Chem. Res. Toxicol. 1991,4, 364-368
364
Mass Spectrometric Analysis of Tobacco-Specific Nitrosamine-DNA Adducts in Smokers and Nonsmokerst Peter G. Foiles,*J Shobha A. Akerkar,t Steven G. Carmella,i Mark Kagan,i Gary D. Stoner,s James H. Resau,II and Stephen S. Hechti Division of Chemical Carcinogenesis, American Health Foundation, 1 Dana Road, Valhalla,
New York 10595, Department of Pathology, Medical College of Ohio, Toledo, Ohio 43699, and Department of Pathology, University of Maryland School of Medicine, Baltimore, Maryland 21201 Received December 19, 1990
A gas chromatography, negative ion chemical ionization mass spectrometry (GC-NICI-MS) based assay for tobacco-specific nitrosamine adducts of DNA is described. T h e assay is based on the observation that acid hydrolysis of DNA from animals treated with tobacco-specific (HPB). HPB and the internal standard nitrosamines releases 4-hydroxy-l-(3-pyridyl)-l-butanone [4,4-D2]HPBare derivatized with pentafluorobenzoyl chloride and the resulting HPB-pentafluorobenzoate is purified by high-performance liquid chromatography prior to GC-NICI-MS analysis. DNA from human peripheral lung and tracheobronchial tissue, collected at autopsy, was analyzed for acid-released HPB. The mean H P B level (fmol/mg of DNA) for peripheral lung DNA was 11 f 16 (SD, n = 9) for smokers and 0.9 f 2.3 ( n = 8) for nonsmokers. Mean adduct levels in tracheobronchus were 16 f 18 ( n = 4) for smokers and 0.9 f 1.7 ( n = 4) for nonsmokers. These are the first measurements of tobacco-specific nitrosamine-DNA adducts in humans. Further studies comparing the levels of DNA and globin adducts will provide a better understanding of the metabolic activation of tobacco-specific nitrosamines in humans and may provide a more accurate indication of an individual's risk of developing tobacco-related cancer.
Introductlon The tobacco-specific nitrosamines, 4-(methylnitrosamino)-1-(3-pyridyl)- 1-butanone (NNK) and "-nitrosonornicotine (NNN), are carcinogens in animal models. NNK causes tumors primarily of the lung, but also of liver, nasal cavity, and pancreas (1,2). NNN causes nasal cavity and esophageal tumors ( I ) . A mixture of NNK and NNN applied to the oral cavity produced oral tumors in rats (3). NNK and NNN are present in unburned tobacco at ppm concentrations and in mainstream and side-stream cigarette smoke a t concentrations of 0.1-1 wg/cigarette (4). The carcinogenic activities of NNK and NNN, and their relatively high concentrations in tobacco products, suggest that they are causative agents for cancers of the lung, oral cavity, esophagus, and pancreas in smokers and others exposed to tobacco products. For NNK and NNN to exert their carcinogenic effects they must undergo metabolic activation to reactive intermediates which bind to cellular macromolecules. As illustrated in Figure 1,methyl hydroxylation of NNK or 2'-hydroxylation of NNN produces a reactive diazohydroxide that modifies DNA or globin, yielding adducts or unknown structure. These adducts release 4-hydroxy1-(3-pyridyl)-l-butanone(HPB) upon hydrolysis. Thus, acid hydrolysis of DNA or base hydrolysis of hemoglobin isolated from rodents exposed to NNK or NNN releases HPB (Figure 1;5-7). On the basis of these observations, a sensitive mass spectrometric method for measuring HPB released by base hydrolysis of hemoglobin has been developed in this laboratory (8). This paper - - describes a This work supported by National Institutes of Health Grants
CA-29580 and ES-5787.
American Health Foundation. Medical College of Ohio. 'I University of Maryland School of Medicine.
DNA
t
OH-$
Globin
Adducts
HPB
Figure 1. Metabolic activation of the tobacco-specific nitrosamines NNK and NNN to reactive species that can pyridyloxobutylate DNA and proteins.
modification of this method to allow the detection of tobacco-specific nitrosamine-DNA adducts by measuring
0~93-22~x/91/2~04-0364~02.50/0 0 1991 American Chemical Society
Chem. Res. Toxicol., Vol. 4, No. 3, 1991 365
GC-NICI-MSAnalysis of TSNA Adducts HPB released by acid hydrolysis of DNA. The ability to measure binding of tobacco-specific nitrosamine metabolites to DNA and protein provides a means of assessing the extent of metabolic activation of tobacco-specific nitrosamines. An individual's ability to metabolically activate tobacco-specific nitrosamines should be a more accurate indicator of that individual's risk of developing tobacco-related cancers than measurn" of exposure such as levels of nicotine and cotinine. While measurement of globin adducts provides a good indication of the overall extent of metabolic activation in an individual, it does not provide information on the level of activation in individual target organs, such as the lung. Measurement of DNA adducts can provide information on the extent of metabolic activation in these target organs. In addition, the development of an assay for tobaccospecific nitrosamine DNA adducts allows us to determine the relationship between DNA and globin adduct levels. This information will permit us to assess the validity of using the more easily made measurements of globin adducts as surrogates for measurements of DNA adducts in target organs such as lung. Experimental Procedures Tissue Samples. Tissue samples were collected from autopsy (leas than 24 h) by Dr. Gary Stoner of the Medical College of Ohio and Dr. James Resau of the University of Maryland School of Medicine. Tissues were placed in L-15 medium and transported to the laboratory where they were frozen immediately at -80 OC. The tissue samples were shipped on dry ice to the American Health Foundation where they were stored at -20 OC. The smoking status of the donor was established from medical histories or from nicotine and cotinine levels in the blood determined as part of a general toxicological screening performed by the Medical Examiner's office in Baltimore, MD. DNA was extracted from the tissue samples essentially as described (9). Peripheral lung samples were from 15 to 25 g in size and tracheobronchussamples from 2 to 4 g. Briefly, the tissue was minced and placed in cold 150 mM NaCl and 15 mM sodium citrate, pH 7.0, buffer and homogenized. The nuclear pellet was collected by centrifugation and lysed. The lysate was extracted twice with chloroform/isoamyl alcohol (51)and the crude DNA precipitated with cold ethanol. The DNA was redissolved in 0.3 M sodium acetate and treated with RNase A. The resulting mixture was extracted with chloroform/isoamyl alcohol and the DNA precipitated with cold ethanol. The concentration of DNA solutions was determined by their absorbance at 260 nm. The AZBO/AzLM ratios of the DNA samples were between 1.8 and 2, indicating the lack of significant protein contamination. Analytical Scheme. The analytical scheme for the assay is outlined in Figure 2. The method is a modification of that described for analysis of HPB released from hemoglobin (8). Because of the low levels of HPB being measured a number of precautions were taken to minimize contamination as outlined in ref 8. These include the use of silanized vials, Teflon and polyethylene cap liners, and a dedicated laboratory. Approximately 1-2 mg of DNA was dissolved in 1 mL of HzO; 12 N HCl was added to bring the final concentration of acid to 0.8 N. The DNA was then hydrolyzed for 3 h a t 80 OC and the hydrolysate extracted twice with an equal volume of CHZCl2.The aqueous phase was saved, the pH adjusted to 7.0 with 1 N NaOH, and the sample spiked with 80 pg (480 fmol) of [4,4-D2]HPB(8). The sample was then extracted twice with equal volumes of CHzCI2, the organic phases were pooled, and the solvent was removed mth a SpeedVac concentrator. The residue, containing HPB, was redissolved in 1 mL of CHzClz for derivatization. Abbreviations: CC-NICI-MS, gas chromatography-negative ion chemical ionization-mass spectrometry;NNK, 4-(methylnitroeamino)1-(3-pyridyl)-l-butanone; NNN, N'-nitrosonornicotine; HPB, 4hydroxy-l-(3-pyridyl)-l-butanone; TSNA, tobacco-specificnitrosamine; THF, tetrahydrofuran.
( D N A Solution
1
I
a) 0.8N HCI 80% 3h
I a) io p~
I
,
7 b) add internal std
O
a) b) c) d)
concentrate derivatire N HPLC add injection standard
F -@F
F
F
F
-vH 0
MW 341)
Figure 2. Scheme for analysis of HPB released by acid hydrolysis of DNA. Derivatization was carried out with a trimethylamine/hexane solution prepared as follows: 240 mg of trimethylamine hydrochloride (Sigma, St. Louis, MO), 2.4 mL of 1N NaOH, 17.6 mL of HzO,and 20 mL of hexane were mixed together, and the hexane layer was removed. The hexane layer was dried with approximately 2 g of anhydrous NazSOI. One milliliter of this solution was added to the CHzClzsolution of HPB. One microliter of pentafluorobenzoyl chloride (Aldrich, Milwaukee, WI) was added and the solution incubated at room temperature for 2 h. The solvent was removed by using a SpeedVac concentrator. Unreacted pentafluorobenzoylchloride was removed from the sample by reversed-phase HPLC under the following conditions. A 4.6 mm X 12.5 cm Whatman Partisil5 ODs-3 cartridge column was eluted with a program of 35% MeOH in HzO for 10 min followed by a linear gradient from 35% to 70% MeOH in 15min a t 1 mL/min. For HPLC analysis the sample was redissolved in 100 pL of a 1:l MeOH/THF solution containing 25 pg/mL pentanophenone and hexanophenone (Pierce Chemical Co., Rockford, IL) as UV markers. The injection volume was 70 fiL, and the region between the apexes of the pentanophenone and hexanophenone peaks was collected and dried by using the SpeedVac concentrator. The residue was taken up with three aliquota of approximately 50 pL of THF and transferred to a silanized 250-pL conical vial. For GC-NICI-MS analysis, a 0.25 mm i.d. X 30 m 50% methylphenylsilicone (DB-17) bonded-phase column (0.15 Ccm film thickness), connected to a 2 m X 0.32 mm i.d. fused silica uncoated deactivated retention gap, was used in an HP 5988A GC/MS (Hewlett Packard, Avondale, PA). Injections were carried out with a septumleas on-column injector (J and W Scientific, Folsom, CA). Injections were made with a 10-pL syringe equipped with a 5-in. fused silica needle. The sample was deposited a t an injection speed of 1 pL/s onto the retention gap in an oven equilibrated at 35 OC. After injection the oven temperature was programmed as follows: 35 OC for 2.5 min, then 20 "C/min to 150 OC, then 4 OC/min to 205 OC, and then hold for 20 min. The flow rate was 1 mL/min He measured a t 35 O C . The sample was taken up in 10 pL of ether containing 60 fmol of the external standard, HPB-2,3,4,5tetrafluorobenzoate (8), and 50% of the sample was injected. The mass spectrometer was operated in the NICI mode with a methane pressure of 1.2 Torr and a source temperature of 150 OC. The molecular ions of HPB-pentafluorobenzoate ( m / e 359), [4,4-DzJHPB-pentafluorobenzoate ( m / e 361), and HPB-2,3,4,5-tetrafluorobenzoate (m/e 341) were monitored. A HzO blank was run for each set of samples. The mean value for the H20 blank was 38 f 16 fmol of HPB. The blank value was subtracted from each sample value in that set of samples. The results are reported as fmol of HPB/mg of DNA with the background subtracted.
Results An outline of the analytical method is given in Figure
"'?r
366 Chem. Res. Toxicol., Vol. 4, No. 3, 1991 500 450
I;
TJ400-
P
-
350+
1000
2 3000
Foiles et al.
4
m 250-
a
-
6000 mle 360.9
I200-
-
-
