Calcium- and cyclic AMP-dependent chloride secretion in human colonic epithelia.

Three stable epithelial cell lines (HCA-7, HCA-7-Col 1 and HCA-7-Col 3) all derived from the same human adenocarcinoma have been cultured on collagen-coated Millipore filters. These epithelial monolayers have been used to record short circuit current (SCC) in response to of secretagogues. Similar monolayers, but grown on plastic dishes, were used for measurements of tissue cyclic AMP. Lysylbradykinin, applied to either side of the monolayers, increased SCC in HCA-7 cells but had little effect on the other two lines. The responses showed rapid desensitization, which could be prevented by cooling to 4 degrees C. Responses to kinin were not significantly attenuated by piroxicam, an inhibitor of cyclo-oxygenase. Other secretagogues, vasoactive intestinal polypeptide (VIP) and carbachol also increased SCC in monolayers. The responses to VIP were greatest in HCA-7-Col 1 monolayers while responses were virtually absent in HCA-7-Col 3. A similar profile was seen with carbachol except that responses of HCA-7 and HCA-7-Col 1 monolayers were more equal. With one exception the responses to VIP and carbachol showed sidedness, acting only from the basolateral side. The effects of the secretagogues were inhibited by piretanide, a loop diuretic, applied basolaterally. It is presumed that SCC responses represent electrogenic chloride secretion. Treatment with forskolin increased SCC in HCA-7 and HCA-7-Col 1 monolayers with little effect in HCA-7-Col 3. Nevertheless cyclic AMP levels were elevated most in HCA-7-Col 3 and least in HCA-7-Col 1 monolayers, in reciprocal relationship to the functional response. A23187 increased SCC when applied to HCA-7 and HCA-7-Col 3 monolayers with little effect on HCA-7-Col 1. The differential responses of the three human cell lines provide unique opportunities to discover the functional responsibilities of entities involved in the chloride secretory process. HCA-7-Col 3 cells which generate high levels of cyclic AMP in response to forskolin but which fail to show a substantial chloride secretory response may be a useful model of some disease conditions.