Ca-dependent K and Cl currents were measured in isolated cells from rat lacrimal glands using the tight-seal whole-cell recording method. Upon application of acetylcholine (ACh), both K and Cl-selective currents were activated. The size of the ACh-activated currents declined after a few minutes of whole-cell recording. The rundown curve was composed of an initial stable period followed by a rather rapid decline. Both the length of the initial plateau and the speed of the falling phase were dependent on cell size and recording pipette resistance. The results suggest that the rundown was due to washout of an unknown cytosolic substance. Another manifestation of washout was an increase in the delay of the response. Plots of the inverse of the delay as a function of time in whole-cell recording showed again an initial plateau and a falling phase, but the stable period lasted less than in amplitude plots. Analysis of the washout time course suggested that the cytosolic substance has a diffusion coefficient of 5.4 x 10(-6) cm2/s, corresponding to a molecular weight of 350. Washed-out cells were insensitive to GTP-gamma-S, but responded normally to an internal application of inositol-trisphosphate (InsP3), introduced through the pipette. Thus, the step of the response which is sensitive to washout is closely related to the production of InsP3. Addition of various exogenous water soluble substances failed to halt washout. Among the inactive substances were GTP (or a combination of Mg and GTP) and small water soluble precursors of InsP3. The results imply that the production of InsP3 by muscarinic agonists in exocrine glands requires the presence of a small molecular weight, water soluble substance. It is suggested that this substance is an unknown co-factor of phospholipase C or of Gp, the GTP binding protein governing the production of InsP3.