A transfected cell model for the renal toxin transporter, rOCT2.

A cDNA for the organic cation transporter (rOCT2) of the rat kidney was inserted into the retroviral plasmid pLXSN. This plasmid was used to stably transfect NIH3T3 cells. The transfected cell line exhibited an enhanced rate of tetraethylammonium (TEA) uptake and efflux compared to wild-type NIH3T3 cells. Uptake of TEA by the transfected cells was markedly reduced upon incubation at 4 degrees C. When the extracellular pH was lowered from 8.1 to 5.9, uptake was also reduced, suggesting inhibition of rOCT2 by extracellular protons. The apparent K(m) for TEA in the transfected cells was 141 microM. The classical organic cation transport inhibitors, cyanine 863 and cimetidine, produced noncompetitive inhibition with apparent Ki values of 0.81 and 198 microM, respectively. Daunomycin, vinblastine, and the deoxyadenosine analogs, 2'-deoxytubercidin and 2-chlorodeoxyadenosine, did not appear to be substrates for rOCT2. However, the anticancer drug, cisplatin, competitively inhibited TEA uptake by rOCT2 with an apparent Ki value of 925 microM, suggesting that rOCT2 may play a role in its renal secretion. In summary, transfected NIH3T3 cells provide a facile system by which this and other organic ion transporters can be studied.

A nucleoside-sensitive organic cation transporter in opossum kidney cells.

Renal secretion of organic cations and anions are pleiotropic, active processes in mammals. Some nucleosides such as deoxyadenosine (dAdo), 2-chlorodeoxyadenosine, and azidothymidine are secreted by human and rodent kidneys. Previous work (J. A. Nelson, J. F. Kuttesch, Jr., and B. H. Herbert. Biochemical Pharmacology 32: 2323-2327, 1983) indicated a role for the classic organic cation transporter (OCT) in the secretion of the dAdo analog, 2'-deoxytubercidin, by mouse kidney. Using [14C]tetraethylammonium bromide ([14C]TEA) as a substrate, we tested several renal cell lines for a nucleoside-sensitive OCT. American opossum kidney proximal tubule cells (OK) express a cimetidine-sensitive and metabolic-dependent ability to efflux TEA. Other classic OCT inhibitors and several nucleosides also inhibit TEA efflux by these cells in a manner reflecting structural specificity for the carrier. Inhibition of OCT by nucleosides is not a universal feature of OCTs, since TEA transport mediated by cloned rat kidney OCT2 in the Xenopus laevis oocyte system was not inhibited by the same nucleosides. In conclusion, OK cells appear to possess an OCT that may also transport some nucleosides by a novel carrier.

The drug verapamil inhibits bystander killing but not cell suicide in thymidine kinase-ganciclovir prodrug-activated gene therapy.

The bystander effect, in which unmodified cells are killed as the result of enzyme-prodrug activation in genetically modified neighboring cells, amplifies the suicide response in a tumor in which only a fraction of the cells are targeted. The drug verapamil (VRP), a calcium channel antagonist that is also used to counteract the multidrug resistance of tumor cells, is shown to inhibit the bystander effect by herpes simplex virus thymidine kinase (HSVtk) enzyme-prodrug therapy with ganciclovir by protecting beta geo marked bystander cells in both in vitro coculture assays and in an in vivo animal tumor model. VRP had no stimulatory or inhibitory effect on the proliferation of CT 26 cells, their tumorigenicity, or prodrug-activated cell death produced by the action of the HSVtk gene. The kinetics of the protection afforded by VRP was time dependent with respect to the time of addition of the prodrug, and protection was ineffective when added two or more days after prodrug administration.

Cysteine conjugate beta-lyase activity in human renal carcinomas.

Cysteine conjugate beta-lyase, an enzyme that converts cysteine S-conjugates to free thiols, pyruvate and ammonia, is normally expressed primarily in the liver and kidney. In theory, this selective distribution affords the opportunity to target thiol-containing drugs to these organs and, perhaps, to tumors derived from them. To assess the potential for delivery of such drugs to kidney-derived tissue, we have used a typical beta-lyase substrate, S-(2-benzothiazolyl)-L-cysteine, to measure the beta-lyase activity in normal and tumor tissue of kidneys removed from patients with renal carcinoma. Although considerable heterogeneity in enzyme activity levels was observed in normal and tumor-derived samples, a high proportion of tumor samples had enzyme activity that was at least 50% of that observed in adjacent normal tissue. Frequently, hypoxanthine-guanine phosphoribosyltransferase activity was observed to be greater in the tumor than in normal tissue. These results may aid in the development of therapy for renal carcinomas.

Enhanced transepithelial flux of cimetidine by Madin-Darby canine kidney cells overexpressing human P-glycoprotein.

Cimetidine has been used as a relatively selective inhibitor of renal organic cation secretion, analogous to the use of probenecid to inhibit organic anion secretion. Many of the substrates for the multidrug transporter P-glycoprotein, which is overexpressed in multidrug-resistant tumor cells, are organic cations. Furthermore, the protein is normally expressed on the apical membranes of proximal tubule cells, the postulated site for active organic cation secretion. To test directly whether P-glycoprotein might serve as a carrier for cimetidine, we measured cimetidine transepithelial movement across Madin-Darby canine kidney cells grown as monolayers on membrane filters. A retrovirally transduced Madin-Darby canine kidney cell line (Madin-Darby canine kidney cells transfected with the human multiple drug resistance 1 cDNA for P-glycoprotein), that expresses the human form of P-glycoprotein on its apical membrane, had an increased capacity to transport cimetidine from the basolateral to apical medium (b-->a) but not in the reverse direction (i.e., a-->b). Qualitatively similar results were observed with daunomycin, a well established substrate for P-glycoprotein. Cellular uptake and energy-dependent efflux experiments further established cimetidine to be a substrate for the human P-glycoprotein. Thus, P-glycoprotein may play a role in the renal secretion of cimetidine and perhaps other organic cations.

Mechanisms of resistance to 6-thioguanine in a murine pancreatic tumor.

PANC02 is a unique experimental animal tumor that fails to respond significantly to any known clinically active antitumor agent. In this regard, the murine ductal adenocarcinoma resembles its human counterpart. To study the mechanism for its intrinsic resistance to 6-thioguanine (TG), we compared the metabolism of the drug in PANC02 and a reference, TG-sensitive adenocarcinoma, CA-755. In comparison with CA-755, PANC02 cells were approximately 6 times less sensitive to TG and CHO cells were 80 times less sensitive in tissue culture. Nevertheless, the incorporation of TG into the DNA of these three cell lines was approximately equal at the lowest concentrations capable of reducing cloning efficiency by 50%, i.e., 3.0-3.8 pmol (dthioGMP)/nmol (dGMP). In mice bearing bilateral implants of CA-755 and PANC02, only CA-755 responded to TG treatment. At various doses used on various schedules, the incorporation of TG into CA-755 DNA readily achieved that observed to be cytotoxic to the cells in vitro, whereas the incorporation into the DNA of PANC02 tumor cells did not. Although the biochemical basis for the poor incorporation of TG into the DNA of PANC02 in vivo is not known, this factor appears to explain the refractoriness of PANC02 as compared with CA-755 to this antitumor antimetabolite.

Intrinsic resistance to anticancer agents in the murine pancreatic adenocarcinoma PANC02.

PANC02 is a ductal adenocarcinoma of the pancreas that is resistant to every known class of clinically active antitumor agent. To study the mechanism(s) underlying the intrinsic drug resistance of this tumor, a mammary adenocarcinoma (CA-755) that also grows in C57/BL mice and is known to be drug sensitive was used for comparison. PANC02 resistance and CA-755 sensitivity to several antitumor agents and to X-ray therapy was confirmed in mice, and PANC02 also demonstrated relative resistance in tissue culture. Relative to Chinese hamster ovary (CHO) and CA-755 cells, PANC02 did not appear to show a higher rate of mutation to drug resistance in culture as based on the 6-thioguanine resistance marker. Although P-glycoprotein characteristic of the multidrug resistance (MDR) phenomenon could be demonstrated at the mRNA level using a sensitive RNAse protection assay, the level of expression found was several orders of magnitude lower than that observed in phenotypic MDR cell lines. Furthermore, quinidine failed to increase the sensitivity of PANC02 cells to Adriamycin under conditions that clearly potentiated the toxicity of the drug to a CHO cell line exhibiting classic MDR traits. The heterogeneity in the distribution of drugs was inferred as being significantly greater in PANC02 versus CA-755 cells in vivo as based on measurements of within-animal, within-tumor variance in the distribution of the marker compounds inulin and antipyrine. Although it may not be the only mechanism involved, this greater intratumor heterogeneity in drug distribution could theoretically play a major role in the intrinsic drug resistance of PANC02 in vivo.

Characterization of the DNA damage in 6-thioguanine-treated cells.

6-Thioguanine (TG) incorporation into DNA has been associated with cytotoxicity and DNA damage in Chinese hamster ovary (CHO) and murine leukemia L1210 cells. According to alkaline elution analysis, single-strand breaks (SSB) occur in both cell types. DNA-protein and interstrand crosslinks are prominent features of TG effects in L1210, CEM, and HL-60 but not CHO cells. To assess which DNA strand experiences SSB in CHO cells, the cells were synchronized by growth to confluence (late G1, S). The cells were then diluted into fresh medium so that they underwent a round of division during a subsequent 16-hr interval. They were treated with TG during this first cell cycle, and mitotic cells were harvested at the end of the first cycle using colcemid. SSB were determined in parental DNA (radiolabeled with thymidine during growth to confluence), TG-containing DNA (radiolabeled with [14C]TG during drug exposure), and daughter DNA (labeled with thymidine during the second cell cycle). SSB occurred in TG-containing DNA late in the second cell cycle after drug exposure and in the DNA synthesized from a TG-DNA template (daughter DNA). This observation is consistent with the known delayed cytotoxicity and chromosomal aberrations seen in CHO cells. The SSB suggest relatively normal elongation of DNA containing TG but altered synthesis and/or ligation from a TG-DNA template. This premise was tested in synchronized CHO cells. The DNA strand incorporating TG elongated naturally; however, DNA elongation was impaired in the cell cycle following TG treatment. The results are consistent with SSB in daughter DNA synthesized from a TG-DNA template due to inability to elongate the newly-synthesized strand.

Selective modulation of antibody response and natural killer cell activity by purine nucleoside analogues.

Analogues that are poor substrates for adenosine deaminase or purine nucleoside phosphorylase may mimic immunodeficiencies associated with the enzyme deficiencies, and their activities may be directed toward selected lymphocyte subpopulations. Four analogues were studied for their effects on primary antibody response to either a T-dependent (sheep erythrocytes) or T-independent (trinitrophenyl-conjugated Escherichia coli lipopolysaccharide) antigen as well as effects on T-cytotoxic and natural killer cell activities in mice. The nucleosides were: an adenosine analogue, tubercidin; two deoxyadenosine analogues, 2-chloro, 2'-deoxyadenosine and 2-fluoroadenine arabinoside-5'-phosphate; and a deoxyguanosine analogue, 9-beta-D-arabinosylguanine. Drugs were given i.p. once daily for 3 consecutive days. Immune responses were determined in spleen cell suspensions 1 day after the last dose. Tubercidin inhibited both T-cytotoxic and natural killer cell activities at doses that did not reduce primary antibody response, whereas the reverse was true for 2-chloro, 2'-deoxyadenosine and 2-fluoroadenine arabinoside-5'-phosphate. At higher doses, T-cytotoxic lymphocytes appeared to be more sensitive than natural killer cells to the deoxyadenosine analogues. 9-beta-D-Arabinosylguanine did not selectively inhibit the immune responses at doses that clearly reduced the yield of spleen lymphocytes. Assuming the analogues mimic endogenous nucleosides, the results suggest that natural killer cells are more sensitive to adenosine than are those cells responsible for primary antibody response, whereas the reverse is true for deoxyadenosine.