(CANCER
RESEARCH
26 Part 1: 323-330, February
Enzyme-alterable
19661
Alkylating
Agents
IX. The Enzymatic Transformation of Some Nitrogen Mustards in
the Presence of Carbon Dioxide: Implications
in Respiration1
CHARLES E. WILLIAMSON,
JAMES G. KIRBY, JACOB
ARNOLD M. SELIGMAN, AND BENJAMIN WITTEN
I. MILLER, SAMUEL
SASS, STANLEY
P. KRAMER,
The Chemical Research Division, Chemical Research and Development Laboratories, Edgewood Arsenal, Maryland, and Departments
Surgery, Sinai Hospital of Baltimore, Inc., and The Johns Hopkins University School of Medicine, Baltimore, Maryland
Summary
In vitro kinetic studies in various biologic media have led to the
observation of an enzyme in blood serum that catalyzes the reac
tion between secondar}' nitrogen mustards and carbon dioxide
to form substituted oxazolidinones.
This reaction causes the very rapid degradation of certain
nitrogen mustards frequently employed in cancer chemotherapy.
Upon contact with blood or animal tissues, agents such as
degranol and nor-HN2 are immediately transformed into prod
more rapid reaction rates are observed in solutions containing
highly nucleophilic reagents such as thiosulfate ion (2). Secondary
nitrogen mustards form ethylenimines (Equation B) in aqueous
media. Alkylation of nucleophilic substances by the uncharged
ethylenimine IV occurs slowly, but alkylation by the charged
si>ecies III (7) occurs much more rapidly.
CO2 ^
R2NCOOe
+ CÃ-e
CH,N
\
CH2CH2C1
CH,
HN2
CH2CH2C1
(A)
I
CH2
CH2CH2C1
+
HN
of carbon dioxide in blood.
R2NH
CH,
CHjCHjCl
CH.N
ucts incapable of undergoing typical alkylation reactions. In
light of these observations réévaluation
of pharmacologie data
thus far collected on these nitrogen mustards is indicated.
Interpretation
of the kinetic data obtained indicates that the
enzyme catalyzes carbamate formation —a reaction of importance
for the transport
of
Cl©
(B)
H®
It is proposed that a normal metabolic function of this enzyme
is to catalyze this reaction during respiration.
Introduction
Because of their cancer chemotherapeutic
effects, the nitrogen
mustards have long been a subject of interest especially from the
viewpoint of their in vivo course of reaction. The biologic effects
of these drugs are generally ascribed to chemical alkylation of
susceptible functions such as sulfhydryl, amino, ring nitrogen
of purines and phosphate (e.g., Refs. 12, 15, 16).
The nitrogen mustards may be divided into 2 general cate
gories: (a) those that cyclize to form aziridinium ions, and (6)
those that cyclize to form ethylenimines.
Studies of the reaction mechanisms in aqueous media of tertiary
nitrogen mustards, such as Ar,Ar-bis(2-chloroethyl)-jV-methylamine (HN2), indicate that the reactions proceed via an aziridin
ium ion I (Equation A). It isthision that reacts with nucleophiles
present in solution. In 0.1 M phosphate buffer, the time to 50%
reaction of HN2 at pH 7.4, 37°C,is approximately 60 min,2 and
1Supported in part by the U.S. Army Edgewood Arsenal Chem
ical Research and Development Laboratories In-House Labora
tory Independent Research Program and in part by a research
grant (CA 02478) from the National Cancer Institute, USPHS,
Bethesda, Maryland.
Received for publication June 21, 1965.
FEBRUARY
196C
H,
CH2CH2C1
CH,CH,C1
IV
Y©= nucleophilic
ion
For this reason, the secondary amine mustards are generally
more reactive in solutions of lower pH, where the concentration
of the protonated form is higher. Two well-known compounds
of this type that have been used extensively in cancer chemo
therapeutic studies are Ar,JV-bis(2-chloroethyl)amine
(nor-HN2),
and 1,6-di-(2-chloroethylamino)-l,6-deoxy-n-mannitol
(mannitol mustard, degranol). The pKa's at 37°Cof these compounds
(pK„of 6.5 for nor-HN2 and of 7.2 for degranol) are such that
the ethylenimine formed on cyclization exists to a considerable
extent in the unprotonated form at physiologic pH. Although the
tertiary nitrogen mustard HX2 hydrolyzes at a reasonably rapid
rate2 when dissolved in an aqueous solution of pH 7.4, the sec
ondary nitrogen mustard, nor-HN2, does not (7).
*Unpublished data, this laboratory.
323
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1966 American Association for Cancer Research.
Williamson,
Kirby, Miller, Sass, Kramer, Seligman,
and Witten
In this report are presented data for a new rapid reaction that
occurs with some secondary nitrogen mustards in fresh human
and animal blood or serum. Although nor-HN2 was used as the
model compound in this investigation,
both nor-HN2 and degranol are degraded rapidly upon contact with fresh serum.
Evidence presented herein indicates that after the administra
tion of certain secondary nitrogen mustards, such as degranol,
to a cancer patient, the drug reacts in at least 2 ways.
As we reported earlier (17) the reactions of these compounds
in blood exhibit characteristics that suggest the participation of
an enzyme. A reaction between nor-HN2 and carbon dioxide to
produce 3-(2-chloroethyl)-2-oxazolidinone
has recently
been
observed (1, 11) (Equation C). Equation C proceeds in buffered
sodium bicarbonate solution without a catalyst. In like manner,
nor-HN2 would be expected to disappear also from fresh blood.
Although the reaction of nor-HN2 in blood would appear to
proceed in accordance with Equation C, the data presented herein
indicate the presence of a factor in blood that accelerates the
rate.
(C1CH2CH2)2NH +
CH2CH2-C1
Step'
CO2
C1CH2CH2N
r
CH2—CH
CÃ-e +
C1CH:2CH2N
O
VI
Materials
and
+
OF REACTION
RATE
OF NITROGEN
MUSTARDS
IN BLOOD.A method for the determination
of mustard in blood
was developed for these experiments. This method, which utilizes
the 4-(4'-nitrobenzyl)pyridine
reagent (NBP) (18), varies some
what from those previously used by other investigators and is
suitable, without substantial modification, for assay of a wide
variety of alkylating agents, including both nitrogen and sulfur
mustards.
Stock solutions containing 5.00 mg/ml of the mustards were
prepared
in A^N-dimethylacetamide
(DMA) or dimethylsulfoxide (DMSO). Aliquots were then introduced into heparinized blood. A concentration of 1% DMA or DMSO in blood was
normally employed.
In a typical measurement of the reaction rate of a mustard in
blood, 0.25 ml of the DMA stock solution was pipeted into 25 ml
of heparinized blood in a reaction vessel maintained at 37°C.
One-mi aliquots were removed at various intervals and de
livered to test tubes containing 4 ml of 1.5% NBP in acetonitrile
and 1 ml of 5% sodium perchlorate in acetonitrile. The contents
of the tubes wrere stirred well and were allowed to remain at
room temperature, for approximately
1 hr. The precipitated
proteins and red blood cells were removed by filtration, and 3
324
piperidine was added to develop a purple color, and the color
intensity was read immediately on a Klett-Summerson
photo
electric colorimeter (No. 59 filter). Suitable blanks were run with
each experiment.
Because of the volatility of the ethylenimine IV, it is important
that the reaction vessel remain closed during the determination
of reaction rates.
These experiments
were conducted in a 125-ml jacketed
Erlenmeyer flask. The internal reaction temperature was main
tained by means of a continuous flow of water at 37°Cthrough
the jacket.
The reaction between nor-HN2 and blood was studied at
various pH levels. The pH of blood prior to reaction with mus
tard was adjusted with either 1.5 N phosphoric acid or 1.5 N
sodium hydroxide for lower or higher pH values, respectively.
This procedure estimates the amount of alkylating agent re
maining in the aqueous phase of blood at any particular time
H®(C) during the determination,
but does not measure the amount of
agent that might be absorbed by various proteins. Such adsorp
tion by proteins and cells would lead to unexpected losses of
alkylating agent upon precipitation of the protein components
by acetonitrile.
The following experiment was therefore con
ducted to eliminate this as a possible mechanism in the inter
pretation of data. Samples of the precipitated proteins and cells
were collected from a typical determination
of the reaction rate
between nor-HN2 and fresh goat blood and analyzed for residual
alkylating agent. These samples were suspended in a volume of
5 N nitric acid that would yield a concentration of protein ap
proximating that found in intact blood. After being heated for
1 hr at 75°C,the mixtures were adjusted to pH 7.4 with 3 N
Methods
DETERMINATION
ml of the clear filtrate were pipeted into calibrated Klett tubes.
Into each tube was introduced 0.5 ml of 0.2 M acetic acid, and it
was immediately incubated in a heated bath to complete the
reaction of the residual mustard with the NBP reagent.3 The
tubes were cooled to approximately 20°Cin an ice bath, 1 ml of
sodium hydroxide, and NBP tests were performed on these
samples in the usual manner. Suitable controls containing norHN2 were employed. An insignificant quantity of alkylating
agent was found in the precipitated cells and proteins, confirming
that the disappearance
of nor-HN2 from fresh blood was not
due to an adsorption phenomonen.
DEACTIVATIONOF BLOOD AT pH 3. Subjecting fresh blood
plasma to an extreme of pH was found to be a convenient
method of obtaining inactive plasma. Fresh plasma was adjusted
to pH 3 with 5% phosphoric acid, evacuated at 20 mm of Hg,
and allowed to remain at room temperature for 30 min. The pH
was then readjusted to pH 7.4 with 3 N sodium hydroxide and 50
/xg/ml of nor-HN2 hydrochloride was added. The level of alkylat
ing activity
within 3 hr.
in the plasma
ABBREVIATED
METHOD
OF
showed
no appreciable
MEASURING
NOR-HN2 IN BIOLOGIC MEDIA. When
temperature,
THE
standard
pH, and time are maintained,
reduction
REACTIVITY
OF
conditions
of
the rate of reaction
3The time and temperature of incubation varied with the reac
tivity of the particular mustard under study. HN2 was heated to
50°Cfor 12 min, whereas nor-HN2 and degranol were heated to
75°Cfor 25 min.
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1966 American Association for Cancer Research.
VOL. 26
Enzymatic
of I101--HN2can be measured indirectly in terms of the quantity
of alkylating agent unreactive towards fresh blood.
The following procedure was utilized for this purpose: One ml
of blood was pipeted into a test tube containing 1 ml of 1.0 M
phosphate buffer, pH 7.4, and 1 ml of water. One ml of nor-HN2
hydrochloride (300 ¿ig/ml)in 0.01 M phosphoric acid solution
was added. The tube was well stirred and incubated for 3 hr at
37°C.A 1-ml aliquot was then removed, and the quantity of
unreactive alkylating agent was determined by the acetonitrileNBP method described herein. With each series of tests ap
propriate blanks were employed in which 1 ml of water or 0.01 M
phosphoric acid was substituted for blood or agent, respectively.
The results of these tests were reported as % of the added norHN2 that was unreaetive in the biologic medium. A higher per
centage value represents the slower disappearance of nor-HN2
from blood.
Stock solutions of nor-HN2 hydrochloride were normall}'
prepared in DMSO or DMA. On the day of test, dilutions of
1101--HX2hydrochloride were prepared such that the concentra
tion was 300 Mg/ml in 0.01 M phosphoric acid solution. The con
centrations of DMSO or DMA in the medium was maintained
below 1%.
CYCLIZATION
OF NOR-HN2 IN AQUEOUSBUFFKE.The cjT'lization rate of nor-HN2 in water was measured at pH 7.4, 37°C,
in 0.1 M phosphate buffer by titration of liberated chloride ion.
A modification of the mercuric nitrate-diphenylcarbazone method
(14) was used.
Mercuric nitrate (chemically pure) was prepared as a 0.002 N
solution in water, and diphenylcarbazone as a 0.5% solution in
ethyl alcohol. The mercuric nitrate was standardized with puri
fied potassium chloride in 80% alcohol, pH 3. The presence of
phosphate buffer showed no significant effect on the normality
determination. Nor-HN2 hydrochloride was dissolved in 0.1 M
phosphate buffer, pH 7.4, 37°C,such that the concentration was
approximately 0.25 mg/ml. From zero time, 5-ml aliquots were
pipeted into small flasks. The solutions were immediately acidi
fied to ])H 3 with nitric acid (approximately 3 drops of concen
trated acid). Ethyl alcohol (20 ml) and diphenylcarbazone (1015 drops) were added, and the aliquots were titrated with
mercuric nitrate solution. Approximately 1 equivalent of chloride
ion was obtained per mole of II, as required for the formation
of III. A plot of chloride ion liberated versus time showed a
half-life of cyclization of approximately 23 min.
PRKCYCLIZATION
OF
NOR-HN2
FOR
BLOOD
STUDIES.
Noi'-HN2
hydrochloride (50 /ug/ml) was dissolved in 0.1 M phosphate
buffer, pH 7.4, maintained at 37°Cby means of a thermostated
water bath. As cyclization proceeded, 1-ml aliquots were re
moved and incubated for 1 hr in test tubes containing 1 ml of
fresh goat blood maintained at pH 7.4, 37°C.(Within these test
tubes, the concentration of nor-HN2 was 25 Mg/ml in 50%
blood and 50% phosphate buffer.) One-mi aliquots were re
moved from the tubes after 1 hr, and the amount of residual
alkylating activity was determined by the NBP method as de
scribed previously. Suitable controls were made without blood
and others without nor-HN2.
Throughout all experimental procedures, it was necessary
to use concentrations of nor-HN2 below 0.1 M (2) in order to
avoid such side reactions as dimerization and polymerization,
which are known to occur in more concentrated solutions (9).
FEBRUARY
Results
Transformation
of Xitrogen Mustards.
IX
and Discussion
Nature of the Reaction between Nor-HXê and Blood
In this kinetic study of mustards in animal and human blood,
it was observed that nor-HN2 reacted at a rate that was much
faster than would be expected from theoretical considerations.
In Equation B, the reaction mechanism of nor-HN2 is illustrated
as it occurs normally in aqueous solution. Step 1 (cyclization)
begins upon contact with water, and nucleophilic ions (Step 2)
act upon the protonated cyclized form. For aliphatic nitrogen
mustards, in general, Step 1 is reasonably fast while Step 2 is
slow and is, therefore, the rate-controlling step in the over-all
process. The cyclic species accumulate to relatively high con
centrations in the reaction media even in the presence of the
highly nucleophilic thiosulfate ion (2). Since in the uncyclized
form nor-HN2 is a poor alkylating agent as compared to the
cyclized form, the over-all reaction rate as depicted in Equation
B can never exceed the rate of cyclization (Step 1). Because the
reaction of nor-HN2 in fresh blood occurs at a much faster rate
than Step 1, it does not proceed through the cyclic intermediate
ion III and, therefore, is not the normal alkylation reaction.
Quantitative measurement of the rate of cyclization of norHN2 in 0.1 M phosphate buffer, pH 7.4, 37°C(Step 1), revealed
a 1st order velocity constant of 3 X 10~2 min"1 (half-life = 23
min). In this system, the cyclized form of nor-HN2 is stable for
many hours. Because of the higher concentration of chloride
ion in blood as compared with the in vitro buffer experiments,
Step 1 in blood would be expected to proceed even slower than
in phosphate buffer. In contrast to its normal behavior in aqueous
media, 50% of the nor-HN2 disappeared from fresh pooled
human blood in approximately 2 min.4
The rate of disappearance of nor-HN2 decreases on dilution
of the blood with 0.1 M phosphate buffer, pH 7.4. The typical
reaction patterns of nor-HN2 in various concentrations of fresh
rabbit blood are shown in Chart 1. This chart illustrates 2 out
standing features of the reaction between nor-HN2 and blood:
(a) the decrease in reaction rate with decrease in concentration
of blood; and (6) the obvious retardation of the reaction rate in
its later stages as well as the fact that the reaction does not
proceed to completion.
Because the NBP test for the determination of remaining
alkylating agent measures both the uncyclized and cyclized
forms of nor-HN2 (Equation B), the alkylating activity remain
ing in the final stages of the reaction in Chart 1 could be one or
both of these species. Each curve in Chart 1, therefore, describes
2 reactions occurring simultaneously in blood: (a) a rapid reac
tion that destroys the nor-HN2, and (6) the formation of a 2nd
alkylating agent that is not destroyed by the active principle in
blood. Furthermore, the quantity of the latter appears to be
dependent on the over-all rate of reaction. Thus in Chart 1,
the slower reaction produces the greater quantity of alkylating
agent in a form unreactive toward blood.
The reaction pattern set forth in Chart 1 suggests a rapid
reaction between uncyclized nor-HN2 and blood. This is most
probably the reaction shown in liquation C, in which carbon
4 The time required for disappearance of half of the agent varied
from less than 1 min to several depending on the particular blood
sample used
1906
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1966 American Association for Cancer Research.
325
Williamson, Kirby, Miller, Sass, Kramer, Seligman, and Wüten
100
T
20
INCUBATION
CHART 1. Variation in reaction rate of nor-HN2 as a function
of the concentration of fresh rabbit blood. Blood was dissolved in
0.1 M phosphate buffer, pH 7.4. The reaction was maintained at
pH 7.4 and 37°C.The concentration of nor-HN2 was SO/iK/ml.
dioxide is responsible for the rapid reaction. Concurrently, the
cyclization process (Equation B), which begins immediately
when nor-H\2 is dissolved in blood, leads to the formation of
cyclized nor-HN2 (IV) that would normally be expected to react
only very slowly with blood.
To test this hypothesis, a series of experiments were per
formed to determine the stability of the eyclized form of norHN2 (IV) in fresh blood. Thus, nor-HN2 was incubated in pH
7. 4 buffer to yield various percentages of IV, which were then
allowed to react with fresh goat blood. Specifically, nor-HN2 (50
¿ig/ml)was dissolved in 0.1 M phosphate buffer, pH 7.4, main
tained at 37°C,and cyclization (Equation B) was allowed to
proceed. Periodically, aliquots containing various percentages of
cyclized product were removed and allowed to react with fresh
goat blood. As shown in Chart 2, the cyclized form of nor-HN2
is essentially unreactive towards fresh goat blood. The experi
mental curve depicted in Chart 2 exhibits a half-life of cycliza
tion of approximately 23 min. This value agrees with the cycliza
tion rate as measured by chloride ion determination herein. It
may therefore be concluded that any cyclized nor-HX2 present
in Chart 1 will also be relatively unreactive in fresh blood.
The reaction rate of uncyclized nor-HN2 in the presence of
blood is therefore best described by subtraction of the rate of
326
30
OF
40
NOR-HN2
50
IN PHOSPHATE
BUFFER,
b'O
VIN.
CHART 2. Inability of fresh goat blood to destroy precyclized
nor-HN2. The plot shows what percentage of alkylating activity
was unreactive towards fresh goat blood after incubation in 0.1 M
phosphate buffer, pH 7.4, for various periods of time at 37°C.The
concentration of nor-HN2 was 50/iK/mI.
cyclization of nor-HN2 from the over-all reaction rate. In this
study, the rates of reaction in blood are approximated in terms
of initial velocity, and it is assumed that the quantity of cyclized
material forming simultaneously during this period is negligible.
As a corollary, the quantity of unreactive alkylating agent
remaining in solution after reaction with blood is inversely pro
portional to the reaction rate. Accordingly, the quantity of al
kylating agent remaining after reaction with blood under stand
ard conditions was utilized as an inverse index of reactivity of
II with blood. It is noteworthy that this remaining alkylating
agent shows no measurable decline in blood after several hours.
To evaluate the role of carbon dioxide, the rate of the reaction
shown in Equation C was measured in sodium bicarbonate solu
tion strongly buffered with phosphate buffer at pH 7.4. The
reaction rate of Equation C versus various concentrations
of
sodium bicarbonate
in 0.25 M phosphate buffer, pH 7.4, at
37°C is plotted in Chart 3. In Table 1 are tabulated similar
values
media
species
course
that a
when nor-HN2 is allowed to react with various biologic
under identical conditions. The large deviations between
are outstanding
and incompatible
with the predicted
of Equation C in animal blood. The data in Table 1 show
concentration of approximately 0.2 M sodium bicarbonate
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1966 American Association for Cancer Research.
VOL. 26
Enzymatic
Transformation
of Nitrogen Mustards.
IX
If these deviations of reaction rates of nor-HN2 are to be as
cribed solely to the concentration of dissolved carbon dioxide
0.40
(e.g., free C02, bicarbonate, or carbamates), the quantities of
CC>2in guinea pig blood must be approximately 10-fold that in
mouse blood. The normal concentrations of total dissolved car
0.35
bon dioxide in mammalian blood is approximately 0.025 M
(5). Deviations as low as 0.015 Mhave been observed in diabetic
coma and nephritic death (13). Deviation as high as 0.035 M
0.30
has been obtained when large quantities of sodium bicarbonate
were experimentally administered to laboratory animals (13).
The differences observed in the reaction rates of nor-HN2 in
mammalian blood cannot, therefore, be attributed solely to a
0.25
change in blood COz concentration. Although the reaction
mechanism for the disappearance of nor-HN2 may be that illus
trated in Equation C, the rate at which this process occurs in
0.20
blood appears to be dependent upon factors other than the total
quantity of carbon dioxide present. Furthermore, mouse blood
does not possess the ability to increase the rate of Equation C
0.15
as shown in Chart 3.
To demonstrate the presence of a blood factor responsible
for this enhanced activity, pooled mouse serum and pooled guinea
0.10
pig serum were brought to pH 3 in separate experiments and
evacuated for 30 min at 20 mm of Hg to remove most of the dis
solved carbon dioxide. The pH of the sera was then returned
to 7.4 with dilute sodium hydroxide solution, and the samples
0.05
were then evaluated. Both were essentially unreactive toward
nor-HN2. To each serum was added 0.025 Msodium bicarbonate
(approximate physiologic concentration), and the sera were
0.04
0.05
0.03
0.02
0.01
again tested. The mouse serum reacted as might be expected
M
in the absence of a rate acceleration factor; it exhibited a rate of
disappearance of nor-HN2 consistent with Chart 3. The reaction
CHART3. Reaction of Nor-HN2 (75 jug/ml) with sodium bicar
bonate in 0.25 Mphosphate buffer, pH 7.4, 37°C.Pseudo 1st order rate of the guinea pig serum, however, was approximately 10
times greater than that of the mouse serum.
rate constants versus molarity of bicarbonate.
It appears, therefore, that the reaction between nor-HN2
and blood serum does not occur in the absence of carbon dioxide
TABLE 1
REACTIONRATE OF Non-HN2 IN VARIOUSBIOLOGICMEDIA but is accelerated beyond its normal rate in the presence of cer
tain mammalian sera. It furthermore appears that the factor
(min)0Biologicmediumc25%9.382.018.574.7Biolouicmedium**100%2.42.00.54.91.71.2MOLARITY
OFBICARBONATEREQUIRED
that promotes the rapid reaction between nor-HN2 and blood
TO
serum is not inactivated when the serum is held at pH 3 for 30
MEDIUMHuman,
BIOLOGIC
PRODUCESIMILARRATE60.0420.050.20.020.0620.085
min. This factor has been observed in the whole blood or serum
of humans, goats, dogs, rabbits, rats, and guinea pigs and in
muscle tissue from rabbits and rats, rat liver, and egg white,
bloodHuman,
whole
but not in mouse whole blood. The concentration of the factor
serumGuinea
in blood varies not only among species, but among animals of
bloodMouse,
pig, whole
the same species.
bloodRat,whole
bloodGoat,
whole
Mechanism of the Transformation of Xor-HNS in Aqueous Bi
whole bloodHALF-LIFE
carbonate and Blood
0 Initial velocity obtained from pooled blood or serum.
Measurement of the rate of disappearance of nor-HN2 by
'Molarity of NaHCOs in 0.25 M phosphate buffer, pH 7.4,
37°C,required to produce a reaction rate similar to that of 100% the NBP method showed that the pseudo 1st order rate con
stant of Equation C increases in a lineai' manner with increasing
biologic medium.
c Experimental values obtained in 0.25 Mphosphate buffer, pH concentration of carbon dioxide, as shown in Chart 3.
7.4, 37°C.
Since the NBP method measures only the rate of disappear
6Extrapolated.
ance of alkylating agent, additional evidence for the mechanism
of the disappearance was sought. For this purpose the rate of
formation of chloride ion was determined at 37°Cin 0.1 M
is required to obtain a reaction rate of the same velocity as ob
tained in fresh guinea pig blood, whereas only approximate!}- phosphate buffer, pH 7.4. In the range of concentration of 0.020.05 M sodium bicarbonate, the rates of disappearance of nor0.02 M sodium bicarbonate is required to obtain the same reac
tion rate as nor-HN2 in mouse blood.
HN2 as measured by the NBP method agreed within experimenFEBRUARY 1900
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1966 American Association for Cancer Research.
327
Williamson, Kirby, Miller, Sass, Kramer, Seligman, and Willen
tal error with the rates of formation of chloride ion (after
correction for Cl~ due to Equation B). Thus, the rates of disap
pearance of nor-HX2 in Chart 3 depict also the rates of 3-(2chloroethyl)-2-oxazolidinone
formation.
The linearity of Chart 3 indicates 2nd order kinetics. The ratecontrolling step in Equation C is therefore the bimolecular Step
1. In these studies only the forward reaction of Step 1 (Equa
tion C) was considered since the backward process is hindered
in basic media.5
In order to accelerate the over-all rate of Equation C, the
active blood factor must alter the rate-controlling
step (or both
steps). Since the rate-controlling step of Equation C iscarbamate
ion formation (Step 1), it appears that the function of the blood
factor is to accelerate this step.
Accordingly, one could logically suppose that the role of the
blood factor is to increase the availability of carbon dioxide in
accordance with the following equation:
Enzyme
H2CO3
H.O
5-
10-
15-
< 20-
(IV)
Equation IV is catalj-zed by the enzyme carbonic anhydrase (3).
The secondary nitrogen mustards might therefore be expected
to react more rapidly in the presence of this enzyme. To test
this hypothesis, several concentrations of carbonic anhydrase
were added to the buffered system, pH 7.4, containing nor-HN2
and sodium bicarbonate. No acceleration in the reaction rate was
observed. Furthermore, the reaction of nor-HN2 in fresh guinea
pig blood serum proceeded undisturbed in the presence of 10~2
30 -
M sodium cyanide, a concentration sufficient to cause serious
inhibition of carbonic anhydrase (4). It is concluded that car
bonic anhydrase plays little or no part in accelerating the reac
tion rate.
Properties of the Factor
THERMALINACTIVACIÓN.
To determine the thermal stability
of the factor, the carbon dioxide was removed from guinea pig
serum as above and the serum was heated to 75°Cfor 5 min.
Upon addition of 0.025 M sodium bicarbonate and reaction with
nor-HN2, a rate of disappearance approximating that shown in
Chart 3 was observed. The blood factor that promotes the reac
tion between nor-HN2 and blood serum appears to have been
destroyed by this treatment. It is, however, quite stable at
lower temperatures. Heating to 50°Cfor 5 days did not destroy
the activity. Thus, this substance exhibits an extremely high
temperature coefficient of thermal inactivation, a characteristic
5 The reaction between ammonia and carbon dioxide has been
studied by Pinsent et al. (10). The reactions involved are:
CO2 + NHs -»NH2COOS + Hs
(I)
H+ + NH3 -»NH4+
(II)
All the hydrogen ions produced in Reaction (I) are instantly trans
formed into ammonium ion by equation (II), and the over-all rate
is governed by the speed of Reaction (I), i.e.,
-4CO-2]
dt
The value of the velocity constant,
sec-' at 40°C.
328
k, was found to be 1130 m~'
40
pH
CHART 4. Variation of the reaction rate of nor-HN2 in freshly
drawn goat blood as a function of pH, 37°C.
of proteinaceous material. The active factor is normally stored as
lyophilized serum at refrigerator temperatures.
DIALYSIS.The factor described herein is not dialyzable. A
sample of guinea pig serum was dialyzed against running tap
water for 18 hr, and the carbon dioxide was removed by treat
ment under the acidic conditions described above. Upon addition
of 0.025 M sodium bicarbonate, the original high activity of the
serum was restored.
Comparison of Reactivity in Blood of Nor-HNe and HN2
The reaction between nor-HN2 and blood serum that has
been inactivated by removal of carbon dioxide is very slow. The
differences between active and inactivated freshly drawn goat
serum more clearly illustrate the effects of the blood reaction
described herein. When nor-HN2 was added to deactivated
freshly drawn goat serum at 37°C,pH 7.4, essentially no reac
tion occurred for at least 1 hr. In contrast, when nor-HN2 was
added to freshly drawn goat serum at 37°C,pH 7.4, a very rapid
reaction ensued. Approximately
peared in the 1st min.4
half of the mustard
disap
CANCER RESEARCH VOL. 26
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1966 American Association for Cancer Research.
Enzymatic
For comparison, the reactivity of the tertian- nitrogen mus
tard HN2 was tested in both inactive and active goat serum at
37°C,pH 7.4. In each, the reaction rate was identical. The halflife was approximately 9 min, a value consistent with the rate
of cyclization of HN2 and the nucleophilic capacity of serum.
Thus, the reaction of HN2 in freshly drawn goat blood serum
proceeds in a predictable manner and exhibits a reasonable reac
tion rate, whereas the reaction of nor-HX2 does not. Although
nor-HN2 would have been expected to be less reactive towards
fresh blood than HN2 heretofore, it reacts much more rapidly.
OPTIMUMpH. The rate of reaction between fresh goat blood
and 1101--HX2at 37°Cversus pH is plotted in Chart 4. The pattern
of variation in reaction rate as a function of pH illustrates an
optimum pH (7.2). It is noteworthy, however, that upon decreas
ing the pH below the pK„of nor-HN2 (6.5 at 37°C)the de
crease in reaction rate is more abrupt, suggesting that the opti
mum ])H shown in Chart 4 is altered owing to protonation of the
substrate, nor-HN2, at the lower pH values. The protonated
form of nor-HN2 is not available for rapid reaction in blood.
Due to the unavailability of free COs at high pH, the reaction
of nor-HX2 in the presence of sodium bicarbonate would be ex
pected to proceed more slowly as the pH is increased. The cy
clization process is pH dependent and proceeds more rapidly at
high pH. Both processes probably contribute to the steep slope
of Chart 4 at high pH.
Interpretation
The data presented reveal an ingredient in blood serum that is
capable of increasing the reaction rate between certain second
ary nitrogen mustards and carbon dioxide. Although this sub
stance does not react directly with nor-HX2, it causes an accel
erated loss of nor-HX2 in the presence of carbon dioxide.
This active principle of blood is evidently an enzyme but is
not carbonic anhydrase, is not inhibited by 10~2 M cyanide ion,
Transformation
of Xifrogen Mustards.
IX
these drugs. As an example, attempts have been made in the
past to design enzyme-activated drugs containing secondary
nitrogen mustards (8). The reaction described herein could
rapidly alter these mustards and prevent the desired effect.
Caution must therefore be exercised in predicting and interpret
ing the in vivo activity of newly designed mustards of this
type.
implications in Respiration
It is postulated herein that an enzyme in blood serum is
capable of accelerating the rate of carbamate ion formation as
depicted in Step 1 of Equation C. That this ingredient should
possess so high a specificity as to accelerate earbamate ion forma
tion of only certain secondary nitrogen mustards is unlikely.
Presently, however, we have no evidence to indicate that nonmustard secondary amines undergo catalysis by this enzyme to
form carbamate ion.
The in vivo formation of carbamate ion is of considerable bio
logic importance and has been shown to be involved in the res
piratory processes (6). A significant fraction of the total carbon
dioxide in the blood is transported in the form of carbamate ion
through the plasma proteins and hemoglobin (6).
Although amines react with CO»to form carbamate ion, the
instability and reversibility of this compound has delayed the
clear demonstration of an enzyme that catalyzes this process.
The discovery of the trapping function through cyclization of a
2-chloroethylamine group to form a stable oxazolidinone (Equa
tion C) has enabled us to demonstrate an enzyme that catalyzes
the 1st step of the reaction. It is probable that the action of this
enzyme is not limited to 2-chloroethylamines but operates with
other naturally occurring amines as well and thereby serves an
important function in CO»transport in the body.
Acknowledgments
and is nondialyzable. It is not inactivated when the blood is
The authors wish to tlmnk Mr. Raymond F. Meli ugh und per
held at pH 3 for 30 min, although gradual inactivation occurs
when it is allowed to stand at this pH for longer periods of time. sonnel of the Biophysics Division, Edgewood Arsenal, Maryland,
The active principle is destroyed when blood is heated to 70°C and Professor William F. Sager, University of Illinois, Chicago,
for 5 min, and it exhibits an extremely large temperature coeffi
cient of thermal inactivation. The amount of the factor in the
blood varies from animal to animal and among species. The
change in activity of the blood in various disease states in man
is now under investigation.
Thus far, only a few nitrogen mustards have been tested.
Degranol, a clinically useful cancer chemotherapeutic agent,
exhibits a half-life of approximately 2 min in fresh human blood.4
Other secondary nitrogen mustards reacted rapidly in fresh
blood only when the pK„was such that a considerable quantity
of the unprotonated amine was present at pH 7.4. The tertiary
nitrogen mustard HX2, however, exhibits a normal reaction
rate in blood and is not affected by the blood factor.
The reaction of certain nitrogen mustards with carbon dioxide
converts them rapidly to compounds that are not potent as
alleviating agents and whose pharmacologie properties have not
been investigated. When these mustards are employed as can
cer chemotherapeutic agents, or in other in vivo biologic studies,
most of the injected drug will be immediately transformed to
another chemical compound. It is therefore likely that mis
leading conclusions could be drawn from studies employing
FEBRUARY
1906
Illinois, for their valuable discussions and suggestions. We are
indebted to Messrs. Michael H. (¡oodwin and Eric Girarci for
technical assistance, and to Drs. Harry 13. Wood, Jr. and Abra
ham Qoldin, Cancer Chemotherapy
National Service Center,
Bethesda, Maryland, for their kind cooperation and for a supply
of degranol.
References
1. Arnold, H. and Bekel, H. Die Oxazolidon Reaktion von 2Chloroathylsubstituierten
sekundären Aminoli. ArzneimittelForsch., 14:750-52, 1904.
2. Cohen, C., Van Artsdalen, E. R., and Harris, J. Reaction
Kinetics of Aliphatic Tertiary 0-Chlorocthylamincs
in Dilute*
Aqueous Solution. I. The Cyclization Process. J. Am. Chem.
Soc., 70:281-85, 1948.
3. Davenport, H. W. Carbonic Anhydrase in Tissues Other than
Blood. Physiol. Rev., 20:5(10-73, 1946.
4. Davis, R. P. Carbonic Anhydrase. In: P. D. Boyer, II. Lardy,
and K. Myrbäck (eds.), The Enzymes, Vol. 5, p. 559. New
York: Academic Press, Inc., 1961.
5. Dittmer, D. 8., and Grebe, R. M. (eds.), Handbook of Res
piration, 93 pp., Philadelphia: W. B. Saundcrs Co., 1958.
329
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1966 American Association for Cancer Research.
Williamson, Kirby, Miller, Sass, Kramer, Seligman, and Wüten
6. Edsall, J. T., and Wyman, J. Biophysical Chemistry, Vol. 1,
pp. 550-90. New York: Academic Press, Inc., 1958.
7. Friedman, O. M., and Boger, E. Colorimetrie Estimation of
Nitrogen Mustards in Aqueous Media. Anal. Chem., 33:906-10,
1961.
8. Friedman, O. M., and Seligman, A. M. Preparation of Sec
ondary Amine Mustards with High Toxicity. J. Am. Chem.
Soc., 76:658-61, 1954.
9. Columbio, C. and Bergmann, M. Chemical Reactions of the
Nitrogen Mustard Gases. II. The Composition of Aged Un
buffered Solutions of Methyl-bis(/3-Chloroethyl)amine.
J.
Org. Chem., ÕÕ.-536-43,
1946.
10. Pinsent, B. R. W., Pearson, L., and Roughton, F. J. W. The
Kinetics of Combination of Carbon Dioxide and Ammonia.
Trans. Faraday Soc., 52:1594-98, 1956.
11. Rauen, H. M. Die Bildungsgeschwindigkeiten
von N-ßHalogenathylaziridiiien
aus Bis ((3-halogenathyl)-aminen
unter physiologischen
Bedingungen.
Arzneimittel-Forsch.,
Õ4:855~59,1904.
12. Ross, W. C. J. Biological Alkylating Agents. London: Butterworth & Co., Ltd., 1962.
330
13. Ruch, T. C., and Fulton, J. F. Medical Physiology and Bio
physics, Ed. 18, pp. 544-46. Philadelphia: W. B. Saunders
Co., 1900.
14. Ungar, J. Mercurimetric Determination of Chlorides. Chem.
Ind. (London), 453-54, 1954.
15. Warwick, G. P. The Mechanism of Action of Alkylating
Agents. Cancer Res., #3:1315-33, 1963.
16. Wheeler, G. P. Studies Related to Mechanisms of Resistance
to Biological Alkylating Agents. Ibid., #3:1334-49, 1963.
17. Williamson, C. E., Kirby, J. G., Miller, J. I., Sass, S., Kramer,
S. P., Seligman, A. M., and Witten, B. A New Biologic Re
action Observed With Secondary Nitrogen Mustards NorHN2 (NSC-10873) and Mannitol Mustard (NSC-9598). Cancer
Chemotherapy Rept., 4/:47-49, 1964.
18. Witten, B., Williamson, C. E., Sass, S., Miller, J. I., Best,
R., Wicks, G. E., Kramer, S. P., Weinberg, T., Solomon,
R. D., Goodman, L. E., and Seligman, A. M. Enzyme Alter
able Alkylating Agents. I. Synthesis, Chemical Properties, and
Toxicities of Sulfur Mustards Containing Enzyme-Susceptible
Amide Bonds. Cancer, Õ5.-1041-55,1962.
CANCER
RESEARCH
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1966 American Association for Cancer Research.
VOL. 26
Enzyme-alterable Alkylating Agents: IX. The Enzymatic
Transformation of Some Nitrogen Mustards in the Presence of
Carbon Dioxide: Implications in Respiration
Charles E. Williamson, James G. Kirby, Jacob I. Miller, et al.
Cancer Res 1966;26:323-330.
Updated version
E-mail alerts
Reprints and
Subscriptions
Permissions
Access the most recent version of this article at:
http://cancerres.aacrjournals.org/content/26/2_Part_1/323
Sign up to receive free email-alerts related to this article or journal.
To order reprints of this article or to subscribe to the journal, contact the AACR Publications
Department at [email protected].
To request permission to re-use all or part of this article, contact the AACR Publications
Department at [email protected].
Downloaded from cancerres.aacrjournals.org on June 18, 2017. © 1966 American Association for Cancer Research.