Abstract
Inhibitory 5-HT(1a) receptors are located on serotonin (5-HT) neurons (autoreceptors) as well as neurons of the respiratory network (heteroreceptors). Thus, effects on breathing of 5-HT(1a) agonists, such as (R)-(+)-8-hydroxy-2-(di-N-propylamino) tetralin (8-OH-DPAT), could either be due to decreased firing of 5-HT neurons or direct effects on the respiratory network. Mice in which the transcription factor LMX1B is genetically deleted selectively in Pet1-1-expressing cells (Lmx1b(f/f/p)) essentially have complete absence of central 5-HT neurons, providing a unique opportunity to separate the effect of activation of downstream 5-HT(1a) heteroreceptors from that of autoreceptors. We used rhythmically active medullary slices from wild-type (WT) and Lmx1b(f/f/p) neonatal mice to differentiate autoreceptor versus heteroreceptor effects of 8-OH-DPAT on hypoglossal nerve respiratory output. 8-OH-DPAT transiently increased respiratory burst frequency in Lmx1b(f/f/p) preparations, but not in WT slices. This excitation was abolished when synaptic inhibition was blocked by GABAergic/glycinergic receptor antagonists. Conversely, after 10 min of application, frequency in Lmx1b(f/f/p) slices was not different from baseline, whereas it was significantly depressed in WT slices. In WT mice in vivo, subcutaneous injection of 8-OH-DPAT produced similar biphasic respiratory effects as in Lmx1b(f/f/p) mice. We conclude that 5-HT1a receptor agonists have two competing effects: rapid stimulation of breathing due to excitation of the respiratory network, and delayed inhibition of breathing due to autoreceptor inhibition of 5-HT neurons. The former effect is presumably due to inhibition of inhibitory interneurons embedded in the respiratory network.