The objective of this study is to determine which inflammatory mediators had direct effects on canine trachealis muscle neuromuscular control to identify candidate mediators of the hyperresponsiveness observed in vitro after O3 exposure. Studies were carried out in the sucrose gap at 29 degrees C and in the muscle bath at 37 degrees C. Leukotriene (LT) B4, LTD4, and prostaglandin (PG) D2 had neither direct nor significant effects on the excitatory junction potentials (EJP's), the secondary membrane potential oscillations, or the associated contractions that followed field stimulation of cholinergic nerves. U 46619, a stable analogue of thromboxane (Tx) A2, enhanced (10(-10)-10(-7) M) the duration and the amplitude of secondary oscillations and associated contractions without affecting the EJP's. In the muscle bath, U 46619 enhanced field-stimulated contractions; this was antagonized competitively by SQ 29548. In both the sucrose gap and the muscle bath, higher concentrations (10(-9) M and higher) caused direct effects, small depolarizations, and contractions. These effects of U 46619 were unaffected by indomethacin or guanethidine but were abolished by SQ 29548, an antagonist selective at TxA2-PGH2 receptors. U 46619 at 10(-9) M did not affect electrical or mechanical responses to acetylcholine and at 10(-9) M did not increase the sensitivity to acetylcholine. Platelet-activating factor (PAF) was inactive in all muscle-bath and most sucrose-gap experiments. In 7 of 20 of the latter, it caused effects qualitatively like those of U 46619, but whether it acted through release of TxA2 could not be tested because of the rapid tachyphylaxis to PAF. We conclude that TxA2 may mediate the hyperresponsiveness found in vitro after O3 treatment.