- Email: [email protected]
Binding of [14C]aflatoxin B1 to cellular macromolecules in the rat and hamster
t&m.-Rio/. Inteructions, ll(1975) 123431 0 ElsevierScientific Publishing Company, Amsterdam- Printed in The Netherlands
123
BINDING OF [W]AFLATOXIN THE RAT AND HAMSTER*
IN
R. COLIN GARNER
AND
Br TO CELLULAR MACROMOLECULES
CATHERINE M. WRIGHT
Demrtment of &.wrimental Pathology and Cancer Research, University of Leeds, 171 W&&we Lane, Leeds LS2 JAR (Great Eritain)
(Received December 28th, 1974) (Revision receivedJanuary 3Oth,1975) (Accepted February 12th, 1975)
SUMMARY
The uptake and binding of ring-labelled [Wlaflatoxin Br (AFBt) by rat and hamster liver and kidney has been studied, the former species being extremely sensitive to the carcinogenic action of AFBt, whereas the latter is resistant. In contrast to an earlier report (Lijinsky et al, Cancer Res., 30 (1970) 2280-2283, binding of the carcinogen to nucleic acids was far greater than that to protein. Rat liver DNA bound ten times and rRNA twenty times more carcinogen than protein. There were also differences in the amount of carcinogen bound to rat liver nucleic acids compared to those of the hamster, the latter species binding lower amounts ofthecarcinogen. Rat liver DNA bound four times and rRNA ten times as much AFBr 6 h after carcinogen administration whereas liver protein bound AFBl was similar for the two species. Not only was there a difference in the amount of AFBi bound but whereas in the rat, liver nucleic acid bound carcinogen decayed with time, no such f& was seen in the hamster, this remaining at a low level throughout the 48-h time period studied. In contrast, reaction of the carcinogen with kidney macromolecules was similar for the two species. The much higher binding of AFBr to nucleic acids than to protein might account for the potent carcinogenicity of this compound in the rat, particularly since liver protein binding does not differ between a susceptible and a resistant species. A further important factor in determining carcinogenic sensitivity may be the removal of nucleic acid bound radioactivity with time, a possible repair pro-s. -
* A preliminarycommunication of this work was presentedat the Ekventh InternationalCancer Congress, Florence, Italy, 1974. Abbreviations: AFBl, aflatoxin 81.
INTRODUCTION
AFBl, one of four related mycotoxins elaborated by AspergilfusJlavus has been shown to be toxid-3 and carcinogenic4 for a wide species range. The compound is ot interest since it is widely distributed as a human and animal food contaminant5-* and its ingestion has been associated with the high primary liver cancer incidence in certain parts of the world 9*10.Its main effects in IaLoratory animals are in the liver, indicating that metabolism in this organ may be important for the biological action of this carcinogen. Previous work has shown that in vitro metabolism of AFBI by liver mixed-function oxidases converts it to a toxic and mutagenic metabolite for microorganismsll-la and that this metabolite reacts with nucleic acids and proteinl”. On the basis of structure-activity studies it was suggested that 2,3-epoxy AFBl was the electrophilic species formed during metabolisml2, a proposal which has recently received strong support from chemical studieslsJ6. One anomalous finding obtained from the in vitro studies using microorganisms was the amount of activated metabolite apparently produced by hamster liver, a species relatively resistant to the carcinogenic action of AFBl (ref. 17). If 2,3-epoxy AFBl is the ultimate carcinogenic form of this compound one might expect this species to produce les#sof the metabolite and not more. To investigate this point further, binding studies of ring-labelled [14C]AFBl to liver and kidney cellular macromolecules have been carried out in the rat and hamster. While this manuscript was in preparation a short communication published by Swenson et al.35 appeared confirming some of the data reported in this paper. MATERIALS AND METHODS
[l*C]AFBl was prepared as previously described and adjudged to be radiochemically pure by silica gel thin-layer chromatography on comparison with an AFBl standardla. All the radioactivity was associated with the fluorescent aflatoxin spot; the specific radioactivity varied from preparation to preparation depending on the amount of aflatoxin produced and was usually between 4-6 mCi/mmole. 200-250 g male Wistar rats or 120-140 g male Golden Syrian hamsters (A. Tuck and Son, Raleigh, Essex) maintained on diet 41B (Oxoid Ltd., London) were injected intraperitoneally with 4Opg/lOOg [14C]AFBl dissolved in dimethylsulphoxide (40 pg AFBl/O.l ml) between 9 and 10 am. 3 animals were used to measure uptake and binding of the labeiled compound for each time point studied. Animals were lightly anaesthetised with ether prior to killing, exsanguinated from the carotid artery and then killed by cervical dislocation. The liver and kidneys from each animal were weighed and then dropped into ice cold 6% w/v para-aminosalicylic acid solution (1 vol. liver + 3 vol. p-aminosalicylic acid solution). After homogenisation an aliquot was removed for stadium hydroxide digestion to determine total organ radioactivity by liquid scintillation counting and the remainder extracted with an equal volume of phenol-cresol solution. The kidney homogenates from each time point were combined prior to phenol-cresol extraction to obtain better yields of
Q---B--r-r-----_e~lr-
O-J
r
I
02
-I 24
I
6 rJME
I 4P
(how)
Fig. I. Loss of radioactivity from rat liver and kidney after a single administration of 40 &lOO g WJAFBr. Each time point is the mean from 3 animals f 1 SD. O-e, liver; CJ- - -0, kidney.
nucleic acid. Ribosomal RNA and DNA were prepared by the method of Irving and Veazeyls. Protein was precipitated by addition of an equal volume of methanol to the phenol-cresol layer and then washed by a modification of the Schmidt-Thannhauser procedure’s, For radioactivity determinations nucleic acids were digested by heating in 0.5 A4 perchloric acid at 80” for 15 min. Protein was digested by overnight incubation in 1 M sodium hydroxide at 37”. Amounts of DNA were estimated by the diphtnylamine reaction20 using calf-thymus DNA as standard, rRNA by the orcinol reactionzl, using yeast tRNA as standard and protein by the Lowry-Folin procedure22, using bovine serum albumin as standard. Radioactivity of these digests was measured in Brays scintillator in a Multimat liquid scintillation counter (Intertechnique, France).
zo-
IO-
OA
r
02
I
I
I
6
24 NME
1
48
(now a)
Fig. 2. Loss of radioactivity from hamster liver and kidney after a single administration of 40 /fgl 100 g [W]AFBI. Each time point is the mean from 3 animals i 1 SD. O-O, liver; C- - -3, kidney.
126 npAFB,bound/w macromolrtulr SO
40 50 20 IO 0i I 02
I
I
I
24
6
I 48
TIME (hours) Fig. 3. Time-coursefor binding of [WIAFBI to rat liver DNA, rRNA and protein after a single administrationof 40 ~g/lOOg. Each time point is the mean from 3 animals rt 1 SD. e-0, rRNA; DNA; A---A, protein. 0 ---0,
Counting efficiency was determined by automatic external standardisation ratio and was greater than 80 %. The standard deviation for counting was 5 % or less. RESULTS
Uptake and binding of [WJAFBl in rat and hamster liver After intraperitoneal injection of [W]AFBl there is a rapid uptake of radioactivity by the liver in both the hamster and rat which reaches a maximum by 2 h (Figs. 1 and 2). The percentage of the dose in rat liver at 4 h lies between that at 2 and 6 h (data not shown). At the early time points approximately 50% of the radioactivity in the rat liver is unbound AFBl or its metabolites while in the hamster some 75% of the radioactivity is not bound to nucleic acid or protein (these calculations obtained by subtraction assume each gram of liver to contain 2 mg DNA, 10 mg RNA and 160 mg protein). Even at 24 h some 30% of the radioactivity in both the rat and hamster is unbound. ng AF8, bound/ mgmacromolecule
I 02
I
I
I
6
24
I
48
tfMf (hours) Fig. 4. Time-coursefor binding of [WIAFBI to hamster liver DNA, rRNA and protein after a single administrationof 40 pg/lOOg. Each time point is the mean from 3 animals f I SD. protein. O-O, rRNA; 0 ---0, DNA; A---A,
127 ng AFB,bound/at9 macromolrculc
@w-‘---mnn”..~.u,**”
0-J I
02
I
I
6
b
---mI--MmamDaIg u-m”..un
I
TlME (hrs)
24
1
40
Fig. 5. Timecourse for binding of [W]AFBl, to rat kidney DNA, rRNA and protein after a single administration of 40 pg/lOOg. Each time point was obtained by pooling kidneys from 3 animals. O-0, rRNA; 0- - - 0, DNA; A*‘,& protein.
Nucleic acid binding of [l’%]AFBl is higher early after carcinogen administration in the rat than in the hamster, although by 48 h there is not a great difference between the two species (Figs. 3 and 4). What is of interest is the fall in rat nucleic acid binding over the period of study compared to the hamster where the binding is relatively stable (the difference between hamster liver DNA binding at 24 and 48 h is not signi~~nt, p > 0.05 by Student’s t test). The total amount of AFBl bound to protein is greater than the total bound to nucleic acid, although on a per mg basis very much less carcinogen is protein-bound. 6 h after AFBl administration protein bound carcinogen is higher in the rat than the hamster but this rapidly falls until by 24 h there is little difference between the two species.
Fig. 6. Time_rourse for binding of [WJAFBl to hamster kidney DNA, rRNA ati Protein after a single ad~iaist~tion of 40 &lcK, g. Eachtime point was obtained by poolingkidneysfrom 3animals*--a, rRNA; O- - -0, DNA; A**- A, protein.
128 Uptake and binding of [W/AFBI by rat and hamster kidney A small percentage of the injected carcinogen is taken up by both rat and hamster kidneys; on a per gram basis less radioactivity is found in this organ than in the liver. 6 h after AFBl administration there is more radioactivity in the hamster kidney than in the rat, a possible reflection of the faster excretion of AFBr in this species (Figs. 1 and 2). Binding to macromolecules in this organ is similar for both species at the time points studied although the loss of bound radioactivity from rat rRNA is slower than from the other macromolecular species. As in the liver there is more reaction of the carcinogen with nucleic acids than with protein in both species (Figs. 5 and 6). DISCUSSIOPI;
Earlier studies using micro-organisms and liver microsomes have shown that AFBl is converted by mixed-function oxidase enzymes to an electrophilic species which is toxic and mutagenic to the micro-organisms 11-13. This activated metabolite will also react with nucleic acids14 and on the basis of chemical and structure-activity studies was postulated to be the 2,3-epoxide of AFBr (refs 12, 15). The data presented here show that a reactive metabolite is also formed in viva after AFBl administration in both the rat and the hamster. Although hamster liver microsomes were more active than those of the rat in converting AFBl into its epoxide, i,n viva there is considerably less macromolecular binding of carcinogen in the hamster liver than in the rat. The percentage of the administered dose of carcinogen taken up by the liver is similar for the two species, and for the rat agrees with the results of other workers23 even though larger rats of a different strain were used. The amount of AFBr bound per mg of nucleic acid was greater in both the liver and kidney than protein binding in either species, a result at variance with that reported on the binding of sH-labelled AFBr to macromoleculesa*. In this earlier paper much higher binding was shown to proteins than to nucleic acids and this binding was very long lived. This difference is probably due to the use of ring labelled [t%Z]AFBr in this study which was radiochemically pure while in the earlier presentation only 80% of the label was associated with the carcinogen. Considerable tritium exchange may have taken place giving a false impression of binding. In a recent communication [sH]AFBr of high purity was used to study macromolecular binding of carcinogen in the rats5. In agreement with the results reported in this paper higher binding was found to liver nucleic acids than to protein. Even so, up to 10% of the tritium label was found to be exchangeable showing the advantage of using ring-labelled material for binding studies. The radioactivity associated with nucleic acids is not due to one carbon fragment incorporation from degraded carcinogen since after gel permeation chromatography of enzyme digested material there is complete separation of the radioactivity from the unreacted nucleosides (Garner and Wright, manuscript in preparation). Also at the early time points after carcinogen administration insufficient time has elapsed for such incorporation to occur. A further interesting finding is that whereas there is a continuous removal of
bound carcinogen from rat liver nucleic acids over the period studied there is little loss of nucleic acid-bound carcinogen in the hamster. The half-life for loss of car&togen from rat liver rRNA and DNA is between 12 and 14 h, and at 6 h, the *ime of maximum binding, approximately I molecule of carcinogen is bound per 2oooo bases in rRNA and one per 4OHlOin DNA. Although there are some differences in the amount of AFBr binding to macromolecules in the liver between the rat and hamster there is little difference in binding in the kidney. In both species there is again more binding to nucleic acids than protein; rat rRNA binding takes longer to fall to a basal level than binding to DNA. Interestingly, whereas in the hamster liver there is little alteration in binding to DNA over the 48-h period, there is a fall in binding to nucleic acids in the hamster ki.Lrey. This might indicate that in the hamster the kidney is more susceptible to biological effects of AF& than the liver, a finding which has already been demonstrated in the mouse, another resistant specie@. There have been a number of studies on the macromolecular binding of labelled carcinogens to macromolecules. None of these have given a clear indication which macromolecular binding is important for carcinogenicity although it is generally true that nucleic acid binding is greater on a per molecule basis than protein binding26-28. Comparisons have also been made on the binding of carcinogens and non-carcinogens to macromolecules and again one finds generally a higher degree of binding of the carcinogen to nucleic acids compared to the non-carcinogerP. Further studies have shown that some carcinogens bind persistently to nucleic acids over long periods after adrninist~tion~~~si whereas others have shown a rapid disap~aran~ of bound careinogenas-34. In the case of AFBl nucleic acid binding is considerably greater than protein binding even on a per mg basis. For example, 6 h after AFBt administration to rats, for every molecule of protein reacted with carcinogen there are 400 molecules of DNA with bound carcinogen (this assumes DNA to have an average molecular weight of I - IO7 and protein of I 105). On the basis of target theory this would indicate that DNA binding must be of greater importance than protein binding for tumour initiation, particularly as the major liver protein alkylated by the activated metabolite appears to be albumin (Ketterer and Garner, manuscript in preparation). Besides the im~rtan~ of nucleic acid binding there is the further inte~sting finding that the labeiled carcinogen is removed over the time period studied in rat liver. There are a number of possible explanations for this: (I) The bound carcinogen is removed by enzymatic excision through some DNA repair process, The fact that ~pair-de~cient ~cteria are more sensitive to the microsomal activated metabolite of AFBr than wild-type bacteria indicates that excision repnir in micro-organisms is important to sensitivityts. Whether this q@ics to liver cells is not known. (2) The bound carcinogen leads to chemical depurination. If this were SO.then there should be some loss of nucleic acid bound ca~inogen in the hamsterus well as the rat unless different positions on the nucleic acid bases were dkylated. This is unlikely. (3) Cell death leads to loss of the bound carcinogen, particularly if only a few l
130 cells are highly alkylated. Although the dose of AFBl administered produces no microscopic evidence of liver necrosis it is not possible to exclude this explanation to account for the loss of carcinogen without more detailed studies, It is interesting that the half-life of loss of bound carcinogen is approximately the same as that found for #6-methylguanine, a methylatsd base thought to be important for the carcinogenic action of the chemical methylating agent@. A comparison of the amount and rate of loss of AFBs binding in the livers of the hamster and the rat shows two fundamental differences between the two species. First, that the overall level of nucleic acid binding is higher in the rat than the hamster and second, that there is a faster removal of bound carcinogen in the former species. The difference in the amount of binding probably reflects a difference in the pharmacokinetics of AFBl metabolism between the two species, there being more metabolism via the detoxification pathways in the hamster than the rat. The faster loss of nucleic acid-bound carcinogen in the rat compared with the hamster may also account for the greater sensitivity of this species to AFBr carcinogenesis. These two major differences may therefore be intimately involved in explaining why the rat is sensitive to the liver carcinogenicity of this compound while the hamster is resistant. ACKNOWLEDGEMENTS
This work was supported by a grant from the Yorkshire Council of the Cancer Research Campaign.
REFERENCES I R. W. Detroy, E. E. Lillehoj and A. Ciegler, Aflatoxin and related compounds, in Microbial Toxins, Chapter 1, Vol. 6, Academic Press, New York, 1971, pp. 3-178. 2 W. H. Butler, Aflatoxicosis in laboratory animals, in L. A. Goldblatt (Ed.), Ajlatoxin, Chapter 8, Academic Press, New York, 1969, pp. 223-234. 3 G. N. Wogan, Experimental toxicity and carcinogenicity of aflatoxins, in G. N. Wogan (Ed.), Mycotoxins in Foodstufi, M.I.T., Cambridge, Mass., 1965, pp. 163-173. 4 P. M. Newbeme, and W. H. Butler, Acute and chronic effects of aflatoxin on the liver of domestic and laboratory animals: a review, Cancer Res., 29 (1969) 236-250. 5 K. Sargeant, S. Sheridan, J. O’Kelly and R. B. A. Carnaghan, Toxicity associated with certain samples of groundnuts, N&we, 192 (1961) 1096-1097. 6 G. A. Bean, J. A. Schillinger and W. L. Klarman, Occurrence of aflatoxins and aflatotixin-producing strains of Aspergillus sop. on Soybeans, Appt. Microbial., 24 (1972) 437-439. 7 M. Sutic, J. C. Ayres and I’. E. Koehler, Identification and aflatoxin production of molds isolated from county cured hams, Appt. Microbial., 23 (1972) 656-658. 8 H. W. Schroeder and R. A. Boller, Aflatoxin production of species and strains of the AspergiUus Paws group isolated from fieldcrops, Appl. Microbial., 25 (1973) 885-889. 9 R. C. Shank, N. Bhamaropravati, J. E. Gordon and G. N. Wogan, Dietary aflatoxins and human liver cancer, IV. Incidence of primary liver cancer in two municipal populations of Thailand, Food Cosmet. Toxicol., 10 (1972) 171-179. 10 F. G. Peers and C. A. Linsell, Dietary aflatoxins and liver cancer - A population-based study in Kenya, Brit. J. Cancer, 27 (1973) 473-484. 11 R. C. Garner, E. C. Miller, J. A. Miller, J. V. Garner and R. S. Hanson, Formation of a factor lethal for S. typhimurium TA 1530 and TA 1531 on incubation of aflatoxin Bl with rat liver
microsomes,Eiochem. Biophys. Res.
Commun., 45 (1971) 774-780.
131 12
R. C. Garner, E. C. Miller and 3. A. Miller, Liver microsomal metabolism of afjatoxin & to a reactive derivative toxic to Salmon&a fyphimurium TA 1530, Cancer Res., 32 (1972) 2058-2066. 13 R. C. Garner and C. M. Wright, Induction of mutations in DNA repair-deficient bacteria by a liver microsomal metabolite of aflatoxin BI, &it. J. Cuncer, 28 (1973) 544-551. 14 R. C. Garner, Microsome-dependent binding of aflatoxin 31 to DNA, RNA, polyribonucleotides and protein in v&u, CIiem.-Biol. Interact., 6 (1973) 125-129. 15 R. C. Garner, Chemical evidence for the formation of a reactive ~atoxin I& metabolite by hamster liver microsomes, FEBS brr., 36 (1973) 261-264. 16 D. H. Swenson, J. A. Miller and E. C. Miller, 2,3-Dihydro-aflatoxin &: An acid hydrolysis product of an RNA-aflatoxin BI adduct formed by hamster and rat liver microsomes in vitro, Biochem. Biop+vs. Res. Comm., 53 (1973) 1260-1267. 17 K. M. Herrold, Aflatoxin induced lesions in Syrian hamsters, &if. J. Cancer, 23 (1969) 655-660. 18 C. C. Irving and R. A. Veazey, Isolation of d~xyr~bon~eic acid from rat liver, Bio~iiir. Biophys, Acra, 166 (1’968) 246-248.
19 G. Schmidt and S. J. Thannhauser, A method for the determination of deoxyribonucleic acid and phospho-proteins in animal tissues, J. Biol. Chem., 161 (1945) 83-89. 20 K. Burton, 4 study of the conditions and mechanism of the diphenylamine reaction for the calorimetric estimation of deoxyribonucleic acid, Biochem. J., 62 (1956) 315-323, 21 G, Ceriotti, Determination of nucleic acids in animal tissue, J. Bid. Chem., 214 (1955) 59-70. 22 H. Lowry, N. J. Ro~brough, A. L. Farr and R. J. Randall, Protein rn~su~nt with the Folin phenol reagent, J. Biot, Chem., 193 (1951} 265-272. 23 G. N. Wogan, G. S. Edwards and P. M. Newberne, Excretion and tissue distribution of radioactivity from aflatoxin B@C in rats, Cancer Res., 27 (1967) 1729-1736. 24 W. Lijinsky, K. Y. Lee and C. H. Gallagher, Interaction of aflatoxins BI and GI with tissues of the rat, Gamer Res., 30 (1970) 2280-2283. 25 M. Akao, K. Kuroda and G. N. Wogan, Afiatoxin BI: the kidney as a site of action in the mouse, Life Sci., 10 (1971) 495-501. 26 P. Brookes, Quantitative aspects of the reaction of some carcinogens with nucleic acids and the possible significance of such reactions in the process of carcinogenesis, Cancer Rex, 26 (1966) 1994-2003. 27 T. A. Lawson, The binding of o-aminoazotoluene to deoxyribonucleic acid, ribonucleic acid and protein in the C57 mouse, Biuchem. J., 109 (1968) 917-920. 28 E. Kriek, Carcinogenesis by aromatic amines, Biorhj/il. Biophys. Ada, 355 (1974) 177-203. 29 0. P. Warwick, The covalent binding of metabolites of tritiated 2-methy~~imethylaminoazobenzene to rat liver nucleic acids and proteins, and the carcinogenicity of the unlabelled compound in partially hepatectomised rats, Eur. J. C’umer, 3 (1967) 227-233. 30 C. C. Irving and R. A, Vsazey, A persistent binding of 2-acetylaminofluorene to rat liver DNA in vivo and consideration of the mechanism of binding of IV-hydroxy-2acetylaminofluorene to rat liver nucleic acids, Cancer Res., 29 (1969) 1799-1804. 31 E. Kriek, Persistent binding of a new reaction product of the carcinogen, ~hydroxy-~-2acetylaminoffuorene with guanine in rat liver in viva, Cuneer Res., 32 (1972) 2042-2048. 32 V. M. Craddock, The pattern of methylated purines formed in DNA of intact and regenerating liver of rats treated with the carcinogen dimethylnitrosamine, Biochim. Biophys. Acta, 312 (1973) 202-210. 33 M. J. Capps, P. J. O’Connor and A. W. Craig, The influence of liver regeneration on the stability of 7-methylguanine in rat liver DNA after t~atment with ~,~~imethylnitro~mine, Biochim. Bioph~~s,Actu, 33 t (1973) 3340. 34 P. J. O’Connor, M. J. Capps and A. W. Craig, Comparative studies on the hepatocarcinogen
N,N-dimethylnitrosamine 1’11 viva: Reaction sites in rat liver DNA and the significance of their J. Crrr~ce~;27 (1973) 153-166. 35 D. H. Swenson, E. C. Miller and S. A. Miller, Aflatoxin BI-2,3-oxide: evidence for its formation in rat liver it1viva and by human liver microsomes b vitro, Biochem. Biophys* Res. Commlm.+ 60 (1974) 1036-1043. relative stabilitics, &it.
123
BINDING OF [W]AFLATOXIN THE RAT AND HAMSTER*
IN
R. COLIN GARNER
AND
Br TO CELLULAR MACROMOLECULES
CATHERINE M. WRIGHT
Demrtment of &.wrimental Pathology and Cancer Research, University of Leeds, 171 W&&we Lane, Leeds LS2 JAR (Great Eritain)
(Received December 28th, 1974) (Revision receivedJanuary 3Oth,1975) (Accepted February 12th, 1975)
SUMMARY
The uptake and binding of ring-labelled [Wlaflatoxin Br (AFBt) by rat and hamster liver and kidney has been studied, the former species being extremely sensitive to the carcinogenic action of AFBt, whereas the latter is resistant. In contrast to an earlier report (Lijinsky et al, Cancer Res., 30 (1970) 2280-2283, binding of the carcinogen to nucleic acids was far greater than that to protein. Rat liver DNA bound ten times and rRNA twenty times more carcinogen than protein. There were also differences in the amount of carcinogen bound to rat liver nucleic acids compared to those of the hamster, the latter species binding lower amounts ofthecarcinogen. Rat liver DNA bound four times and rRNA ten times as much AFBr 6 h after carcinogen administration whereas liver protein bound AFBl was similar for the two species. Not only was there a difference in the amount of AFBi bound but whereas in the rat, liver nucleic acid bound carcinogen decayed with time, no such f& was seen in the hamster, this remaining at a low level throughout the 48-h time period studied. In contrast, reaction of the carcinogen with kidney macromolecules was similar for the two species. The much higher binding of AFBr to nucleic acids than to protein might account for the potent carcinogenicity of this compound in the rat, particularly since liver protein binding does not differ between a susceptible and a resistant species. A further important factor in determining carcinogenic sensitivity may be the removal of nucleic acid bound radioactivity with time, a possible repair pro-s. -
* A preliminarycommunication of this work was presentedat the Ekventh InternationalCancer Congress, Florence, Italy, 1974. Abbreviations: AFBl, aflatoxin 81.
INTRODUCTION
AFBl, one of four related mycotoxins elaborated by AspergilfusJlavus has been shown to be toxid-3 and carcinogenic4 for a wide species range. The compound is ot interest since it is widely distributed as a human and animal food contaminant5-* and its ingestion has been associated with the high primary liver cancer incidence in certain parts of the world 9*10.Its main effects in IaLoratory animals are in the liver, indicating that metabolism in this organ may be important for the biological action of this carcinogen. Previous work has shown that in vitro metabolism of AFBI by liver mixed-function oxidases converts it to a toxic and mutagenic metabolite for microorganismsll-la and that this metabolite reacts with nucleic acids and proteinl”. On the basis of structure-activity studies it was suggested that 2,3-epoxy AFBl was the electrophilic species formed during metabolisml2, a proposal which has recently received strong support from chemical studieslsJ6. One anomalous finding obtained from the in vitro studies using microorganisms was the amount of activated metabolite apparently produced by hamster liver, a species relatively resistant to the carcinogenic action of AFBl (ref. 17). If 2,3-epoxy AFBl is the ultimate carcinogenic form of this compound one might expect this species to produce les#sof the metabolite and not more. To investigate this point further, binding studies of ring-labelled [14C]AFBl to liver and kidney cellular macromolecules have been carried out in the rat and hamster. While this manuscript was in preparation a short communication published by Swenson et al.35 appeared confirming some of the data reported in this paper. MATERIALS AND METHODS
[l*C]AFBl was prepared as previously described and adjudged to be radiochemically pure by silica gel thin-layer chromatography on comparison with an AFBl standardla. All the radioactivity was associated with the fluorescent aflatoxin spot; the specific radioactivity varied from preparation to preparation depending on the amount of aflatoxin produced and was usually between 4-6 mCi/mmole. 200-250 g male Wistar rats or 120-140 g male Golden Syrian hamsters (A. Tuck and Son, Raleigh, Essex) maintained on diet 41B (Oxoid Ltd., London) were injected intraperitoneally with 4Opg/lOOg [14C]AFBl dissolved in dimethylsulphoxide (40 pg AFBl/O.l ml) between 9 and 10 am. 3 animals were used to measure uptake and binding of the labeiled compound for each time point studied. Animals were lightly anaesthetised with ether prior to killing, exsanguinated from the carotid artery and then killed by cervical dislocation. The liver and kidneys from each animal were weighed and then dropped into ice cold 6% w/v para-aminosalicylic acid solution (1 vol. liver + 3 vol. p-aminosalicylic acid solution). After homogenisation an aliquot was removed for stadium hydroxide digestion to determine total organ radioactivity by liquid scintillation counting and the remainder extracted with an equal volume of phenol-cresol solution. The kidney homogenates from each time point were combined prior to phenol-cresol extraction to obtain better yields of
Q---B--r-r-----_e~lr-
O-J
r
I
02
-I 24
I
6 rJME
I 4P
(how)
Fig. I. Loss of radioactivity from rat liver and kidney after a single administration of 40 &lOO g WJAFBr. Each time point is the mean from 3 animals f 1 SD. O-e, liver; CJ- - -0, kidney.
nucleic acid. Ribosomal RNA and DNA were prepared by the method of Irving and Veazeyls. Protein was precipitated by addition of an equal volume of methanol to the phenol-cresol layer and then washed by a modification of the Schmidt-Thannhauser procedure’s, For radioactivity determinations nucleic acids were digested by heating in 0.5 A4 perchloric acid at 80” for 15 min. Protein was digested by overnight incubation in 1 M sodium hydroxide at 37”. Amounts of DNA were estimated by the diphtnylamine reaction20 using calf-thymus DNA as standard, rRNA by the orcinol reactionzl, using yeast tRNA as standard and protein by the Lowry-Folin procedure22, using bovine serum albumin as standard. Radioactivity of these digests was measured in Brays scintillator in a Multimat liquid scintillation counter (Intertechnique, France).
zo-
IO-
OA
r
02
I
I
I
6
24 NME
1
48
(now a)
Fig. 2. Loss of radioactivity from hamster liver and kidney after a single administration of 40 /fgl 100 g [W]AFBI. Each time point is the mean from 3 animals i 1 SD. O-O, liver; C- - -3, kidney.
126 npAFB,bound/w macromolrtulr SO
40 50 20 IO 0i I 02
I
I
I
24
6
I 48
TIME (hours) Fig. 3. Time-coursefor binding of [WIAFBI to rat liver DNA, rRNA and protein after a single administrationof 40 ~g/lOOg. Each time point is the mean from 3 animals rt 1 SD. e-0, rRNA; DNA; A---A, protein. 0 ---0,
Counting efficiency was determined by automatic external standardisation ratio and was greater than 80 %. The standard deviation for counting was 5 % or less. RESULTS
Uptake and binding of [WJAFBl in rat and hamster liver After intraperitoneal injection of [W]AFBl there is a rapid uptake of radioactivity by the liver in both the hamster and rat which reaches a maximum by 2 h (Figs. 1 and 2). The percentage of the dose in rat liver at 4 h lies between that at 2 and 6 h (data not shown). At the early time points approximately 50% of the radioactivity in the rat liver is unbound AFBl or its metabolites while in the hamster some 75% of the radioactivity is not bound to nucleic acid or protein (these calculations obtained by subtraction assume each gram of liver to contain 2 mg DNA, 10 mg RNA and 160 mg protein). Even at 24 h some 30% of the radioactivity in both the rat and hamster is unbound. ng AF8, bound/ mgmacromolecule
I 02
I
I
I
6
24
I
48
tfMf (hours) Fig. 4. Time-coursefor binding of [WIAFBI to hamster liver DNA, rRNA and protein after a single administrationof 40 pg/lOOg. Each time point is the mean from 3 animals f I SD. protein. O-O, rRNA; 0 ---0, DNA; A---A,
127 ng AFB,bound/at9 macromolrculc
@w-‘---mnn”..~.u,**”
0-J I
02
I
I
6
b
---mI--MmamDaIg u-m”..un
I
TlME (hrs)
24
1
40
Fig. 5. Timecourse for binding of [W]AFBl, to rat kidney DNA, rRNA and protein after a single administration of 40 pg/lOOg. Each time point was obtained by pooling kidneys from 3 animals. O-0, rRNA; 0- - - 0, DNA; A*‘,& protein.
Nucleic acid binding of [l’%]AFBl is higher early after carcinogen administration in the rat than in the hamster, although by 48 h there is not a great difference between the two species (Figs. 3 and 4). What is of interest is the fall in rat nucleic acid binding over the period of study compared to the hamster where the binding is relatively stable (the difference between hamster liver DNA binding at 24 and 48 h is not signi~~nt, p > 0.05 by Student’s t test). The total amount of AFBl bound to protein is greater than the total bound to nucleic acid, although on a per mg basis very much less carcinogen is protein-bound. 6 h after AFBl administration protein bound carcinogen is higher in the rat than the hamster but this rapidly falls until by 24 h there is little difference between the two species.
Fig. 6. Time_rourse for binding of [WJAFBl to hamster kidney DNA, rRNA ati Protein after a single ad~iaist~tion of 40 &lcK, g. Eachtime point was obtained by poolingkidneysfrom 3animals*--a, rRNA; O- - -0, DNA; A**- A, protein.
128 Uptake and binding of [W/AFBI by rat and hamster kidney A small percentage of the injected carcinogen is taken up by both rat and hamster kidneys; on a per gram basis less radioactivity is found in this organ than in the liver. 6 h after AFBl administration there is more radioactivity in the hamster kidney than in the rat, a possible reflection of the faster excretion of AFBr in this species (Figs. 1 and 2). Binding to macromolecules in this organ is similar for both species at the time points studied although the loss of bound radioactivity from rat rRNA is slower than from the other macromolecular species. As in the liver there is more reaction of the carcinogen with nucleic acids than with protein in both species (Figs. 5 and 6). DISCUSSIOPI;
Earlier studies using micro-organisms and liver microsomes have shown that AFBl is converted by mixed-function oxidase enzymes to an electrophilic species which is toxic and mutagenic to the micro-organisms 11-13. This activated metabolite will also react with nucleic acids14 and on the basis of chemical and structure-activity studies was postulated to be the 2,3-epoxide of AFBr (refs 12, 15). The data presented here show that a reactive metabolite is also formed in viva after AFBl administration in both the rat and the hamster. Although hamster liver microsomes were more active than those of the rat in converting AFBl into its epoxide, i,n viva there is considerably less macromolecular binding of carcinogen in the hamster liver than in the rat. The percentage of the administered dose of carcinogen taken up by the liver is similar for the two species, and for the rat agrees with the results of other workers23 even though larger rats of a different strain were used. The amount of AFBr bound per mg of nucleic acid was greater in both the liver and kidney than protein binding in either species, a result at variance with that reported on the binding of sH-labelled AFBr to macromoleculesa*. In this earlier paper much higher binding was shown to proteins than to nucleic acids and this binding was very long lived. This difference is probably due to the use of ring labelled [t%Z]AFBr in this study which was radiochemically pure while in the earlier presentation only 80% of the label was associated with the carcinogen. Considerable tritium exchange may have taken place giving a false impression of binding. In a recent communication [sH]AFBr of high purity was used to study macromolecular binding of carcinogen in the rats5. In agreement with the results reported in this paper higher binding was found to liver nucleic acids than to protein. Even so, up to 10% of the tritium label was found to be exchangeable showing the advantage of using ring-labelled material for binding studies. The radioactivity associated with nucleic acids is not due to one carbon fragment incorporation from degraded carcinogen since after gel permeation chromatography of enzyme digested material there is complete separation of the radioactivity from the unreacted nucleosides (Garner and Wright, manuscript in preparation). Also at the early time points after carcinogen administration insufficient time has elapsed for such incorporation to occur. A further interesting finding is that whereas there is a continuous removal of
bound carcinogen from rat liver nucleic acids over the period studied there is little loss of nucleic acid-bound carcinogen in the hamster. The half-life for loss of car&togen from rat liver rRNA and DNA is between 12 and 14 h, and at 6 h, the *ime of maximum binding, approximately I molecule of carcinogen is bound per 2oooo bases in rRNA and one per 4OHlOin DNA. Although there are some differences in the amount of AFBr binding to macromolecules in the liver between the rat and hamster there is little difference in binding in the kidney. In both species there is again more binding to nucleic acids than protein; rat rRNA binding takes longer to fall to a basal level than binding to DNA. Interestingly, whereas in the hamster liver there is little alteration in binding to DNA over the 48-h period, there is a fall in binding to nucleic acids in the hamster ki.Lrey. This might indicate that in the hamster the kidney is more susceptible to biological effects of AF& than the liver, a finding which has already been demonstrated in the mouse, another resistant specie@. There have been a number of studies on the macromolecular binding of labelled carcinogens to macromolecules. None of these have given a clear indication which macromolecular binding is important for carcinogenicity although it is generally true that nucleic acid binding is greater on a per molecule basis than protein binding26-28. Comparisons have also been made on the binding of carcinogens and non-carcinogens to macromolecules and again one finds generally a higher degree of binding of the carcinogen to nucleic acids compared to the non-carcinogerP. Further studies have shown that some carcinogens bind persistently to nucleic acids over long periods after adrninist~tion~~~si whereas others have shown a rapid disap~aran~ of bound careinogenas-34. In the case of AFBl nucleic acid binding is considerably greater than protein binding even on a per mg basis. For example, 6 h after AFBt administration to rats, for every molecule of protein reacted with carcinogen there are 400 molecules of DNA with bound carcinogen (this assumes DNA to have an average molecular weight of I - IO7 and protein of I 105). On the basis of target theory this would indicate that DNA binding must be of greater importance than protein binding for tumour initiation, particularly as the major liver protein alkylated by the activated metabolite appears to be albumin (Ketterer and Garner, manuscript in preparation). Besides the im~rtan~ of nucleic acid binding there is the further inte~sting finding that the labeiled carcinogen is removed over the time period studied in rat liver. There are a number of possible explanations for this: (I) The bound carcinogen is removed by enzymatic excision through some DNA repair process, The fact that ~pair-de~cient ~cteria are more sensitive to the microsomal activated metabolite of AFBr than wild-type bacteria indicates that excision repnir in micro-organisms is important to sensitivityts. Whether this q@ics to liver cells is not known. (2) The bound carcinogen leads to chemical depurination. If this were SO.then there should be some loss of nucleic acid bound ca~inogen in the hamsterus well as the rat unless different positions on the nucleic acid bases were dkylated. This is unlikely. (3) Cell death leads to loss of the bound carcinogen, particularly if only a few l
130 cells are highly alkylated. Although the dose of AFBl administered produces no microscopic evidence of liver necrosis it is not possible to exclude this explanation to account for the loss of carcinogen without more detailed studies, It is interesting that the half-life of loss of bound carcinogen is approximately the same as that found for #6-methylguanine, a methylatsd base thought to be important for the carcinogenic action of the chemical methylating agent@. A comparison of the amount and rate of loss of AFBs binding in the livers of the hamster and the rat shows two fundamental differences between the two species. First, that the overall level of nucleic acid binding is higher in the rat than the hamster and second, that there is a faster removal of bound carcinogen in the former species. The difference in the amount of binding probably reflects a difference in the pharmacokinetics of AFBl metabolism between the two species, there being more metabolism via the detoxification pathways in the hamster than the rat. The faster loss of nucleic acid-bound carcinogen in the rat compared with the hamster may also account for the greater sensitivity of this species to AFBr carcinogenesis. These two major differences may therefore be intimately involved in explaining why the rat is sensitive to the liver carcinogenicity of this compound while the hamster is resistant. ACKNOWLEDGEMENTS
This work was supported by a grant from the Yorkshire Council of the Cancer Research Campaign.
REFERENCES I R. W. Detroy, E. E. Lillehoj and A. Ciegler, Aflatoxin and related compounds, in Microbial Toxins, Chapter 1, Vol. 6, Academic Press, New York, 1971, pp. 3-178. 2 W. H. Butler, Aflatoxicosis in laboratory animals, in L. A. Goldblatt (Ed.), Ajlatoxin, Chapter 8, Academic Press, New York, 1969, pp. 223-234. 3 G. N. Wogan, Experimental toxicity and carcinogenicity of aflatoxins, in G. N. Wogan (Ed.), Mycotoxins in Foodstufi, M.I.T., Cambridge, Mass., 1965, pp. 163-173. 4 P. M. Newbeme, and W. H. Butler, Acute and chronic effects of aflatoxin on the liver of domestic and laboratory animals: a review, Cancer Res., 29 (1969) 236-250. 5 K. Sargeant, S. Sheridan, J. O’Kelly and R. B. A. Carnaghan, Toxicity associated with certain samples of groundnuts, N&we, 192 (1961) 1096-1097. 6 G. A. Bean, J. A. Schillinger and W. L. Klarman, Occurrence of aflatoxins and aflatotixin-producing strains of Aspergillus sop. on Soybeans, Appt. Microbial., 24 (1972) 437-439. 7 M. Sutic, J. C. Ayres and I’. E. Koehler, Identification and aflatoxin production of molds isolated from county cured hams, Appt. Microbial., 23 (1972) 656-658. 8 H. W. Schroeder and R. A. Boller, Aflatoxin production of species and strains of the AspergiUus Paws group isolated from fieldcrops, Appl. Microbial., 25 (1973) 885-889. 9 R. C. Shank, N. Bhamaropravati, J. E. Gordon and G. N. Wogan, Dietary aflatoxins and human liver cancer, IV. Incidence of primary liver cancer in two municipal populations of Thailand, Food Cosmet. Toxicol., 10 (1972) 171-179. 10 F. G. Peers and C. A. Linsell, Dietary aflatoxins and liver cancer - A population-based study in Kenya, Brit. J. Cancer, 27 (1973) 473-484. 11 R. C. Garner, E. C. Miller, J. A. Miller, J. V. Garner and R. S. Hanson, Formation of a factor lethal for S. typhimurium TA 1530 and TA 1531 on incubation of aflatoxin Bl with rat liver
microsomes,Eiochem. Biophys. Res.
Commun., 45 (1971) 774-780.
131 12
R. C. Garner, E. C. Miller and 3. A. Miller, Liver microsomal metabolism of afjatoxin & to a reactive derivative toxic to Salmon&a fyphimurium TA 1530, Cancer Res., 32 (1972) 2058-2066. 13 R. C. Garner and C. M. Wright, Induction of mutations in DNA repair-deficient bacteria by a liver microsomal metabolite of aflatoxin BI, &it. J. Cuncer, 28 (1973) 544-551. 14 R. C. Garner, Microsome-dependent binding of aflatoxin 31 to DNA, RNA, polyribonucleotides and protein in v&u, CIiem.-Biol. Interact., 6 (1973) 125-129. 15 R. C. Garner, Chemical evidence for the formation of a reactive ~atoxin I& metabolite by hamster liver microsomes, FEBS brr., 36 (1973) 261-264. 16 D. H. Swenson, J. A. Miller and E. C. Miller, 2,3-Dihydro-aflatoxin &: An acid hydrolysis product of an RNA-aflatoxin BI adduct formed by hamster and rat liver microsomes in vitro, Biochem. Biop+vs. Res. Comm., 53 (1973) 1260-1267. 17 K. M. Herrold, Aflatoxin induced lesions in Syrian hamsters, &if. J. Cancer, 23 (1969) 655-660. 18 C. C. Irving and R. A. Veazey, Isolation of d~xyr~bon~eic acid from rat liver, Bio~iiir. Biophys, Acra, 166 (1’968) 246-248.
19 G. Schmidt and S. J. Thannhauser, A method for the determination of deoxyribonucleic acid and phospho-proteins in animal tissues, J. Biol. Chem., 161 (1945) 83-89. 20 K. Burton, 4 study of the conditions and mechanism of the diphenylamine reaction for the calorimetric estimation of deoxyribonucleic acid, Biochem. J., 62 (1956) 315-323, 21 G, Ceriotti, Determination of nucleic acids in animal tissue, J. Bid. Chem., 214 (1955) 59-70. 22 H. Lowry, N. J. Ro~brough, A. L. Farr and R. J. Randall, Protein rn~su~nt with the Folin phenol reagent, J. Biot, Chem., 193 (1951} 265-272. 23 G. N. Wogan, G. S. Edwards and P. M. Newberne, Excretion and tissue distribution of radioactivity from aflatoxin B@C in rats, Cancer Res., 27 (1967) 1729-1736. 24 W. Lijinsky, K. Y. Lee and C. H. Gallagher, Interaction of aflatoxins BI and GI with tissues of the rat, Gamer Res., 30 (1970) 2280-2283. 25 M. Akao, K. Kuroda and G. N. Wogan, Afiatoxin BI: the kidney as a site of action in the mouse, Life Sci., 10 (1971) 495-501. 26 P. Brookes, Quantitative aspects of the reaction of some carcinogens with nucleic acids and the possible significance of such reactions in the process of carcinogenesis, Cancer Rex, 26 (1966) 1994-2003. 27 T. A. Lawson, The binding of o-aminoazotoluene to deoxyribonucleic acid, ribonucleic acid and protein in the C57 mouse, Biuchem. J., 109 (1968) 917-920. 28 E. Kriek, Carcinogenesis by aromatic amines, Biorhj/il. Biophys. Ada, 355 (1974) 177-203. 29 0. P. Warwick, The covalent binding of metabolites of tritiated 2-methy~~imethylaminoazobenzene to rat liver nucleic acids and proteins, and the carcinogenicity of the unlabelled compound in partially hepatectomised rats, Eur. J. C’umer, 3 (1967) 227-233. 30 C. C. Irving and R. A, Vsazey, A persistent binding of 2-acetylaminofluorene to rat liver DNA in vivo and consideration of the mechanism of binding of IV-hydroxy-2acetylaminofluorene to rat liver nucleic acids, Cancer Res., 29 (1969) 1799-1804. 31 E. Kriek, Persistent binding of a new reaction product of the carcinogen, ~hydroxy-~-2acetylaminoffuorene with guanine in rat liver in viva, Cuneer Res., 32 (1972) 2042-2048. 32 V. M. Craddock, The pattern of methylated purines formed in DNA of intact and regenerating liver of rats treated with the carcinogen dimethylnitrosamine, Biochim. Biophys. Acta, 312 (1973) 202-210. 33 M. J. Capps, P. J. O’Connor and A. W. Craig, The influence of liver regeneration on the stability of 7-methylguanine in rat liver DNA after t~atment with ~,~~imethylnitro~mine, Biochim. Bioph~~s,Actu, 33 t (1973) 3340. 34 P. J. O’Connor, M. J. Capps and A. W. Craig, Comparative studies on the hepatocarcinogen
N,N-dimethylnitrosamine 1’11 viva: Reaction sites in rat liver DNA and the significance of their J. Crrr~ce~;27 (1973) 153-166. 35 D. H. Swenson, E. C. Miller and S. A. Miller, Aflatoxin BI-2,3-oxide: evidence for its formation in rat liver it1viva and by human liver microsomes b vitro, Biochem. Biophys* Res. Commlm.+ 60 (1974) 1036-1043. relative stabilitics, &it.