[CONTRIBVTION FROM THE LABORATOBIES OF THEROCKEFELLER INSTITUTE FOR MEDICAL RESEARCH]
CHEMICAL REACTIONS OF T H E NITROGEN MUSTARD GASES.' VI. T H E REACTIONS OF T H E NITROGEN MUSTARD GASES WITH CHEMICAL COMPOUKDS OF BIOLOGICAL INTEREST JOSElPH S. FRUTON? WILLIAM €I. STEIN, MARK A. STAHMANE: CALVIN GOLUMBIC'
AND
Received March 28, 1946
In a previous paper of this series (1) it was shown that the nitrogen mustards react readily in vitro with a number of functional groups of protein constituents. It was concluded, therefore, that the nitrogen mustards would be likely to react with cellular proteins in vivo. The experiments to be reported in this communication were undertaken to ascertain whether chemical substances of biological interest other than proteins could also be expected to react with the nitrogen mustards. It has been found that the nitrogen mustards can ccmbine with a large number of chemical ccmpounds essential to the econcmy of the living cell. Among the more significant of such compounds, in addition to proteins, may be mentioned several vitamins (nicotinic acid or its amide, pyridoxine, and thiamin), and organic phosphate compounds. The reaction of various substances with the nitrogen mustards as measured by their competition with alanine. Since it was desired to investigate a number of substances of widely different chemical structure, it was necessary to devise a simple general method to determine the reactivity of the nitrogen mustards towards a given substance. Certain of the observations reported in paper V of this series (1)form the basis for such a method. It will be recalled that the amino group of 2-alanine reacts a t pH 7.5-8 with the water-soluble transformation products formed when the nitrogen mustards are dissolved in an aqueous bicarbonate solution. When to the system containing alanine and a nitrogen mustard there is added a substance that also reacts with the nitrogen mustard, the amount of nitrogen mustard available for reaction with alanine is decreased, and consequently, the extent of the disappearance of amino nitrogen is reduced. Thus, the reaction between a nitrogen mustard and alanine may be used to determine whether the nitrogen mustard or its transformation products react with a given substance, and further to obtain an estimate of the rate of this reaction relative to the rate of the reaction of the nitrogen mustard with alanine. Obviously, alanine can be replaced as the reference compound by other substances that react with the nitrogen mustard a t a suitable rate. This method, in principle, is similar 1 This work was done in whole under Contract No. OEMsr-313 between The Rockefeller Institute for Medical Research and the Office of Scientific Research and Development, which assumes no responsibility for the accuracy of the statements contained herein. The experiments were performed during the period June 1942-January 1944. * Present address, Yale University, New Haven, Connecticut. a Present address, University of Wisconsin, Madison, Wisconsin. 4 Present address, University of Pittsburgh, Pittsburgh, Pennsylvania.
571
572
FRUTON, YTEIE;, STAHMANN, AND QOLUMBIC
to the competition method used by Ogston in his study of the chemical reactions of mustard gas (2). The competition method was previously used to demonstrate that the imidazole group of acetyl-dl-histidine reacts with methyl-bid@-chloroethy1)amine (MBA). The reaction of several other substances with the three nitrogen mustards, MBA, ethyl-bis(&chloroethyl)amine (EBA), and trid&chloroethyl)amine (TBA), has also been studied by this method and the results are presented in Table I. The data in Table I show that in the presence of 2-proline, the extent of the reaction of any of the above nitrogen mustards with alanine is slightly more than half that of the value obtained in the absence of proline. This indicates that the three nitrogen mustards react with proline, EBA being the most, and TBA the least, reactive. Although the possibility of esterification at the carboxyl cannot be excluded, it seems most likely that the reaction involves the imino group of proline. In view of the important role played by pyridine derivatives in biological systems, the strong competitive effect of pyridine, nicotinic acid, nicotinic acid amide, and pyridoxine is striking. The data in Table I indicate that all of the nitrogen mustards combine very readily with the pyridine nitrogen atom, MBA and EBA reacting with greater ease than TBA. The product of the reaction must be a quaternary pyridinium derivative. A strong competitive effect was also observed with 2-thiopyridine, while piperidine competes with alanine only to a slight extent. It is of interest that MRA reacts with adenosine and that thiamin also competes effectively with alanine for reactions with both MBA and EBA. In the case of thiamin, the possibility should be borne in mind that the reaction may involve secondary products formed from the vitamin at p H 8. Imidazole also competes effectively with the amino group of alanine in the reaction with each of the nitrogen mustards. MBA has the greatest tendency to combine with imidazole, EBA is slightly less reactive, and TBA is definitely the least reactive. The results given in Table I confirm the conclusion already reached (1) that the nitrogen mustards are capable of reacting with the imidazole nucleus of histidine. Moreover, the order of reactivity of the various nitrogen mustards towards imidazole itself, is the same as that postulated earlier for the reactivity of these agents towards the imidazole group of histidine (1). In paper V of this series, experiments were presented on the reactivity of the nitrogen mustards towards carboxyl groups. The results of further experiments bearing on this point are presented in Table I. It will be rioted that acetic acid exerts a negligible competitive eflect upon the reaction of MEA or EBA with alanine, but does compete effectively in the case of TBA. Hence, TBA is definitely the most reactive of the nitrogen mustards towards the carboxyl group of acetic acid. A comparison of the data on MBA reveals, however, that the reactivity of a carboxyl group depends upon the structurs of the carboxylic acid. Thus, hippuric acid is a slightly better competitor for MBA than is acetic acid, and carbobenzoxy-1-glutamic and carbobenzoxy-l-aspartic acids are both rather
NITROGEN MUSTARD GAS.
573
VI
effective competitors for MBA. In the latter cases it seems probable that it is the 7-carboxyl of glutamic acid and the ,%carboxyl of aspartic acid which are inTABLE I COMPETITIVE EFFECTO F VARIOUS SUBSTANCES ON THE REACTION O F THE NITROGEN MUSTARDS WITH THE AMINO GROUPOF ALANINE Concentration of reactants per cc.: For MBA; 0.127 m M of MBA, 0.536 m M of I-alanine, 0.536 m M of competing substance, 0.526 mM of NaHCOt; for EBA; 0.133 m M of EBA, 0.534 m M of 2-alanine, 0.534 m M of competing substance, 0.534 mM of NaHCOs; for TBA; 0.127 mM of TBA, 0.762 m M of 1-alanine, 0.762 m M of competing substance, 0.526 m M of NaHCOa. Temperature, 25.4'; pH 7.5-8.0; reaction period, for MBA and EBA, 15 minutes shaking and 20 hours standing; for TBA, 20 hours shaking. NHrN
DISAPPEARED PEX m u 01
SUBSTANCE ADDED
MBA,
EBA,
TBA,
&.EQWIV
6.SQUN
1.EQUIV.
- None . . . . 0.75 I-Proline .................... .41 Pyridine .................... .04 Nicotinic acids. . . . . . . . . . . . . . .09 Pyridoxine. . . . . . . . . . . . . . . . . . 2-Thiopyridine, . Piperidine. . . . . . . . . . . . . . . . . . .
................ Thiamineb. . . . . . . . . . . . . . . . . . . Imidazmole. . . . . . . . . . . . . . . . . . . Hippuric acida. . . . . . . . . . . . . . Acetic acids.. . . . . . . . . . . . . . . . Carbobenzoxy - I - glutamic acid- . . . . . . . . . . . . . . . . . . . . . Carbobenzoxy - 1 - aspartic acida . . . . . . . . . . . . . . . . . . . . . Carbobenzoxy - d l - methio, nine. . . . . . . . . . . . . . . . . . . . . .
Ethyldiethanolaminec.. . . .
.04
.20
1.15 0.68 * 37 .24 .36
.30 .05 .61 $46
.02 .05
.38
.ll
.29 .18
.63
79
.85
-12 .65 .72 .44
0.86 * 44 .15 .17
*
MBA [.EQUIV % --
0.34 .71 -66 .71 * 45 .70 .14 .29 .64 .63 .10 .03
45 95 88 95 60 93 19 39 85 84 13 4
.31 41
.84
EBA !.EQVIV.
TBA % -
% -
0.42 .71 .69 .66
49 83 80 77 .84 98 .81 94
0.47 40 .78 70 .91 80 .79 70 .75 65
.57 66 -68 79
.52
45
.07
8
.30
25
02
2
*
.31 41
.44 .66 -50 .50 .32
DECREASE I N ALANINE PZACTTNG WITE
-54 .53 .46
.67 .80 .85
.09 12 .25 33 .25 33 .43 57
.32 37 .33 38
.40
.48 40 .35 30 .30 25
47
Employed as sodium salts. The amino group of this substance does not react appreciably with nitrous acid in the 5-minute period required for complete deamination of alanine. c Th'a solutions of these substances were brought t o pH 8 by the addition of HCI prior to the addition of NaHCOs. d Separate experiments have shown that when adenosine reacts with MBA, the amount of adenine amino nitrogen is decreased. a
volved. In contrast to the results with MBA, carbobenzoxy-Z-glutamic acid does not compete at all with alanine for reaction with EBA. The results presented here, coupled with those given earlier (l),make it appear probable that
574
FRUTON, STEIN, STAHMANN,
AND
GOLUMBIC
the order of reactivity of the nitrogen mustards towards carboxyl groups is TBA > MBA > EBA. The competitive effect of a derivative of methionine was also studied. It will be noted that, in the reaction with MBA, carbobenzoxy-dl-methionine has the same competitive effect as hippuric acid. It cannot be decided on the basis of this experiment alone, therefore, whether the slight reaction of carbobenzoxy-dlmethionine with MBA involves the thioether sulfur or the free carboxyl group. The experiments with thiodiglycol [his@-hydroxyethyl)sulfide], however, prove definitely that all of the nitrogen mustards are capable of reacting with thioether sulfur, with the formation of a sulfonium salt [cf. also (3)]. TABLE I1
THEREACTION OF
(MBA) AND ETHYL-BIS(~-CHLORO(EBA) wIrH PHOSPHATE Concentration of reactants per cc.: 0.10 m M of MBA.HC1 or EBA.HC1; 0.40 mM of Na2HP04;0.32 mM of NaHCOJ. Temperature 25'; pH 7.5. METHYL-BIS(P-CHLOR0ETHYL)AMINE ETHYL)AMINE
~
~~
~~
INORGANIC PHOSPHATE'
TIME, HOURS
B.EhC1E.D PER
I
MBA. m M
0.7
3 6 24 48
mu
OF
EBA, m M
1.0 1.3 1.4 1.5
1.0 1.3 1.3
Determined by the Fiske-SubbaRow method.
The fact that the nitrogen mustards are capable of reacting with tertiary amines is indicated by the results obtained with triethanolamine, methyldiethanolamine, and ethyldiethanolamine. The reactivity of the various nitrogen mustards towar& these bases seems to be about the same, with the exception of MBA which reacts readily with methyldiethanolamine. The product of this reaction has been isolated, and has the formula (I) given below. CHa
HO CH, CH2
CHS
CHa
\ I NCE-12CHZNCH, CH2NI / +
HO CH, ch2
c1-
+ \
\
CHz CHI OH
c1-
(1) Since at one time it was suspected that this product might be formed during the reaction of MBA with water, the toxicity of compound I was investigated. On intraperitoneal injection into mice, compound I was found to possess an LDoo of about 1 g./kg. In addition to the compounds listed in Table I, it should be mentioned that, a t pH 7.5, MBA and EBA react with inorganic phosphate (TBA was not investigated). In Table I1 data are presented on the rate of the disappearance of
NITROGEN MUSTARD QAS.
575
VI
inorganic phosphate when MBA or EBA are treated with aqueous NaaHPO,. Gilman, Goodman, and Philips (4) have reported that MBA and EBA react readily with phosphate and, under their experimental conditions, consume about one equivalent of phosphate per mole of nitrogen mustard. Additional evidence for the reaction of MBA with phosphate-containing compoundis has been obtained by use of the alanine competition method. Secondary sodium phosphate, sodium pyrophosphate, sodium glycerophosphate, several hexose phosphates, and two nucleotides (barium salts of cytidine diphosphate and adenosine triphosphate) were allowed to compete with the a-amino group of TABLE I11 T H ECOMPETITIVE EFFECTOF PHOSPHATES AND OF SUBSTANCES RELATED TO NUCLEIC ACID ON THE REACTION OF METHYL-BIS(~~-CHLOROETHYL)AMINE (MBA) WITH THE AMINOGROUPOF ALANINE Concentration of reactants per cc.: 0.133 m M of MBASHCl; 0.534 m M of Lalanine; 0.534 m.M of competing substance; 0.526 m M of NaHC08. pH ;'.5-8.0. Reaction period, 20 hours a t 25". -.
"7-N SUBSTANCE
................................ Na glycerophosphate Fructose 6-phosphate Glucosi. 3-phosphate.. . . . . . . . . . . . . . . . . . . . Glucose 6-phosphate. . . . . . . . . . . . . . . . . . . . . Cytidine diphosphate Ba. . . . . . . . . . . . . . . . Adenosine triphosphate Ba. . . . . . . . . . . . . . Theophylline glucoside. . . . . . . . . . . . . . . . . . Desoxyribose. . . . . . . . . . . . . . . . . . . . . . . . . . . .
WITH
MBA
m M MBA, M.EQWV.
None. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
DECREASE IN ALANINE PEACTINO
DISAPPEARED P E P
0.77 .41 -46 .53 .20 .04 .58 .20 .31 .31 .79 .74
Y.EQWIV.
0.36 .31 .24 .57 -73 .I9 .57 .46 .46 - .02 .03
t 47 40 30 74 95 25 74 60 60 0 4
alanine for reaction with MBA. The results are presented in Table 111. It will be noted that, in addition to sodium phosphate and pyrophosphate, the organic phosphate compounds also show evidence of reaction with MBA. Ogston (2) has shown that mustard gas reacts with pyrophosphate and glycerophosphate. It is of interest that the nucleoside, theophylline glucoside, which does not contain a phosphate group does not appear to react with MBA; neither does desoxjnibose, which is a constituent of certain nucleic acids. It may be concluded, therefore, that nucleic acids are capable of reaction with the nitrogen mustards through their phosphate groups as well as through the 6-amino group of the adenine residue (cf. Table I, footnote d). The extensive reaction of MBA with hexose phosphates suggests the possibility that the nitrogen mustards may influence the course of carbohydrate metabolism by reaxting not only with some of the enzymes involved [cj. references 10-18 in the preceding paper of this series (l)],but with some of the substrates as well.
576
FRUTON, S T E I N , S T A H M A N N , AND GOLUMBIC
Organic phosphates such as triose and hexose phosphates, adenylic acid, adenosine triphosphate, thiamin pyrophosphate, creatine phosphate, etc., are of importance in the maintenance of normal metabolic processes. Moreover, nucleic acids, phospholipids, and certain proteins are esters of phosphoric acid. The fact that MBA reacts with such organic phosphates is, therefore, of interest for the problem of the physiological action of the nitrogen mustards. The use of thiosulfate in the study of the reaction of the M B A system with various substances. In addition to the alanine competition method described above, a second method for determining whether a given substance reacts with a nitrogen mustard has been employed. The method has only been worked out and applied in the case of MBA, although it undoubtedly is applicable to other nitrogen mustards as well.
TABLE I V FORMS O F METHYL-BIS(@-CI~LOROETHYL AMINE (MBA) I N THE PRESENCE OF VARIOUSSUBSTANCES Concentration of reactants per cc.: 0.127 m M of MBAeHCl; 0.536 mM of added substance; 0.526 mM of NaHCOs; 0.127 mM of KaOH. p H , 7.5-8.0. Temperature, 25.4".
THEDISAPPEARANCE O F THE
ETHYLENIMOKIUM
M D E D SUBSTANCE
None. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . E -Alanine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I-Proline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Hexamethylene tetramine. . . . . . . . . . Pyridine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Nicotinic acid.. . . . Imidazole. . . . . . . . . . . . . . . . . . . . . . . . Histidine . . . . . . . . . . . . . . . . . . . . . . . . Sodium acetate.. ....................... Ammonium chloride.. . . . . . . . . . . . . . . . . . .
1
I
IO-MINUTETHIOSULFATE TITEX PEP TC. REACTION YIXTURE APTEX
40 min.. M.EQUIV.
0.126 .loo .os0 .Ooo
!40 min
0.098 .048 .032 .000
.004
.om
.002 .014 .036 .118 .114
.Ooo .Ooo .006
.OX .076
, Y EQUIV.
0.074 .030
.OlO
.056 .054
As has been shown earlier ( 5 ) , when MBA reacts with an aqueous bicarbonate solution, there occurs a progressive decrease in the concentration of ethylenimonium ions in the solution. The approximate ethylenimonium content is determined after various time intervals by measuring the thiosulfate consumption of aliquots of the reaction mixture. When alanine is also present in the reaction mixture, the rate of the disappearance of the ethylenimonium ion is accelerated. Thus, by measuring the rate a t which the ethylenimonium ion disappears in the presence of a given substance, such as alanine, it is possible to determine whether the substance reacts with the MBA system. Several substances have been tested in this manner. The results are presented in Table IV. Like alanine, proline produces an accelerated disappearance of the ethylenimor,ium ion. The effect of proline is even somewhat greater than that observed with alanine. This result, when combined with the finding in Table I, may be taken to indicate that proline reacts with the ethylenimonium form of
KITROQEN MUSTARD GAS.
VI
577
MBA. The presence of pyridine, nicotinic acid, imidazole, and histidine results in a nearly complete disappearance of the ethylenimonium ion in 40 minutes. These findirgs are in accord with those reported in Table I, and indicate the high reactivity of these compounds towards MBA. Sodium acetate and ammonium chloride, on the other hand, cause but a slight acceleration in the rate of disappearance of the ethylenimonium ion. As will be noted from Table IV, hexamethylene tetramine causes a complete disappearance of the ethylenimonium form in 40 minutes. It became of interest t o determine, therefore, rvhether hexamethylece tetramine was more reactive than thiosulfate towards the nitrogen mustards. By means of 8 competition experiment using thiosulfate as the reference substance, it was found that hexamethylene tetramine is an effective competitor in the reaction of thiosulfate with MBA and EBA. The extent of the competition indicated, however, that thiosulfate was the more reactive of the substances. It should be mentioned that the reaction of hexamethylece tetramine with MBA has been investigated in detail by Gurin and co-workers (6). The thiosclfate method has also been applied to the study of the reactions of l-methyl-l-~p-hydroxyethyl)ethyle~~imonicm picrylsulfonate (11). This compound reacts ccmpletely vith thiosulfate in 10 minutes, one equiralent of thiosulfate beirg cocst-med ( 5 ) . Since none of the reaction products of I1 cocsume thiosulfate, the 10-minute thiosulfate titer is a direct measure of the concentration of I1 in the solution. Hence, if a given substance has reacted with 11, the thiosulfate titer of the reaction mixture must differ from that of a solution of II to which the substance has not been added. By the use of this method, I1 was found to react rapidly in bicarbonate solution with Z-prohe, nicotinic acid, imidazole, hexamethylene tetramke, methyldiethanolamine, and thiodiglycol (Table V). No appreciable reaction with sodium acetate was observed. Periodides of MBA and TBA. In the course of the investigations described in this arid preceding papers, it was observed that the nitrogen mustards form insoluble periodides. It has long been known that both organic and inorganic halides ccmbine n ith one or more molecules of free halogen to give perhalides, of which KI3, [ICH3)4SlI3. [(CH3)4N]ClI4,and [R3NH]I3are examples. Kith some t,ertiary amines, iodine forms products of the formula [R3NI]I (7, 8). Quaternary ammonium halides form perhalides of the type RdNeX, in ivhich X represents halogen and n may be 3, 5 , 7 , 9 or more (9). Likewise, the formation from mustard gas of perhalides of the type [(ClCH&H&S]X, in which n may be 2,4, or 6 has been noted (10). The formation of a periodide of MBA and TBA was demonstrated by adding a solution of the amine hydrochloride to an excess of 0.1 N iodine-potassium iodide solution. A purple crystalline solid separated which was removed by filtration. Titration of the filtrate with sodium thiosulfate showed that about 2 atoms of iodine had been removed per mole of nitrogen mustard hydrochloride. “Aged” solutions of MBA produced similar crystalline precipitates with iodine-potassium iodide solution, but consumed somewhat less iodine.
578
FRUTON, STEIN, STAHMANN, AND QOLUMBIC
11 REACTED
11 REACTED WITE ADDED SUBSTANCE
APTEP
ADDED SUBSTANCE
20 min.,
% 60 rnin., % 180min.,%
20 mio.,
__________--None.. ....................... 2-Proline...................... Nicotinic acido.. . . . . . . . . . . . . . . Imidazole.. . . . . . . . . . . . . . . . . . . Hexamethylene tetramine.. , . . Methyldiethanolamineb . . . . . . . Thiodiglycol . . . . . . . . . . . . . . . . . . Sodium acetate., . . . . . . . . . . . . .
3 61 55 35 100
35 32 9
17 66 75 52
32 87 100
58 52 32 97 32 29 6
84
100 60 38 16
% W min., % 180 min., %
93 64 35
49 58 35
55 68 52
43 21 -1
61 32 3
TABLE VI THELIMIT OF PRECIPITATION OF NITROGEN MUSTARDS AND RELATED SUBSTANCES BY 12-KI LIy1T OF PPEClPlTATION AFTEP ADINO PERIOD CO~OUND
I
30 min P.P.V."
1 hr.b P.P.Y.
MBAeHCl MBA
100 100 10,Ooo 6 equiv. NaHCOa MBA.HC1 100 1 - Methyl - 1 - (8-chloroethy1)ethyleni100
+
monium picrylsulfonate Methyl - p - chloroethyl - p - hydroxy4W ethylamine. HCl 400s Methyl - 8 - chloroethyl - p - hydroxyethylamine picrylsulfonate Methyl - p - chloroethyl - p - hydroxy4OOb 6 equiv. NaHC03 ethylamineeHC1 100s N,N' - Dimethyl - N,N' - bis(p-chloroethy1)piperazinium dichloride 100s N,N' - Dimethyl - N,N' - bis(p-chloroethy1)piperazinium dipicrylsulfonate 10,Ooo Methyldiethanolamine TBA-HCl 1,000 1,Ooo TBAaHCl 6 equiv. NaHCOa Triethanolamine 20,000 Tetraethanolammonium hydroxide 5,000
+
+
I
I
20 hrs;, P.P.Y.
12 25 12 6 12
6
25 12
100
100
100
200
100
100
6
6
6
6
5,000
5, 000 120 500 20,000
16 20,000 2,500
120
40 hrs. P3.V.''
12 25 12 50
120 1,0oO
NITROGEN MUSTARD GAS.
VI
579
It seemed of interest to determhe whether the formation of insoluble periodides could be used for the detection of the nitrogen mustards and their transformation products,. To determine the lowest concentration a t which iodine-potassium ioiide solution could be used to detect the nitrogen mustards, their transformation products or closely related compounds, an iodine solution was added to portions of the progressively diluted solution to be tested. The greatest dilution a t which a precipitate was formed was determined. The results listed in Table VI were determined as follows: To 5 cc. of the solution to be tested, 0.1 g. of sodium bicarbor!ate and 1.O cc. of 1 N iodine-potassium iodide solution were added. The highest dilution a t which a turbidity was produced was determined by ccmparison with a control test tube. This comparison was best made within about 10 mifiutes after adding the reagents by observing the tubes while they were strongly illuminated from the side. EXPERIMEKTAL
The reaction rf various substances with the nitrogen mustards as measured by thsir competition with alanine. The requisite r?rrount of nitrogen mustard.HC1 was added t o the solution containing K;s,HCOI, alanine, and the competing substance as well as the calculated quantity of KEOH to liberate the nitrogen mustard free base. The reaction mixture was shaken at 25.4" for 15 minutes and then left for 20 hours a t 25.4". In the case of TBA the shaking UP.S continued for 20 hours. One-cc. aliquots were withdrawn, diluted to 10 cc., and amino nitrogen determinations were carried out on I-cc. samples of the diluted solution (5 minutes' shaking in the Van Slyke apparatus). The pH of the undiluted reaction mixture was measured by means of the glass electrode. The reaction cf MBA with methyldiethanohwiine. A solution (10 cc.) of MBAeHCl containing 8 mM of MBA was shaken with 4 cc. of methyldiethanolamine (34mM) for one hour. The clear solution was left a t room temperature for 20 hours and then concentrated under reduced pressure. The syrup Bhich resulted Bas dissolved in 20 cc. of absolute alcohol and 100 cc. of acetone was added. The oil which separated crystallized on scratching and chilling. The substance was recrystallized twice from absolute ethanol; yield 2.7 g. (84%); m.p. 93-95". Anal. Calc'd for C I ~ H I , C I ~ N I O C,~45.6; : H, 9.5; K , 10.6; C1, 18.0. Found: C, 45.7; H , 9.6; N, 10.4; C1, 17.9. The use of thiosulfate in the study of the reaction of the M B A system with various substances. The requisite amount of MBASHCl was added t o the solution of IL'aHC03 to which had been added the test substance and the calculated quantity of NaOH to liberate the MBA base. The reaction mixture was shaken for 10-15 minutes a t 25.4" until a clear solution resulted a n a kept at this temperature during the experiment. Five-cc. aliquots were withdrawn after 40 minutes, 2 hours, and 4 hburs, and added t o 10 cc. of 0.1 N thiosulfate. After exactly 10 minutes, the unreacted thiosulfate was titrated with 0.1 N iodine. The experiments ~ i t compound h I1 were performed in the same manner except that no XaOH v a s added. Competition between thiosulfate and hezamethylene tetramine ( H M T )for reaction with the nitrogen mustards. MRA.HC1 (3.2 m M ) plus 3.2 mM of NaOH was treated with 12.8 mM of sodium thiosulfate in the presence of bicarbonate (pH 8:. After 2 hours at 25.4", the disappearance of thiosulfate had ceased; 6.29 m M of thiosulfate had reacted (98% of theory for the formation of the "Bunte salt"). When 3.2 m M of MBA was treated with a solution containing 12.8 m M of thiosulfate and 12.8 mM of HPvlT at pH 8, a t the end of the reaction (2 hours) only 4.27 m M of thiosulfate had reacted (67%). This result indicates that although HMT shows great reactivity with MBA, i t is not as reactive as thiosulfate.
580
FRUTON, STEIN, STAHMANN, AND QOLUMBIC
When 1.6 mM of EBA.HCI (plus 1.6 mM of NaOH) was allowed t o react with 5 mM of Na&Os in the presence of bicarbonate, the amount of thiosulfate that had reacted after 2 hours at 25’ was determined by iodometric titration and was found to be 3.2 mM (100% of the theory for the formation of the “Bunte salt”). However, when 1.6 mM of EBA was treated with 5 mM of HMT, only 2.24 mM of thiosulfate had reacted after 2 hours (70%). This result shows that, as in the case of MBA, EBA reacts readily with HMT, but more readily Rith thiosulfate. It is of interest that, although the reaction of EBA with thiosulfate is more rapid than is that of MBA, the competitive effect of HMT is nearly the same for EBA and for MBA (30 and 33Yc respectively). Periodidss of MBA and T B A . The reaction mixtures were made up t o contain per cc.: 0.02 mM of nitrogen mustard or nitrogen mustard hydrochloride and 0.06 m.equiv. iodine in potassium iodide. I n each case, the mixture was shaken for 6 hours, filtered, and the purple solid washed with water. The unused iodine in the filtrate was titrated with thiosulfate. The MBA hydrochloride (481 mg., 2.5 mM) consumed 5.07 m.equiv. (0.643 g.) of iodine and yielded 0.994 g. of a dark purple solid. The crude product melted a t 94-98’. The TBA hydrochloride (482 mg., 2.0 m M ) consumed 5.1 m.equiv. (0.652 9.) of iodine and yielded 0.968 g. of a dark purple solid. The crude product melted a t 74-78’. A 2% solution of 2.5 mM MBA was aged for 3 hours. The It-KI solution was then added. After a further 6 hours, 4.48 m.equiv. of iodine was consumed and 0.834 g. of purple solid was obtained. This crude product formed a dark melt at 15C-160°. Another sample of the 2% MBA solution was aged for 6 hours. After the addition of I,-KI and shaking for 6 hours, i t consumed 3.89 m-equiv. of iodine and 0.659 g. of purple solid was obtained. This crude product formed a dark melt a t 120-13O0. These melting points of the crude periodides are mentioned only to indicate that the periodides obtained from the fresh solutions and from the aged solutions were different.
The authors wish t o express their indebtedness to the late Dr. Max Bergmann for the constant advice and encouragement which he gave in the course of this research. Thanks are due also t o Miss Rosalind E. Joseph for assistance in the conduct of these experiments and t o Mr. Stephen M. Nagy, who performed the microanalyses reported in this paper. XEW YORK,N . Y. REFERENCES (1) FRUTON, STEIN,AND BERQMANN, J. Org.Chem., (paper V this series).
(2) OGSTON,(1941).O (3) GOLUMBIC, FRUTON, AND BERGMANN, J . Org. Chem., (paper VI1 this scries). (4) GILMAN,GOODMAN, AND PHILIPS (1943).b (5) GOLUMBIC, FRUTON, AND BEROYANN, J . Org.Chem., (paper I this series). ( 6 ) WILSON, V A R S , GURIN,BROWN,CRANDALL, AND DELLUVA(1943).b (7) RENSENAND NORRIS, A m . Chem. J., 18, 90 (1896’. (8) NohRIs AND FRANKLIN, Am. Chem. .I., 21, 499 (1899). (9) CHATTAWAY A N D HOYLE, J . Chem. SOC.,133, 654 (1923). (IO) SARTORI,“The War Gases,” pp. 23&232, New York 1940.
Unpublished data obtained in Great Britain.
* Unpublished data obtained in the United States.