Cryopreserved human hepatocytes: characterization of drug-metabolizing enzyme activities and applications in higher throughput screening assays for hepatotoxicity, metabolic stability, and drug-drug interaction potential.

Cryopreserved human hepatocytes were extensively characterized in our laboratory. The post-thaw viability, measured via dye exclusion, ranged from 55 to 83%, for hepatocytes cryopreserved from 17 donors. Post-thaw viability and yield (viable cells per vial) were found to be stable up to the longest storage duration evaluated of 120 days. Drug-metabolizing enzyme activities of the cryopreserved hepatocytes (mean of ten donors) as percentages of the freshly isolated cells were: 97%, for cytochrome P450 isoform (CYP) 1A2, 78% for CYP2A6, 96% for CYP2C9. 86% for CYP2Cl9, 90% for CYP2D6, 164% for CYP3A4, 76% for UDP-glucuronidase, and 88% for umbelliferone sulfotransferase. Known species-differences in 7-ethoxycoumarin (7-EC) metabolism were reproduced by cryopreserved hepatocytes from human, rat, rabbit, dog, and monkey, illustrating the utility of cryopreserved hepatocytes from multiple animal species in the evaluation of species-differences in drug metabolism. Higher throughput screening (HTS) assays were developed using cryopreserved human hepatocytes for hepatotoxicity, metabolic stability, and inhibitory drug-drug interactions. Dose-dependent cytotoxicity, measured using MTT metabolism as an endpoint, was observed for the known hepatotoxic chemicals tamoxifen, clozapine, cadmium chloride, diclofenac, amiodarone, tranylcypromine, precocene II, but not for 2-thiouracil. Cell density- and time-dependent metabolism of 7-EC and dextromethorphan were observed in the HTS assay for metabolic stability. Known CYP isoform-specific inhibitors were evaluated in the HTS assay for inhibitory drug-drug interactions. Furafylline, sulfaphenazole, quinidine, and ketoconazole were found to be specific inhibitors of CYP1A2, CYP2C9, CYP2D6, and CYP3A4, respectively. Tranylcypromine and diethyldithiocarbamate were found to be less specific, with inhibitory effects towards several CYP isoforms, including CYP2A6, CYP2C9, CYP2C19, and CYP2E1. These results suggest that cryopreserved human hepatocytes represent a useful experimental tool for the evaluation of drug metabolism, toxicity, and inhibitory drug-drug interaction potential.

Preparation of precision-cut renal slices and renal proximal tubular fragments for evaluating segment-specific nephrotoxicity.

Previous research in animals and humans has demonstrated that many nephrotoxic chemicals induce selective injury within the kidney affecting either renal proximal straight (PST) or proximal convoluted (PCT) tubules. Selective injury has also been observed following in vitro nephrotoxicant exposure to precision-cut renal slices and isolated PCT and PST segments. These in vitro models provide a means of comparing and contrasting basic mechanistic differences which render these segments innately susceptible to nephrotoxicant injury. In this article, methods for preparing precision-cut slices and isolating PST and PCT segments will be reviewed.

Bulk isolation of renal PCT and PST. I. Glucose-dependent metabolic differences.

A new procedure for separately isolating milligram quantities of rabbit renal proximal straight (PST) or convoluted (PCT) tubules is described, and the differential abilities of these segments to utilize glucose as a metabolic substrate are investigated. Separate dissection of the cortical cortices and the outer medullary stripe, followed by collagenase digestion and discontinuous Percoll centrifugation, provide enriched populations (greater than 98% pure) of PCT (37 mg) and PST (14 mg), respectively, per rabbit. The purity of PCT and PST fractions was quantitated morphologically and by comparing the enriched activity of the proximal tubular marker leucine aminopeptidase and deenriched activity of the distal marker hexokinase to previously published values reported from microdissection studies. To investigate glucose-dependent metabolic differences, PCT and PST suspensions (1 mg/ml) were preincubated in Dulbecco's modified Eagle's-Ham's F-12 medium for 1 h before being incubated for 30 min in buffer with or without glucose as the only available metabolic substrate. In glucose-containing buffer, PST segments maintained their oxygen consumption and ATP contents at levels significantly higher than PCT segments. These differential responses between PST and PCT were glucose-dependent because they were abolished when segments were incubated under glucose-free conditions. Because responses in PCT were glucose-independent, these results suggest that PCT cannot utilize glucose to support oxidative metabolism, whereas PST segments can oxidatively metabolize this substrate. These differences in glucose utilization do not correlate with the distribution of glycolytic enzyme activities, suggesting that differential metabolic regulation of these enzymes may determine the ability of each segment to utilize glucose.

Bulk isolation of renal PCT and PST. II. Differential responses to anoxia or hypoxia.

Innate biochemical responses of rabbit renal proximal convoluted (PCT) and straight (PST) segments following in vitro exposure to anoxia or hypoxia were investigated to delineate the mechanisms responsible for segment-selective injury in vivo. After bulk isolation, suspensions (1 mg/ml) enriched in either PCT or PST were preincubated in Dulbecco's modified Eagle's-Ham's F-12 medium for 1 h before being exposed to either 40 min of anoxia (N2) or 120 min of hypoxia (1% O2) and 1 h of recovery under air-CO2 conditions. After recovery from anoxia, the percent of control values for each viability indicator in PCT and PST, respectively, were as follows: O2 consumption (QO2), 30/50; ATP content, 22/49; K+ content, 60/70; and percent lactate dehydrogenase (LDH) release, 66/45. Likewise, following recovery from hypoxia, the percent of control values for PCT and PST, respectively, were as follows: QO2, 50/90; ATP, 16/57; K+, 52/79; LDH, 45/17. These differential responses indicate that PCT segments were innately more susceptible to anoxic and hypoxic injury than PST segments. Because ATP content was significantly higher in PST segments immediately after anoxia and hypoxia, we investigated glucose-dependent responses during anoxia by exposing these segments to 30 min of anoxia in nutrient buffer with or without glucose. Results from these experiments demonstrate that the PST protection from anoxia was glucose dependent because removal of glucose from the nutrient buffers during anoxia abolishes the differential responses between PCT and PST. The in vitro PCT sensitivity observed here contrasts with the PST sensitivity observed following in vivo ischemia, suggesting that hemodynamic factors present in vivo may ultimately determine the overall susceptibility of PST segments in situ.

Differential patterns of injury to the proximal tubule of renal cortical slices following in vitro exposure to mercuric chloride, potassium dichromate, or hypoxic conditions.

The innate susceptibility of renal cell types to these agents was investigated using precision-cut rabbit renal cortical slices made perpendicular to the cortical-papillary axis. Slices were incubated in DME/F12 medium containing 10 microM, 100 microM, or 1 mM concentrations of either metal for 12 hr or in Krebs-Hepes buffer gassed with nitrogen (100%) for 0.75 to 5 hr of hypoxic exposure. To simulate postischemic reperfusion, some slices were transferred to vessels gassed with oxygen after an initial hypoxic period. Mercuric chloride (100 microM) exposure resulted in damage to the straight regions of proximal tubules by 12 hr leaving convoluted regions unaffected. Hypoxia (2.25 hr) and potassium dichromate (100 microM for 12 hr) both caused injury to the convoluted proximal tubules without affecting straight proximal tubular regions. Mercury concentrations of 10 microM and 1 mM had no effect or injured all cell types within the slice, respectively. Similar results were observed for hypoxic periods less than 1.5 hr or greater than 3 hr of exposure. Potassium dichromate had no measurable affect at 10 microM, but at 1 mM focal lesions were observed after 4 hr of exposure, and by 12 hr all cell types within the slice were affected. Intracellular potassium content normalized to DNA correlated well, but always preceded the pathological lesions observed. These results demonstrate that injury to specific regions of the proximal tubule by these agents relates to an innate susceptibility of the intoxicated cell type independent of physiologic feedback or blood delivery patterns proposed as mechanisms of selective injury from in vivo studies.

Preparation of positional renal slices for study of cell-specific toxicity.

To reduce structural complexity, rabbit kidneys were sliced perpendicular to their cortical-papillary axis to isolate four distinct cell groupings. This positional orientation allows identification of each renal cell type based on its location within the slice. A mechanical slicer was used to make several precision-cut slices rapidly from an oriented cylindrical core of renal tissue, with minimal tissue trauma. Slices were then submerged under a gently circulating oxygenated media in a fritted glass support system that maintains viability (intracellular K+/DNA ratio) and structural integrity (histology) for at least 30 h. A high dose of mercuric chloride (10(-3) M) was used to demonstrate the structural and biochemical changes of intoxicated slices. This method provides a controlled subchronic in vitro system for the study of the individual cell types involved in cell-specific renal toxicities and may also be a useful tool for addressing other pharmacological and physiological research questions.