Division of Biochemistry and Molecular Biology, Barker Hall, University of California, Berkeley, CA 94720 ...... H. J., Jacob, R. A. & Ames,-B. N. (1991) Proc. Natl.
Proc. Nail. Acad. Sci. USA Vol. 89, pp. 3375-3379, April 1992 Medical Sciences
Assay of excised oxidative DNA lesions: Isolation of 8-oxoguanine and its nucleoside derivatives from biological fluids with a monoclonal antibody column (cancer/mutation/endogenous DNA adducts/oxygen radicals)
EUN-MI PARK*, MARK K. SHIGENAGA, PAOLO DEGANt, TOMMY S. KORN, JEFFREY W. KITZLERt, CAROL M. WEHR, PREMA KOLACHANA, AND BRUCE N. AMES§ Division of Biochemistry and Molecular Biology, Barker Hall, University of California, Berkeley, CA 94720
Contributed by Bruce N. Ames, December 30, 1991
An immunoinity column is described that ABSTRACT facilitates the analysis of oxidative damage products of DNA and RNA in urine, blood plasma, and medium isolated from cultures of Escherichia coli. In intact animals, lesions (adducts) excised from DNA are transported from the cell through the circulation and excreted in urine. In bacteria, DNA adducts are excreted directly into the medium. In either case, the adducts can be assayed as a measure of oxidative damage to DNA. A monoclonal antibody that recognizes 8-oxo-7,8-dihydro-2'deoxyguanosine(oxo dG; 8-hydroxy-2'-deoxyguanosine), abiomarker of oxidative damage to DNA, has been isolated, and its substrate binding properties have been characterized. The relative binding affinities of this monoclonal antibody for oxo8dG, unmodified nucleosides, or derivatives of Gua made it suitable for the preparation of immunoaffinity columns that greatly facilitate the isolation of oxo8dG, 8-oxo-7,8dihydroguanine, and 8-oxo-7,8-dihydroguanosine from various biological fluids. Quantitative analysis of these adducts in urine of rats fed a nucleic acid-free diet and in the medium from cultures of E. coli suggests that oxo8-7,8-dihydroguanine is the principal repair product from oxo8dG in DNA of both eukaryotes and prokaryotes. The results support our previous estimate of about 105 oxidative lesions to DNA being formed and excised in an average rat cell per day.
guanosine) and 8-oxo-7,8-dihydroguanine (oxo8Gua) by HPLC with electrochemical detection (EC) (11, 12). In particular, oxo8dG serves as an excellent marker for DNA damage produced by oxidants because it represents one of the major products generated by a wide array of treatments associated with oxidant damage such as that produced by irradiation and various carcinogens (13-16), because it is implicated in spontaneous transversion mutagenesis (17, 18), and because it can be measured with very high sensitivity by HPLC-EC (19). This report describes the preparation and characterization of a monoclonal antibody (mAb)-based immunoaffinity column that is used to isolate oxo8dG, oxo8Gua, and 8-oxo-7,8dihydroguanosine (oxo8G) rapidly from complex biological fluids prior to their analysis by high sensitivity HPLC-EC (19). The described mAb improves greatly upon the solidphase extraction (SPE) and polyclonal antibody-based immunoaffinity methods (11, 12) and offers, in combination with HPLC-EC, an analytical approach with the specificity, sensitivity, selectivity, and relative simplicity required to measure oxidative DNA and RNA damage noninvasively and routinely.
Oxidants are generated continuously as normal by-products of mitochondrial electron transport, inflammatory reactions, and many other processes associated with aerobic metabolism (1). Although elaborate enzymatic and nonenzymatic defenses exist that protect the cell against oxidant-induced damage (2-7), the oxidants that escape the defenses damage cellular macromolecules. Oxidant damage escaping repair may be a major cause of the age-dependent decline in normal cellular function. Unrepaired oxidative damage to DNA, for example, has been proposed to be a major contributor to aging and other age-related degenerative diseases including cancer (2, 8, 9). When DNA is damaged, nonspecific DNA repair enzymes excise DNA lesions to release deoxynucleotides, and basespecific repair glycosylases excise the corresponding base. Deoxynucleotides are enzymatically hydrolyzed to stable deoxynucleosides, and these repair products are transported through the blood and excreted in the urine. Damage to RNA is reflected in nucleoside adducts. We have previously described methods for noninvasively quantitating various DNA adducts produced by oxidative damage, including those that measure urinary thymidine glycol and thymine glycol by HPLC with UV detection (8, 10) and urinary 8-oxo-7,8dihydro-2'-deoxyguanosine (oxo8dG; 8-hydroxy-2'-deoxy-
Chemicals. Alkaline phosphatase-goat anti-mouse IgG conjugate, Sigma 104 phosphatase substrate tablets, and CNBr-activated Sepharose 4B were obtained from Sigma. [3H]oxo8dG, [14C]oxo8Gua, and [14C]oxo8G were prepared as described (11, 12). Animal and Human Samples. Three- to 6-month-old male, outbred albino Fischer 344 rats (Simonsen Laboratories, Gilroy, CA) were used. Animals were fed either a standard diet (Purina rat chow) or nucleic acid-free diet (Dyet no. 510007; Dyets, Bethlehem, PA) ad libitum and maintained in a temperature- and photoperiod (12 h/day)-controlled room. For urine collections, the rats were housed individually in metabolic cages (Nalge), and 24-h urine outputs were collected and stored at -20°C until analyzed. Human urine samples (24-h output) were obtained from 20 healthy individuals (26-76 years of age) on an unrestricted diet.
MATERIAL AND METHODS
Abbreviations: mAb, monoclonal antibody; EC, electrochemical detection; oxo8dG, 8-oxo-7,8-dihydro-2'-deoxyguanosine (8-
hydroxy-2'-deoxyguanosine); oxo8G, 8-oxo-7,8-dihydroguanosine; oxo8Gua, 8-oxo-7,8-dihydroguanine; BSA, bovine serum albumin; SPE, solid-phase extraction; oxo8dA, 8-oxo-7,8-dihydro-2'-
deoxyadenosine; RT, retention time. *Present address: Department of Chemistry, Incheon University, 177, Dohwadong Namgu, Incheon, Republic of Korea. tPresent address: IST Istituto Scientifico Tumori, via le Benedetto XV, 10 16132 Genoa, Italy. tPresent address: National Institute of Environmental Health Sciences, P.O. Box 12233, Research Triangle Park, NC 27709. §To whom reprint requests should be addressed.
The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact.
3375
3376
Medical Sciences: Park et al.
Preparation of Protein-Nucleoside Conjugates. Bovine serum albumin (BSA) and casein conjugates of periodatetreated oxo8G were prepared as described (12). Production and Characterization of anti-oxo8dG mAb. Two Swiss-Webster, two Biozzi, and two BALB/c mice (6-8 weeks of age) were injected subcutaneously at two dorsal sites with 50 A.l each of a suspension of the oxo8G-BSA conjugate (at 0.5 Ag/IAl) prepared in monophosphoryl lipid A plus trihalose dimycolate adjuvant (Ribi Immunochem). A second set of mice was injected with the oxo8G-casein conjugate. The mice received two booster doses of 50 .g in 100 1.d, 7 and 21 days after the initial immunization. Sera were isolated on day 27 by tail vein bleeding and tested by competition ELISA for binding specificity and affinity. A Swiss-Webster mouse (BSA conjugate) and a Biozzi mouse (casein conjugate) were selected for hyperimmunization. Four days before the hybridomas were prepared, the mice were treated to prevent a potentially lethal hypersensitivity response by subcutaneous injection of a mixture of 1 ,Ag of Robinul-V (glycopyrrolate; A. H. Robbins, Richmond, VA) and 100 ,ug of A-H injection (doxylamine succinate; Coopers Animal Health, Kansas City, KS) in a total volume of 50 1.l (A. E. Karu, personal communication). This treatment was followed -30 min later by injection of the mice, via the tail vein, with the corresponding conjugate (50 ,ug in 50 A.l of sterile saline). Four days later, the mice were sacrificed; splenocytes were isolated, and hybridomas were prepared with myeloma cell line P3X63 AG8.653 by electrofusion (A. E. Karu, personal communication). Culture media from aminopterin-resistant hybridoma clones were assayed by ELISA (20) for recognition of the oxo8G hapten by screening against the alternate conjugate. Clones exhibiting favorable binding affinities and binding specificities for oxo8dG and oxo8Gua were recloned at limiting dilution. Ascites fluid was prepared from clone 15A3 by inoculating 1-2 x 106 hybridoma cells intraperitoneally into immunosuppressed, pristane-primed Swiss-Webster mice. Ascites fluid was pooled and stored at -80°C. Competition ELISA. Plates were coated with 2.5 ng of the oxo8G-casein conjugate per well and incubated overnight at 4°C. The inhibitors oxo8dG, oxo8Gua, dG, and Gua were each mixed at amounts ranging from 0.025 to 1000 pmol with a 1:100,000 dilution of 15A3 ascites fluid in a total volume of 100 ,ul and incubated overnight at 4°C. These mixtures were applied to the wells on the following day, and the competition ELISA was performed as described (21). The mAb isotype was determined by ELISA using a commercial kit (Pierce). The specificity of the mAb was characterized by comparing binding affinity constants (Kff) (12) that had been determined by competitive RIA using an ascites dilution of 1:1000 (22, 23). Preparation of Immunoaffinity Columns. The immunoaffinity matrix was prepared with unpurified ascites fluid (12) by covalently linking proteins in 5 ml of ascites fluid to 5 g of CNBr-activated Sepharose 4B. Immunoaffinity columns (1to 2-ml bed volume) were prepared by packing 1.0 x 10 cm Econo-Columns (Bio-Rad) with a mixture of antibodyconjugated Sepharose 4B (0.2 parts) and unmodified Sepharose 4B (0.8 parts). Columns that contained a 2-ml bed volume bound -2 nmol of the oxo8dG standard. Sample Preparation and Isolation from Urine, Plasma, and Media. Urine samples (1-5 ml) were diluted with an equal volume of 1 M NaCl, spiked with the appropriate radiolabeled tracer(s) ([3H]oxo8dG, [14C]oxo8Gua, or [14C]oxo8G), and applied to a preconditioned C18/OH SPE column (11). The SPE column was then washed with 5 ml of 50 mM KH2PO4 buffer (pH 7.5), and retained compounds were eluted with 3 ml of 15% MeOH in the same buffer. The eluate was applied to the immunoaffinity column at 40C [this temperature favors nonequilibrium binding and reduces cross-
Proc. Natl. Acad Sci. USA 89 (1992)
reactivity (24)]. The column was washed in sequential order with 5 ml of each of the following: H20, 1 M NaCl, H20, and acetonitrile. Less than 10%o of the applied radioactivity, corresponding to the radiolabeled internal standards for oxo8dG, oxo8Gua, and oxo8G, was lost during this series of washes. Immediately after the acetonitrile wash, the antibody binding compounds were eluted with MeOH (5 ml) into a polypropylene culture tube. The resulting MeOH eluate was concentrated to dryness under a stream of nitrogen in a 40-450C water bath for 1-2 h. The sample was resuspended in 200 pu of water, and a 20- to 50-p, sample was analyzed by HPLC-EC as described below. Recoveries of oxo8dG, oxo8Gua, and oxo8G from the immunoaffinity columns were typically 60-90%o. The urinary excretion rate values are expressed as pmol of oxo8dG excreted in a 24-h urine void per kg of body weight (pmol kg-1 day-1). The column could be used for at least 10 cycles without apparent loss of binding capacity. The column was regenerated immediately after elution by applying MeOH (10 ml). The bed volume was reequilibrated and resuspended twice in water (10 ml) with a Pasteur pipet to remove air bubbles. Human blood plasma (5-10 ml) was spiked with the appropriate radiolabeled tracer(s), and the proteins were precipitated by adding an equal volume of acetonitrile. The precipitated proteins were separated by centrifugation at 3000 x g for 15 min, and the supernatant was transferred to a new culture tube and mixed with 8 volumes of water (40 ml). The resulting sample was applied directly to an immunoaffinity column and processed according to the protocol described above. Alternatively, after centrifugation, the sample could be stored at -200C for 2 h to allow separation of the organic and aqueous phases. After removal of the organic phase with a Pasteur pipet, the remaining aqueous layer was transferred directly to the immunoaffinity column and processed according to the protocol described above. Wild type Escherichia coli K-12 was cultured in VogelBonner citrate medium with 0.4% glucose. The 100-ml cultures were inoculated with 3 ml of an overnight culture and incubated at 370C for 8 h (ANO = 1.0). Cells were removed from the medium by centrifugation at 3000 x g at 40C, and the medium was filtered through a 0.22-t&m membrane filter unit (Nalgene, Rochester, NY). The medium (100 ml) was applied directly to the antibody column (2-ml bed volume, diluted 1:2) and processed as above. HPLC-EC Analysis. The HPLC conditions used in the present study have been described (12). Compounds were analyzed by flow-through EC employing an ESA model 5100 Coulochem detector equipped with a 5011 or 5010 highsensitivity analytical cell with the oxidation potentials of electrodes 1 and 2 adjusted to 0.10 and 0.35 V, respectively.
RESULTS Characterization ofmAb 15A3. Screening of the hybridomas after immunization of mice with oxo8G-BSA and oxo8Gcasein conjugates resulted in selection of 80 stable clones from which 15A3, a fusion cell line produced from spleen cells of an oxo8G-BSA conjugate-immunized Swiss-Webster mouse, was selected for further characterization. The isotype of mAb produced by hybridoma 15A3 was IgG2 K. Binding specificity of hybridoma 15A3 was examined by competition ELISA and shown to be highest for oxo8dG followed by oxo8Gua, dG, and Gua: 50%6 inhibition of oxo8Gcasein conjugate-antibody complex formation at 0.4, 40, 2000, and 2000 pmol, respectively, was observed. However, since stock solutions of commercial dG and Gua contain measurable amounts of oxo8dG and oxo8Gua, respectively (1 part per 20,000-100,000 unmodified residues), the calculated binding specificity of the antibody for dG and Gua is a considerable overestimate.
Medical Sciences: Park et al. The mAb exhibited Ka values (see Materials and Methods) for oxo8dG, oxo8Gua, and oxo8G of 9.2 x 108 M-', 4.7 x 107 M-1, and 2.1 x 108 M-1, respectively, as determined by competitive RIA. Table 1 lists Kaff values and relative affinities for the various compounds and compares them to the polyclonal antibody reported previously (12). The mAb exhibits an '-400-fold higher affinity for oxo8Gua than does the polyclonal antibody (12). This permits the isolation of oxo8Gua from samples without enzymatically digesting uric acid, a major urinary excretion product present in millimolar concentrations that was found previously to interfere with the binding of oxo8Gua to the polyclonal antibody-based immunoaffinity column (12). A comparison ofthe affinities listed in Table 1 suggests that modifications of Gua at the C-8, N-7, C-6, and C-2 positions influence antibody binding and indicate that the epitope recognized by the antibody is comprised of the portion of the purine ring (the C-2, N-1, C-6, C-5, N-7, C-8 backbone; see Fig. 1 Inset) distal to the deoxyribose residue. HPLC-EC of mAb Immunffinity-Purified Urine. Fig. 1 shows a representative chromatogram of human urine processed by the immunoaffinity column and analyzed by HPLC-EC. In addition to oxo8dG (Fig. 1, peak 6), which elutes at 20.9 min, other 6,8-dioxo purines, which include uric acid [peak 1, retention time (RT) = 3.2 min], oxo8Gua (peak 2, RT = 5.1 min), 7-methyl-8-oxoguanine (peak 3, RT = 12.2), N2-methyl-8-oxoguanine (peak 4, RT = 13.2; H. Helbock, J. Thompson, and B.N.A., unpublished results), and oxo8G (peak 5, RT = 14.5 min), were also observed. Similar chromatographic profiles were observed for rat urine and human blood plasma. Validation of the Urinary oxo8dG Assay. The urinary oxo8dG values obtained by HPLC-EC after sample processing using the mAb-based immunoaffinity columns, compared sample for sample to that obtained by using either SPE [mAb vs. SPE (n = 8); range = 74-1021 pmol-kg-lday-1] or anti-oxo8dG polyclonal antibody-based immunoaffinity columns (mAb vs. polyclonal (n = 4); range = 82-1043 pmol kg-1 day-1), were essentially identical (r2 = 0.99) for both sets of comparisons. To determine the reproducibility of the assay, urine samples obtained from 20 healthy individuals
Proc. Natl. Acad. Sci. USA 89 (1992)
3377
Co 2
5
Co
~~~~~~~~~~~6
0
5
10 15 Time, min
20
25
FIG. 1. HPLC-EC chromatogram of human urine processed by C18/OH SPE and mAb-based immunoaffinity columns. Human urine was processed according to the procedures described in Materials and Methods and analyzed by HPLC-EC as described (12). The sample, which was equivalent to 0.42 ml of urine, contained 3.98 pmol of oxo8dG (9.48 nM). Since the SPE step was optimized for oxo8dG, recoveries of earlier eluting compounds, which were not calculated, were in most cases significantly lower. Identities of the peaks are as follows: 1, uric acid; 2, oxo8Gua; 3, 7-methyl-oxo8Gua; 4, N2-methyl-oxo8Gua; 5, oxo8G; and 6, oxo8dG.
processed and assayed for urinary oxo8dG. The procedure was repeated with the same samples but on a separate day, and the urinary oxo8dG values were compared (r2 = 0.91) as shown in Fig. 2A. To examine the variation in excretion rates of oxo8dG of samples analyzed on different days, urine samples from the same group of individuals that had been collected 130 days apart were processed and assayed. The correlation coefficient for this comparison (Fig. were
Table 1. Binding affinities (Kff), determined by competitive RIA, of the anti-oxo8dG mAb 15A3 compared with the anti-oxo8dG polyclonal antibodies described previously (12) mAb 15A3 Polyclonal Ab Inhibitor Binding Kff(oxo8dG)/ Kaff(oxo8dG)/ inhibition* (I) Kaf, M-1 Kaff(I) Kaff(I) x x 108 9.2 oxo8dG 5.0 10-13 1 1 oxo8G 2.0 x 10-12 2.1 x 108 4.4 5.3 4.7 x 107 oxo8Gua 8.5 x 10-12 19.6 7.6 x 103 sh8G 2.0 x 10-11 2.0 x 107 46 86 5.7 x 105 Uric acid 7.0 x 10-10 1.6 x 103 Br8G 1.5 x 10-9 2.6 x 105 >3.5 x 103 dGt 3.0 x 10-9 1.3 x 105 7.1 x 103 7.4 x 104 Gua 1.4 x 105 2.9 x 10-9 6.6 x 103 >3.5 x 105 3.3 x 10-9 1.2 x 105 7.7 x 103 oxo8dI dT 2.5 x 104 1.6 x 10-8 3.7x 104 5.3 x 105 m7G 8.0 x 10-8 5.0 x l03 1.8 x 105 1.7 x 104 >5.0 x 10-8 1.2x 105 >>1.0 x 10-6 2.3 x 106 8.6 x 104 dAt dCt >>1.0 x 10-6 2.1 x 107 2.3 x 106 sh8G, 8-mercaptoguanosine; Br8G, 8-bromoguanosine; m7G, 7-methylguanosine; oxo8dI, 8-oxo-7,8-
dihydro-2'-deoxyinosine; oxo8dA, 8-oxo-7,8-dihydro-2'-deoxyadenosine (oxo8dA); Ab, antibody; I,
inhibitor. *Moles of I that inhibit binding by 50%o. tThe commercial dG used in the competitive RIA contained 1 oxo8dG residue per 137,440 unmodified dG residues. tThe percent inhibitions at the indicated concentrations of oxo8dA, dA, and dC were 45%, 37%, and 10%1, respectively.
3378
Medical Sciences: Park 300
et
Proc. Natl. Acad Sci. USA 89 (1992)
al.
.B
A y
=
3.4046 + 0.95994x R, 0.908
0
CO) 200
y = 47.583 + 0.68406x
30D
R2= O.530
CZ
.a
en
0
20
0 . .
0~~~~~~~~~~~
c
C: loo
U1)
10
0
"0
100
200
Analysis A
TO
300
A ._I pmol oxowdG/kg/day
100
200
300
Collection Day 1
FIG. 2. Reproducibility of the immunoaffinity-based urinary oxo8dG assay (A) and day-to-day variability of oxo8dG urinary excretion rates in humans (B). Urine samples from 20 healthy individuals were processed and assayed forurinary oxo8dG as describe4 in Materials and Methods. (A) Comparison of the oxo8dG urinary excretion rates of the same sample processed and analyzed on separate days (analysis A vs. analysis B). (B) Comparison of the oxo8dG urinary excretion rates of urine samples collected 130 days apart (collection day 1 vs. collection day 130). All axis values are expressed as pmol of oxo8dG kg-l day-1.
2B) was r2 = 0.53, indicating some consistency in the excretion rates of oxo8dG of samples collected 4 months apart.
oxoedG, oxo8Gua, and oxo8G in Urine, Plasma, and Medium. The average oxo8dG excretion rate in humans (172 79 pmol-kg-.day-1) was comparable to the rate obtained from separate groups of urine samples that had been processed by SPE' alone or by SPE in combination with the polyclonal antibody-based immunoaffinity column (11, 12). The urinary excretion rate of oxo8G, the corresponding 125 oxidative damage product of RNA, was 333 pmol kg. Iday-l. Analysis of urine obtained from rats reveals an oxo8dG excretion rate of =400 pmol kg-l day-1. This rate is not influenced by diet since the average oxo8dG vales obtained from nucleic acid-free diet-fed rats were not si hificantly different' from those obtained from chow-fed rats (Table 2). In contrast, the urinary excretion rate of oxo8Gua and oxo8G are affected greatly by the diet. Nucleic acid-free diet-fed rats excrete oxo8Gua and oxo8G at rates that are roughly 90%o and'65% lower than that of chow-fed rats. Urinary excretion rates of oxo8Gua and oxo8G in rats were noted to reach these lower levels within 2-3 days of switching from normal chow to a nucleic acid-free diet. The average concentrations of oxo8dG and oxo8G in human plasma were 69 and 127 pM, respectively. Similar values were observed in serum samples. For comparison, the average urinary concentrations of oxo8dG and oxo8G were 10 and 25 nM, respectively; thus, the kidney concentrates blood plasma oxo8dG and oxo8G by factors of =150 and 200, respectively. The concentrations of oxo8dG, oxo8Gua, and oxo8G in spent medium isolated from cultures of E. coli (A6N = 1.0) were =11, 204, and 94 pM, respectively. The higher concen-
tration of oxo8Gua in the medium suggests that the reported oxo8Gua glycosylase is the principal repair enzyme for oxo8dG lesions in DNA.
DISCUSSION The mAb isolated from the hybridoma cell line 15A3 binds with high specificity and affinity to oxo8dG, oxo8G1a, and oxo8G present in biological fluids with Kaffvalues of 9.2 x 108 M-1, 4.7 x' 17 M-1, and 2.1 x 108 M-1, respectively. Immunoaffinity' columns rapidly and efficiently isolate these compounds from urine, blood plasma, and medium recovered from bacterial cultures. The binding properties of the anti-oxo8dG mAb are distinct from that of the anti-oxo8dG polyclonal antibody described by us in a previous report (12). The mAb exhibits an affinity for oxo8dG comparable to that ofthe polyclonal antibody, yet it exhibits a much higher affinity 'for oxo8Gua. That is, in contrast to the anti-oxo8dG polyclonal antibody, substitution of deoxyribose or ribose for the hydrogen atom at the N-9 position of oxo8Gua has almost no influence, as judged by the comparable Kff values observed in the competitive RIA, on the binding affinity of the mAb for oxo8dG and oxo8G versus that for oxo8Gua. Characterization of this antibody by competitive RIA indicates that lack of an oxygen at C-8 (i.e., dG and Gua), substitution of C-6 with a primary amino group (oxo8dA), substitution of the primary amino group at C-2 with an oxygen atom (i.e., uric acid, the deamination product of oxo8Gua), or absence of the primary amino group at C-2 (8-oxo-7,8-dihydro-2'-deoxyinosine) all lead to a significant decrease in antibody binding affinity (Table 1). We conclude from this data that the epitope that the antibody recognizes
Table 2. Excretion rates or concentrations of oxo8dG, oxo8Gua, and oxo8G in urine, plasma, and medium Excretion rate or concentration oxo8Gua/oxo8dG oxo8G* (n) oxoSGua* (n) oxo~dG* (n) Sample 333 125 (53) 172 79 (63) ND Human urine 2810 830 (40) 370 + 63 (30) 50,545 24,832 (18) Rat urine (normal chow) 9.6 908 357 (27) 414 227 (27) 1,307 (25) 3,975 Rat urine (nucleic acid-free diet) 14 (4) 127 ND 69 15 (6) Human blood plasma 18.5 11 + 4 (4) 94 63 (4) 49 (2) 204 E. coli spent media Ratios of oxo8Gua/oxo8dG were calculated for those conditions where diet was controlled for nucleic acids. ND, not determined. *For human and rat urine, the excretion rates of oxo8dG, oxo8Gua, and oxo8G are given in pmol kg-l day-1. For human blood plasma and E. coli spent medium, the concentrations of oxo8dG, oxo8Gua, and oxo8G are given in fmol/ml. ±
Medical Sciences: Park
et
al.
is comprised of the surface of the purine base (C-2, C-6, N-7, and C-8) distal to the N-9 position. The mAb 15A3, in contrast to the anti-oxo8dG polyclonal antibody reported previously (12), is not suitable for quantitating oxo dG itt enzymatic digests of DNA or for immunohistochemistry due to its higher cross-reactivity to the normal deoxynucleoside dG. The high cross-reactivity of the antibody for other 6,8-dioxo purines (Table 1 and Fig. 1 Inset) limits the use of ELISAs to quantitate the levels ofeach ofthe described biomarkers in biological fluids. 'This crossreactivity is illustrated in an experiment that directly compares the quantitation of oxo8dG by ELISA and HfPLC-EC in spent medium recovered from cultures of E. coli (data not shown). The amount of oxo8dG measured by ELISA overestimates the values obtained by HPLC-EC by -300-fold, indicating that it is not feasible to quantitate oxo8dG or its derivatives by ELISA without further purification. Estimates of oxo8dG and oxo8Gua in urine from rats fed a nucleic acid-free diet or in spent medium isolated from cultures of E. coli grown in minimal media (Table 2) suggest that oxo8Gua is the major excision product of oxo8dG DNA repair in both eukaryotes and prokaryotes. As the urinary oxo8dG and oxo8Gua detected by our assay appear to be derived from DNA repair, we can estimate that at least 90o of the oxo8dG lesions present in DNA of both eukaryotes and prokaryotes are excised by a glycosylase activity such as that reported to exist in E. coli (25-28). The availability of immunoaffinity columns that permit the quantitation of oxo8dG and oxo8Gua from bacterial medium should greatly facilitate efforts to understand the role of this oxidative DNA damage product and the importance of the described prokaryotic enzyme in mutagenesis. These DNA adducts also have been measured in spent medium from mammalian cell cultures (Q. Chen and B.N.A., unpublished results). This avoids the problem of dietary absorption for certain biomarkers and should complement the studies performed with bacterial medium by providing insights on the importance of oxidative DNA damage and its repair in mutagenesis, cell transformation, and cellular senescence in a eukaryotic model system. The described immunoaffinity technique will allow one to examine the effects of various conditions that induce oxidative DNA damage and the dietary and endogenous factors that defend the cell or the organism from these insults. Such a tool can be used to facilitate our understanding of the many genetic, physiological, and pathological processes that influence, through oxidative mechanisms, cancer and the rate of aging as well as to investigate inhibition of oxidative DNA damage by dietary antioxidants (29). The hybridoma cell lines described in this paper were developed at the University of California Berkeley College of Natural Resources Hybridoma Facility, which was supported in part by National Science Foundation Grant DIR-90-40161: we are indebted to A. E. Karu for his generous advice and help and M. Bigelow for technical assistance. This work was supported by National Institute Outstanding Investigator Grant CA 39910 to B.N.A. E.-M.P. was supported by the University of California Toxic Substances Re-
Proc. Natl. Acad. Sci. USA 89 (1992)
3379
search and Teaching Program. M.K.S. was supported by a National Institute of Aging Postdoctoral Fellowship (AG 05489). T.S.K. was supported by a University of California President's Undergraduate Fellowship (696KOR). J.W.K. was supported by a National Institutes of Health Postdoctoral Fellowship (ES 05428). 1. Halliwell, B. & Gutteridge, J. M. C. (1985) Free Radicals in Biology and Medicine (Clarendon, Oxford). 2. Ames, B. N. (1989) Free Radic. Res. Commun. 7, 121-128. 3. Frei, B., England, L. & Ames, B. N. (1989) Proc. Natl. Acad. Sci. USA 86, 6377-6381. 4. Stocker, R., Yamamoto, Y., McDonagh, A. F., Glazer, A. N.
& Ames, B. N. (1987) Science 235, 1043-1046. 5. Frei, B., Kim, M. C. & Ames, B. N. (1990) Proc. Natl. Acad.
Sci. USA 87, 4879-4883. 6. Frei, B., Stocker, R. & Ames, B. N. (1988) Proc. Natl. Acad.
Sci. USA 85, 9748-9752. 7. Frei, B. & Ames, B. N. (1992) in The MolecularBiology ofFree Radical Scavenging Systems, ed. Scandalios, J. (Cold Spring Harbor Lab., Cold Spring Harbor, NY), in press. 8. Adelman, R., Saul, R. L. & Ames, B. N. (1988) Proc. Natl. Acad. Sci. USA 85, 2706-2708. 9. Fraga, C. G., Shigenaga, M. K., Park, J. W., Degan, P. & Ames, B. N. (1990) Proc. Natl. Acad. Sci. USA 87,4533-4537. 10. Cathcart, R., Schwiers, E., Saul, R. L. & Ames, B. N. (1984) Proc. Natl. Acad. Sci. USA 81, 5633-5637. 11. Shigenaga, M. K., Gimeno, C. J. & Ames, B. N. (1989) Proc. Natl. Acad. Sci. USA 86, 9697-9701. 12. Degan, P., Shigenaga, M. K., Park, E. M., Alperin, P. E. & Ames, B. N. (1991) Carcinogenesis 12, 865-871. 13. Dizdaroglu, M. (1985) Biochemistry 24? 4476-4481. 14. Kasai, H., Nishimura, S., Kurokawa, Y. & Hayashi, Y. (1987) Carcinogenesis 8, 1959-1961. 15. Fiala, E. S., Conaway, C. C. & Mathis, J. E. (1989) Cancer Res. 49, 5518-5522. 16. Floyd, R. A. (1990) Carcinogenesis 11, 1447-1450. 17. Wood, M. L., Dizdaroglu, M., Gajewski, E. & Essigmann, J. M. (1990) Biochemistry 29, 7024-7032. 18. Shibutani, S., Takeshita, M. & Grollman, A. P. (1991) Nature (London) 349, 431-433. 19. Floyd, R. A., Watson, J. J., Wong, P. K., Altmiller, D. H. & Richard, R. C. (1986) Free Radic. Res. Commun. 1, 163-172. 20. Voller, A., Bidwell, D. & Bartlett, A. (1976) in Manuals of Clinical Immunology, eds. Rose, N. & Friedman, H. (Am. Soc. Microbiol., Washington), pp. 506-512. 21. Liddell, J. E. & Cryer, A. (1991) A Practical Guide to Monoclonal Antibodies (Wiley, Chichester, U.K.). 22. Muller, R. & Rajewsky, M. F. (1980) Cancer Res. 40, 887-896. 23. Muller, R. (1980) J. Immunol. Methods 34, 345-352. 24. Prevost, V., Shuker, D. E., Bartsch, H., Pastorelli, R., Stillwell, W. G., Trudel, L. J. & Tannenbaum, S. R. (1990) Carcinogenesis 11, 1747-1751. 25. Boiteux, S., O'Connor, T. R., Lederer, F., Gouyette, A. & Laval, J. (1990) J. Biol. Chem. .265, 3916-3922. 26. Chung, M. H., Kasai, H., Jones, D. S., Inoue, H., Ishikawa, H., Ohtsuka, E. & Nishimura, S. (1991) Mutat. Res. 254, 1-12. 27. Tchou, J., Kasai, H., Shibutani, S., Chung, M.-H., Laval, J., Grollman, A. P. & Nishimura, S. (1991) Proc. Natl. Acad. Sci. USA 88, 4690-4694. 28. Michaels, M. L., Pham, L., Cruz, C. & Miller, J. H. (1991) Nucleic Acids Res. 19, 3629-3632. 29. Fraga, C. G., Motchnik, P. A., Shigenaga, M. K., Helbock, H. J., Jacob, R. A. & Ames, -B. N. (1991) Proc. Natl. Acad. Sci. USA 88, 11003-11006.
