Document pmyz2v45gOkb6jV2BB5YwBE2a

utngencsis vo1.8 IM 2 pp.121-126. 1993 tandardizatian of the 32P-postlabeling assay for polycyclic omatii hydrocarbon-DNA adducts p.G.Shields, C.C.Harris', SPetruzielli, E.D.Bowman and A.Weston Laboratory of Hunian Cariinogenesis. Satianal Cancer Institute. National Institutes of Health. Bethe.&. \1D 20892. USA 'To whom correspondence should 'kaddressed A modified method for the 3'P-postlabeling assay that permits standardized quantitation for specific polycyclic aromatic hydrocarbon ( P W - D N A adducts is described. This method has been designed to test the components of the 32P-postlabeling assay for use in the chemically specific detection of individual adducts vith the ultimate goal of testing the biological significanceof PAAH-DNA adducts in humans. The approach relies upon high performance liquid chromato- graphy (HPLC),concomitant labeling of 2'-deoxyguanosine (dGp) as an internal standard and thin layer chromatography m C ) , which identifies unmodified nucleotides along with PAH-DNA adducts on the same TLC plate. This method assesses labeling eftlcienc!. detects the presence of unknown inhibitors, assesses the adequac! of digestion (when combined with HPLC) and allows for the development of calibration curves for directly determined molar ratios (adducthternal standard). Chemically synthesized adduct standards and quantitative 32P-postlabelingdata have been corroborated by W spectroscop! , fluorescence spectroscopy and liquid scintillation counting where radiolabeled materials were available. Labeling efficiencies Of the PAH-DNA adducts were found to be up to 100-fold less than expected and depended upon both the adduct and adduct levels (lower levels being less efficiently detected). The presence of unmodified nucleotides resulted in a 1.5-fold lower labeling efficiency. &tures of selected PAH-DNA adducts did not affect the labeling efficiencies of each other. The data suggest that Previous 32P-postlabeling assay studies for PAH -DNA adducts may have underestimated adduct levels due to Vhtions in labeling efficiencr. Introduction Molecular epidemiology is a multi-disciplinary field that seeks to identify and quantitate cancer risk in individuals, incorporating h s o n s from epidemiology, molecular biology and toxicology (Shields and Harris, 1991). Cancer risk is a function of DNA exposure to carcinogens as well as acquired and inherited genetic Predispositions. Exposure can be measured by carcinogen-DNA adducts (or their surrogates)and has been the focus of investigations by numerous laboratories. DNA adduct levels reflect exogenous carcinogenic environmental exposure, absorption and metabolic activation, as well as DNA repair. Thus, accurately determined levels will assess interindividualvariation for exposure to xenobiotics and enhance cancer risk assessments. Carcinog e n - D N ~ adducts can be measured by a variety of methods (immunoassays, electrochemical detection, 32P-postlabeling fluorescencespectroscopy and mass spectroscopy) (Harris @ Oxford University et al., 1987; Weston et al., 1989b; Randerath et al., 1981). The accuracy of these assays, however, are challenged when applied to human samples due to low levels of adducts (frequently at the detection limit) and are confounded by multiple carcinogen exposures. The 32P-postlabeling assay is among the most sensitive methods used to detect carcinogen-DNA adducts (Randerath et al., 1981). Postlabeling analysis of human DNA samples, for example, can detect adducts at a reported level of up to one adduct per 10" nucleotides (5 fig DNA) (Phillips et al., 1988a; Foiles et al., 1989). Commonly, many putative adducts are detected as a diagonal radioactive zone (Randerath et al., 1986; Phillips el al., 1988a,b; Foiles et al., 1989; Schoket er al., 1988; Savela et al., 1989). Quantitation is based upon specific activity calculations or by relative adduct labeling efficiencies compared to vast excesses of normal nucleotides. Modifications of the original procedure (Randerath et al., 1981): nuclease PI (Randerath er al., 1981; Reddy and Randerath, 1986, 1987; Gupta and Randerath, 1988), butanol extraction (Gupta, 1985; Gupta and Randerath, 1988), digestion to dinucleotides (Randerath et al., 1989) or 'adduct intensification' (adenosine triphosphate deficient conditions, assuming adducts label better than unmodified nucleotides) (Randerath et al., 1985; Gupta and Randerath, 1988) have been used to increase the resolution and sensitivity. However, even with these modifications, these assays still lack chemical specificity, labeling efficiencies are typically not determined and co-chromatography remains a problem. Addressing these limitations, modificatians of the 32Ppostlabeling assay have been used to detect specific alkyl adducts in human samples (Wilson et al., 1988; Shields et al., 1990). These have utilized high performance liquid chromatography (HPLC) and co-chromatography of standards to enhance chemical specificity. Internal and adducted standards, and determination of calibration curves with direct molar rztios were also used to enhance quantitative reliability. We now seek to develop a similar 32P-postlabelingassay for PAH-DNA adducts. The goal of this method is to attain chemical specificity for use in corroborating the32P-postlabelingassay and other detectioii methods as they relate to exogenous exposures. Chemically specific assays would also be useful for testing the biological significance of newly identified genetic polymorphisms or interindividual variations related to arylhydrocarbon hydroxylase (CYP1A 1) inducibility (Kouri et al., 1982; Hawke and Farrell, 1986; Kawajiri et al., 1990; McLemore et al., 1990). Materials and methods Chemicals r-7.1-8-dihydroxy-1-9I,O-epoxy-7,8,9,IO-tetrahydrobenzo[a]pyrene(BPDE), r-1,f2-dihydroxy-t-3,4-epoxy-l,2,3,4-tetrahydrochryse(CnehDE), r-4,1-3dhydroxy1-1, 2-epoxy-1,2,3,4-tetrahydroben[a]anthraccnc (1,2,3,4-BADE), r-8,?-9dihydroxy- 1-10,lI-epoxy-8,9,10,11-tctrahydrobenz[a]anthracene (8,9,10,11BADE), 7R,8S,9~-trihydroxy-I1~R-(N2-dcoxyguanosyl-3'-phosphate)-7,8,9,10tctrahydrobcnzo[a]pyrcnc (BF'DE-dGp), 1R,2S,3S-trihydroxy-4R-(N2deoxyguanosyl-3'-phosphate)-],2,3,4-tctrahydrochrysene(ChDE-dGp) and r-7,f8.1-9.r-IO-retrahydroxy-7,8,9,10~trahydrobenzo[a]pyrenewere purchased from Midwest Research Institute (Kansas City, MO). Tritiun~labeled benzo[a]pyrcne 121 . .P.C.Shhlds et aI. @[alp) (96Ci/mM)was obtained from Amersharc (Arlington Heights, IL).Micro- coccal nuclease and nuclease P1 were purchasul from Sigma (St Louis, MO). Calf spleen phosphodiesterase was purchased frcm Boehringer (Indianapolis, IN). T4- polynucleotide kinase (<5.0 nmoVunit 3' phosphatase activity) was procured from New England Nuclear (Boston, MA). [Y-~*P]AT(Pspec. act. >5000 Ci/ mM) was obtained from Amenham (Arlington Heights, IL).Triethylamineacetate was purchased from Applied Biosystems (Foster City, CA). All solvents were reagent or HPLC grade. Stanhrds Deoxyribonucleoside-3'-monophosphateadducts were prepared from modified DNA as previously described (Arnin et ai., 1989; Weston et ai., 1989a). Briefly, either BPDE (1 mg), ChDE (1 mg), 1,2,3,4-BADE (3 tng) or 8.9,lO.Il-BADE (3 mg) in tetrahydrofuran:ethanol (4 ml, 1:3) were allowed to react with calf thymus DNA (1.7 mg/ml; 7 ml; Tris-HCl, 100 mM, pH 7.4) for 16 h at 24-26C. The samples were extracted with diethyl ether (eight times) and isoamyl alcohol (twice) and then precipitatedby the addition of NaCl(100 mM) and ethanol (2.5 volumes) (weston et al., 1989a). Modified DNA samples were then digested to deoxynucleoside-3'-monophosphates using micrococcal nuclease and calf spleen phosphodiesterase (Wilson et ai., 1986). Synthesized adducts were purified by HPLC Method 1, which was reverse-phase (Beckman Altex Ultrasphere ODS 5 pm: 4.6 mm X 2.5 cm) with gradient elution (0-80% methanol in water over 55 minat 1 mllmin). Adducts were alsopurified by HPLC Method 2 (also utilized for isolation of adducts from DNA of SV40-immortalized human liver epithelial cells treated with B[a]P as described below), which was ion-pair reverse-phase (Beckman Altex Ultraspherc ion pair column 5 pm; 4.6 mm x 25 cm) in an isocratic system (triethylamine acetate 0.1 M; pH 7.0 and 2 % acetonitrile at 1 ml/min for 15 min) and subsequent gradient elution (2 - 95% acetonitnle over 65 min). 2'-Deoxyguanosine-3'-monophosphate(Sigma) (dGp) was purified by HPLC according to Method 2 above. '2P-post/abeiing cssay The 32P-postlabelingassay (Randerath et al., 1981) was modified to use dGp as an internal standard. Reaction mixtures contained different quantities of adducts (700 amol to 700 fmol), dGp (0.7-70 pmol), water (5 pl), postlabeling buffer (1.2 pl; bicine 0.1 M, spermidine LO mM MgC1,O.l M; pH 9.5). dithiothreitol P(1.0 pl; 0.1 M), adenosine triphos hate (ATP) (0.05-0.6 nmol; ATP limiting to ATP excess conditions), [y3PIATP (2 pl; >5000 Ci/mM) and poly- nucleotide kinase (PNK) (2 pl; 5 unitslpl). The ATP and the [r-'?P]ATP were premixed in order to maintain identical specific activity among samples. The samples were incubated at 37C for 15 min to 6 h. Polyethyleneiminecellulose thin layer chromatography (TLC) plates (Machery Nagel, Germany; 20 x 20 cm) were prespotted with UV detectable (254 nm) quantities of non-radiolabeled 2'-deoxyguanosine-3',5'-bisphosphate (pdGp) (Phamacia, Piscataway, NJ) before loading 9- 10pl of postlabeled sample (PCR Pipettes; Drummond Scientific, Brogmull, PA) and developed according to S c k I. The essential difference from other TLC systems is that, after cutting the plate with scissors, unmodified nucleotides are retained and developed in a m n d dimension rather than discarding them on a wick. The U V marker aids in the determination of where to cut the plate and confirms the location of postlab&j dGp. The portion of the plate containing unmodified nucleotides was thendeveloped a.in D2a to segregate ATP from pdGp and exposed to X-ray plates for 2 The portion of the plate containing adducts was developed overnight in D2b after attaching a wick (Whatman 3 mm CHR) with staples. Following removal ofthe wick, the plates were washed in water and developed in D3 and D4 as d e s c n m in Scheme I. Autoradiography was carried out for 1 h to 5 days depending on adduct lev.& (-70C; Intensiijing screens). Unxnodif~edand adducted nucleotides were from the plates and the radioactivity was measured by mixing with HCI (1 d; 0.1 N) and liquid scintillation counting. Direct molar ratios based upon counts per minute for PAH-DNA adduct levels to dGp were then determined. Immortalii SV40 large T antigenmfe=ted human Liver epithelialcellswere treated with [3H]B[a]P (1.5 mM: 52 mCi; Amersham, Arlington, IL).These cells have been shown to express CYPlAl mRNA by Northern Blot analysis (Pfeiffer et ai., in press). DNA was extracted as previously reported ( W m n et al., 1989d). Incorporation of tritium into DNA was determined using liquid scintillation counting. DNA was enzymatically digested to nucleotides with calf spleen phosphodiesterase(Boehringer-Mannheim, Piscataway, NJ) and nuclease (Sigma) (Shields et al., 1990). The de~xynucleoside-3'-monophosphate digest was subjected to HPLC Method 2 and then the 'LP-postlableing assay. In this case, the dGp fraction was collected. diluted 1/1OOO (to equal 70 pmol) and added back to a 3 ml fraction containing BPDE-dGp. Results Purijcation of adducts HPLC elution profiles for the BPDE-. ChDE-. 8.9.IO. 1 1-BADEand 1,2,3,4-BADE-modified- DSX are shown in Figure 1. Method 1 was used to rapidlt purify adducts while method 2 was utilized to assess adequacy of enzymatic digestion, as evidenced by the area under the peaks for each nucleotide. The major adduct peak for each PAH treatment was collected and analyzed. The authenticity of thz products aere confirmed by fluorescence excitation, synchronous fluorescence spectroscopy (SFS) and UV spectroscopy b> comparison with published spectra for the corresponding tetrols (Karcher er 01.. 1983). Isolated Step 6 7 8 9 '0 Description Solvent Prewash D1 Wash cut D2a D2b Wash D3 Wash 04 Methanol (5 min) 1.25 M Lithium Chloride Methanol (20 min) Sat. Ammonium Sulfate/ Isopropanol/lM Sodium Acetate (80:2:18 v/v) 2.3M Sodium Phosphate pH6 (with wick by stapling) Water (15 min twice) 3.5M Lithium Formatel 8.5M Urea pH 3.5 Water (15 min twice) 1.2M Lithium Chloride/0.5M Tris-HCI/8.5M Urea pH 8.0 f \ \ \ \ \ \ \ \I ULU \ I1 I I I .Scheme 1. Cut and develop for unmodified dGp determination A a ib B -> 2 0 0 t Fig. 1. HPLC profiles for the majx ~eosynucleoside-3'-monophosphate adducts found in enzymatic digess of DNA samples that had been treated With diolepoxides of various P.-wC. The major peak for each diol-epoxide treatment was collected and utilized for development of the 32P-postlabelkkz m y . Method 1 (A) and method 2 iB) are described in the text. Insets are blown up profiles from 40 to 50 min for BPDE-DNA (a), ChDE-DNA (b), 8,9,10,11-BADE-DNA (c) aDd 1.2.3,4-BADE-DNA (d). adducts \?ere quanlit;ircii h) LTV 111:)l;ii. L . \ t i i i i , i i t i s :\'t>. i ,<'I, : I : \ (Karchcr ct c i l . , 1983) snd corrohoratc*tl ti!, SI:S ;I~'ILY ;ii.id hydrolysis (Wcston ct (11. . 198%) and coiiip;u.isoii to I\iio\vn quantities of the corrcsponding tctrols. The qiiaiiti1ati\)caccuracy of SFS for thcsc adducts \vas prc\.iously confiriiicd b ~ n,i;iss spectroscopy (Manchester er ( I / . . 1988). The BPDE-dGp and ChDE-dGp "P-postlabeled products had idcntical TLC retention properties compared with postlabeled products of authentic reference materials purchased from the National Cancer Institute Carcinogen Standard Repository (1.2,3,4-BADE- and 8,9,10.11-BADE-dGp were not examined). 31P-pos~lubelingassuj Figure 2 shows representative autoradiograms for each of the adducted and unmodified nucleotides. Labeling efficiency of adducts was maximized by testing incubation periods and ATP concentrations (utilizing 0.7 pmol of adduct and 70 pmol of dGp, equivalent to one adduct in Id dGp). Incubation time did not have a significant impact on labeling efficiency. The incubation time chosen for subsequent experiments was 30 min. The wide range of ATP conditions (excess to ATP limiting) did not result in preferential adduct labeling. The amount used in subsequent experiments was 0.5 ~1 of 0.5 mM ATP to allow for highest specific activity without using ATP-limiting conditions. Radio- activity incorporated into unmodified dGp (70 pmol) was typically 2 - 6 X lo6 c.p.m., depending on the initial specific activity. Failure of the reaction to work properly, for example due to the presence of inhibitors in the labeling mixture or defective kinase, would result in significantly lower counts (100- to 1000-fold). Different lithium chloride coilcentrations were tested (0-2 M) to assess potential migration of adducts during D1 into the c u t h g zone, but no migration was observed. Methanol washing of the plates to remove LiCl does not result in adduct loss because washing with water gave the same results. The labeling efficiencies were found to vary both by adduct concentration and adduct type (Figure 3). For the range of levels previously observed in human samples (Shields ef al., 1991), the labeling efficiency of all the adducts was < loo%, with a 2- to 100-folddiscrepancy in known versus detected ratios (adduct to dGp). The slopes of the curves for BPDE-dGp and ChDE-dGp were similar in that there is almost a one-to-one labeling efficiency for higher adduct concentrations, but a 2- to AB C *9t 8,9,10,11 BADE-dGp BPDE-dGp t CHDE-dGp 1,2,3,4 BADE-dGp F *' 2* "P-WStk~beling assay autoradiograms. Unmodified dGp together with unincorporated ATP on the portjon of the plate that is cut after development of Dl and then development in D2a (A). Autoradiograms showing "P-postlabeling BPDE-dGp and 8,9.10,1 I-H4DE-dGp (R), ChDE-dGp (C) and 1'2'314-BADE-dGp (D) were de\eloped according to Schemc 1. 123 P.G.ShielBs et nl. A 0.010 0.001 o.Ooo1 -/0 . m 1 o.Ooo1 0.001 0.010 0. A Level of BPDE-dGp/dGp 1.OE-6 1.OE-7 r+A -~ * 0 0.007 0.035 0.07 0.07 Level of ChDE-dGp (pmol) a:U o.Ooo1 0.001 0.010 0.10 r-1O.L'l0 oz;[#0 . m 1 u o.ooo1 0.001 0.010 0.10 Ratio - Known Fig. 3. Calibration curves for known (UV and fluorescence spectroscopy) versus determined (32P-postlabelingassay) molar ratios of BPDE -dGp/dGp (A), ChDE-dGp/dGp (B) and 8,9,10.11-BADE-dGp/dGp (C). Seventy picomoles of dGp is mixed with varying amounts of adduct (700 amol to 700 fmol). The detection limit was one adduct in lo7 unmodified dGp. Each point represents a mean of six separate sample determinations. Error bars indicate SEM. Table I. Labeling efficiency for BPDE-dGp adducts with and without dGp' Known ratiob +dGpb -dGpb *Ratio dGp 0.01 0.001 o.Ooo1 11 868 1936 254 63 278 626 I 384 5.3 3.2 1.5 aLevels represent the mean of six samples expressed as counts per minute. bDetermined by UV detection. 4-fold lower labeling efficiency for lower adduct levels. The slope of the curve for 8,9,10,11-BADE-dGp indicated a 10-fold lower labeling efficiency at all adduct levels. A calibration curve for 1,2,3,4-BADE-dGp could not be determined because the highest level tested (0.7 pmol) was at the detection limit (100-fold lower efficiency). The effect of the internal standard on labeling efficiency was also tested by "P-postlabeling of BPDE-dGp in the presence and absence of dGp. Presence of unmodified nucleotides resulted in reduced incorporation of radioactivity (lower counts per minute) but the effect was less at lower adduct levels (Table I). Because unmodified nucleotides can affect the labeling efficiency of PAH-DNA adducts, the effects of individual PAH - DNA adducts on each other were also examined. Mixtures of BPDE-dGp and ChDE-dGp, in the presence o f K i p . did not alter the labeling cfficiency over 2 log orders of concentrations (range of 7- 700 fmol) (Figure 4). Similar results LVCW Kouncl l o r BPDE -dGp and 8.9. I O , I I BADE -dGp 123 B Level of ChDE-dGp/dGp 1.OE-6 5.OE-7 1.OE-7 I 1 --- , T 0 0.007 0.07 0.07 0.07 Level of BPDE-dGp (pmol) Fig. 4. Labeling efficiency for DNA adduct mixtures. The ratio of known (UV and fluorescence spectroscopy) versus determined (32P-postlabeling assay) are shown for two different BPDE-dGp levels with 0.007-0.07 pmol ChDE-dGp or without CHDE-dCp c.4) or for three levels of ChDE-dGp using 0.007-0.07 pmol or uithout BPDE-dGp. Each point represents a mean of six separate sample determinations. Error bars indicate SEM. The resultant limit of detection uas 7 fmol of adduct. When combined with micropreparative techniques. e.g. HPLC or immunoaffinity chromatograph! and dilution of dGp as previously described for alkyl adducts (Wilson et al., 1988; Manchester et al., 1990; Shields er 01.. 1990). the overall limit of detection for this system. therefore. is at least one adduct in 10' dGp for 100 pg of DSA. This level is the same as previously detected by SFS (Xlancnester et (11.. 1988; Weston et al.. 1989~)D. NA from immortalized human liver cells treated with [3H]B[a]P was enzymatically digested and purified by HPLC as described. Levels of adduction identified by specific activity was 60 and 10 per IO- JGp. respectively. 3'P-postlabeling analysis detected adduci Ie\els of 19 and 8 per lo7dGp, respectively. Discussion The 32P-postlabeling assay has been widely used for the detection of adducts in single exposure and complex biological systems (Randerath et al., 1981. 1986: Phillips et al., 1988a,b; Schutte et al., 1988; Savela rt 01.. 1989). The applicability Of the method for screening human tissues and determining total adduct levels has been well established. although adduct identity is not possible and labeling efficiency is assumed to be 1 0 % . The current work modifies the "P-postlabeling assay wherein an unmodified nucleotide is concomitantly labeled to act as an internal stndard and a TLC separation is utilized that identifies modified and unmodified nucleotides on the same TLC plate. The assay can then be calibrated by directly determining molar ratios. The mcthod assesses labeling efficiency, confirms a complete kinase reaction, rules out the presence of unknown enzyrne inhibitors and detects adequacy of digestion (wh combined with HPLC). It further demonstrates that the devel ment of a chemically specific "P-postlabelingassay needs to based upon the analysis of synthesized standards. determinati dibration cunes for known Lersus determined molar ratios of micropreparative techniques that isolate adducts. These susart,that adduct levels are probably higher than previously (corroborated in this study by SFS, UV detection and ra&&emical specific activir! 1. Decreased labeling efficiencies for PAH-adducts have been p~viouslyreported. Traditionally. calibration curves for adduct labling has been generall! limited to BPDE-trcatcd DNA. Both Rddy and Randerath ( 1987) and Gupta et N/. (1982) rcported of 40-805 for f 'H]BPDE-treated DNA (Gupta et 1982). which is consistent with our findings. Moreover, &e determination that o\erall recovery is based upon level of adduction has been noted for BPDE-DNA adducts and fluomnthene-DNA adducts (Gorelickand Wogan, 1989; Jahnke et al., 1990). Thus. detected adduct levels depend on both preparative methods and labeling efficiencies (indicating up to 1o&ld lower reco\ e? at some adduct levels for some adducts). A variety of experiments \\ere carried out to determine factors that might alter labeling efficiencies. Conditions of limiting ATP (defined here as less ATP than nucleotides on a molar basis). in Contrast to previous repons of DNA -adduct preferential labeling (Randerath et al.. 1985). did not affect the determined ratios of adduct to unmodified nucletoides in the range tested. It should also be noted that n hile ATP limiting conditions may enhance detection. o\ erall le\ els of recovery. when compared with procedures such as butanol extraction, are < 10%(Gupta and Earley, 1988). It IS also of interest that for other adducts. such as the 80H-deoxlguanosine-3'-monophosphate.ATP excess increases labeling eficienc! (Poley et al., 1989)and a minimum of excess ATP is needed for fluoranthene-DNA adducts (Gorelick and Wogan. 1989).For alkyl adducts, ATP levels did not appreciably alter labeling efficiency (Shields et al., 1990). Several modifications to the "P-postlabeling assay have led to a decreased sensiti\it! for this assay. This is likely due to labeling conditions and the addition of the internal standard with its differentiallabeling efficient! . Use of different polynucleotide enzymes from other manufacturers. buffers. ATP concentrations. incubation times and less amounts of internal standard did not improve sensitivity. Separatelj . a minor shortcoming of the -method reported here is that dGp as an internal standard decreases the sensitivity of the assay b) 1.5-fold at levels of adducts typically observed in humans. However, the use of calibration Curve accounts for this. Gorelick and Wogan (1989) also noted a decreased yield of adduct in the presence of unmodified nucleotides. Importantly. this effect has implications for 32P-Postlabeling assays that might concomitantly label unmodified nucleotides as a matter of course (kinase/excess ATP and kinase/limiting ATP) and may account for the increased sensitivity of enrichment procedures over that which is expected merely by the increased amount of DNA used (Reddy and b d e r a t h , 1987; Gorelick and Wogan, 1989). n e data reported here indicate that mixtures of these adducts do not affect the individual labeling efficiencies of each other suggest that adduct classes may be labeled together, e.g. after the separation by HPLC or immunoaffinity chromatography. Giventhe demonstration that unmodified nucleotides affect the labelingefficiencyof the PAH-DNA adducts, it is plausible that of other classes may affect the labeling efficiencies of the PAH-DNA adducts, especially those with smaller side chains.For example, it has been reported that the imidazole-ring O Y n d form of N7-methyl-2'-deoxyguanosine is nuclease-P1 reslsbnt (Hemminki, 1989).If this adduct is present in labeling mixturcs. existing cndogenousiy o r occ~urririg;i1-11t'iL,i:ill> . 11 1iuy affect the quantitation of PAH -DNA a i k l u c ~ s . In summary, these experiments modit). tlw '-'P-postlabcling assay to usc an internal standard for conlir-mat ion of labeling efficiency and quantitate PAH -DNA adducts using direct molar ratios with calibration curves for adducts of interest. The data suggest that the labeling efficiency of PAH -DNA adducts is not 100%at levels observed in human samples by other assays and that the presence of unmodified nucleotides further affects the labeling efficiency of adducts. Given differential labeling efficiencies for PAH-adducts, micropreparative techniques need to be developed so that specific PAH-DNA adducts undergo detection by the 3'P-postlabeling assay. The data also demonstrate that the use of micropreparative techniques is required for the quantitative recovery of individual adducts. For this assay, such techniques are required to purify adducts from starting unmodified nucleotides. The detection level of this assay, when combined with HPLC, is one adduct in lo7 unmodified dGp. This level is within the range of observed values reported for other assays such as SFS and immunoassays (Shields et al., 1991). The use of chemically specific assays will be useful for correlations of adduct formation and the ability to metabolically activate carcinogens and develop mutations. These types of correlations will ultimately enhance cancer risk assessment as they relate to genetically susceptible individuals. Acknowledgements Our thanks to Bob Julia and Dorothea Dudek far their skillful editorial assistance, and to Julie Riiska and Van Doan for their technical assistance. References A m i n S , Misra.B., Desai.D., Hu1e.K. and Hecht.S.S. (1989) Chromatographic conditions for separation of "P-labeled phosphates of major polynuclear aromatic hydrocarbon -deoxyribonucleoside adducts. Carcinogenesis, 10, 1971 - 1974. Foi1es.P.G.. Mig1ietta.L.M..Quan.A.M., Quart,E.,Kabat,G.C.and Hecht,S.S. (1989) Evaluation of 32P-postlabelinganalysis of DNA from exfoliated oral mucosa cells as a means of monitoringexposure of t!!e oral cavity to genotoxic agents. Carcinogenesis, 10, 1429- 1434. Gorelick,N.J.and Wogan,G.N. (1989) Ruomthene-DNA edducts: identifcation and quantification by an HPLC-"P-postlabelig method. Carcinogenesis, 10, 1567-1577. Gupta,R.C. (1985) Enhanced sensitivity of 32P-postlaklinganalysis of aromatic carcinogen-DNA adducts. Cancer Res., 45, 5656-5662. Gupta,R.C. and Earley,K. (1988) 32P-adductassay: comparative recoveries of structurally diverse DNA adducts in the vzrious enhancement procedures. Carcinogenesis, 9, 1687- 1693. Gupta,R.C. and Randerath,K. (1988) Ana!ysis of DNA adducts by 32Plabeling and thin layer chromatography. In Fr;edberg,E.C. and Hanawalt,P.C. (eds), DNA Repair, A Laboraroy Man& ofResearch Prccedures. Marcel Dekker, New York, pp. 399-417. Gupta,R.C., R4dy.M.V. and Randerath,K. (1982) 32P-postlabelinganalysis of non-radioactive aromatic carcinogen-DNA adducts. Carcinogenesis, 3, 1081 - 1092. Harris,C.C., Weston,A.. Willey,J.C., Trivers,G.E. and Mann,D.L. (1987) Biochemical and molecular epidemiology of human cancer: indicators of carcinogen exposure., DNA damage, and genetic predisposition. Environ. Health Perspecr., 75, 109-119. Hawke,L.J. and Farrel1,G.C. (1986) Increased binding of benzo(a1pyrene metabolites to lymphocytes from patients with lung cancer. Cancer Len.,30, 289-297. Hemrninki,K. (1989) Ring-opened 7-methylguanine nucleotides are resistant to nuclease PI digestion and good substrates to polynucleotide kinase. Carcinogenesis, 10, 1761 - 1763. Jahnke,G.D., Thompson,C.L., Walker,M.P., Gal1agher.J.E.. Lucier,G.W. and DiAugustine,R.P. (1990) Multiple DNA adducts in lymphocytes of smokers and nonsmokers determined by 32P-postlaklinganalysis. Carcinogenesis, 11, 205-21 I . 125 I i !, i i * I 1 i i t i ! I -P2G.ShieTds et al. Karcher,W., Fordham,R.J., Dubois.J.J., Glaude,?.G.J.M. and LighIhart,J.A.M. (1983) Spectral Atlar of Polycyclic Aromutic `hnpounds. Reidel, Boston. Kawajiri,K., Nakachi,K., Imai,K..YoshU,A., Shi.mda,N. and WatanabeJ. (1990) Identification of genetically high risk individuals to lung cancer by DNA polymorphism of the cytochrome P450IA 1 gene. FEES Lett., 263, 13I - 133. Kouri,R.E., McKinney,C.E., S1omiany.D.J., Snodgrass,D.R., Wray,N.P. and McLemore,T.L. (1982) Positive correlation between high aryl hydrocarbon hydroxylase activity and primary lung cancer as analyzed in cryopreserved lymphocytes. Cancer Res., 42, 5030-5037. Manchester,D.K., Weston,A., Choi,J.S., Trivers,G.E., Fennessey,P., Quintana,E., Farmer,P.B., Mann,D.L. and Harris,C:C. (1988) Detection of benzo[a]pyrene diol epoxide-DNA adducts in human placenta. Proc. Natl. Acad. Sci. USA, 85, 9243-9247. Manchester,D.K., Wilson,V.L., Hsu,I.C., Choi,J.S., Parker.N.B., Mann,D.L., Weston,A. and Harris,C.C. (1990) Synchronous fluorescence spectroscopic, immunoaffinity chromatographic, and 32P-postlaheling analysis of human placental DNA known to contain benzo[a[pyrene diol epoxide adducts. Carcinogenesis, 11, 553-559. McLemore,T.L., Adelberg,S., Liu,M.C., McMahon,N.A., Yu,S.J., Hubbard,W.C., Czznvinski,M., Woo\l,T.G., Storeng,R., Lubet,R.A., Eggleston,J.C., Boyd,M.R. and Hines,R.N. (1990) Expression of CYPlAl gene in patients with lung cancer: evidence for cigarette smoke-induced gene expression in normal lung tissue and for altered gene regulation in primary pulmonary carcinomas. J. Nctl. Cancer Inst., 82, 1333- 1339. Pfeiffer,A.M., Cole,K.E., SmrJt,D.J., Weston,A., Gmopman,J.A., Shields,P.G., Vignaud,J.M., Juillerat,M., Lipsky,M.M., Trump,B.F., Lechner,J.F., Hanis,C.C. (1993) SUM T-antigen immortalizednormal human liver epithetial cells express hepatocyte characteristics and metabolize chemical carcinogens. Proc. Natl. Acad. Sci. USA, in press. Phillips,D.H., Hemminki,K., Alhonen,A., Hewer,A. and Gr0ver.P.L. (1988a) Monitoring occupational exposure to carcinogens: detection by 32P-postlabelling of aromatic DNA adducts in white blood cells from iron foundry workers. Mutat. Res., 204, 531 -541. Phillips,D.H., Hewer,A.. Martin,C.N., Gmer,R.C. and King,M.M. (1988b) Correlation of DNA adduct levels in human lung with cigarette- smoking. N m e , 336, 790-792. Povey,A.C., Wilson,V.L., Taffe,B.G., Wood,M.L., Essigmann,J.M. and Harris,C.C. (1989) Detection and quantitation of 8-hydroxydeoxyguanosine residues by 32P-postlabeling. Proc. Am. Ass. Cancer Res., 30, 201 (Abstract). Randerath,K., Reddy,M.V. and Gupta,R.C. (1981) 32P-postlabeling test for DNA damage. Proc. Natl. A c d . Sci. USA. 78, 6126-6129. Randerath,E., Agrawal,H.P., Weaver,J.A., Borde1on.C.B. and Randerath.K. (1985) 32P-Postlabelinganalysis of DNA adducts persisting for up to 42 weeks in the skin, epidermis and dennis of mice treated topically with 7,12-dimethylbenz[a]anthracene. Carcinogenesis, 6 . 1117-1126. Randerath,E., Avitts,T.A., Reddy,M.V., Miller,R.H., Evers0n.R.B. and Randerath,K. (1986) Comparative 32P-analysis of cigarette smoke-induced DNA damage in human tissues and mouse skin. Cuncer Res., 46, 5869-5877. Randerath,K., Randerath,E., Danna,T. F., van Go1en.L. and Punnan.K.L. (1989) A new sensitive 32P-postlabeling assay based on the specific enzymatic conversion of bulky DNA lesions to radiolabeled dinucleotides and nucleoside 5'-monophosphates. Carcinogenesis, 10, 1231-1239. Reddy,M.V. and Randerath,K. (1986) Nuclease P1-mediated enhancement of sensitivity of 32P-postlabeling test for structurally diverse DNA adducts. Curcinogenesis, 7, 1543-1551. Reddy,M.V. and Randerath,K. (1987) 3'P-postlabeling assay for its carcinogen -DNA adducts: nuclease P1-mediated enhancement of its sensitivity and applications. Environ. Health Perspect.. 76. 41 -47. Savela,K., Leppala,S. and Hemminki,K. (1989) 32P-postlabeling of DNA adducts in white blood cells of foundry workers. Arch. To.ricol. SuppL. 13. 101- 103. Schoket,B., Hewer.A.. Gr0ver.P.L. and Phil1ips.D.H. (1988) Covalent binding of components of coal-tar, creosote and bitumen to the DNA of the skin and hags of mice following topical application. Carcinogenesis. 9, 1253- 1258. Schutte.H.H., Van der Schans.G.P. and Lohman,P.H. (1988) Comparison of induction and repair of adducts and of alkali-labile sites in human lynihocytes and granulocytes after exposure to ethylating agents. Mutat. Res., 1W.23-37. Shields.P.G., Povey,A.C., Wilson.V.L., Weston.A. and Harris.C.C. ( 1990) Combined high performance liquid chromato_rraphy/-7'P-postlabelin::assay of N7-methyldcoxy~uanosincC. uncer Rcs.. 50. 6580-6584. Shields.P.G., Weston.A.. Suginutm.H.. Bowman.E.D., Caporaso.N.E.. Manchestcr.D.K.. Trivcrs.C.E., Tan1ai.S.. Rcsau.J.H.. Trun1p.B.F. and Harris.C.C. (1991) MoIccuIx cpidcrniology: tlosinittry, susceptibility and cancer risk. In Vmdcrlxui.M.. Stankcr.L.H.. W;itkins.B.E. and Rokrts.D.W. (eds). Itiiiiii~!1v~ i.\.\ily,/vi IV TV(I<CY/icmicul. h i l w i . s . Monitotin,y fi~.riiC. ltemitrils i n ff~ii~iuii.~.t;oo(/, ~ i i i J r / i ~ , Eni,;rmtn~wr. Anicrican Chemical Society. \V:isliin~tm,I ) ( ` , lip I S(1 lh. 126 Shie1ds.P.G. and Harris,C.C. (1991) Molecular epidemiology and the genetics of environmental cancer. J. Am. Med. Ass., 266, 681-687. West0n.A.. Manchester,D.K., P0irier.M .C., Ch0i.J.S.. TriversGE., M ~ , D . L . and Hams,C.C. (1989a) Derivative fluorescencespectral analysis of p01yq,-&~ aromatic hydrocarbon-DNA adducts in human placenta. Chon. Res. Taic01., 2, 104-108. Weston,A., Manchester,D.K., P0vey.A.C. and Harris,C.C. (1989b) Dewon of carcinogen-macromolecular adducts in humans J. Am. COLI.Toxicol., 8, 913-932. Weston,A., Rowe,M.L., Manchester.D.K., Farmer,P.B., Mann,D.L. and Hanis,C.C. (1989~)Fluorescence and mass spectral evidence for the f o m m of benzo[a]pyrene anti-diol-epoxide-DNA and -hemoglobin addurn i,, humans. Carcinogenesis, 10, 251 -257. Weston,A., Willey,J.C., Modali,R., Su-dura,H., McDowel1,E.M.. Resau,J., Light,B., Haugen,A., Mann,D.L., Trump,B.F. and Hanis,C.C. (198%) Differential DNA sequence deletions from chromosomes 3, 11, 13 and 17 i,, squamous cell carcinoma, large cell carcinoma and adenocarcinoma of fie human lung. Proc. Natl. Acad. Sci. CU,86, 5099-5103. Wilson,V.L., Smith,R.A., Autrup,H.. Krokan,H., Musci,D.E., Le,N.N., Longoria,J., Ziska,D. and Harris,C.C. (1986) Genomic 5-methylcytosine bmination by 32P-postlabeling analysis. ..lnal. Biochem., 152, 275-284. Wilson,V.L., Basu,A.K., Essigmann.J.SI.. Smith,R.A. and Hanis,C.C. (1988) @-alkyldeoxyguanosine detection by 32P-postlabeling and nucleotide chromatographic analysis. Cancer Res.. 48,2156-2161. Received on July 3, 1992; accepted on Ocrober 1. 1992