Methods Studies were performed on tissue obtained from nine adult human kidneys (HK): six cadaver donor kidneys judged unsuitable for transplantation, and normal renal tissue obtained from three surgical nephrectomy samples. All donor kidneys were initially flushed with ice-cold Collin's solution and then perfused via the renal artery with purified protein fraction and immersed in the same solution (in the perfusion apparatus) up until the time of transfer to the laboratory. The surgical samples were obtained immediately following nephrectomy. The data available on the nine patients includes: HK I was from a 51-year-old patient who underwent nephrectomy for a large hematoma involving the middle portion and the upper pole of the kidney; the uninvolved lower pole of the kidney was used. HK 2 was a cadaver donor solitary kidney from a 56-year-old patient who died in an auto accident. HK 3 was a cadaver donor kidney from a 44-year-old patient; the cause of death was not listed. HK 4 was a donor kidney from a 41-year-old patient who died from a stroke. HK 5 was from a 58-year-old patient who underwent nephrectomy for a nonobstructing tumor of the renal pelvis. The renal parenchyma was not involved. HK 6 was a cadaver donor kidney from a 35-year-old patient who also died in an auto accident. HK 7 was a cadaver donor kidney from a 31-year-old patient who died from head injuries in an auto accident. HK 8 was a donor kidney from a 5-year-old patient who died from a traumatic subdural hematoma. HK 9 was from a patient who underwent nephrectomy for a large hypernephroma. An uninvolved piece (as judged by light microscopy on a frozen section) of the renal cortex most distant from the tumor was used.

As soon as the tissue samples became available, they were immersed in ice-cold modified Kreb's buffer (KRB) of the fol-lowing composition: 140 mrvt NaCl, 5 ms KCI, 1.2 mM MgSO4, 2.0 mrvi CaCI2, 10 m glucose, 10 m sodium acetate, 2 mM sodium phosphate, and 20 mM Tris (pH= 7.4), and transferred to the laboratory. The renal capsule was stripped off the parenchyma, the kidney was bisected, and the cortical tissue was dissected with a pair of scissors. All subsequent steps were performed at 4 C. Glomeruli and nonglomerular cortical tissue samples referred to here as" tubules" were prepared by a sieving technique as described [5, 18] with minor modifications. The cortical tissue samples were minced and then pressed through a stainless steel sieve (212 m) with a metal spoon. The resulting suspension was passed in succession through 390-, 250-, and l50-m nylon sieves; glomeruli were collected from the top of the l50-m sieve. The glomerular suspension was then allowed to settle by gravity for 5 mm and the super-natant containing small tubular fragments or portions of glomeruli was aspirated with a Pasteur pipette. Glomeruli were then resuspended in fresh KRB. The glomerular suspension was greater than 95% pure as judged by light microscopy. Unlike rat cortical tissue [5, 7], the human nonglomerular fraction of the renal cortex fragmented into small tubular segments that passed through all the sieves, with relatively little material collecting on top of the sieves. Therefore, all the washings subsequent to the first sieving (212 Im) were collected and com-