Acquired renal cysts derive from terminally differentiated tubular epithelium in adults as a consequence of increased epithelial cell proliferation, fluid accumulation and extracellular matrix remodelling. To understand better how human epithelial cysts may be initiated and progressively expand, cells from primary cultures of normal human adult renal cortex were dispersed in polymerized type I collagen. The transparent matrix permitted repeated observation by light microscopy of cyst formation from individual renal cells. The cyst cells reacted strongly with distal nephron histochemical markers (cytokeratin antibodies AE1/AE3, epithelial membrane antigen, and Arachis hypogaea lectin) but inconsistently or not at all to markers of proximal tubules (Tetragonolobus purpureas lectin and Phaseolus vulgaris erthroagglutinin lectin). The number of spherical, fluid-filled epithelial cysts that developed in a standardized microscope field quantified cyst initiation. Cyst progression was determined from the increase in the diameter (surface area) of cysts and represents a hyperplastic event. EGF or TGF alpha, were required in serum-free defined medium to cause cysts to develop from individual epithelial cells dispersed in the matrix; insulin was required as a co-factor. The EC50 for EGF was approximately 0.1 ng/ml, and for insulin 1 microgram/ml. Early cultures of normal cortex formed cysts more efficiently when dispersed in collagen matrix than cells passaged several times before suspension in the gel. Agonists of adenylate cyclase (PGE1, AVP, VIP, PTH, forskolin, cholera toxin), methylisobutylxanthine, and 8-Br-cAMP, though incapable of causing cyst formation alone in defined medium, enhanced cyst initiation and progression in the presence of EGF and insulin. Angiotensin II, TNF alpha, beta-estradiol, and pertussis toxin had no effect in the absence or presence of EGF and insulin. Pertussis toxin inhibited cyst initiation and expansion caused by EGF and forskolin but potentiated cyst initiation and expansion caused by EGF and PGE1. Cyst formation and expansion were inhibited by TGF beta 1 and 2-chloroadenosine. Polarized monolayers of human renal cortical cells grown on permeable membranes were used to independently quantify the effects of agonists on the net secretion of solute and water from the basolateral to the apical surface of the cells. PGE1, forskolin, and 8-Br-cAMP stimulated net fluid secretion that was sustained for several days; EGF enhanced forskolin-stimulated fluid secretion. We conclude that the formation and expansion of in vitro cysts derived from solitary human cortex cells depends on the coordinated interplay between cellular proliferation and fluid secretion.(ABSTRACT TRUNCATED AT 400 WORDS)

SUMMARY:We describe an unambiguous and highly reproducible method to measure net fluid secretion across cultured renal epithelial monolayers. The method can also be used to measure the constituents of secreted fluid and the effect of agonists and inhibitors of net solute and fluid transport.

Transepithelial fluid secretion is an important process in the progressive enlargement of certain types of renal cysts. Arginine vasopressin (AVP) increases the rate of cyst formation and expansion in an in vitro model of renal cysts that uses Madin-Darby canine kidney (MDCK) cells grown in a gelled matrix of Type 1 collagen. In this study, it was determined if AVP promoted net fluid secretion by MDCK cells. The rate of volumetric fluid secretion was determined from the net movement of water across epithelial layers of MDCK cells grown on permeable, collagen-coated membranes. AVP in the basolateral medium (but not in apical medium) at concentrations exceeding 10(-9) M caused sustained basolateral to apical transepithelial fluid secretion (approximately 0.6 microL/cm2/h). 1-Desamino-8-D-AVP, a V2 receptor agonist, had a similar effect. The secreted fluid was hyperosmotic compared with the bath (5.7 to 9.7 mosM). Chloride was consistently secreted, but the absolute level in the secreted fluid was variable. Intracellular cAMP content was increased 187% by a 2-h exposure to AVP and 10(-4) M methylisobutylxanthine. Net fluid secretion was augmented by methylisobutylxanthine and theophylline and was inhibited by ouabain, bumetanide, and a sodium-dependent Cl-/HCO3- exchange inhibitor (L-645,695) but was not altered by clonidine, guanabenz, or indomethacin. AVP-induced fluid secretion was not accompanied by a change in transepithelial hydraulic conductivity. It is suggested that AVP stimulates fluid secretion of MDCK epithelial monolayers by activating V2 receptor-mediated adenylate cyclase. The regulation of net fluid secretion by AVP would appear to depend on modulation of solute transport, rather than on water permeability.