Chem. Res. Toxicol. 1993,6, 269-276
269
8-Aminoguanine: A Base Modification Produced in Rat Liver Nucleic Acids by the Hepatocarcinogen 2-Nitropropane Rama S. S o d u m , G u o Nie, and Emerich S. Fiala* Division of Biochemical Pharmacology, American Health Foundation, 1 Dana Road, Valhalla, New York 10595 Received December 7,1992
2-Nitropropane (2-NP), an important industrial chemical and a hepatocarcinogen in rats, had previously been found to produce several modifications of nucleosides in rat liver RNA and DNA that are discernible using HPLC with electrochemical detection. While one of these modifications has been identified as an increase in the levels of 8-oxoguanosine and 8-oxo-2'deoxyguanosine in RNA and DNA, respectively, the others had not been identified. We now present evidence that a major modification in rat liver nucleic acids due to the administration of 2-NP is the amination of guanine a t C8, apparently a completely novel in vivo reaction. 8-Aminoguanosine, isolated from hydrolysates of liver RNA from 2-NP-treated rats, cochromatographed with synthetic or commercially-obtained standard on reverse-phase as well as cation-exchange HPLC, and its UV spectral characteristics at acidic, neutral, and basic pH were identical to those of the standard. Acid hydrolysis produced 8-aminoguanine, which had a retention time and fragmentation pattern identical to that of the standard on gas chromatographymass spectrometry of the trimethylsilyl derivatives. Evidence for the presence of S-aminodeoxyguanosine in liver DNA of rats treated with 2-NP was also obtained by cochromatography with synthetic standard on HPLC. Hydroxylamine-O-sulfonic acid was found t o react with RNA and DNA to give 8-oxo- and 8-amino-substituted guanines. We propose, as a working hypothesis, that 2-NP may be metabolized t o hydroxylamine-0-sulfonate or acetate, which yield the reactive nitrenium ion, NH2+,capable of aminating cellular macromolecules in vivo.
Introduction 2-Nitropropane (2-NP),' an important industrial chemical (1)and a component of cigarette smoke (21, was shown to be a hepatocarcinogen in male Sprague-Dawley rats when administered either by inhalation (3) or by gavage (4). Treatment with 2-NP produces a characteristic pattern of nucleoside modifications in liver RNA and DNA of Sprague-Dawley or F344 rata that is discernible using HPLC with electrochemical detection (EC) (5-8). These modifications include increased levels of 8-oxoguanosine and 8-oxodeoxyguanosine and the formation of thus far unidentified nucleosides designated as RX1 and RX2 in RNAand DX1 and DX22in DNA (5-8). RX1, RX2, DX1, and DX2 are not detectable in the liver nucleic acids of untreated rats (5-8). There are several lines of evidence favoring the relevance of these modifications to the hepatocarcinogenicity of 2-NP. For instance, they are produced to only a minimal extent in liver nucleic acids of rata treated with l-nitropropane (51, which is either noncarcinogenic or only weakly carcinogenic ( 4 ) , and significantly lower levels of these modifications are produced in liver nucleic acids of female rata (7), which are
* To whomcorrespondenceshould beaddressed at the AmericanHealth Fax: Foundation, 1Dana Road, Valhalla, NY 10595. Tel: 914-789-7128; 914-592-6317. 1 Abbreviations: 2-NP, 2-nitropropane; EC, electrochemical detection; ECD, electrochemical detector; BSTFA-1% TMCS, N,O-bis(trimethylsily1)trifluoroacetamide plus 1% trimethylchlorosilane; HMDS, hexamethyldisilazane; Tri-Si1 TBT, a Pierce Chemical Co. (Rockford, IL) formulation of N-(trimethylsilyl)imidazole,N,O-bis(trimethylsily1)acetamide, and trimethylchlorosilane; EI, electron impact; CI, chemical ionization; FAB, fast-atom bombardment. 2 E. S. Fiala, G. Nie, L. Reinhart, B. S. Hussain, and C. C. Conaway, unpublished results.
less susceptible to the carcinogenicity of 2-NP than male rats (9). These modifications are produced to even lesser extents in the livers of male New Zealand White rabbits,2 which have been reported to be resistant to the hepatotoxicity and hepatocarcinogenicity of 2-NP (3). Furthermore, only very small amounts of these modifications are produced in the nucleic acids of the rat kidney (3, an organ which is not a target for the carcinogenicity of 2-NP (3, 4). Interestingly, several other secondary, but not primary, nitroalkanes such as 2-nitrobutane, as well as the liver tumorigen (10) acetone oxime, produce a qualitatively apparently identical pattern of nucleic acid modifications in rat liver (6,8).These secondary, but not primary, nitroalkanes are also mutagenic in Salmonella typhimurium (11)and induce unscheduled DNAsynthesis in rat liver hepatocytes in vitro2 (12). In this report, we provide chromatographicand spectral evidence for the identity of RX2 as 8-aminoguanosineand chromatographic evidence for the identity of DX2 as 8-aminodeoxyguanosine, and we suggest a possible mode of formation of these species. Although arylamination at the C8 position of guanine is a preferred pathway for the formation of activated aromatic amine adducts in nucleic acids (ref 13 and references therein), to our knowledge, the production of 8-aminoguaninein rat liver nucleic acids represents an entirely novel in vivo base modification caused by a chemical carcinogen.
Experimental Section Materials. 2-NP and hydroxylamine-0-sulfonic acid were purchased from Aldrich Chemical Co. (Milwaukee, WI). N,OBis(trimethylsily1)trifluoroacetamide-1 % trimethylchlorosilane
QS93-228~/93/27Q6-0269~04.QO/Q 1993 American Chemical Society
270 Chem. Res. Toxicol., Vol. 6, No. 3, 1993 (BSTFA-1% TMCS), Tri-Si1 TBT [a Pierce Chemical Co. formulation of N-(trimethylsilyl)imidazole,N,O-bis(trimethy1silyl)acetamide, and trimethylchlorosilane], and acetonitrile were obtained from Pierce Chemical Co. (Rockford, IL). Enzymes, calf thymus DNA, and nucleoside and nucleotide standards were purchased from Sigma Chemical Co. (St. Louis, MO), unless stated otherwise. Ultrasphere ODS columns and Ultrasphere ODS guard columns were purchased from Beckman Instruments, Inc. (Berkeley, CA). Emulphor 620, a polyoxyethylated vegetable oil emulsifier, was obtained from GAF Corp. (Wayne, NJ). Optisil 10 SCX columns were obtained from Phenomenex (Torrance, CA), and Hamilton PRP-1 columns were purchased from Rainin Instrument Co. (Woburn, MA). Nucleoside Standards. 8-Aminoguanosine was either synthesized using the procedure described by Long et al. (14) or obtained from Sigma Chemical Co. In both cases, the nucleoside was characterized by UV and NMR spectroscopy as well as by electron impact (EI), chemical ionization (CI), and fast-atom bombardment (FAB) mass spectra. 8-Aminodeoxyguanosine was prepared by reacting hydrazine with 8-bromodeoxyguanosine (14). The latter was prepared from deoxyguanosine and bromine water by the method of Cho et al. (15). 8-Oxoguanosine and 8-oxodeoxyguanosine were prepared by oxidation of guanosine and deoxyguanosine, respectively, with hydrogen peroxide in the presence of ascorbic acid (16) and were purified by HPLC. Animals. MaleF344 rats (200-300 g), purchasedfrom Charles River Laboratories, Inc. (Kingston, NY), were treated intraperitoneally with 2-NP at a dose of 100 mgikg body weight (25 mg/mL solution in Emulphor/HzO, 1:4) or with Emulphor/HzO alone (4 mL/kg) and killed 18 h later. Livers were immediately excised, rinsed in ice-cold 0.15 M NaC1-0.015 M sodium citrate (pH 7.0), and frozen a t -70 "C. Isolation and Hydrolysis of RNA and DNA. As detailed previously (5),DNA was isolated by a modification of Marmur's method ( I 7) and RNA was isolated by a slight modification of the method of Chomczynski and Sacchi (18). DNA and RNA were hydrolyzed to nucleosides by sequential incubations with nuclease P1 and alkaline phosphatase (Sigma type 111)a t pH 5.4 and 7.0, respectively. The nucleic acid hydrolysates were analyzed immediately by HPLC-EC. Judging from the only minute amounts of UV-absorbing material eluting before cytosine or deoxycytosine and after adenosine or deoxyadenosine upon reverse-phase HPLC, the hydrolysis of both the RNA and DNA was essentially complete under these conditions. HPLC Systems. They were as follows: system 1, reversephase HPLC using a 0.46- X 25-cm Ultrasphere ODS column with a 0.46- X 4.6-cm Ultrasphere ODS guard column eluted with a 12.5 mM citric acid and 25 mM sodium acetate buffer (pH 5.1), 5 % methanol; system 2, reverse-phase HPLC using two 0.46- X 25-cm Ultrasphere ODS columns in series with a 0.46x 4.6-cm Ultrasphere ODS guard column eluted with a 12.5 mM citric acid and 25 mM sodium acetate buffer, 9.5% methanol (pH 5.1); system 3, reverse-phase HPLC using two 0.46- X 25-cm Ultrasphere ODS columns in series with a 0.46- X 4.6-cm Ultrasphere ODS guard column eluted with 12.5 mM citric acid and 25 mM sodium acetate buffer, 3 % methanol (pH 5.1); system 4, cation-exchange HPLC using two 0.46- X 25-cm Optisil 10 SCX columns in series with a guard column containing a 3-cm Aquapore CX-300 cation-exchange cartridge (Brownlee Labs, Inc., Santa Clara, CA), eluted with 12.5 mM sodium citrate and 25 mM sodium acetate buffer a t pH 5.1; system 5, a0.41- X 25-cm Hamilton PRP-1 column with a 0.46- X 2.7-cm PRP-1 guard column eluted with water. The flow rate was 1 mLimin in all cases. It is important to note that, in the absence of buffer salts a t neutral pH, 8-aminoguanine, 8-aminoguanosine, and 8-aminodeoxyguanosine are very strongly adsorbed to the reversephase HPLC silica-based packings (but not the PRP-1 packing) used in the present work, presumably due to ion exchange with silanol hydrogens. Instrumentation. Waters Model 510 HPLC pumps with Model LP-21 pulse dampeners from SSI (Scientific Systems, Inc., State College, PA) or the Shimadzu Model LC600 HPLC
Sodum et al. pumps were used. Separations were monitored with a Waters Model 990+ photodiode array detector or an SSI Model 500 variable-UV/vis detector a t 253 nm. Amperometric detectors used were the BAS (Bioanalytical Systems, West Lafayette, IN) Model LC-4B/CC-4 or the Princeton Applied Research (Princeton, NJ) Model 400. In either case, the potential was +600 mV versus the Ag/AgCl/3 M NaCl reference electrode. UV spectra were recorded with a Beckman DU-7 spectrophotometer (Beckman Instruments Inc., Fullerton, CA) in a mixed buffer of ammonium formate, potassium phosphate, and ammonium acetate. NMR spectra in Me$O-ds andD20 were obtained with a 360-MHz Bruker AM 360 WB spectrometer. E1 and CI (methane) mass spectra were acquired using a H P (HewlettPackard, Avondale, PA) Model 5971A mass-selective detector interfaced to a H P 5890 Series I1gas chromatograph. Injections were made with a H P 7673 automatic sampler. FAB mass spectra were obtained with the VGZAB-2F mass spectrometer using xenon gas fast-atom bombardment through the courtesy of Dr. C. Metra1 of the Chemical Synthesis and Analysis Laboratory of Program Resources Inc., Frederick, MD. Purification, Hydrolysis, and Derivatization of RX2. Enzymatic hydrolysates of liver RNA from rats treated with 2-NP were repeatedly subjected to HPLC using system 2 with the electrochemical detector (ECD) turned off, and fractions containing RX2 were collected and concentrated by rotary evaporation under vacuum. Purified samples were desalted by HPLC using system 5, and UV spectra were determined a t various pH values. For hydrolysis and derivatization, desalted RX2 was concentrated, transferred to a 2-mL amber vial, and lyophilized. HCl, as a 10% solution in methanol, was added, and the vial was tightly closed with a screw cap with a Teflon-coated septum and heated in an oil bath at 95 "C for 20 min. The solvent was evaporated under vacuum, and traces of HC1 were removed by repeated additions of methanol followed by evaporations. The residue was derivatized with 20 rL of a mixture of BSTFA-1% TMCS, hexamethyldisilazane (HMDS), and acetonitrile (1:0.2: 0.3 v/v) a t 100 "C for 90 min in a 0.1-mL Reacti-Vial (Pierce Chemical Co., Rockford, IL). All glassware was silanized with dichlorodimethylsilane prior to use. GC-MS Analyses. 8-Aminoguanosine standard was hydrolyzed to the base in 1mL of methanol, 10% in HC1, at 95 OC for 20 min. After removal of solvent under reduced pressure and drying under vacuum overnight, the products were derivatized using a 2:l (v/v) mixture of BSTFA-1% TMCS/CH&N at 100 "C for 90 min. For use as GC standards, TMS derivatives of various purines, including 8-oxoguanine, guanine, and adenine, were prepared by similar procedures. After appropriate dilution of trimethylsilylated samples with CH3CN, separations were performed using an H P Ultra-2 capillary column (5% phenylmethylsilicone), 12 m in length, 0.2 mm i.d., and 0.11-pm film thickness. Helium, grade 5.0 purity, was used as carrier. Injections were performed in the splitless mode. The temperature was programmed as follows: 3 min at 100 "C, increase to 250 "C a t a rate of 25 "Cimin, and then hold a t 250 "C for 10 min. Reaction of Hydroxylamine- 0-sulfonic Acid with Nucleic Acids and Nucleosides. Hydroxylamine-0-sulfonic acid, as a 0.28 M aqueous solution, was added dropwise to 1 mg/mL solutions of RNA isolated from livers of untreated male F344 rats, or calf thymus DNA, in 0.1 M potassium acetate. The final ratios of reactants were 8 or 5.3 mg of hydroxylamine-0-sulfonic acidimg of RNA or DNA, respectively. The pH was adjusted to 7.3, and the reaction mixture was incubated at 37 "C for 2 h. Aliquots were diluted 10-fold with HzO and hydrolyzed to nucleosides by sequential incubations with nuclease P1 a t pH 5.4 and alkaline phosphatase a t pH 7.0. Reactions of hydroxylamine-0-sulfonic acid with guanosine and deoxyguanosine were carried out in a similar manner for 7.5 h, to give a yield of 0.5% of 8-aminoguanosine and 1.5 % of 8-aminodeoxyguanosine, respectively.
Chem. Res. Toxicol., Vol. 6, No. 3, 1993 271
8-Aminoguanine Produced by 2-Nitropropane Table I. UV Absorption Spectra of Certain 8-Substituted Guanine Derivatives wavelength (nm) pH 2.0 pH 7.3 pH 10.5 Guo 8-oxo-Guo 8-NHz-Guo 8-NHz-dGuo 8-NH2-Gua
sh,
Amax
Amin
Amax
Amin
Amax
Amin
254 278Sh 295 246 289 251 288 251 288 247
227
252 272Sh 293 247 293 257 291 256 292 250 222
224
261
230
270
280 250 259 280Sh 258 280Sh 283 250Sh
265
270 273 229 272 227 268
280 227 281 227 267 241
i 0.36 AU Q 253 n m ( 1 ) 0.6 nA ( 2 ) 1.5 nA ( 3 - 5 )
i
231
--
5
227 240
shoulder.
Results Verification of Standards. 8-Aminoguanosine standard, either commercially obtained or prepared according to the procedure of Long et al. (14), was found to be homogeneous by reverse-phase (system 1) and cationexchange (system 4) HPLC. The structure of the nucleoside was confirmed by its UV and NMR spectra, and FAB mass spectrum. The UV spectral characteristics of 8-aminoguanosine and of certain other guanines are given in Table I. The FAB mass spectrum of 8-aminoguanosine showed the expected molecular ion (M + H)+ at mlz 299 and base plus hydrogen after sugar cleavage (B + H)+ at mlz 167. Hydrolysis in methanol-HC1 gave 8-aminoguanine, which was characterized by HPLC, UV spectra, and E1 and CI mass spectra. The E1 mass spectrum after trimethylsilylation with BSTFA-1% TMCSlCH3CN or with Tri-Si1 TBT gave the expected molecular ion with four TMS groups at mlz 454 (M+)and (M - CHd+ at mlz 439 due to loss of a methyl group. The CI mass spectrum showed the molecular ion plus a hydrogen at mlz 455, along with other characteristic peaks (M + CzHs)+at mlz 483 and (M + C2H5 + H)+at mlz 484. When BSTFA-l% TMCS/CH&N was used for the derivatization, an additional GC peak was observed that corresponded to 8-aminoguanine.5TMS, with a molecular ion M+ at mlz 526 and (M - CH3)+ at mlz 511. The formation of the penta-TMS derivative of 8-aminoguanine is attributed to the tautomerization of the 8-amino group to an imine in the presence of acid. Identification of 8-Amiaoguanosine in Rat Liver RNA. As shown in Figure 1,HPLC-EC analysis (system 1)of nucleosides obtained by enzymatic hydrolysis of liver RNA from rats treated with 2-NP indicates the presence of three electrochemically active species designated as RX1,8-oxoguanosine, and RX2. Although corresponding preparations from rats treated with vehicle (Emulphorl water) only (trace 2 in Figure 1) or from untreated rats (data not shown) also contain small amounts of 8-oxoguanosine, the species designated as RX1 (elution volume -12.8 mL) and RX2 (elution volume -21.5 mL) are undetectable in rat liver RNA from control or untreated rats (5-8). RX1 appears to be unstable; when RNA hydrolysates are allowed to stand at 0 OC, the amount of this presumed nucleoside decreases with time. Addition of 8-oxoguanosine standard to the RNA hydrolysate resulted in an increase in the size of the peak eluting at -18 mL (Figure 1, trace 41, whereas addition of &aminoguanosine standard to the mixture resulted in an
1
.......................... 10
20
30
ELUTION VOLUME, mL
F i g u r e 1. Cochromatography (HPLC-EC system 1) of 8-oxoguanosine and 8-aminoguanosine standards with 8-oxoguanosine and RX2, respectively, in an enzymatic hydrolysate of liver RNA from a 2-NP-treated rat. 1, UV detector response; 2-5, electrochemical detector responses. 3, liver RNA hydrolysate from 2-NP-treated rat without additions; 4, the same RNA hydrolysate after addition of 8-oxoguanosine standard only; 5, the same RNA hydrolysate after addition of 8-aminoguanosine standard only. For comparison, the EC profile of an equal amount of liver RNA hydrolysate (no additions) from a vehicle-treated rat is shown in trace 2. Note that trace 2 was obtained at a higher ECD sensitivity. A sporadic artifactual ECD peak, probably originating from the buffers used, is indicated by "?".
increaseonlyofRX2, thepeakelutingat -21.5mL (Figure 1,trace 5). Thus, RX2 coelutes with 8-aminoguanosine standard on reverse-phase HPLC. Assuming identity of RX2 with 8-aminoguanosine, the level of this base modification determined by HPLC-EC in the liver RNA of male F344 rats 18 h after the intraperitoneal administration of 100 mglkg 2-NP is approximatly 1.1 mol of 8-aminoguanosine/103 mol of guanosine. When submitted to cation-exchange HPLC-EC (system 4), RNA hydrolysates obtained from livers of rats treated with 2-NP yield two distinct electrochemically active species at elution volumes of -8.5 and -21 mL (Figure 2), identified as 8-oxoguanosine and 8-aminoguanosine, respectively, on the basis of elution volumes. As shown in Figure 2B, 8-aminoguanosine standard as well as RX2 purified by reverse-phase HPLC coeluted with the peak at -21 mL. Using cation-exchange HPLC-EC, we did not observe a peak corresponding to RX1; this may be due to the lower resolution of the cation-exchange columns or to a greatly increased EC base-line noise obtained when using the ion-exchange system. Alternatively, it may be that the instability of RX1 is enhanced under the conditions of ion-exchange HPLC used. The UV spectral characteristics of 8-aminoguanosine standard and purified RX2 under acidic, neutral, and basic conditions are compared in panels A and B of Figure 3. 8-Aminoguanosine standard shows bathochromic shifts
272 Chem. Res. Toxicol., Vol. 6, No. 3, 1993
Sodum et al.
A f
;/
o'6
1 n A Q 600 m V or
0.4
Lu
0
i
4
21 0'
300
350
20
10
-0 210
-....,J-4 0
250
I
-___ 1
,. \----, 10 20
ELUTION VOLUME, mL
Figure 2. Cation-exchange HPLC-EC (system 4) profiles of nucleosides from enzymatic hydrolysis of liver RNA from rats treated with 2-NP. (A) Liver RNA hydrolysate from control or vehicle-treated rats: 1,UV detector response; 2, ECD response. (B) Liver RNA hydrolysate from 2-NP-treated rats: 1, UV detector response; 2-4, ECD response; 2, RNA hydrolysate alone; 3, cochromatography of RNA hydrolysate and purified RX2; 4, cochromatography of RNA hydrolysate and 8-aminoguanosine.
as well as increases in its two absorption maxima as the pH is decreased from 7.2 to 2.2. This is ascribed to the protonation of N7. At the basic pH of 10.3,8-aminoguanosine shows a decrease in the maxima at 259 and 291 nm; this is ascribed to the dissociation of the N1 proton of the guanine ring. As documented in Figure 3B, purified RX2 shows essentially identical UV spectral characteristics. In ancillary experiments (results not shown) both 8-aminoguanosine and RX2 were stable at 25 "C at pH 10.3 for up to 24 h at room temperature. Upon neutralization of such basic solutions, lack of alteration of the nucleoside could be demonstrated by reappearance of the original UV spectra in neutral or acidic solutions, and also by HPLC-EC. Both 8-aminoguanosine and purified RX2 reacted with nitrous acid with loss of electrochemical activity at +600 mV (pH 5.1). Both RX2 and 8-aminoguanosine underwent hydrolysis of the glycosidic bond when heated in methanol containing 10% HC1 at 95 "C for 20 min. After neutralization, the hydrolytic product of RX2 eluted at 1 2 mL on reverse-phase HPLC-EC (system 3) and showed a UV spectrum with absorption maxima at 292, 250, and 220 nm using the photodiode array detector; these characteristics were identical to those of 8-aminoguanine obtained by acidic hydrolysis of 8-aminoguanosine standard. Followingacid hydrolysisof RX2 and trimethylsilylation as described in the Experimental Section, GC-MS analysis indicated a peak eluting at 7.61 min corresponding to the
250
300
350
WAVELENGTH, nm
Figure 3. UV spectra of (A) 8-aminoguanosine standard and (B) purified RX2, at neutral, acidic, and basic pH values.
tetra-TMS derivative of 8-aminoguaninewith the expected molecular ion, M+, at mlz 454, and (M - CH3)+at mlz 439. A larger peak at 7.84 min was also observed, which corresponds to the penta-TMS derivative of 8-aminoguanine, with M+ at mlz 526 and (M - CH3)+at mlz 511,The mass spectra and the retention times of these peaks, as shown in Figure 4, are essentially identical to those of the correspondingtrimethylsilyl derivatives of 8-aminoguanine standard in Figure 5. Evidence for 8-Aminodeoxyguanosine in Liver DNA. HPLC-EC analysis of deoxyribonucleosides obtained from the livers of 2-NP-treated rats showed three electrochemically active peaks (Figure 6). The deoxynucleoside eluting at 30 mL was identified as 8-oxodeoxyguanosine by cochromatography with standard. While small amounts of 8-oxodeoxyguanosineare always present in corresponding preparations from the livers of control (vehicle-treated) or untreated rats, the other two species eluting at 17 mL (designated as DX1) and at 27.5 mL (designated as DX2) are invariably undetectable in liver DNA preparations from control animals. As in the case of RX1, the unknown DX1 appeared to be unstable. The species designated as DX2 eluted with 8-aminodeoxyguanosine standard upon cochromatography, suggesting identity. The properties of DX2 have not been studied as extensively as those of RX2 because of limitations on sample quantity. The yield of DX2 from liver DNA of male F344 rats treated with 100 mg/kg 2-NP is only -2.9 m01/105 mol of dGuo. Efforts to isolate amounts of DX2 sufficient for a more rigorous characterization are underway. Formation of 8-Aminoguanosine and 8-Aminodeoxyguanosine in a Model System. Hydroxylamine-0sulfonic acid, an aminating reagent, reacts with guanosine to produce 8-aminoguanosine at a yield of 20 % in the pH range of 2.0-4.0 (19). We find that, at neutral pH,
Chem. Res. Toxicol., Vol. 6, No. 3, 1993 273
8-Aminoguanine Produced by 2-Nitropropane undanc.
TIC: RSRX299.D
TIC: HH2CN56.D
4000000 1
B
i!
1
B
3500000 1 3000000. 2500000.
A
2000000
1 1 1 ww,q,h
1000000
~
!
5000001
ime
-* 0
7.20
T i m ->
sundance I
RSRX299.D
12000j I
1 1
250000
7.40
7.60
6000
7.80
8.20
8.00
Scan 300 (7.602 m i n ) : HH2GN56.D 4
/ A
OTMS
'I.
8000
I
E
I 1500000
scan 311 (7.605 mi"):
~
MW 454
366
253 277
IIIII undance
Ill Ill
II
381
309 334
I I
I I
I
scan 330 (7.840 a i n ) : RSRX299.D
I
&
'Z -> 511 I
50
150
200
250
1
'
350
300
Scan 317 (7.834 m i " ) :
400
4
HH2GN56.D
I
I
B
500000
1 I 1
100
5ooooj
I
400000 171
40000{ 300000
325
365 397
/Z -D
50
100
150
200
14@4
L L L ,
L
250
300
350
400
450
I1 I
500
GC-MS analysis of the trimethylsilylated acid hydrolysis product of purified RX2. Top: Total ion chromatogram; middle; E1 mass spectrum of peak A; bottom: E1 mass spectrum of peak B.
Figure 5. GC-MS analysis of trimethylsilylated 8-aminoguanine standard. Top: Total ion chromatogram; middle: E1 mass spectrum ofpeak A (8-NHz-G.4TMS);bottom: E1 mass spectrum
hydroxylamine-0-sulfonic acid reacts with guanosine to give not only 8-aminoguanosine but also 8-oxoguanosine (Figure 7). As the pH is increased from neutrality, some N1-aminoguanosine begins to form, and in 1 N NaOH, N1-aminoguanosine is formed in quantitative yield (results not shown) as previously reported by others (20). Hydroxylamine-0-sulfonic acid also reacts with rat liver RNA (Figure 7) and with calf thymus DNA (Figure 8) at pH 7.3 to give 8-oxo- and 8-aminoguanosines and -deoxyguanosines. The yields under the conditions described in the Experimental Section were approximately 3.3, 14.8, 7.9, and 39.7 mol of 8-oxoguanosine,8-aminoguanosine, 8-oxodeoxyguanosine, and 8-aminodeoxyguanosine,respectively, per lo3 mol of guanosine or deoxyguanosine.
&aminodeoxyguanosine/105mol of deoxyguanosinein liver DNA 18 h after the intraperitoneal administration of 100 mg/kg 2-NP to male F344 rats. 8-Aminoguanine is not a naturally occurring minor base in nucleic acids, and its formation in vivo in response to treatment with chemical agents other than 2-NP and certain secondary nitroalkanes (5,8)and ketoximes (6,8) has not been previously reported. Since very little information exists on the in vivo metabolism of 2-NP, especially on the fate of the nitro moiety, it is not easy to provide a plausible explanation for the formation of 8-aminoguanine in rat liver nucleic acids following 2-NP administration. 2-NP is known to undergo oxidative denitrification to nitrite and acetone (21-27); thus, it is conceivable that reactive nitrogen oxide intermediates or byproducts in this process might nitrate or nitrosate guanine and that this might be followed by reduction to the amine. However,the biological reduction of a nitro or a nitroso group attached to a base within a nucleic acid appears unlikely to occur in vivo. The formation of both 8-aminoguanine and 8-oxoguanine as observed in the in vitro reactions of hydroxylamine0-sulfonic acid with rat liver RNA and calf thymus DNA suggests an alternative metabolic pathway capable of generating an aminating species from 2-NP. The most plausible mechanism would be one in which the nucleo-
Figure
4.
Discussion In this work, we have provided chromatographic and spectral evidence for the identity of a major 2-NP-induced base modification in rat liver RNA as 8-aminoguanine. Moreover, our results suggest that this modified base is also present in the liver DNA of 2-NP-treated rats. This base modification is not detectable in the liver nucleic acids of untreated or vehicle-treated rats using HPLCEC. We estimate the amount of 8-aminoguanosine as 1.1 mol/lO3 mol of guanosine in liver RNA, and 2.9 mol of
of peak B (8-NHz-G-5TMS).
274 Chem. Res. Toxicol., Vol. 6, No. 3, 1993
Sodum et al.
/I
8-NH *-dGuo
I\
d;
1
2 5 n A Q 600 m V OT
0 025 AL' D 253 n m
I
0 . 8 AU Q 253 nm ( I ) 0 . 2 nA ( 2 ) 0 . 8 nA ( 3 - 5 )
8,
I
r
I
,
,
8 '
I
I
I
,
I
I
I
I
'
'
I
10 20 ELUTION V O L U M E , mL
0
I
,
'
1
30
Figure 8. Reverse-phase HPLC profiles of the in vitro reactions of hydroxylamine-0-sulfonic acid: 1 and 2, with calf thymus DNA (1,UV response at 253 nm; 2, ECD response); and 3 and 4, with dGuo (3, UV response a t 253 nm; 4, ECD response).
l
l
l
l
l
l
0
l
r
l
l
l
10
1
1
1
1
1
1
1
1
1
1
30
20
1
1
1
40
1
1
1
i
50
ELUTION VOLUME, mL
Figure 6. Reverse-phase HPLC-EC (system 2) profiles of deoxynucleosidesfrom enzymatic hydrolysates of liver DNA from vehicle-treatedrata and rata treated with 2-NP. 1,UVabsorbance at 253 nm; 2-5, ECD responses. 2, liver DNA hydrolysate from vehicle-treated rat; 3, liver DNA hydrolysate from 2-NP-treated rat; 4, cochromatography of DNA hydrolysate from liver of 2-NPtreated rat with 8-oxodeoxyguanosine only; 5, cochromatography of the same DNA hydrolysate with 8-aminodeoxyguanosine only. Note that trace 2 was obtained using a higher ECD sensitivity.
I
1
1
\
H3C\ C=O
+
0 025 Ab' Q 253 nm
p-X H-N VI
1
'
\
- +NH, +
OX-
H
I
8-NHz-Guo
1
Or
IV
H3C
H
I
2 5 n A 0 600 mi/
HOH
X = - S O s H ; -COCH3
8 -AMI NOG U AN IN E
Figure 9. Proposed (hypothetical) activation pathway for 2-NP.
i
I (
,
&cL;'/.::
...
r
I,,
o
i
I
'
i"
_____..^__
~
,-'
, , ' ' 10 20 ELUTION VOLUME, mL I
I
I
I
'& -...- 1 ' '
Figure 7. Reverse-phase HPLC-EC profiles of the in vitro reactions of hydroxylamine-0-sulfonic acid 1 and 2, with rat liver RNA (1,UV response a t 253 nm; 2, ECD response); and 3 and 4, with Guo (3, UV response a t 253 nm; 4, ECD response).
philic carbon-2 of 2-NP nitronate would be converted to an electrophilic carbon and the nitrogen of the nitro group to a nucleophilic center. This could be achieved by deoxygenation of 2-NP, enzymatically or nonenzymatically, and by enzymatic conjugation of the second oxygen to a sulfonyl or acetyl function, as shown in Figure 9. The resulting intermediate IV, a derivative of acetone oxime, could undergo addition of a water molecule followed by hydrolysis to acetone and hydroxylamine-0-sulfonate, or hydroxylamine-0-acetate (VI). The latter would dissociate
to give the highly reactive intermediate NH2+, capable of aminating biological macromolecules. Whether or not the metabolic activation of 2-NP is indeed mediated through these pathways is currently being examined in our laboratory through in vivo and in vitro approaches. In the model reactions of hydroxylamine-0-sulfonic acid with RNA and DNA, the formation of both 8-aminoguanine and 8-oxoguanine nucleosides was observed. Both these products can be derived from one intermediate if the nitrenium ion NH2+could add across the " I 4 8 double bond of guanosine (or deoxyguanosine) to give an aziridine intermediate, VII, which would rearrange to the 8-aminoguanine (VIII) or N7-aminoguanine (IX) nucleosides. In the case of the formation of aromatic amine adducts with guanine at C8, an analogous three-membered ring intermediate has been suggested (13). N7-Aminoguanosine can also be formed by direct reaction of NH2+ at the N7 position of guanosine and then either rearrange to 8-aminoguanosine or, because of the presence of a positive charge on N7, react with water at C8 and lose NH3 to give 8-oxoguanosine (X). As pointed out by Kohda et al. (28) for analogous reactions, this would be one of the few examples of the formation of 8-oxoguanineby mechanisms other than attack by hydroxyl radical or reaction with singlet oxygen (29). Formation of 8-oxoguanosinewas also reported in the reaction of the aminating agent 2,4-
Chem. Res. Toxicol., Vol. 6, No. 3, 1993 275
8-Aminoguanine Produced by 2-Nitropropane r
U
t 0
r
I
t O
II -ut ' H, ,H N
X
K R
K
= R I B O S Y L OR DEOXYRIBOSYL
Figure 10. A possible mechanism for the formation of 8-aminoguanosine (VIII) and 8-oxoguanosine (X) from the same intermediate.
dinitrophenoxyamine with guanosine (30). If NH2+ acts as a cation radical (311, another route for the formation of 8-oxoguanosine could be abstraction of a proton from C8 of guanosine, followed by reaction with H20. 8-Aminoguanosine could also be formed by other reaction pathways, such as (1)insertion of NH2+ at the C8 position of guanosine followed by the loss of a proton and, (2) assuming "ninoguanosine forms a stable ylide by loss of the C8 proton (13),a shift of the NH2+ from the N7 to the C8 position. Although NH2+ was shown by theoretical calculations to be more stabilized in the triplet state (32), its reactive nature has not yet been studied. Substituted nitrenium ions were shown to react in both singlet and triplet states (33). It is also possible that 8-aminoguanosine is formed by one pathway, for instance, by addition of NH2+ across the N7-C8 double bond, followed by rearrangement to the C8-amino product, and that 8-oxoguanosine is formed by another pathway, as illustrated in Figure 10. At the present, experimental evidence for the pathways discussed is not available; however, it appears that the proposed mechanisms can be readily tested since the preparation of "ninoguanosine by a published method (30) is feasible. To what extent the suggested pathways for the formation of 8-oxoguanine in the livers of 2-NP-treated rats supplant the hydroxyl radical-mediated mechanism proposed in an earlier publication from our laboratory (5)remains to be determined. I t has previously been suggested that the tumorigenicity of acetone oxime for rat liver might be due to its metabolic conversion to 2-NP (10). In apparent support of this suggestion, acetone oxime was found to produce a pattern of nucleoside modifications in rat liver DNA and RNA qualitatively identical to that produced by 2-NP ( 6 ) . Indeed, the ability of rat liver microsomes to carry out the N-oxidation of acetone oxime to 2-NP has been described,
and 2-NP has been detected in the urine of rats treated with acetone oxime (34). The hypothetical scheme in Figure 9 provides an alternative mechanism, whereby acetone oxime could be activated directly by conjugation with sulfate, acetate, or other good leaving groups. In this case, 2-NP would represent the precursor of acetone oxime, rather than the oxidation product. In this connection, the metabolic reduction of 2-NP to acetone oxime has also been reported (35). Although the lesser ability of acetone oxime to produce nucleic acid modifications in rat liver, as compared to 2-NP (61, would appear to contradict the precursor-product relationship depicted in Figure 9, it is possible that other factors such as differing rates of absorption, distribution, and excretion might account for the quantitative differences in the biological effects of the two chemicals. In this work we have provided evidence for the formation of 8-aminoguanine, a unique in vivo base modification, in rat liver RNA and DNA as a result of 2-NP administration. Because of their electrochemical activity, nucleosides and nucleotides of 8-aminoguanine, like those of 8-oxoguanine (36), are relatively easy to detect using HPLC-EC. However, it is obvious that 2-NP may induce other, or analogous, nucleic acid base modifications not detectable under the HPLC-EC conditions used here. This possibility, the identification of RX1 and DX1, and finally, the role of these base modifications in the tumorigenicity of 2-NP ( 3 , 4 )and of acetone oxime (10) are topics presently under investigation in our laboratory.
Acknowledgment. The excellent technical assistance of Adriana Surace in parts of this work is gratefully acknowledged. We thank Drs.C. Metral and C. Michejda, of the Frederick Cancer Center, Frederick, MD, for FAB mass spectral analyses and Dr. B. Misra for obtaining the NMR spectra. This work was supported in part by Grant ES03257 from the National Institute of Environmental Health Sciences. References (1) International Agency for Research on Cancer (1982) IARC Monographs on the Evaluation of Carcinogenic Risk of Chemicals to Man, Some Industrial Chemicals and Dyestuffs, Vol. 29, pp 331343, WHO, Lyon, France. (2) Hoffman, D. H., and Rathkamp, G. (1968) Chemical studies on tobacco smoke. 111. Primary and secondary nitroalkanes in cigarette smoke. Beitr. Tabakforsch. 4, 124-134. (3) Lewis, T. R., Ulrich, C. E., and Busey, W. M. (1979) Subchronic inhalation toxicity of nitromethane and 2-nitropropane. J.Enuiron. Pathol. Toxicol. 2, 233-249. (4) Fiala, E. S., Czerniak, R., Castonguay, A,, Conaway, C. C., and Rivenson, A. (1987) Assay of I-nitropropane, 2-nitropropane, 1-azoxypropane, and 2-azoxypropane for carcinogenicity by gavage in Sprague-Dawley rats. Carcinogenesis 8, 1947-1949. (5) Fiala, E. S., Conaway,C. C., Mathis, J. E. (1989) Oxidative DNA and RNA damage in the livers of Sprague-Dawley rats treated with the hepatocarcinogen 2-nitropropane. Cancer Res. 49, 5518-5522. (6) Hussain, N. S., Conaway, C. C., Guo, N., Wagdy, A., and Fiala, E. S. (1990) Oxidative DNA and RNA damage in rat liver due to acetoxime: similarity to effects of 2-nitropropane. Carcinogenesis 11, 1013-1016. (7) Guo, N., Conaway, C. C., Hussain, N. S., and Fiala, E. S. (1990)Sex and organ differences in oxidative DNA and RNA damage due to treatment of Sprague-Dawleyrats with acetoxime or 2-nitropropane. Carcinogenesis 11, 1659-1662. (8) Conaway, C. C., Guo, N., Huasain, N. S., and Fiala, E. S. (1991) Comparison of oxidative damage to rat liver DNA and RNA by primary nitroalkanes, secondarynitroalkanes, cyclopentanoneoxime, and related compounds. Cancer Res. 51, 3143-3147. (9) National Research Council (1981) Selected Aliphatic Amines and Related Compounds: An Assessment of the Biological and Environmental Effects, pp 143-168, National Academy Press, Washington, DC.
276 Chem. Res. Toxicol., Vol. 6, No. 3, 1993 (10) Mirvish, S. S., Salmasi, S.,and Runge, R. G. (1982) Carcinogenicity test of acetoxime in MRC-Wistar rats. J . Natl. Cancer Inst. 69, 961-962. (11) Conaway, C. C., Hussain, N. S., Way, B. M., and Fiala, E. S. (1991) Evaluation of secondary nitroalkanes, their nitronates, primary nitroalkanes, nitrocarbinols, and other aliphatic nitrocompounds in the Ames Salmonella assay. Mutat. Res. 261, 197-207. (12) Andrae, U., Homfeldt, H., Vogl, L., Lichtmannegger, J.,and Summer, K. H. (1988) 2-Nitropropane induces DNA repair synthesis in rat hepatocytes in vitro and in vivo. Carcinogenesis 9, 811-815. (13) Humphreys, W. G., Kadlubar, F. F., and Guengerich, F. P. (1992) Mechanism of CHalkylation of guanine residues by activated arylamines: Evidence for initial adduct formationat the NSposition. Proc. Natl. Acad. Sci. U.S.A. 89, 8278-8282. (14) Long, R. A., Robins, R. K., and Townsend, L. B. (1967) Purine nucleosides. XV. The synthesis of 8-amino- and 8-substituted aminopurine nucleosides. J. Org. Chem. 32, 2751-2756. (15) Cho, B. P., Kadlubar, F. F., Culp, S. J., and Evans, F. E. (1990) 15N nuclear magnetic resonance studieson thetautomerism of 8-hydroxy2’-deoxyguanosine, 8-hydroxyguanosine, and other C8-substituted guanine nucleosides. Chem. Res. Toxicol. 3,445-452. (16) Kasai, H., and Nishimura, S. (1984)Hydroxylation of deoxyguanosine at the C-8 position by ascorbic acid and other reducing agents. Nucleic Acids Res. 12, 2137-2145. (17) Marmur, J. (1961)A procedure for the isolation of deoxyribonucleic acid from microorganisms. J. Mol. Biol. 3, 208-218. (18) Chomczynski,P., and Sacchi, N. (1987)Single-step method of RNA isolationby guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156-159. (19) Kawazoe, Y., and Huang, G.-F. (1972) Direct C-amination of guanosine with hydroxylamine-0-sulfonic acid. Simplest model reaction possibly involved in DNA damage leading to carcinogenesis by N-arylhydroxylamines. Chem. Pharm. Bull. 20, 2073-2074. (20) Broom, A. D., and Robins, R. K. (1969)Purine nucleosides. XXIII. Direct amination of purine nucleosides. J . Org. Chem. 34, 10251029. (21) Sakurai, H., Hermann, G., Ruff, H. H., and Ullrich, V. (1980) The interaction of aliphatic nitro compounds with the liver microsomal monooxygenase system. Biochem. Pharmacol. 29, 341-345. (22) Ullrich, V., Hermann, G., and Weber, P. (1978) Nitrite formation from 2-nitropropane by microsomal monoxygenases. Biochem. Pharmacol. 27, 2301-2304.
Sodum et al. (23) Marker, E. K., and Kulkarni, A. P. (1985) Enzymatic denitrification of 2-nitropropane in uninduced mouse liver microsomes. Toxicol. Lett. 26, 181-185. (24) Russell, G. A. (1954) The autoxidation of 2-nitropropane in basic solution. J . Am. Chem. SOC.76, 1595-1601. (25) Kido, T., and Soda, K. (1984) Oxidation of anionic nitroalkanes by flavoenzymes,and participation of superoxide anion in the catalysis. Arch. Biochem. Biophys. 234, 468-475. (26) DeRycker, J., and Halliwell, B. (1978) Oxidation of 2-nitropropane by horseradish peroxidase. Biochem. J . 175, 601-606. (21) Porter, D. J., and Bright, H. J. (1983) The mechanism of oxidation ofnitroalkanes by horseradishperoxidase. J.Biol. Chem. 258,99139924. (28) Kohda, K., Tada, M., Kasai, H., Nishimura, S., and Kawazoe, Y. (1986) Formation of 8-hydroxyguanine residues in cellular DNA exposed to the carcinogen 4-nitroquinoline-1-oxide.Biochem. Biophys. Res. Commun. 139,626-632. (29) Floyd, R. A., West, M. S., Eneff, K. L., and Schneider, J. E. (1990) Mediation of 8-hydroxy-guanine formation in DNA by thiazin dyes plus light. Free Radical Biol. Med. 8, 327-330. (30) Kohda, K., Baba, K., and Kawazoe, Y. (1986) Synthesis of 7-aminoguanosine and its conversion to 8-substituted derivatives. Chem. Pharm. Bull. 34, 2298-2301. (31) Gassmann, P. G., and Cryberg, R. L. (1969) The discrete existence of singlet and triplet nitrenium ions. J . Am. Chem. SOC. 91, 51765177. (32) Lee, S. T., and Morokuma, K. (1971) A b initio molecular orbital studies on the structure of the nitrenium ion and its implications. J . Am. Chem. SOC. 93, 6863-6866. (33) Gassman, P. G. (1970) Nitrenium ions. Acc. Chem. Res. 3, 26-33. (34) Kohl, C., Schiller, C. D., Gescher, A., Farmer, P. B., and Bailey, E. (1992) Acetoxime is metabolized by human and rodent hepatic cytochrome P450 enzymes to the genotoxicant and carcinogen propane 2-nitronate. Carcinogenesis 13, 1091-1094. (35) Haas-Jobelius, M., Ziegler-Skylakakis, K., and Andrae, U. (1991) Nitroreduction is not involved in the genotoxicity of 2-nitropropane in cultured mammalian cells. Mutagenesis 6, 87-91. (36) Floyd,R. A., Watson, J. J.,Wong, P. K., Altmiller, D. H., andRickard, R. C. (1986) Hydroxyl free radical adduct of deoxyguanosine: sensitive detection and mechanism of formation. Free Radical Res. Commun. 1, 163-172.
