IH exposures induce sustained increases in normoxic ventilation in PAFR(+/+) mice (LTVF), and these changes may in fact represent a type of PAFR-dependent metaplasticity (for definition, please refer to Ref. 33) induced by the prolonged IH exposures. It is unclear whether LTVF is mediated by modification of peripheral chemoreceptor tonic activity (42, 44) or by alterations in the overall net neural output of central pattern ventilatory generators. This will require additional experiments, which are clearly beyond the scope of the present study.
Other studies using transgenic mice deficient in the dopamine D2 receptor have demonstrated impaired time-dependent adaptations to chronic sustained hypoxia (CH; see Refs. 18 and 19); however, we have previously shown that IH and CH are inherently different stimuli that produce unique ventilatory effects in rats (48). Despite such intrinsic differences between the two exposure paradigms, notable similarities also exist. Both exposures induce time-dependent increases in the baseline normoxic ventilation (LTVF) and time-dependent increases in pHVR, although the underlying mechanisms for each may differ (48). It is also noteworthy that, in certain models of plasticity, such as phrenic and ventilatory long-term facilitation elicited by a reduced number of recurrent hypoxic-normoxic cycles, the LTVF is undetectable after comparable durations of CH (1, 38). Furthermore, long-term IH exposures have been demonstrated to enhance the expression of both phrenic and ventilatory long-term facilitation in reduced, anesthetized preparations and in awake, freely behaving rats (29–31, 43, 48). Although the mechanism by which long-term IH exposures enhance LTVF remain unclear, evidence for the involvement of central serotoninergic receptors and brain-derived neurotrophic factor (2, 14, 15, 31) as well as involvement of reactive oxygen species (43) have recently emerged. Our findings now implicate the need for PAFR integrity to elicit LTVF after IH.
We should point out that the PAFR(−/−) mice used in these experiments were not conditional knockouts, i.e., the targeted gene ablation was not performed acutely in the adult animal, and therefore fetal ablation of PAFR may have allowed for adaptive mechanisms to be established. Therefore, if no differences in ventilatory responses had emerged between the two groups of mice studied herein, it could be argued that the absence of such differences in the ventilatory responses after long-term IH may reflect fetal cellular and molecular adaptations to PAFR deficiency. Although we cannot definitively rule out that such adaptive phenomena may have occurred, they were clearly insufficient to overcome the LTVF deficits incurred by the genetic ablation of PAFR. However, it can be postulated that the adaptive mechanisms that may underlie the IH-induced modification of both pHVR and HVD after long-term IH are PAFR-independent and do not require PAFR activity, even though PAFR activity is required for the normal expression of HVR in normoxic animals. These discrepancies in pHVR in normoxic and long-term intermittently hypoxic mice are not necessarily in conflict with the putative role of PAFR, given that mechanisms of adaptation required for survival in long-term IH may not be sufficiently induced by acute hypoxia and are thus not obvious in normoxic animals during hypoxic challenges. Similarly, inherent differences in PAFR(+/+) and PAFR(−/−) mice may be obscured by compensatory mechanisms that are PAFR independent and induced by long-term IH.
In summary, PAF has been implicated as playing an important role in the modulation of neural activity in general and in respiratory control in particular. The present study further substantiates that PAFR activation selectively mediates important components of the ventilatory response to hypoxia and is critically involved in metaplastic changes brought about by long-term intermittent hypoxic exposures.
This study was supported by National Institutes of Health (NIH) Grants HL-63912 and HL-69932 and The Commonwealth of Kentucky Research Challenge Trust Fund. S. R. Reeves was supported by NIH summer research stipend NCI R25 CA-44789.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2004 the American Physiological Society