4627 ~

Fd Cosmer. To:dcol. Vol. 18. pp. 159 to 172 I){) 15.6264/80/0401·01 59502.1){);0 © Pergamon Press Ltd 1980. Printed in Great~Brjtain

TOXICOLOGY OF , RELATED COMPOUNDS (NITRILES, R-GOITRIN, ) AND VITAMIN U FOUND IN CRUCIFERAE

K. NISHIE us Department of Agriculture. Science and Education Administration, Russell Agricultural Research Center, P.O. Box 5677, Athens. GA 30604 and M. E. DA.XENBICHLER US Department of Agriculture, Science and Education Administra,.on, Northern Regional Research Center, 1815 N. University Street, Peoria, IL 61604, USA

(Received 9 August 1979)

Abstract-The cruciferous glucosinolates (epiprogoitrin, and ). nine of their derivatives (nitriles, isothiocyanates, R-goitrin) and vitamin U were not teratogenic to the rat foetus when adminis­ tered sc to the dams on days 8 and/or 9 of gestation. However, l-cyano-3,4-epithiobutane, 3-methylsul­ phinylpropyl and caused embryonal death and decreased foetal weight, whereas five other derivatives only decreased foetal weight. Under identical conditions, the reference teratogen, trypan blue, caused a significant incidence of malformations and resorptions. 3-Methylsulphinylpropyl isothiocyanate was the most toxic compound, followed by allyl isothiocyanate, l-cyano-3,4-epithiobutane (nephrotoxic) and S-1-cyano-2-hydroxy-3-butene. Some nitriles caused hyper­ plasia of the thyroid, an effect which often was more pronounced in nonpregnant than in pregnant rats.

INTRODUCTION brick & Hill, 1975; VanEtten, Daxenbichler & Wolff, 1969a; VanEtten, Gagne, Robbins, Booth, Daxen­ The long search for the cause of neural tube defects bichler & Wolff, 1969b). These adverse effects of the in humans has been described in a review by Leck feeds were attributed to the high content of glucosino­ (1974). In the 1960's, potato blight was suspected as lates and their derivatives, which include goitrin, iso­ the cause of foetal anencephaly and spina bifida cys­ and nitriles. Very little is known about tica in England and Wales. However, this hypothesis the toxicity of the individual compounds. was rejected because potato-free diets did not prevent In this study, the potential teratogenicity and acute these malformations (Renwick, 1974). Births recorded toxic actions of certain glucosinolates and derivatives from 1965 to 1972 in Sweden (Sandahl, 1977) showed are described. a significant increase in the incidence of anencephaly in foetuses conceived in March and of spina bifida in July. In South Africa, Singer, Nelson & Beighton EXPERIMENTAL (1978) reported an increase in central nervous system defects in foetuses conceived in May. The chemicals used in this study and tested for Certain plant foods are the prime suspects for the toxic and teratogenic action are listed in Table 1. seasonal rise in foetal malformations due to their Listed below each (epiprogoitrin, sini­ seasonal abundance and availability. In addition, grin and sinalbin) are its decomposition products. seasons and climate are known to influence the Owing to unavailability, was not chemical composition of plants. tested; but its derivatives, 3-indolylacetonitrile and Vegetables and seeds of the Cruciferae family, such -3-methanol were tested. Vitamin U was as crambe, kale, mustard, rape, , turnips, etc., included due to its coexistence with glucosinolates in are rich in glucosinolates (Ahmed, 1972; VanEtten, Cruciferae, and trypan blue was used as the reference 1969). Many feeds containing a high concentration of teratogenic compound. The chemistry of these gluco­ rape or crambe seed meal have been shown to de­ sinolates and their derivatives has been described by crease feed intake and growth rate and to cause Cole (1975, 1976), Daxenbichler, VanEtten, Brown & enlarged liver, kidneys, thyroid and adrenal glands Jones (1964), Daxenbichler, VanEtten & Spencer (Bell, Benjamin & Giovannetti, 1972; Clandinin & (1977), Daxenbichler, VanEtten & Wolff (1966, 1968), Bayly, 1960; Hussar & Bowland, 1959; Kennedy & Kirk & MacDonald (1974), VanEtten (1969) and Vir­ Purves, 1941; Kirk, Mustakas, Griffin & Booth, 1971; tanen (1965). Manns, Bowland, Mendel & Zivkovic, 1963; Nord­ Pregnant Holtzman rats weighing 250-300 g were feldt, Gellerstedt & Falkmer, 1954; Srivastava, Phil- used to test the potential teratogenic effect of the . compounds listed in Table 1. The animals were main­ *The mention of this product does not imply recommenda­ tained on commercially available Teklad* 4% mouse/ tion over similar products that may be available. rat diet (obtained from Mogul Corp., Winfield, IA)

159 '"o

Table I. Compollllds screelled/i)1' toxicity a/l(l terato!Jellicity

Chcmical Slruclure Source

Epiprogoitrin CH1 :CH,CHOH,CH 2,qS-glucose):N,OS03'Na/K Cl'ambe abyssillica (crambe) seed S-I-Cyano-2-hydroxy-3-butene CH 2:CI+CHOI+CH2'CN C. abyssillica seed

R-5- Vinyl-2-oxazolidinethione (R-goitrin) CH2:CH,CH'CH1'NH ,CS C. abyssillica secd L-o--l (" Z ;:;; I-Cyano-3,4-epithiobutane CH1,CH 'CH1,CH2,CN Bl'assica rapa L. :r: (tender grcen mustard) seed ;;:; LSJ f'l ::> 0- Sinigrin CH1:CH'CH1,qS-glucosc): N'OS03K'H20 Pfaltz & Bauer, Stamford, CT ~ Allyl isothiocyanatc CH2:CH,CH2'N:C:S Aldrieh Chemical Co" Milwaukee, WI rn t:I Sinalbin hydrate OH'C"H4 'CH1,qS-glucosc): N'OS03-sinapin Aldrich Chemical Co, ;p. >< p-Hydroxyphenylacetonitrile OH,C6 H4 'CH2'CN ICN Pharmaceuticals, Inc" m ~ Cleveland, OH n I-Cyano-3-methyl,sulphinylpropanc (iberin nitrilc) CH3'SO'CH 2,CH 2'Cl-1 1'CN Lesl/lIel'ella !Jol'dollii seed :r: e­ 3-Methylsulphinylpropylisothiocyanate (ibcrin) CH3-50,C1-1 2'CH2-cH2'NCS L. !/ol'dollii seed m

~CH -CN '" 3-lndolylacetonitrile ~ 2 Aldrich Chemical Co, H ~'CH2-0H Indole-3-methanol ~ Aldrich Chemical Co, H Vitamin U (cabagin) CH3'S(CH3)'C1-12'CH2,CH(NH1)'COOH ICN Pharmaceuticals Inc, Trypan blue (reference teratogcnic compound) Allied Chemical Corp" Morristown, NJ Toxicology of glucosinolates and derivatives 161

with 1~..~ iodized salt, and two rats were placed in each p-hydroxyphenylacetonitrile and l-cyano-3-methyl­ suspended wire-bottomed cage with an automatic sulphinylpropane (iberin nitrile), increased foetal flushing system to clean the cages, The compounds weights. From nine pregnant rats treated with iberin were dissolved in either propylene glycol or distilled nitrile (200 mg/kg, sc) on days 8 and 9 of pregnancy. water and administered sc or orally in volumes of ' only one embryo out of 87 had skeletal malformations, 0'5-1'Oml/kg, One dose/day was administered on day namely, talipes and curled tail. Alizarin stain revealed 8 and/or 9 of gestation, although trypan blue was also . that this malformed, underweight foetus (2'14 g) had given on day 7 of gestation, Normally, six pregnant less calcified tail and feet bones than did control (sperm positive) rats were used per dose, Six pregnant foetuses. rats from each shipment of 24-30 were treated sc with Table 2 shows that single-dose treatments of the either saline or propylene glycol and served as the reference teratogenic compound trypan blue control group, (100 mg/kg, sc) given on day 7 of pregnancy caused The treated and control rats were weighed daily, the highest rate of embryonal death (resorption) and and the weight gains or losses were calculated as per­ the lowest rate of malformations. Trypan blue injec­ centages of the body weight on day 3 of gestation, On tion on day 8 resulted in higher incidence of malfor­ day 20 of gestation (11-12 days after treatment), all mations ofthe head (cranioschisis, meningocele, hydro­ rats were anaesthetized with sodium pentobarbital, cephalus, unilateral-microphthalmia, -anophthalmia. and the foetuses and resorbed embryos were counted, -anotia, -agnathia). Injection on day 9 resulted in Each foetus was weighed and examined for external a higher incidence of malformations in the hind malformations, For examination of the skeletal mal­ quarters (spina bifida, curled tail, talipes, omphalo­ formations, the viscera and skin of one third of the cele, hernia etc.). When the same total amount of foetuses were removed; and, sequentially, the foetuses 100 mg trypan blue/kg was given in 2 days (50 mg/kg/ were fixed overnight in 95~~ alcohol, placed in 1~~ day), on days 8 and 9 of gestation, the incidence of KOH for 8 hr and then placed in 0,01 alizarin red, malformation was comparable to that caused by a according to Dawson (1926), The internal malforma­ single high dose (100 mg/kg) given either on day 8 or tions in two-thirds of the foetuses were examined after 9 of gestation. The high yield of malformed foetuses they had been fixed in Bouin's fluid and serially sec­ caused by one or two sc doses of trypan blue on day 8 tioned (Wilson, 1965), Also on day 20, following re­ and/or 9 of pregnancy indicate that these injection moval of the foetuses, the organs (liver, kidneys, adre­ days were suitable for tests to screen the potential nals and thyroids) of the dams were removed, weighed teratogenic action of the compounds listed in Table 1. and fixed in formalin solution for histological examin­ ation. Effects on the body weight of pregnant rats In order to find the approximate lethal doses of the Control pregnant rats had a weight gain of test compounds that were available in greater supply, 50·13 ± 9'2~j~ from day 3 to day 20 of pregnancy. relatively high single doses were administered sc to Nonpregnant female controls with almost identical male Sprague-Dawley rats (50-210 g). These treated initial body weights gained 9-4 ± 3'9~~ during the rats were weighed daily and observed for 12 days and same period. Two sc injections of epiprogoitrin, sini­ killed. The same organs were removed, weighed and grin, sinalbin hydrate, 1-cyano-3-methylsulphinyl­ fixed in formalin solution for histological examin­ propane or vitamin U on days 8 and 9 of gestation ation. had no significant effect on the weight gain of preg­ Periodic acid-Schiff and haematoxylin were used as nant rats (Table 2). 3-Indolylacetonitrile and indole-3­ the histological stains. Student's t test was used to methanol depressed weight gains from 2 to 3 days determine the significance of weight gains, foetal following injections. S-1-Cyano-2-hydroxy-3-butene, weights, litter size and organ weight changes caused R-goitrin, l-cyano-3,4-epithiobutane, allyl isothio­ by the test compounds. The chi-square test was used cyanate, and p-hydroxyphenylacetonitrile depressed to determine the significance of the difference in the weight gains from 5 to 12 days following sc injections. incidence of resorbed and malformed foetuses. This depression of weight gain was associated with a decrease in food intake. In addition, l-cyano-3,4­ RESULTS epithiobutane caused mild diarrhoea of short dur­ ation. Embryotoxicity and teratogenicity Of the 13 compounds tested, none was teratogenic Effects on relative organ weights to rat foetuses. However, seven compounds decreased Sinigrin (sc), sinalbin and trypan blue increased the foetal weight at certain doses (Table 2). Also, three of relative liver weight of pregnant rats, whereas epipro­ them at high doses increased embryonic death goitrin and R-goitrin had the same effect in male rats (resorption). In comparison with control foetuses, (Table 3). S-1-Cyano-2-hydroxy-3-butene decreased those with significantly lower weights were less devel­ the relative liver weight of pregnant rats. 1-Cyano-3,4­ oped, as indicated by the lower number of calcified epithiobutane was nephrotoxic and increased the small bones of the feet, tail and sternum. Generally, kidney weight to a greater extent in pregnant rats the control foetuses had"slx--small calcified hones than in male rats. 1-Cyano-3,4-epithiobutane, S-l­ forming the sternum; but underweight foetuses had cyano-2-hydroxy-3-butene and iberin increased the three to five bones calcified, and some of the bones adrenal weights of pregnant rats; and R-goitrin were split or abnormally shaped. increased the adrenal weights of nonpregnant rats. In Epiprogoitrin and S-1-cyano-2-hydroxy-3-butene this series of test compounds, only p-hydroxyphenyl­ increased or decreased foetal weight depending on acetonitrile increased thyroid weight significantly in dosage (Table 2). Three nitriles, 3-indolylacetonitrile, pregnant rats; sinigrin (orally) decreased the thyroid F.e.T. 18/2-D 0'> 1-,0 Table 2. Etfl'cts Orl//I/cosillo/atcs ami reill/ctl coml'oll/I/Is Oil tlams alltljiwlllscs, qficr sc or ora/atlmillistratioll Oil tla.\' 7,,\ alltl/or l) (!rl/cstalioll

Percentage increase Dose (mg/kg body Days'!' of or decrease in Percentage of Percentage weight) and (day'!' No. of weight loss No. of live Total no. of foetal. weight compared foetuses that of foetuses Test compounds of administration) dams in dams foetuses/daml live foetuses with control~ were resorbed with malformations

Control 54 9·39 ± 2·9 507 8·15 1·77 Epiprogoitrin 159(8,9) 5 0 %±3-8 48 + 14'35** 5·88 0 320(8,9) 8 0 10·5 ± 2·2 84 -9** 7-69 1·19 S-I-Cyano-2-hydroxy-3-butene 76(8,9) 7 917 10·1 ± 1·6 71 + 604** 11·25 1·4 150 (8) 4 920 9·5 ± 2·6 38 15'5** 7-32 0 175 (8) 5 1020 9·0 ± 4.9 45 -18'0** 13-46 2·2 R-Goitrin 100 (8,9) 6 0 9·0 ± 3-6 54 -5'6** 3·57 0 200(8) 12 9 15 9·92 ± 2·2 119 -0,12 9·85 2·52 I-Cyano-3,4-epithiobutane 44(8,9) 5 10 16 10·6 ± 1·6 53 -20** 5·35 0 ?' 95 (8) 5·0 ± 4,34** -27** 10 9 20 50 48'4** 2 Z Sinigrin 322 (8,9) 12 0 8'92 ± 4·5 107 - 5,6** 7-69 1·18 r;; :r 32211 (8,9) 10 0 10·0 ± 2·5 100 - 5,74** 3-85 3·0 m Allyl isothiocyanate 50(8,9) 6 0 10·0 ± 2·0 60 -11** 3·22 0 ::l'" 100 (8,9) 8 II 17 9·87 ± 3-6 79 +3-9 1904** 1·27 0- Sinalbin hydrate 300(8) 6 0 8·83 ± 4·7 53 3·5 5·36 1·75 ~ 300 (8,9) 3 0 11·3 ± 2·1 34 +104 2·85 2·94 fTl 100(8,9) 1'-Hydroxyphenylacet onitri Ie 4 0 11·25 ± 1·5 45 +4·25 4·25 4044 0 200(8) 6 9 14 9·33 ± 4·37 55 +6'2* 5·17 5-45 ;l-x 100(8,9) m I-Cyano-3-methylsulphinylpropane 6 0 11·7 ± 1·75 70 +3-6 0 0 Z (iberin nitrile) 200(8,9) 9 0 %6 ± 3-4 87 +4'58* 2·24 3-44 n'" 3-Methylsulphinylpropyl :r r- isolhiocyanate (iberin) 100(8) 5 0 7·0 ± 6·55 35 -1·1 28'6** 0 m :

"Day of gcstation. lValues arc mcans ± SD. ~The mean (± SD) weight of control foetuses was 4·18 ± 0047 g. ilThis dose was given orally; all other doses were given sc. All foetal data were collected on day 20 of gestation. Values marked with astcrisks dil1ered significantly (Student's I test or chi-square test) from the control value: *p < 0·05; ** P < 0·01. Table 3. Effl'cts of fJlllcosilloiates alld related compollllds 011 relatil'e orgall weiy1Jts of prey/l(lIlt allilllollpreYlllllJl females alld male rats killed 11-:12 da.\"s ajier sc or oraladmillistratioll

Relative organ weights Dose (mg/kg body weight) No. of Initial Compound x no. of dosest Condition rats body weight Liver Kidney Adrenal Thyroid

Epiprogoitrin 320 x 2 P 5 NS NS NS NS 320 x 2 M 3 + 15·9·· NS NS NS S-I-Cyano-2-hydroxy-3-butene 175 x I P 5 -29** NS +57·· + 14·75 175 x I NP 7 NS NS NS NS 200 x I M 3 NS NS NS NS R-Goitrin 200 x I P 6 NS NS NS NS 200 x I NP 8 NS NS + 13·8·· +11·7 200 x I M 5 + 12·2·· NS NS +10·4 I-Cyano-3.4-epithiobutane 95 x I P 5 NS + 78·7·· +53·6·· NS -l 95 x I M 6 NS +13-1· +24'7· NS 0 "(S- Sinigrin 322 x 2 P 7 + \7·1· NS -14· -15 o 322 x 2 NP 2 -II NS -10 -22·3· 0- Otl 322 x 2t P 7 NS NS NS -23· '< 322 x 2t NP 2 NS -10· -13 -28·· £, 322 x 2t M 3 NS NS + 10·9 +10'3 Otl () Sinalbin hydrate 300 x I P 6 + 14'3·· NS NS -10 0 300 x 2 P 3 NS NS + 10·3 + 13·2 S''" 300 x 2 NP 3 NS NS NS NS 0 ii> p-Hydroxyphenylacetonitrile 200 x I P 6 NS NS + 12-4 + \7-4· <> 100 x 2 P 4 NS NS NS -10 '"po 200 x I M 3 NS NS NS NS ::> 0.. 3-Methylsulphinylpropyl 100 x 1 P 5 NS NS + 15·25· +14 0.. 100 x 1 NP 7 NS NS NS NS ..,C1l isothiocyanate (iberinl "::;- 88 x I M 4 NS NS + 16·5·· NS !:? 50 x 2 P 6 + 10·6· + 17· +14·4 +23'8· ::;. Trypan blue: day 8,9 C1l day 7 100 x I P 6 + 17·5· +21'8·· NS + 23-4·· '" day 8 100 x I P 6 +22'8·· + 17·2·· NS NS day 9 100 x I P 6 +24·3·· + 15·4·· NS NS Controls~ P 54 268 ± 14·42 35·24 ± 14·4 5·03 ± 0·6 180·4 ± 21·5 51·2 ± 10·1 NP 15 267 ± 19 32·0 ± 1·8 6·71 ±0·55 240·8 ± 23 65·3 ± 10·8 M 6 202 ± 5·5 37·7 ± 2·5 7·93 ± 0·41 189·3 ± 23 67·0 ± 4'81

P = Pregnant NP = Nonpregnant M = Male NS = Value not significant and under lOy', tPregnant rats were injected on day 8 and/or 9 of pregnancy except for trypan blue which was injected on the days indicated. The doses were given sc except where indicated. tThese doses were given orally. *Control values are means ± SD; body weight is expressed in g; liver and kidney weights are expressed in g/kg body weight and adrenal and thyroid weights are expressed in mg/kg body weight. a- Values marked with asterisks differed significantly (Student's t test) from the control value (.P < 0'05; ••P < 0'01). w Q

164 K. NISHIE and M. E. DAXENBICHLER weight in pregnant and nonpregnant rats and sini­ ney tubular cells was observed in rats that died 4 to grin (sc) decreased the thyroid weight in nonpreg­ 5 hr after 150 mg/kg, sc, of iberin. nant rats alone. The following compounds had no Thyroid. The results of thyroid histological examin­ effect on organ weights: allyl isothiocyanate, iberin ation are summarized in Table 5, and representative nitrile, 3-indolylacetonitrile, indole-3-methanol and photomicrographs are presented in Figs 2 & 3. There vitamin U. was considerable variation in the control thyroid with respect to the size of the follicles, height of the epithe­ Lethal e.fJects lial cells lining the follicles, and the absence or pres­ Because of the limited availability of some com­ ence of vacuoles in the colloidal material filling the pounds, a systematic LDso determination was not feas­ lumen. The thyroid gland of pregnant rats treated ible: but only seven compounds in this series caused with the only compound in this series that increased death at the doses tested (Table 4). S-I-Cyano-2­ the thyroid weight. p-hydroxyphenylacetonitrile, had hydroxy-3-butene induced intermittent 'rolling seiz­ enlarged follicles or a slightly hyperplastic appear­ ures' and a loss of righting reflex preceding death (I-{j ance. The hyperplastic response to the test com­ days). The rats treated with l-cvano-3.4­ pounds seemed more intense in nonpregnant female epithiobutane, 3-indolylacetonitrile(iberin, -allyl iso­ rats than in the pregnant or male rats. ) and indole-3-methanol were comatose before death. Lethal and near lethal doses of p-hyd­ General observations roxyphenylacetonitrile caused blinking of the eyelids The following compounds caused lachrymation: and lachrymation within 10 min. In 30 min, clonic leg S-I-cyano-2-hydroxy-3-butene, l-cyano-3,4-epithio­ movements started, and the rats did not respond to butane, 3-indolylacetonitrile, p-hydroxyphenylaceto­ painful stimuli and lost the righting reflex capability. nitrile, allyl isothiocyanate and iberin. These com­ When righted, the legs could not support the body, pounds also caused intense local oedema at the sc and the animals rested on the abdomen with legs injection site, Indole-3-methanol and 3-indolylaceto­ spread and twitching. This clonic movement co~­ nitrile induced sedation, ataxia, loss of righting reflex tinued as long as an hour and was followed by a and sleep, Treatment with 300 mg/kg, sc, of indole-3­ motionless period preceding death or recovery. Older methanol induced loss of righting reflex which lasted male rats were more resistant to p-hydroxyphenyl­ about 4 hr. acetonitrile than young male rats (Table 4). The sur­ vivors did not show anv adverse residual neurological DISCUSSION effects. Young male r~ts \vere more sensitive than pregnant rats, which were older. to the lethal effects of The previously demonstrated absence of teratogeni­ allyl isothiocyanate and iberin. city of goitrin (Khera, 1977: Rakalska & Dzierzawski, 1972) and of allyl isothiocyanate (Ruddick, Newsome Histology & Nash, 1976) was confirmed by using different doses, Liver. The pregnant rats that had rolling seizures a different route of administration and different fre­ and died 2 to 6 days after sc injection of S-I-cyano-2­ quencies and times of administration (Table 2). hydroxy-3-butene had liver necrosis in the subcapsu­ The occurrence of neurolathyrism and osteolathyr­ lar. or in certain midzonal areas spreading ism due to the ingestion of certain leguminous plants toward the centrilobular veins, which were intact. (peas) has been known for centuries: but the identifi­ Some survivors had localized liver necrosis, replaced cation of the aminonitriles responsible for those toxic by scar tissue with disorganized proliferation of new signs is more recent as described by Selye (1957), biliary ducts and blood vessels when examined on the Gardner (1959) and others. We expected foetal mal­ 12th day after receiving 200 mg/kg of this compound. formations due to osteolathyrism which involves Rats treated with the other compound in this series bones and cartilages, when we treated the pregnant had liver histology similar to that of the controls, with rats with various nitrile derivatives of the glucosino­ occasional clusters of lymphocytes. lates. However. the nitriles did not cause a significant Kidney. A section of normal kidney from a control number of skeletal malformations. From nine preg­ rat is shown in Fig. 1a for the purpose of comparison. nant rats treated with two doses of iberin nitrile The kidney sections of rats that died within 7 hr after (200 mg/kg, sc), only one foetus of 87 had defects treatment with 300 mg/kg of S-I-cyano-2-hydroxy-3­ involving the skeleton, namely, talipes and a curled butene had swollen kidney tubular cells with swollen tail. nuclei, Rats that died overnight after treatment had The neurolathyrism caused by aminonitriles such necrosis of the kidney tubular cells. The survivors of as 2,2'-iminodipropionitrile in experimental animals is the group treated with 200 mg/kg of this compound characterized by a lifelong hyperexcitability, chorei­ did not show any kidney abnormality on the 12th day form and circling movements starting 3 to 4 days after after treatment. The kidney of rats that died just 2-3 treatment The nitrile S-I-cyano-2-hydroxy-3-butene days after receiving 95 mg/kg, sc, of l-cyano-3.4­ in lethal doses (200 mg/kg, sc) induced intermittent epithiobutane had vacuolation and necrosis of the rolling seizures before death (I-{j days), but lower tubular cells (Fig. Ib,c). On the 12th day, some sur­ doses did not cause any visible permanent neurologi­ vivors of this group had kidneys with extremely cal effect in pregnant rats. Treatment with p-hydroxy- dilated tubules, lined with flat epithelial cells, and had . phenylacetonitrile in high doses caused clonic leg focal infiltration of lymphocytes in the interstitial movements, but the survivors were normal. without space (Fig. Id) of the kidney. Some kidneys that were any residual neurologic effects. extremely enlarged and pale had dilated tubules By the action of the plant enzyme thioglucosidase filled with a pink-staining material. Necrosis of kid- (myrosinase), glucobrassicin forms 3-indolylaceto- Table 4. Lethal "oses of COIllPOll/ll/S "eril'e"FOIII {I/Ilcosil/o/ates (rats)

Death LDso ± SD Dose Test chemical (mg/kg body weight) Sex (mg/kg body weight) Rate Time (hI') Remarks

3-Methylsulphinylpropyl isothiocyanate (iberin) 90 ± 2·89 F 150 6/6 46 Comatose, oedema at the F injec!ion site -J 100 0/12 o ;.< M 100 3/8 824 n' M(Y) 87,5 2/6 824 o Allyl isothiocyanate 92 ± 2·3 M(Y) 75 0/3 o fI",< F 100 x 2 1/13 48 Comatose, oedema at thc M(Y) 100 7 injection site £, 3/3 (IQ M(Y) 90 2/6 5 ne- I-Cyano-3,4-epithiobutane 109 ± 4·8 M(Y) 75 0/3 ~ F 132 6/6 48 Comatose 5' o F 95 1/13 27 Comatose §: 72 Comatose <> M 95 1/7 V> S-I-Cyano-2-hydroxy-3-butene 200±12·5' F 88 1/6 96 Comalose P> ::J M 250 3/5 1--6 Rolling seizures, liver necrosis 0- F 200 9/18 24-144 Rolling seizures, liver necrosis 0- ~ 3-lndolylacetonitrile 255 ± 14 F 175 0/12 :::' M(Y) 300 7/8 24 Comatose ~ <' M(Y) 250 4/9 1'5-24 Comalose D: Indole-3-melhanol M(Y) 225 0/3 M(Y) 500 3/4 1-3 Comatose M(Y) 300 x 2 0/3 p-Hydroxyphenylacetonitrile F 250 6/6 1·5 Clonic Icg movements M(Y) 350 0/6 Clonic leg movements M(YY) 300 2/3 12 Clonic leg movements M(YY) 260 1/1 1·5 Clonic Icg movements

F = female rats (pregnant + nonpregnant) weighing 250--280 g M = Male rats weighing 200 -230 g M(Y) = Young males, 100-150 g M(YY) = Younger male, 50-75 g

v,0­ '"

Table 5. Thyroid histology ofrats, 12 days after sc or oral administration of one or two doses ofgll/cosinolates allli related compol/nds

Thyroid histologyt Condition Compounds* of rats .... Pregnant Nonpregnant Male

Epiprogoitrin Fig. 2b; normal Normal; slightly hyperplastic Normal; smaller follicles, columnar epithelial cells S-1-Cyano-2-hydroxy-3-butene Normal; hyperplastic, Fig. 2c; very hyperplastic; Very hyperplastic; small smaller follicles smaller follicles; foamy, follicles, columnar cells; scanty colloid vacuoles in the colloid R-Goitrin Fig. 2e; normal; smaller Normal; hyperplastic; smaller Fig. 2d; small follicles; follicles; slightly follicles; scanty colloid tall cells; scanty colloid hyperplastic or foamy colloid ~ Z l-Cyano-3,4-epithiobutane Fig. 2f; normal; smaller Normal; smaller follicles Normal <;; follicles, cuboidal cells ::x: Sinigrin (sc administration) Normal; smaller follicles; Fig.3b Fig. 3a; large follicles iTi § mitotic cells columnar cells; papillary 0- formations ~ Sinigrin (oral administration) Normal; slightly Normal; slightly hyperplastic Normal hyperplastic vacuoles (lumen) !'1 Allyl isothiocyanate Fig.3c Hyperplastic areas; columnar o cells ~ ~ Sinalbin hydrate Normal; slightly Normal; slightly hyperplastic; tIl small follicles n hyperplastic; Fig. 3d ::x: p-Hydroxyphenylacetonitriie Normal; slightly t"' hyperplastic ~ Iberin nitrile Normal; smaller follicles; Fig.3e columnar cells; scanty colloid Iberin Normal Normal 3-Indolylacetonitriie Normal; slightly Fig. 3f; very hyperplastic; hyperplastic foamy or scanty colloid Indole-3-methanol Normal; smaller follicles; Small follicles, columnar columnar cells cells Vitamin U Normal; smaller follicles;

*All the compounds were administered sc except where indicated. tHyperplastic refers to an increase in the number of cells: cells are epithelial cells lining the thyroid follicles. Fig.!. Kidney sections of (aj a control rat; (b) a pregnant rat that died 27 hr after treatment sc with 95 mg l-cyano-3.4-epithiobutane/kg showing vacuolated tubular cells; (cl a pregnant rat that died on the third day after treatment with 95 mg l-cyano-3.4-epithiobutane/kg showing dilated tubules and necrosis of tubular cells; (d) a pregnant rat killed 12 days after treatment with 95 mg l-cyano-3.4-epithiobu­ tane/kg showing dilated tubules. lined with flat epithelial cells, and interstitial infiltration of lymphocytes. All the kidney sections were stained with periodic acid-Schiff and haematoxylin. a,b,c. x 400. d. x 160.

167 Fig. 2. Thyroid sections of (a) a control pregnant rat on day 20 of gestation; (b) a pregnant rat on day 20 of gestation, 12 days after two daily sc doses of 320 mg epiprogoitrin(kg showing an essentially normal thyroid; (c) a nonpregnant rat 12 days after an sc dose of 175 rng S-I-cyano-2-hydroxy-3-butene(kg showing columnar epithelial cells lining the follicles and lumens without colloid or with foamy colloid; (d) a male rat, 12 days after an sc dose of 200 mg R-goitrin(kg showing epithelial cell hyperplasia and scanty colloid; (e) a pregnant rat on day 20 of gestation, 12 days after an sc dose of 200 mg R-goitrin(kg showing smaller more cellular follicles than those of the control thyroid; (f) a pregnant rat on day 20 of gestation, 12 days after an sc dose of 95 mg l-cyano-3,4-epithiobutane showing an essentially normal thyroid. All thyroid sections were stained with periodic acid-Schiff and haematoxylin x 400.

168 • Fig. 3. Thyroid sections of (a) a male rat 12 days after two daily sc doses of 322 mg sinigrin/kg showing relatively large follicles lined with columnar epithelial cells; (b) a nonpregnant rat 12 days after two sc doses of 322 mg sinigrin/kg showing smaller follicles than the control lined with slightly taller epithelial cells; (c) a pregnant rat on day 20 of gestation, 12 days after two daily sc doses of 100 mg allyl thiocyanate/kg showing uniformly smaller follicles lined with tall epithelial cells; (d) a pregnant rat on day 20 of gestation, 12 days after two daily sc doses of 300 mg sinalbin hydrate/kg showing slightly smaller. more cellular follicles than the controls; (e) a nonpregnant rat, 12 days after two daily sc doses of 200 mg iberin nitrile/kg showing uniformly tall epithelial cells lining the follicles, scanty colloid, and some follicles with vacuoles; (I) a nonpregnant rat 12 days after an sc dose of 150 mg 3-indolylacetonitrile/kg showing extremely hyperplastic and cellular thyroid and scanty colloid. All thyroid sections were stained with periodic acid-Schiff and haematoxylin, x 400.

169 Toxicology of glucosinolates and derivatives 171

nitrile at pH 3-4 and indole-3-methanol and other (1973). Synergy of ethanol and alcohol-like metabolites: compounds at pH 7 (Michajlovskij & Langer, 1967: Tryptophol and 3.4-dihydrophenylethanol. Phanna­ Virtanen. 1965). 3-Indolylacetonitrile is a plant cology 9, 294. growth hormone (Jones, Henbest, Smith & Bentley, Bosin, T. R. Campaigne, E., Dinner, A., Rogers, R. B. & 1952). but in in vitro experiments it inhibited the Maickel. R. P. (1976). Comparative toxicological studies growth of ascites sarcoma cells (Pilotti, Ancker, Arr­ of indole, benzo(b)thiophene, and l-methylindole deriva­ tives. J. Toxicol. en vir. Hlth 1, SIS. henius & Enzel, 1975). 3-Indolylacetonitrile is present Chesney, A. M., Clawson, T. A. & Webster. B. (1928). in many plants, e.g. cabbage (Henbest, Jones & Smith, Endemic goitre in rabbits. Incidence and characteristics. 1953). and in tobacco smoke (Pilotti et al. 1975). In Johns Hopkins Hasp. Bull. 43, 261. rats. 3-indolylacetonitrile and indole-3-methanol Clandinin, D. R. & Bayly, L. (1960). Rapeseed oil meal induced sedation, ataxia, loss of righting reflex and studies (RSOM). Effect of feeding rapeseed oil meal on sleep. These responses were not unexpected, because the structure of the thyroid glands and chickens. Poult. tryptophol (indole-3-ethanol). a homologue of Sci. 39, 1239. indole-3-methanol, is known to cause sedation (Bosin, Cole, R. A. (1975). l-Cyanoepithioalkanes: major products Campaigne. Dinner, Rogers & Maickel, 1976) and of alkenylglucosinolate hydrolysis in certain cruciferae. Phytochemistry 14, 2293. enhanced ethanol-induced sleep in mice (Blum, Cal­ Cole, R. A. (1976). Isothiocyanates, nitriles and thiocya­ houn. Merritt & Wallace, 1973). 3-Indolylacetonitrile nates as products of autolysis of glucosinolates in Cruci­ caused marked hyperplasia in nonpregnant rat thy­ ferae. Phytochemistry 15, 759. roid (Fig. 31), but its effect on the thyroid of the preg­ Dawson, A. B. (1926). A note on the staining of the skele­ ... nant rat \vas less pronounced. This diminished suscepti­ ton of cleared specimens with Alizarin Red S. Stain bility of pregnant rats to the goitrogenic (hyperplastic) Tee/lIlol. 1, 123. action of the nitriles and other compounds may have Daxenbichler, M. E., VanEtten, C H., Brown. F. S. & been due to the higher concentration of oestrogens Jones, Q. (1964). Oxazolidinethiones and volatile isothio­ .. during pregnancy, since Eskin & Bogdanove (1956) cyanates in enzyme-treated seed meals from 65 species of cruciferae. J. agric. Fd Chenl. 12, 127. have shown that oestrogens reduced the hyperplastic Daxenbichler, M. E., VanEtten, C H. & Spencer, G. F. response of the thyroid to propylthiouracil (goitro­ (1977). Glucosinolates and derived products in cruci­ gen). ferous vegetables. Identification of organic nitriles from The goitrogenic effect of cabbage in rabbits was cabbage. J. agric. Fd Chem. 25, 121. discovered by Chesney et al. (Chesney, Clawson & Daxenbichler, M. E., VanEtten, C. H. & Wolff, 1. A. (1966). Webster. 1928; Webster & Chesney, 1930). The active (S)- and (R)-I-Cyano-2-hydroxy-3-butene from myrosi­ compound L-5-vinyl-2-thiooxazolidone (goitrin) was nase hydrolysis of epiprogoitrin and . Bio­ identified by Astwood, Greer & Ettlinger (1949); and chemistry, N. Y. 5, 692. later, Greer isolated progoitrin, the precursor of goi­ Daxenbichler, M. E., VanEtten. C A. & Wolff. 1. H. (1968). Diastereomeric episulfides from epiprogoitrin upon trin (1956). Other goitrogenic compounds include autolysis of crambe seed meal. Phytochemistry 7, 989. allyl isothiocyanate and thiocyanate (Langer, 1966; Eskin, B. A. & Bogdanove, E. M. (1956). The influence of VanEtten, 1969). estrogen upon goiter induction in adult immature rats. Whitelaw (1947) reported a case of human foetal Endocrinology 59, 688. anencephaly due to the maternal treatment of Gardner, A. F. (1959). Experimental lathyrism. Review of hyperthyroidism with thiouracil. Since thiouracil has the literature. Am. J. din. Nutr. 7, 213. goitrogenic and teratogenic properties, the possibility Greer, M. A. (1956). Isolation from rutabaga seed of pro­ arose that the seasonal increases in human foetal goitrin, the precursor of the naturally occurring anti­ anencephaly and spina bifida were due to the goitro­ thyroid compound, goitrin (L-5-vinyl-2-thiooxazoli­ done). J. Am. chem. Soc. 78, 1260. • gens present in Cruciferae and other plant-foods. How­ Henbest, H. B., Jones, E. R. H. & Smith, G. F. (1953). ever, the limited number of compounds we tested Isolation of a new plant-growth hormone 3-indolyl­ were not teratogenic, although they did cause varying acetonitrile. J. chem. Soc. p. 3796. degrees of thyroid hyperplasia. Further investigations Hussar, N. & Bowland, 1. P. (1959). Rapeseed oil meal as a are needed to resolve their possible role in birth de­ protein supplement for swine and rats. 1. Rate of gain, fects. efficiency of food utilization, carcass characteristics and thyroid activity. Can. J. Aninl. Sci. 39, 84. Jones, E. R. H., Henbest, H. B., Smith, G. F. & Bentley, Acknoll'ledgemenrs-We thank Mrs. Pamela Hayes and Mr J. A. (1952). 3-Indolylacetonitrile: a naturally occurring William Schroeder for their excellent technical assistance. plant growth hormone. Nature, Land. 169,485. Kennedy, T. H. & Purves, H. D. (1941). Studies on experi­ mental goitre. 1. The effect of Brassica seed diets on rats. Br. J. expo Path. 22, 241. REFERENCES Khera, K. S. (1977). Nonteratogenicity of D- and L-goitrin Ahmed. Z. A. (1972). Naturally occurring glucosinolates in the rat. Fd Cosmet. Toxicol. 15, 61. with special reference to those of family Capparidaceae. Kirk, J. T. O. & MacDonald, C G. (1974). l-Cyano-3,4­ Planta med. 21, 35. epithiobutane: a major product of glucosinolate hy­ Astwood, E. B., Greer. M. A. & Ettlinger. M. G. (1949). drolysis in seeds from certain varieties of Brassica cam­ L-5-Vinyl-2-thiooxazolidone. an antithyroid compound pestris. Phytochemistry 13, 2611. from yellow turnip and from Brassica seeds. J. bioI. Kirk, L. D.. Mustakas, G. C, Griffin, E. L. & Booth, A. N. Chem. 181, 121. (1971). Crambe seed processing: decomposition of gluco­ Bell, 1. M.. Benjamin, B. R. & Giovannetti. P. M. (1972). sinolates (Thioglucosides) with chemical' additives. J. Histopathology of thyroids and liver of rats and mice fed Am. Oil Chem. Soc. 48, 845. diets containing Brassica glucosinolates. Can. J. Anim. Langer, P. (1966). Antithyroid action in rats of small doses Sci. 52, 395. of some naturally occurring compounds. Endocrinology Blum. K.. Calhoun. W.. Merritt, J. H. & Wallace. J. E. 79,1117. 172 K. NISHIE and M. E. DAXENBICHLER

Leck. I. (1974). Causation of neural tube defects: clues Spina bifida and anencephaly in the Cape. S. Afr. med. J. from epidemiology. Br. med. Bull. 30, 158. 53,626. Manns, 1. G .. Bowland, J. P.. Mendel, V. E. & Zivkovic, S. Srivastava, V. K., Philbrick, D. J. & Hill, D. C. (1975). (1963). Solvent-extracted rapeseed oil meal as a protein Response of rats and chicks to rapeseed meal subjected source for pigs and rats. Can. J. Anim. Sci. 43, 271. to different enzymatic treatments. Can. J. Anim. Sci. 55, Michajlovskij. N. & Langer, P. (1967). Identity of the 331. goitrogenic effects of glucobrassicin and the equivalent VanEtten. C. H. (1969). Toxic constituents of plant food­ amount of thiocyanate in rats. Endocrinologia Experi­ stuffs. In Goitrogens. Edited by I. E. Liener. p. 103. Aca­ mentalis 1,229. demic Press. New York. Nordfeldt. S.. Gellerstedt. N. & Falkmer, S. (1954). Studies VanEtten. C. H., Daxenbichler, M. E. & Wolff, I. A. on rape-seed meal and its goitrogenic effect on pigs. Acta (1969a). Natural glucosinolates (Thioglucosides) in foods path. microbiol. scaml. 35, 217. and feeds. J. agric. Fd Chem. 17, 483. Pilotti, A.. Ancker. K., Arrhenius. E. & Enzell, C. (1975). VanEtten, C. H., Gagne. W. E., Robbins. D. J., Booth, Effect of tobacco smoke constituents on cell multipli­ A. N .. Daxenbichler, M. E. & Wolff, I. A. (1969b). Bio­ cation in vitro. Toxicology 5, 49. logical evaluation of crambe seed meal and derived Rakalska, Z. & Dzierzawski, A. (1972). Effect of 5-vinyl-2­ products by rat feeding. Cereal Chem. 46, 145. thiooxazolidone on the development of golden hamster Virtanen, A. I. (1965). Studies on organic sulphur com­ fetuses. Bull. Vet. Inst. PIIlaw}' 77,95. pounds and other labile substances in plants. Phytochem­ Renwick, 1. H. (1974). The epidemiology of natural terato­ istr}' 4, 207. gens. Proc. Univ. Mo. Ann. Conf. Trace SlIbst. EnL Webster, B. & Chesney, A. M. (1930). Studies in the etio­ Health 8, 3. logy of simple goiter. Am. J. Path. 6, 275. Ruddick. 1. A., Newsome. W. H. & Nash, L. (1976). Corre­ Whitelaw, M. 1. (1974). Thiouracil in the treatment lation of teratogenicity and molecular structure: ethyl­ of hyperthyroidism complicating pregnancy and its ene thiourea and related compounds. Teratology 13, 263. effect on the human foetal thyroid. J. din. Endocr. 7, Sandahl, B. (1977). Seasonal incidence of some congenital 767. malformations in the central nervous svstem in Sweden. Wilson. J. G. (1965). Embryological consideration in tera­ .. Acta paediat. Stockh. 66, 65. - tology. In Teratology: Principles and Techniques. Edited Seyle. H. (1957). Lathyrism. Revile can. Bioi. 16, 1. by 1. G. Wilson and 1. Warkany. p. 262. University of Singer, H. A.. Nelson. M. M. & Beighton, P. H. (1978). Chicago Press, Chicago.

•