Cellular mechanism of aminoglycoside tolerance in long-term gentamicin treatment.

In the rat, nephrotoxicity results from uptake of gentamicin at the apical membrane of proximal tubule (PT) cells. However, during continuous gentamicin treatment, the PT epithelium has been shown to recover. The mechanism(s) of cellular recovery and development of tolerance remains unknown. Therefore, we undertook studies designed to characterize cellular adaptations that occur during long-term gentamicin (LTG) treatment. After 19 days of gentamicin treatment, electron microscopy morphological evaluation revealed cellular recovery with an apparent mild decrease in height and number of microvilli. Enzymatic analysis of LTG PT membranes showed that apical and basolateral membranes had essentially returned to normal. Analysis of apical membrane lipid content revealed persistent statistically significant (P < 0.01) elevations in phosphatidylinositol (PI). In vivo immunogold morphological studies and biochemical studies in LTG rats revealed that endocytosis of gentamicin was selectively reduced, whereas the markers of fluid-phase (horseradish peroxidase) and receptor-mediated (beta2-microglobulin) endocytoses were unaffected or increased. Biochemical analysis showed that, although gentamicin binding to apical membranes isolated from LTG rats increased greater than twofold (P < 0.05) over membranes from untreated rats, in vivo cellular uptake, quantified with [3H]gentamicin, was reduced. Western blot analysis of LTG apical membranes and immunofluorescent staining of perfusion-fixed LTG kidneys showed no change in megalin levels or its apical membrane localization. These data imply that recovery of PT cells from and tolerance to LTG treatment involve a selective inhibition of gentamicin uptake across the apical membrane. They indicate that the mediators of gentamicin endocytosis were affected differently: PI levels increased, whereas megalin levels did not change. We conclude that selective inhibition of gentamicin uptake during LTG treatment is not affected by a reduction in PI or megalin levels. We postulate that trafficking of gentamicin and/or gentamicin-containing endocytic structures is reduced in LTG rats, allowing cells to develop tolerance to gentamicin.

Mechanism of ischemia-enhanced aminoglycoside binding and uptake by proximal tubule cells.

Preceding ischemia or concurrent hypotension is known to enhance aminoglycoside nephrotoxicity; however, the underlying mechanisms responsible have not been determined. To investigate the effect of preceding mild ischemia on cellular gentamicin handling, brush-border membrane vesicle binding and in vivo cellular gentamicin uptake were quantified using [3H]gentamicin as a tracer. Fifteen minutes of ischemia resulted in a marked increase in apical membrane gentamicin binding (2.8 +/- 0.4 vs. 4.9 +/- 0.8 nmol/mg protein, P < 0.01). This increase was associated with an increased number of binding sites (3.7 +/- 0.3 vs. 9.1 +/- 2.3 nmol/mg protein, P < 0.01) and a reduced binding affinity (11.8 +/- 2.2 vs. 27.7 +/- 10.4 microM, P < 0.01). This increase in gentamicin binding was accompanied by alterations in apical membrane phospholipids including a doubling of phosphatidylinositol (PI) levels (13.8 +/- 0.4 vs. 27.5 +/- 3.1 nmol/mg protein, P < 0.01). Furthermore, treatment of apical membrane vesicles with PI-specific phospholipase C markedly reduced the difference in gentamicin binding between paired control and ischemic membrane fractions. Increased gentamicin binding was associated with increased in vivo uptake of gentamicin by S1/S2 and S3 cells. Outer cortical uptake of gentamicin increased from 2.18 +/- 0.39 to 2.68 +/- 0.27 nmol/mg protein (P < 0.01) after 15 min of ischemia and 4 h of reperfusion. Juxtamedullary uptake also increased from 1.39 +/- 0.31 to 1.75 +/- 0.12 nmol/mg protein (P < 0.01). Immunocytochemical techniques, utilizing immunogold labeling, showed gentamicin was taken up via the receptor-mediated endocytic pathway by S1/S2 and S3 cells. After ischemic injury gentamicin was localized in abnormal intracellular accumulations in S3 but not S1 or S2 cells. Taken together, these data indicate ischemia results in a marked increase in apical gentamicin binding due to increases in apical PI content. This is associated with increased internalization by S1/S2 and S3 cells and abnormal intracellular compartmentalization of gentamicin within S3 cells.

Cytoskeleton disruption and apical redistribution of proximal tubule Na(+)-K(+)-ATPase during ischemia.

The polar distribution of Na(+)-K(+)-ATPase to the basolateral membrane of proximal tubule cells is essential for the efficient and vectorial reabsorption of Na+ and may be dependent on the formation of a metabolically stable, detergent-insoluble complex of Na(+)-K(+)-ATPase with the actin membrane cytoskeleton. The present studies utilized immunocytochemical techniques to demonstrate and quantify the apical redistribution of Na(+)-K(+)-ATPase during mild ischemia (15 min) that occurred in proximal (1.3 +/- 0.9 vs. 4.5 +/- 1.1 particles/100 microns surface membrane, P less than 0.01) but not distal tubule cells. Treatment of control apical membranes with 2-(2-methoxyethoxy)ethyl 8-(cis-2-n-octylcyclopropyl)octanoate (A2C), a fluidizing agent, markedly increased membrane fluidity without any effect on Na(+)-K(+)-ATPase activity. In brush-border membrane vesicles isolated after ischemia, however, A2C further increased an already elevated Na(+)-K(+)-ATPase activity. During ischemia, total cellular Na(+)-K(+)-ATPase activity remained unaltered, but the Triton X-100-soluble (noncytoskeleton associated) fraction of Na(+)-K(+)-ATPase increased significantly following 15 and 30 min. There was a corresponding decrease in the Triton X-100-insoluble fraction of Na(+)-K(+)-ATPase, with the ratio of detergent-soluble to -insoluble Na(+)-K(+)-ATPase increasing from 13 +/- 2 to 32 +/- 5% (P less than 0.01) during 30 min of ischemia. Western blot analysis of the Triton X-100-soluble fraction, following 30 min of ischemic injury, revealed the presence of Na(+)-K(+)-ATPase, actin, fodrin, and uvomorulin. However, in a fraction highly enriched for Na(+)-K(+)-ATPase, neither actin, fodrin, nor uvomorulin was detected.(ABSTRACT TRUNCATED AT 250 WORDS)

Effect of reversible ATP depletion on tight-junction integrity in LLC-PK1 cells.

To further understand and investigate how ischemia affects the tight junction we have developed a 2-h model of rapidly reversible ATP depletion and cellular injury in confluent LLC-PK1 monolayers. ATP depletion was achieved utilizing substrate-free medium containing 0.1 microM antimycin A (AA). Cellular ATP levels dropped rapidly to less than 5% of control values, but recovery of ATP and cell morphology was possible even after 2 h of exposure to AA. Ruthenium red, an electron-dense marker of tight-junction integrity, was excluded from the tight junctions of control monolayers but penetrated cellular tight junctions during ATP depletion in a duration-dependent manner. Electrical resistance across the monolayers remained unchanged in control monolayers but decreased linearly during ATP depletion to 59% of control values. Transmonolayer movement of [3H]mannitol increased from a control level of 7 to 13.5% during ATP depletion. Recovery of tight-junction integrity was demonstrated by a slowing of [3H]mannitol transfer from the basolateral to the apical medium. The transfer rate in control monolayers was 0.0126%/min. During the initial 120 min of cellular recovery from 2 h of ATP depletion, the transfer rate was 0.0789%/min, but this decreased to 0.0045%/min between 2 and 4 h of recovery. In summary, physiology, biochemical, and morphological evidence indicates that reversible ATP depletion results in rapid opening of cellular tight junctions. After ATP-repletion physiological studies indicate a recovery of tight-junction integrity.

Dissociation and redistribution of Na+,K(+)-ATPase from its surface membrane actin cytoskeletal complex during cellular ATP depletion.

Establishment and maintenance of a polar distribution of Na+,K(+)-ATPase is essential for efficient Na+ reabsorption by proximal tubule cells and is dependent upon the formation of a metabolically stable, detergent-insoluble complex of Na+,K(+)-ATPase with the actin membrane cytoskeleton. The present studies show that cellular ATP depletion results in a rapid duration-dependent dissociation of Na+,K(+)-ATPase from the actin cytoskeleton and redistribution of Na+,K(+)-ATPase to the apical membrane. During ATP depletion, total cellular Na+,K(+)-ATPase activity was unaltered, but the Triton-X-100-insoluble fraction (cytoskeleton associated) of Na+,K(+)-ATPase activity decreased (P less than 0.01), with a corresponding increase in the detergent-soluble fraction of Na+,K(+)-ATPase (P less than 0.01). Indirect immunofluorescent studies of cells with depleted ATP revealed a redistribution of Na+,K(+)-ATPase from the basolateral membrane into the apical membrane and throughout the cytoplasm. ATP depletion also resulted in the redistribution of F-actin from a primarily cortical concentration to a perinuclear location. There was also a rapid, duration-dependent conversion of monomeric G-actin to F-actin starting during the first 5 min of ATP depletion. Taken together, these data suggest that ATP depletion causes profound alterations in cell polarity by inducing major changes in the actin cytoskeletal architecture.