Chapter 3 Metabolism, kinetics and genetic variation
Observations in humans is about 17 min in plasma and about 3 amount of urinary N-acetylcystei ne min in cells (Xu &Thornalley, 2000a). –ITC might have been underestimated, The 05H conjugates undergo fur- because these conjugates are unsta- Isothiocyanates ther enzymatic modification, in the ble and readily dissociate to isothio- Metabolism and disposition GSH portion, to give rise sequentially cyanates (Conaway et al., 2001). The The –N=C=S group of most isothio- to the cysteinyiglycine, cysteine and N- metabolic disposition of other isothio- cyanates is electrophilic and can react acetylcysteine conjugates, which are cyanates, including allyl-ITC and cul- readily with various nucleophiles, excreted in urine (Brüsewitz et al., foraphane, are similar in humans, as including thiols. Studies in both 1977), as shown in Figure 10. Urinary discussed below. humans and experimental animals excretion of N-acetylcysteine conju- have established that isothiocyanates gates is the principal route of disposi- Measurement of isothiocyanates and are metabolized in vivo to various tion of ingested isothiocyanates. For isothiocyanate metabolites dithiocarbamates principally by the example, when benzyl-ITC was admin- The cyclocondensation assay is a mercapturic acid pathway. An initial istered orally to six men, 54% of a dose highly sensitive quantitative method for conjugation with glutathione (aSH, the of 14.4 mg was recovered in the urine measuring isothiocyanates and their most abundant cellular thiol), which as N-acetyl-S-(N- ben zylthiocarba- metabolites (referred to collectively takes place spontaneously but is fur- moyl)-L-cysteine (benzyl-ITC--N-ace- here as isothiocyanate equivalent) ther promoted by glutathione S-trans- tylcysteine). This compound was (Zhang et al., 1996) and is a valuable ferase (GST), gives rise to the corre- excreted rapidly: maximum excretion tool for studying dietary consumption sponding conjugates. occurred within 2-6 h, and the com- of isothiocyanates and their metabo- In the spontaneous reaction, isoth- pound was essentially undetectable lism. The assay is based on the almost iocyanates react reversibly with cys- 10-12 h after dosing (Mennicke et al., universal ability of isothiocyanates to teinyl thiols, forming dithiocarbamates: 1988). Watercress is rich in gluconas- react quantitatively with 1,2-ben- zenodithiol, a vicinal dithiol, to give rise to a five-membered cyclic product and a free amine (Zhang et 1992a), as cystine residue—SH + RN=C=s - cystine residue -S—C —NHR al., 11 shown in Figure 11. Reactive isothio- s cyanates are converted to the same cyclic product, 1 ,3-benzodith iole-2- The equilibrium position and the chem- turtiin (constituting more than 30% of thione, and conversion is complete in ical relaxation time for equilibrium have total glucosinolates), the precursor of the presence of excess 1,2-ben- been estimated for phenethylisothio- phenethyl-ITC. When four volunteers 7enedithiol. The resultant 1,3-bon- cyanate (ITC) reacting with cysteinyl ate 30 g of watercress containing 21.6 zodithiole-2-thione can be measured residues in blood plasma (cysteinyl mg of gluconasturtiin, an average of accurately in amounts as low as a few thiol concentration, about 500 pmol/l) 47% of the amount was recovered as picomoles, with a simple high-perfor- and cells (cysteinyl thiol concentration, the N-acetylcysteine conjugate in 24-h mance liquid chromatography (HPLC) 5 mmol/l). The equilibrium constant, /Ç urine (Chung etaL, 1992). The amount procedure. Conjugation products of was 730 per mol. At equilibrium, about of phenethyl-ITC–N-acetylcysteine in isothiocyanates with thiols (dithiocar- 27% of phenethyl-ITC is bound to urine is probably greater when bamates), including N-acetyl-cys- plasma thiols in blood, and about 88% phenethyl-ITC is administered, because teine–isothiocyanate, also react quan- is bound to thiols in cells. For a plasma the my rosin ase-catalysed conversion titatively with 1 ,2-benzenedithiol, giving concentration of 5 pmol/l of phenethyl- of gluconasturtiin to phenethyl-ITC rise to the same cyclic product. Thus, ITC, the equilibrium relaxation time for may be incomplete after ingestion of this assay allows detection of the total formation of adducts with protein thiols watercress. In these studies, the amount of isothiocyanates and their
25
IARC Handbooks of Cancer Prevention Volume 9 Cruciferous vegetables, isothiocyanates and indoles
Glu-Cys-Gly Glu-Cys-Gly Cys-Gly SH GST I GGT S +
R—N C=S R—NH—C=S R—NH—C=S
lsothiocyanate Dithiocarbamate CG
NAC Cys
S s
R—NH—C---S
Mercapturic acid
Figure 10 Isothiocyanates are conjugated to glutathione by glutathione S-transferase (GST), metabolized sequentially by y-glutamyl transpeptidase (GGT), cysteinyiglycinase (CG) and N-acetyltransferase (AT) to form, ultimately, mercapturic acid. NAC, N-acetylcysteine
R N=C=S R-NH2 Isothiocyanate
SH
+ LJL>=S
1 ,3-Benzodithiole-2-thione 1 ,2-Benzenedithiol
S / R—NH--C R—NH2 + R1—SH S—R,
Dithiocarbamate
Figure 11 Cyclocondensation reaction of isothiocyanate and dithiocarbatrate with 1,2-ben7enedithiol
thiol conjugates. The assay has isothiocyanates. Even though individ- Use of the cyclocondensation allowed quantitative assessment of ual isothiocyanates and isothiocyanate assay is nevertheless subject to some total isothiocyanates in cruciferous metabolites can be quantified sensi- caveats. Although there are no endo- vegetables and isothiocyanate equiva- tively by other methods, including genous sources of urinary isothio- lents in human fluids, including blood HPLC-mass spectrometry (Ji & cyanates or dithiocarbamates in humans and urine. Because it reveals both Morris, 2003; Vermeulen et al., 2003), (Shapiro et cL, 1998), non-dietary isothiocyanates and their thiol metabo- such analyses might not provide accu- sources may exist. The contributions of lites, the total urinary concentrations of rate values, since the metabolites are cigarette smoke (possibly due to the these compounds are not affected by unstable and can dissociate to isothio- presence of carbon disulfide, which is dissociation of thiol conjugates to cyanates. also detected in the assay), agricul-
26 Metabolism, kinetics and genetic variation tural chemicals, chemicals used in the equivalent was recovered in urine rect conclusions about long-term rubber industry and medical com- within 10h (Shapiro of al., 1998). Thus, dietary intake of isothiocyanates. pounds such as disulfiram and anti- isothiocyanates are rapidly and effi- Cellular uptake of isothiocyanates septics, should be considered (Zhang ciently absorbed, rapidly cleared from is enhanced by GST isozymes (Zhang, etaL, 1996; Shapiro of al., 1998; Ye of blood and eliminated almost exclu- 2001); however, export of intracellular aL, 2002). Shapiro of al. (1998) found sively in urine. isothiocyanate—GSH conjugates is that when urine samples were stored The molecular basis for the rapid also highly efficient without GST at —20 °C, the cyclocondensation reac- metabolic disposition of isothio- involvement (Zhang & Calloway, 2002). tivity remained stable, but 35% of the cyanates in humans was elucidated by While GSTs are likely to be important dithiocarbamate content was lost in studying cultured human cells with the in isothiocyanate metabolism, current samples stored for 18 months at the cyclocondensation assay (Zhang & understanding does not allow a predic- same temperature. Talalay, 1998; Zhang, 2000, 2001; tion of the consequences of GST poly- Zhang & Callaway, 2002). Exposure of morphism on exposure of tissues or Measuring metabolic disposition of cells to an isothiocyanate led to rapid the rates and extent of urinary excre- isothiocyanates accumulation of the compound tion of isothiocyanates and isothio- Studies on dietary intake of isothio- through conjugation with cellular thiols. cyanate conjugates. cyanates are still extremely limited, f3SH, the most abundant intracellular and the information below was thiol, is the major driving force for Bioavailability of isothiocyanates obtained from studies of a small num- isothiocyanate accumulation, and cel- from cru ciferous vegetables ber of volunteers. Ye et al. (2002) deter- lular GST enhances isothiocyanate In a study designed to compare the mined the metabolic disposition of accumulation by promoting conjuga- bioavai ability of isothiocyanate in isothiocyanates from broccoli sprouts tion reactions. Peak intracellular isoth- fresh and steamed broccoli, 12 men by giving four volunteers a single serv- iocyanate accumulation was achieved were asked to eat 200 g of fresh or ing of a myrosinase-treated extract of within 0.5-3 h after the beginning of steamed broccoli, and urine samples 3-day-old broccoli sprouts containing exposure to isothiocyanate, reaching were collected during the subsequent 200 pmol total isothiocyanate. The 100-200 times the extracellular isoth- 24 h. The total content of isothio- isothiocyanate composition of the iocyanate concentration. Intracell ularly cyanate in fresh and steamed broccoli preparation was 77.2% sulforaphane accumulated isothiocyanate and con- after treatment with myrosinase was or iberin and 22.8% iberin (aliphatic jugates were also rapidly exported by virtually identical (1.1 and 1.0 pmol/g isothiocyanates with closely related membrane transporters, including the wet weight); however, the average 24- chemical structures). The isothio- multidrug resistance-associated pro- h urinary excretion of isothiocyanate cyanates were absorbed rapidly and tein (MRP)-1 and P-glycoprotein-1: the equivalent was 32.3 ± 12.7% and 10.2 reached peak plasma concentrations half-life of the accumulated isothio- ± 5.9% of the amount ingested in fresh of 0.94-2.27 pmol/l 1 h after ingestion, cyanate and conjugates in human and steamed broccoli, respectively. which declined according to first-order prostate cancer LNCaP cells was only Thus, the bioavailability of isothio- kinetics (half-Fife, 1.77 ± 0.13 h). The about 1 h. cyanates from fresh broccoli was cumulative urinary excretion of isothio- These studies have begun to approximately three times greater cyanate equivalent at 8 h was 58.3 ± reveal the pharmacokinetics of than that from steamed broccoli, in 2.8% of the dose. In a similar experi- ingested isothiocyanates in humans; which myrosinase was inactivated by ment, the cumulative urinary excretion however, the question remains as to heat (Conaway of al., 2000). In of isothiocyanate equivalent at 72 h when urine should be collected for another study (Getahun & Chung, was 88.9 ± 5.5% (Shapiro etal., 2001). assessment of dietary isothiocyanate 1999), 350 g of watercress (containing Similar results were obtained in volun- intake. As dithiocarbamates are 475 pmol total glucosinolates) was teers given horseradish isothio- excreted rapidly in urine after ingestion cooked in boiling water for 3 min to cyanates (Shapiro of al., 1998), most of isothiocyanates, continuous urine inactivate myrosinase and then eaten of which are allyl- or phenylethyl-ITC collection over a sufficient period (e.g. by nine volunteers. The 24-h urine (Liebes of al., 2001; Ji & Morris, 2003). 8 or 24 h) after isothiocyanate inges- samples showed a total urinary excre- For example, after a single serving of tion might be important for comparing tion of isothiocyanate equivalent rang- 20 ml of horseradish juice containing urinary total isothiocyanate equivalents ing from 5.6 to 34.8 p.mol, correspond- 74 pmol of isothiocyanates to 10 vol- among individuals. Analysis of random ing to 1.2-7.3% of the total glucosino- unteers, 42 ± 5% of isothiocyanate spot urine samples might lead to incor- lates ingested. In contrast, ingestion of
27 IARC Handbooks of Cancer Prevention Volume 9: Cruciferous vegetables, isothiocyanates and indoles
150 g of uncooked watercress resulted cyanates was consistent between vol- tionnaire on vegetable and fruit intake in excretion of 172-77.7% of the total unteers in different studies and that the and collected a 24-h urine sample ingested glucosinolates in 24-h urine. urinary concentration of dithiocarba- (Fowke et al., 2001), the urinary isoth- Clearly, myrosinase in the original veg- mates was proportional to an escalat- iocyanate correlated well with the 24-h etable contributed to the release of ing dose regimen (r2 = 0.976). The uri- consumption of cruciferous vegetables isothiocyanates from glucosinolates. It nary concentrations peaked within < 8 (Pearson correlation coefficient, 0.57; can also be inferred from these results h of consumption, but complete excre- p < 0.01) and with an 'unknown but that human myrosinase, known to exist tion took 24-72 h. In studies in which true intake' calculated from a struc- in the intestinal flora (Shapiro et al., volunteers were given glucosinolates tural equation (Ocké & Kaaks, 1997). 1998; Getahun & Chung, 1999; Krul et and isothiocyanates from broccoli Single-void urine samples are clearly aL, 2002), might hydrolyse only a frac- sprouts, measurements of urinary less reliable markers than urine col- tion of the glucosinolates ingested and isothiocyanate equivalent with this lected over 8 or 24 h (Scow et al., might also vary widely in activity assay were highly reproducible (coeffi- 1998; London et al., 2000). The former among individuals. Nevertheless, cient of variation, < 10%), and their are useful mainly for assigning partici- when enteric flora were reduced by a concentrations accurately reflected the pants to reasonable categories of combination of mechanical cleansing absorption, metabolism and excretion intake (e.g, negative, positive, low, and oral antibiotics after ingestion of of the isothiocyanate consumed. A high), while the latter, especially 100 iimol of a broccoli glucosinolate strict linear relationship was observed repeated 8- or 24-h urine collection, preparation in which myrosinase had between isothiocyanate dose and allow realistic assessments of con- been inactivated, a dramatic reduction isothiocyanate equivalent in 72-h urine sumption of cruciferous vegetables or was seen in 72-h urinary excretion of over an eightfold range of isothio- isothiocyanates. isothiocyanate metabolites, falling from cyanate dose (25-200 pmol) (Shapiro 11.3 ± 3.10% of the dose before treat- et al., 2001), with no evidence of a lndoles ment to 1.3 ± 13% after treatment threshold. Fowke and colleagues Metabolism and disposition (Shapiro et al., 1998). Not surprisingly, (2001) reported that an acceptable The mildly acid environment of the the extent of glucosinolate conversion dose—response relationship between stomach induces chemical modifica- to isothiocyanates by the myrosinase intake of cruciferous vegetables and tion of indole-3-carbinol, which is in vegetables is also affected by the urinary dithiocarbamate was found dehydrated in acidic solutions and is length of time the vegetable is chewed. only at intake levels of 100 200 g/day. converted to its active derivates In a study in which four volunteers The intake levels were achieved in a (Grose & Bjeldanes, 1992). Thus, were each given 12 g of fresh broccoli planned dietary intervention and were indole-3-carbinol administered intra- sprouts containing 109 i.tmol total glu- much higher than the 0.2 serving per peritoneally does not exert metabolic cosinolates and were asked to either day (one serving one cup of raw change in the host (Bradfield & swallow without chewing or chew thor- leafy or one-half cup of cooked or Bjeldanes, 1987a). The main products oughly before swallowing, thorough chopped raw) estimated to be the aver- of this chemical reaction are 3,3'-df n- chewing resulted in significantly age intake of cruciforous vegetables by dolylmethane, triindolylmethane and greater excretion of isothiocyanate residents of the USA in 1994-96 indolo[3,2-b]carbazole (Figure 12). equivalent (42.4.t: 7.5 pmol and 28.8 ± (Johnston et al., 2000). Many other indole-3-carbinol-derived 2.6 pmol, respectively) in 24-h urine In an Asian population (n = 246) compounds were detected in experi- (Shapiro et al., 2001). who frequently ate cruciferous vegeta- mental animals exposed to this com- bles (mean daily intake of cooked, 40.6 pound (Stresser et al., 1995a; Assessing consumption of crucifer- g; mean daily isothiocyanate intake, Anderton et al., 2003). ous vegetables by measuring isoth- 9.1 pmol), concentrations of total isoth- Most of the information derives locyanate equivalents in urine iocyanate in spot urine samples were from studies in vitro in which conden- Controlled metabolic studies have significantly associated with crucifer- sation products were collected without shown that urinary isothiocyanate ous vegetable intake and isothio- ascorbic acid. When glucobrassicin equivalent assayed by the cyclocon- cyanate intake, estimated from a semi- and neoglucobrassicin are degraded derisation reaction reflects the amount quantitative food-frequency question- in the presence of ascorbic acid, vari- of cruciferous vegetables eaten. naire (Seow et al., 1998). In 34 post- ous ascorbigens are formed. It has Shapiro and colleagues (1998) found menopausal women who provided 24- been estimated that 20-60% of indolyl that the disposition of ingested isothio- h dietary recalls, responded to a ques- glucosinolates are converted to
28 Metabolism, kinetics and genetic variation
CH,
3,3'-D ii n dolylm etha ne
Ind olo(3 ,2-b)carbazo I e
C H 20H jj 2NH
IndoIe-3-çarbino!
NH
[2- (Ind ol-3-ylme thy I)indoI3-yI]indol-3-ylmethiane
5,6,11,12,17,18 -Hexa h yd rocyc I on on a - (1 ,2-b:4,5-b':7,8-b)triindole
Figure 12 Ind6e-3-carbinol and its by-products ascorbigens, depending on the pH carbinol, and it is the most stable, with of higher-order oligomers (Grose & (Buskov et al., 2000; Hrncirik et aL, the longest biological lite (Broadbent & Bjeldanes, 1992). In vitro at physiolog- 2001). Broadbent, 1998a,b). pH affects the ical pH, only 3,3'-diindolylmethane Asoorbigen is unstable in acidic oligomerization process: only 5% of was detected after 48 h (Niwa et al., media (as in the stomach), giving rise indole-3-carbinol was converted to 1994). In contrast, Staub et ai. (2002) to indolo[3,2-b]carbazole. About 20 3,3'-diindolylmethane at neutral pH found that indole-3-carbinol was stable times more indolo[3,2-b]carbazole was (Amat-Guerri et al., 1984), but 80% in cell-free medium or in cultured MCF- formed in vitro after incubation with was transformed at pH 4.5 (de Kruif et 7 cells, with a half-life of about 40 h. ascorbigen than after incubation with al., 1991). As the pH falls, the produc- 3,3'-Diindolylmethane can be indole-3-carbinol under identical condi- tion of indolo[3,2-b]carbazole and extracted from the urine of volunteers tions (Preobrazhenskaya etal., 1993a). larger oligomers increases over that of given indole-3-carbinol.This assay has 3,3-DiindoIylmethane is the most 3,3 '-diindolylmethane. Dilution of been used to estimate indole-3- common of derivative of indole-3- indole-3-oarbinol favours the formation carbinol intake, in order to correlate it
29 IARC Handbooks of Cancer Prevention Volume 9: Cruciferous vegetables, isothiocyanates and indoles with the severity of cervical dysplasia. intake (Morse et al., 1990a;Taioli et al., Modulation of phase I and Il In one volunteer given 150 mg of the 1997). Thirteeen healthy smokers enzymes' indole, from whom urine was collected treated with 400 mg of indole-3- The effects of cruciferous vegetables over the next 18 h, peak urinary excre- carbinol for 5 days showed a change in on phase I and Il enzymes have been tion was observed after 7 h. In the the urinary levels of two metabolites of studied extensively (see also the same study, 10 women were treated 4-(methylnitrosamino)-1 -(3-pyridyl)-1- section on intermediary biomarkers). with indole-3-carbinol for 4 weeks. In butanone (NNK): 4-(methylnitro- Some of the studies addressed the five treated with 200 mg/day, the mean samino)-1 -(3-pyridy-1 -butanol and its effects of the vegetables themselves, urinary concentration of 3,3-diindolyl- glucuronide. These changes suggest while others focused on the effects of methane was 12.1 ± 2.5 pg/mg creati- that indole-3-carbinol increases NNK specific compounds in vegetables nine; in the remaining five women, metabolism in smokers. It may also (e.g., glucosinolates, precursors of treated with 400 mg/day, the concen- affect the metabolism of several other isothiocyanates and indoles). The veg- tration was 15.6 ± 22.2 pg/mg creati- substances, including aflatoxins, alco- etables studied were usually Brussels nine (Sepkovic et aI., 2001). 3,3'- hol and certain drugs (Fong et al., sprouts, cabbage, broccoli and cauli- Di indolyl methane was also measured 1990; Chung et ai., 1993; Manson et flower. The diets were often prepared in plasma and blood of four women al., 1997; Oganesian et al., 1999). by mixing lyophilized vegetables into treated with 400 mg of indole-3- In a study of structure—activity rela- the regular diet at concentrations of carbinol orally. The concentrations in tionships of 3,3-diindoiylmethane ana- 2.5% up to 50% (wlw). in some stud- plasma and serum were 0.1-0.4 logues in human cells, substitutions at ies, vegetables extracts were mixed pg/mi. No indole-3-carbinol was recov- the 5 and 5' positions induced the into the regular diet after evaporation ered in blood, confirming that it under- activity of both aromatase and of the solvent, which was usually goes oligomerization in the stomach ethoxyresorufin O-deethylase (EROD) water, methanol or ethanol. Regard- (Arneson et al., 1999). activity, whereas analogues with a sub- less of the study protocol and diet When indolo[3,2-b]carbazole was stitution in the bridging methylene car- preparation, most of these studies measured in two samples of faeces of bon induced neither enzyme showed that feeding cruciferous volunteers eating controlled diets, both (Sanderson et al., 2001) vegetables to animals generally samples showed a chromatographic Dimethyl 3,3'-diindolylmethane induces phases I and II enzymes in the peak in the range 2-20 pg/kg (dry and tetramethyl 3,3 '-diindolyl methane liver and intestinal tract. weight, w/w) (Kwon et al., 1994). derivatives did not induce bacterial Many biologically active com- The presence of other metabolites chloramphenicol acetyltransferase pounds in cruciferous vegetables are in human blood, urine or faeces has (CAT) reporter gene activity in T47D likely to contribute to the effects on not been tested. cell lines, and no aryl hydrocarbon (Ah) phase I and Il activities, and the receptor antagonist activity was increased enzyme activities reflect a Induction and inhibition of metabo- observed; however, no comparison combined effect of these active com- lizing enzymes was made with the induction ability of pounds. Isothiocyanates and indoles Indoles modulate the activity of genes 3,3 '-diindolylmethane under these are likely to have opposite effects on involved in the metabotsm of both conditions (McDougal et al., 2001). phase I enzymes, as indoles are endogenous and exogenous compounds strong inducers of cytochrome P450 (Nho & Jeffery, 2001). Differences in Experimental studies enzymes, whereas isothiocyanates dietary habits, in conjunction with meta- usually inhibit these enzymes (see bolic gene polymorphisms, may partially Cruciferous vegetables below). Indoles also inhibit cytochrome explain ethnic differences in the ratio of Metabolism and disposition P450 activity, however (Stresser et al., 2-hydroxy-: 1 6-hydroxyestradiol in Few studies have been reported on the 1995b; Takahashi et al., 1995a). The women's urine (Taioli et al., 1996). metabolism of isothiocyanates and induction of phase Il enzymes such as The metabolism of tobacco can indoles in experimental animals fed GST by cruciferous vegetables is prob- also be influenced by dietary indole cruciferous vegetables. ably due to the concerted action of
'Usually, xenobiotic metabolizing enzymes are subdivided by toxicologists into phase I and phase il enzymes on the basis of whether they catalyse oxidation—reduction or conjugation reactions. Thus, NAD(P)H:quinone oxidoreductase 1 is sometimes classified as a phase I enzyme. In this Handbook, however, it is referred to as a phase II enzyme becuase it is coordinately regulated with xenobiotic conju- gating enzymes through an antioxidant response element.
30 Metabolism, kinetics and genetic variation
isothiocyanates and indoles. In all Bradfield and Bjeldanes (1984) fed observed after feeding allyl-ITC, which these studies, the contribution of isoth- rats a diet containing 25% (w/w) caused stronger enhancement of sub- iocyanates was difficult to assess Brussels sprouts for 10 days and com- unit 2. When the average sinigrin con- quantitatively because of the presence pared the effects on intestinal and centration in the sprouts was as low as of other active compounds in the veg- hepatic GST activity with those of a that of progoitrin (about 540 pmol/kg), etables (Loub et ai., 1975; McDanell et diet containing indole-3-carbinol at however, a goitrin-like induction pat- al., 1989). Nevertheless, there is little 50-500 mg/kg. The Brussels sprouts tern was observed. The authors con- doubt that the observed effects are diet increased GST activity by 1.9-fold cluded that at least two compounds due in part to isothiocyanates, as in the gut and by 1.6-fold in the liver, (probably allyl-ITC and goitrin) are shown in studies with isothiocyanate but neither activity was significantly responsible for the induction of OSTs as the only agent (see below). affected by purified indole-3-carbinol, in rat liver and small intestine by Some reports have emphasized even at the highest dose. The diet Brussels sprouts. both the complexity of the phase I containing sprouts significantly Thus, much of the induction of response to Brass/ca vegetables in rat increased the activities of Ah hydroxy- phase I and phase Il enzymes by models and the fact that it depends lase (by 3.6-fold) and ethoxycoumarin vegetables may be due to indoles. The strongly on differences in the chemical Q-deethylase (by 3.2-fold). The same vegetables studied, Brussels sprouts, composition of vegetables, yang et al. group (Bradfield et al., 1985) included cabbage, broccoli and cauliflower, (1991) studied the effects of broccoli hepatic GST in a study of the effects of have high concentrations of indole glu- on CYP1A1 and CYPIIB mRNA and 12 vegetables on phase I and phase Il cosinolates (glucobrassicin) and proteins in rats fed freeze-dried activity in mice. Diets containing 20% relatively little isothiocyanate glucosi- vegetables for 7 days. In the colon, freeze-dried, powdered Brussels nolates (van Etten & Tookey, 1979; both CYP1A1 mRNA and protein were sprouts or cauliflower significantly Fenwick & Heaney, 1983). lsothio- induced by the broccoli diet. CYPFA2 increased the activities of GST (by 2.0- cyanate-rich vegetables with a small protein was present in the colon, but it and 1.2-fold, respectively) and epoxide amount of indoles might reduce phase was not altered by the diet, and the hydratase (both by 1.6-fold): the activ- I enzyme activity, as was observed in mRNA was not detectable. Paradoxi- ity of ethoxycoumarin O-deethylase studies with pure isothiocyanate cally, CYPIIB mRNA was reduced, but was increased significantly (by 2.2- compounds (see below). The prepara- the protein was increased. In the liver, fold) by cauliflower but that of Ah tion of the vegetable diets for this type CYPIHB and CYPIIE1 proteins were hydroxylase was not affected. The of experiment is an important consid- increased by the broccoli diet, but effects on enzyme activities were not eration, as it might considerably alter CYPIIB mRNA was not affected. In a confined to Brass/ca vegetables, the bioavailability of the hydrolysed later study in which different varieties although the high doses used make products. The free isothiocyanates in of broccoli were fed to rats, it was these results difficult to interpret. freeze-dried broccoli samples are less also shown that modulation of Bogaards et al. (1990) studied the bioavailable, as estimated from the cytochrome P450 and other phase I effects of dietary supplements of mercapturic metabolite excreted in enzymes is critically dependent on the Brussels sprouts (2.5-30%), the sini- urine, and less effective in inducing concentrations of glucosinolates grin breakdown product ally]-ITC (0.03 quinone reductase activity in colon than and other biologically active phyto- and 0.1%) and goitrin (0.02%) on the freeze-dried broccoli samples contain- chemicals in plant tissue (yang et al., GST subunit pattern in the liver and ing intact glucosinolates (Keck et al., 2001). Various strategies have been small intestinal mucosa of male rats. A 2003). Equally important is the method used to overcome the problems due statistically significant linear relation- of storage, which could have significant to the complexity of Brass/ca vegeta- ship was found between the amount of effects on the stability of compounds bles. Sorensen et al. (2001) fed rats a Brussels sprouts in the diet and induc- such as isothiocyanates. Brussels sprout extract containing tion of GST, with similar increases in A question of practical importance a complex, incompletely characterized the total amounts of GST subunits. is whether cooked vegetables affect mixture of glucosinolates and their When the average concentrations of phase H and 11 enzymes. Cooked vege- breakdown products for 4 days. sinigrin and progoitrin in the sprouts tables are devoid of myrosinase, the No significant modulation of phase I were 1835 pmol/kg and 415 pmol!kg, enzyme responsible for hydrolysing enzymes but significant up-regulation respectively, the subunit induction pat- glucosinolates to release bioactive of phase II enzymes was observed. terns in the liver and the small intesti- aglucones. In a study in which Wistar nal mucosa were similar to those rats were fed a diet supplemented with
31 IARC Handbooks of Cancer Prevention Volume 9: Cruciferous vegetables, isothiocyanates and indoles cooked Brussels sprouts at 2.5-20% while suiforaphane statistically signifi- excreted 50-55%. The rats had (w/w), various phase I and II enzymes cantly reduced both enzyme activities retained 20-25% of the dose in the were measured, including CYP1A1, (Matusheski & Jeffery, 2001). In a sub- carcass after 96 h. Faecal 14C CYP1A2 and CYP2B, GST, UDP-giu- sequent study, groups of three male accounted for 6-12% and expired 002 curonosyl transferases (UGTs) and Fischer 344 rats [age not specified] for 4-7% of the dose. In this study, NAD(P)H-quinone reductase (Wortel- were given the rapide seed nitrile 67-85% of the 140 in rat urine samples boer et al., 1992a). Almost all the crambene or suiforaphane at a dose of was identified as aliyi-1TC--N-acetyi- enzymes were induced by the diet con- 50 mg/kg bw by gavage in corn oil for cysteine by HPLC and positive-ion taining Brussels sprouts, and effects 7 days (515 pmol/kg bw for crambene, electrospray mass spectroscopy. Three were seen as soon as 2 days after 282 pmol/kg bw for suiforaphane) metabolites were found in mice: feeding. Although no direct compari- (Keck et al., 2002). Controls were 48-85% of the urinary 140 was on son was made of the effects on these given corn oil only. Hepatic cytosolic thiocyanate, 7-52% on aHyi-1TC- enzymes of cooked and uncooked quinone reductase activity was cysteine (Cys) conjugate and 8-12% vegetables, they appeared to have induced 1.5-fold by crambene and 1.7- on aHyi-ITC-N-acetyicysteine (males similar effects. This suggests that glu- fold by suiforaphane. only). in B6C3F1 mice, aiIyi-1TC-N- cosinolates are hydrolysed to agly- acetyicystei ne represented less than cones after ingestion of Brussels 20% of total urinary radioactivity; three sprouts by a myrosinase-like microflo- Isothiocyanates other major and two minor unidentified rai activity in the rat intestinal tract. In a Metabolism and disposition urinary metabolites were also detected study of the influence of the ntestinal The metabolism and tissue disposition (loannou et al., 1984). in both rats and microflora and dietary glucosinolates of several isothiocyanates have been mice, 70-85% of the administered on hepatic cytochrome P450 enzymes investigated extensively in rodents. dose was collected in the urine by (Nugon-Baudon et al., 1998), conven- These studies showed that the mer- 72 h, while 13-15% was trapped as tional rats and germ-free rats were fed capturic acid pathway is the main route CO2 (in rats only), and 3-6% of the a diet containing myrosinase-free of metabolism of isothiocyanates, 140, consisting of a single unidentified rapeseed. While the effects of the glu- involving initial conjugation with GSH metabolite, was found in faeces. The cosinolate-rich diet on these enzymes mediated by GSTs, followed by enzy- urinary bladders of animals given appeared to be complex, the rapeseed matic degradation to its N-acetyicys- [14C]ally[-ITC showed sex and species meal decreased total cytochrome teine conjugate (Meyer et al., 1995; differences, more 140 being found in P450 activity in conventional rats but Whalen & Boyer, 1998). Other, minor the bladder tissue of male rats. The not in germ-free rats, suggesting that metabolites of isothiocyanates with an amount of 140 detected in bile from the microflora present in the gut play a alkyl or aryl moiety have also been cannulated rats was greater than that role. identified in animals. This section sum- in faeces, indicating possible entero- The ability of nitriles to induce marizes the data from these studies hepatic circulation of metabolites and phase Il enzymes was investigated in (Table 8). eventual urinary excretion (Borghoff & two studies. In a comparison of the Birnbaum, 1986; Bollard et al., 1997). effects of the major broccoli isothio- Biliary metabolites of allyl-ITC from cyanates and nitrile, groups of five Allyl isothiocyanate Fischer 344 rats given [U-14C]aHyi-ITC 4-week-old male Fischer 344 rats were In Wistar rats, the main urinary by gavage were similar to urinary given 5-(methylsu lfinyl)pentane nitrile metabolite of aHyi-ITC was the mer- metabolites, as identified by HPLC, but in saline at 200, 500 or 1000 pmol/kg capturic acid N-acetyi-S-(N-aflyfthio- the relative proportions differed consid- bw, sulforaphane in saline at 500 carbamoyl)-L-cysteine (aHyi-1TC-N- erably (loannou et al., 1984). pmol/kg bw or saline, by gavage for acetyicysteine) (Mennicke et al., 5 days. Controls and test animals given 1983). Studies of pharmacokinetics Benzyl isothiocyanate nitrile were pair-fed with semi-synthetic and metabolism were conducted after Brüsewitz et al. (1977) identified the food throughout the study. The animals oral administration of [14C]aHyi-ITC mercapturic acid of benzyi-FTC (ben- were killed 24 h after the last dose, and labelled in the isothiocyanate moiety to zyi-ITC--N-acetyicysteine) as its the activities of quinone reductase and Fischer 344 rats and B6C3F1 mice dicyciohexylam me salt after adminis- GST were determined in cytosolic frac- (Bollard et al., 1997). Within 96 h, male tration of benzyi-ITC and various thiol tions from the liver, pancreas and and female mice had excreted -80% of conjugates orally or by intraperitoneal colon mucosa. The nitrile had no effect, the 140 in urine, and the rats had or intravenous injection to rats. The
32 Metabolism, kinetics and genetic variation
Table B. Urinary metabolites in experimental animals after administration of isothiocyanates (ITCs)
Isothiocyanate Route Main metabolite(s) Administered Strain and Reference dose recovered (%) species
Allyl-ITC Oral Ally -ITC-NAC 40-50 Wistar rat Mennicke et al. (1983) Ally ITC NAC 76-82 Fischer 344 rat loannou eta). (1984) Ally -ITC-NAC 8-20 B6C3F1 mouse loarinou eta]. (1984) Ally -ITC-NAC 67-85 Fischer 344 rat Bollard et al. (1997) Ally -ITC-NAC 8-12 B6C3F 1mouse Bollard et al. (1997) Ally ITC Cys 7 52 B6C3F1 mouse Bollard et al. (1997) -SCN-- 15-33 Fischer 344 rat Bollard et al. (1997) -SON- 48-85 B6C3F mouse Bollard et al. (1 997) Ally-ITC 0.3-0.4 Fischer 344 rat loannou eta). (1984) Ally -ITC 1.9 B6C3F1 mouse loannou et at. (1984)
Butyl-ITC Oral Butyl-ITC-NAC 10-20 Wistar rat Mennicke et al. (1983)
Benzyl-ITC Oral Benzyl ITO NAC Only metabolite Wistar rat BrOsewitz et al. (1977) Benzyl-ITC-NAC Trace Guinea-pig GorIer et a). (1982) 4H4CBTT 23 ± 3 Guinea-pig GOrIer et cl. (1982)
Benzyl-ITC-Cys Oral Benzyl-ITC-NAC 62 W star rat BrOsewitz et al. (1977) Benzyl-ITC < 1 Wstar rat BrOsewitz et al. (1977) Benzyl-ITC Cys < 1 w star rat BrOsewitz et al. (1977) Hippuric acid 40 Beagle dog BrOsewitz etal. (t 977) 4H4CBTT 33 ± 4 Guinea-pig GorIer et cl. (1982)
Benzyl-ITC-NAC Oral Benzyl-ITC-NAC Trace Guinea-pig GOrIer et al. (1982) 4H4CBTT 4-9 Guinea-pig GOrier et cl. (1982)
Phenethyl-ITC Oral Phenethyl-ITC-NAC 9-10 A1J mouse Eklind etal. (1990) 4H4CFETI 25 NJ mouse Eklind et cl. (1990) Phen ethyl-ITC-NAC > 90 Fischer 344 rat Conaway et al. (1999) Phenethyl-ITC < t Fischer 344 rat Conaway et al. (1999)
Gluconasturtiin Diet Phen ethyl- ITC-NAC 22+8 NJ mouse Chung et at. (1992)
6-Phenylhexyl-ITC Oral Unidentified 7 ± 1 Fischer 344 rat Conaway et al. (1999) 6-Phenylbexyl-ITC-NAC Tracea Fischer 344 rat Conaway et aI. (1999)
Sulforaphane Intraperitoneal Suiforaphane-NAC - 60 Sprague-Dawley rat Kassahun eta]. (1997) Erucin-NAC - 12 Sprague-Dawey rat Kassahun et al. (1997)
Erucin Intraperitoneal Sulforaphane-NAC - 67 Sprague-Daw ey rat Kassahun et al. (1997) Erucin-NAC -29 Sprague-Dawey rat Kassahun et al. (1997)
NAC, N-acetylcysteine; Cys, cysteine; 4H4CBTT, 4-hydroxy-4-cardoxy-3- ben zyJthiazo[dine-2-thione 4H4CPETT, 4-hydroxy-4-cardoxy- 3-phenylethylthiazolidine-2-thione Most metabolites found in faeces
thiol conjugates studied were GSH tion of benzyl-ITC was studied [14C]benzyl-ITC-Cys was adminis- (benzyl-ITC-GSH), cysteinylglycine with 14C-labelled compound synthe- tered orally to fasted male (benzyl-ITC-Cys-Gly) and cysteino sized by radiolabelling the carbon adja- and female rats, the mean peak (benzyl-ITC Cys). The tissue disposi- cent to the isothiocyanate group. When plasma concentration of 140 occurred
33 IARC Handbooks of Cancer Prevention Volume 9 Cruciferous vegetables, isothiocyanates and indoles within 45 min. The plasma radioactivity Figure 1313). The mechanism for its in urine and 13.2% in faeces. Rather declined rapidly, with a half-time of 1-2 formation is thought to involve the cys- than the mercapturic acid metabolite, h. A total of 92.4% of the 14C was teine conjugate, which is transaminat- [14C]hippuric acid (40% of the dose) collected in urine, while 5.6% ed to the S-substituted mercaptopyru- was identified (see Figure 13C). appeared in faeces, and 0.4% was vate, followed by enolysis and cycliza- Moreover, when benzyl-ITC-[ 5S]Cys detected in expired air over 3 days. tion to form the cyclic metabolite was administered, the radiolabelled Metabolites in urine were analysed by instead of the expected N-acetylated metabolite was not recovered in urine, mass spectrometry after isolation by product. Benzyl- ITC-N-acetylcysteine indicating that the conjugate had lost thin-layer chromatography, and most of was, however, also identified as a the cysteine moiety by dissociation. the recovered 14C in the urine was minor metabolite. Similar results were Components in dog urine with reten- identified as the mercapturic acid observed after administration of ben- tion times on thin-layer chromatogra- (see Figure 13A). Free [14C]benzyl-ITC zyl-ITC or benzyl-ITC-Cys to rabbits. phy corresponding to the unchanged or [14C]benzyl-ITC-Cys accoun-ted for Thus, guinea-pigs and rabbits excrete cysteine conjugate, free benzyl-ITC or less than 1% of the administered dose little mercapturic acid after isothio- other metabolites represented less in urine. No other metabolite was iden- cyanate metabolism and apparently than 5% of the administered dose. The tified in rat urine after oral administra- do not readily form mercapturic acids fate of the free [11S]Cys formed was tion of benzyl-ITC or its thiol conju- after receiving other types of xenobiotic. not reported. gates. Small amounts of free benzyl- In beagle dogs given a single dose In summary, rats metabolized ben- ITC were recovered in urine after of ["Cbenzyl-ITC-Cys by gavage, zyl-ITC via the mercapturic acid path- administration of the conjugates by the absorption occurred more slowly than way, and the major urinary metabolite other routes, those presumably having in rats, with a peak mean plasma formed was ben zyl -ITC-N-acetylcys- been formed by dissociation of the concentration of 14G within 1.5-6 h teine. Guinea-pigs and rabbits conjugates in vivo. When the [35S]Cys (Brûsewitz et al., 1977). After 3 days, excreted primarily the cyclic mercap- conjugate of benzyl-ITC was adminis- 86.3% of the dose had been excreted topyruvate conjugate 4-hydroxy-4-car- tered orally to the rats or unlabelled benzyl-!TC-Cys was given to rats pre- H H S H A treated with [35S]Cys to label the body sulfur pool, [35S]mercapturic acid was recovered in the urine. Excretion of H HCCH benzyl-ITC-[35S]N-acetylcysteine after 11 dosing with benzyl-ITC-[35S]Cys indi- O cates that the conjugate was at least partially absorbed unchanged, subse- s quently acetylated and then excreted. H c-_s 8 / \ The excretion of benzyl-ITC-[35SN- acetylcysteine after administration of H /\ unlabelled benzyl-ITC-Cys and HO COOH -[35S]Cys suggests that the benzyl- ITC-Cys was deconjugated and then reconjugated with [35S]Cys or GSI-1. The metabolism of benzyl-ITC in C other species differs from that in the rat. When male guinea-pigs were given a single dose of benzyl-ITC or benzyl-ITC-Cys by gavage, 23.3 ± 2.8% and 33.1 ± 4.0 %, respectively, Figure 13 Variations in urinary metabolites of benzyl-isothiocyanate (ITC) and benzyl-ITC-cysteine of the dose excreted in 24-h urine was A) N-Acetyl-S-(N-benzylthiocarbamoy-L-cysteine (benzyl-ITC-N-acetylcysteine), the mercapturic identified as the cyclic 4-hydroxy-4- acid of benzyl-ITC carboxy-3-benzylthiazol idi ne-2-thione, B) 4-Hydroxy-4-carboxy-3-benzothiazolidine-2-tbicne, major benzyl-ITC--cysteine metabolite in a mercaptopyruvic acid conjugate of urine of guinea-pigs and rabbits (Gorier et al., 1982) benzyl-ITC (Grler et al., 1982; see C)Hippuric acid, major metabolite in urine of beagle dogs (Brüsewitz et al., 1977)
34 Metabolism, kinetics and genetic variation boxy-3-benzythiozolidine-2-th ione. In lism of [ 4C]phen ethyl- ITC were investi- tissues is presented in Figure 15. By 48 contrast, dogs excreted the glycine gated in Fischer rats. After receiving 10 h, 88.7% of the 14C was detected in conjugate, hippuric acid, possibly imol in corn oil by gavage, three rats urine, 9.9% in faeces and 0.1% trapped because they readily hydroxylate the were killed at various times over 48 h, as CO2. More than 90% of the urinary benzylic moiety of benzyl-ITC, which is and 14C was evaluated in whole blood 4D was identified by HPLC as subsequently oxidized to benzoic acid and in 13 major organs at each time, phenethyl-ITC-N-acetylcysteine on the and conjugated. while 14C in expired 002, urine and basis of co-chromatography with faeces was assayed at 8, 24 and 48 h. authentic standards, but less than 1% Phenethyl isothiocyanate The 14C in whole blood peaked at 2.9 occurred as phenethyl-ITC in the urine. The metabolism and tissue disposition h, with two compartments (a and 1) Measurement of radioactivity in homo- of phenothyl-ITC was first investigated with half-times of 2.4 and 21.7 h. The genates of liver and lung by HPLC in NJ mice, a model of lung carcino- maximum activity of 14C appeared in showed the presence of an additional genesis (Shimkin & Stoner, 1975) the liver at 2.5 h, the lungs at 4.5 h major unidentified metabolite (other widely used used to study the efficacy and brain at 6.5 h. The time course of than phenethyl-ITC---GSH an phenethyl- and mechanisms of action of chemo- absorption and elimination in selected ITC-Cys). A considerable amount of the preventive agents (Eklind et al. 1990; Chung, 2001). When [14C]phenethyl- S ITC, synthesized from 2-phenyl[1- 14C]ethylami ne hydrochloride, was CS H \H administered by gavage (EkIind at al., /\ / 1990), 55.2% of the dose was excreted 1. (.) C—c—Nc in urine within 72 h, and 23.3% appeared in faeces. Two major urinary HO COOH metabolites were isolated in urine and identified by and 13C-nuclear magnetic resonance spec- HH S HH trometry and by HPLC co-chromatog- 2. C •-d-NH--S--C•--COOH raphy with ultraviolet standards, as the H H H NH-C--CH3 cyclic mercaptopyruvic acid conjugate 11 4- hyd roxy-4-carboxy-3-phenethylthia- zolidine-2-thione (25% of the dose) and the mercapturic acid of phenethyl- I ITC, N-acetyl- S-(N-phenethylth iocar- bamoyl)-L-cysteine (phenethyl-ITC-N- acotylcystoine) (10% of dose) (see Figure 14). Radioactivity in major organs was determined up to 72 h after dosing; the maximum 14C activity t in liver and lung occurred within 2-8 h and 4-8 h, respectively. When glu- conasturtiin, the glucosinolate precur- '~J'J 2 sor of phenethyl-ITC, was mixed into P AIN-76 diet with myrosinase and fed to 0 10 20 40 6 0 10 20 30 40 50 Retention time (mm) NJ mice, the same two metabolites were Retention time (mm) identified in urine, with only 22% recovery (Chung et al., 1992). These results indi- Figure 14 Detection of urinary metabottes of phenetbyl-isothiocyanate (ITC) in NJ mice by high- cate that gluconasturtiin is hydrolysed to performance liquid chromatography with radioflow and confirmation of the identif es of metabolites phenethyl-ITC endogenously, presum- by nuclear magnetic resonance and mass spectroscopy ably mediated by micro-flora in the gut, as 1. 4-Hydrcxy4-carboxy-3-phenethylthiazolidine-2-thione, a cyclic nercaptopyruvic acid conjugate described above. 2, N-Acetyl S-(N-phenethylthiocarbamoy-L-cysteine (phenethyl-ITC-W-acetylcysteine, a mercap- The pharmacokinetics and metabo- turic acid) (Eklind et ai., 1990)
35 IARC Handbooks of Cancer Prevention Volume 9: Cruciferous vegetables, isothiocyanates and indoles