A major shortcoming of renal proximal tubular cells (RPTC) in culture is the gradual modification of their energy metabolism from the oxidative type to the glycolytic type. To test the possible reduction of glycolysis by naturally occurring long-chain fatty acids, RPTC were cultured in a two-chamber system, with albumin-bound palmitate (0.4 mM) added to the basolateral chamber after confluency. Twenty-four hours of contact with palmitate decreased glycolysis by 38% provided that carnitine was present; lactate production was decreased by 38%, and the decrease in glycolysis resulted from a similar decrease of basolateral and apical net uptake of glucose. In contrast to the previously described effect of the nonphysiological oxidative substrate heptanoate, palmitate promoted a long-term decrease in lactate production and sustained excellent cellular growth. After 4 days of contact, decreased glycolysis was maintained even in the absence of carnitine and resulted from a decrease of basolateral uptake only, suggestive of long-term regulation different from the earlier effects. Thus, although cultured RPTC lost their oxidative phenotype, they exhibited a type of regulation (Randle effect) that is found in the oxidative-type but not in the glycolytic-type tissues, therefore unmasking a regulative capacity barely detectable in fresh RPTC. Low Po 2 (50 mmHg in the apical chamber) could be a major cause of elevated glycolysis and could hinder the effects of palmitate.

Among the cellular models used in in vitro renal pharmacotoxicology, isolated kidney tubules, used as suspensions mainly of proximal tubules, offer important advantages. They can be prepared in large amounts under nonsterile conditions within 1-2 h; thus, it is possible to employ a great number of experimental conditions simultaneously and to obtain rapidly many experimental results. Kidney tubules can be prepared from the kidney of many animal species and also from the human kidney; given the very limited availability of healthy human renal tissue, it is therefore possible to choose the most appropriate species for the study of a particular problem encountered in man. Kidney tubules can be used for screening and prevention of nephrotoxic effects and to identify their mechanisms as well as to study the renal metabolism of xenobiotics. When compared with cultured renal cell, a major advantage of kidney tubules is that they remain differentiated. The main limitations of the use of kidney tubules in pharmacotoxicology are (1) the necessity to prepare them as soon as the renal tissue sample is obtained; (2) their limited viability, which is restricted to 2-3 h; (3) the inability to expose them chronically to a potential nephrotoxic drug; (4) the inability to study transepithelial transport; and (5) the uncertainty in the extrapolation to man of the results obtained using animal kidney tubules. These advantages and limitations of the use of human and animal kidney tubules in pharmacotoxicology are illustrated mainly by the results of experiments performed with valproate, an antiepileptic and moderately hyperammonemic agent. The fact that kidney tubules, unlike cultured renal cells, retain key metabolic properties is also shown to be of the utmost importance in detecting certain nephrotoxic effects.

No abstract

Rabbit renal proximal tubular cells, cultured to confluency on a permeable collagen film in a two-chamber system, were exposed for 72 h to various concentrations of the nephrotoxic antibiotic, cephaloridine (CLD). A decrease in cellular proteins, leakage of lactate dehydrogenase and morphological changes appeared at CLD concentrations of 0.1,1.0, and 0.5 mg/ml, respectively. The permeability of the monolayer to Lucifer yellow (LY), a dye that does not cross cell membranes, was increased by 1 or 2 mg/ml but not by lower concentrations of CLD. The large basolateral/apical glucose concentration gradient established by the cells was decreased by CLD. However, the fact that, at the CLD concentration of 1 mg/ml, LY totally equilibrated by diffusion across the monolayer, whereas the injured monolayer was still able to maintain a detectable glucose gradient, shows that damage caused by CLD to the diffusion barrier prevails over that affecting glucose uptake. Consistent with the data in the literature concerning the mechanism of CLD accumulation in renal cells, our results show that CLD was more toxic when it was added at the basolateral than at the apical side of the cultured cells. These results illustrate the advantages of using a two-chamber system of cell culture in nephrotoxicity studies.