- Email: [email protected]
Decrease in blood coagulation factors II (prothrombin), VII, IX and X in the rat after a single oral dose of butylated hydroxytoluene
Fd Chem. Toxic. Vol. 25, No. 3, pp. 219-224, 1987 Printed in Great Britain.All rights reserved
0278-6915/87 $3.00+0.00 Copyright ~ 1987PergamonJournals Ltd
DECREASE IN BLOOD COAGULATION FACTORS II (PROTHROMBIN), VII, IX A N D X IN THE RAT AFTER A SINGLE ORAL DOSE OF BUTYLATED HYDROXYTOLUENE O. TAKAHASHI
Department of Toxicology, Tokyo Metropolitan Research Laboratory of Public Health, 24-1, Hyakunincho 3 chome, Shinjuku-ku, Tokyo 160, Japan (Received 10 June 1986) Abstract--Male Sprague-Dawley rats were given butylated hydroxytoluene (BHT) in a dose of 800 mg/kg body weight orally, and 0.5-72 hr later plasma concentrations of factors II, VII, IX and X and hepatic levels of BHT and BHT quinone methide (2,6-di-tert-butyl-4.methylene-2,5-cyclohexadienone) were determined. Levels of factors II, VII, X and IX were reduced 36-60 hr after BHT treatment, but by 72 hr, those most affected (VII and IX) showed some recovery and X had returned to normal. Hepatic levels of BHT reached a maximum 3 hr (a major peak) and 24 hr after BHT dosing and BHT quinone methide reached a maximum at 6 and 24 hr (a major peak). In rats given BHT orally in doses of 200, 400 and 800 mg/kg, factors II, VII and X decreased after 48 hr only in rats given the highest dosage, but factor IX was more susceptible to BHT and showed a dose-dependent decrease. Phylloquinone (1 me/rat) injected ip 24 hr after the administration of 800 mg BHT/kg maintained normal levels of factors VII and X and an almost normal level of factor IX, but had little effect on the level of factor II. In studies of the effects of drug-metabolizing-enzyme modifiers, neither ip pretreatment with 75 mg phenobarbital sodium/kg for 3 days nor the feeding of 1% cysteine in the diet throughout the experiment prevented the decrease in vitamin-K-dependent factors by 800 mg BHT/kg, but 2-day ip pretreatment with 60 mg cobaltous chloride/kg/day maintained normal levels of factors II and VII and reduced the BHT effect on factors IX and X. SKF 525A (50 mg/kg) injected ip either 30 min before or 12 hr after BHT treatment partially prevented the decrease in factors II, VII and X, or in all four factors, respectively. Thus the decrease in vitamin K-dependent factors may be the same with a single oral dose of BHT as with dietary BHT, and the anticoagulant effect may require the metabolic activation of BHT.
INTRODUCTION
When administered to male rats in the diet, butylated hydroxytoluene (BHT), a sterically hindered phenolic antioxidant, can cause haemorrhagic death in 9-37 days and a dose-related decrease in the prothrombin index (Takahashi & Hiraga, 1978a,b). In a search for the critical defect in this bleeding, we examined several haemostatic functions in rats given 1.2% BHT in the diet for 7 days (Takahashi, 1985; Takahashi & Hiraga, 1981a, 1982 & 1984) and found that BHT decreases the vitamin K-dependent coagulation factors and changes platelet function, vascular permeability and the activity of the kallikrein-kinin system, but does not affect fibrinolysis or the classical complement fixation process. Of those effects, the most remarkable are a decrease in factors II (prothrombin), VII, IX and X, and a reduction in aggregation of the washed platelets. Previously, an attempt was made to divide the period of pathogenesis, from the start of BHT feeding to the occurrence of hacmorrhaglc death, into four stages (Takahashi, 1986): stage I - - a decrease in blood coagulation factors II, VII, X and IX (from days 1-2 to days 3-4), stage I I - - a decrease in these coagulation factors and hacmorrhages in the epididymis (from days 3-5 to days 6-8), stage I I I - - a decrease in the four coagulation factors with hacmorrhages in several organs and a defect in platelet function (from days'6-8 to
days 8-10) and stage IV--massive haemorrhages and death (after days 9-10 and up to approximately l month), the median lethal time being about day 15 (Takahashi & Hiraga, 1978c). Our interests are now focused on a search for the mechanism of early bleeding, i.e. the haemostatic functions at stage 1. The previous finding of a significant decrease in factors on day 2 suggested that factors II, VII, IX and X were rapidly reduced immediately after the ingestion of BHT. The work reported here considered the time course of changes in the coagulation factors after a single oral BHT dose in relation to the metabolism of BHT, the dose response and the effect of drug-metabolizing-enzyme modifiers. MATERIALS AND METHODS
Chemicals. BHT, obtained from Tokyo Kasei Kogyo Co. (Tokyo) was dissolved in soya-bean oil for administration. 2,6-Di-tert-butyl-4-methylene2,5-cyclohexadienon¢ (BHT quinone methide) was synthesized as described previously (Takahashi & Hiraga, 1979a). Kaywan ®, a phylloquinone emulsion, was purchased from Eisai Co. (Tokyo). Phenobarbital sodium, cobaltous chloride and SKF 525A (proadifen hydrochloride), from Wako Pure Chemical Industry (Osaka) and Smith Klinc &
219
220
O. TAKAHASHI
French Laboratories (Philadelphia, PA, USA) were Spraguc-Dawley male rats weighing approximately dissolved in 0.9% NaCI solution. L-Cysteine was 300 g and fed a laboratory ration (CLEA CE-2). purchased from Tokyo Kasei Kogyo Co. ThromboTissue analyses. The livers were mixed with anhyplastin, cepharin (a platelet factor), 0.02 M-calcium drous sodium sulphate and homogenized with acetone (1 g tissue in 10mi acetone) using a Polytron chloride, Owren's Veronal buffer pH 7.35 and human and/or bovine plasma lacking factor II, VII, IX or X homogenizer (Kinematic GmbH, Lucerne, Switzerwere purchased from Dade Diagnostics Inc. (Miami, land). The acetone solution was evaporated under FL, USA). nitrogen and the residue was dissolved in hexanc for Animals and treatments. Male Sprague-Dawley analysis by gas chromatography in a Shimadzu rats, obtained when 4 wk old from CLEA Japan Inc. GC-9A gas chromatograph equipped with a flame(Tokyo), were fed a laboratory ration (CLEA CE-2) ionization detector (Shimadzu Inc., Kyoto). A glass for 1-2 wk, and then received a purified diet for 3 'column (3m x 3ram ID) was packed with Gas days before and after the administration of BHT. The Chrom Q containing 10% Silicone DC-200, while a purified diet, which has been described previously glass cartridge packed with a 3.5-cm length of silanized glass wool and ifiserted into the port chamber (Takahashi & Hiraga, 1978a), contained 24.0% casein, 25.0% sucrose, 41.4% starch, 5.0% soya-bean served as a pre-column. The operating temperatures oil, 4.0% salts, 0.45% vitamins and 0.15% choline were 180°C (column) and 250°C (detector and injecchloride. The purified diet was fed for 3 days before tion port). The carrier gas (nitrogen) flow rate was BHT administration because it took 3 days for the 50ml/min (Takahashi & Hiraga, 1979a). levels of blood coagulation factors, which were Statistical analysis. Variation of results was routinely expressed as the standard error of the mean. affected to some extent by the change of diet, to become stable (Takahashi, 1986). In the first experi- Significance of differences was tested on the data for ment, rats were given BHT orally by gavage in a dose coagulation times by Student's t test, taking P < 0.05 of 800 mg/kg and were killed 0.5, 3, 6, 12, 24, 48, 60 as the limit of significance (Gad & Wcil, 1982). and 72 hr later. Control rats were given soya-bean oil. There were between four and ten rats in each group. RESULTS The concentrations of blood coagulation factors II, VII, IX and X in the plasma and of hepatic BHT and Body-weight gain was reduced 24-72 hr after the BHT quinone methidc were determined. In the administration of BHT. Relative liver weights of second experiment, four or five rats/group were given control rats varied regularly between a larger size in BHT in doses of 200, 400 or 800 mg/kg and killed the morning and a smaller one at night, but in rats 48 hr later, for determination of plasma levels of the given BHT this regular oscillation disappeared and coagulation factors. In one group given 800mg 36-72 hr after BHT administration the relative liver BHT/kg, each rat also received by ip injection l mg weights were significantly larger than those in the phylloquinone 24hr after the BHT. In the third controls. No necrotic changes were observed, howexperiment, rats were pretreated ip with 75mg ever. Levels of factors II and IX showed a decrease phenobarbital/kg daily for 3 days, and 24 hr after the 36-72 hr after BHT administration, while factor VII final injection, the BHT was administered. Other rats was slightly decreased at 3 and 12 hr and markedly were fed purified diet containing I% cysteine 3 days decreased at 36-60hr. Factor X was decreased before administration of BHT and throughout the 36-60 hr after dosing but returned to normal by experiment and another group received 60 mg cobalt- 72hr. All four factors reached a minimum at ous chloride/kg sc daily for 2 days, followed 24 hr 48-60 hr (Fig. la). Hepatic levels of BHT showed after the second injection by BHT. A further group peaks at 3 and 24 hr after dosing, while BHT quinone was treated with SKF 525A 30 rain before or 12 hr methide peaked at 6 and 24hr (Fig. lb). The major after BHT administration. For all these five groups peaks were at 3 and 24 hr for BHT and BHT quinone (four or five rats/group), the dose of BHT was methide, respectively. 800 mg/kg, and the rats were killed 48 hr after BHT Factors II, VII and X were decreased 48 hr after treatment for the determination of plasma concen- BHT treatment only in rats given the highest dosage trations of blood coagulation factors. (800mg/kg), but factor IX decreased in a doseDetermination of factors H (prothrombin ), VII, IX dependent manner (Fig. 2). Phylioquinone (1 murat) and X. Whole blood was collected (under ether injected ip 24 hr after the administration of 800 mg anaesthesia) from the inferior vena cava into plastic BHT/kg completely prevented the decrease in the syringes containing 0.147 u-sodium citrate solution levels of factors VII and X, and partially overcame (citrate-blood, 1:9, v/v). Plasma was prepared by the effect on factor IX, but the reduction in factor II centrifugation at 740 g for 10 min. Factors II, VII and was not significantly affected (Fig. 2). X were measured by a prothrombin time assay. Each Treatment with phenobarbital or dietary cysteine 0.l-mi volume of a 1:5 dilution of rat plasma in did not prevent the BHT-dependent decrease in vitaOwren's Veronal buffer pH 7.35 was mixed with min K-dependent factors (Fig. 3). Cobaltous chloride 0.1 ml human and/or bovine plasma lacking factor II, completely prevented the decrease in levels of factors VII or X and the prothrombin time was measured. II and VII and partially prevented the decrease in the Factor IX was determined by the activated partial levels of factors X and IX, while SKF 525A given thromboplastin time assay technique, again using 30 rain before or 12 hr after BHT partially prevented human plasma lacking factor IX (Owen & Bowie, the decrease in factors II, VII and X, or in all four 1978; Takahashi, 1986; Takahashi & Hiraga, 1981a). factors, respectively (Fig. 3). The mean relative concentrations of each factor were In addition, the liver enlargement caused by BHT determined using plasma dilution curves of normal was not prevented by pretreatment with cobaltous
221
Rat-blood coagulation factors and BHT
2~°I
(a)
_
! ~ool~r,_.,'l~--",X-~ P."o"
_ : ~
/ )'o - ~ ' ' ~
~X ~~~ ~ ~ o
X
lie
,I
~
+o-
" ~
'k't
.....
-
I~'/"
~
~
~
~
8 ~
~
°
~
~. ......
~
~$
e
./. ~
~
o K
+.. . . . . . . . . ~
+x ~
./ ~
/
/ ~
~ ~" ~
~
~
~ ~
~x,~z
10
I
I
I
I
I
0
2O
~l/#ill \\x\k%~
~15
,oL t
III ,,
/
L/l\lt ,,, o L t ,~ 0
I#
[
%%% ,, ~,~
\t'"-..
NT/
~
l
12
%x
"~.~
~t-----__~x 24
36
48
........ 60
72
Time after BHT treatment ( h r )
Fig. 1. Changes in (a) the plasma concentrations of blood coagulation factors II (--) VII (--.) X (---) and IX ( - - - ) and (b) the mean hepatic levels (+SEM) of BHT ( ) and BHT quinone methide (---) in rats given a single oral dose of 800 mg BHT/kg. In (a) values have been calculated from the mean clotting times of five to ten rats, using dilution curves of normal rat plasma, and are expressed as a percentage of the control concentration. The values differing significantlyfrom the controls (tested using the data for coagulation times) are those for factor II at 36, 48, 60 and 72 hr, for factor VII at 3, 12, 24, 48 and 60 hr, for factor X at 3, 36, 48 and 60 hr and for factor IX at 36, 48, 60 and 72 hr. chloride or by SKF 525A given 30 min before BHT, but was completely inhibited by the administration of SKF 525A 12 hr after BHT.
DISCUSSION
When BHT is administered to rats in the diet, food intake and therefore BHT intake decrease temporarily as a result of the change from the control diet. This decrease may obscure any analysis of the early response to BHT. In a previous report, it was noted that the changes in blood coagulation factors in rats on day 1 of consumption of a BHT-containing diet were not significantly different from the control levels, but it appeared that some decrease in factors II, VII, X and IX occurred immediately after the ingestion of BHT (Takahashi, 1986).
In the case of a single oral dose of BHT, the plasma levels of blood coagulation factors II, VII, X and IX did not show any obvious change for 24 hr and it took about 48 hr after BHT administration for any notable decrease in these factors to occur. If we compare the present data with the previous results for dietary administration (although oral intubation by stomach tube may not necessarily be equal to administration in the diet), the decrease in factors must begin to occur on day 2 of BHT feeding. The plasma half-life of rat prothrombin is 6-8 hr and it may be totally replaced with new prothrombin in about 20 hr (Knauer et al. 1976; Siegfried et al. 1979). Warfarin can induce hypoprothrombinaemia 8-18 hr after its administration (Shah & Suttie, 1979; Siegfried et al. 1979). A tingle oral dose of R- or S-warfarin diminished the levels of factors VII, X and II in 16-24 hr (Fasco & Principe, 1982). The levels of
222
O. TAKAHASHI 1000
-
0
X
100
~
IIII1:::1 ~
Eii
~ :!:! /~ :.:. ,~ .:.: ,,~ :.:. z/.:.: /~ :.:.
o
z ~::~: 10
~ :'2~ ~ ":'i ~ ~.y
i~ ,:.: .. :.:.:
m
~
,,,,,...,!!!!!::i::[z ",~:,z
z z,
,,::: ~i~t !!~ Z~i:.i..t
11111:::I ~
""
rii='i
iiiiiiiii,, ~...........,, ~, ~ ...........
~, ,,,,~,~ ........ ..
.. .. .. ........... .
'. ". .".:.:.:. .
:':'1
IIII17771~ ,. ,. ~,.: : :,. ¢ ~ ~.:;~
.........
..
."~::::~ ..... '
I[III;;~Z ........
,~::::s z~::::~
- ~- ' • • ~
¢ ~;:;'i
Ill[I:::
,~...a
IIII1~i1~::i~ . . . . . . . . . . . . . . : ~
111 I~: : :I "~ ::::~
iii:
"
!H~!~::i]~:~l
'
H,,..
~ ~:':'~
~
'""'"'
II:::
, ~ ~q.:
Z~ ~ ~: :~: :::::
IIII~:::I ~
:::~
~
:':':1 I [I II: : :I "~ :.:.:1
~
~[~ii!~Z~ I II II :::I ~
~"
**>F
Z ~ ~ :::: /~ .:.: ~ ~ :::: */:.:.:
~ ~i~ ~ :2:: ¢~ .-..
A
8
C
0
Fig. 2. Plasma concentrations of factors II (llI~),VII (1~1),X (1~) and IX (l~l) in rats 48 hr after treatment with a single oral dose of (A) 200 mg BHT/kg, (B) 400 mg BHT/kg, (C) 800 mg BHT/kg and (D) 800 mg BHT/kg with phylloquinone (1 rag/rat ip). Data have been calculated from the mean clotting times of four or five rats, using dilution curves of normal rat plasma, and are expressed as a percentage of the control concentration. Significant differences from the control (*P < 0.05) or from the corresponding value in group C ($P < 0.05) were identified using the data for coagulation times. factors II, VII, X and IX reached a minimum 36-60 hr after the administration of BHT. Taking the turn-over time of prothrombin into account, the effective period of activity of BHT or a BHT metabolite may be 16-40 hr after administration. This time coincides with the time at which a maximum hepatic level of the quinone methide of BHT occurs. This is also the time at which the level of BHT itself in the liver reaches its secondary peak. However we must rule out the effect of BHT itself, because no effect was detected at 27 hr--24 hr after the BHT level in the liver reached its maximum. Since tri-tert-butylphenol, a stronger anticoagulant to rats, cannot be converted to the corresponding quinone methide (Takahashi & Hiraga, 1983), it is not known whether BHT quinone methide has a direct effect on the synthesis of blood coagulation factors in vivo. However a close relationship between BHT quinone methide and the mechanism of BHT-induced haemorrhage is undeniable. Moreover, BHT quinone methide inhibits phylloquinone epoxide-dependent protein carboxylation in vitro (Takahashi & Hiraga, 1981b). The dose-response relationship between dietary BHT and haemorrhagic effects or the decrease in
prothrombin index was reported earlier (Takahashi & Hiraga, 1978a). Factor IX showed a significantly dose-related decrease when over 200 mg BHT/kg was ingested, but other factors were reduced only by 800 mg BHT/kg. When a purified diet was used as the basal diet, feeding of 120rag BHT/kg/day (0.2% in the diet) for 7 days caused hypoprothrombinaemia. In the case of a single oral administration, a much higher dosage is required to cause hypoprothrombinaemia. The injection of phylloquinone 24 hr after BHT administration could prevent the decrease in these factors, suggesting that the decrease in coagulation factors caused by a single oral dose of BHT may be a specific anticoagulant effect similar to the effect of dietary BHT (Takahashi & Hiraga, 1979b). The relation between haemorrhagic effects and BHT metabolism has also been reported (Takahashi & Hiraga, 1978c). Some metabolites generated through the major pathway, notably BHT alcohol (2,6-di-tert-butyl-4-hydroxymethylphenol), BHT aldehyde (3,5-di-tert-butyl-4-hydroxybenzaldehyde) and BHT acid (3,5-di-tert-butyl-4-hydroxybenzoic acid), do not have the same effect as BHT itself (Fig.
Rat-blood coagulation factors and BHT
223
1OO t •
10
O
A
B
C
D
F
G
Fig. 3. Effects of phenobarbital sodium pretreatment (B), 1% dietary cysteine (C), cobaltous chloride pretreatment (E), SKF 525A pretreatment (F) and SKF 525A post-treatment on the decrease in factors II (1~), VII (1~), X (~1) and IX (1~) measured in rat plasma 48 hr after a single oral dose of 800 mg BHT/kg (A and D). Data have been calculated from the mean clotting times of four or five rats, using dilution curves of normal rat plasma, and are expressed as a percentage of the control concentration. Significant differences from the control (*P < 0.05) or from the corresponding value in group A or D (tP < 0.05) were identified using the data for coagulation times.
H 0
: ~ cy~./,-4s0_ I CH~
~,,~ ~
(I)
-;
.~. 0
-.....
~
O
CH~
CH~
J
0
O
CH~
O
(vl CH~
Cov~lenl bindin i I o
-
SH o~GSH o~ proteins
CH~
OH
CH~OH (n)
(iv)
8 H O
~
l~G~,,~,o.id~ COOH (m~
Fig. 4. Molecular structures of BHT (1) and of 2,6-di-ter:-butyl-4-hydroxymethylphenol (]]), 3,5-ditert-buty]-4-hydroxybenzoic acid (]]]), 2,6-di-tert-butyl-4-methy[-4-hydroxy-2,5-cyclohexadienone (IV)
and 2,6-di-tert-butyl-4-methylene-2,5-cyclohexadienone (BHT quinone methide; V), the major metabolites of BHT in the rat. tlndicates a tert-butyl group.
224
O. TAKAHASHl
4). From studies of the hepatic metabolism of BHT In Principles and Methods of Toxicology. Edited by A. W. Hayes. p. 237. Raven Press, New York. and of species, strain and sex differences in the antioxidant's haemorrhagic effects and hepatic me- Knauer T. E., Siegfried C. M. & Matschiner J. P. (1976). Vitamin K requirement and the concentration of vitamin tabolism, we found that an amount of free BHT K in rat liver. J. Nutr. 106, 1747. quinone methide was generated in the liver of rats Nakagawa Y., Suga T. & Hiraga K. (1984a). Preventive given BHT. Since rats were the species most suseffect of cysteine on butylated hydroxytoluene-induced ceptible to BHT, we suspect that this quinone mepulmonary toxicity in mice. Biochem. Pharmac. 33, 502. thide may play an important role in the effects of Nakagawa Y., Tayama K., Nakao T. & Hiraga K. (1984b). BHT recorded in rats (Takahashi et al. 1980; TakaOn the mechanism of butylated hydroxytoluene-induced hashi & Hiraga, 1979a). hepatic toxicity in rats. Biochem. Pharmac. 33, 2669. We have not previously studied the effect of drug- Nakagawa Y., Tayama K., Nakao T. & Hiraga K. (1985). Effect of cobaltous chloride on butylated hydroxytoluenemetabolizing-enzyme modifiers. Nakagawa et al. induced hepatic necrosis in rats. Toxicology Lett. 24, 85. (1984a,b; 1985) found that dietary cysteine prevented BHT-induced lung injury in mice and demonstrated Owen C. A., Jr & Bowie E. J. W. (1978). Rat coagulation factors, V, VIII, XI and XII: vitamin K dependent. an inhibitory effect of cobaltous chloride on liver Haemostasis 7, 189. necrosis caused by BHT in rats. In our studies, Shah D. V. & Suttie J. W. (1979). Vitamin K-dependent neither pretreatment with phenobarbital nor an excarboxylase: liver activity in various species. Proc. Soc. cess intake of cysteine prevented the decrease in exp. Biol. Med. 161, 498. blood coagulation factors, but cobaltous chloride and Siegfried C. M., Knauer G. R. & Matschiner J. T. (1979). Evidence for increased formation of preprothrombin and SKF 525A completely or partially inhibited the the noninvolvement of vitamin K-dependent reactions in decrease. The lack of any detectable action by phenosex-linked hyperprothrombinemia in the rat. Archs Biobarbital may indicate that any active metabolite is chem. Biophys. 194, 486. generated by a minor and branched pathway. The sulphydryl group originating from dietary cysteine Takahashi O. (1985). Reduced aggregation capacity of washed platelets and dysfibrinogenaemiain the rat given cannot prevent the formation of the active metabolite butylated hydroxytoluene. Fd Chem. Toxic. 23, 937. of BHT. The induction mechanism for hypoproTakahashi O. (1986). Feeding of butylated hydroxytoluene thrombinaemia may be different from that of necrosis to rats caused a rapid decrease in blood coagulation by BHT since the effective doses and times for the two factors II (prothrombin), VII, IX and X. Archs Toxicol. 58, 177. effects differ (Nakagawa et aL 1984b). In the situation preceding the development of haemorrhage, hepatic Takahashi O., Hayashida S. & Hiraga K. (1980). Species differences in the haemorrhagic response to butylated glutathione decreases but is not totally depleted (O. hydroxytoluene. Fd Cosmet. Toxicol. 18, 229. Takahashi, unpublished results, 1986). Under these Takahashi O. & Hiraga K. (1978a). Dose-response study of conditions, the supply of additional sulphydryl hemorrhagic death by dietary butylated hydroxytoluene groups appears to have no significant effect on the (BHT) in male rats. Toxic. appl. Pharmac. 43, 399. activity of BHT quinone methide. The preventive or Takahashi O. & Hiraga K. (1978b). Effects of low levels of ameliorating effects of cobaltous chloride indicate a butylated hydroxytoluene on the prothrombin index of role for cytochrome P-450 in the generation of an male rats. Fd Cosmet. Toxicol. 16, 475. active metabolite. SKF 525A can also partially pre- Takahashi O. & Hiraga K. (1978c). The relationship between hemorrhage induced by butylated hydroxytoluene vent the decrease in coagulation factors, being more and its antioxidant properties or structural charactereffective when injected 12hr after BHT adminisistics. Toxic. appl. Pharmac. 46, 811. tration than when given as a pretreatment. Since the later administration of SKF 525A can reduce only Takahashi O. & Hiraga K. (1979a). 2,6-Di-tert-butyl-4methylene-2,5-cyclohexadienone:a hepatic metabolite of the level of the second (major) peak of the BHT butylated hydroxytoluene in rats. Fd Cosmet. Toxicol. 17, metabolite, this 24-hr peak must play an important 451. part in the reduction in the blood coagulation factors. Takahashi O. & Hiraga K. (1979b). Preventive effects of It is interesting that SKF 525A was effective in phylloquinone on hemorrhagic death induced by butylated hydroxytoluene in male rats. J. Nutr. 109, 453. maintaining normal liver weight only when injected Takahashi O. & Hiraga K. (1981a). Haemorrhagic toxicosis 12 hr after the ingestion of BHT. in rats given butylated hydroxytoluene. Acta pharmac. This experiment on drug-metabolizing-enzyme tox. 49, 14. modifiers indicated that BHT metabolism may play Takahashi O. & Hiraga K. (1981b). Inhibition of phylloan important role and an active metabolite may be quinone epoxide-dependent carboxylation of microsomal involved in the haemorrhagic effect of BHT. proteins from rat liver by 2,6-di-tert-butyl-4-methyleneAlthough we cannot conclude definitely that this 2,5-cyclohexadienone. Fd Cosmet. Toxicol. 19, 701. active metabolite is BHT quinone methide, there Takahashi O. & Hiraga K. (1982). Effects of butylated must be a close relationship between the haehydroxytoluene (BHT) on Fitzgerald factor, Fletcher factor and complement fixation activity of rat plasma. morrhagic effect of BHT and the formation of this Ann. Rep. Tokyo Metr. Res. Lab. publ. Hlth 33, 423. quinone methide. Takahashi O. & Hiraga K. (1983). Metabolic studies in the rat with 2,4,6-tri-t-butylphenol: a haemorrhagic antioxidant structurally related to butylated hydroxytoluene. REFERENCES Xenobiotica 13, 319. Fasco M. J. & Principe L. M. (1982). R- and S-warfarin Takahashi O. & Hiraga K. (1984). Effects of dietary inhibition of vitamin K and vitamin K 2,3-epoxide reducbutylated hydroxytoluene on functional and biochemical tase activities in the rat..L biol. Chem. 257, 4894. properties of platelets and plasma preceding the occurGad S. C. & Weil C. S. (1982). Statistics for toxicologists. rence of haemorrhage in rats. Fd Chem. Toxic. 22, 97.
0278-6915/87 $3.00+0.00 Copyright ~ 1987PergamonJournals Ltd
DECREASE IN BLOOD COAGULATION FACTORS II (PROTHROMBIN), VII, IX A N D X IN THE RAT AFTER A SINGLE ORAL DOSE OF BUTYLATED HYDROXYTOLUENE O. TAKAHASHI
Department of Toxicology, Tokyo Metropolitan Research Laboratory of Public Health, 24-1, Hyakunincho 3 chome, Shinjuku-ku, Tokyo 160, Japan (Received 10 June 1986) Abstract--Male Sprague-Dawley rats were given butylated hydroxytoluene (BHT) in a dose of 800 mg/kg body weight orally, and 0.5-72 hr later plasma concentrations of factors II, VII, IX and X and hepatic levels of BHT and BHT quinone methide (2,6-di-tert-butyl-4.methylene-2,5-cyclohexadienone) were determined. Levels of factors II, VII, X and IX were reduced 36-60 hr after BHT treatment, but by 72 hr, those most affected (VII and IX) showed some recovery and X had returned to normal. Hepatic levels of BHT reached a maximum 3 hr (a major peak) and 24 hr after BHT dosing and BHT quinone methide reached a maximum at 6 and 24 hr (a major peak). In rats given BHT orally in doses of 200, 400 and 800 mg/kg, factors II, VII and X decreased after 48 hr only in rats given the highest dosage, but factor IX was more susceptible to BHT and showed a dose-dependent decrease. Phylloquinone (1 me/rat) injected ip 24 hr after the administration of 800 mg BHT/kg maintained normal levels of factors VII and X and an almost normal level of factor IX, but had little effect on the level of factor II. In studies of the effects of drug-metabolizing-enzyme modifiers, neither ip pretreatment with 75 mg phenobarbital sodium/kg for 3 days nor the feeding of 1% cysteine in the diet throughout the experiment prevented the decrease in vitamin-K-dependent factors by 800 mg BHT/kg, but 2-day ip pretreatment with 60 mg cobaltous chloride/kg/day maintained normal levels of factors II and VII and reduced the BHT effect on factors IX and X. SKF 525A (50 mg/kg) injected ip either 30 min before or 12 hr after BHT treatment partially prevented the decrease in factors II, VII and X, or in all four factors, respectively. Thus the decrease in vitamin K-dependent factors may be the same with a single oral dose of BHT as with dietary BHT, and the anticoagulant effect may require the metabolic activation of BHT.
INTRODUCTION
When administered to male rats in the diet, butylated hydroxytoluene (BHT), a sterically hindered phenolic antioxidant, can cause haemorrhagic death in 9-37 days and a dose-related decrease in the prothrombin index (Takahashi & Hiraga, 1978a,b). In a search for the critical defect in this bleeding, we examined several haemostatic functions in rats given 1.2% BHT in the diet for 7 days (Takahashi, 1985; Takahashi & Hiraga, 1981a, 1982 & 1984) and found that BHT decreases the vitamin K-dependent coagulation factors and changes platelet function, vascular permeability and the activity of the kallikrein-kinin system, but does not affect fibrinolysis or the classical complement fixation process. Of those effects, the most remarkable are a decrease in factors II (prothrombin), VII, IX and X, and a reduction in aggregation of the washed platelets. Previously, an attempt was made to divide the period of pathogenesis, from the start of BHT feeding to the occurrence of hacmorrhaglc death, into four stages (Takahashi, 1986): stage I - - a decrease in blood coagulation factors II, VII, X and IX (from days 1-2 to days 3-4), stage I I - - a decrease in these coagulation factors and hacmorrhages in the epididymis (from days 3-5 to days 6-8), stage I I I - - a decrease in the four coagulation factors with hacmorrhages in several organs and a defect in platelet function (from days'6-8 to
days 8-10) and stage IV--massive haemorrhages and death (after days 9-10 and up to approximately l month), the median lethal time being about day 15 (Takahashi & Hiraga, 1978c). Our interests are now focused on a search for the mechanism of early bleeding, i.e. the haemostatic functions at stage 1. The previous finding of a significant decrease in factors on day 2 suggested that factors II, VII, IX and X were rapidly reduced immediately after the ingestion of BHT. The work reported here considered the time course of changes in the coagulation factors after a single oral BHT dose in relation to the metabolism of BHT, the dose response and the effect of drug-metabolizing-enzyme modifiers. MATERIALS AND METHODS
Chemicals. BHT, obtained from Tokyo Kasei Kogyo Co. (Tokyo) was dissolved in soya-bean oil for administration. 2,6-Di-tert-butyl-4-methylene2,5-cyclohexadienon¢ (BHT quinone methide) was synthesized as described previously (Takahashi & Hiraga, 1979a). Kaywan ®, a phylloquinone emulsion, was purchased from Eisai Co. (Tokyo). Phenobarbital sodium, cobaltous chloride and SKF 525A (proadifen hydrochloride), from Wako Pure Chemical Industry (Osaka) and Smith Klinc &
219
220
O. TAKAHASHI
French Laboratories (Philadelphia, PA, USA) were Spraguc-Dawley male rats weighing approximately dissolved in 0.9% NaCI solution. L-Cysteine was 300 g and fed a laboratory ration (CLEA CE-2). purchased from Tokyo Kasei Kogyo Co. ThromboTissue analyses. The livers were mixed with anhyplastin, cepharin (a platelet factor), 0.02 M-calcium drous sodium sulphate and homogenized with acetone (1 g tissue in 10mi acetone) using a Polytron chloride, Owren's Veronal buffer pH 7.35 and human and/or bovine plasma lacking factor II, VII, IX or X homogenizer (Kinematic GmbH, Lucerne, Switzerwere purchased from Dade Diagnostics Inc. (Miami, land). The acetone solution was evaporated under FL, USA). nitrogen and the residue was dissolved in hexanc for Animals and treatments. Male Sprague-Dawley analysis by gas chromatography in a Shimadzu rats, obtained when 4 wk old from CLEA Japan Inc. GC-9A gas chromatograph equipped with a flame(Tokyo), were fed a laboratory ration (CLEA CE-2) ionization detector (Shimadzu Inc., Kyoto). A glass for 1-2 wk, and then received a purified diet for 3 'column (3m x 3ram ID) was packed with Gas days before and after the administration of BHT. The Chrom Q containing 10% Silicone DC-200, while a purified diet, which has been described previously glass cartridge packed with a 3.5-cm length of silanized glass wool and ifiserted into the port chamber (Takahashi & Hiraga, 1978a), contained 24.0% casein, 25.0% sucrose, 41.4% starch, 5.0% soya-bean served as a pre-column. The operating temperatures oil, 4.0% salts, 0.45% vitamins and 0.15% choline were 180°C (column) and 250°C (detector and injecchloride. The purified diet was fed for 3 days before tion port). The carrier gas (nitrogen) flow rate was BHT administration because it took 3 days for the 50ml/min (Takahashi & Hiraga, 1979a). levels of blood coagulation factors, which were Statistical analysis. Variation of results was routinely expressed as the standard error of the mean. affected to some extent by the change of diet, to become stable (Takahashi, 1986). In the first experi- Significance of differences was tested on the data for ment, rats were given BHT orally by gavage in a dose coagulation times by Student's t test, taking P < 0.05 of 800 mg/kg and were killed 0.5, 3, 6, 12, 24, 48, 60 as the limit of significance (Gad & Wcil, 1982). and 72 hr later. Control rats were given soya-bean oil. There were between four and ten rats in each group. RESULTS The concentrations of blood coagulation factors II, VII, IX and X in the plasma and of hepatic BHT and Body-weight gain was reduced 24-72 hr after the BHT quinone methidc were determined. In the administration of BHT. Relative liver weights of second experiment, four or five rats/group were given control rats varied regularly between a larger size in BHT in doses of 200, 400 or 800 mg/kg and killed the morning and a smaller one at night, but in rats 48 hr later, for determination of plasma levels of the given BHT this regular oscillation disappeared and coagulation factors. In one group given 800mg 36-72 hr after BHT administration the relative liver BHT/kg, each rat also received by ip injection l mg weights were significantly larger than those in the phylloquinone 24hr after the BHT. In the third controls. No necrotic changes were observed, howexperiment, rats were pretreated ip with 75mg ever. Levels of factors II and IX showed a decrease phenobarbital/kg daily for 3 days, and 24 hr after the 36-72 hr after BHT administration, while factor VII final injection, the BHT was administered. Other rats was slightly decreased at 3 and 12 hr and markedly were fed purified diet containing I% cysteine 3 days decreased at 36-60hr. Factor X was decreased before administration of BHT and throughout the 36-60 hr after dosing but returned to normal by experiment and another group received 60 mg cobalt- 72hr. All four factors reached a minimum at ous chloride/kg sc daily for 2 days, followed 24 hr 48-60 hr (Fig. la). Hepatic levels of BHT showed after the second injection by BHT. A further group peaks at 3 and 24 hr after dosing, while BHT quinone was treated with SKF 525A 30 rain before or 12 hr methide peaked at 6 and 24hr (Fig. lb). The major after BHT administration. For all these five groups peaks were at 3 and 24 hr for BHT and BHT quinone (four or five rats/group), the dose of BHT was methide, respectively. 800 mg/kg, and the rats were killed 48 hr after BHT Factors II, VII and X were decreased 48 hr after treatment for the determination of plasma concen- BHT treatment only in rats given the highest dosage trations of blood coagulation factors. (800mg/kg), but factor IX decreased in a doseDetermination of factors H (prothrombin ), VII, IX dependent manner (Fig. 2). Phylioquinone (1 murat) and X. Whole blood was collected (under ether injected ip 24 hr after the administration of 800 mg anaesthesia) from the inferior vena cava into plastic BHT/kg completely prevented the decrease in the syringes containing 0.147 u-sodium citrate solution levels of factors VII and X, and partially overcame (citrate-blood, 1:9, v/v). Plasma was prepared by the effect on factor IX, but the reduction in factor II centrifugation at 740 g for 10 min. Factors II, VII and was not significantly affected (Fig. 2). X were measured by a prothrombin time assay. Each Treatment with phenobarbital or dietary cysteine 0.l-mi volume of a 1:5 dilution of rat plasma in did not prevent the BHT-dependent decrease in vitaOwren's Veronal buffer pH 7.35 was mixed with min K-dependent factors (Fig. 3). Cobaltous chloride 0.1 ml human and/or bovine plasma lacking factor II, completely prevented the decrease in levels of factors VII or X and the prothrombin time was measured. II and VII and partially prevented the decrease in the Factor IX was determined by the activated partial levels of factors X and IX, while SKF 525A given thromboplastin time assay technique, again using 30 rain before or 12 hr after BHT partially prevented human plasma lacking factor IX (Owen & Bowie, the decrease in factors II, VII and X, or in all four 1978; Takahashi, 1986; Takahashi & Hiraga, 1981a). factors, respectively (Fig. 3). The mean relative concentrations of each factor were In addition, the liver enlargement caused by BHT determined using plasma dilution curves of normal was not prevented by pretreatment with cobaltous
221
Rat-blood coagulation factors and BHT
2~°I
(a)
_
! ~ool~r,_.,'l~--",X-~ P."o"
_ : ~
/ )'o - ~ ' ' ~
~X ~~~ ~ ~ o
X
lie
,I
~
+o-
" ~
'k't
.....
-
I~'/"
~
~
~
~
8 ~
~
°
~
~. ......
~
~$
e
./. ~
~
o K
+.. . . . . . . . . ~
+x ~
./ ~
/
/ ~
~ ~" ~
~
~
~ ~
~x,~z
10
I
I
I
I
I
0
2O
~l/#ill \\x\k%~
~15
,oL t
III ,,
/
L/l\lt ,,, o L t ,~ 0
I#
[
%%% ,, ~,~
\t'"-..
NT/
~
l
12
%x
"~.~
~t-----__~x 24
36
48
........ 60
72
Time after BHT treatment ( h r )
Fig. 1. Changes in (a) the plasma concentrations of blood coagulation factors II (--) VII (--.) X (---) and IX ( - - - ) and (b) the mean hepatic levels (+SEM) of BHT ( ) and BHT quinone methide (---) in rats given a single oral dose of 800 mg BHT/kg. In (a) values have been calculated from the mean clotting times of five to ten rats, using dilution curves of normal rat plasma, and are expressed as a percentage of the control concentration. The values differing significantlyfrom the controls (tested using the data for coagulation times) are those for factor II at 36, 48, 60 and 72 hr, for factor VII at 3, 12, 24, 48 and 60 hr, for factor X at 3, 36, 48 and 60 hr and for factor IX at 36, 48, 60 and 72 hr. chloride or by SKF 525A given 30 min before BHT, but was completely inhibited by the administration of SKF 525A 12 hr after BHT.
DISCUSSION
When BHT is administered to rats in the diet, food intake and therefore BHT intake decrease temporarily as a result of the change from the control diet. This decrease may obscure any analysis of the early response to BHT. In a previous report, it was noted that the changes in blood coagulation factors in rats on day 1 of consumption of a BHT-containing diet were not significantly different from the control levels, but it appeared that some decrease in factors II, VII, X and IX occurred immediately after the ingestion of BHT (Takahashi, 1986).
In the case of a single oral dose of BHT, the plasma levels of blood coagulation factors II, VII, X and IX did not show any obvious change for 24 hr and it took about 48 hr after BHT administration for any notable decrease in these factors to occur. If we compare the present data with the previous results for dietary administration (although oral intubation by stomach tube may not necessarily be equal to administration in the diet), the decrease in factors must begin to occur on day 2 of BHT feeding. The plasma half-life of rat prothrombin is 6-8 hr and it may be totally replaced with new prothrombin in about 20 hr (Knauer et al. 1976; Siegfried et al. 1979). Warfarin can induce hypoprothrombinaemia 8-18 hr after its administration (Shah & Suttie, 1979; Siegfried et al. 1979). A tingle oral dose of R- or S-warfarin diminished the levels of factors VII, X and II in 16-24 hr (Fasco & Principe, 1982). The levels of
222
O. TAKAHASHI 1000
-
0
X
100
~
IIII1:::1 ~
Eii
~ :!:! /~ :.:. ,~ .:.: ,,~ :.:. z/.:.: /~ :.:.
o
z ~::~: 10
~ :'2~ ~ ":'i ~ ~.y
i~ ,:.: .. :.:.:
m
~
,,,,,...,!!!!!::i::[z ",~:,z
z z,
,,::: ~i~t !!~ Z~i:.i..t
11111:::I ~
""
rii='i
iiiiiiiii,, ~...........,, ~, ~ ...........
~, ,,,,~,~ ........ ..
.. .. .. ........... .
'. ". .".:.:.:. .
:':'1
IIII17771~ ,. ,. ~,.: : :,. ¢ ~ ~.:;~
.........
..
."~::::~ ..... '
I[III;;~Z ........
,~::::s z~::::~
- ~- ' • • ~
¢ ~;:;'i
Ill[I:::
,~...a
IIII1~i1~::i~ . . . . . . . . . . . . . . : ~
111 I~: : :I "~ ::::~
iii:
"
!H~!~::i]~:~l
'
H,,..
~ ~:':'~
~
'""'"'
II:::
, ~ ~q.:
Z~ ~ ~: :~: :::::
IIII~:::I ~
:::~
~
:':':1 I [I II: : :I "~ :.:.:1
~
~[~ii!~Z~ I II II :::I ~
~"
**>F
Z ~ ~ :::: /~ .:.: ~ ~ :::: */:.:.:
~ ~i~ ~ :2:: ¢~ .-..
A
8
C
0
Fig. 2. Plasma concentrations of factors II (llI~),VII (1~1),X (1~) and IX (l~l) in rats 48 hr after treatment with a single oral dose of (A) 200 mg BHT/kg, (B) 400 mg BHT/kg, (C) 800 mg BHT/kg and (D) 800 mg BHT/kg with phylloquinone (1 rag/rat ip). Data have been calculated from the mean clotting times of four or five rats, using dilution curves of normal rat plasma, and are expressed as a percentage of the control concentration. Significant differences from the control (*P < 0.05) or from the corresponding value in group C ($P < 0.05) were identified using the data for coagulation times. factors II, VII, X and IX reached a minimum 36-60 hr after the administration of BHT. Taking the turn-over time of prothrombin into account, the effective period of activity of BHT or a BHT metabolite may be 16-40 hr after administration. This time coincides with the time at which a maximum hepatic level of the quinone methide of BHT occurs. This is also the time at which the level of BHT itself in the liver reaches its secondary peak. However we must rule out the effect of BHT itself, because no effect was detected at 27 hr--24 hr after the BHT level in the liver reached its maximum. Since tri-tert-butylphenol, a stronger anticoagulant to rats, cannot be converted to the corresponding quinone methide (Takahashi & Hiraga, 1983), it is not known whether BHT quinone methide has a direct effect on the synthesis of blood coagulation factors in vivo. However a close relationship between BHT quinone methide and the mechanism of BHT-induced haemorrhage is undeniable. Moreover, BHT quinone methide inhibits phylloquinone epoxide-dependent protein carboxylation in vitro (Takahashi & Hiraga, 1981b). The dose-response relationship between dietary BHT and haemorrhagic effects or the decrease in
prothrombin index was reported earlier (Takahashi & Hiraga, 1978a). Factor IX showed a significantly dose-related decrease when over 200 mg BHT/kg was ingested, but other factors were reduced only by 800 mg BHT/kg. When a purified diet was used as the basal diet, feeding of 120rag BHT/kg/day (0.2% in the diet) for 7 days caused hypoprothrombinaemia. In the case of a single oral administration, a much higher dosage is required to cause hypoprothrombinaemia. The injection of phylloquinone 24 hr after BHT administration could prevent the decrease in these factors, suggesting that the decrease in coagulation factors caused by a single oral dose of BHT may be a specific anticoagulant effect similar to the effect of dietary BHT (Takahashi & Hiraga, 1979b). The relation between haemorrhagic effects and BHT metabolism has also been reported (Takahashi & Hiraga, 1978c). Some metabolites generated through the major pathway, notably BHT alcohol (2,6-di-tert-butyl-4-hydroxymethylphenol), BHT aldehyde (3,5-di-tert-butyl-4-hydroxybenzaldehyde) and BHT acid (3,5-di-tert-butyl-4-hydroxybenzoic acid), do not have the same effect as BHT itself (Fig.
Rat-blood coagulation factors and BHT
223
1OO t •
10
O
A
B
C
D
F
G
Fig. 3. Effects of phenobarbital sodium pretreatment (B), 1% dietary cysteine (C), cobaltous chloride pretreatment (E), SKF 525A pretreatment (F) and SKF 525A post-treatment on the decrease in factors II (1~), VII (1~), X (~1) and IX (1~) measured in rat plasma 48 hr after a single oral dose of 800 mg BHT/kg (A and D). Data have been calculated from the mean clotting times of four or five rats, using dilution curves of normal rat plasma, and are expressed as a percentage of the control concentration. Significant differences from the control (*P < 0.05) or from the corresponding value in group A or D (tP < 0.05) were identified using the data for coagulation times.
H 0
: ~ cy~./,-4s0_ I CH~
~,,~ ~
(I)
-;
.~. 0
-.....
~
O
CH~
CH~
J
0
O
CH~
O
(vl CH~
Cov~lenl bindin i I o
-
SH o~GSH o~ proteins
CH~
OH
CH~OH (n)
(iv)
8 H O
~
l~G~,,~,o.id~ COOH (m~
Fig. 4. Molecular structures of BHT (1) and of 2,6-di-ter:-butyl-4-hydroxymethylphenol (]]), 3,5-ditert-buty]-4-hydroxybenzoic acid (]]]), 2,6-di-tert-butyl-4-methy[-4-hydroxy-2,5-cyclohexadienone (IV)
and 2,6-di-tert-butyl-4-methylene-2,5-cyclohexadienone (BHT quinone methide; V), the major metabolites of BHT in the rat. tlndicates a tert-butyl group.
224
O. TAKAHASHl
4). From studies of the hepatic metabolism of BHT In Principles and Methods of Toxicology. Edited by A. W. Hayes. p. 237. Raven Press, New York. and of species, strain and sex differences in the antioxidant's haemorrhagic effects and hepatic me- Knauer T. E., Siegfried C. M. & Matschiner J. P. (1976). Vitamin K requirement and the concentration of vitamin tabolism, we found that an amount of free BHT K in rat liver. J. Nutr. 106, 1747. quinone methide was generated in the liver of rats Nakagawa Y., Suga T. & Hiraga K. (1984a). Preventive given BHT. Since rats were the species most suseffect of cysteine on butylated hydroxytoluene-induced ceptible to BHT, we suspect that this quinone mepulmonary toxicity in mice. Biochem. Pharmac. 33, 502. thide may play an important role in the effects of Nakagawa Y., Tayama K., Nakao T. & Hiraga K. (1984b). BHT recorded in rats (Takahashi et al. 1980; TakaOn the mechanism of butylated hydroxytoluene-induced hashi & Hiraga, 1979a). hepatic toxicity in rats. Biochem. Pharmac. 33, 2669. We have not previously studied the effect of drug- Nakagawa Y., Tayama K., Nakao T. & Hiraga K. (1985). Effect of cobaltous chloride on butylated hydroxytoluenemetabolizing-enzyme modifiers. Nakagawa et al. induced hepatic necrosis in rats. Toxicology Lett. 24, 85. (1984a,b; 1985) found that dietary cysteine prevented BHT-induced lung injury in mice and demonstrated Owen C. A., Jr & Bowie E. J. W. (1978). Rat coagulation factors, V, VIII, XI and XII: vitamin K dependent. an inhibitory effect of cobaltous chloride on liver Haemostasis 7, 189. necrosis caused by BHT in rats. In our studies, Shah D. V. & Suttie J. W. (1979). Vitamin K-dependent neither pretreatment with phenobarbital nor an excarboxylase: liver activity in various species. Proc. Soc. cess intake of cysteine prevented the decrease in exp. Biol. Med. 161, 498. blood coagulation factors, but cobaltous chloride and Siegfried C. M., Knauer G. R. & Matschiner J. T. (1979). Evidence for increased formation of preprothrombin and SKF 525A completely or partially inhibited the the noninvolvement of vitamin K-dependent reactions in decrease. The lack of any detectable action by phenosex-linked hyperprothrombinemia in the rat. Archs Biobarbital may indicate that any active metabolite is chem. Biophys. 194, 486. generated by a minor and branched pathway. The sulphydryl group originating from dietary cysteine Takahashi O. (1985). Reduced aggregation capacity of washed platelets and dysfibrinogenaemiain the rat given cannot prevent the formation of the active metabolite butylated hydroxytoluene. Fd Chem. Toxic. 23, 937. of BHT. The induction mechanism for hypoproTakahashi O. (1986). Feeding of butylated hydroxytoluene thrombinaemia may be different from that of necrosis to rats caused a rapid decrease in blood coagulation by BHT since the effective doses and times for the two factors II (prothrombin), VII, IX and X. Archs Toxicol. 58, 177. effects differ (Nakagawa et aL 1984b). In the situation preceding the development of haemorrhage, hepatic Takahashi O., Hayashida S. & Hiraga K. (1980). Species differences in the haemorrhagic response to butylated glutathione decreases but is not totally depleted (O. hydroxytoluene. Fd Cosmet. Toxicol. 18, 229. Takahashi, unpublished results, 1986). Under these Takahashi O. & Hiraga K. (1978a). Dose-response study of conditions, the supply of additional sulphydryl hemorrhagic death by dietary butylated hydroxytoluene groups appears to have no significant effect on the (BHT) in male rats. Toxic. appl. Pharmac. 43, 399. activity of BHT quinone methide. The preventive or Takahashi O. & Hiraga K. (1978b). Effects of low levels of ameliorating effects of cobaltous chloride indicate a butylated hydroxytoluene on the prothrombin index of role for cytochrome P-450 in the generation of an male rats. Fd Cosmet. Toxicol. 16, 475. active metabolite. SKF 525A can also partially pre- Takahashi O. & Hiraga K. (1978c). The relationship between hemorrhage induced by butylated hydroxytoluene vent the decrease in coagulation factors, being more and its antioxidant properties or structural charactereffective when injected 12hr after BHT adminisistics. Toxic. appl. Pharmac. 46, 811. tration than when given as a pretreatment. Since the later administration of SKF 525A can reduce only Takahashi O. & Hiraga K. (1979a). 2,6-Di-tert-butyl-4methylene-2,5-cyclohexadienone:a hepatic metabolite of the level of the second (major) peak of the BHT butylated hydroxytoluene in rats. Fd Cosmet. Toxicol. 17, metabolite, this 24-hr peak must play an important 451. part in the reduction in the blood coagulation factors. Takahashi O. & Hiraga K. (1979b). Preventive effects of It is interesting that SKF 525A was effective in phylloquinone on hemorrhagic death induced by butylated hydroxytoluene in male rats. J. Nutr. 109, 453. maintaining normal liver weight only when injected Takahashi O. & Hiraga K. (1981a). Haemorrhagic toxicosis 12 hr after the ingestion of BHT. in rats given butylated hydroxytoluene. Acta pharmac. This experiment on drug-metabolizing-enzyme tox. 49, 14. modifiers indicated that BHT metabolism may play Takahashi O. & Hiraga K. (1981b). Inhibition of phylloan important role and an active metabolite may be quinone epoxide-dependent carboxylation of microsomal involved in the haemorrhagic effect of BHT. proteins from rat liver by 2,6-di-tert-butyl-4-methyleneAlthough we cannot conclude definitely that this 2,5-cyclohexadienone. Fd Cosmet. Toxicol. 19, 701. active metabolite is BHT quinone methide, there Takahashi O. & Hiraga K. (1982). Effects of butylated must be a close relationship between the haehydroxytoluene (BHT) on Fitzgerald factor, Fletcher factor and complement fixation activity of rat plasma. morrhagic effect of BHT and the formation of this Ann. Rep. Tokyo Metr. Res. Lab. publ. Hlth 33, 423. quinone methide. Takahashi O. & Hiraga K. (1983). Metabolic studies in the rat with 2,4,6-tri-t-butylphenol: a haemorrhagic antioxidant structurally related to butylated hydroxytoluene. REFERENCES Xenobiotica 13, 319. Fasco M. J. & Principe L. M. (1982). R- and S-warfarin Takahashi O. & Hiraga K. (1984). Effects of dietary inhibition of vitamin K and vitamin K 2,3-epoxide reducbutylated hydroxytoluene on functional and biochemical tase activities in the rat..L biol. Chem. 257, 4894. properties of platelets and plasma preceding the occurGad S. C. & Weil C. S. (1982). Statistics for toxicologists. rence of haemorrhage in rats. Fd Chem. Toxic. 22, 97.