Comparison of Human and Rat Hepatocyte Metabolism and Mutagenic Activation of 2-Acety Laminofluorene Kenneth Rudo, William C

Home , Rat

[CANCER RESEARCH 47, 5861-5867, November 15, 1987] Comparison of Human and Rat Hepatocyte Metabolism and Mutagenic Activation of 2-Acety laminofluorene Kenneth Rudo, William C. Meyers, Walter Dauterman, and Robert Langenbach1

Cellular and Genetic Toxicology Branch, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 2 7709 [K. R., K. L.J; Department of Surgery, Duke University Medical Center, Durham, North Carolina 27710 [W. C. MJ; Toxicology Program, North Carolina State University, Raleigh, North Carolina 27695 (W. D.J

ABSTRACT a wide variety of animal species (16-19). These studies have also illustrated species differences in the metabolism and mu A method has been developed to assess the metabolism and mutagenic tagenesis of these compounds. The metabolism and mutagene activation of carcinogens using human and rodent hepatocytes in vitro. A sis of AAF have been studied with animal microsomal, S9, and slicing technique which was especially useful for nonperfusable biopsy and resected surgical human liver tissue was used to prepare the hepa intact hepatocyte preparations (7, 20, 21) and with human tocytes. Metabolites of the model carcinogen 2-acetylaminofluorene subcellular preparations (22-25). Previous studies comparing (AAF) produced by human and rat hepatocytes were similar and consisted subcellular and intact cell metabolism and mutagenicity of primarily of 2-aminofluorene with ring hydroxylated products at the 1-, aromatic amines in animals have illustrated quantitative and to 3-, 5/9-, 7-, and 8-positions produced in addition to /V-hydroxy-AAF. a lesser extent qualitative differences between the two prepara Sulphate and glucuronide conjugates of ring-hydroxylated metabolites tions (26-29). It has been postulated from these experiments and 2-aminofluorene were detected. Metabolism and cell-mediated Sal that intact hepatocytes may simulate the overall metabolic monella mutagenicity illustrated interindividual variation with human process (activation and detoxification) as it occurs in the liver hepatocytes. Levels of metabolism and mntagenesis were generally higher better than cell homogenate preparations (26-29). with human hepatocytes compared to rat hepatocyte results. The in Therefore, the present study focuses on the differences in creased levels of metabolism and mutagenesis of AAF by human hepa tocytes compared to rat hepatocytes probably indicates a different sen AAF metabolism and in its metabolic activation to mutagenic sitivity to hepatocarcinogenic effects of AAF on humans as compared to products by hepatocytes from humans and the rat. A modified rats. Understanding differences and similarities between human and slicing technique (30) for hepatocyte preparation from rat and rodent carcinogen activation capabilities should be useful in the extrap human liver allowed the use of nonperfusable tissue samples olation of rodent carcinogenesis data to humans. for experiments. The Salmonella typhimurium assay has proven to be a sensitive measure of aromatic amine mutagenesis (27, 31) and is used in this study to measure cell-mediated mutagen INTRODUCTION esis of AAF. Levels of metabolism, specific metabolite forma In order to understand the carcinogenesis process in addition tion, and extent of mutagenicity were measured to determine to testing for agents which cause its occurrence, the factors interindividual differences in humans and the extent of species involved in the development of this disease must be delineated. variation. The metabolism and mutagenicity data are discussed Animal models have been widely utilized for mechanistic stud in relation to the rodent to human extrapolation process used ies and as bioassay systems for potential human carcinogens. in the prediction of carcinogenic risk. However, species differences between humans and animals may lead to problems in the quantitative extrapolation of the effect of specific environmental chemicals in the development of MATERIALS AND METHODS cancer in humans. Differences between human and rat enzy Chemicals. [WC]AAF (specific activity, 57.0 fiCi/pmol) was pur matic pathways can affect the rates of formation as well as the chased from Amersham (Arlington Heights, IL). It was found to be levels of reactive metabolites of environmental chemicals. Me >99% radiochemically pure by HPLC. Nonradiolabeled AAF, AF, 1- tabolism to reactive intermediates represents one of the signif OH-AAF, 3-OH-AAF, 5-OH-AAF, 7-OH-AAF, 8-OH-AAF, 9-OH- icant events in the carcinogenic process which can affect the AAF, and N-OH-AAF were purchased from the National Cancer accuracy of animal to human extrapolation (1-9). Therefore, a Institute (Bethesda, MD). Desferal mesylate was obtained from Ciba knowledge of xenobiotic metabolic differences between humans Pharmaceutical Co. (Summit, NJ). 0-Glucuronidase (Type B-10, 100,000 units/mg), ethylene glycol bis(/3-aminoethyl i-iher).v,.V -ti-ira- and test animals should contribute to the reliability of the extrapolation process. However, in cases in which human tis acetic acid, arylsulfatase (33 units/mg, Type VIII), calcium chloride, bovine serum albumin, and r>saccharic-l,4-lactone were purchased sues have been utilized in metabolism studies, a high degree of from Sigma Chemical Co. (St. Louis, MO). Dulbecco's PBS (without individual variation is evident (10-15) and this must also be calcium and magnesium), Eagle's Minimum Essential Medium (with considered in the extrapolation process. out L-glutamine), collagi-nasi-, gentamicin, and HBSS (without calcium Certain aromatic amines, including the model chemical of and magnesium) were obtained from GIBCO (Grand Island, NY). this class, AAF2, have been identified as hepatocarcinogens in HPLC spectral grade ethylacetate, methanol, acetone, diethyl ether, ethanol, and isopropanol were purchased from Burdick and Jackson Received 3/11/87; revised 6/19/87, 8/10/87; accepted 8/18/87. Laboratories, Inc. (Muskegon, MI). Aquasol-2 and Protosol were ob The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in tained from Dupont/New England Nuclear (Boston, MA). Trichloro- accordance with 18 U.S.C. Section 1734 solely to indicate this fact. acetic acid and dimethyl sulfoxide were purchased from J. T. Baker 1To whom requests for reprints should be addressed, at Cellular and Genetic Chemical Co. (Phillipsburg, NJ). All treatments, extractions, evapora Toxicology Branch, National Institute of Environmental Health Sciences, Re tions, and HPLC procedures were carried out under yellow light. search Triangle Park, NC 27709. 2The abbreviations used are: AAF, 2-acetylaminofluorene; AF, 2-aminoflu Isolation of Human and Rat Hepatocytes. Human liver tissue was orene; 1-OH-AAF, l-hydroxy-2-acetylaminofluorene; 3-OH-AAF, 3-hydroxy-2- obtained from nine surgical specimens at Duke University Medical acetyl-aminofluorene; 5/9-OH-AAF, a combination of 5- and 9-hydroxy-2-acet- Center (Durham, NC). Table 1 lists the ages, sex, cause of liver tissue ylaminofluorene; 7-OH-AAF, 7-hydroxy-2-acetylaminofluorene; 8-OH-AAF, 8- hydroxy-2-acetylaminofluorene; N-OH-AAF, AMiydroxy-2-acetylaminofluorene; removal, and viability of liver cells from each individual. Tumor tissue 11l'I ( , high pressure liquid chromatography; HBSS, Hanks' balanced salt solu was separated from normal tissue by the pathology department at Duke tion. University Hospital. Tissue specimens for this study were transported 5861

Table I Liver donor patient data combined, evaporated to dryness under a stream of nitrogen, and stored at â€”80Â°C.Sampleswere reconstituted in methanol for HPLC analyses. of Cell Patient1IIIIIIVVVIVIIVIIIIXAge(yr)304330617131343845SexMaleMaleMaleMaleFemaleFemaleFemaleFemaleFemaleDisease/surgicalprocedureOrgan viability92.266.981.670.779.179.092.181.080.0Aliquots of the organic and aqueous layers were counted in a Packard accident)Rectaldonor (motorcycle Tri Labs 2000CA Liquid Scintillation Counter (United Packard Indus metastasis;partialcarcinoma, liver tries, Downers Grove, IL) to determine the distribution of radioactivity. hepatectomyFibromellarleft lateral Water-soluble AAF metabolites were identified after /3-glucuronidase cancer,liverhepatocellular metastasis; subtotal left re or aryl sulfatase treatment of the aqueous phase. To ensure the removal sectionHemangioma; of all organic-soluble metabolites, a final reextraction of the aqueous partial left hepatec phase with diethyl ether was done before the pH was adjusted to 5.0 tomyAdenocarcinoma with 0.5 N acetic acid. One-half of each aqueous sample (2.5 ml) was livermetastasis; of the colon, treated with /3-glucuronidase (1000 units) while the other half was resectionRetroperitonealsubtotal left cancer.partial and liver treated with aryl sulfatase (250 units) plus 19.2 mg of D-saccharic-1,4- hepatectomyAdenocarcinomaright lactone (30). The samples were mixed thoroughly and were incubated livermetastasis; of the colon, at 37Â°Cfor 24 h. After incubation, the pH of the aryl sulfatase samples domectomyRight rightlateralhepatic mass; subtotal was readjusted to 7.5 with l N NaOH and all samples were extracted resectionOrgan as described above with the organic layer being removed following donor% centrifugation. The samples were then analyzed by HPLC. The amount of AAF bound to cellular macromolecules was deter mined as follows. Bovine serum albumin (10 mg) was added to the to the lab in either 4"( ' Euro-Collins solution or Williams E supple combined enzymatically hydrolyzed aqueous phase containing cell de mented with sucrose within 30-45 min of surgical removal. Tissue bris after the final organic extraction. Both bovine serum albumin and digestion was initiated 2-3 h after removal except in Individuals / and cell debris were precipitated by the addition of 1 ml 20% trichloroacetic IX in which cases unused organ donor tissue was obtained after trans acid to 5 ml of aqueous phase. Each tube was mixed for 2 min and port from Pittsburgh, PA, and resulted in another 3 h added to the centrifuged for 10 min at 1100 x g. The precipitate was washed twice time between removal and isolation. Rat hepatocytes were isolated from with PBS, 4 times with diethyl ether, and 4 times with 95% ethanol at 6- to 10-wk-old male Fischer 344 rats which were sacrificed via CO2 which time the extractable radioactivity was at background levels. The asphyxiation and the livers immediately removed surgically. pellet was solubilized in 1 ml Protosol overnight in a 37'C water bath Hepatocytes were isolated from rat and human livers by a modifica and neutralized with 1 ml 0.5 N HC1. Aquasol-2 (10 ml) cocktail was tion of a slicing technique previously described (30). The tissue cubes added and each sample counted in the liquid scintillation counter. were sliced to a thickness of about 0.5 mm with a sterile microtome High Pressure Liquid Chromatography of AAF Metabolites. The blade. About 5-20 g of liver slices per flask were rinsed 6-9 times for analysis of AAF metabolites and HPLC conditions is a modification of 10 min per rinse in 75 ml phosphate buffered saline containing 0.2 m\\ methods previously described by Smith and Thorgeirsson (21) and ethylene glycol bis(/3-aminoethyl ether)N,N'-tetraacetic acid in a shak modified by Langenbach et al. (30). Metabolite identification was ing water bath (60 rpm) at 37Â°C.The slices were first digested for 20 accomplished by co-chromatography with authentic standards. Samples min and then 90 min followed by a third digestion of 20 min with were chromatographed on a Dupont Model 8800 Chromatograph collagenase (250 units/ml) in 200 ml HBSS containing 0.12 mmol/ml equipped with a UV (254 nm) absorbance detector (Dupont Instru <';i( I. plus 2 ml gentamicin. The cells were filtered through a 100 ments, Wilmington, DE). A flow rate of 1.5 ml/min using a Zorbax Micron Nitex nylon screen to remove undigested tissue. Cells from the C8 column, 4.6 mm (inside diameter) x 15 cm (Dupont Instruments, initial digest were discarded because of low viability and debris, while Wilmington, DE) was used. Solvent A consisted of a mixture of 0.01 cells from the second and third digests were combined and used. M acetic acid and 0.01% desferal mesylate and solvent B was isopro Viability was determined by trypan blue exclusion staining. Hepatocytes panol with 0.01% desferal mesylate. The desferal mesylate prevented were carefully transferred by pipets with large bore tips. Approximately chemisorption of W-hydroxy-AAF to the column (21). The gradient 5-12 million hepatocytes per gram of liver were obtained from each system used and the separation of standard and radiolabeled metabolites preparation for both human and rat. Isolated human hepatocytes pre are shown in Fig. 1. The top profile is the HPLC chromatogram of pared by the perfusion technique (32) were generously supplied by Dr. each standard metabolite. The lower profile represents the radioactivity S. Strom (Duke University Medical Center) and compared for viability collected corresponding to each separate metabolite produced by rat or and mutagenic activation to hepatocytes generated via the slicing tech human hepatocytes. The AAF metabolites identified were the 1-, 3-, 5/ nique. All of the cells were suspended in complete minimum essential 9-, 7-, and 8-hydroxy AAF, N-OH-AAF, and AF. Attempts to separate medium for metabolism studies or in HBSS for mutagenesis studies the 5- and 9-hydroxy-AAF as previously reported (21) were unsuccessful with Salmonella. and therefore these metabolites are combined in subsequent tables. The Extraction of Organic-soluble and Water-soluble Metabolites. Liver 8-OH-AAF has been previously reported as a metabolite of AAF in the cells suspended in media were exposed to [I4C]AAF for 0.25, 0.5, 1.0, rat (33), but the present study is the first HPLC detection of this 2.0, and 4.0 h for the time-course study and for 4 h in all other metabolite. The 8-OH-AAF and 5/9-OH-AAF were clearly separated metabolism experiments. During a 4-h incubation, there was approxi as seen in the radiochromatogram. Eluted fractions were collected mately a 10% decrease in viability of rat and human hepatocytes as directly into scintillation vials with variable collection times: 1 min/ determined by trypan blue exclusion. Cell densities for measuring AAF fraction for the first 2 min, 6 sec/fraction from 2 to 6 min to resolve metabolism by human and rat hepatocytes were 5 million/5 ml in a 50- closely eluting early metabolites, and 30 sec/fraction until the end of ml screw top tube. Each study was conducted with three tubes per cell the Chromatographie procedure. The fractions were counted in a Pack preparation and the tubes were maintained sealed in an atmosphere of ard Tri Carb 2000 CA Liquid Scintillation Counter (Packard Instru 5% CO2 in air at a pH of 7.0-7.4 at 37*C in a shaking water bath (30 ment Co., Downer's Grove, IL) in Aquasol-2 (Dupont/New England oscillations/minute). In the dose-response studies with rat hepatocytes, Nuclear, Boston, MA). Medium controls with [I4C]AAF but not con the concentration of [I4C]AAF was 0.44, 4.4, 44, and 440 nM with taining cells were run in each experiment. Recovery of radioactivity specific activities of 57 /iCi//imol for 0.44 and 4.4 ^M, 6.52 iiC\/tano\ from the column was greater than 90%. for 44 /IM, and 1.16 /iCi//miol for 440 Ã•Ã•M.Theconcentration of AAF S. typhi muri um Assay. The S. typhimurium assay was performed in all other rat and human hepatocyte metabolism experiments was 4.4 according to methods reported by Ames et al. (31) as modified by Hix MM.The incubations were terminated by freeze-thawing of the cells 3 et al. (34) for using isolated hepatocytes. The following were added to times with liquid nitrogen and a 37*C water bath. The pH was adjusted 2 ml of top agar at 45'C: 100 ^1 of a 16-h nutrient broth culture of S. to 6.5 with 0.5 N HC1 and the media extracted 3 times with diethyl typhimurium TA98; the test chemical dissolved in dimethyl-sulfoxide ether. During each extraction, the tubes were vortexed vigorously for 2 (50 *il) and HBSS containing the isolated hepatocytes (1 ml). The min and centrifuged for 20 min at 2200 x g. The organic layers were volumes varied with cell concentration and amount of chemical per

5862

Therefore, all hepatocytes including those from rat liver were prepared by the slicing technique in subsequent studies. The dose responses and cell number responses for human and rat hepatocyte-mediated Salmonella mutagenesis are shown in Figs. 2 and 3. Each Roman Numeral in the figures represents a separate human donor (see Table 1) with results from Indi viduals lll-Vll illustrated because only these cases yielded a sufficient number of cells necessary to do complete dose and cell number response experiments. In addition, mutagenesis data for Individuals /, //, and IX were obtained using 10 ng/ O TJ AF/plate at a cell concentration of 1 x IO6 cells/plate. Hepa Z tocytes from these individuals induced 651, 226, and 481 rev- X O. I DOSE RESPONSE OF HUMAN ANDRAI HEPATOCYTE co MEDIATED MUTAGENESIS (AF-lxlO CELLS/PLATE) 1350

TIME(min.) Fig. 1. Profiles of HPLC chromatogram ( ) and corresponding radioactiv ity chromatogram ( ) of AAF metabolites. Separation via HPLC was achieved under the following condition; Segment 1, 28% isopropanol with 0.01% desferal mesylate:72% 0.01 M acetic acid with 0.01% desferal mesylate for 20 min; Segment 2, 100% isopropanol with 0.01% desferal mesylate for 10 min.

Table 2 Viability and Salmonella mutagenesis comparison ofAF and AAF by hepatocytes prepared by the slicing and perfusion technique Both techniques were performed on liver from each individual in the table. TA98 revenants"

Individual I Individual III Sliced Perfused Sliced Perfused AFconcentration0* jig/Plate1 ^g/plate10 Â»ig/plate25 (ig/plateAAF Fig. 2. Effect of AF concentration and human and rat hepatocyte-mediated '922197182250ND90.147725827927182245815624022373 (25jig/plate)Viabilit/(%)1863203298NO'. mutagenesis of Salmonella TA98. Results are from individual human cases and the average of three rat experiments using 1 x 10* hepatocytes/plate. The revenant values are the average of 2 plates/concentration. The variation between Â°TA98 revenant counts based on 2 plates/point at a cell concentration of 0.5 the 2 averaged revenant values is Â±15%of the stated value. Zero axis values, x 10* hepatocytes/plate. dimethyl sulfoxide controls. '' Dimethyl sulfoxide control for background revenant measurements. c ND, not done. 1750-1 '' Viability determined by trypan blue exclusion procedure. 1350- AAF(25|ig/pJOU 950- â€¢1300 â€¢TOO plate. The agar mix was vortexed gently and poured on Vogel-Bonner Medium E plates and incubated inverted at 37Â°C.The mixture of test chemicals, hepatocytes, and bacteria was preincubated for 30 min at 37Â°C.Each experiment was carried out with duplicate plates per treat Â¿ 500 ment point. A negative control with dimethyl sulfoxide instead of AF or AAF was run as well as a rat liver S9 positive control (using 2- aminoanthracene) to monitor the quality of each assay. Colonies were counted 48 h later. â€¢300w

RESULTS 200 m Cell-mediated Salmonella Mutagenesis. Because preliminary studies indicated that many human liver samples received were not sufficiently perfusable to yield adequate numbers of viable hepatocytes, the possibility of using the slicing technique on CELLS PER PLATE X 10Â° these samples was investigated. The data in Table 2 indicate Fig. 3. Effect of human and rat hepatocyte number on Salmonella TA98 that hepatocytes obtained by the slicing technique (30) and the revenants, a, revenant values for AF at 25 *ig/plate; b, revenant values for AAF perfusion technique (32) from liver samples from the same at 25 Â¿ig/plate.Results are from individual human cases and the average of three rat experiments. The revenant values are the average of 2 plates/cell number. individual were comparable in their viabilities and in their The variation between the 2 averaged revenant values is Â±15%of the stated value. abilities to activate AAF and/or AF to Salmonella mutagens. Zero axis values, dimethyl sulfoxide controls.

5863

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1987 American Association for Cancer Research. HEPATOCYTE METABOLISM AND MUTAGENIC ACTIVATION OF AAF ertants/plate, respectively. As shown in Fig. 2, levels of Sal- TIME COURSE METABOLISM OF AAF monella revenants increased with AF concentration for both 40OO human and rat hepatocytes. Each human case exhibited a higher mutagenic response at 10 and 25 ng/plate than was observed 6000 â€¢3000 with rat hepatocytes, although at the 1 fig/plate dose the differ ences were smaller. In Fig. 3, a and Â¿>,thecell number responses 4000 20OO for AF and AAF, respectively, indicate that at all cell concen trations, the mutagenic response was higher with human he patocytes than with rat hepatocytes. In Figs. 2 and 3, the -1000 variation in the ability to activate the aromatic amine by the different human samples is apparent. However, rat hepatocytes 3000 prepared from different animals exhibited a relatively consistent level of mutagenic activity. For example, in studies at 10 /ug/ 2000 ! plate AF at a cell concentration of 1 x 106cells/plate, the range for eight human samples was about 6-fold (198-1225 rever- IOOO tants/plate) with the variation by rat hepatocytes being less than 1.5 fold. Metabolism of AAF. To determine concentration and times 600 for AAF metabolism studies, dose-response studies with rat hepatocytes and a time course experiment with human and rat hepatocytes were conducted. In Table 3, the overall percentage of metabolism and percentage of metabolism to water-soluble ZOO products by rat hepatocytes were measured at four AAF con centrations. With increasing AAF levels the percentage of dose metabolized decreased. In addition, the ratio of percentage of 120 240 15 30 60 120 24O water-soluble:percentage of organic-soluble metabolism de TIME(minutes) creased with increasing AAF concentration. Fig. 4. Effect of time on the metabolism of AAF by human and rat hepatocytes. Results are from individual human hepatocyte cases (HL) IV, V.and VI and from Fig. 4 illustrates the time-dependent rat and human hepato- the average of three rat experiments. Ether-extractable and water-soluble metab cyte metabolism of AAF separated into different metabolite olite values represent the sum of individual metabolites for each group. C,-OH- AAF values represent the sum of all carbon-ring hydroxylated metabolites. AF, fractions. Rat and human hepatocytes showed an increase in N-OH-AAF, and d-OH-AAF values represent the sum of ether-extractable the amount of organic-soluble metabolite formation with time; products and glucuronide plus sulfate-conjugated products. All experiments were however, hepatocytes from Humans (HL) IV and V demon conducted for 4 h at a [14C]-AAFconcentration of 4 nM using 5x10' hepatocytes strate peaks in ether extractable metabolites at 1 and 2 h, in 5 ml medium with 3 flasks/experimental point. respectively (Fig. 4a). Levels of human hepatocyte organic- soluble metabolite formation from these three cases were higher Table 4 Metabolism by human and rat hepatocytes All experiments were done at 4 n\i | "( | V\ I for 4 h in flasks containing 5 x at all time points than in the rat. Both rat and human hepato 1d" hepatocytes in 5 ml medium. cytes also showed a time-dependent increase for water-soluble Human" metabolites and for covalently bound products (Fig. 4, b and c). Rat" Despite lower metabolism, rat hepatocytes had equal or higher II III IV VI VII VIII % of total AAF 49.4 36.4 45.9 19.8 42.5 38.4 40.1 31.8 19.6 Â±3.3 levels of covalent binding of AAF than the three human cases. metabolized* Total AF production increased with time although human % of water-soluble 45.4 23.2 28.0 7.3 16.5 19.6 8.2 19.8 8.1 Â±0.77 hepatocytes deacetylated AAF to a greater extent than rat metabolism hepatocytes (Fig. 4d). CX-OH-AAF (summation of all carbon " Each human hepatocyte value represents 1 experiment and rat hepatocyte values represent the average of 3 experiments. hydroxylated metabolites) formation exhibited variability h The percentage of AAF metabolized and percentage of water-soluble metab among human hepatocytes with peaks at l h for Individual IV, olism were based on the initial amount of [I4C]AAF added. The total AAF 2 h for Individual V, and an increase over the 4-h time period metabolized represents the percentage of organic-soluble and water-soluble me for Individual f7and by rat hepatocytes. N-OH-AAF formation tabolism. varied in the three different human hepatocyte preparations, was lower in the rat hepatocytes than in Individual V and VI, with 4 UMAAF was measured at 4 h. In Table 4, total metab but was similar to the hepatocytes from Individual IV. Based olism by hepatocytes from each individual human case and the on the results from the dose-response (Table 3) and time-course average of 3 experiments for the rat are shown. Hepatocytes experiments (Fig. 4), human and rat hepatocyte metabolism from Individual / had the highest overall metabolism (49.4%) of AAF and highest metabolism to water-soluble products Table 3 Dose response metabolism of AAF by rat hepatocytes (45.4%) with Individual IV exhibiting the lowest metabolism All experiments were 4-h incubations containing 5x10* hepatocytes in 5 ml (19.8% and 7.3%, respectively). The range of conjugative me medium/flask with 3 flasks/dose. tabolism by human hepatocytes was greater (7.3-45.4%) than Concentration0Â»M)0.44 of total AAF of water-soluble the range of overall metabolism (19.8-49.4%). Except for In metabolized"26.7 metabolism20.2 dividual IV, human hepatocyte metabolism of AAF to organic and water-soluble products was greater than rat hepatocyte 4.4 19.6 8.1 44 9.9 3.2 metabolism. 440% 3.3% 0.50 The ranges of specific metabolite levels formed by hepato Â°The percentage of total AAF metabolized and percentage of water-soluble metabolism were based on the initial amount of [MC]AAF added. Total percentage cytes from the eight human cases and the rat are shown in of AAF metabolized includes the percentage of organic-soluble and the percentage Table 5. The major organic-soluble metabolites formed by both of water-soluble metabolism. human and rat hepatocytes were AF, 7-OH-AAF, and the 5864

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1987 American Association for Cancer Research. HEPATOCYTE METABOLISM AND MUTAGENIC ACTIVATION OF AAF unknown metabolite(s). The primary glucuronide- and sulfate- studies because of their advantages when compared to isolated conjugated product formed by human and rat hepatocytes was subcellular preparations (20, 27, 34, 35). Intact hepatocytes the 7-OH-AAF although sulfate and glucuronide conjugates catalyze Phase I and Phase II reactions at rates and in the were formed with 8-OH-AAF, 5/9-OH-AAF, 1/3-OH-AAF, N- sequence that they occur in vivo (28), as well as preserving their OH-AAF, and AF. Levels of organic-soluble and water-soluble metabolizing capabilities for a longer period of time than sub- N-OH-AAF formation were low compared to other metabolites cellular fractions (36). Differences in levels of mutagenic activ formed by both human and rat hepatocytes. AF formation was ity are also apparent between subcellular fractions and intact highest in Individuals V, VI, and VU and lowest in Individuals hepatocytes (27, 34). Levels of AF and AAF mutagenic activity / and VIII (data not shown). It is interesting that while total previously measured (20, 23, 30, 35, 37-39 and data not shown) metabolism or water-soluble metabolism with human hepato are higher in S9 preparations than intact hepatocytes from cytes varied 2.5- and 6-fold (Table 4), respectively, the range of humans and rats possibly due to a lack of detoxifying pathways individual metabolites such as AF (organic-soluble) varied 35- that are present in intact hepatocytes but not in subcellular fold while glucuronide and sulfate conjugates of 7-OH-AAF preparations. Previous investigation of AAF metabolism by varied 30- and 36-fold, respectively, in eight human cases (Table human liver microsomes (23, 40) did not detect AF, although 5). deacetylase activity has been measured in microsomal fractions from guinea pigs, dogs, hamsters, mice, and rats (41) and with DISCUSSION intact hepatocytes in the present study. Other studies have illustrated extensive species differences in the metabolism and In the present study, human hepatocytes were compared to mutagenesis of many environmental compounds (7-10, 30, 38, rat hepatocytes for their ability to metabolize AAF and AF and 39, 42-45) and these results also may be useful in the extrapo to gain insight into interindividual variations in the human lation of animal data to represent hazard levels in humans (45). population for AAF metabolism. The slicing technique used to In the present study, rat and human hepatocytes metabolized prepare hepatocytes in the present study is valuable because it AAF in a qualitatively similar fashion. The major metabolite utilizes samples from biopsy or resected tissue that are too produced by rat and human hepatocytes was AF. Evidence of small or are otherwise not perfusable. In six of nine cases in the importance of AF as an AAF metabolite is provided by our this study, the tissues obtained were not perfusable. The slicing findings that levels of AF mutagenesis were substantially higher technique therefore allows studies to be done on a more frequent than with AAF and that AF was the major metabolite formed basis with human hepatocytes and yields hepatocytes with a while N-OH-AAF, which is a mutagenically active metabolite viability and mutagenic activity comparable to hepatocytes pre isolated primarily in microsomal preparations (21, 23, 40, 46) pared via the perfusion technique (Table 2). was a minor metabolite in both human and rat hepatocytes. It Intact hepatocytes are useful in metabolism and genotoxicity should be noted, however, that in hepatic microsomal prepa rations N-OH-AAF was more mutagenically active than AF Table 5 Metabolism of AAF to ether-extractable products and glucuronide and (47). Because AF may be further metabolized (48, 49) and N- sulfate-conjugated metabolites by human and rat hepatocytes OH-AAF can be metabolically changed primarily via deacety- All experiments were done using 4 nM [MC]AAF for 4 h. Hepatocytes were incubated with 5 x IO6cells in 5 ml medium. lation to a reactive form (18, 38, 50), the relative contribution (pumi.1 of these two compounds to the mutagenic process cannot be MetaboliteUnknownEther-extractableGlucuronideSulfate7-OH-AAFEther-extractableGlucuronideSulfate8-OH-AAFEther-extractableGlucuronideSulfate5/9-OH-AAFEther-extractableGlucuronideSulfate1/3-OH-AAFEther-extractableGlucuronideSulfateN-OH-AAFEther-extractableGlucuronideSulfateAFEther-extractableGlucuronideSulfateHuman(pmol/assay)"43-89328-42618-53736-87727-98117-58077-35823-16918-18077-47810-547-7186-50825-17517-17551-44322-28419-312490-7824261-2552205-2213Ratassay)*551-70482-10887-192126-506148-315196-32839-14147-13774-7838-7914-5314-42144-38621-2418-2819-6120-2917-762025-3683399-774493-852clearly established. The C-hydroxylation of AAF by human and rat hepatocytes represents a step in the detoxification process. In addition to the normal C-hydroxylated products the tentative identification of 8-OH-AAF was reported in rat hepatocytes in this study and previously (30) and in human hepatocytes for the first time. The 8-OH-AAF had also been detected in rat urine (33). The unknown peak in the HPLC chromatogram and 7-OH-AAF were the major organic-soluble C-hydroxylated metabolites with the 5/9-OH-AAF, 3-OH-AAF, 1-OH-AAF, and 8-OH- AAF appearing in roughly equivalent levels in hepatocytes from human individuals. The 7-OH-AAF was the primary glucuro nide and sulfate conjugate formed by human and rat hepato cytes. It has been previously reported that C-hydroxylated glu- curonides are the only detectable water soluble metabolites formed in rat hepatocytes (7). In the present study both conju gates were detected using longer enzymatic times than in pre vious sulfatase and 0-glucuronidase hydrolysis experiments (51). AF-conjugated products were also identified (Table 5) based on the extractable radioactivity after sulfatase and ,Â¡- glucuronidase treatment, but the structural nature of the con jugates could not be positively identified. Quantitative differences in rat and human hepatocyte metab olism were evident when comparing levels of overall organic- soluble and water-soluble metabolite formation with human hepatocyte metabolism higher than that by rat hepatocytes. " Ranges of human hepatocyte metabolites from 8 individuals. This was also true for water-soluble metabolite formation in 7 * Ranges of rat hepatocyte metabolites from 3 animals. of 8 human cases in which the specific metabolite levels, espe- 5865

cially AF, 7-OH-AAF, and 5/9-OH-AAF, were appreciably assays can be useful for risk assessment of environmental chem higher in humans than rats. However, there was a higher icals to humans. conversion of total water-soluble metabolites by rat hepatocytes to glucuronide and sulfate conjugates (>85%) than by human hepatocytes (40-67%), although humans exhibited a higher REFERENCES overall water-soluble metabolite level. Although it was not 1. Glowinski, I. B., Radtke. H. E.. and Weber, W. W. Genetic variation in JV- acetylation of carcinogenic arylamines by human and rat liver. Mol. Phar- investigated in the present study, a possible reason for this macol., 14: 940-949, 1978. discrepancy is that more glutathione and amino acid conjugates 2. Drayer, D. E., and Reidenberg, M. M. Clinical consequences of polymorphic of AAF metabolites are produced by human hepatocytes than acetylation of basic drugs. Clin. Pharmacol. Ther., 22: 251-258, 1977. 3. Boobis, A. R., and Davies, D. S. Human cytochromes P-450. Xenobiotica, are produced by rat hepatocytes. The higher level of such 14: 151-185, 1984. conjugates produced by human hepatocytes may indicate a 4. Boobis, A. R., Murray, S., Kahn, G. C., Roberti, G. M., and Davies, D. S. species difference between humans and rats in the sensitivity Substrate specificity of the form of cytochrome P-450 catalyzing the 4- hydroxylation of debrisoquine in man. Mol. Pharmacol., 23:474-481, 1983. and susceptibility to the hepatocarcinogenicity of aromatic 5. Otton, S. V., Inaba, T., Manon, W. A., and Kalow, W. In vitro metabolism amines. Rat, guinea pig, and hamster differences in hepatocar- of sparteine by human liver: competitive inhibition by debrisoquine. Can. J. cinogenic susceptibility have been reported (18, 19) and Holme Physiol. Pharmacol., 60: 102-105, 1982. 6. Irving, C. C. Species and tissue variations in the metabolic activation of and co-workers (7) have illustrated differences in the ability of aromatic amines. In: A. C. Griffin and C. R. Shaw (eds.), Carcinogenesis: these species to detoxify AAF. Identification and Mechanisms of Action, pp. 211-227. New York: Raven Press, 1979. Another factor to be considered when extrapolating animal 7. Holme, J. A., Trygg, B., and Soderlund, E. Species differences in the data to humans is the wide degree of interindividual variation metabolism of 2-acetylaminofluorene by hepatocytes in primary monolayer in humans. The extent of interindividual variation in humans culture. Cancer Res., 46: 1627-1632, 1986. 8. Gillette, J. R. Solvable and unsolvable problems in extrapolating toxicological is well documented with the emphasis divided on factors related data between animal species and strains. In:}. R. Mitchell and M. G. Horning to environment or heredity (9, 52, 53). Genetic factors, nutri (eds.). Drug Metabolism and Drug Toxicity. pp. 237-260. New York: Raven tion, drug intake, cigarette smoking, and exposure to pollutants Press, 1984. 9. Idle, J. R., and Ritchie, J. C. Probing genetically variable carcinogen metab are but a few of the factors that modify the way human enzy olism using drugs. In: C. C. Harris and H. Autrup (eds.). Human Carcino matic systems and in particular the cytochrome P-450 system genesis, pp. 857-881. New York: Academic Press, 1983. metabolize foreign and endogenous substances (52, 54-56). The 10. Vahakangas, K., Autrup, H., and Harris, C. C. Interindividual variation in carcinogen metabolism, DNA damage and DNA repair. In: A. Berlin, N. metabolic polymorphisms and substrate specificities of certain Draper, K. Hemminki, and H. Vainio (eds.). Monitoring Human Exposure P-450 isozymes in the hydroxylation of compounds such as to Carcinogenic and Mutagenic Agents. IARC Scientific Publication no. 59: 85-98. Lyon, France: International Agency for Research on Cancer, 1984. debrisoquine and spancine (4, 5, 9, 37), populations divided 11. Booth, S. C., Bosenberg, H., Garner, R. C., Hertzog, P. J., and Norpoth, K. into slow and fast acetylators (38, 57), and findings that have The activation of aflatoxin B, in liver slices and in bacterial mutagenicity identified multiple P-450s all attest to a widespread variation assays using livers from different species including man. Carcinogenesis (Lond.), 2:1063-1068, 1981. in humans. Such variation may not be seen in inbred laboratory 12. Harris, C. C. Chemical carcinogenesis and experimental models using human animals or animals maintained under identical environmental tissues. Beitr. Pathol., 75Â«:389-404, 1976. 13. Harris, C. C., Autrup, H., Stoner, G. D., McDowell, E. M., Trump, B. F., conditions. and Schafer, P. Metabolism of dimethylnitrosamine and 1,2-dimethylhydra- The variability in individual AAF metabolites by human zine in cultured human bronchi. Cancer Res., 37: 2309-2311, 1977. hepatocytes probably results from the polymorphism of P-450 14. Harris, C. C., Grafstrom, R. C., Lechner. J. F., and Autrup, H. Metabolism of /V-nitrosamines and repair of DNA damage in cultured human tissues and isozymes. It has been shown that in human liver microsomes cells. In: P. N. Magee (ed.), Nitrosamines and Human Cancer, pp. 121-139. two phenotypes, slow and fast metabolizers, exist in the metab New York: Cold Spring Harbor. 1982. olism of AAF (22). However, the extent of the variability of 15. Stoner, G. D., Harris, C. C., Autrup, H., Trump. B. F., Kingsburg, E. W., and Myers, G. A. Explant culture of human peripheral lung. 1. Metabolism human hepatocyte-mediated metabolism and mutagenesis of of benzo(a)pyrene. Lab. Invest., 38: 685-692, 1978. AAF was markedly lower than the variation reported in pre 16. Weisburger, J. H., and Weisburger, E. K. Biochemical formation and phar macological, toxicological and pathological properties of hydroxylamines vious studies of human metabolism of other compounds by and hydroxamic acids. Pharmacol. Rev., 25: 1-66, 1973. various tissues (11-15, 58). This lower variability in the present 17. Kriek, E. Carcinogenesis by aromatic amines. Biochem. Biophys. Acta., 355: study may be due to the consistent use of fresh suspensions of 177-203, 1974. 18. Miller, E. C., and Miller, J. A. Search for the ultimate chemical carcinogens hepatocytes from surgical tissue instead of cultured cells from and their reactions with cellular macromolecules. Cancer (Phila.), 47: 2327- autopsy samples or from expiant cultures which were utilized 2345, 1981. 19. Miller, E. C., Miller, J. A., and Enomoto, M. The comparative carcinogen- in many of these previous studies (10, 11, 59). The human icilies of 2-acetylaminofluorene and its \ Indri>\> metabolite in mice, ham variability in deacetylase activity is interesting as it reflects sters, and guinea pigs. Cancer Res.. 24: 2018-2031, 1964. individual polymorphism and the possibility of phenotypic ac 20. Dybing, E.. Soderlund. E., Haug, L. T., and Thorgeirsson. S. S. Metabolism and activation of 2-acetylaminofluorene in isolated rat hepatocytes. Cancer tivity comparable to slow versus fast acetylator populations (38, Res.. 39: 3268-3275. 1979. 57). The high levels of AF produced by human and rat hepa 21. Smith, C. L.. and Thorgeirsson, S. S. An improved high-pressure liquid Chromatographie assay for 2-acetylaminofluorene and eight of its metabolites. tocytes suggest evidence for the role of deacetylation in the Anal. Biochem.. /13:62-67. 1981. formation of genotoxic AAF products in rats and humans. A 22. Minchin, R. F., McManus. M. E.. Boobis. A. R., Davies, D. S., and Thor possible relationship between metabolism to AF and AAF geirsson. S. S. Polymorphic metabolism of the carcinogen 2-acetylaminoflu mutagenesis was seen in human hepatocytes but was not evident orene in human liver microsomes. Carcinogenesis (I orni.). 6: 1721-1724, 1985. with rat hepatocytes (data not shown). This possible difference, 23. Dybing, E., vonBahr, C., Aune, T., Glaumann, H., Levitt, D. S., and Thor combined with the higher levels of AAF mutagenesis and me geirsson. S. S. In vitro metabolism and activation of carcinogenic aromatic amines by subcellular fractions of human liver. Cancer Res., 39:4206-4211, tabolism by human hepatocytes, may alter the direct extrapo 1979. lation of rat carcinogenesis data as far as an effect from aromatic 24. Hammons. G. J.. Guengerich, F. P., Weis, C. C., Beland, F. A., and Kadlubar. F. F. Metabolic oxidation of carcinogenic arylamines by rat, dog, and human amines in humans would be concerned. hepatic microsomes and by purified flavin-containing cytochrome P-450 The current study shows the practicality and utility of inves monooxygenases. Cancer Res.. 45: 3578-3585, 1985. tigating the metabolism and activation of carcinogens with 25. Guengerich, F. P. Isolation and purification of cytochrome P-450. and the human tissues. Such data when used in combination with rodent existence of multiple forms. Pharmacol. Ther., 6: 99-121, 1979. 26. Bigger. C. A. H.. Tomaszewski, J. E.. Dipple, A., and Lake, R. S. Limitation bioassay data and other results from frequently used short-term of metabolic activation systems used with in vitro tests for carcinogens. 5866

Science (Wash. DC), 209: 503-505, 1980. genotoxic effects of 2-acetylaminofluorene and its primary metabolites 2- 27. Langenbach, K.. and Oglesby, I.. The use of inlacl cellular activation systems aminofluorene and ÃV-OH-2-acetylaminofluorene. Carcinogenesis (Lond.), 6: in genetic toxicology assays, Â¡n:F. deSerres (ed.). Chemical Mutagens. pp. 421-425, 1985. 55-93. New York: Plenum Publishing Corp., 1983. 43. Neis, J. M., Yap, S. H., vanGemert, P. J. L., Roelofs, H. M. J., Bos, R. P., 28. Billings, R. E., McMahon, R. E., Ashmore, J., and Wagles, S. R. The and Henderson, P. T. Activation of mutagens by hepatocytes and liver 9000 metabolism of drugs in isolated rat hepatocytes. A comparison with in vivo x g supernatant from human origin in the Salmonella typhimurium muta drug metabolism and drug metabolism in subcellular liver fractions. Drug. genicity assay. Comparisons with rat liver preparations. MutÃ¢t.Res., 164: Metab. Dispos., 5: 518-526, 1977. 41-51, 1986. 29. Schmeltz, I., Tosk, J., and Williams, G. M. Comparison of the metabolic 44. Weber, W. W. Genetic variability and extrapolation from animals to man: profiles of benzo(a)pyrene obtained from primary cell cultures and subcellular some perspectives on susceptibility to chemical Carcinogenesis from aromatic fractions derived from normal and methyl cholanthrene-induced rat liver. amines. Environ. Health Perspect., 22: 141-144, 1978. Cancer Lett., 5: 81-89, 1978. 45. Harris, C. C. Human tissues and cells in Carcinogenesis research. Cancer 30. Langenbach. R.. Rudo, K., Ellis, S., Mix, C., and Nesnow, S. Species Res., 47: 1-10, 1987. differences in bladder and liver cell activation of 2-acetylaminofluorene. Cell 46. McManus, M. E., Minchin, R. F., Sanderson, N. D., Wirth, P. J., and Biol. Toxicol.. 3: 303-319, 1987. Thorgeirsson, S. S. Kinetic evidence for the involvement of multiple forms 31. Ames, B. N., McCann, J., and Yamasaki, E. Methods for detecting carcino of human liver cytochrome P-450 in the metabolism of acetylaminofluorene. gens and mutagens with 5a/mon?//a/mammalian microsome mutagenicity Carcinogenesis (Lond.), 4:693-697, 1983. test. MutÃ¢t.Res., 31: 347-369, 1975. 47. Robertson, I. G. C. The importance of 2-aminofluorene in the mutagenic 32. Strom, S. C., Jirtle, R. L., Jones, R. S.. Novicki, D. L., Rosenberg. M. R., activation of 2-acetylaminofluorene. MutÃ¢t.Res., 775: 153-157, 1986. Novotny, A., Irons, G., McLain, J. R., and Michalopoulos, G. Isolation, 48. Vanderslice, R. R., Boyd, J. A.. Eling, T. E., and Philpot, R. M. The culture and transplantation of human hepatocytes. J. Nati. Cancer Inst., 68: cytochrome P-450 monooxygenase system of rabbit bladder mucosa: enzyme 771-778, 1982. components and isozyme 5-dependent metabolism of 2-aminofluorene. Can 33. Weisburger, J. H., Weisburger, E. K.. Morris. H. P., and Sober, H. A. cer Res., 45: 5851-5858, 1985. Chromatographie separation of some metabolites of the carcinogen N-2- 49. Boyd, J. A., and Eling, T. E. Evidence for a one-electron mechanism of 2- fluorenylacetamide. J. Nati. Cancer Inst., 17: 363-365, 1956. aminofluorene oxidation by prostaglandin H synthase and horseradish per- 34. Hix, C., Oglesby. L.. MacNair, P.. Sieg. M., and Langenbach. R. Bovine oxidase. J. Biol. Chem., 259: 13885-13896. 1984. bladder and liver cell and homogenate-mediated mutagenesis of Salmonella 50. Miller, J. A., and Miller, E. C. Some historical aspects of N-aryl carcinogens lyphimurium with aromatic amines. Carcinogenesis (Lond.). 4: 1401-1407, and their metabolic activation. Environ. Health Perspect.. 49: 3-12, 1983. 1983. 51. Rudo, K., Ellis, S., Lawrence, K., Curtis, G., Garland. H., and Nesnow, S. 35. Holme, J. A., Haug, L. T., and Dybing. E. Modulation of aromatic amine Quantitative analysis of the metabolism of benzo(o)pyrene by transformable mutagenicity in Salmonella typhimurium with rat-liver 9000 g supernatant C3H10T'/2CL8 mouse embryo fibroblasts. Teratog. Carcinog. Mutagen., 6: or monolayers of rat hepatocytes as an activation system. MutÃ¢t.Res., 117: 307-319, 1986. 113-125,1983. 52. Kalow, W. The metabolism of xenobiotics in different populations. Can. J. 36. Bridges. J. W.. and Hubbard, S. A. Problems in providing an appropriate Physiol. Pharmacol., 60: 1-12, 1981. drug metabolizing system for short term tests for carcinogens. In: G. M. 53. Eichelbaum, M. Polymorphic drug oxidation in humans. Fed. Proc., 43: Williams (ed.). The Predictive Value of Short Term Screening Tests in 2298-2302. 1984. Carcinogenicity Evaluation, pp. 69-87. Elsevier/North Holland: BiomÃ©dical 54. Vahakangas, K., Pelkonen, <). and Sotaniemi, E. Cigarette smoking and Press, 1980. drug metabolizing. Clin. Pharmocol. Ther., 33: 375-380, 1983. 37. Bos, R. P., Neis, J. M., van Gemert, P. J. L., and Henderson. P. T. 55. Anderson. K. E., Conney, A. H., and Kappas, A. Nutrition as an environ Mutagenicity testing with the Salmonella/hepatocyte and the Salmonella/ mental influence on chemical metabolism in man. In: W. Kalow, H. W. microsome assays. MutÃ¢t.Res., 124: 103-112, 1983. Goedde, and D. P. Agarwal (eds.). Ethnic Differences in Reactions to Drugs 38. Razzouk, C., and Roberfroid. M. Comparative study of rat, dog, and human and Xenobiotics, pp. 39-54. New York: Alan R. Liss, Inc., 1986. liver microsomal arylamide /V-hydroxylases. Toxicol. Letters, 16: 339-345, 56. Nei. M., and Saitou, N. Genetic relationship of human populations and 1983. ethnic differences in reaction to drugs and food. In: W. Kalow, H. W. Goedde, 39. Phillipson, C. E., and loannides, C. Activation of aromatic amines to muta and D. P. Agarwal (eds.). Ethnic Differences in Reactions to Drugs and gens by various animal species including man. MutÃ¢t.Res., 124: 325-336, Xenobiotics, pp. 21-37. New York: Alan R. Liss, Inc., 1986. 1983. 57. Weber, W. W., and Hein, D. W. N-acetylation pharmacogenetics. Pharmacol. 40. McManus, M. E., Boobis, A. R.. Minchin. R. F., Schwartz, D. M., Murray, Rev., 37: 25-79, 1985. S.. Davies, D. S., and Thorgeirsson, S. S. Relationship between oxidative 58. Siegfried, J., Rudo, K., Ellis. S., Mass, M., and Nesnow, S. Metabolism of metabolism of 2-acetylaminofluorene, debrisoquine, bufuralol, and aldrin in benzo(a)pyrene in monolayer cultures of human bronchial epithelial cells human liver microsomes. Cancer Res., 44: 5692-5697, 1984. from a series of donors. Cancer Res., 46:4368-4371, 1986. 41. King, C. M.. and Glowinski, I. B. Acetylation, deacetylation and acyltransfer. 59. Moore, C. J., and Gould, M. N. Metabolism of benzo(a)pyrene by cultured Environ. Health Perspect., 49:43-50, 1983. human hepatocytes from multiple donors. Carcinogenesis (Lond.), 5: 1577- 42. Holme, J. A., and Soderlund, E. J. Species differences in the cytotoxic and 1582, 1984.

5867

Downloaded from cancerres.aacrjournals.org on October 2, 2021. © 1987 American Association for Cancer Research. Comparison of Human and Rat Hepatocyte Metabolism and Mutagenic Activation of 2-Acetylaminofluorene

Kenneth Rudo, William C. Meyers, Walter Dauterman, et al.

Cancer Res 1987;47:5861-5867.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/47/22/5861

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/47/22/5861. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.