Methods. Animals. Male or female Sprague-Dawley (CD strain, Charles River Breeding Labs, Wellesley, Massachu-setts) weighing 50 to 60 g were used. Animals were anesthetized with ether then given lethal doses of sodium pentobarbiturate. Isolation of glomeruli. The kidneys were excised and the cortices cut away from the medullae. These cortical pieces were chopped into millimeter square pieces and passed through a series of steel sieves (WS Tyler, Inc.) with decreasing pore sizes—200 j-pore (60 mesh), 150 p-pore (150 mesh), and 75 p-pore (200 mesh)—with the glomeruli resulting on top of the 200-mesh sieve [31. Previous studies using the same isolation method [2] have shown by scanning electron microscopy that the glomeruli are stripped of their capsules and that the prepara-tions are almost 100% free of any tubular tissue. The capsules and the tubular material become stuck on a larger pore sieve through which the glomeruli pass. Cell culture. The isolated glomeruli were rinsed twice in Hanks' balanced salt solution, buffered with HEPES, pH 7.4 (HBSS), containing antibiotics (penicillin, 100 U/mI, streptomycin, 100 tg/ml, and amphotericin, 0.25 tg/ml), and incubated with trypsin (0.2%) for 20 mm at 37 C followed by an incubation with 0.1% collagenase (189 U/mI, Worthington Diagnostics Systems, Inc.) for 40 mm at 37 C. This procedure loosens up the glomeruli but gives few single cells. After it was washed once in buffered HBSS and antibiotics, the pellet was divided, resuspended in the appropriate media, and plated under the appropriate conditions to permit proliferation of either mesangial or epithelial cells. Polygonal cells were obtained by resuspending the enzyme-treated glomeruli in four 60-mm Petri dishes in a medium (Kl-3T3) consisting of equal parts of:(1), a defined medium (Kl)[4] containing 5% NuSerum (Collaborative Research, Waltham, Massachusetts) and (2), 24-hr conditioned medium taken from cultures of Swiss 3T3 fibroblasts grown in Dulbecco's minimum essential medium plus 10% fetal calf serum. After a week, the predominantly polygonal cells were passaged at low density (102/ml) by mild trypsinization. The dishes were washed twice in calcium-magnesium free HBSS, followed by a trypsin (0.025%)/EDTA (0.5 mM) solution which was immediately poured off. After 5 mm at 37 C, the cells were detached by agitation and plated at a concentration of 102/ml onto collagen gel-coated 100 mm Petri plates in Kl-3T3 medium. The collagen (Flow Lab., Inc.) gel-coated tissue culture plates were prepared as follows: The acid soluble collagen was mixed 8: 1: 1 with lOX RPMI-1640 medium and 0.1 N NaOH. Three to four milliliters of this solution were added per 100 mm Petri dish and allowed to gel at 37 C for 60 mm. Once colonies appeared (—7 days), some subsequently reaching a diameter of 1 cm, their position was marked on the underside of the dish. The cloning efficiency was approximately 10%; however, only about ten clones of varying sizes were picked per dish. This was done by cutting around each of the marked colonies with a hypodermic needle, and placing each in a tube containing 1 ml of 0.2% collagenase. This was incubated at 37 C until the collagen was completely solubilized (about 30 mm). Enough Kl-3T3 medium was then added to fill the tube and the cells were pelleted by centrifugation (x bog for 5 mm). Each cell pellet was resuspended in I ml of the Kl-3T3 medium and plated into one well of a 24-well cluster dish coated with collagen gel. Once the cells reached confluency, they could be passaged into larger dishes—always using the K1-3T3 medium and collagen-coated dishes.