Sites and mechanisms of action of neuropeptides on canine gastric motility differ in vivo and in vitro.

Motilin, pentagastrin and substance P (SP), injected intra-arterially into the canine gastric corpus in vivo increased the amplitude of contractions by an action dependent on activation of cholinergic nerves; i.e. atropine or tetrodotoxin (TTX) completely blocked the responses to motilin and pentagastrin and increased the ED50 of SP. TTX and atropine were not equally effective in increasing the ED50 for SP in vivo and the effect of combining them depended on the order of their addition. Both were much more effective than the SP analog D-Pro2, D-Trp7,9 SP (DSP) which appeared to be a weak antagonist of actions dependent on neural activity. In strips from the same region in vitro no receptors dependent on cholinergic nerve activation could be demonstrated for any peptide; i.e., all were atropine- and TTX-insensitive. Motilin, as expected in the absence of such receptors caused no contractile response in vitro. SP, also as predicted, caused contractions suggesting that a smooth muscle receptor, independent of nerve activation was present. However contrary to expectation pentagastrin induced an atropine and TTX-insensitive increase in the amplitude and frequency of contractions. These results show that 1) the most sensitive sites of action of a number of excitatory peptides depend on cholinergic nerve function in vivo; 2) such sites or the nerve activity on which they depend cannot be demonstrated in vitro; 3) SP has an additional site of action on smooth muscle demonstrable in vivo and in vitro, but motilin does not; 4) pentagastrin has only an action dependent on nerve function in vivo, but manifests an action independent of nerve function in vitro. We conclude that sites and mechanisms of action of peptides cannot be assumed to be identical in vivo and in vitro. Actions dependent on nerves are often lost in vitro and not all smooth muscle actions can be demonstrated in vivo.