molecular dosimetry of aflatoxin-n7-guanine in human urine ... · coefficient of 0.82 with p <...

Vol. 1, 221-227, March/April 1992 Cancer Epidemiology, Biomarkers & Prevention 221

Molecular Dosimetry of Aflatoxin-N7-guanine in Human UrineObtained in The Gambia, West Africa1

John D. Groopman,2 Andrew J. Hall, Hilton Whittle,Geoffrey J. Hudson, Gerald N. Wogan, RuggeroMontesano, and Christopher P. WildDepartment of Environmental Health Sciences, School of Hygiene and

Public Health, Johns Hopkins University, Baltimore, Maryland 21205[I- D. C.]; International Agency for Research on dancer, 150 coursAlbert-Thomas, 69372 Lyon dedex 08, France [A. I. H., R. M., C. P. W.];Medical Research Council, Dunn Nutrition Unit, Downham’s Lane,Milton Road, Cambridge dB4 lXl, England [H. W., C. I. H.]; and Massa-chusetts Institute of Technology, Division of Toxicology, Cambridge,Massachusetts02l39 [C. N. W.]

Abstract

Hepatocellular carcinoma is one of the major humancancers, causing at least 250,000 deaths each year.Two of the major risk factors for this disease areaflatoxin exposure and hepatitis B virus. This study wasundertaken to explore the relationship between dietaryexposure to aflatoxins and the excretion of the majoraflatoxin-DNA addud and other metabolites into theurine of chronically exposed people who were eitherhepatitis B virus surface antigen-positive or -negative.The diets of 20 individuals, 10 males and 10 females,with ages ranging from 15 to 56 years, were monitoredfor 1 week, and aflatoxin intake levels weredetermined for each day. Starting on the fourth day,total 24-h urines were consecutively obtained for 4days. The subjects were generally paired for hepatitis Bvirus status. Preparative monoclonal antibody affinitychromatography/high-performance liquidchromatography and competitive enzyme-linkedimmunosorbent assays were carried out on each of theurine samples, and the relationship between aflatoxinintake values and the excretion of (a) total aflatoxinmetabolites and (b) aflatoxin-N7-guanine (AFB-N7-guanine) was determined.

The average intake of total aflatoxins was 12.0 zgfor the entire study group during the 1-week collectionperiod. However, there was considerable day-to-dayvariation in exposures, from a low of zero to a high of29.6 �zg total aflatoxins/day. Initial efforts tocharacterize total aflatoxin metabolites in the urinesamples were made by competitive enzyme-linked

Received 8/26/91.1 This research was supported by USPHS Grant UO1-CA48409. J. D. C.is a recipient of a USPHS Research Career Development Award (K04-CA01517). The work was also supported in part by the Department ofCooperation and Development of the Ministry of Foreign Affairs of Italy.Preliminary data from this study were reported at the 1989 AnnualMeeting of the American Association for Cancer Research.2 To whom requests for reprints should be addressed, at Department of

Environmental Health Sciences, School of Hygiene and Public Health,Johns Hopkins University, 615 N. Wolfe Street, Baltimore, MD 21205.

immunosorbent assay. The correlation coefficient forthe analysis was 0.65, with P < 0.001. Thesecompetitive immunoassays determine a compositerepresentation of the aflatoxin compounds in urine,and a separation method using high-performanceliquid chromatography was required to resolve theindividual aflatoxin components in urine. High-performance liquid chromatography revealed thepreponderance of aflatoxin G1 in many of the urinesamples in addition to the oxidative metabolitesaflatoxin Pi, aflatoxin Q’, and AFB-N7-guanine.

One of the objectives of this study was todetermine the molecular dosimetry of AFB-N7-guaninein people who were either hepatitis B virus surfaceantigen-positive or -negative. Comparison of total AFB-N7-guanine excretion in the urine of all subjects overthe complete collection period plotted with totalaflatoxin B1 exposure in the diet yields a correlationcoefficient of 0.82 with P < 0.0001. Separation of thepopulation into hepatitis B virus carriers andnoncarriers revealed no difference in aflatoxin-DNAadduct levels in urine for a given dietary exposure.

Introduction

Molecular dosimetry has developed because of the dif-ficulties encountered by traditional epidemiologicalmethods in assessing the exposure status of an individualto a chemical carcinogen. The evolving molecular dosim-etry methods are going to be important for identifyingindividuals at high risk when multiple etiological agentsexist for a human cancer, such as hepatocellular carci-noma, where both hepatitis B virus and dietary aflatoxinintake are major risk factors. The association betweenAFB13 exposure, HBV status, and human liver cancer hasbeen established through epidemiological studies (1-4).Since primary liver cancer causes at least 250,000 deathsannually worldwide, preventive measures to limit bothexposure to and the adverse health effects of aflatoxinsand HBV are clearly warranted.

Aflatoxin B1 is converted to its toxic and carcinogenicforms by the cytochrome P-450 enzyme superfamily.These enzymes change AFBI to various oxidative deny-atives, including AFM1, AFQ1, and AFP1; a reduced afia-toxin species, aflatoxicol; and an unstable, highly reactive8,9-epoxide metabolite, the ultimate carcinogenic com-pound. These processes have been reviewed extensively(5). The reactive electrophilic epoxide can covalently

3 The abbreviations used are: AFB, aflatoxin B,; HBV, hepatitis B virus;HB,Ag, HBV surface antigen; HPLC, high-performance liquid chromatog-raphy; ELISA, enzyme-linked immunosorbent assay; DMSO, dimethylsulfoxide.

on January 12, 2021. © 1992 American Association for Cancer Research. cebp.aacrjournals.org Downloaded from

http://cebp.aacrjournals.org/

222 Aflatoxin-N7-guanine in Human Urine from The Gambia, West Africa

bind to DNA (6) and serum albumin (7), forming anaflatoxin-N7-guanine and lysine adduct, respectively.Both of these adducts have been used in animal andhuman molecular dosimetry investigations, and thesestudies were recently reviewed (8). The role of the majoraflatoxin-DNA adduct, AFB-N�-guanine, in the cancerinitiation process provides the justification for studyingthis agent as a molecular dosimeter in humans (9, 10).Other molecular dosimetry methods for measuring afla-toxins have been developed that use serum albuminadduct formation, and the aflatoxin-lysine adduct hasbeen found to be a good dose-dependent marker ofexposure in humans (1 1 , 1 2). The relatively small numberof studies on both human DNA adduct measurements inurine and serum albumin adducts suggests that bothmethods are going to be more reliable than dietarysurveys for assigning risk for aflatoxins. It is well knownthat dietary surveys are difficult to perform and have highstatistical coefficients of variation in analysis (13). Fur-thermore, in studies of people living in regions of Africaand Asia where the frequency of liver cancer is high,appropriate sample sizes for proper controlled food anal-yses are not easily attained. Thus, the use of DNA andprotein adduct measurements are mechanistically justi-fied and, in practical terms, will be more feasible inpopulation studies.

This paper describes the characterization of AFB-N7-guanine adduct excretion in human urine and its rela-tionship to the dietary intake of aflatoxins. The popula-tion studied resides in The Gambia, a country of highliver cancer incidence. We have been able to obtainurine samples from individuals paired for HB�Ag carrierstatus to investigate any differences in adduct formationbetween carrier and noncarnier states. These same mdi-viduals have been assessed for aflatoxin-albumin adductformation, which is reported elsewhere.4 The moleculardosimetry studies described here support the conceptthat measurement of the major, rapidly excised AFB-N7-guanine adduct in urine is an appropriate dosimeter forestimating exposure status and possibly risk in individualsconsuming this mycotoxin.

Materials and Methods

Human Sample Collection Design

Urine, blood, and food samples were obtained from 20subjects, 10 adult males and 10 adult females, living inKeneba, a village of approximately 1500 people, in theWest Kiang District of The Gambia. The collections weremade from October 10 to 17, which was at the end ofthe rainy season. One field worker was assigned to su-pervise each pair of study subjects to ensure compliancewith the protocol. The subjects were primarily selectedby HBV surface antigen status. Keneba residents whowere positive for serum HB�Ag during three earlier stud-es extending back to 1970 were selected and where

possible were matched for age, sex, and common dietaryexposures with noncarniers of HBV. Each subject was

4 C. P. Wild, C. I. Hudson, C. Sabbioni, B. Chapot, A. Hall, C. N. Wogan,H. Whittle, R. Monlesano, and J. D. Groopman. Dietary intake of aflatox-ins and the level o albumin bound aflatoxin in peripheral blood in The

Gambia, West Africa. dancer Epidemiol., Biomarkers & Prey., 1: 229-

234, 1992.

examined for HBV carrier status; however, over 5O% ofthe previously identified HBV chronic carriers were nolonger positive. Therefore, subjects shown to be HB�Ag-positive in 1980 and 1984 were selected to complete thestudy population. Five pairs of females were successfullyrecruited, and each pair ate from the same food bowl.Four pairs of males were also included in the study, butthey did not eat from the same bowl. The differencearises because females who are wives of the same manoften share the same food bowl, but adult males usuallyeat alone. One additional pair of males (brothers), eatingfrom the same food bowl, were recruited into the study,although both of these men were noncarniers of HBVsurface antigen. The study population is listed in Table1 . Informed consent was obtained from all subjects asapproved by the Medical Research Council/Cambiangovernment ethical committee.

Samples of cooked food were collected at each mealfrom the bowl that each subject was to share. The mealsalways consisted of sauces and a staple, such as boiledrice. Each subject normally ate 2 meals/day, at middayand in the evening. There were very few instances ofbreakfast meals among the monitored subjects. The maindietary staple for these people was rice, which wasconsumed during more than 95% of the meals. Thesauces consumed were kucha and naa (leaf sauces),bukolo (flour sauce), and tia durango, a roast groundnutsauce. The weight in each food bowl was recorded anddivided by the number of people, mathematicallyweighted for sex and age, who were eating from thebowl to provide an estimate of individual consumptionto be used in the calculation of daily aflatoxin exposures.It should be noted that the sample size of the saucetaken for analysis from the food bowl represented ap-

Table 1 Study po pulation, Ke neba, The Gambi a, West Africa

HB,Ag Age

Females

AB

L’L

+-

2921

CD

LL

+-

2829

EF

+

-

3338

CH L

+-

4630

II

+

-

5456

MalesKL

+-

4322

MN

-

-

5456

TP

+

-

2828

QR

+-

1616

0S

+

-

3915

a L, lactating.



Cancer Epidemiology, Biomarkers & Prevention 223

proximately 5% of the total meal of the person, andusually the dry weights of these samples were 5 g or less.Field workers were present at all food collections andsampled foods directly from the bowl of the individual.In some cases, when two subjects shared the same bowl,only one set of food samples was obtained. All foodsamples were lyophilized in the laboratory in Keneba forshipment. In total, 400 individual food samples werecollected over the 7-day monitoring period.

Twenty-four-h urine samples were obtained, and thepH of the urine was adjusted to about pH 5 with 1 N

HCI. Two separate 25-mI samples were applied to mdi-vidual C18 Sep-Pak cartridges, as described below(Waters Associates, Milford, MA). In addition, 50 ml ofurine from the samples were stored at -20#{176}C.Finally, 1-ml subsamples were taken from every urine sample forcreatinine determination. The range of volumes in the24-h urine samples was 200 to 2350 ml, and a total of180 urine samples were collected from days 4 through 7

of the study.Blood was drawn in most cases by venipuncture into

hepaninized tubes on days 1 and 8. In the case of oneindividual, only a fingerprick sample was obtained. Theanalysis of the plasma samples for aflatoxin-albumin ad-

duct levels is the subject of another publication.4

Analysis of Aflatoxin Contents in Foods

The aflatoxin levels in the food samples were quantifiedby preparative immunoaffinity clean-up of a food extract,and both HPLC and fluorescence quantification of theaflatoxins were performed (14). Briefly, 10 volumes of70% methanol:water (v/v) were added to each gram ofthe food sample. NaCI was added to 2% (w/v) finalconcentration, and the sample was blended at high speedfor 1 mm in a Waring blender. The extract was filteredthrough a paper filter. The filtrate was adjusted to 20%methanol:water by the addition of water, and an amountof filtrate equal to 1 g of the food sample was applied toa preparative monoclonal antibody column. The afiatox-ins were eluted from the column and submitted to HPLCor fluorescence analysis.

Analytical reverse-phase HPLC analysis was per-formed on all samples containing more than 20 ppb (�zg/kg). The detectors were a Beckman Model 160 fixedwavelength detector set at 365 nm, 0.001 absorbanceunits full scale, in series with a Hewlett-Packard Model1040A diode-array detector to quantify aflatoxin B1, B2,G1, and G2 in the food samples. The HPLC column wasa C18 5-jzm 25-cm Ultrasphere analytical column (RaininInst. Co., Woburn, MA), and chromatographic separationwas obtained by isocratic elution for 20 mm with 23%(v/v) ethanol. All mobile phases were buffered with 0.01M tniethylammonium formate, pH 3.0, and the columntemperature was maintained at 35#{176}C.The flow rate was1 mI/mm. Authentic aflatoxin standards were used todetermine chromatographic retention times.

Total aflatoxin levels in the food samples were alsodetermined by fluorescence. Eluates from the prepara-tive monoclonal antibody affinity columns were meas-ured by fluorescence with 360 nm excitation and 450nm emission spectra. Standard curves were generatedusing aflatoxin B1 and G. Aflatoxin intakes were calcu-lated using an average of 38% (w/w) solids for all saucesand an average portion size of 100 g cooked sauce (themean weighed sauce intake in the study was 100.3 g).

Human Urine AnalysisAfter collecting the human urine, samples of 25 ml eachwere loaded onto activated C18 Sep-Pak cartridges(Waters Associates). The urine was applied to the Sep-Pak cartridge at a flow rate of about 10 mI/mm, and theeffluent was discarded. The cartridge was washed with10 ml of 5% methanol:water, the ends of the cartridgewere sealed with Parafilm, and samples were stored at4#{176}Cprior to shipment for analysis.

AFB-N7-guanine in Urine Analysis. Before the elution ofthe aflatoxins from the Sep-Pak cartridges, the columnwas first rewetted with 5 ml of 5% methanol:water. Theaflatoxin metabolites were eluted from the column with10 ml of 80% methanol:water, and then the eluate wasrotary evaporated to dryness under reduced pressureand reconstituted by adding 0.3 ml of 0.1 N HCI withheating at 50#{176}Cfor 10 mm. This ensured that the rela-tively insoluble aflatoxin-DNA adducts were in solution.The sample was cooled to room temperature, and 0.5 mlof 1 M ammonium formate, pH 4.5, was added. Thevolume was adjusted to 10 ml with water, and the samplewas applied to a 4-mI preparative monoclonal antibodyaffinity column at a flow rate of about 0.3 mI/mm (11).The effluent from the loading of this sample was saved.The affinity column was washed with two 7-mI washesusing phosphate-buffered saline, pH 7.4, and one 7-mIwash with water to remove any nonspecifically boundmaterials. The aflatoxins were released from the prepar-ative monoclonal antibody affinity column by elutionwith 7 ml of 70% DMSO:water. The column was furtherwashed with two 7-mI phosphate-buffered saline washes,and the DMSO fraction and the last two phosphate-buffered saline washes were combined. The DM50 con-centration in these combined fractions was reduced toless than 3.5% by the addition of water. This sample wasthen applied to a C1� Sep-Pak cartridge, prepared asdescribed above, to remove the DMSO. The 80% meth-anol:water fraction from the Sep-Pak cartridge was re-duced to dryness under reduced pressure. The samplewas first redissolved in 100 jzl of 0.1 N HCI with heatingat 50#{176}Cfor 10 mm and then adjusted to about pH 3.0by adding an equal volume of 1 M ammonium formate,pH 4.5. At this point the sample was ready for injectiononto the HPLC column for analysis.

Analytical reverse-phase HPLC analysis used in se-nies a Beckman model 160 fixed-wavelength detector setat 365 nm, 0.001 absorbance units full scale, and aHewlett-Packard model 1040A diode-array detector toquantify aflatoxin metabolites in the human urine sam-pIes. The HPLC column was a C18 5-zm 25-cm Ultra-sphere analytical column (Rainin Inst. Co., Woburn, MA),and chromatographic separation was obtained by elutionfor 20 mm at 13% ethanol followed by a 13 to 25%ethanol linear gradient generated over 25 mm and thenisocratic elution at 25% ethanol. All mobile phases werebuffered with 0.01 M tniethylammonium formate, pH 3.0,and the column temperature was maintained at 35#{176}C.The flow rate was 1 mI/mm. Authentic aflatoxin standardswere used to determine chromatographic retentiontimes.

Aflatoxin Metabolites in Urine Analysis. Aflatoxins wereeluted from the Sep-Pak cartridges with 80% metha-nol:water, following a water wash of the cartridge, andwere purified on aflatoxin antibody affinity columns (Easi-



0z

w

z

zx0H-

-J

-J

H-0H-

I 0000

1 000

1 00

.

.

0 1 2 3 4 5 6 7 8

DAY OF COLLECTION

Fig. 1. Calculated daily dietary aflatoxin intake during the 7-day collec-tion period. Each day is depicted using a box and whisker plot analysis.The box and whisker plot represents a description of the data showingthe 10th and 90th percentiles of data (ends ofthe whiskers), the 25th and75th percentile (ends of the box), and a line for the 50th percentile (linewithin the box). Any outlier points were depicted as individual values.

uJzz

H-

LU

0

0

0

z

>(0I-

-JLL�

< 0.1

AFLATOXIN INTAKE (NC PER DAY)

Fig. 2. Linear regression analysis of the total aflatoxins in urine measured

by a polycbonal antiserum compared with average dietary aflatoxin intakeper day.

100 10,00 10000


extract; Oxoid, Baningstate, England) as described above.A competitive ELISA was used to measure the totalaflatoxin content extracted by the affinity column. Theantibody used in the ELISA was a polyclonal serum, andthe method has been described in detail (15, 16). Quan-tification was made against a standard curve for AFB1over the range 0.25 to 10 ng AFB1/ml/ELISA well. Eachsample or standard AFB1 inhibitor was assayed in qua-druplicate, and the raw data were transcribed into acomputer program that calculates and plots the standardcurve and corresponding value of test samples.

Statistical Analysis of Data: Box and Whisker Plots

Initial graphical data analysis was done by box and whis-ken plot analysis. The box and whisker plot represents adescription of the data showing the 10th and 90th per-centiles of data (ends of the whiskers), the 25th and 75thpercentiles (ends of the box), and a line for the 50thpercentile (line within the box). Any outlier points weredepicted as individual values. This type of data analysisreveals useful distribution and outlier characteristics ofthe data set, which can be lost or suppressed if theinformation is calculated using mean and SE analysis.Other data analyses were done by standard linear regres-sion methods.

Results

Aflatoxin Exposure in the Study Population. This studywas undertaken to explore the relationship betweendietary exposure to aflatoxins and excretion of the majoraflatoxin-DNA adduct, AFB-N7-guanine, and other me-tabolites into the urine of chronically exposed peoplewho were either HB5Ag-positive or -negative. The dietsof 20 individuals, 10 males and 10 females, with agesranging from 15 to 56 years (see Table 1 ), were monitoredfor 1 week, and aflatoxin intake levels were estimated foreach day. Starting on the fourth day, total 24-h urinecollections were obtained for 4 consecutive days. Thesubjects were generally paired for HBV status, and eachof the pairs of female subjects ate from the same foodbowl. Preparative monoclonal antibody affinity chroma-tography/HPLC and competitive ELISA analyses weredone on each of the urine samples, and the relationshipbetween aflatoxin intake and excretion of (a) total afla-toxin metabolites and (b) AFB-N7-guanine was de-termined.

Fig. 1 depicts the box and whisker plot analysis foraflatoxin intakes for all the study subjects. The averageintake of total aflatoxins was 12.0 jzg for the entire studygroup during the 1-week collection period. However,when the data were dissociated into male and femaleexposures, the average exposures for the period were8.2 and 15.7 �zg for the males and females, respectively.As seen in Fig. 1, there was considerable day-to-dayvariation in exposures from a low of zero to a high of29.6 �g total aflatoxins/day. Since the data in Fig. 1 arepresented on a logarithmic scale, zero values have beenomitted from the depicted analysis. It is notable thatnone of the subjects was an outlier on more than oneday. Details of the food intake will be publishedelsewhere.

Initial Analysis of the Urine Samples by CompetitiveELISA. Initial efforts to characterize total aflatoxin metab-

olites in the urine samples were made by competitiveELISA. The rabbit antiserum used in this analysis has beencharacterized extensively by Wild et a!. (17). This poly-clonal antiserum has significant recognition for all of themajor aflatoxin metabolites and parent compounds, ex-dept AFP, and AFQ. Fig. 2 depicts the linear regressionanalysis comparing total aflatoxin metabolites in urine/mg of creatinine with the intake of total aflatoxins/day.The aflatoxin contents in urine were normalized for cre-atinine values to adjust for disparities in degree of com-pleteness of urinary collection. The correlation coeffi-cient for the analysis was 0.65, with P < 0.001. Theseresults showed that aflatoxins were present in the urinesamples, and further studies were undertaken to deter-mine specific metabolites, especially AFB-N7-guanine.

High-Pressure liquid Chromatography Separation of Af-latoxin Metabolites in Human Urine. Competitive mi-munoassays revealed a composite representation of theaflatoxin compounds in urine, because no antibody orantiserum available had the specificity to recognize only



30

0

lB 2B 30 40 50Time (mm.)

N7-guanine in urine for the total study group were 48.2and 7099.2 ng/day. Given this analysis it is clear that asummation of excretion and exposure status over thecollection period provides strong evidence for the utilityof AFB-N7-guanine in urine as an appropriate moleculardosimetry marker of exposure.

Finally, this study has provided the opportunity tobegin the investigation of the effect of chronic hepatitisB virus infection on the metabolism of aflatoxins in peo-pIe. Fig. 6 shows the linear regression analysis of theassociation of hepatitis B virus status and the excretionof the AFB-N7-guanine adduct in urine for all subjects.These data reveal that no difference existed betweenHBV carrier status with respect to the ability to convertAFB1 to its genotoxic metabolite. This is also the case ifwe consider only the data from the five female pairswhere food intake of aflatoxin was the same. This analysisclearly demonstrates that hepatitis B virus status does notalter the excretion kinetics of the DNA adduct. Furtherstudies remain to be done to determine the relative risk

one of the aflatoxin metabolites. Thus, a separationmethod using HPLC was required to resolve the individ-ual aflatoxin components in urine. Samples (25 ml) ofurine were analyzed by a combined preparative mono-clonal antibody affinity chromatography/HPLC method,and nearly 80 individual analytical HPLCs were run. Toensure the validity of the method, four separate controlhuman urine samples were spiked with 50 ng of AFB-N7-guanine, AFM1, AFQ1, and AFB1. These samples wereindividually applied to the preparative monoclonal anti-body column and then analyzed by analytical HPLC. Therecovery of each aflatoxin was greater than 90% for thesereference samples.

Fig. 3 shows a representative HPLC of a human urinesample following the preparative monoclonal antibodyaffinity procedure. This chromatogram was selected be-cause, in contrast to previous results (18), the majoraflatoxin found in the urine was aflatoxin G1. In othermolecular dosimetry studies of aflatoxin exposures (18),the diets did not contain any aflatoxin G1; however, inthis study the foods were contaminated with substantialamounts of this parent aflatoxin. The preponderance ofafiatoxin G1 in many of the urine samples may explainthe strong positive association between diet and urinaryanalysis by ELISA seen in Fig. 2. In addition to aflatoxinG1, the oxidative metabolites aflatoxin P1, aflatoxin Qi,and AFB-N7-guanine were measured. Aflatoxin M1 wasalso found in some of the HPLC profiles. All HPLCchromatograms were integrated by area, and each of theidentified aflatoxins was quantified using standard curvesgenerated with authentic standards.

Excretion of AFB-N7-guanine as a Biomarker. Fig. 4shows the total AFB-N7-guanine excretion in the urine ofthe male and female subjects over the complete urinecollection period plotted against the total aflatoxin ex-posure in the diet for each ofthe individuals. This analysissmooths the day-to-day variations in both intake of afla-toxins and excretion of AFB-N7-guanine. Linear regres-sion analysis reveals a correlation coefficient of 0.82,P<0.0001. Fig. 5 shows the box and whisker plot analysisof the day-to-day variation in the excretion of AFB-N7-guanine. The minimum and maximum excretions of AFB-

12

10

8

6

4

2

0100 1000 10000

AFLATOXIN INTAKE (NC PER DAY)

Fig. 4. Linear regression analysis of the association between total afla-toxin-N’-guanine adduct excretion in urine and dietary aflatoxin intakeduring the study period.


Fig. 3. Representative HPLC profile of ahuman urine sample preparatively purifiedusing a monoclonal antibody immunoaf-finity column as described in “Materialsand Methods.”

Lu

z

� 14I-

LU

0

0

0z

LU

zz

0

N

1��



0

z

�. 10000

z

0

z

LUJ

� 1000z

00

z

� 100

H-0

12

.

.

.

� � � I

� � I HI �...

zz

LU

LU

0

0z

LUz

z

0

0

z

LU4 5 6 7

DAY OF COLLECTION

Fig. 5. Box and whisker plot analysis of the total aflatoxin-N7-guanineexcretion per day.

�)O 1000 10000

AFLATOXIN INTAKE PER DAY (No)

Fig. 6. Linear regression analysis of the aflatoxin-N-guanine excretionin urine compared with the total aflatoxin intake per day, dissociated for

HB,Ag carrier and noncarrier status.


14

V HBV (-) V

1O:H:±�.v.

8

6

4

2

0

of individuals with high levels of AFB-N7-guanine in urineand hepatitis B virus exposure.

Discussion

The development of molecular dosimetry markers toassess human exposure to environmental carcinogens isa rapidly expanding field. The potential utility of thesemeasurements is in the assignment of risk from exposure,and the identification of high-risk groups should permitthe allocation of resources for prevention strategies. Anumber of molecular dosimetry markers have been in-vestigated, and many aflatoxin studies have focusedupon the DNA and protein adducts formed by chemicalcarcinogens (5). In relatively few studies has there beena systematic comparison of these markers from the sameindividuals, together with a study of their relationshipswith individual dietary intakes of the carcinogen. Thishas been a major advance in the work done in theGambian population described here and elsewhere.4 Inaddition, this study in The Gambia has afforded one ofthe first comparisons of aflatoxin-DNA and serum albu-mm adduct formation in HBV-positive and -negativepeople. This is important because of the recognition ofhepatitis B virus infection as a major risk factor for humanliver cancer.

In this study, the dietary intakes of aflatoxins in the7-day study period were relatively constant. Initial urine

analyses were done using a polyclonal antiserum thatwas different from the antibody used in other studiesfrom our laboratory, where composite immunoassayanalysis of a urine sample revealed no statistical associa-tion between total aflatoxin metabolite content in theurine and dietary exposure (18). The antibody used inthose investigations had a high level of recognition foraflatoxin Pi. HPLC analysis of the urine samples showedthat AFP1 was a major metabolite, but the urinary con-centration of this metabolite was not dose related. In thepresent study, a polyclonal antiserum was used that didnot have recognition for aflatoxin P1 (15). This differingantibody specificity could partially explain the positiveassociation between urinary metabolites and intake dem-onstrated in these Gambian urine analyses. Another find-ing that may contribute to the strong positive associationbetween aflatoxin metabolites in urine and intake is

attributed to the very high levels of aflatoxin C, a me-tabolite recognized by the antiserum, in these urinesamples. The influence of antibody specificity on thequantitative analysis of a mixture of urinary aflatoxinmetabolites has been discussed (19).

The data on AFG in the urine strongly indicate that

this compound is poorly metabolized in these individ-uals. As was seen in the HPLC profiles of the urinesamples, the contribution of aflatoxin C to the totalcontent in urine was almost 8O% in the highest samples.The mechanistic basis for the lack of metabolism of AFGis intriguing and suggests that the K,,, for AFC metabolismconcurrent with AFB exposure in humans is very low.Some evidence for this possibility may be derived fromthe in vitro metabolism data presented by of Baertschi eta!. (20), who found AFC-DNA adduct formation to bemuch lower than AFB on a molar basis. While theunderlying mechanism is important, the observation ofmajor unmetabolized aflatoxin derivatives in urine, un-related to the biological process mediated by aflatoxinB1, illustrates the problems associated with compositeurinary measurements for dosimetry assessment. Fur-thermore, the AFG1 measurements in the urine demon-strate that an underestimation of the total aflatoxin ex-posure in the food samples had occurred in this study.For example, a number of human urine samples con-tamed 20-30 �zg AFG1 in the total urine volume/day,even though the food intake data indicated that theexposure in the diet was far lower. Even accounting fordiffering half-lives in the excretion of different aflatoxins,the levels of AFG1 in the urine clearly indicate that thefood intakes were underestimated. There are a numberof reasons to account for this that are discussed morefully elsewhere.4 They include food consumed by theindividuals other than at the two main meals and therelatively small sample sizes of the foods that can resultin assigning nondetectable values, hence zero exposure,for a food. However, given the high level of consumptionof a staple food such as rice, even low levels of aflatoxincontent can lead to substantial daily exposures. Theseobservations mainly serve to emphasize the problemsassociated with determining food exposures (1 3) that willbe overcome by use of the molecular dosimetry markers.




The major aflatoxin-DNA adduct is rapidly excisedand exclusively excreted into the urine of experimentalanimals (21) and humans (18). AFB-N7-guanine is stablein urine under long-term storage conditions. Similar toprevious data from China (18), a highly significant positivecorrelation between AFB-N7-guanine in urine and intakewas found in this study. In contrast with the AFG1 infor-mation, clearly AFB1 is metabolized to the highly reactive8,9-epoxide by human cytochrome P-450 enzymes, re-suiting in DNA damage and adduct removal. The dosedependency of excretion of the DNA adduct demon-strates that parallel damage to tissues has occurred inthese people. While the rapidly excreted adduct reflectsonly relatively recent exposures to the parent compound,it is still not understood what the temporal relationshipis between DNA damage and long-term risk from expo-sure in people. This will be examined in future studiesand eventually verified through prospective epidemio-logical investigations.

Finally, this study has provided a small set of datadescribing the impact of HBV infection on the metabo-Iism of aflatoxin to DNA adducts, and no difference wasfound between HB5Ag carriers and noncarniers with re-spect to metabolism. In a larger study of children in TheGambia, significantly higher aflatoxin-albumin adductlevels were seen in the HB�Ag carrier individuals (22).This question clearly merits further study. While theeventual risk of developing cancer is higher in the HBV-positive/afiatoxin-exposed population, further studies areneeded to examine possible interactions of these twofactors that may synergistically amplify liver cancer de-velopment. Several experimental model studies designedto investigate this possible interaction have been re-ported in Peking ducks (23-26) and in transgenic micecarrying the human HBV surface antigen gene (27). Thelatter study showed a marked increase in liver nodules,adenomas, and hepatocellular carcinoma in HBV trans-genic mice treated with AFB in comparison to controlgroups. Thus, prevention methods that lower aflatoxin-DNA adduct formation in HBV-positive people may bean important prevention goal to ameliorate synergy be-tween these two potent carcinogens.

References

1. Van Rensburg, S. J., Cook-Mozaffari, P., Van Schalkwyk, D. J., Van derWatt, I. J., Vincent, T. J., and Purchase, I. F. Hepatocellular carcinomaand dietary aflatoxin in Mozambique and Transkei. Br. J. Cancer, 51: 713-726, 1985.

2. Peers, F., Bosch, x., Kaldor, J., Linsell, A., and Pluijmen, M. Aflatoxinexposure, hepatitis B virus infection and liver cancer in Swaziland. Int. J.Cancer, 39: 545-553, 1987.

3. Yeh, F-S., Yu, M. C., Mo, C-C., Luo, S., Tong, M-J., and Henderson,B. E. Hepatitis B virus, aflatoxins, and hepatocellular carcinoma in south-em Guangxi, China. Cancer Res., 49: 2506-2509, 1989.

4. Beasley, R. P., Hwang, L. Y., Lin, C. C., and Chien, C. S. Hepatocellularcarcinoma and hepatitis B virus. A prospective study of 22,707 men inTaiwan. Lancet, 2: 1129-33, 1981.

5. Busby, W. F., and Wogan, C. N. Aflatoxins. In: C. E. Searle (ed),Chemical Carcinogens, Ed. 2, pp. 945-1136. Washington, D. C.: Amen-can Chemical Society, 1985.

6. Essigmann, J. M., Croy, R. C., Nadzan, A. M., Busby, W. F., Jr.,Reinhold, V. N., Buchi, C., and Wogan, C. N. Structural identification ofthe major DNA adduct formed by aflatoxin B in vitro. Proc. NatI. Acad.Sci. USA, 74: 1870-1874, 1976.

7. Sabbioni, C., Skipper, P., Buchi, C., and Tannenbaum, S. R. Isolationand characterization of the major serum albumin adduct formed byaflatoxin B, in vivo in rats. Carcinogenesis (Lond.), 8: 819-824, 1987.

8. Groopman, J. D., Sabbioni, C., and Wild, C. P. Molecular dosimetryofaflatoxin exposures. In: J. D. Croopman and P. Skipper (eds.), MolecularDosimetry of Human Cancer: Epidemiological, Analytical and SocialConsideration Considerations, pp. 302-324. Boca Raton, FL: CRC Press,1991.

9. Groopman, J. D., Donahue, P. R., Zhu, J., Chen, J., and Wogan, C. N.Aflatoxin metabolism in humans: detection of metabolites and nucleicacid adducts in urine by affinity chromatography. Proc. NatI. Acad. Sci.USA, 82: 6492-6497, 1985.

10. Groopman, J. D., Donahue, P. R., Zhu, J., Chen, J., and Wogan, C.N. Temporal patterns of aflatoxin metabolites in urine of people living inCuangxi province, P. R. C. Proc. Am. Assoc. Cancer Res., 28: 36, 1987.

11 . Can, L-S., Skipper, P. L., Peng, X-C., Croopman, I. D., Chen, J-S.,Wogan, C. N., and Tannenbaum, S. R. Serum albumin adducts in themolecular epidemiology of aflatoxin carcinogenesis: correlation with af-latoxin B, intake and urinary excretion of aflatoxin Ml. Carcinogenesis(Lond.), 9: 1323-1325, 1988.

12. Sabbioni, C., Ambs, S., Wogan, C. N., and Croopman, J. D. Aflatoxin-lysine adducts, quantified by high-pressure liquid chromatography fromhuman serum albumin. Carcinogenesis (Lond.), 1 1: 2063-2066, 1990.

13. Campbell, A. A., Whitaker, T. B., Pohland, A. E., Dickens, J. W., andPark, D. L. Sampling, sample preparation, and sampling plans for food-stuffs for mycotoxin analysis. Pure AppI. Chem., 58: 305-314, 1986.14. Croopman, J. D., and Donahue, K. F. Aflatoxins, a human carcinogen:analyzing foods and biological samples using monoclonal antibody affinitychromatography. J. Assoc. Off. Anal. Chem., 71: 861-867, 1988.

15. Wild, C. P., Umbenhauer, D., Chapot, B., and Montesano, R. Moni-toring of individual human exposure to aflatoxins (AF) and N-nitrosamines(NNO) by immunoassays. J. Cell. Biochem., 30: 171-179, 1986

16. Srivatanakul, P., Parkin, D. M., Jiang, Y. Z., Khlat, M., Kao-Ian, U. T.,Santipong, S., and Wild, C. P. The role of infection by Opisthorchisviverrini, hepatitis B virus and aflatoxin exposure in the etiology of livercancer in Thailand. Cancer (Phila.), 68: 241 1-2417, 1991.

17. Wild, C. P., Jiang, Y. Z., Sabbioni, C., Chapot, B., and Montesano, R.

Evaluation of methods for quantitation of aflatoxin-albumin adducts andtheir application to human exposure assessment. Cancer Res., 50: 245-251, 1990.

18. Croopman, J. D., Zhu, J., Donahue, P. R., Pikul, A., Zhang, L., Chen,I-s., and Wogan, C. N. Molecular dosimetry of urinary aflatoxin DNAadducts in people living in Cuangxi Autonomous Region, People’s Re-public of China. Cancer Res., 52: 45-52, 1992.

19. Olubuyide, I. 0., Makarananda, K., Judah, D. I., and Neal, C. E.Investigation of the assay of AFB,-albumin adducts using proteolysisproducts in ELISA. Int. J. Cancer, 48: 468-72, 1991.

20. Baertschi, S. W., Raney, K. D., Shimada, T., Harris, T. M., andCuengerich, F. P. Comparison of rates of enzymatic oxidation of aflatoxinB, aflatoxin C, and stenigmatocystin and activities of the epoxides informing guanyl-N’ adducts and inducing different genetic responses.Chem. Res. Toxicol., 2: 114-122, 1989.

21. Bennett, R. A., Essigmann, J. M., and Wogan, C. N. Excretion of anaflatoxin-guanine adduct in the urine of aflatoxin B-treated rats. CancerRes., 41: 650-654, 1981.

22. Wild, C. P., Jiang, Y. z., Allen, S. J., Jansen, L. A., Hall, A. J., andMontesano, R. Aflatoxin-albumin adducts in human sera from differentregions ofthe world. Carcinogenesis (Lond.), 11: 2271-2274, 1990.

23. Uchida, M., Esumi, M., and Shikata, T. Molecular cloning and se-quence analysis of duck hepatitis B virus genomes of a new variantisolated from Shanghai ducks. Virology, 173: 600-606, 1989.

24. Cova, L., Wild, C. P., Mehrotra, R., Turusov, V., Shirai, T., Lambert,V., Jacquet, C., Tomatis, L., Trepo, C., and Montesano, R. Contributionof aflatoxin B1 and hepatitis B virus infection in the induction of liventumors in ducks. Cancer Res., 50: 2156-2163, 1990.

25. CulIen, J. M., Marion, P. L., Sherman, C. J., Hong, X., and Newbold,J. E. Hepatic neoplasms in aflatoxin B-treated, congenital duck hepatitisB virus-infected, and virus-free Peking ducks. Cancer Res., 50: 4072-4080, 1990.

26. Qian, C-S., Croopman, J. D., Wogan, C. N., xu, H-)L, and Yu, Y.Effect of duck hepatitis B virus infection on aflatoxin-DNA adductsformation in duck liver, Tumor (Shanghai), 10: 207-210, 1990.

27. Sell, S., Hunt, j. M., Dunsford, H. A., and Chisani, F. V. Synergybetween hepatitis B virus expression and chemical hepatocarcinogens intransgenic mice. Cancer Res., 51: 1278-1285, 1991.



1992;1:221-227. Cancer Epidemiol Biomarkers Prev J D Groopman, A J Hall, H Whittle, et al. obtained in The Gambia, West Africa.Molecular dosimetry of aflatoxin-N7-guanine in human urine

Updated version

http://cebp.aacrjournals.org/content/1/3/221

Access the most recent version of this article at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cebp.aacrjournals.org/content/1/3/221To request permission to re-use all or part of this article, use this link



http://cebp.aacrjournals.org/cgi/alerts

mailto:[email protected]



molecular dosimetry of aflatoxin-n7-guanine in human urine ... · coefficient of 0.82 with p <...

Documents