442
Chem. Res. Toxicol. 1989,2, 442-448
Labeling of Albumin Secreted from Isolated Rat Hepatocytes during the Metabolism of N-Nitrosodimethyl- and N-nitrosodiethy lamine Lee D. Gorskyt and Paul F. Hollenberg* Departments of Pathology and Molecular Biology and Cancer Center, Northwestern University Medical School, Chicago, Illinois 60611, and Department of Pharmacology, Wayne State University School of Medicine, Detroit, Michigan 48201 Received December 14, 1988
The metabolic activation of N-nitrosodimethylamine (NDMA) and N-nitrosodiethylamine (NDEA) to reactive intermediates which covalently bind to cellular proteins was investigated. Isolated hepatocytes were used to determine whether protein alkylation is random in nature or whether it results in the alkylation of specific proteins. Isolated hepatocytes from rats treated with either ethanol (EtOH) or phenobarbital were incubated with the I4C-labelednitrosamines for 1-3 h, after which the cells were separated from the incubation medium in order to distinguish secreted proteins from intracellular proteins. SDS-PAGE of the proteins in the medium followed by fluorographic analysis of the gels revealed that a heavily labeled protein was secreted into the medium which represents the predominantly labeled protein. Intracellularly, the major portion of the covalently bound label was found in the region of the gel where the cytochromes P-450 migrate. Pretreatment of the hepatocytes with diethyl maleate and buthionine sulfoximine to decrease the intracellular levels of glutathione had no effect on the labeling, indicating that glutathione does not protect cellular proteins from labeling by these carcinogens. Pretreatment of the cells with D-(+)-galactosamine to inhibit UDP-glucuronyltransferases resulted in a significant decrease in protein labeling by NDEA, suggesting that a glucuronide intermediate may be involved in the activation of NDEA to an alkylating species. The heavily labeled protein secreted into the incubation medium was identified as albumin on the basis of its apparent molecular weight of 66K, as determined by SDS-PAGE, and its cross-reactivity with anti-rat albumin IgG.
Introduction The activation of dialkylnitrosamines is believed to involve the initial oxygenation of the a-carbon to form an a-hydroxy intermediate which spontaneously decomposes to yield a carbonium ion as the ultimate alkylating agent (1-3). The a-hydroxylation of nitrosamines is catalyzed by specific isozymes of the cytochrome P-450family which may therefore be primary targets for alkylation by the activated metabolite of these nitrosamines. Suicide inactivation of the hepatic cytochrome P-450isozyme(s) which catalyzes (catalyze) the initial hydroxylation may explain the loss of dimethylnitrosamine demethylase activity in the livers of mice which received dimethylnitrosamine (4). The present study was undertaken to examine the nature of the proteins that are alkylated during the metabolism of two simple dialkylnitrosamines, N-nitrosdimethylamine (NDMA)' and N-nitrosodiethylamine (NDEA), by isolated rat hepatocytes. We report here that the metabolism of the 14C-labeled nitrosamines by hepatocytes results in the incorporation of a significant amount of the label into intracellular proteins which migrate in the cytochrome P-450 region of SDS-polyacrylamide gels while the majority of the label
associated with proteins secreted from the hepatocytes is covalently bound to albumin.
Materials and Methods Chemicals. ['*C]NDMA (54 mCi/mmol) was purchased from New England Nuclear (Boston, MA). ['CINDEA (57 mCi/mmol) was obtained from Amersham (Chicago, IL). Unlabeled NDMA and NDEA were purchased from Aldrich Chemical Co. (Milwaukee, WI). Collagenase was purchased from Cooper Biomedical (Freehold, NJ). Anti-albumin IgG was obtained from Cappel (Cochranville, PA). The SDS-PAGE molecular markers were from Bio-Rad (no. 1610304). All other chemicals and solvents used were reagent grade from commercial sources. Fixed Staphylococcus aureus was a gift from Dr. M. K. Rundell of Northwestern University. Animals. Male Fischer F344 rats (150-200 g) were obtained from Harlan Industries (Indianapolis, IN) and fed a standard rat chow diet and water ad libitum. EtOH-pretreated rats were given 10% EtOH in their drinking water for 2-3 weeks prior to the isolation of hepatocytes. Phenobarbital-pretreated rats received 0.1% phenobarbital in their drinking water for 1week prior to the isolation of hepatocytes. Preparation of Hepatocytes. Hepatocytes were isolated by the two-step Seglen modification (5) of the Berry and Friend method (6). Pentobarbital (0.65 mgf g body weight) was used as Abbreviations: NDMA, N-nitrosodimethylamine(dimethylnitros-
* To whom correspondence should be addressed at the Department of Pharmacology, Wayne State University School of Medicine, 540 E. Canfield, Detroit, MI 48201. 'Present address: Abbott Laboratories, Abbott Park, IL 60064. 0893-228xf 89f 2702-0442$01.50 f 0
amine); NDEA, N-nitrosodiethylamine(diethylnitrosamine);GSH, glu-
tathione; BSO, buthionine sulfoximine; DEM, diethyl maleate; SDSPAGE, sodium dodecyl sulfate-polyacrylamidegel electrophoresis;Pb, phenobarbital;PBS, phosphate-buffered saline; BSA, bovine serum albumin. 0 1989 American Chemical Society
Albumin Labeling during Nitrosamine Metabolism the anesthetic. Pentobarbital was chosen as the anesthetic rather than ether because ether has been shown to strongly inhibit the metabolism of NDMA by F344 rata (7)and this inhibition was observed for 3 h after the initial exposure to the anesthetic. The inhibitory effect of ether anesthesia on the metabolism of NDMA and N-nitrosomorpholinewas also demonstrated by Spiegelhalder et al. (8). On the other hand, when pentobarbital was used as the anesthetic in F344 rata, the activities of several Phase I and Phase I1 enzymes were not significantly different in the liver just prior to perfusion, in dissociated hepatocytes,or after 2 h in culture (9). The viability of the hepatocytes after isolation was generally 90-95% as judged by the exclusion of trypan blue. Microscopic examination of the suspended cells showed well-rounded cells without the appearance of blebs. Hepatocyte preparations from animals in different states of induction showed no significant differences in viability. Metabolism of NDMA and NDEA by Isolated Hepatocytes. The metabolism of NDMA and NDEA by the hepatocytes was measured by using reverse-phase C18 HPLC. The isolated hepatocytes (4 X lo6 cells) were incubated in 1.0 mL of Hams F10 with 25 mM HEPES, pH 7.4, at 37 "C with constant shaking. The reactions were initiated by the addition of the 14C-labelednitrosamine. After termination of the reaction with 100 pL of 0.6 N perchloric acid and the addition of the internal standard @nitroanisole, 25 nmol) the incubation mixture was centrifuged for 5 min at 15000g. The supernatant, containing unmetabolized parent nitrosamine and acid-soluble metabolites, was filtered through a 0.2-pm Acro LC 13 filter (Gelman),and an 80-pL aliquot was injected onto the HPLC. The reaction mixtures were separated on an Altex 5-pm Ultrasphere ODS column using a gradient program starting with 5% acetonitrile/water for 1.0 min followed by a linear gradient to 75% acetonitrile/water over 15 min. The flow rate was 2 mL/min, and detection was by absorption at 220 nm using a Beckman Model 164 detector. Fractions were collected at 15-9 intervals, and the radioactivity of each fraction was determined after the addition of 10 mL of scintillationfluid. Under these conditions, the retention times for NDMA and NDEA were 2.8 and 8.3 min, respectively. Measurement of Alkylation. For the measurement of alkylation, the reactions were terminated by the addition of 0.35 mL of 20% sulfosalicylic acid. The acid-precipitated material was extracted by suspension and mixing in 95% EtOH followed by centrifugation until no counts above background could be detected in the wash solvent. The pellets were then dissolved in 250 MLof BTS-450 (Beckman). The samples were neutralized with acetic acid, decolorized with 30% HzOz,and counted in a liquid scintillation counter after the addition of 10 mL of scintillation fluid. Polyacrylamide Gel Electrophoresis and Fluorography. The cells were separated from the medium by centrifugation at 15000g for 5 min. The samples were then stored at -70 "C until the time of analysis. Details of the sample preparation for electrophoresis are given in the appropriate figure legends. The samples were electrophoresed on 7.5% polyacrylamide gels (IO) in a water-cooled chamber. After the separation, the gels were fixed in 50% water/methanol for a minimum of 2 h with three changes of the fixative solution. Then either the gels were silver stained (11)to detect protein or they were prepared for fluorographic detection of 14C-labeled proteins. Fluorography was performed as follows. The fixed gels were impregnated with EN3HANCE (Amersham) for 1 h followed by a water rinse for 30 min. The gels were then dried and used to expose a piece of Kodak X-OMAT film. After 2 weeks of exposure at -70 "C, the films were developed in an automated developer. The films were scanned with an LKB Ultrascan XL laser densitometer. Immunoprecipitation of Albumin. After incubation of the hepatocytes with [I4C]NDEA,the medium was separated from the cells by centrifugation at 15000g for 5 min and the medium was removed. An aliquot (0.25 mL) of the medium was incubated with 10 pL of anti-albumin IgG at 37 "C for 1h. Fixed S. aureus (100 pL of a 10% solution) was then added, and the mixture was incubated overnight at 4 "C. The S. aureus was washed three times with PBS by pelleting and resuspension before use, and the final reconstitution was done with 100 p L of PBS containing 100 pg of BSA. After incubation with the S. aureus, the mixture was centrifuged at 15000g for 5 min and the supernatant was
Chem. Res. Toxicol., Vol. 2, No. 6, 1989 443
[NDMA] or [NDEA],WM
Figure 1. Covalent labeling of acid-precipitable cellular macromolecules during metabolism of NDMA or NDEA by hepatocytes from EtOH-pretreated rats. One milliliter of hepatocyte suspensions (4 x lo6 cells/mL) was incubated at 37 OC with constant shaking. Reactions were initiated by the addition of the indicated concentrations of either [14C]NDMAor [14C]NDEA, and metabolism was terminated after 1h by the addition of 0.35 mL of 20% sulfosalicylicacid. The precipitated material was then pelleted and extracted exhaustively with EtOH and the radioactivity determined as described under Materials and Methods. NDMA ( 0 ) ;NDEA (A). removed and labeled as the nonbound fraction. The precipitate was then washed three times with fresh PBS and was resuspended in 250 pL of PBS. This fraction was labeled as the antibody-bound fraction. Determination of Intracellular Oxidized and Reduced Glutathione. The method used was from Farris et al. (12) with slight modification. At the appropriate times, 0.5 mL of the cell suspension was layered carefully onto 0.4 mL of dibutyl phthalate which had previously been layered over 0.5 mL of 20% sulfosalicylic acid containing y-glutamylglutamic acid (20 nmol) as an internal standard. The tubes were centrifuged briefly at 15000g, which caused the cells to rapidly move through the phthalate cushion, enter the acid layer, and lyse. An aliquot (400 pL) of the acid layer was removed, and 50 pL of a solution containing 25 mM 1,lO-phenanthroline and 20 mg/mL iodoacetic acid was added. To this was added 0.5 mL of 2 M KOH-2.4 M KHC03, and this was incubated for 1h in the dark. One milliliter of 1% dinitrofluorobenzene in EtOH (v/v) was added, and the resulting mixture was incubated for 24 h at 25 "C in the dark. Aliquots (100 pL) were analyzed by HPLC using a propylamine column (Beckman) and the solvent system described by Reed et al. (9).
Results Labeling of Cellular Macromolecules by [14C]NDMA and [WINDEA in Hepatocytes Isolated from Ethanol-Pretreated Rats. During t h e course of a previous series of experiments (13) we had observed that the metabolism of NDMA and NDEA by isolated hepatocytes led to t h e incorporation of label from the parent nitrosamine into the acid-insoluble fraction of t h e incubation mixture, in agreement with the reports of others (I, 14-20). Figure 1 illustrates the concentration-dependent covalent labeling of acid-insoluble hepatic macromolecules by NDMA and NDEA in hepatocytes isolated from the livers of EtOH-pretreated rats. I n order t o determine whether t h e labeling of cellular macromolecules by NDMA and NDEA was relatively nonspecific or represented selective binding of the label to specific macromolecules, t h e following experiment was carried out. Hepatocytes were incubated with either [I4C]NDMA or [14C]NDEA for 3 h. At t h e e n d of t h e incubation, the suspensions were centrifuged a t 15000g for
444 Chem. Res. ToxicoZ., VoZ. 2, No. 6, 1989
Gorsky and Hol Zenberg Table I. Glutathione Levels and NDEA Metabolism by Isolated HeDatOCYteS" GSH concn,b NDEA metabolized, treatment % of control % of control control 100 100 DEM/BSO 22 95 D-(+)-galactosamine 93 104 "The hepatocytes were treated as described in the text. Glutathione levels and NDEA metabolism were measured as described under Materials and Methods. * GSH concentrationwas measured immediately prior to the addition of NDEA (160pM).
1
2
3 4
Figure 2. SDS-PAGE and l% fluorography of intracellular and extracellular proteins after metabolism of [14C]NDMA and [14C]NDEAby hepatocytes isolated from EtOH- retreated rats. One milliliter of hepatocyte suspensions (10 X 18cells/mL) was incubated for 3 h with either NDMA (68 p M , 54 mCi/mol) or NDEA (88 pM; 57 mCi/mol). Reactions were terminated by pelleting the cells, removing the supernatant, and freezing both supernatant and the cell pellet at -70 "C. Prior to electrophoresis, 1.0 mL of 2% SDS was added to the cell pellet. After homogenization, the cell sample was diluted 10-foldwith sample-diluting buffer and 30 pL was loaded on the gel. The supernatant fractions were diluted 10-fold with sample-diluting buffer and 30-pL aliquots were loaded on the gel. Electrophoresisand fluorography were performed as described under Materials and Methods. Lane 1: Medium fraction from the incubation with NDMA. Lane 2: Cell fraction from NDMA incubation. Lane 3: Cell fraction from the incubation with NDEA. Lane 4 Medium fraction from the incubation with NDEA. 5 min and the supernatant was removed. The cells and the medium were then separately prepared for SDS-PAGE as described under Materials and Methods. The resulting gels are shown in Figure 2. There are several conclusions which can be derived from the fluorograph illustrated in Figure 2. First, the metabolism of both NDMA and NDEA results in the labeling of both intracellular and extracellular protein. Second, although there is some general labeling in the cellular fractions, the major labeled band migrates in the P-450 region, and there is a second band that runs close to the dye front. Third, in the medium fractions there is a major band which is heavily labeled and has an apparent molecular weight of approximately 66K as determined by calibration of the gel with molecular weight standards. Under the conditions used for these experiments, cell viability decreased by less than 5% during the incubation. Therefore, the proteins contained in the supernatant after sedimentation of the cells consist primarily of proteins secreted by the hepatocytes and some proteins that may have leaked from dying cells. If the proteins in the supernatant were derived primarily from the small proportion of cells that became permeable to trypan blue during the incubation period, then it would be expected that the fluorographs of the medium and the cellular fractions would have essentially identical labeling patterns. That this is not the case can best be seen by comparing the relative amounts of the 66K protein in the medium and the cellular fractions. Labeling of Cellular Macromolecules by [14C]NDEA in Hepatocytes Isolated from PhenobarbitalPretreated Rats. During the course of previous experi-
ments we had observed that NDEA was metabolized more rapidly by microsomes isolated from Pb-pretreated rats than by microsomes isolated from either control or EtOH-pretreated rats (13). Therefore, hepatocytes were isolated from rats given phenobarbital in their drinking water, and the metabolism of NDEA was measured under several conditions. Cells were isolated as described under Materials and Methods, and they were then either not treated (control), treated with diethyl maleate and buthionine sulfoximine (DEM/BSO) to decrease the intracellular levels of glutathione (21),or treated with D-(+)galactosamine to inhibit UDP-glucuronyltransferasesas described by Ullrich and Bock (22). The cells were incubated for 30 min with shaking at 37 "C in the presence or absence of the indicated compounds and then washed three times with fresh medium to remove the DEM/BSO or the D-(+)-galactosamine. The cells were finally resuspended in fresh medium, and the reactions were initiated by the addition of the NDEA. The incubations with the D-(+)-galactosamine-pretreated cells contained D-( +)galactosamine (5 mM) during the incubation with NDEA. Table I shows that neither of these treatments significantly interfered with the metabolism of the NDEA by the hepatocytes. However, treatment with the DEM/BSO decreased the intracellular level of glutathione to 22% of the control value. Figure 3 shows the fluorographs and the silver-stained gels for both the medium and the cellular fractions for this experiment. From the fluorograph and the densitometer scans of the fluorograph it can be seen that there was no diminution in the amount of covalent label incorporated into the proteins from the NDEA in either the cellular fraction or the medium fraction (compare lanes 1 and 4 with lanes 2 and 5 ) even though the intracellular glutathione level was decreased. On the other hand, the preincubation of the hepatocytes with D-(+)galactosamine resulted in a dramatic decrease in the incorporation of label from the NDEA into the proteins in both the cellular fraction and the medium fraction (compare lanes 1 and 4 with lanes 3 and 6)even though the extent of the NDEA metabolism in these cells was the same as in the control group of cells (Table I). A comparison of the silver-stained gel and the fluorograph illustrates two important points. First, the medium lanes each contain an equivalent amount of protein as do each of the cellular lanes. Therefore, the decrease in the amount of label incorporated into the proteins in the presence of galactosamine is not due to differential loading of the gel. Second, in the cellular fractions, the intensity of the protein band at 66K in the silver-stained gel is much greater than the intensity of the same band in the medium fractions. However, the fluorograph shows that even though there appears to be less of the 66K protein in the medium lane, it is much more heavily labeled. The preferential labeling of the 66K protein in the medium suggests that this protein is secreted from the hepatocytes. Effect of D-(+)-Galactosamineon Protein Synthesis. One possible route for the incorporation of label from
Albumin Labeling during Nitrosamine Metabolism
Chem. Res. Toxicol., Vol. 2, No. 6,1989 445
A I
u
I
Lane 2
1
.
1
2
3
4
5
6
-
-
2
3
4
5
6
60
100 PMiihn)
140
le0
fluorography of intracellular and extracellular proteins after metabolism of [WC]NDEAby hepatocytes isolated from phenobarbital-pretreatedrats. Ten milliliters of hepatocytes (10 X lo6cells/mL) was preincubated with either no addition (control), a combination of 1 mM DEM and 1mM BSO, or 5 mM D-(+)-galactosamine for 30 min a t 37 "C with constant shaking. The suspensions were then diluted to 40 mL with fresh medium (no addition), pelleted, and then resuspended. This wash procedure was repeated three times. The cells were resuspended to a concentration of 10 x lo6 cells/mL and were incubated with NDEA (88 pM; 57 mCi/mmol) at 37 OC with shaking for 3 h. For the cells preincubated with D-(+)-galactosamine, 5 mM D-(+)-galactosamine was added to the incubation medium with the NDEA. Aliquots were removed prior to the addition of the NDEA for the determination of glutathione as described under Materials and Methods. Reactions were terminated by pelleting the cells, removing the supernatant, and freezing at -70 "C. Prior to electrophoresis, 1.0 mL of 2% SDS was added to the cell pellet. After homogenization, the sample was diluted 5-fold with sample diluting buffer and 60 pL of each sample was loaded on the gel. The supernatant fractions were diluted 5-fold with sample-diluting buffer, and 60-pL aliquots were loaded on the gel. (A) Fluorograph of SDS-PAGE. (B)Densitometer scans of the fluorograph. (C) Protein silver stain of the SDS-PAGE. Lane 1: Medium from control incubation. Lane 2: Medium from DEM/BSO incubation. Lane 3: Medium from D-(+)-galactosamine incubation. Lane 4 Cells from control incubation. Lane 5: Cells from DEM/BSO incubation. Lane 6: Cells from D-(+)-galactosamine incubation.
Figure 3. SDS-PAGE and
NDEA into the hepatocyte proteins is through the entry of the label into the protein through normal protein synthesis. When [14C]leucinewas added to isolated hepatocytes, proteins were rapidly labeled, indicating the presence of ongoing protein synthesis. Figure 4 shows the results obtained when the cells are incubated with [14C]leucine in the presence and absence of D-(+)-galact"ine. Under the conditions used, D-(+)-galactosaminehad no significant effect on the incorporation of [14C]leucineinto hepatocyte proteins. The lack of an effect of the D-(+)-galactosamine on the overall incorporation of the label into the hepatocyte proteins suggests that the effect of the D-(+)-galactosamine on the transfer of the 14C-labelfrom the NDEA to proteins is by a mechanism other than by blocking metabolic incorporation through the entry of the label into the amino acid pool of the hepatocyte.
Identification of the Labeled Protein in the Medium. The data presented above indicated that the protein in the medium that was heavily labeled by NDEA was a protein which was secreted by hepatocytes and had an apparent molecular weight of 66K. These characteristics are consistent with the protein being albumin. In order to investigate this possibility, an experiment was performed in which hepatocytes were incubated with ['%]NDEA and the medium was then separated from the cells. The medium fraction was then incubated with anti-albumin IgG and an immunoprecipitation was performed as described under Materials and Methods. As shown in Figure 5, immunoprecipitation of the I4C-labeled medium with anti-albumin IgG results in a selective loss of the 66K protein from the medium, SDS-PAGE of the IgG-protein precipitate shows full recovery of the labeled 66K protein.
446 Chem. Res. Toxicol., Vol. 2, No.6,1989
Gorsky a n d Hollenberg
A
1
&1
B
24
Lane I
2
2
3 I
B
I8
9
14
Lane 2
C
3 Lane 2
Figure 4. Incorporation of [l%]leucine into intracellular proteins in the absence and presence of p(+)-galact"ine. Hepatocytes were isolated from a phenobarbital-pretreatedrat. Untreated cells were labeled with [14C]leucinefor 3 h (lane 1). Pretreated cells were preincubated with 5 mM D-(+)-gdaCtoSaminefor 30 min, washed three times by suspension and repelleting, and then labeled with [14C]leucinefor 3 h in the presence of 5 mM p(+)galactosamine (lane 2). The samples were prepared for SDSPAGE as described in the legend to Figure 3. (A) Fluorograph of the SDS-PAGE. (B)Densitometer scan of the fluorograph.
Thus, the results of this experiment demonstrate that the identity of the labeled protein that is secreted from the hepatocytes is albumin.
Discussion During the metabolism of dialkyhitrcsamines, activated intermediates are generated which are capable of alkylating cellular macromolecules including the nucleic acids and
Figure 5. Immunoprecipitation of albumin secreted from hepatocytes incubated with [l"c]NDEA. Hepatocytes were isolated from a phenobarbital-pretreated rat and were incubated with ['TINDEA as described in the legend to Figure 2. The medium was separated from the cells, and the immunoprecipitation was performed as described under Materials and Methods. SDSPAGE, fluorography, and densitometry were performed as described under Materials and Methods. (A) Fluorograph of the SDS-PAGE. (B)Densitometer scan of the fluorograph. Lane 1: Control (medium not treated with anti-albumin IgG). Lane 2 Supernatant after precipitation with anti-albumin IgG. Lane 3: Anti-albumin IgG bound fraction.
Albumin Labeling during Nitrosamine Metabolism
proteins. The alkylation of cellular macromolecules by metabolites of NDMA has been demonstrated in vivo in the rat (14,16) and in vitro with guinea pig hepatocytes (18),rat hepatocytes (15),and hepatic microsomes derived from the rat (1). The alkylation of cellular macromolecules by metabolites of NDEA has been demonstrated in vivo in hamsters and rats (17, 20, 23-25) and with hepatic microsomes derived from the rat (19). The results presented here demonstrate that the metabolism of NDMA and NDEA results primarily in the alkylation of albumin in hepatocytes isolated from EtOHpretreated rats. In addition, the metabolism of NDEA in hepatocytes isolated from phenobarbital-pretreated rats also results in the labeling of albumin. These experiments also demonstrate that the covalent binding of 14C-label from the nitrosamines to albumin is a relatively specific process and that cellular protein is not randomly alkylated by metabolites of NDMA and NDEA. Although there are other labeled proteins present in the medium from the hepatocyte incubations, the majority of the label is found associated with the albumin. The alkylation of macromolecules during the metabolism and activation of dialkylnitrosamines by subcellular fractions rules out the possibility that incorporation of label into proteins may occur only through protein synthesis. It has been reported, however, that albumin secreted from primary cultures of rat hepatocytes over a 24-h period contained label that had been derived from NDMA through the entry of label into the amino acid pools (26). We do not believe that metabolic incorporation through the amino acid pools can account for a significant proportion of the incorporation of label that was seen in this study since the incubations were for 3 h as opposed to the 24-h period used by Seago et al. (26). If the protein labeling were primarily due to metabolic incorporation of I4C from NDEA into the amino acid pool, then we would expect that the labeling patterns observed with ['%]NDEA would be similar to those observed when a 14C-labeled amino acid was incubated with the hepatocytes under similar conditions. However, comparison of the labeling pattern in Figure 3B, lane 4 ([14C]NDEA),with that in Figure 4B, lane 1(['%]Leu), shows that they are markedly different. Although D-(+)-galaCtOSaminedid not inhibit the incorporation of label from 14C-labeledleucine into protein (Figure 4), it did inhibit incorporation of label from [14C]NDEA (Figure 3), thereby demonstrating another difference between incorporation from [14C]NDEAand incorporation from the amino acid pool. Therefore, these results suggest that there is a pathway other than incorporation of label into the amino acid pools of the cells that accounts for the predominant portion of the incorporation of label into protein observed in this study. The definitive proof, isolation and structural characterization of the alkylated portion of the albumin, requires further work. Studies have demonstrated that the binding of activated intermediates to proteins during the metabolism of NDMA and NDEA by microsomes can be stimulated by the addition of cytosolic components to the incubation systems (19).The stimulation of protein alkylation by the addition of cytosol to microsomal fractions implies the participation of non-P-450 elements and raises the possibility that Phase I1 conjugation reactions may be involved in the metabolic activation of the nitrosamine carcinogens to alkylating species. Mutagenicity assays utilizing coincubation of isolated hepatocytes with Salmonella typhimurium have demonstrated that activated intermediates formed by the hepatocytes are capable of leaving the hepatocyte and inducing
Chem. Res. Toxicol., Vol. 2, No. 6, 1989 447
mutations in the bacteria. This was demonstrated with a series of N-nitrosodialkylamines, (27) and oxidized derivatives of N-nitrosodi-n-propylamine(28). Furthermore, the metabolism of NDMA by hepatocytes was also found to result in the formation of reactive intermediates mutagenic to cocultured Chinese hamster V79 cells (29). These mutagenicity assays indicate that hepatocytes are capable of metabolically activating procarcinogenic nitrosamines to forms that may be released from the hepatocyte and are capable of entering other cells before exerting their biological effect. In addition, the alkylation of exogenous DNA during the metabolism of NDMA by isolated hepatocytes has been reported by Umbenhauer and Pegg (15). The evidence for the inhibition of protein labeling by galactosamine suggests that a glucuronide derivative may play a role in the pathway of the activation of NDEA. The implication of a glucuronide as being involved in alkylation during the metabolism of NDEA raises the possibility of a conjugation-deconjugation cycle as being a determinant in the alkylation potential of this nitrosamine. A conjugation-deconjugation cycle has been proposed as a determinant in the net production of glucuronide metabolites from substrates generated by cytochrome P-450 (30). Although glucuronic acid conjugation is generally considered to be a detoxifying metabolic pathway, it has been suggested that it may play a role in the activation of nitrosamine carcinogens (31). The formation of glucuronides of hydroxylated nitrosamines has also been reported by Wiessler and Rossnagel(32), Suzuki and Okada (33), and Kokkinakis and co-workers (34, 35). More than one pathway for the activation of NDEA to an alkylating agent is very possible. However, from the results presented above, it is clear that conjugates of glutathione are not involved in the activation pathway of NDEA in this model system. This is consistent with the lack of effect of glutathione depletion on DNA methylation by NDMA in the rat in vivo as reported by Tacchi et al. (14). However, one must bear in mind that the pathways involved in the overall disposition of NDEA may differ dramatically from that of NDMA as evidenced by the marked differences in selectivity for different isozymes of P-450 previously demonstrated for these two structurally related nitrosamines (13, 36, 37). Further studies are in progress to isolate and characterize the putative glucuronide conjugate. In addition, studies are in progress to identify the labeled proteins in the P-450 region of the SDS-PAGE gels and to characterize the nature of the alkylation.
Acknowledgment. These studies were supported in part by Grant CA 16954 from the National Cancer Institute, USPHS, DHHS. We thank Dr. Janardan Reddy for his advice and encouragement and for generously providing the anti-rat albumin IgG.
References (1) Lai, D.Y.,and Arcos, J. C. (1980) Dialkylnitrosamine bioactivation and carcinogenesis. Life Sci. 27, 2149-2165. (2) Magee, P. N.,and Farber, E., (1962) Toxic liver injury and carcinogenesis: methylation of rat liver nucleic acids by dimethylnitrosamine in uiuo. Biochem. J. 83, 114. (3) Magee, P.N., and Hultin, T. (1962) Toxic liver injury and carcinogenesis. Methylation of proteins of rat liver slices by dimethylnitrosamine in uitro. Biochem. J. 83, 106 (4) Anderson, L. C.,Harrington, G. W., Pylypiw, H. M., Jr., Hagiwara, A., and Magee, P. N. (1986) Tissue levels and biological effects of N-nitrosodimethylamine in mice during chronic low or high dose exposure with or without ethanol. Drug Metab. Dispos. 14,733-739.
Gorsky and Hollenberg
448 Chem. Res. Toxicol., Vol. 2, No. 6, 1989 (5) Seglen, P. 0.(1976)Preparation of isolated rat liver cells. In Methods in Cell Biology (Prescott, D. M., Ed.) Vol. XIII, pp 29-83, Academic Press, New York. (6) Berry, M. N., and Friend, D. S. (1969)High-yield preparation of isolated rat liver parenchymal cells. J. Cell Biol. 43,506-520. (7) Keefer, L. K., Garland, W. A., Oldfield, N. F., Swagzdis, J. E., and Mico, B. A. (1985)Inhibition of N-Nitrosodimethylamine metabolism in rats by ether anesthesia. Cancer Res. 45, 5457-5460. (8) Spiegelhalder, B., Eisenbrand, G., and Preussmann, R. (1982) Urinary excretion of N-nitrosamines in rata and humans. In N-Nitroso Compounds: Occurence and Biological Effects, IARC Scientific Publication No. 41, pp 443-449. (9) Reed, D. J., Babson, J. R., Beatty, P. W., Brodie, A. E.,Ellis,W. W., and Potter, D. W. (1980)High-performance liquid chromatography analysis of nanomole levels of glutathione, glutathione disulfide, and related thiols and disulfides. Anal. Biochem. 106, 55-62. (10) Laemmli, U. K. (1970)Cleavage of structural proteihs during the assembly of the head of bacteriophage T4. Nature (London) 227,680-685. (11) Wray, W., Boulikas, T., Wray, V. P., and Hancock, R. (1981) Silver staining of proteins in polyacrylamide gels. Anal. Biochem. 118,197-203. (12) Farris, M. W., Brown, M. K., Schmitz, J. A., and Reed, D. J. (1985)Mechanism of chemical-inducedtoxicity. I. Use of a rapid centrifugation technique for the separation of viable and non-viable hepatocytes. Toxicol. Appl. Pharmacol. 79,283-295. (13) Gorsky, L. D., and Hollenberg, P. F. (1989)Metabolism of N-nitrosodimethyl- and N-nitrosodiethylamine by rat hepatocytes: Effects of pretreatment with ethanol. Chem. Res. Toxicol. (preceding paper in this issue). (14) Tacchi, A. M., Jensen, D. E., and Magee, P. N. (1987)Effect of glutathione modulation using buthionine sulfoximine on DNA methylation by dimethylnitrosamine in the rat. Biochem. PharmQCOl. 36,881-885. (15) Umbenhauer, D. R.,and Pegg, A. E. (1981)Alkylation of intracellular and extracellular DNA by dimethylnitrosamine following activation by isolated rat hepatocytes. Cancer Res. 41, 3471-3474. (16) Schwarz, M., Buchmann, A., Klormann, H., Schrenk, D., and Kunz, W. (1985)Effect of phenobarbital and other liver monooxygenase modifiers on dimethylnitrosamine-induced alkylation of rat liver macromolecules. Cancer Res. 45,202C-2024. (17) Becker, R. A., Lau, S. S., Bresil, H., Shank, R. C., and Montesano, R. (1985)DNA ethylation in target and nontarget organs of hamsters and rats treated with diethylnitrosamine. Cancer Lett. 26, 17-24. (18) Hauber, G., Frommberger, R., Remmer, H., and Schwenk, M. (1984)Metabolism of low concentrations of N-nitrosodimethylamine in isolated liver cells of the guinea pig. Cancer Res. 44, 1343-1346. (19) Lai, D. Y.,Arcos, J. C., and Argus, M. F. (1979)Factors influencing the microsome- and mitochondria-catalyzed in uitro binding of diethylnitrosamine and N-nitrosopiperidine to deoxyribonucleic acid. Biochem. Pharmacol. 28, 3545-3550. (20) Letnansky, K. (1982)Alkylation of nuclear proteins in rat liver by diethylnitrosamine. Cell. Mol. Biol. 28,353-361. (21) Drew, R., and Miners, J. 0. (1984)The effects of buthionine sulphoximine (BSO) on glutathione depletion and xenobiotic biotransformation. Biochem. Pharmacol. 33,2989-2993.
(22) Ullrich, D., and Bock, K. W. (1984)Inhibition of glucuronide formation by D-galactosone of D-galactosamine in isolated hepatocytes. Biochem. Pharmacol. 33,1827-1830. (23) Swenberg, J. A., Dyroff, M. C., Bedell, M. A., Popp, J. A., Huh, N., Kirstein, V., and Rajewsky, M. F. (1984) 0'-Ethyldeoxythymidine, but not OB-ethyldeoxyguanosine,accumulates in hepatocyte DNA of rata exposed continuously to diethylnitrosamine. Proc. Natl. Acad. Sci. U.S.A. 81,1692-1695. (24) Dyroff, M. C., Richardson, F. C., Popp, J. A., Bedell, M. A., and Swenberg, J. A. (1986)Correlation of 0'-ethyldeoxythymidine accumulation, hepatic initiation and hepatocellular carcinoma induction in rats continuously administered diethylnitrosamine. Carcinogenesis 7,241-246. (25) Boucheron, J. A., Richardson, F. C., Morgan, P. H., and Swenberg, J. A. (1987)Molecular dosimetry of O'-ethyldeoxythymidine in rata continuously exposed to diethylnitrosamine. Cancer Res. 47,1577-1581. (26) Seago, A., Shuker, D. E.G., and Paine, A. J. (1986)Interaction of 14CC-dimethylnitrosamine with albumin produced by rat hepatocytes in culture. Toxicol. Lett. 30, 41-48. (27) Rumruen, K., and Pool, B. L. (1984)Metabolic activation capabilities of S9 and hepatocytes from uninduced rats to convert carcinogenic N-nitrosamines to mutagens. Mutat. Res. 140, 147-153. (28) Langenbach, R., Gingell, R., Kuszynski, C., Worker, B., Nagel, D., and Pour, P. (1980)Mutagenic activities of oxidized derivatives of N-nitrosodipropylamine in the liver cell-mediated and Salmonella typhimurium assays. Cancer Res. 40, 3453-3467. (29) Langenbach, R., Nesnow, S., Tompa, A,, Gingell, R., and Kuszynski, C. (1981)Lung and liver cell-mediated mutagenesis systems: specificities in the activation of chemical carcinogens. Carcinogenesis 2,851-858. (30) Schollhammer, I., Poll, D. S.I., and Bickel, M. H. (1975)Liver microsomal @-glucuronidaseand UDP-glucuronyl transferase. Enzyme 20,269-275. (31) Pastorelli, R., Ancidei, A., Fanelli, R., and Airoldi, L. (1988) Effect of acute and chronic butylated hydroxyanisole administration on in uiuo glucuronidation of N-nitroso(4-hydroxybutyl)amine in rata. Food Chem. Toxicol. 26,419-423. (32) Wiessler, M., and Rossnagel, G. (1987)a-Glucuronides of N nitrosomethylbenzylamine, IARC Scientific Publication No. 84, pp 170-172. (33) Suzuki, F., and Okada, M. (1981)Metabolic fate of N,N-dipropylnitrosamine and N,N-diamylnitrosamine in the rat, in relation to their lack of carcinogenic - effect on the urinary bladder. Gann 72,552-559. (34) Kokkinakis. D. M.. Hollenbere. P. F.. and Scam& D. G. (1985) Major urinary metabolites in h&sters and rats treated with Nnitroso(2-hydroxypropyl)(2-oxopropyl)amine. Cancer Res. 45, 3586-3592. (35) Kokkinakis, D. M., Scarpelli, D. G., Sabbarao, V., and Hollenberg, P. F. (1987)Species differences in the metabolism of Nnitroso(2-hydroxypropyl)(2-oxopropyl)amine.Carcinogenesis 8, 295-303. (36) Magour, S.,and Nievel, J. G. (1971)Effect of inducers of drug-metabolizing enzymes on diethyl nitrosamine metabolism and toxicity. Biochem. J. 123,8p-9p. (37) Arcos, J. C., Bryant, G. M., Pastorand, K. M., and Argus, M. F. (1976)Structural limits of specificity of methylcholanthrenerepressible nitrosamine N-dealkylases. Inhibition by analog substrates. Z. Krebsforsch. 89,171-183. ~
