structure-activity relationships of isothiocyanates as mechanism

[CANCER RESEARCH 54, 4327-4333, August 15, 1994]

Structure-Activity Relationships of Isothiocyanates as Mechanism-based Inhibitors

of 4-(Methy Initrosamino) -1-(3-py ridy 1)-1-butanone-inducedLung Tumorigenesis in A/J Mice1

Ding Jiao, Karin I. Eklind, Chang-In Choi, Dhimant H. Desai, Shantu G. Amin, and Fung-Lung Chung2

Division of Chemical Carcinogcnesis, Naylor Dana Institute for Disease Prevention, American Health Foundation, Valhalla, New York 10595

ABSTRACT

A structure-activity relationship study was carried out to identify struc

tural features in arylalkyl and alkyl isothiocyanates that are associatedwith the inhibitory potency of these compounds against lung tumorigen-esis induced in A/J mice by the tobacco-specific nitrosamine 4-(methylni-trosamino)-l-(3-pyridyl)-l-butanone (NNK). These features include the

alkyl chain length, phenyl substitution, and secondary isothiocyanates.The naturally occurring ally! isothiocyanate, phenethyl Â¡sothiocyanate,and the synthetic analogues such as 6-phenylhexyl isothiocyanate, 8-phe-nyloctyl isothiocyanate, 10-phenyldecyl isothiocyanate, 1,2-diphenylethylisothiocyanate, 2,2-diphenylethyl isothiocyanate, and alkyl isothiocyanates(with 1-hexyl, 2-hexyl, and 1-dodecyl as alkyl moieties) were assayed in

mice for their tumor inhibitory potential. The isothiocyanates were givenin corn oil by gavage at doses of either 0.04, 0.1, and 0.2 juniol or 1 and 5/imol 2 h prior to a single i.p. injection of 10 jumol NNK. Mice weresacrificed 16 weeks later and lung adenomas were counted. At 0.2 /Â¿mol,8-phenyloctyl isothiocyanate and 10-phenyldecyl isothiocyanate werestronger inhibitors than the previously tested 6-phenylhexyl isothiocya

nate, but the difference in potency was not obvious at the lower doses. Atboth 1 and 5 /xmol, ally! isothiocyanate was inactive, while the other fivesynthetic isothiocyanates were considerably more potent than phenethylisothiocyanate. In the alkyl isothiocyanate series, 2-hexyl isothiocyanatewas more potent than 1-hexyl isothiocyanate, while 1-dodecyl isothiocya

nate was the most potent at 1 itmol, reducing tumor multiplicity in thegroup treated with NNK alone from 11.1 to the background level. Also,1,2-diphenylethyl isothiocyanate appeared to be a stronger inhibitor than2,2-diphenylethyl isothiocyanate. In this study we have shown that the

phenyl moiety is not essential for the inhibitory activity since alkyl isothiocyanates exhibit strong inhibitory effects against lung tumorigenesis.We have also shown that secondary isothiocyanates possess a higherpotency than their structural isomers bearing a primary isothiocyanate.From results of this study and of seven previously studied isothiocyanates,we conclude that the observed inhibitory potency of isothiocyanates in theA/J mouse lung tumor model is correlated with their partition coefficients(log /') and the pseudo first order rate constants for the reaction of

isothiocyanates toward glutathione (A,,,,J. These results reveal that bothhigh lipophilicity and low reactivity of isothiocyanates are important forinhibitory activity toward NNK-induced lung tumorigenesis. These obser

vations provide a structural basis for the discovery of more effectivechemopreventive agents.

INTRODUCTION

Naturally occurring isothiocyanates of cruciferous origin have beenshown to protect against tumorigenesis induced by environmentalcarcinogens such as polycyclic aromatic hydrocarbons and nitro-samines in rodents (1-3). For example, BITC3 inhibited B(o)P-in-

Received 3/25/94; accepted 6/6/94.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1This work was supported by Grant CA 46535 from the National Cancer Institute.This is Paper 19 in the series. "Dietary Inhibitors of Chemical Carcinogenesis."

2 To whom requests for reprints should be addressed.' The abbreviations used are: BITC, benzyl isothiocyanale; NNK, 4-(methylnitro-

samino)-l-(3-pyridyl)-l-butanone; B(a)P, benzo(a)pyrene; AITC, allyl isothiocyanate;PITC, phenyl isothiocyanate; PEITC, phenethyl isothiocyanate; PPITC, 3-phenylpropylisothiocyanate; PBITC, 4-phenylbutyl Â¡solhiocyanate; PPelTC, 5-phenylpentyl isothio-

duced tumor development in the lung and forestomach and N-nitroso-diethylamine-induced neoplasms in the forestomach in A/J mice (1,

2). PEITC showed significant inhibitory effects against lung tumorinduction by NNK in both F344 rats and A/J mice (3, 4). PEITC alsoeffectively inhibited W-nitrosobenzylmethylamine-induced esopha-

geal carcinogenesis in rats (5). However, PEITC showed no effect onlung tumorigenesis induced by B(a)P in A/J mice (6). Since bothNNK and B(a)P are important tobacco carcinogens and lung cancer isattributed to tobacco usage (7), further development of more effectivechemopreventive agents in the isothiocyanate family is warranted forthe prevention of lung cancers (8-11).

We have previously shown that synthetic analogues of PEITC witha longer alkyl chain length, up to 6 carbons as in PHITC, possessgreater inhibitory activity against NNK-induced lung tumors in A/J

mice (12). These natural and synthetic isothiocyanates are potentcompetitive inhibitors of cytochrome P-450 isozymes in mouse lung

microsomes that are responsible for the metabolic activation of NNKto methylating and pyridyloxobutylating species (13-15). Inhibitionof the metabolic activation of NNK reduces the formation of O6-

methylguanine and consequently decreases the number of neoplasmsin the lungs of NNK-treated A/J mice (3, 12). The greater inhibitions

by arylalkyl isothiocyanates with a longer alkyl chain are qualitativelycorrelated with the decreasing level of O6-methylguanine in the DNA

of lung tissues (12). Consistent with these observations, in vitrostudies have shown that increasing the alkyl chain length enhanced thebinding affinity of isothiocyanates to cytochrome P-450 isozymes,

which resulted in greater inhibition of NNK metabolism (15). It hasbeen suggested that the increased lipophilicity of the longer chainanalogues enhances the binding affinity of the isothiocyanates tocytochrome P-450 isozymes (12). Factors such as chemical stability,

metabolism, absorption, and molecular geometry of these isothiocyanates may also play a role in the relative inhibitory potency of thesecompounds. Since the major metabolic pathway of isothiocyanates inrodents and humans proceeds via conjugation with GSH (16-18),

their chemical reactivity toward GSH would seem to be important forin vivo potency. The identification of the structural features of isothiocyanates is important to explain their inhibitory efficacy and toguide the design of more effective chemopreventive agents. Our goalsin this study were: (a) to determine the optimal chain length formaximal tumor inhibition by testing synthetic arylalkyl isothiocyanates with longer alkyl chain lengths than those previously assayed;(b) to examine other structural features that are critical to the inhibitory activities by replacing the aromatic moiety with allyl and alkylgroups, by increasing the number of aromatic rings, or by includingsecondary isothiocyanates; (c) to correlate lipophilicity and chemicalreactivity of isothiocyanates with their relative inhibitory potency in

cyanate, PHITC, 6-phenylhexyl isolhiocyanate; POITC, 8-phenyloctyl isothiocyanate;PDITC, 10-phenyldecyl isothiocyanate; OPBITC, 4-oxo-4-(3-pyridyl)bulyl isothiocyanate; HITC, 1-hexyl isothiocyanate; 2HITC, 2-hexyl isothiocyanate; DDITC, 1-dodecylisothiocyanate; 12DPEITC, 1,2-diphenylelhyl isothiocyanate; 22DPEITC, 2,2-diphenyl

ethyl isothiocyanate; GSH. glutathione; HPLC. high performance liquid chromalography;DCI, desorption chemical ionization; NMR, nuclear magnetic resonance; m, multiplet; t,triplet.

4327

Research. on December 7, 2018. © 1994 American Association for Cancercancerres.aacrjournals.org Downloaded from

http://cancerres.aacrjournals.org/

CHEMOPREVENTIVE ACTIVITIES. LIPOPIIIIJCITY. AND REACTIVITY

order to establish a convenient guide for predicting their inhibitoryactivities against lung tumorigenesis (Figure 1).

MATERIALS AND METHODS

Animals. Female A/J mice were obtained from The Jackson Laboratory(Bar Harbor, ME). The mice were fed the semipurified diet AIN-76A with 5%

corn oil (Dyets, Bethlehem, PA) and were maintained under the followingstandard conditions: 20 Â±2Â°C(SD), 50 Â±10% relative humidity, and a 12-h

light. 12-h dark cycle. Animals were used in bioassay at 7 weeks of age after

2 weeks of quarantine.Instrumentation. A HPLC system equipped with an automatic gradient

controller, two Model 501 pumps, and a Waters 990 photodiode array detectorin conjunction with a 4.6 x 250-mm Partisi! 5 ODS-3 reverse-phase C|K

analytical column (Whatman, Clifton, NJ) was used in the determination of thecapacity factor k' of isothiocyanates. In the kinetic studies, we used a Hewlett-

Packard 8452A diode array spectrophotometer (Hewlett-Packard, Palo Alto,CA) interfaced with a computer and the MS-DOS UV/VIS software. NMRspectra were recorded on a Bruker AM 360 WB spectrometer (Bruker, Bil-lerica, MA) using CDCI, as solvent. DCI-mass spectra were obtained on aHewlett-Packard 5988A mass spectrometer.

Chemicals. AITC, BITC, P1TC, and PEITC of highest purity were purchased from Aldrich Chemical Co. (Milwaukee, WI). PPITC, 8-phenyloctylchloride, and 10-phenyldecyl chloride were obtained from Fairfield Chemical

Co. (Blythewood, SC). HITC, 2HITC, DDITC, 12DPEITC, and 22DPEITCwith greater than 98% purity were obtained from Trans World Chemicals(Rockville, MD). GSH (reduced form) was purchased from Sigma (St. Louis,MO). PBITC, PPelTC, OPBITC, PHITC, and the tobacco-specific lung car

cinogen NNK were previously synthesized and characterized in our laboratories (12, 19,20).

POITC and PDITC were newly synthesized. The general procedures forsynthesizing isothiocyanates through the reduction of azide and reacting theresulting amine with thiophosgene were used (19). POITC and PDITC wereboth obtained with HPLC purity greater than 98%. The spectral data are asfollows: the DCI-mass spectrometry of POITC showed a base peak at mie 248,

(CH2)n N C

n=0 . P lTC

1 , BITC

2 , PE ITC

3, PPITC

4 , PBITC

5, PPelTC

6 , PH ITC N=C=:=S

POI TC8

10, PDITC

O

H lTC

2H ITC

DDITC

OPB ITC

N ~OiN

CHj

NNK

12DPEITC

which is consistent with the M+ 1 of the molecular formula C|5H,|NS (calculated, 247.405). 'H-NMR (360 MHz, CDCI,, 6 in ppm, tetramethylsilane as

internal reference): 0 1.30-1.45 (m, 8H), 1.55 - 1.75 (m, 4H), 2.60 (t, J = 7.6Hz, 2H), 3.50 (t, J = 6.6 Hz, 2H), 7.18-7.32 (m, 5H). "C-NMR (92.52 MHz,

CDCI,, S in ppm, tetramethylsilane as internal reference): 8 26.51, 28.70,29.11, 29.22, 29.93, 31.39, 35.92, 45.03, 125.59, 128.23, 128.38, 129.61(-NCS), 142.75. The DCI-mass spectrometry of PDITC showed a base peak at

mie 276, which is consistent with the M + l of the molecular formulaC^H^NS (calculated, 275.4ft). 'H-NMR: Â«1.30-1.45 (m, 12H), 1.55-1.75

(m, 8H), 2.70 (t, J = 7.6, 2H), 3.5 (t, J = 6.6, 2H), 7.18-7.32 (m, 5H).13C-NMR: 26.54, 28.77, 29.27, 29.35, 29.40, 29.40, 29.94, 31.49, 35.96, 45.05,

125.56, 128.21, 128.38, 129.59 (-NCS), 142.87.Pulmonary Adenoma Bioassay in the A/J Mouse. The single-dose model

for NNK lung tumor induction in the A/J mouse was used in this study (21).As an extension of the previous bioassay with PHITC, POITC, and PDITCwere examined with reference to PHITC to determine the optimal chain lengthin arylalkyl isothiocyanates for maximal potency (12). The doses chosen werebased on the lowest dose (0.2 /unni i for PHITC (the most potent inhibitor

among the previously tested arylalkyl isothiocyanates) to be effective. In thesecond bioassay, six other isothiocyanates of various structures were assayedusing PEITC as a reference compound at doses tested previously (3). Groupsof 20 A/J mice were each given isothiocyanate by gavage at a dose of 5, 1, 0.2,

0.1, or 0.04 fimol/mouse depending upon the isothiocyanate tested. Two hlater, mice were given 10 /xmol NNK (in 0.1 ml saline) by i. p. injection. Micein control groups were treated with either corn oil alone or isothiocyanate in

corn oil, followed by i.p. injection of saline. Sixteen weeks after these treatments the mice were killed, and pulmonary adenomas were counted. Thestatistical significance of bioassay data was determined by Student's t test.

Measurement of Lipophilicity by HPLC. The fully silanized WhatmanPartisil ODS-3 C,s analytical column was used under the same HPLC condi

tions for all isothiocyanates to minimize variability. The mobile phase consistsof a mixture of 70% acetonitrile and 30% water eluting at a flow rate of 1ml/min. Sample solutions of isothiocyanates were prepared by dissolving 2 ftl

of each isothiocyanate in 1 ml of acetonitrile with 0.1% formamide. Ten fil ofeach sample were used for each HPLC analysis. The peak of formamide (logP = - 1.69) was used as an unretained reference peak (22). The capacity factork' was calculated as follows:

K' = (tg - Â«â€ž)//,â€ž

where Â¡Kis the retention time of the isothiocyanate and /â€žis the retention timeof the unrctaincd peak. For isothiocyanates with known log P values, a linearcorrelation between log P and log k' was obtained with a correlation coefficient of 0.97 (Fig. 2), log P = 2.322 + 3.675 log k'. With this linear equation,the log P values of all isothiocyanates were calculated from their k' values.

Measurement of the First Order Reaction Rate Constants of Isothiocyanates. A stock solution of each isothiocyanate in acctonitrile (10 mM) wasprepared. GSH solution (10 mM) was freshly prepared by dissolving 30.7 mgof GSH (reduced form) in a 10-ml mixture of 0.2 M phosphate buffer (pH 7.4)

and methanol (1:1). Both solvents were degassed under vacuum before use.This high concentration of methanol is necessary to ensure miscibility for thosehighly lipophilic isothiocyanates. In each UV cell, 1 ml of GSH solution wasadded and a UV spectrum was recorded as a background. After addition of 10fil of isothiocyanate solution to the UV cell, the product formation wasmonitored by the programmed UV scans with selected time intervals. Thechange of the absorbance at wavelength 270 nm, an absorption wavelength forthe conjugate at which there is essentially no absorption for isothiocyanates,except for PITC, was used for calculations of the observed first order reactionrate constants Â¿obb.In the case of PITC, absorbance at 300 nm was used. Thefollowing equation was fitted with observed data by using the built-in regres

sion method.

Fig. 1. Structures of isothiocyanates and NNK.

where A(>,A,, and Aa are the absorbance at time zero, during the run, and at theend of the run, respectively. Three separate measurements were made for eachisothiocyanate.

4328



CHEMOPREVENTIVE ACTIVITIES, LIPOPHILICITY, AND REACTIVITY

RESULTS

Inhibition of NNK-induced Pulmonary Adenoma by LongChain Arylalkyl Isothiocyanates. POITCand PDITCwere tested inthis assay at 0.2, 0.1, and 0.04 /nmol/mouse. PHITC was included asa positive control because our previous studies had shown that 0.2jLimolof PHITC was inhibitory (12). The tumor multiplicities andincidences are summarized in Table 1. During the bioassay, all groupsshowed normal body weight gains. The mice in the control groupspretreated with 0.2 /xmol of isothiocyanates or with corn oil prior tosaline administration developed only background tumors normallyfound in the A/J mouse; in some cases, no tumor was found. Micepretreated with corn oil, prior to NNK administration, developed10.2 Â±0.6 tumors/mouse with an incidence of 100%. These responses

5.0

Table 2 Effects of synthetic and naturally occurring i.mthiocyanates an NNK-inducedlung lumorigenicity"

4.5

4.0

O) Q rO 3.5

3.0

2.5

2.0

PPelTC HITC

PBITC

PPITC

PE ITC

-0.1 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Log k'

Fig. 2. Plot of log P against log k' for isothiocyanates. A linear correlation between thetwo parameters was obtained: log P = 2.322 + 3.675 log k' (r = 0.97).

Table 1 Effects of synthetic long chain arylalkyl isothiocyanates on NNK-induced lungtumorigenicity"

DoseTreatment(^mol/mouse)Corn

oil/NNKPHITC/NNKPOITC/NNKPDITC/NNKPOITC/salinePDITC/salineCorn

oil/saline0.2

0.10.040.2

0.10.040.2

0.10.040.20.2Tumor

Tumor incidenceNo. of animals multiplicity*(%)r4020

20202019"2020

20201(1101010.2Â±0.65.2Â±oy5.5Â±0.6''10.9Â±0.93.3Â±0.5d

7.9Â±0./9.0Â±0.62.8Â±0.6'/

6.3Â±0.81'7.9Â±0.6/'Q.2Â±Q.ld0.0Â±0.0rfOÃŸÂ±0.(fl100100

100100<*/1001007(/

100100Iff</(/

a Tested in a single-dose model; experimental details given in the text.h Mean Â±SE; those significantly different from corn oil/NNK control according to

Student's I test are denoted by superscripts.' Tumor incidence was compared between each treatment group versus the corn

oil/NNK group using a procedure analogous to the Dunnett test for multiple comparison.''/>< 0.001.'' One accidental death occurred during the experiment.!P < 0.05.

DoseTreatment(/xmol/mouse)Corn

oil/NNKPEITC/NNK1AITC/NNK''

51HITC/NNK

512HITC/NNK

51DDITC/NNK

5122DPEITC/NNK

5112DPEITC/NNK

51AITC/saline''

5HITC/saline5

2HITC/saline 5DDITC/saline 522DPEITC/saline512DPEITC/saline5Corn

oil/salineTumor

TumorNo. of animals multiplicity'' incidence(%)'40202019"2(12(\2(1

19"202(1202020201010

10

10

10III10ll.lÂ±0.710.2Â±0.99.5

Â±1.211.6Â±1.0i.sÂ±o.y4.6Â±OV0.3

Â±O.I71.6Â±0.4/0.3Â±0.1/

0.6Â±0.2/0.6Â±0.2/'i.iÂ±o.y0.5

Â±O.I70.4Â±0./0.2Â±0.2/0.2Â±0.1/

0.0Â±0.0^o.oÂ±o.o''0.6Â±0.270.3Â±0./10010010010075"10030*70s25*

35*45*

60*45*25*10*30*

20*

1150*20*

a Tested in a single-dose model; experimental details given in the text.' Mean Â±SE; those significantly different from corn oil/NNK control according to

Student's t tests are denoted by superscripts.' Tumor incidence was compared between each treatment group versus the corn

oil/NNK group using a procedure analogous to the DunneÂ»test for multiple comparisons.Tested with the same protocol in a separate bioassay. Positive control (corn oil/

NNK): 9.2 Â±0.6 nodules/mouse, 100% incidence. Negative control (corn oil/saline):1.0 Â±0.4 nodules/mouse, 50% incidence.

" One accidental death occurred during the experiment.1'P < 0.001.

x P <0.05.

are similar to those observed previously (19). All three isothiocyanatescaused significant reduction of this tumor multiplicity, except for PHITC andPOITC at the dose level of 0.04 p,mol/mouse. Previously we have shown thatPHJTC was the most potent inhibitor among all the arylalkyl isothiocyanatestested (12) and that it is about 50-100 times more potent than the naturally

occurring PEITC (23). However, at 0.2 /unol/mouse in the current experiment, longer chain homologues POITC and PDITC appear to be more potentthan PHITC (P < 0.05). Such a trend is not obvious at the lower doses. At thelowest dose, only PDITC showed significant reduction of tumor multiplicity.Both POITC and PDITC caused a clear dose-dependent inhibition, but they

are not significantly different in their potencies.Inhibition of NNK-induced Pulmonary Adenoma in A/J Mice

by Isothiocyanates of Other Structural Features. Six isothiocyanates were tested together with PEITC in this bioassay at twodifferent dose levels, 5 and 1 /nmol/mouse. The tumor multiplicityand tumor incidence in all treated groups are summarized in Table2. In these experiments, all animal groups showed body weightgains comparable to those of the control groups. The controlgroups pretreated with the highest dose of isothiocyanates (5/nmol/mouse) or with corn oil prior to saline administration developed no tumors or only background tumors. As expected, micepretreated with corn oil, prior to NNK administration, developed11.1 Â±0.7 tumors/mouse with an incidence of 100%. Most of theisothiocyanates caused significant reduction of this tumor multiplicity at both dose levels. AITC, a main ingredient in brownmustard, was inactive at both doses. In agreement with previousresults, PEITC was shown to be effective at 5 jxmol/mouse orhigher doses, but not at 1 jxmol/mouse (12).

4329



CHEMOPREVENTIVE ACTIVITIES, LIPOPHILICITY, AND REACTIVITY

Table 3 Partition coefficients (log P) and pseudo first order rate constants <kobs)ofisolhiocyanates

IsothiocyanatesAITCPITCBITCPEITCPPITCPBITCPPelTCPHITCPOITCPDITCHITC2HITCDDITC12DPEITC22DPEITCOPBITCk'

=

(IK -'uV'o0.991.951.501.602.262.683.554.778.8916.762.832.7520.343.332.630.72LogPLog

k'(calculated)-0.010.290.180.210.360.430.550.680.951.220.450.441.310.520.42-0.152.303.392.973.083.623.904.344.825.816.823.983.947.134.243.871.79LogP(literature)2.11*3.28*2.83*3.4C3.8C4.0C4.4C4.6CVi(X10^sf3.58Â±0.524.60Â±0.402.19Â±0.432.44Â±0.351.14Â±0.271.31Â±0.050.80

Â±0.020.79Â±0.080.56Â±0.020.42Â±0.101.32Â±0.230.72

Â±0.020.32Â±0.052.40Â±0.142.28Â±0.381.12Â±0.16

" Mean Â±SE, based on three separate measurements." From Ref. 26.cFrom Ref. 12.

In the aliphatic isothiocyanate series, HITC at 5 Â¿Â¿mol/mousewasvery effective in reducing both tumor multiplicity and incidence. Atthe lower dose, where PEITC is not effective, HITC still showed astrong inhibition with an approximately 58% reduction of tumormultiplicity. As predicted on the basis of the trend observed in thearylalkyl isothiocyanate series, its longer chain homologue DDITCshowed a much stronger inhibition, reducing both tumor multiplicityand incidence essentially to the background levels. Interestingly,2HITC, a branched structural isomer of HITC, caused greater inhibition than HITC at both dose levels (P < 0.001). This provides anexample of a secondary isothiocyanate that has an improved efficacyover the primary isothiocyanate as an inhibitor of NNK-induced lung

tumorigenesis in A/J mice. These results also show for the first timethat the phenyl group in the arylalkyl isothiocyanate series is notcritical for tumor inhibition.

In a comparison of 12DPEITC and 22DPEITC with PEITC, bothanalogues showed significantly greater inhibitory potency than PEITC(P < 0.001). 12DPEITC was more inhibitory than 22DPEITC at 1/xmol/mouse (P < 0.05). However, at 5 /j.mol, the difference in theirpotency is not obvious.

Lipophilicity of Isothiocyanates. The capacity factor k' for 16

isothiocyanates studied in the present and previous works was obtained using a single HPLC system (details in "Materials and Methods"). A list of measured k' values and log P (partition coefficientsbetween 1-octanol and aqueous phases) calculated on the basis of k'values is presented in Table 3. Since HPLC retention time and k'

values have been used to obtain log P and, in some cases, the retentiontimes of drugs are directly correlated with biological activity (22, 24,25), it seems reasonable to use the log P values as a measure ofrelative lipophilicity of isothiocyanates. The lipophilicity scale measured by these log P values for isothiocyanates is consistent with thatpredicted on the basis of chemical structures. For instance, the increase of lipophilic property of arylalkyl isothiocyanates with alkylchain length is well reflected by log P values. The change Alog P/CH2is about 0.5 for longer chain analogues [(CH2)4 to (CH2)10], which isin agreement with the trend observed for fatty acids and alcoholhomologues with Alog P/CH2, being 0.53 and 0.52, respectively (26).For HITC and DDITC, the Alog P/CH2 = (7.13 - 3.98)/6 = 0.52,

also falls into the same range. The log P values for 2HITC and12DPEITC are relatively smaller than those for HITC and 22DPEITC,respectively. This also agrees with the trend that the branched isomershave smaller log P values than their unbranched homologues (26).

Chemical Reactivity of Isothiocyanates toward GSH. Since theconcentration of GSH used in these reactions was 10 rriMwhich is 100

times greater than the initial concentration of isothiocyanates, it can beconsidered to be constant during the course of its reaction. Thus, onemay treat it as a pseudo first order reaction with respect to theconcentration of isothiocyanates. The experimentally determined reaction rate constants for the interaction of isothiocyanates and GSHare listed in Table 3. The results clearly show that all isothiocyanatesare reactive toward GSH. Under the conditions used, the reactionsusually completed within 15 min. The conjugation between isothiocyanates and the sulfhydryl group of GSH appears to be the predominant reaction. This is evidenced by the appearance of the characteristic UV spectra of the products, the characteristic 13C-NMR chemicalshifts of the products,4 and the fact that isothiocyanates react withthiol groups 103-104 times faster than with amino or hydroxyl groups

(27). Among the isothiocyanates examined, PITC and AITC were themost reactive. In the arylalkyl isothiocyanate series, the rate constantsdecreased with increasing alkyl chain length. The rate constantsdropped significantly when the alkyl chain length reaches 3 carbons orlonger; thus, POITC and PDITC appeared to be the least reactiveones. Similarly, alkyl isothiocyanates with longer alkyl chains showedlower fcabs.The Â£obsof the longer chain isothiocyanate DDITC isabout one-fourth of that of its shorter chain homologue HITC. Inter

estingly, the structural isomers HITC and 2HITC have an alkyl chainwith same number of carbons but HITC reacted with GSH twice asfast as its isomer, 2HITC, which bears a secondary isothiocyanategroup. The diphenylethyl isothiocyanates, 12DDPEITC and22DPEITC, had rate constants similar to those of PEITC and BITC.The second phenyl group at either position 1 or position 2 did notsignificantly alter the chemical reactivity. In summary, both the longeralkyl chain and the secondary alkyl isothiocyanates examined by ushad lower reactivity toward the thiol group of GSH. Their rateconstants may differ by as much as one order of magnitude due tosteric and electronic influences.

Correlation of Lung Tumor-inhibitory Activities with Lipophilicity and Reactivity of Isothiocyanates. To assist structure-activity

relationship study, the relative potency of all 16 isothiocyanates in thelung tumorigenesis assays in A/J mice was rated on the basis of theirpercentile inhibition of NNK-induced tumors at 3 doses tested,

namely 5, 1, and 0.2 /xmol/mouse (Table 4). To further delineate therelationships of inhibitory potency with the physical and chemicalproperties of isothiocyanates, the relative inhibitory potencies areplotted against log P and Â£obsobtained in this study (Fig. 3). The plotcan be divided into three regions. The first region is the one locatedat the bottom of the right side of the plot and is defined by low log Pvalues, as seen for compounds such as AITC, PITC, BITC, PEITC,and OPBITC. Some of these isothiocyanates such as AITC, PITC, andBITC are inactive even at the highest doses, 5 or 25 jxmol/mouse (27).Most of these compounds are characterized by high reactivity and, inparticular, low lipophilicity. For example, OPBITC has the lowest logP value among these 16 isothiocyanates. The second region, in themiddle area of the plot, encircles compounds with modest log P and^obs- Most isothiocyanates with modest to strong tumor inhibitorpotency reside in this region, such as arylalkyl isothiocyanates withmedium alkyl chain length or diphenylethyl isothiocyanates and hexylisothiocyanates. The front left corner of the plot, where compoundsare associated with large log P and small Â£obsis the third region andis occupied by highly potent inhibitors such as PHITC, POITC,PDITC, and DDITC.

4 D. Jiao and F-L. Chung, unpublished data.

4330



CHEMOPREVENTIVEACTIVITIES, LIPOPHILICITY,AND REACTIVITY

Table 4 Relative inhibitory potencies of isothiocyanates against NNK-induced lung tumorigenicity in A/J mouse

% of inhibition"

5 /Â¿mol 1 /Ã•IlHll 0.2 /Â¿mol

CompoundAITCPITCB1TCOPBITCPEITCPPITCPBITCPPelTCPHITCPOITCPDITC12DPEITC22DPEITCHITC2HITCDDITCMultiplicity00014769898%99NDND9594859797Incidence00II411m897595NDND5555257(175Multiplicity0NDNDND085911mi100NDNDÂ«X,mi51Xh95IncidenceÃœNDNDND0254253100NDND754003065MultiplicityND1NDNDND05347fi2856X73NDNDNDNDNDIncidencepotency*NDNDNDND0

+10+ ++0+ ++0+ ++30

+ + ++10+ + ++30+ + ++ND+ + ++ND+ ++ND++ND

+ ++ND+ + + +

" Data are compiled from the present and the previous studies (Refs. 12, 19, and 28). Data for PITC, BITC, OPBITC, PEITC, PPITC, PBITC. PPelTC, and PHITC are from previous

studies.6 The rating of the inhibitory potency is based on the percentage of inhibition. - and +, negative and positive inhibitory activity, respectively. -, inhibition of tumor multiplicity

<20% at 5 /i nini +, inhibition of tumor multiplicity <80% at 5 /Â«noibut no inhibition at the lower doses; + +, inhibition of tumor multiplicity <90% at 5 /Â¿moland <60% at 1 /Â¿mol;+ + +, inhibition of tumor multiplicity >90% at 5 /Â¿moland <9()% at 1 /Â¿mol;+ + + +, inhibition of tumor multiplicity >9()% at 5 /Â¿molor >65% at 0.2 /Â¿mol.In some cases, theinhibition of tumor incidence is also taken into consideration in the rating.

' ND. not determined.

DISCUSSION

The selection of isothiocyanates for the bioassays in this study wasbased on the following rationales. The evaluation of arylalkyl isothiocyanates with longer alkyl chains (C8 and C,,,), together with resultsfrom previously tested isothiocyanates, may provide informationabout the optimal chain length required for maximal inhibition. Sincelipophilicity appears to be an important factor in the inhibition ofNNK-induced lung tumorigenesis by arylalkyl isothiocyanates, several lipophilic isothiocyanates without aryl moiety, such as HITC,2HITC, and DDITC, were also tested in this study. Thus, we wereable to examine whether the phenyl group is essential for the inhib-

Fig. 3. Plot of relative inhibitory potency of isothiocyanates against log P and *l)b,. Forconvenience, the relative inhibitory potency is expressed as 0 (â€”), 1 (+), 2 (++), 3

( + + +), and 4 ( + + + + ). Three regions of the plot, upper front left corner, the middle andthe lower right region, are associated with high, intermediate, and low potency ofisothiocyanates. See text for detailed discussion.

itory activity. These evaluations are expected to provide alternatestructure leads for future studies. While previously tested isothiocyanates are primary isothiocyanates, a secondary isothiocyanate, 2HITC(a structural isomer of HITC), was tested for the first time. Since amajor route of metabolism for isothiocyanates is via conjugation withGSH, the chemical reactivity of isothiocyanates toward GSH wouldconceivably be an important factor in determining their potency (16,17, 29, 30). The secondary isothiocyanate group is expected to be lessamenable to conjugation because of greater steric hindrance andelectron-donating effect due to the a-substitution. For similar reasons,the biphenyl analogues of PEITC, 12DPEITC and 22DPEITC, weretested in this bioassay and were also found to be more lipophilic thanPEITC. Finally, AITC occurs naturally in many of the vegetables,food additives and flavoring agents consumed by humans, and it is astrong phase II enzyme inducer (31-33).

Our earlier data had suggested that the tumor-inhibitory potency ofisothiocyanates appears to be associated with increased lipophilicityand chemical stability (12). Now we have obtained further evidencethat an increase of the alkyl chain length diminishes the reaction ratesof isothiocyanates toward GSH. The reactivity of the NCS functionalgroup toward nucleophiles declines with an increase of electrondensity around the carbon atom of the NCS group. The observed lowreactivity of long chain arylalkyl isothiocyanates may be explained bythe reduced electron-withdrawing effect of the phenyl ring along thealkyl chain. In some instances, the high lipophilicity or low reactivityalone may enhance the inhibitory activity of the isothiocyanate. Forexample, HITC and 2HITC have similar log P values, 2HITC beingslightly less lipophilic (Table 3). However, the reaction rate constantof 2HITC is only one-half that of HITC, which is likely related to thesignificantly stronger inhibitory activity of 2HITC. In another case,PEITC, 22DPEITC, and 12DPEITC all have similar reaction rateconstants (Table 3), but their tumor-inhibitory activity is in the order:PEITC < 22DPEITC < 12DPEITC. This trend is consistent with theorder of their log P values. However, with low lipophilicity and lowreactivity, OPBITC is yet an inactive compound. Evidence obtainedthus far strongly supports the notion that the chemopreventive efficacy of isothiocyanates is mainly determined by structural characteristics associated with high lipophilicity and low reactivity. Thus,future studies of isothiocyanates as inhibitors of NNK-induced lung

4331



CHEMOPREVENTIVE ACTIVITIES, LIPOPH1LICTTY, AND REACTIVITY

tumorigenesis should be focused on various lipophilic isothiocyanateswith low reactivity. For example, it has been reported that the rateconstants of the reaction of ^-substituted aromatic isothiocyanates

with thioglycolate are correlated with the Hammett constant cr of thesubstituents (34). It would be interesting to test compounds withelectron-donating groups, such as a methyl or methoxy group at the p

position of the phenyl moiety which will result in low reactivity ofisothiocyanates.

In the tumor bioassay, the reactive isothiocyanates with low li-

pophilicity, such as AITC, BITC and PITC, were poor inhibitors.These compounds interact readily with nuleophiles in vivo; consequently little of them reaches the target cytochrome P-450s. As shown

in this study, these isothiocyanates react with GSH at a substantiallyhigh rate even without glutathione S-transferases. It is possible thattheir reactions with functional proteins or other important biomol-

ecules are the basis for their toxicity. In previous studies, PITC givenat 25 jumol/mouse for two consecutive daily administrations resultedin a 100% mortality, while BITC caused nearly 25% (28). In contrast,PEITC was not toxic in A/J mice after four consecutive daily administrations at the same dose (28). The long chain analogue PHITC wasfound to be even less toxic than PEITC in rats in a short term acutetoxicity study.5

Evidence obtained from in vitro and in vivo studies indicates thatthe inhibition of NNK-induced lung tumorigenesis by arylalkyl iso

thiocyanates is mainly due to inhibition and inactivation of cytochrome P-450 enzymes that are required for NNK activation (13-15).

The inhibition of microsomal NNK metabolism in lung tissue byarylalkyl isothiocyanates appears to involve two types of mechanisms,namely competitive and noncompetitive components (13-15). The

competitive inhibition is reversible; whereas the noncompetitive inhibition is probably due to irreversible inactivation of the enzymes.The relative importance of these mechanisms depends on the time ofexposure, dose, and structures of isothiocyanates (13, 15). PHITC is avery potent competitive inhibitor of NNK activation with an apparentKÂ¡of 11-16 nM, which is considerably more potent that PEITC (15),

and is among the most potent competitive inhibitors known forcytochrome P-450 enzymes. The chemical reactivity of isothiocya

nates, therefore, is not a main determinant of enzyme binding, sincePHITC is much less reactive than PEITC. The binding of isothiocyanates with cytochrome P-450 isozymes appears to be governed largely

by lipophilicity, and it has long been speculated that the active site ofthe mammalian cytochrome P-450 is hydrophobic and that hydropho-

bicity is a major factor in the binding of substrates to cytochromeP-450 (35). Although the crystal structures of mammalian cytochromeP-450 isozymes have not yet been available, the X-ray crystallographystudies of the two bacterial P-450 enzymes revealed a long hydro-

phobic tunnel leading to the active site (36, 37). This pocket iscomposed of nonaromatic hydrophobic residues, which was considered to be commonly present in most of mammalian cytochromeP-450 isozymes (37, 38). It is reasonable to assume that this hydro-

phobic pocket may harbor the binding sites for the isothiocyanates.The correlation between the inhibitory activity and the lipophilicity ofthese chemopreventive agents can be rationalized at least in part bythe interaction of the isothiocyanates with this hydrophobic bindingpocket on the enzymes. Besides enzyme binding, the increased lipophilicity of certain isothiocyanates may also facilitate the absorption and tissue disposition in vivo.

Based on the cytochrome P-450 enzyme binding site model, it

would be tempting to address the question regarding the geometricrequirements for binding with isothiocyanates. However, this is com

plicated by the fact that there are multiple isozymes involved in NNKmetabolism. For instance, at least cytochrome P-450 2A1 and 2B1

have been shown to be partially responsible for NNK metabolism inthe lung of the A/J mouse (13). The binding specificity of isothiocyanates toward cytochrome P-450 isozymes must be examined in order

to assess the potential of isothiocyanates to inhibit bioactivation ofother carcinogens. In this context, the isothiocyanates of variousstructural features will be useful probes for studying cytochromeP-450 binding site structures and their substrate recognition.

ACKNOWLEDGMENTS

We thank C. S. Yang for helpful discussions.

REFERENCES

16.

17.

18.

19.

20.

5F-L.Chung,unpublisheddata. 21.

4332

Wattenberg. L. W. Inhibition of carcinogenic effects of polycyclic hydrocarbons bybenzyl isothiocyanate and related compounds. J. Nail. Cancer Inst.. 58: 395-398,

1978.Wattenberg, L. W. Inhibitory effects of benzyl isothiocyanate administered shortlybefore diethylnitrosamines or benzo(a)pyrene on pulmonary and forestomach neoplasia in A/J mice. Carcinogenesis (Lond.), 8: 1971-1973, 1987.Morse, M. A., Wang, C-X., Stoner, G. D., Mandai. S.. Conran, P. B., Amin, S. G.,Hecht, S. S., and Chung, F-L. Inhibition of 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone-induced DNA adduci formation and tumorigenicity in the lung of F344 ratsby dietary phenethyl isothiocyanate. Cancer Res., 49: 549-553, 1989.Morse, M. A., Reinhardt, J. C, Amin, S. G., Hecht, S. S., Stoner, G. D., and Chung,F-L. Effect of dietary aromatic isothiocyanates fed subsequent to the administrationof 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone on lung tumorigenicity in mice.Cancer Lett.. 49: 225-230, 1990.Stoner, G. D.. Morrissey, D. T.. Heur, Y-H., Daniel, E. M., Calati, A. J.. and Wagner,S. A. Inhibitory effects of phenethyl isothiocyanate on A/-nitrosobenzylmethylaminecarcinogenesis in rat esophagus. Cancer Res., 5/: 2063-2068, 1991.Adam-Rodwell, G., Morse, M. A., and Stoner, G. D. The effects of phenethylisothiocyanate on benzo[u]pyrene-induced tumors and DNA adducts in A/J mouselung. Cancer Lett., 71: 35-42. 1993.Shopland. D. R., Eyre, H. J., and Pechacek. T. F. Smoking-attributable cancermortality in 1991: is lung cancer now the leading cause of death among smokers inthe United States? J. Nail. Cancer Inst., 83: 1142-1148. 1991.Chung. F-L. Chemoprcvention of lung Carcinogenesis by aromatic isothiocyanates.In: L. Wattenberg, M. Lipkin. C. W. Boone, and G. J. Kelloff (eds.). CancerChcmoprevention. pp. 227-245. Boca Raton, FL: CRC Press, Inc., 1992.Wattenberg, L. W. Chemoprevention of cancer. Cancer Res., 45: 1-8, 1985.Sporn, M. B. Chemoprevention of cancer. Lancet (N. Am. Ed.), 342: 1211-1213,1993.Kelloff, G. J., Boone, C. W., Crowell, J. A., Steele, V. E., Lubet, R., and Sigman,C. C. Chemopreventive drug development: perspectives and progress. Cancer Epi-demiol. Biomarkers Prev., 3: 85-98, 1994.Morse, M. A., Eklind, K. I., Hecht, S. S.. Jordan. K. G., Choi, C-I., Desai, D. H.,Amin, S. G.. and Chung. F-L. Structure-activity relationships for inhibition of4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone lung tumorigenesis by arylalkyl isothiocyanates in A/J mice. Cancer Res., 57: 1846-1850, 1991.Smith. T. J., Guo, Z., Thomas, P. E., Chung, F-L., Morse, M. A., Eklind, K., andYang, C. S. Metabolism of 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone in mouselung microsomes and its inhibition by isothiocyanates. Cancer Res., 50: 6817-6822,1990.Guo, Z., Smith, T. J., Wang, E., Sadrieh, N.. Ma, Q., Thomas, P. E.. and Yang, C. S.Effects of phenyl isolhiocyanate, a Carcinogenesis inhibitor, on xenobiotic-metabo-lizing enzymes and nitrosamine metabolism in rats. Carcinogenesis (Lond.), 13:2205-2210, 1992.Guo, Z., Smith, T. J., Wang, E., Eklind, K. I., Chung, F-L., and Yang, C. S.Structure-activity relationships of arylalkyl isolhiocyanates for the inhibition of4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone: metabolism and the modulation ofxenobiotic-metabolizing enzymes in rats and mice. Carcinogenesis (Lond.), 14:1167-1173, 1993.

Brusewitz, G., Cameron, B. D., Chasseaud, L. F., GÃ¶rler,K., Hawkins, D. R.. Koch,H., and Mennicke, W. H. The metabolism of benzyl isothiocyanate and its cysteineconjugate. Biochem. J., 162: 99-107, 1977.Eklind. K. I.. Morse, M. A., and Chung, F-L. Distribution and metabolism of the

natural anticarcinogen phenethyl isothiocyanate in A/J mice. Carcinogenesis (Lond.),11: 2033-2036. 1990.Chung, F-L., Morse, M. A.. Eklind, K. I., and Lewis, J. Ouantitation of human uptake

of the anticarcinogen phenethyl isothiocyanate after a watercress meal. CancerEpidemiol. Biomarkers Prev., 1: 383-388. 1992.Morse, M. A.. Eklind. K. I.. Amin. S. G.. Hecht, S. S., and Chung, F-L. Effects ofarylalkyl chain length on the inhibition of NNK-induced lung neoplasia in A/J miceby arylalkyl isothiocyanates. Carcinogenesis (Lond.), 4: 1757-1759, 1989.Hecht, S. S., Lin, D., and Castonguay, A. Effects of a-deuterium substitution on themutagenicity of 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone (NNK). Carcinogenesis (Lond.), 4: 305-310, 1983.Hecht. S. S., Morse, M. A., Amin, S. G., Stoner, G. D., Jordan, K. G.. Choi, C. I., and



CHEMOPREVENTIVE ACTIVITIES, LIPOPHILICITY. AND REACTIVITY

Chung, F-L. Rapid single dose model for lung tumor induction in A/J mice by4-(methylnitrosamino)-l-(3-pyridyI)-l-butanone and the effect of diet. Carcinogene-sis (Lond.), 10: 1901-1904, 1989.

22. McCall. J. M. Liquid-liquid partition coefficients by high-pressure liquid chromatog-raphy. J. Med. Chem., Â¡8:549-552, 1975.

23. Chung, F-L., Morse, M. A., and Eklind, K. I. New potential chemopreventive agentsfor lung carcinogenesis of tobacco-specific nitrosamine, Cancer Res., 52 (Suppl.):2719s-2722s, 1992.

24. Baker, J. K., Rauls. D. O., and Borne, R. Correlation of biological activity andhigh-pressure Chromatographie retention index for a series of propranolol, barbiturate.and anthranilic acid analogues. J. Med. Chem., 22: 1301-1306, 1979.

25. Mirrlees, M. S., Moulton, S. J., Murphy, C. T., and Taylor, P. J. Direct measurementof octanol-water partition coefficients by high-pressure liquid chromatography.J. Med. Chem., 19: 615-619, 1976.

26. Leo, A., Mansch, C., and Elkins, D. Partition coefficients and their uses. Chem. Rev.,71: 525-616, 1971.

27. Drobnica, L., Kristian, P., and Augustin, J. The chemistry of the -NCS group. In: S.

Palai (ed.). The Chemistry of Cyanates and Their Thio Derivatives, Part 2, pp.1003-1221. New York: Wiley-Liss, Inc., 1977.

28. Morse, M. A., Amin, S. G., Hecht, S. S., and Chung, F-L. Effects of aromaticisothiocyanates on tumorigenicity, O6-melhylguanine formation, and metabolism of

the tobacco-specific nitrosamine 4-(methylnitrosamino)-l-(3-pyridyl)-l-butanone inA/J mouse lung. Cancer Res., 49: 2894-2897, 1989.

29. Mennicke, W. H., GÃ¶rler,K., Krumbiegel. G., Lorenz. D., and Rittmann, N. Studieson the metabolism and excretion of benzyl isothiocyanate in man. Xenobiotica, 78:441-447, 1988.

30. Jiao, D., Ho, C-T., Foiles. P., and Chung F-L. Identification and quantification of the/V-acetylcysteine conjugate of allyl isothiocyanate in human urine after ingestion of

mustard. Cancer Epidemiol. Biomarkers Prev., in press, 1994.31. Daxenbichler. M. E., Spencer, G. F., Carlson, D. G., Rose, G. B., Brinker, A. B., and

Powell. R. G. Glucosinolate composition of seeds from 297 species of wild plants.Phytochemistry (Oxf.), 30: 2623-2638, 1991.

32. Life Sciences Research Office. Evaluation of the Health Aspects of Mustard and Oilof Mustard as Food Ingredients, SCOGS-16. Bethesda, MD: Life Sciences Research

Office, 1975.33. HarÃ¡, A., Sakai, N., Yamada, H., Tanaka, T., Kalo, K., Mori, H., and Sato, K.

Induction of glutathione 5-transferase, placenta! type in T9 glioma cells by dibutyry-ladenosine 3',5'-cyclic monophosphate and modification of its expression by natu

rally occurring isothiocyanates. Acta Neuropathol., 79: 144-148, 1989.

34. Brobnica, L.. and Augustin, J. Reaction of isothiocyanates with amino acids, peptidesand proteins. III. Kinetics and mechanism of the reaction of aromatic isothiocyanateswith thioglycolic acid. Coll. Czechoslv. Chem. Commun., 30: 1618-1625, 1965.

35. Yang, C. S., and Lu, A. Y. H. The diversity of substrates for cytochrome P-450. In:f. P. Guengerich (ed.). Mammalian Cytochromes P-450. Vol. 2, pp. 1-17. Boca

Raton, FL: CRC Press, Inc., 1987.36. Poulos, T. L.. Finzel, B. C., and Howard, A. J. High-resolution crystal structure of

cytochrome P450cam. J. Mol. Biol., 795: 687-700, 1987.

37. Ravichandran. K. G., Boddupalli, S. S., Hasemann, C. A., Peterson, J. A., andDeisenhofer, J. Crystal structure of hemoprotein domain of P450BM-3, a prototypefor microsomal P450's. Science (Washington DC), 267: 731-736, 1993.

38. Poulos, T. L. Modeling of mammalian P450s on basis of P450cam X-ray structure.Methods Enzymol., 206: 11-31, 1992.

39. Gotoh, O. Substrate recognition sites in cytochrome P450 family 2 (CYP2) proteinsinferred from comparative analyses of amino acid and coding nucleotide sequences.J. Biol. Chem., 267: 83-90, 1992.

4333



1994;54:4327-4333. Cancer Res Ding Jiao, Karin I. Eklind, Chang-In Choi, et al. Tumorigenesis in A/J Mice4-(Methylnitrosamino)-1-(3-pyridyl)-1-butanone-induced LungMechanism-based Inhibitors of Structure-Activity Relationships of Isothiocyanates as

Updated version

http://cancerres.aacrjournals.org/content/54/16/4327

Access the most recent version of this article at:

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

Subscriptions

Reprints and

[email protected] at

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

Permissions

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

.http://cancerres.aacrjournals.org/content/54/16/4327To request permission to re-use all or part of this article, use this link



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

mailto:[email protected]



structure-activity relationships of isothiocyanates as mechanism

Documents