Reversal by phosphodiesterase-4 blockers of in vitro apnea in the isolated brainstem-spinal cord preparation from newborn rats

Abstract

Ventilation of the lungs is mediated by neurons of the respiratory network in the lower brainstem. The activity of rhythmogenic respiratory network neurons seems to depend greatly on cellular levels of the second messenger cAMP. Accordingly, depression of breathing in (preterm) infants associated with clinical administration of opioids and prostaglandins results likely from a fall of cAMP in these cells caused by Gi/o proteins that are activated via µ-opiate or EP3 prostanoid receptors, respectively. Typically, such drug-induced depression of infant breathing is treated with high doses of methylxanthines that have notable adverse effects. It was the aim of our study to investigate whether clinically applicable blockers of cAMP-hydrolyzing phosphodiesterase-4 counteract the inhibition of the respiratory network associated with a drug-induced fall of cAMP. For this purpose, inspiratory-related cervical nerve activity was measured in isolated brainstem-spinal cord preparations from newborn rats. Respiratory frequency was depressed by >80% (from >5 bursts/min to <1 burst/min) with nociceptin (1 µM) which decreases cAMP via a Gi/o protein-coupled opioid-like receptor. The nociceptin-induced respiratory depression was reversed by the activator of adenylyl cyclase, forskolin (5–25 µM) and the phosphodiesterase-4 blockers rolipram (0.1–1 µM) and RO-201724 (1–5 µM). Blocking phosphodiesterases 3 and 5 with milrinone (25–100 µM) and zaprinast (25–100 µM), respectively, was not effective. The results indicate that phosphodiesterase-4 blockers are strong stimulants of the respiratory network. We hypothesize that these or related agents may be potent tools for a treatment of drug-induced disturbances of breathing in (preterm) infants.

Keywords: Brain stem; cAMP; PDE; pre-Bo¨tzinger complex; Rhythm generation; Rolipram

Breathing in infants is often depressed by opioids and prostaglandin-E, which are administered for analgesia and treat- ment of congenital heart diseases, respectively [2,16,24]. In newborn rat models of the neuronal control of breathing, a very similar depression of inspiratory rhythm by opioids or prostaglandin-E is reversed by agents that increase the cel- lular concentration of cAMP, such as forskolin or agonists for D1 dopamine or 5-HT4 serotonin receptors [2,4,17,27]. This may explain why drug-induced respiratory depression in (preterm) infants as well as spontaneous apneas of prematurity are reversed by methylxanthines only at high doses, at which the agents unspecifically block phosphodiesterase-4 (PDE4) which decreases cellular cAMP levels [6,11]. Administration to infants of methylxanthines, specifically theophylline or caf- feine, is associated with adverse effects including agitation, increased heart rate, exaggerated diuresis, metabolic acidosis or decreased peripheral circulation [2,8,9]. It has not been studied so far whether clinically applicable specific blockers of PDE4, such as rolipram and RO-201724 [5,19,22,26], stimulate the respiratory network effectively and probably with less adverse effects. In the present study, we have tested in brainstem-spinal cord preparations from newborn rats the extent to which these PDE4 blockers reverse the potent depression of in vitro fic- tive breathing related to a fall of cAMP induced by nociceptin [28,29].

Experiments were performed on brainstem-spinal cord preparations from 0- to 3-days-old rats. All procedures were approved by the Animal Welfare Committee of the Univer- sity of Alberta. Under deep (2–3%) isoflurane anaesthesia, the brainstem and spinal cord were isolated at room temperature
[3,4]. The brainstem was decerebrated rostrally between cra- nial nerve root VI and the lower border of the trapezoid body. The spinal cord was cut at the cervical (C6–8) level. After trans- fer to a superfusion chamber (volume 2 ml), the preparations were fixed (ventral side upwards) with insect pins at the ros- tral brainstem and spinal cord level and superfused (flow rate 5 ml/min) at 26 ◦C with a solution containing (mM) 118 NaCl; 3 KCl; 1.5 CaCl2; 1 MgCl2; 25 NaHCO3; 1.2 NaH2PO4 and 30 D-glucose (pH adjusted to 7.4 by gassing with 95% O2, 5% CO2).

Discharge of spinal motoneurons providing inspiratory out- put of the respiratory network in the lower brainstem [3] was recorded extracellularly with suction electrodes from cervical ventral roots (C2–C5). This neuronal population activity was bandpass-filtered (0.3–1 kHz) and integrated. During the exper- iments, nerve activities were displayed on a chart recorder (Win- doGraf, Gould, Ohio, USA) and fed into a computer (sampling rate: 0.1–1 kHz) via a digital recording system (Powerlab, ADIn- struments, Castle Hill, Australia).

Nociceptin (1 µM) was bath-applied for 2 min. In a previous study [29], we have shown that this protocol depresses respi- ratory frequency from >5 to <2 burst/min within 5–10 min after start of the application. Within 60 min following removal of nociceptin from the superfusion chamber, the frequency depression gets more pronounced, resulting in an ‘in vitro apnea’ in ∼50% of cases, with no spontaneous recovery of rhythm during at least 90 min [29]. In the present study, test agents were applied 20 min after removal of nociceptin from the superfusion system. At that time, the progressive block of rhythm by nociceptin that con- tinued to increase in strength after washout (see above) resulted in a mean inspiratory frequency <1 burst/min. Effects on res- piratory frequency and burst pattern were analyzed for 5 min at steady-state of stimulatory drug action that was typically reached within 10 min after start of the application. Maximally three tests were done on a single preparation. The following stock solu- tions were used: nociceptin (1 mM in H2O), forskolin, rolipram (1 mM in dimethylsulfoxide), milrinone and zaprinast (10 mM in dimethylsulfoxide), RO-201724 (1 mM in 1:1 dimethylsul- foxide/ethanol). All drugs were purchased from Sigma–Aldrich (Canada) except RO-201724 obtained from Tocris (Ellisville, USA).

The specific PDE4 blockers rolipram and RO-201724 [7,19,22,26] had a stimulating action very similar to that of forskolin. Upon pronounced depression of the respiratory net- work by nociceptin, rolipram (1 µM) reactivated a rhythm of almost identical frequency and inspiratory burst pattern (Fig. 1B and C). The bar graph in Fig. 3A shows that rolipram significantly reincreased respiratory frequency from 0.31 ± 0.09 (n = 14) to 2.4 ± 0.59 bursts/min (n = 9) already at a concentra- tion of 0.1 µM. Upon application of 0.25 µM rolipram, inspira- tory frequency stabilized at 3.9 ± 0.51 bursts/min (n = 10) while 1 µM of the drug induced 5.5 ± 0.35 bursts/min (n = 12), very similar to control (5.6 ± 0.23 bursts/min, n = 14).

In a different set of experiments, slightly higher concentrations of RO-201724 were necessary to induce a stimulatory action similar to that of rolipram. As summarized in Fig. 3B, 1 µM of the drug significantly reincreased respiratory frequency after nociceptin-induced depression from 0.64 ± 0.22 (n = 8) to 2.9 ± 0.38 (n = 4) bursts/min while 5 µM induced a rhythm at a frequency (4.5 ± 0.45, n = 8) almost identical with control (4.9 ± 0.33, n = 8). At 0.5 µM, the agent did not have a signifi- cant stimulatory effect.

The above results showed that the depressing action of noci- ceptin is effectively reversed by the cAMP-elevating agents forskolin, rolipram and RO-201724. Next, we tested the effects of 25 and 100 µM milrinone, a blocker of PDE3 which also hydrolyzes cAMP [6,11]. Milrinone induced a significant, but minor reincrease of respiratory frequency from 0.31 ± 0.26 (n = 6) to 1.5 ± 0.42 (n = 5) only at a concentration of 100 µM (Figs. 1D and 4A). Also application of zaprinast, a blocker of cGMP-specific PDE5 [6,11], was ineffective to restimulate res- piratory rhythm at either 25 and 100 µM (Fig. 4B).

In three of five preparations, 100 µM of the drug induced irregular nerve bursts with a frequency of <1 burst/min, a burst duration <200 ms and an amplitude notably smaller than that of respiratory bursts. cAMP appears to be a critical cellular factor for rhyth- mic activity of the respiratory network in the lower brainstem which initiates and controls breathing movements. This view is based upon the findings that respiratory rhythm is depressed by neurotranmitters or drugs that lower cAMP via Gi/o pro- teins coupled to, e.g., α2, EP3 prostanoid or µ receptors, while rhythm is augmented when cAMP levels are raised by agents that activate Gs proteins via, e.g., A2, β1, D1 or 5-HT4 recep- tors [1–4,13–15,18,21]. Further, respiratory depression due to µ receptors or prostaglandin-E1 activated EP3 receptors is reversed by 5-HT4 and D1 agonists, respectively [4,17]. Also neurotrans- mitter receptor-independent elevation of cAMP by forskolin or the unselective PDE blocker isobutylmethylxanthine antago- nizes respiratory depression induced by α2, µ or EP3 receptors [1,4,23,27].

In respiratory neurons in vivo, phosphorylation following activation of the protein kinase A pathway by cAMP blocks K+ channels and also affects glutamate, GABAA and glycine receptors [15,21]. It is currently studied, which classes of rhythmogenic respiratory neurons are the primary target for cAMP- modulating agents. A combined in vivo/in vitro study on new- born rats showed that opiates depress inspiratory activity while pre/post-inspiratory rhythm persists [13]. Proposedly [18], the inhibitory opiate effect on inspiratory activity is due to transmis- sion failure from the rhythmogenic parafacial respiratory group [20] to the inspiratory centre, the pre-Bo¨tzinger complex [25]. Nociceptin may depress the in vitro inspiratory rhythm in the brainstem-spinal cord preparation [28,29] via a similar (quan- tal) block of synaptic transmission as it decreases cAMP via activation of an opioid-like receptor [10]. However, it is also possible that nociceptin directly inhibits inspiratory neurons of the pre-Bo¨tzinger complex, which produces inspiratory rhythm in brainstem slices lacking the parafacial respiratory group [18,25].

We found that the PDE4 blockers rolipram and RO-201724 reverse the depressing effects of nociceptin very effectively. The agents exert a potent stimulatory action at nanomolar and low micromolar concentrations, respectively. Thus, rolipram is effective at doses similar to those improving memory in intact rodents [5]. Probably, such concentrations are lower than those causing nausea and emesis as a major side effect in clinical treatment of neuropsychiatric disorders, such as major depres- sive disorder, Alzheimer’s disease or multi-infarct dementia [11,19,22]. As explanation of their action, the specific PDE4 blockers may stimulate respiratory rhythm by reincreasing cAMP levels in rhythmogenic pre-Bo¨tzinger complex neurons and/or antagonizing a putative transmission failure between neu- rons of the parafacial respiratory group and the pre-Bo¨tzinger complex. That the latter process may be indeed important is indi- cated by the observation that cAMP-increasing agents directly stimulate bursting of parafacial respiratory group neurons [1].

Hypothetically, the agents may also either directly affect other PDEs or modulate their activity via the evoked rise of cAMP [6,7,19,22,26]. However, the blockers of PDE3, milrinone, and PDE5, zaprinast, did not reestablish regular respiratory rhythm even at a concentration of 100 µM which is much higher than the dose for a specific action [6,7,11,26]. Since PDE3 expression levels seem to be low in the brain [6], the weak stimulating effect of 100 µM milrinone may be due to cross-reaction with PDE4 as reported for cerebellar neurons [7]. Milrinone may also produce a small increase in cAMP via PDE3, which has similar affinity for cAMP and cGMP although cGMP is poorly hydrolyzed and is reported to rather act as an inhibitor of this enzyme [6,11]. cGMP itself appears to play a minor role as elevating its cellular levels with zaprinast can not reverse the depressing effect of nociceptin.
In conclusion, we provide for the first time evidence that specific PDE4 blockers antagonize respiratory depression asso- ciated with a Gi/o protein induced fall of cAMP. Currently, more specific PDE4 blockers with potentially less adverse effects are being developed which target the four subtypes of the enzyme [19,26]. These novel agents, but possibly also the drugs in the present study, rolipram and RO-201724, may be suited for a treatment of drug-induced or spontaneous ZK-62711 respiratory depres- sion of infancy.