Proc. Nati. Acad. Sci. USA
`Vol. 85, pp. 8261-8265, November 1988
`Medical Sciences
`
`Identification of a common chemical signal regulating the induction
`of enzymes that protect against chemical carcinogenesis
`(Michael acceptors/quinone reductase/glutathione S-transferase/anticarcinogens)
`
`PAUL TALALAY, MARY J. DE LONG, AND HANS J. PROCHASKA
`Department of Pharmacology and Molecular Sciences, The Johns Hopkins University School of Medicine, Baltimore, MD 21205
`Contributed by Paul Talalay, July 19, 1988
`
`ABSTRACT
`Carcinogenesis is blocked by an extraordi-
`nary variety of agents belonging to many different classes-
`e.g., phenolic antioxidants, azo dyes, polycyclic aromatics,
`flavonoids, coumarins, cinnamates, indoles, isothiocyanates,
`1,2-dithiol-3-thiones, and thiocarbamates. The only known
`common property of these anticarcinogens is their ability to
`elevate in animal cells the activities of enzymes that inactivate
`the reactive electrophilic forms of carcinogens. Structure-
`activity studies on the induction of quinone reductase
`[NAD(P)H:(quinone-acceptor) oxidoreductase, EC 1.6.99.2]
`and glutathione S-transferases have revealed that many anti-
`carcinogenic enzyme inducers contain a distinctive and hitherto
`unrecognized chemical feature (or acquire this feature after
`metabolism) that regulates the synthesis of these protective
`enzymes. The inducers are Michael reaction acceptors char-
`acterized by olefinic (or acetylenic) bonds that are rendered
`electrophilic (positively charged) by conjugation with electron-
`withdrawing substituents. The potency of inducers parallels
`their efficiency in Michael reactions. Many inducers are also
`substrates for glutathione S-transferases, which is further
`evidence for their electrophilicity. These generalizations have
`not only provided mechanistic insight into the perplexing
`question of how such seemingly unrelated anticarcinogens
`induce chemoprotective enzymes, but also have led to the
`prediction of the structures of inducers with potential chemo-
`protective activity.
`
`An astonishing variety of chemical agents protects rodents
`against the toxic and neoplastic effects of carcinogens (1, 2).
`Many lines of evidence (2-5) strongly suggest that the
`elevation ofenzymes concerned with carcinogen inactivation
`is one mechanism of critical importance in achieving chemo-
`protection. Anticarcinogenic enzyme inducers are of two
`types: (i) bifunctional inducers (polycyclic aromatic hydro-
`carbons, azo dyes, and flavonoids), which elevate phase II*
`xenobiotic metabolizing enzymes-e.g., glutathione S-
`transferases (GST), UDP-glucuronosyltransferases, and qui-
`none reductase [QR; NAD(P)H:(quinone-acceptor) oxidore-
`ductase, EC 1.6.99.21-as well as inducing phase I activities
`aryl hydrocarbon hydroxylase); and (ii) mono-
`(e.g.,
`functional inducers (diphenols, thiocarbamates, 1,2-dithiol-
`3-thiones, isothiocyanates, cinnamates, and coumarins),
`which elevate phase II enzymes without inducing aryl hy-
`drocarbon hydroxylase (3). Since phase I enzymes are the
`principal activators of carcinogens to their ultimate reactive
`forms, monofunctional inducers are more promising candi-
`dates than bifunctional inducers as useful anticarcinogens.
`Although bifunctional inducers appear to elevate phase II
`enzymes in part by binding to the Ah (aryl hydrocarbon)
`receptor, the molecular mechanisms by which monofunction-
`
`The publication costs of this article were defrayed in part by page charge
`payment. This article must therefore be hereby marked "advertisement"
`in accordance with 18 U.S.C. §1734 solely to indicate this fact.
`
`al inducers act are less clear (3). This paper extends our
`earlier efforts to identify structural features important for
`phase II enzyme induction by diphenols and phenylenedia-
`mines (5). Since the specific activity of QR in the Hepa lclc7
`murine hepatoma cells is raised by virtually all compounds
`that produce coordinate elevations of phase II enzymes in
`vivo (4, 5, 7), this system was used to determine the potency
`of various types of enzyme inducers. Some inducers identi-
`fied in cell culture were also tested as inducers of QR and
`GST in mouse tissues.
`We report here that phase II enzyme inducers contain, or
`acquire by metabolism, a hitherto unrecognized and distinc-
`tive chemical and structural feature-i.e., an electrophilic
`olefin or related electron-deficient center. They are, there-
`fore, Michael reaction acceptors.t This generalization has led
`to the identification of a number of phase II enzyme inducers
`that are Michael acceptors and are potential chemoprotec-
`tors. Most inducers are also substrates for GST, but whether
`this merely reflects their electrophilic nature or is an intrinsic
`aspect of the mechanism of induction is unclear.
`
`MATERIALS AND METHODS
`QR activities of Hepa lclc7 murine hepatoma cells were
`measured in cells grown in microtiter plates (8). The potency
`of compounds was determined from plots relating the ratio of
`treated to basal (vehicle only) of specific activities of QR to
`the concentration of inducer. Potencies are expressed as the
`concentrations required to double (designated CD) the basal
`specific activity of QR. Induction of cytosolic QR and GST
`activities in female CD1 mouse tissues was assessed by a
`standard protocol (9, 10). All compounds were of the highest
`quality obtainable commercially and were purified when
`necessary.
`
`Abbreviations: QR, quinone reductase; GSH, glutathione; GST,
`glutathione S-transferase; CD, concentration of a compound that
`specific
`activity of QR; BHA, 2(3)-tert-butyl-4-
`doubles the
`hydroxyanisole; CDNB, 1-chloro-2,4-dinitrobenzene; DCNB, 1,2-
`dichloro4nitrobenzene.
`*Enzymes of xenobiotic metabolism are of two types: (i) phase I
`enzymes (e.g., cytochromes P450) functionalize compounds usu-
`ally by oxidation or reduction; and (ii) phase II enzymes conjugate
`functionalized compounds with endogenous ligands (e.g., gluta-
`thione, glucuronic acid). Quinone reductase may be considered a
`phase II enzyme since it does not introduce new functional groups,
`is often induced coordinately with other phase II enzymes, and
`protects cells against toxic agents.
`tIn 1887 A. Michael reported that olefins conjugated with electron-
`withdrawing groups (Z) are susceptible to attack by nucleophiles.
`These so-called "Michael acceptors" have the structures
`CH2=CH-Z, Z'-CH=CH-Z (including quinones), or
`R-CLC-Z (acetylenes). The or4er of reactivity of CH2=CH-Z
`with morpholine or pyrrolidine is Z = NO2 > COAr > CHO >
`COCH3 > CO2CH3 > CN > CONH2 > CONR2 (6). Furthermore,
`alkyl substituents on the olefin decrease reactivity by electronic and
`steric effects.
`
`8261
`
`Page 1 of 5
`
`

`

`8262
`
`Medical Sciences: Talalay et al.
`
`Proc. Natl. Acad. Sci. USA 85 (1988)
`
`their partial aromaticity, which lowers the electrophilicity of
`the a,,B-unsaturated carbonyl function.
`Comparison of the potencies of lactones with those of
`corresponding carbocyclics (compare Compound 4 with 8
`and 7 with 9) showed that the bridge oxygen of the lactones
`weakened inductive activity, either because of hydrolysis of
`lactones to acids or because lactones (i.e., esters) are weaker
`Michael acceptors than ketones.t Moreover, 2-methylene-4-
`butyrolactone (Compound 5; CD = 22 ,uM) and 3-methylene-
`2-norbornanone (Compound 10; CD = 6.8 ,uM) were more
`potent inducers than other cyclic olefins, further strengthen-
`ing the evidence that these compounds induce by behaving as
`Michael acceptors. The higher potencies of Compounds 5 and
`10 may result from the absence of acidic hydrogens on the
`carbon atom adjacent to the electrophilic center of the olefin
`(compare 5 with 4 or 7 and 10 with 8 or 9, respectively). Such
`acidic protons can interfere with Michael' addition by neu-
`tralizing the attacking nucleophile and by destroying the
`electrophilic character of the acceptor (15):
`R)- X
`Nuc:-
`C X
`Nc
`Z
`
`>
`
`z
`
`R
`
`+ Nuc-H
`
`+ Nuc:-
`
`Although these protons are only weakly acidic in aqueous
`systems, they may decrease the reactivity of Michael accep-
`tors in the presence of biological macromolecules that could
`promote deprotonation. The relationship between potency of
`phase II enzyme induction and Michael acceptor efficiency is
`also supported by the high activity of 1-nitro-1-cyclohexene
`(Compound 12; CD = 1-2 IM) since nitroolefins are highly
`efficient Michael acceptors.t Furthermore, all cyclic com-
`pounds lacking an olefin conjugated to an electron-
`withdrawing group were inactive as inducers (Compounds 6
`and 11).
`-Acrylate, Crotonate, and Cinnamate Analogues. Identifica-
`tion of an a,,p-unsaturated carbonyl function as essential for
`inductive activity among lactones and carbocyclics raised
`two questions: (i) must these olefins be part of a ring system,
`and (ii) what electron-withdrawing groups can replace the
`ketone function? Studies with acrylate, crotonate, and cin-
`namate analogues clarified both questions (Table 2).
`The methyl esters of acrylic (Compound 13) and cinnamic
`(Compound 28) acids were both inducers, showing unequiv-
`ocally that the activated olefin need not be cyclic. In contrast,
`free crotonic (Compound 22) and cinnamic (Compound 27)
`acids were inactive, as anticipated from the fact that free
`carboxyl groups weaken Michael acceptor efficiency, but
`more efficient cellular uptake of esters than free acids also
`may be important. The inactivity of methyl methacrylate
`(Compound 14), methyl crotonate (Compound 23), and
`methyl tiglate (Compound 15) in comparison to methyl
`acrylate (Compound 13; CD = 20,M) suggested that methyl
`substituents on either the a- or B-carbon (or both) interfered
`with inductive activity, presumably because of electronic and
`steric effects (6), as predicted by recent molecular orbital
`calculations (16). Aldehydes of all three series are inducers,
`although crotonaldehyde (Compound 24; CD = 9 ,uM) was
`far more potent than either acrolein (Compound 16; CD =
`130 AM) or cinnamaldehyde (Compound 29; CD = 110 ,uM).
`The variation in inductive potencies of these aldehydes might
`reflect differences in rates of their metabolism.
`In the acrylate series, inductive activity is preserved if the
`oxygen function is part of a ketone (i.e., methyl vinyl ketone,
`Compound 17; CD = 40 ,uM) or of a sulfone (i.e., methyl
`vinyl sulfone, Compound 18; CD = 25 AM), but not of an
`amide (i.e., acrylamide; Compound 20). The carbonyl func-
`tion may also be replaced by a nitrile (acrylonitrile, Com-
`pound 19; CD = 50 AM), although not in the crotonate
`
`RESULTS AND DISCUSSION
`Diphenols, Phenylenediamines, and Quinones. Studies with
`2(3)-tert-butyl-4-hydroxyanisole (BHA) and several ana-
`logues led to the conclusion that the inductive capacity of
`these compounds depended on their conversion to diphenols
`(10). Furthermore, among such diphenols only catechols
`(1,2-diphenols) and hydroquinones (1,4-diphenols), but not
`resorcinols (1,3-diphenols), were inducers, and the presence
`or position of other substituents was of minor importance (5).
`Analogous results were obtained with phenylenediamines.
`These findings suggested that chemical (oxidative) reactivity
`rather than unique structural features was the crucial deter-
`minant of inductive activity, since 1,2- and 1,4-diphenols and
`corresponding phenylenediamines readily undergo oxidative
`conversion to quinones or quinoneimines, respectively,
`whereas the 1,3- analogues do not (5). We speculated that the
`inductive process depended on oxidative lability of the
`inducers. However, the present studies disclosed that the
`signaling of enzyme induction depended on electrophilic
`centers, and the inductive ability of diphenols and phenylene-
`diamines could therefore be ascribed to their conversion to
`electrophilic quinones or quinoneimines, respectively.
`Coumarin Analogues. Coumarin (Table 1, Compound 1) is
`a widely distributed, naturally occurring flavoring agent. It
`protects against hydrocarbon-mediated carcinogenesis and
`elevates GSTs in rodent tissues (11-14). We reported that
`coumarin raised the QR levels in Hepa lclc7 cells grown on
`75-cm2 plates (CD = 100 ,uM; ref. 7), although it was inactive
`in the microtiter assay system (CD > 800 ,M).
`Examination of coumarin analogues to decipher the struc-
`tural features required for QR induction showed that 2H-
`pyran-2-one (Compound 2) and 4H-pyran-4-one (Compound
`3) were inactive, whereas 5,6-dihydro-2H-pyran-2-one (Com-
`pound 4) had considerable inductive activity (CD = 45 jAM).
`The latter finding suggested that the structural feature critical
`for inductive activity was an electrophilic olefin conjugated
`with a carbonyl group-i.e., a Michael reaction acceptor.t
`Hence, lack of activity of Compounds 1-3 could be related to
`
`Concentration (CD) of coumarin and pyran analogues
`Table 1.
`required to double quinone reductase in Hepa lc1c7 cells
`Compound
`Name
`
`No.
`
`CD, AM
`
`Coumarin
`1
`2H-Pyran-2-one
`2
`4H-Pyran-4-one
`3
`5,6-Dihydro-2H-pyran-2-one
`4
`2-Methylene-4-butyrolactone
`5
`2-Methylbutyrolactone
`6
`4-Hydroxycrotonolactone
`7
`1-Cyclohexen-2-one
`8
`1-Cyclopenten-2-one
`9
`3-Methylene-2-norbornanone
`10
`y-Valerolactone
`11
`1-Nitro-1-cyclohexene
`12
`-2
`I, inactive; <20% increase in specific activity at 200 MM.
`The structures of 1-12 are as follows:
`
`45
`22
`I
`210
`28
`100
`6.8
`
`11
`
`I I
`
`0
`
`3
`
`CH2
`
`CH3
`
`4eso 5s
`4
`5
`
`6o
`6
`
`:
`
`H2 ]:
`O 10O C
`12
`1 1
`10
`°
`
`yNO2
`
`1
`
`2
`
`o7- )
`
`7 0
`7
`
`8
`
`9
`
`Page 2 of 5
`
`

`

`Medical Sciences: Talalay et al.
`
`Proc. Natl. Acad. Sci. USA 85 (1988)
`
`8263
`
`Concentration (CD) of acrylate, crotonate, and
`Table 2.
`cinnamate analogues required to double quinone
`reductase in Hepa 1clc7 cells
`Compound
`
`No.
`
`Name
`
`CD,,M
`
`13
`14
`15
`16
`17
`18
`19
`20
`21
`
`22
`23
`24
`25
`26
`
`27
`28
`29
`30
`31
`32
`
`Crotonic acid
`Methyl crotonate
`Crotonaldehyde
`Crotononitrile
`Methyl tetrolate
`
`dimethyl esters of fumarate (Compound 35) and maleate
`(Compound 38) might result from the lower electrophilicity of
`the double bond or from steric effects. The relatively low
`inductive potency of dimethyl acetylenedicarboxylate (Com-
`pound 41; CD = 87 MxM) in comparison to the acetylenic
`monocarboxylates (Compounds 21 and 26) and the olefinic
`dicarboxylates (Compounds 34 and 37), correlates with
`decreased Michael acceptor efficiency, possibly because of
`the tendency of one carbonyl group to destabilize the gen-
`eration of a positive center by the other carbonyl function.
`These observations are of potential importance since
`fumarate and itaconate occur naturally and are used as food
`additives. Furthermore; fumaric acid is the active principle
`responsible for some of the many pharmacological effects of
`extracts of shepherd's purse (Capsella bursa-pastoris), a
`crucifer widely used for medicinal purposes in Asia. Such
`extracts, or fumaric acid itself, protect against chemical
`carcinogenesis in liver of rats and in forestomach and lungs
`of mice (17, 18).
`Induction of QR and GST in Mouse Tissues. The identifi-
`cation- of a number of Michael reaction acceptors that
`elevated QR levels in Hepa 1c1c7 cells raised the question
`whether such compounds were (i) active when administered
`to mice in vivo, (ii) active in tissues other than liver, and (iii)
`whether GSTs (which are normally not significantly induced
`in $epa cells) were also induced. Hence, we administered
`several of these newly recognized inducers-dimethyl mal-
`eate (Compound 37), dimethyl itaconate (Compound 40),
`methyl acrylate (Compound 13), 5,6-dihydro-2H-pyran-2-one
`(Compound 4), and 2-methylene-4-butyrolactone' (Com-
`pound 5)-by gavage (25-75 ,umol daily for 5 days) to female
`CD1 mice and measured the specific activities ofQR and GST
`activities with 1-chloro-2,4-dinitrobenzene (CDNB) and 1,2-
`dichloro-4-nitrobenzene (DCNB) in the cytosols of liver,
`forestomach, and glandular stomach (Table 4) according to a
`standard protocol (9,
`10). Mice treated with BHA and
`tert-butylhydroquinone were used as positive controls and
`gave inductions similar to those reported previously (see ref.
`9; values -not shown in Table 4).
`The results in mice in vivo are similar to the findings
`obtained with Hepa 1c1c7 cells. With few exceptions, all
`compounds induced all three enzyme activities and, for the
`most part, coordinately. Furthermore, several of the new
`electrophiles were'more potent and more effective than BHA
`and tert-butylhydroquinone. For example, in the glandular
`stomach, phase 1I enzymes are only slightly responsive to
`BHA and tert-butylhydroquinone (9, 10), whereas five doses
`of 25 jLmol of dimethyl itaconate raised the enzyme specific
`activities 2.63- to 6.73-fold. Examination of the least toxic of
`these compounds with widest tissue specificity will, there-
`fore, be of great interest in cancer protection experiments in
`vivo. Chemoprotection by these new compounds would
`further strengthen the view that enzyme induction is a central
`mechanism of chemoprotection.
`Suffur Compounds: Isothiocyanates, 1,2-Dithiol-3-thiones,
`Thiocarbamates, and Allyl Sulfides
`These sulfur-containing compounds: (i) protect rodents
`against the toxic and neoplastic effects of carcinogens; and
`(ii) induce QR and GST in rodent tissues and QR in the Hepa
`lclc7 cell line. Moreover, many sulfur compounds are
`present in commonly consumed vegetables 'that protect
`against cancer (1, 13, 14, 19-28). We therefore considered
`whether electrophilic centers might likewise be responsible
`for the enzyme inducing activity of these sulfur-containing
`compounds.
`Isothiocyanates. The inductive ability of various alkyl and
`aromatic isothiocyanates depended on the presence of at
`least one hydrogen on the carbon adjacent to the isothiocya-
`
`20
`I
`I
`130
`40
`25
`50
`I
`
`5
`
`I
`I
`
`9
`
`I
`15
`
`I
`125
`110
`I
`600
`25
`
`Structure
`Acrylates
`CHf=CHCOOCH3
`Methyl acrylate
`Methyl methacrylate CH2=C(CH3)COOCH3
`CH3CH-=C(CH3)COOCH3
`Methyl tiglate
`CH2=CHCHO
`Acrolein
`CH2--CHCOCH3
`Methyl vinyl ketone
`Methyl vinyl sulfone CH2=CHSO2CH3
`CH2=CHC-N
`Acrylonitrile
`CH2=CHCONH2
`Acrylamide
`CHaCCOOCH3
`Methyl propiolate
`Crotonates
`CH3CH=CHCOOH
`CH3CH=-CHCOOCH3
`CH3CHI=CHCHO
`CH3CH=CHC=N
`CH3(>CCOOCH3
`trans-Cinnamates
`C6H5CH=CHCOOH
`Cinnamic acid
`C6H5CH=CHCOOCH3
`Methyl cinnamate
`C6H5CH=CHCHO
`Cinnamaldehyde
`C6H5CH=CHC=N
`Cinnamonitrile
`Cinnamanmide
`C6H5CH=CHCONH2
`f3-Nitrostyrene
`C6H5CH=CHNO2
`I, inactive; <20% increase in specific activity at 200 MM.
`(Compound 25) and cinnamate (Compound 30) series. Elec-
`trophilic acetylenes, such as methyl propiolate (Compound
`21) and methyl tetrolate (Comp'ound 26), are very efficient
`Michael acceptors and were also very potent inducers. As
`predicted from the high potency of 1-nitro-l-cyclohexene in
`relation to other cyclic olefins, ,/-nitrostyrene (Compound 32;
`CD = 25 ,uM) was by far the most efficient inducer among the
`cinnamates. We conclude that the electrophilic olefin need
`not be cyclic and that inductive activity generally parallels
`the potency of the electron-withdrawing group.t
`Unsaturated Dicarboxylic Acids. Fumaric (Table 3, Com-
`poiand 33) and maleic (Compound 36) acids' were inactive as
`inducers of QR in Hepa lclc7 cells, whereas their dimethyl
`esters (Compounds 34 and 37, respectively) and dimethyl
`itaconate (Compound 40) were moderately potent inducers
`with CD values of 20-35 AM. The olefinic structure is
`essential since the saturated dimethyl succinate (Compound
`39) was inactive. The esters of the unsaturated dicarboxylic
`acids were more active than were the free acids. Further-
`more, the lower potency of the diethyl compared with the
`Concentration (CD) of fumarate, maleate, and itaconate
`Table 3.
`derivatives required to double quinone reductase in Hepa
`1clc7 cells
`
`Compound
`
`No.
`33
`34
`35
`36
`37
`38
`39
`40
`41
`
`Structure
`Name
`HOQCCH=CHCOOH (trans)
`Fumaric acid
`Dimethyl fumarate CH300CCH=CHCOOCH3 (trans)
`C2H500CH=CHCOOC2H5 (trans)
`Diethyl fumarate
`HOOCCH=CHCOOH (cis)
`Maleic acid
`Dimethyl maleate CH300CCH=CHCOOCH3 (cis)
`C2H50OCCH=CHCOOC2H5 (cis)
`Diethyl maleate
`Dimethyl succinate CH300CCH2CH2COOCH3
`Dimethyl itaconate CH300CC(=CH2)CH2COOCH3
`Dimethyl acetylene CH3O-C0CCOOCH3
`dicarboxylate
`I, inactive; <20%o increase in specific activity at 200 MM.
`
`CD
`AM
`I
`22
`100
`I
`20
`40
`I
`35
`87
`
`Page 3 of 5
`
`

`

`8264
`
`Medical Sciences: Talalay et al.
`
`Proc. Natl. Acad. Sci. USA 85 (1988)
`
`Table 4.
`
`Induction patterns of QR and GSTs measured with CDNB and DCNB in mouse tissues
`Dose, ,umol
`Ratio of specific activities (treated/control)
`per
`mouse per day
`25
`
`Inducing agent
`Dimethyl maleate (37)
`
`Glandular stomach
`Liver
`Enzyme
`Forestomach
`2.42 ± 0.18
`QR
`2.68 ± 0.14
`2.44 ± 0.13
`CDNB
`1.44 ± 0.08
`3.42 ± 0.17
`3.61 ± 0.12
`6.35 ± 0.76
`DCNB
`3.96 ± 0.16
`1.11 ± 0.02
`2.43 ± 0.10
`2.82 ± 0.20
`QR
`2.89 ± 0.23
`3.61 ± 0.09
`6.15 ± 0.52
`CDNB
`4.36 ± 0.36
`2.09 ± 0.03
`5.05 ± 0.84
`DCNB
`5.34 ± 0.49
`2.37 ± 0.22
`2.74 ± 0.11
`2.63 ± 0.13
`QR
`1.79 ± 0.06
`3.47 ± 0.06
`5.07 ± 0.53
`CDNB
`4.34 ± 0.30
`6.73 ± 0.71
`DCNB
`0.90 ± 0.06
`2.73 ± 0.15
`2.90 ± 0.07
`QR
`3.30 ± 0.05
`4.27 ± 0.28
`6.71 ± 0.30
`CDNB
`2.79 ± 0.07
`1.71 ± 0.08
`4.90 ± 0.22
`7.75 ± 0.57
`DCNB
`2.43 ± 0.10
`2.11 ± 0.07
`2.94 ± 0.10
`QR
`1.64 ± 0.10
`2.78 ± 0.07
`3.65 ± 0.29
`CDNB
`0.93 ± 0.03
`2.98 ± 0.17
`4.72 ± 0.22
`DCNB
`3.40 ± 0.12
`2.11 ± 0.23
`2.55 ± 0.03
`QR
`2.55 ± 0.03
`2.58 ± 0.35
`CDNB
`2.50 ± 0.20
`1.51 ± 0.06
`DCNB
`4.37 ± 0.24
`2.04 + 0.39
`7.93 ± 0.33
`1.04 ± 0.05
`QR
`2.39 ± 0.06
`2.30 ± 0.06
`2.29 ± 0.03
`CDNB
`3.11 ± 0.16
`1.32 ± 0.11
`1.61 ± 0.09
`DCNB
`3.30 ± 0.13
`The SEM values of the treated/control ratios were obtained by dividing the SEM of the mean of each treated group by the control value.
`The enzyme specific activities (nmol/min/mg of protein ± SEM) of vehicle-treated controls were as follows. Liver: QR, 126 + 3.3; CDNB,
`1800 ± 30; DCNB, 37.9 + 3.9. Forestomach: QR, 1454 ± 88; CDNB, 1130 + 69; DCNB, 15.6 ± 0.8. Glandular stomach: QR, 3955 + 187;
`CDNB, 684 + 40; DCNB, 8.35 ± 1.06. All compounds were administered to 6-week-old female CD1 mice (three to six mice per group) by gavage
`in indicated single daily doses in 0.1 ml of Emulphor EL620P for 5 days. Cytosols were prepared from the tissues 24 hr after the last treatment
`and assayed for enzyme activities as described (9, 10).
`nate group (i.e., R1R2CH-N=C=S). Thus, benzyl (CD =
`1.8 AM), phenethyl (CD = 2.0 ,uM), ethyl (CD = 30 AtM),
`n-propyl (CD = 14 AM), n-butyl (CD = 15 ,uM), allyl (CD =
`4 ,uM), and' cyclohexyl (CD = 14 ,uM) 'isothiocyanates were
`potent enzyme inducers. In contrast, tert-butyl, phenyl,
`2-fluorophenyl, 3-fluorophenyl, 4-fluorophenyl, 4-chloro-
`phenyl, 4-tolylphenyl, and a-naphthyl isothiocyanates did
`not induce. Since an a-hydrogen is essential for inductive
`activity, it is tempting to speculate that tautomerization ofthe
`methylene-isothiocyanate moiety to a structure resembling
`an aB3-unsaturated thioketone may be important for induc-
`tive activity:
`
`Dimethyl itaconate (40)
`
`Methyl acrylate (13)
`
`5,6-Dihydro-2H-pyran-2-one (4)
`
`2-Methylene-4-butyrolactone (5)
`
`75
`
`25
`
`75
`
`25
`
`50
`
`50
`
`though these compounds do not appear to be Michael
`acceptors per se, nucleophilic attack on the electrophilic
`terminal carbon adjacent to the sulfur, with the resultant
`elimination of a thiol, may be a mechanism by which these
`compounds are active:
`
`Nuc:
`
`b Nuc
`
`+ X
`
`Indeed, we found that allyl bromide (CD = 200 ,uM) is
`approximately equiactive with allyl disulfide (CD = 150 ,uM)
`as an inducer of QR.
`Thiocarbamates. There is no obvious Michael-based mech-
`anism for phase II enzyme induction by chemoprotective
`anticarcinogenic thiocarbamates such as disulfiram [(C2-
`H5)2NCQ=S)S]2, diethyldithiocarbamate [(C2H5)2NCQ=S)-
`SHI, 'and bisethylxanthogen [(C2H5OC(=S)S)2]. Possibly,
`thiocarbamates require metabolic activation to form the
`required electrophiles since at least some thiocarbamates are
`conjugated with glutathione (see ref. 30).
`Inducers Are Substrates for GSTs. GSTs catalyze the
`conjugation ofglutathione (GSH) with compounds ofthe type
`RARBC=CRcZ, in which Z is an electron-withdrawing
`group. The enzymatic reactivity of such compounds depends
`on the electron-attracting power of the Z group and on the
`electronic (attraction or repulsion) and steric effects of the
`RA, RB, and Rc groups (30-32). GSTs catalyze Michael
`additions of GSH to hydrophobic electrophiles by increasing
`the nucleophilicity of GSH (30, 33).
`A remarkable similarity is immediately apparent between
`the structural features of substrates for GSTs and those
`required for phase II enzyme induction. Many a,3-
`unsaturated enzyme inducers such as esters, aldehydes,
`ketones, lactones, nitriles, nitroalkenes, and sulfones, share
`these properties (30-32). Thus, the ethyl esters of maleate,
`fumarate, and acrylate, and dimethyl itaconate share the
`capacity to induce and to serve as substrates for GSTs.
`Furthermore, the very active nitroalkene inducer, /3-
`
`R1R2CH-N
`
`C
`
`S
`
`R1R2C= N
`
`C
`
`S
`
`1,2-Dithiol-3-thiones. Substituted 1,2-dithiol-3-thiones
`were initially recognized to be potential anticarcinogens
`because of their ability to induce phase'II enzymes in mice
`(25, 29). Since unsubstituted 1,2-dithiol-3-thione (Compound
`42) and 1,2-dithiol-3-one (Compound 43) are efficient induc-
`ers of phase II enzymes, signaling of phase II enzyme
`induction does not require ring substitution (26, 27). The
`1,2-dithiol-3-thione nucleus contains an olefin conjugated to
`an electron-withdrawing thioketone or conventional ketone,
`a structural feature common to all the inducers discussed
`here:
`
`X = S (Compound 42)
`X = 0 (Compound 43)
`Allyl Mono-, Di-, and Trisulfides. Recently, Wattenberg
`and colleagues (28) have described the anticarcinogenic
`effects of allyl mono-, di-, and trisulfides and their ability to
`induce GST. An unsaturated allylic moiety is essential for
`both enzyme inductive and anticarcinogenic effects. Al-
`
`Page 4 of 5
`
`

`

`Medical Sciences: Talalay et A
`
`Proc. Natl. Acad. Sci. USA 85 (1988)
`
`8265
`
`3.
`
`4.
`
`5.
`
`6.
`
`8.
`
`Prochaska, H. J. & Talalay, P. (1988) Cancer Res. 48, 4776-
`4782.
`Talalay, P. & Prochaska, H. J. (1987) in DT-Diaphorase: A
`Quinone Reductase with Special Functions in Cell Metabolism
`and Detoxication, eds. Ernster, L., Estabrook, R. W., Hoch-
`stein, P. & Orrenius, S. (Cambridge Univ. Press, Cambridge,
`UK), pp. 62-66.
`Prochaska, H. J., De Long, M. J. & Talalay, P. (1985) Proc.
`Natl. Acad. Sci. USA 82, 8232-8236.
`Shenhav, H., Rappoport, Z. & Patai, S. (1970) J. Chem. Soc.
`B469-B476.
`7. De Long, M. J., Prochaska, H. J. & Talalay, P. (1986) Proc.
`Natl. Acad. Sci. USA 83, 787-791.
`Prochaska, H. J. & Santamaria, A. B. (1988) Anal. Biochem.
`169, 328-336.
`9. De Long, M. J., Prochaska, H. J. & Talalay, P. (1985) Cancer
`Res. 45, 546-551.
`Prochaska, H. J., Bregman, H. S., De Long, M. J. & Talalay,
`P. (1985) Biochem. Pharmacol. 34, 3909-3914.
`Feuer, G., Kellen, J. A. & Kovaks, K. (1976) Oncology 33, 35-
`39.
`Wattenberg, L. W., Lam, L. K. T. & Fladmoe, A. V. (1979)
`Cancer Res. 39, 1651-1654.
`Sparnins, V. L. & Wattenberg, L. W. (1981) J. Natl. Cancer
`Inst. 66, 769-771.
`Sparnins, V. L., Chuan, L. & Wattenberg, L. W. (1982) Can-
`cer Res. 42, 1205-1207.
`March, J. (1985) Advanced Organic Chemistry: Reactions,
`Mechanisms, and Structure (Wiley, New York), 3rd Ed.
`Osman, R., Namboodiri, K., Weinstein, H. & Rabinowitz,
`J. R. (1988) J. Am. Chem. Soc. 110, 1701-1707.
`Kuroda, K., Kanisawa, M. & Akao, M. (1982) J. Nat!. Cancer
`Inst. 69, 1317-1320.
`Kuroda, K., Terao, K. & Akao, M. (1983) J. Natl. Cancer Inst.
`71, 855-857.
`Kjwr, A. (1%1) in Organic Sulfur Compounds, ed. Kharash, N.
`(Pergamon, New York), Vol. 1, pp. 409-420.
`Wattenberg, L. W. (1977) J. Natl. Cancer Inst. 58, 395-398.
`Sparnins, V. L., Venegas, P. L. & Wattenberg, L. W. (1982) J.
`Natl. Cancer Inst. 68, 493-496.
`Wattenberg, L. W. (1983) Cancer Res. Suppl. 43, 2448s-2453s.
`Benson, A. M. & Barretto, P. B. (1985) Cancer Res. 45, 4219-
`4223.
`Benson, A. M., Barretto, P. B. & Stanley, J. S. (1986) J. Natl.
`Cancer Inst. 76, 467-473.
`Ansher, S. S., Dolan, P. & Bueding, E. (1986) Food Chem.
`Toxicol. 24, 405-415.
`26. De Long, M. J., Dolan, P., Santamaria, A. B. & Bueding, E.
`(1986) Carcinogenesis 7, 977-980.
`Kensler, T. W., Egner, P. A., Dolan, P. M., Groopman, J. D.
`& Roebuck, B. D. (1987) Cancer Res. 47, 4271-4277.
`Sparnins, V. L., Barany, G. & Wattenberg, L. W. (1988)
`Carcinogenesis 9, 131-134.
`Wattenberg, L. W. & Bueding, E. (1986) Carcinogenesis 7,
`1379-1381.
`Chasseaud, L. F. (1979) Adv. Cancer Res. 29, 175-274.
`Boyland, E. & Chasseaud, L. F. (1967) Biochem. J. 104, 95-
`102.
`Boyland, E. & Chasseaud, L. F. (1968) Biochem. J. 109, 651-
`661.
`Jakoby, W. B. (1978) Adv. Enzymol. 46, 383-414.
`Brusewitz, G., Cameron, B. D., Chasseaud, L. F., Gorler, K.,
`Hawkins, D. R., Koch, H. & Mennicke, W. H. (1977) Bio-
`chem. J. 162, 99-107.
`Mennicke, W. H., Gorler, K. & Krumbiegel, G. (1983) Xeno-
`biotica 13, 203-207.
`
`nitrostyrene, is also an excellent substrate for GST, as are
`methyl vinyl sulfone (Compound 18) and ethyl vinyl ketone
`(related to Compound 17). Whereas methyl acrylate (Com-
`pound 13) is active in both systems, the closely related methyl
`methacrylate (Compound 14) is not. Moreover, free fumaric
`(Compound 33), cinnamic (Compound 27), and crotonic
`(Compound 22) acids, and a,,3-dimethyl acrylate (Compound
`15) are inactive in both systems (30-32).
`In a series of aromatic and alkyl isothiocyanates, only
`analogues bearing a free hydrogen on the carbon adjacent to
`the isothiocyanate group were inducers of QR in Hepa lclc7
`cells (see above). It is striking that the ability of these
`compounds to form GSH conjugates also required the pres-
`ence of an a-hydrogen (34, 35).
`Are members of other classes of inducers also conjugated
`with glutathione in animal tissues? Quinones are good Mi-
`chael acceptors and their reaction with GSH is enzyme
`promoted. Some organic isothiocyanates are metabolized to
`products that suggest an initial conjugation with GSH (34,
`35), and some thiocarbamates appear also to undergo initial
`sulfoxidation followed by conjugation with GSH (30).
`Furthermore, several common substrates used to assay
`GST were found to be inducers. Thus, the CD for QR
`induction by DCNB was 150,uM. CDNB produced a 1.6-fold
`induction at 10 AM but was toxic at higher concentrations.
`Two other substrates for GST were also inducers: ethacrynic
`acid [2,3-dichloro-4-(2-methylenebutyryl)phenoxyacetic
`acid] (CD = 30,uM) and 1,2-epoxy-3-(p-nitrophenoxy)pro-
`pane (CD = 68 MM).
`It is perhaps not surprising that inducers of phase II
`xenobiotic metabolizing enzymes should prove to be sub-
`strates for GST (and vice versa) since many xenobiotics (or
`even endobiotics) induce enzymes for their own metabolism.
`Both induction and substrate activity require the presence of
`an electrophilic Michael acceptor function. Whether these
`processes are related through a requirement for electrophi-
`licity, or whether there is a more fundamental causal rela-
`tionship between them is presently unclear.
`In conclusion, it is gratifying that the capacity of an
`extraordinary variety of seemingly unrelated anticarcinogens
`to induce protective enzymes can be attributed to the
`presence, or acquisition by metabolism, of a simple and
`hitherto unrecognized chemical property: that of a Michael
`reaction acceptor.
`
`We are grateful for many valuable and pleasurable discussions
`with our colleagues: Jih Ru Hwu, Gary H. Posner, and Cecil H.
`Robinson. We thank Annette B. Santamaria for expert technical
`assistance. These studies were supported by grants from the National
`Institutes of Health (CA 44530), the American Cancer Society (SIG-3
`and RDP-30), and the American Institute for Cancer Research.
`H.J.P. was supported by National Institutes of Health Training Grant
`CA 09243.
`
`1.
`2.
`
`Wattenberg, L. W. (1985) Cancer Res. 45, 1-8.
`Talalay, P., De Long, M. J. & Prochaska, H. J. (1987) in
`Cancer Biology and Therapeutics, eds. Cory, J. G. & Szenti-
`vani, A. (Plenum, New York), pp. 197-216.
`
`10.
`
`11.
`
`12.
`
`13.
`
`14.
`
`15.
`
`16.
`
`17.
`
`18.
`
`19.
`
`20.
`21.
`
`22.
`23.
`
`24.
`
`25.
`
`27.
`
`28.
`
`29.
`
`30.
`31.
`
`32.
`
`33.
`34.
`
`35.
`
`Page 5 of 5
`
`