ANALYTICAL

BIOCHEMISTRY

Method Mette

205,

337-341

(19%)

for Estimation

Tingleff

Skaanild

of Benzo[a]pyrene

and Jmgen

Institute

for Life Sciences and Chemistry,

Received

December

Clausen*l Roskilde

University

Center, Roskilde

DK-4000,

Denmark

27, 1991

Polycyclic aromatic hydrocarbons, e.g., benzo[alpyrene (B(a)P) are known carcinogens/mutagens. These compounds may be metabolized by the P450 mixed function monooxygenase to more nucleophilic compounds which may form adducts to the cellular macromolecules, e.g., DNA, RNA, and proteins. We have used synchronous fluorescence scanning for the assay of DNA adduct formation. In our earlier work with in vitro exposed human lymphocytes we estimated the adduct formation (femtomoles B(a)P per microgram DNA) to be higher than that estimated by other workers. We suggested that this difference may be related to the DNA isolation method used. In order to elucidate these differences we compared DNA adduct formation in human lymphocytes where DNA was isolated by the two different methods, i.e., using phenol extraction or the Gene Clean method. The data demonstrate that the phenol extraction procedure gives a yield of adducts per microgram DNA lower than that obtained by the Gene Clean method. The principle of the Gene Clean method for DNA isolation is protein denaturation by means of NaI followed by catching of DNA by absorption on silica particles. In contrast, the phenol extraction method is based upon phenol-mediated denaturation of proteins in the cell lysate leaving the hydrophilic nucleotides in the aqueous phase. However, during adduct formation more lipophilic adducts derived from DNA may redistribute between the aqueous phase and the phenol phase. In support of this theory we found higher adduct concentration per microgram DNA by the Gene Clean method 40 to 60 times than that found by the phenol method. Furthermore, assaying the distribution of adducts in both phases separated by the phenol method revealed an adduct concentration per microgram DNA in the phenol phase 34 times higher than that in the aqueous phase. Finally, it was shown by means of alkaline elution that the Gene Clean-isolated DNA contained single-stranded (adducted) DNA

‘To whom Life Sciences 260, DK-4000

DNA Adducts

correspondence should and Chemistry, Roskilde Roskilde, Denmark.

0003.269’7192 $5.00 Copyright 0 1992 by Academic Press, All rights of reproduction in any form

be addressed University

at Institute Center, P.O.

for Box

whereas the aqueous phase from the phenol method did not. The adducted ssDNA showed an inhomogeneous distribution of specific adduct level, i.e., picomoles B(a)P per microgram ssDNA. 8 1992 Academic PESS, IUC.

Polycyclic aromatic hydrocarbons are a class of chemicals that contain many known carcinogens, e.g., benzo[a]pyrene (B(a)P).’ B(a)P is metabolized by the P450 mixed function monooxygenase whereby more nucleophilic compounds, e.g., epoxides and hydroxyepoxides, are formed (1). These metabolites may bind to macromolecules, i.e., DNA, RNA, and proteins, and this binding, or adduct formation, may be carcinogenic or mutagenic, if the adducts are not removed (2). The DNA adducts are repaired mostly by excision repair (3). DNA adduct formation has been analyzed using different methods such as HPLC and mass spectroscopy (4,5), the 32P-postlabeling technique (6), or the enzymelinked immunosorbent assay (7) and the synchronous fluorescence scanning assay. DNA adduct formation was quantitatively estimated by Autrup et al. (9) and Burke et al. (5) using a HPLC procedure. Furthermore, Vahakangas et al. (8) and Skaanild and Clausen (10) used synchronous fluorescence scanning. The principle of this method was presented by Lloyd in 1975 (11) and further developed by Vo-Dihn (12), who presented the following equation for the syncronous intensity, I&l, l) = k x c x d x I&(1’) x I&(1’ + dl), where I, in the syncronous spectrum is proportional with: 1. the concentration of the compound (c), 2. the intensity of the emission spectrum (E,) at emission wavelength 1’+ dl (dl is the wavelength differ-

’ Abbreviations sential medium;

used: B(a)P, benzo[a]pyrene; SDS, sodium dodecyl sulfate.

MEM,

minimal

es-

337 Inc. reserved.

338

SKAANILD

ence between peaks in the exitation trum), and 3. the intensity of the exitation wavelength 1.

and emission spectrum

AND

spec(E,)

at

Syncronous scanning makes it possible to distinguish different aromatic components because each component has at least one peak at a specific wavelength corresponding to a specific dl value. The results of our study gave yields of adduct formation comparatively higher than those obtained by other studies. The aim of the present communication was therefore to find the reason for these differences in levels of adducts. The DNA analyzed for adducts was in all previous studies isolated using phenol extraction. We used the rather new Gene Clean kit (Bio. 101, Inc., La Jolla, CA). This kit was used because it was an easy and fast method for preparation of DNA without proteins and RNA. MATERIAL

AND

METHODS

Chemicals All chemicals were of highest obtainable purity and were purchased from Sigma Chemicals Co. (St. Louis, MO) or Merck (Damstadt, Germany). Eagle’s minimal essential medium (MEM) was from GJBCO (Uxbridge, Middelsex, UK), NADP (as the Na salt) from Boehringer-Manheim, Ficoll-Paque from Pharmacia (Copenhagen, Denmark), and the Gene Clean kit from Bio. 101, Inc. The radioactive labeled B(a)P was obtained from Amersham (UK). Assay of in Vitro Adduct Formation of Calf Thymus DNA with [3H]B(a)P Calf thymus DNA (50 pg) suspended in 50 ~1 TE buffer (cf. below) was added to 2 ml of MEM, microsomes were added to a final concentration of 8 mg/ml, NADP+ and glucose 6-phosphate were added to final concentrations of 0.4 mM and 1 mM, respectively, and finally a 200-~1 suspension of [3H]B(a)P (sp act of 70 Ci/mmol or 272 mCi/mg and 5 mCi/ml) was added. This suspension was then incubated overnight in the dark at 37°C. Isolation and Incubation of Human Lymphocytes Lymphocytes were isolated from blood taken from a normal individual not suffering from any acute infections and not taking any drugs 14 days prior to sampling. The lymphocytes were isolated using Ficoll-Paque and vital stained with trypan blue to ensure survival of the cells prior to incubation (viability, 99%). They were grown in MEM containing 0.3 mg/ml L-glutamine, 100 IU/ml penicillin, and 100 pg/ml streptomycin and stimulated with 5 rg phytohemagglutinin per 1 ml. The

CLAUSEN

cells were resuspended in this medium to a cell density of about 3 X lo6 cells/ml. One milliliter of this suspension was then seeded per well in a Tech-Nunc 24-well culture dish, after which B(a)P in 300 ~1 autologous serum was added to a final concentration of 1 mM, and the cells were incubated for about 30 h. Eight wells were incubated and the DNA was isolated using the Gene Clean procedure on four of the cultures and phenol extraction on the remaining four cultures.

Isolation of DNA DNA was isolated from human lymphocytes and from incubation mixtures by using the Gene Clean kit or phenol extraction. Cells were sucked up in an Eppendorph tube and isolated by centrifugation for 4 min at 45Og, resuspended in 300 ~1 lysis buffer (100 mM NaCl, 50 mM Tris, pH 8.0, l%, w/v, SDS, 10 mM EDTA), and frozen at -20°C overnight or longer. After thawing, proteins were digested with proteinase K (0.4 pg/ml sp act lo-20 U/mg) for 1 h at 45°C. These first steps of DNA isolation were identical for the two procedures. DNA isolated by the Gene Clean kit method. The procedure of the manufacturer was followed: The digested sample (300 ~1) was mixed with 2 vol of 6 N NaI and incubated on ice (15 min). Next, 10 ~1 of the glass milk solution was added and mixed carefully every minute for 10 min. The sample was then centrifuged and the glass milk was washed three times with 1 ml of the wash buffer (1 vol ethanol + 1 vol 10 mM Tris buffer). The DNA was now resuspended in 0.5 or 1 ml TE buffer (10 mM Tris and 1 mM EDTA, pH 8.0) by heating for 3 min at 50°C. The suspension was then centrifuged and the supernatant containing the DNA was processed for further analysis. The DNA concentration could then be measured using diaminobenzoic acid (13). DNA isolated by the phenol method. To the digested sample (300 ~1) equal volumes of phenol and chloroform:isoamyl alcohol (24:l) were added. The sample was mixed well and centrifuged. After centrifugation the water phase was collected and the DNA precipitated by addition of 2 vol of ethanol. The DNA was then sedimented at 4°C after which the DNA was resuspended in 400 ~1 TE buffer and incubated for 1 h at 37°C. After incubation 25 ~1 of 3.5 M NaCl was added and DNA was again precipitated by addition of 850 ~1 cold ethanol. The solution was frozen for 30 min, after which the DNA was sedimented at 4°C. The DNA was then resuspended in TE buffer. Isolation of Microsomes Microsomes were isolated from liver cells according to Benford and Hubbert (14).

ENZYMATIC

ASSAY

ESTIMATING

FIG. 1. The peak height in the synchronous scan plotted versus the concentration of B(a)P adducts in DNA samples isolated by means of the Gene Clean method from an incubation mixture of 50 ~1 calf thymus DNA + microsomes + a NADPH-generating system + 1 mCi [3H]B(a)P (for detaiis cf. the text). Abscissa, pmol [3H]B(a)P per ml incubation mixture; ordinate, peak height of syncronous signal in ar* - - *, slit 10.10, bitrary units. n * * . W, slit 10.10, prior to acidification; after acidification; m- --¤, slit 5.5, prior to acidification; n - n , slit 5.5, after acidification.

Alkaline Et&ion Analysis The DNA isolated was alkaline eluted according to Kohn et al. (15) briefly as follows: The DNA isolated was disposited onto a 25mm-diameter, 1.2-pm-poresize polyvinyl chloride filter (Millipore) washed with 3 ml of 0.02 M Na,EDTA, pH 10, and eluted in the dark with a solution consisting of 0.04 M H,EDTA plus tetrapropylammonium hydroxide added in the amount required to give the desired pH of 12.1. The flow rate was about 0.07 ml/min, and fractions of 2 ml were collected in LKB fraction collector Model Ultra rat No. 7000 (LKB, Stockholm, Sweden). The fractions were further analyzed for DNA (13) and for adducts by synchronous scanning. Fluorescence Measurement All measurements were made at 22°C using a PerkinElmer LS-50 fluorescence spectrophotometer. A l-ml quartz cuvette with a light path of 0.5 cm was used for all the measurements, both the conventional and the synchronous fluorimetry. Data were standardized by the use of an aqueous quinine solution. Both emission and excitation slits were set at 5 or 10 mm for synchronous fluorimetry. According to Skaanild and Clausen (10) the B(a)P signal was found in the syncronous spectrum at about 387 nm using a delta lambda of 20 nm.

BENZO[a]PYRENE

was done as previously

Assay of DNA The assay was the diaminobenzoic

described acid method

(10). (13).

RESULTS

Initially DNA from the incubation mixture with calf thymus DNA, microsomes, and [3H]B(a)P was isolated

339

ADDUCTS

after 16 h incubation using the Gene Clean procedure. After isolation DNA was determined and the amount of B(a)P adducted was estimated by scintillation counting. During the 16-h incubation period 0.64 pmol B(a)P/pg DNA was found adducted. In order to assay the “fluorescence coefficient” relating height of syncronous signal to picomoles B(a)P under different experimental conditions B(a)P-adducted DNA was diluted from 0 to 17 pg DNA/ml TE buffer (pH). These samples were then synchronously scanned at emission and excitation slit of 5 and 10 mm. Finally, l-ml samples were hydrolyzed for 3 h with 100 ~10.1 N HCl at 9O”C, because Vahakangas et al. (8) have stated that activation will increase the fluorescence of the sample. Figure 1 shows data obtained at slit widths of 10 and 5 mm, respectively. The results reveal a higher syncronous fluorescence signal of the samples prior to hydrolysis than after. It can also be seen that there is a linear relationship between the concentration of B(a)P and the peak height in the scan. The fluorescence coefficients were estimated (Table 1). The numbers reveal that the fluorescence signals in the samples which are not hydrolyzed are about 13 times higher than those in the hydrolyzed samples. Comparison Adducted

of the Two Methods DNA

for Isolation

of

Half of the incubation mixture of calf thymus DNA + [3H]B(a)P + microsomes + the NADPH-generating system was used to isolate DNA by means of the Gene Clean system. The other half of the incubation mixture was used for DNA isolation by the phenol method. Gene Clean-isolated DNA contained adducts at a level of 125 fmol/pg DNA whereas the phenol-extracted DNA contained 2.31 fmol/wg DNA, which is about 60 times less. In order to explain the differences between Gene Clean-isolated DNA and DNA isolated by phenol extraction the Gene Clean-isolated DNA from the second

TABLE

1

“Fluorescence Coefficients,” i.e., the Coefficients Relating the Peak Height of the Syncronous Signal to picomoles B(a)P Adducted to DNA

E,

LO counting The counting

DNA

slit 10 5 10 5

Note. Data acid hydrolysis E,, exitation; peak.

E, slit 10 5 10 5 are calculated from (for experimental IU, internal Units,

Hydrolysis + +

Fluorescence coefficients (fmol/IU) 19.2 94.3 234.2 1368.7

Figs. 1 and 2 both prior to and after details cf. the text). Em, emission; i.e., peak height of the syncronous

340

SKAANILD

AND

CLAUSEN

tions from phenol-extracted at all.

DNA

contained

no adducts

DISCUSSION

* ‘,C

FIG. ing tion and ml;

2. The elution during alkaline elution of 22 pg DNA contain353 pmol B(a)P adducts. (For details cf. the text.) Abscissa, fracnumber (2 ml); ordinate, pg DNA/ml, pmol B(a)P per fraction, specific DNA adducts level of pmol B(a)P/pg DNA. W, pg DNA/ *, pmoles B(a)P/pg DNA; 9]DO], pmoles B(a)P/ml.

incubation mixture was reisolated using phenol extractions and the amount of B(a)P adducted to DNA was determined both in the TE phase and in the phenol phase after elimination of phenol by dialysis against TE buffer. The highest concentration of adducts per microgram DNA was found in the phenol phase having 176.9 fmol/pg DNA whereas the TE phase contained 5.16 fmol/pg DNA. Eight cultures containing 3 X lo6 freshly isolated human lymphocytes were then incubated with B(a)P for about 30 h. The DNA from four of the cultures was isolated with the Gene Clean kit and the DNA from the remaining four cultures by phenol extraction. After DNA isolation the DNA and the B(a)P were estimated. Next the adduct concentration was determined by synchronous scanning without hydrolysis, giving an adduct concentration of 16.1 pmol/pg DNA when isolated with Gene Clean and 0.34 pmol/pg DNA when isolated using phenol extraction. Then 22 fig DNA isolated by the Gene Clean method, containing 353 pmol B(a)P adducts, was separated in ssDNA and dsDNA by alkaline elution. A similar separation in ss- and dsDNA was made on phenol-extracted DNA. The eluted fractions (2 ml) (ssDNA) were then assayed for DNA and synchronous scanned in order to determine if they contained any DNA adducts. Figure 2 shows the elution profile of the Gene Clean-isolated DNA. Ail soluble DNA was eluted within the first seven fractions. The first two fractions contained the highest amounts of DNA. Fifty eight percent of the total amount of DNA applied to the filter was eluted. In contrast the B(a)P adducts peaked at fractions 3 and 4. Of the total 353 pmol B(a)P adducts applied to the filter, 27% were eluted with ssDNA. “The B( a)P adducted specific adduct level,” i.e., picomoles per microgram DNA, showed a maximum level in fraction 4. Thus the eluted DNA showed a heterogenous population of DNA adducts. The aqueous frac-

The classical method for isolation of DNA by means of phenol (16) makes use of phenol as a protein-denaturating agent, leaving the hydrophilic polynucleotides in the aqueous phase. However, another principle is used in the Gene Clean method (17), where NaI is used as a denaturing agent and silica spheres are used to catch the DNA from the aqueous phase by adsorption. The aim of the present communication was to explain the differences in adduct levels found in our work (10) and the works of Vahakangas et al. (8) and Burke et al. (5). First DNA isolated from the incubation mixture of calf thymus + microsomes + [3H]B(a)P and the NADPH-generating system was syncronously scanned both prior to and after acid hydrolysis. Opposite to what we expected from the work of Vahakangas et al. (8) the syncronous fluorescence signal of the DNA sample decreased after hydrolysis. This difference may perhaps be related to the DNA isolation procedure used since iodide may increase the quantum yield of fluorescence (18). The syncronous fluorescence signal in B(a)P may thus in our system of unhydrolyzed DNA samples be enhanced by iodide whereas acid hydrolysis may cause evaporation of HI no such effects are present when using the phenol method. Since B(a)P is transformed by a two-step reaction catalyzed by the P450 complex to 7,8dihydroxy-9,10-epoxide derivative, which may be covalently bound to guanosine (19). This derivative has phenolic properties and may therefore react like phenol. Bridges (20) investigated the fluorescence of phenol and found that the undissociated form exhibited fluorescence whereas the ionized form was nonfluorescent. The data thus support our findings of a linear relationship between the peak height in the synchronous scan and the concentration of adducts (assayed by isotope tracing). This was also expected given the equation developed by Vo-Dihn (12). In colliparing the two methods for DNA isolation we found that the DNA isolated with the Gene Clean kit had an adduct concentration (estimated by scintillation counting) higher than that of DNA isolated using phenol extraction. The reason for this may be that the single-stranded DNA was extracted into the phenol phase due to the covalent binding of B(a)P to DNA, transforming DNA to a more lipophilic state. Our findings are in agreement with those of Morris and Shertzer (21), who found that single-stranded DNA could be removed by extraction with aqueous phenol. The formation of single-stranded fragments may be caused by DNA strand breaks induced by the adducts. In support

ENZYMATIC

ASSAY

ESTIMATING

of this theory Gene Clean-isolated DNA was reisolated using phenol extraction. Here it was obvious that the concentration of adduct DNA found in the phenol phase was much higher than the concentration found in the aqueous TE phase, indicating that ssDNA-containing adducts probably were extracted into the phenol phase. Finally in order to support that theory we compared the amounts of DNA adducts obtained by the two methods for DNA isolation from lymphocytes incubated with B(a)P. Again the concentration of DNA adducts per microgram DNA (estimated by synchronous scanning) in the Gene Clean-isolated DNA was 40 times higher than that in the phenol-extracted DNA, again indicating that the single-stranded DNA was extracted into the phenol phase. Since Kohn et al. (15) demonstrated that the alkaline elution method measures the rate at which ssDNA is released to an alkaline solution whereby it is able to pass through a filter, we used this method to trace ss- and dsDNA in the DNA sample by isolating the two methods. The aqueous phase of phenol-extracted DNA contained no ssDNA adducts, whereas the DNA fractions adsorbed on silica by the Gene Clean method contained high concentrations of ssDNA adducts (as well as dsDNA). These data are in agreement with those of Bradley et al. (22), who showed that nitrosamines induced a dose-dependent increase in ssDNA in rat hepatocytes. At the highest exposure levels about 90% of the total DNA was alkaline eluted (22). ssDNA may be formed from dsDNA especially during cell replication. A long patch system may explain the high level of adducted ssDNA (3). Our data also show that the adducted ssDNA is a highly inhomogeneous population of molecules. The present results may thus indicate that the Gene Clean procedure for DNA isolation gives a higher yield of adducts per microgram DNA than does the phenol extraction procedure. The Gene Clean method is made at a final concentration of 4 M NaI, at which proteins leave DNA (23). Bio. 101, Inc., producing Gene Clean, claim that the silica particles (glass milk) efficiently extract DNA with more than 100 bp. However, DNA with lower numbers of base pairs are also extracted, but with somewhat lower yield (24). Further work may reveal to what extent the two methods for DNA isolation are supplementary.

BENZO[a]PYRENE

DNA

341

ADDUCTS

REFERENCES F. P., and Liebler, D. C. (1985) Hepatic and Extrahe1. Guengerich, patic Mixed-Function Oxidases, pp. 419-440, Handbook of Physiology, American Physiology Society, Washington, DC. 2. Bohr, V. A., Evans, M. K., and Fornace JR, A. J. (1989) Lab. Znuest. 61, 143-161. 3. Bernheim, N. J., and Falk, H. (1983) J. Am. Coil. Toxicol. 2, 2354. 4. Rahn, (1982)

R. O., Chang, S. S., Holland, J. M., Biochem. Biophys. Res. Commun. 25,

5. Burke, (1977)

M. S., Skipper, P. L., Jernstrom, J. Biol. Chem. 25,6424-6431.

6. Philips, Grover, I.

D. H., Hemminki, P. L. (1988) Mutat.

and Shugart, 262-268. B., and

K., Alhonen, A., Res. 24, 531-541.

L. R.

Orrenius,

Hewer,

S.

A.,

and

Harris, C. C., Vahakangas, K., Newman, M. J., Trivess, G. E., Shemsuddin, A., Sinopol, N., Mann, D. L., and Wright, W. (1985) Proc. Natl. Acad. Sci. USA 82, 6672-6676.

8. Vahakangas, K., Haugen, 6(8), 1109-1116. 9. Autrup, (1986) 10. Skaanild,

H., Arch.

A., and Curtis,

Grafstrom, R., Vahakangas, Tonicol. Suppl. 9, 147-153.

M. T., and Clausen,

J. (1991)

C. (1985) K.,

and

Harris,

C. C.

in preparation.

11. Lloyd, J. B. F. (1975) J. Forensic Sci. Sot. 11, 12. Vo-Dihn, T. (1978) Anal. Chem. 50, 396-401. 13. Kissane, J. M., and Robins, F. H. (1958) J. Biol. 188. 14. Benford, D. J., and Hubbert, ogy: A Practical Approach, gland.

Carcinogenesis

83-94. Chem.

233,184-

S. A. (1985) Biochemical Toxicolpp. 64-72, IRL Press, Oxford, En-

15. Kohn, K. W., Erichson, L. C., Ewig, R. A. G., and Friiedman, (1976) Biochemistry 15,4629-4637. 16. Mamur, J. (1961) J. Mol. Biol. 3, 208-218. 17. Vogelstein, B., and Gillespie, USA. 76(2), 615-619. 18. Hood, L. V. S., and Winefordner, 1927. 19. Weinstein, B. I. (1976) Science

D. (1979)

PFOC. Natl.

J. D. (1966)

Anal.

Acad. Chem.

C. A.

Sci. 1922-

193,592-595.

20. Bridges, J. W. (1964) Methods in Polyphenol, pp. 59-72, Pergamen, Oxford. 21. Morris, S. T., and Shertzer, H. G. (1985) Enuiron. Mutagen. 7, 871-880. 22. Bradley, M. O., Dysart, J., and Wolf, G. (1982)

G., Fitzsimmons, Cancer Res. 42,

K., Harbach, 2592-2597.

23. Hamaguchi, K., and Geiduschek, E. P. (1962) 84, 13-29. 24. Wilson, V. G. (1988) Bio Techniques 6, 733.

J. Am.

P., Leurni, Chem.

Sot.