Biochemical Pharmacology, 1961, Vol. 7, pp 88-95. Pergamon Press Ltd., Printed in Great Britain.

A NEW AND RAPID COLORIMETRIC DETERMINATION

OF ACETYLCHOLINESTERASE ACTIVITY
GEORGE L. ELLMAN, K. DIANE COURTNEY, VALENTINO ANDRES, JR.
and ROBERT M. FEATHERSTONE
Department of Pharmacology, University of California Medical Center,
San Francisco, California
(Received 14 November 1960)
Abstract-A photometric method for determining acetylcholinesterase activity of tissue
extracts, homogenates, cell suspensions, etc., has been described. The enzyme activity
is measured by following the increase of yellow color produced from thiocholine when
it reacts with dithiobisnitrobenzoate ion. It is based on coupling of these reactions:
(enzyme)
acetylthiocholine -----+ thiocholine + acetate
thiocholine + dithiobisnitrobenzoate -+ yellow color
The latter reaction is rapid and the assay is sensitive (i.e. a 10 pl sample of blood is
adequate). The use of a recorder has been most helpful, but is not essential. The method
has been used to study the enzyme in human erythrocytes and homogenates of rat
brain, kidney, lungs, liver and muscle tissue. Kinetic constants determined by this
system for erythrocyte cholinesterase are presented. The data obtained with acetyl-
thiocholine as substrate are similar to those with acetylcholine.
INTRODUCTION
A FEW years ago Bonting and Featherstonel introduced a modification of the Hestrin
hydroxamic acid method2 suitable for the determination of cholinesterase levels in
small quantities of cells cultured in vitro. This modification was used successfully in
several studies involving the control of enzyme levels in cells by manipulating the
levels of substrates or closely related compounds present in the medium.3 Several
interesting areas of research were indicated by these studies. However, this modified
method of enzyme assay, although scaled down to the micro-level, had several dis-
advantages. Among these was the fact that only a terminal figure could be obtained
from the material in one tube of cultured cells. Thus, the time course of the reaction
could not be followed without resorting to separate experimental tubes for each time
interval desired. The method also had the disadvantage that the color measured *was
developed from the remainder of an added substrate-a procedure in which the possi-
bility of error is relatively great when the level of enzyme activity is small.
Consideration of the relative merits of various methods which might be useful in
studying the time-course of acetylcholinesterase activity in very small tissue samples
led us to combine a method reported by Koelle4 with a sulfhydryl reagent studied by
Ellman.5 This new method, which is presented here, is extremely sensitive and is
applicable to either small amounts of tissue or to low concentrations of enzyme. It
makes detailed kinetic studies of acetylcholinesterase activity possible. The progress of
the hydrolysis is followed by the measurement of a product of the reaction.
88
A new and rapid calorimetric determination of acetylcholinesterase activity
89
Acetylthiocholine is used as the substrate. This analog of the natural substrate has
been used most extensively by Koelle4 for histochemical localization. Other workerss,
have used the sulfur analog in the enzyme assay. Their work, in addition to data we
shall present, suggests that this compound is a satisfactory substitute for the natural
substrate, and differs much less than some of the synthetic substrates frequently used
as in assays of phosphatases, trypsin, chymotrypsin, pepsin, etc.
The principle of the method is the measurement of the rate of production of thio-
choline as acetylthiocholine is hydrolyzed. This is accomplished by the continuous
reaction of the thiol with 5 : Sdithiobis-2-nitrobenzoate ion (I)5
(enzyme)
H,O + (CH,),&CH,CH,SCOCH, h (CH&CH2CH2S- +
+ CH,COO- + 2H+
(CH&H$H$- + RSSR + (CH&&CH&H,SSR + RS-
(I) (II)
to produce the yellow anion of 5-thio-2-nitro-benzoic acid (II). The rate of color
production is measured at 412 rnp in a photometer. The reaction with the thiol has been
shown to be sufficiently rapid so as not to be rate limiting in the measurement of the
enzyme, and in the concentrations used does not inhibit the enzymic hydrolysis. By
recording the output of the photometer continuously, records of the complete assay can
be obtained (Fig. 1).
We considered it desirable to establish that this method yields results comparable
with other procedures. For this reason, the effects of inhibitors were examined; the
kinetic constants were calculated and compared with those obtained by other methods.
In addition, we were able to compare assays on blood samples by the ferric-
hydroxamate methods and the present one.
METHODS
The reaction rates were recorded with a Beckman DU spectrophotometer equipped
with a Beckman adapter and a Minneapolis-Honeywell recorder.
The general method used was to place buffer in the photocell and add concentrated
solutions of reagents by means of micropipettes. The mixture was stirred by continued
blowing through the pipettes while moving them around the bottom of the photometer
cells. In this way, reagents were added, mixed, and the cover of the cell compartment
replaced within 10-15 sec. Our photometer required 3040 set to become stabilized
to new light conditions. Thus, there was about 1 min when the readings were due to a
combination of factors (e.g. bubbles rising through the solutions, sulfhydryl material
in the enzyme, etc.) which were unrelated to the desired measurements. Subsequent
readings were strictly dependent on the absorption of the solution under considera-
tion and even rapid changes were followed by the recorder faithfully, as evidenced
by the reproducibility of time-transmission curves (Fig. 1).
90 GEORGE L. ELLMAN et. al.
Solutions
Buffer. Phosphate, 0.1 M, pH 8.0.
Substrate. Acetylthiocholine iodide, * O-075 M (21.67 mg/ml). This solution was
used successfully for lo-15 days if kept refrigerated.
Reagent. Dithiobisnitrobenzoic acid (DTNB) 0.01 M of the 5 : 5-dithiobis-2-
nitrobenzoic acid7 prepared as described previously,5 39.6 mg were dissolved in
10 ml pH 7.0 phosphate buffer (O-1 M) and 15 mg of sodium bicarbonate were added.
The reagent was made up in buffer of pH 7 in which it was more stable than in that of
pH 8.
Enzyme. Bovine erythrocyte cholinesterase (Nutritional Biochem. Corp., 20,000
units) was dissolved in 20 ml of 1% gelatin. This solution was diluted 1 : 200 with
water for use, yielding a solution of 5 units/ml.
General method
A typical run used : 3.0 ml pH 8.0 buffer, 20.0 ~1 substrate, 100.0 ~1 DTNB (reagent),
50.0 ~1 enzyme. The results of several runs are shown in Fig. 1. The blank for such a
run consists of buffer, substrate, and DTNB solutions. The absorbances: were read
from the strip charts and plotted on a rectangular graph paper, the best line drawn
through the points and the slope measured. In a run such as that described above, the
linear portion of the curve describing the hydrolysis was observed during the first
15-20 min of the reaction; the slope is the rate in absorbance units/min. At this pH
level, there is an appreciable non-enzymic hydrolysis of the substrate, and for long
runs it was necessary to correct for this. The rate of non-enzymic hydrolysis of
acetylthiocholine at 25 was 0.0016 absorbance units per min.
The procedures have been extended to micro-size. A run comparable to those in
Fig. 1 was made in a micro-cell (total solution volume was O-317 ml). The rate was
0*102/min, the same as that determined in the larger cuvettes.
Since the extinction coefficient of the yellow anion (II) is known5 the rates can be
converted to absolute units, viz. :
A absorbance/min
rate (moles/l. per min) = -
1.36 x lo4
In dealing with cell extracts or suspensions, a blank consisting of extract or sus-
pension, DTNB, and buffer may be required to correct for release of thiol material
from the cells and the absorbance of the other materials in the suspension.
Method for blood
A fairly stable suspension was formed from whole blood or washed human erythro-
cytes. Since the acetylcholinesterase is on the cell membrane, hemolysis was not
necessary. The assay of blood was carried out as follows:
(1) A suspension of the blood cells4 in phosphate buffer (pH 8.0, 0.1 M) was pre-
pared. The most practical dilution was 1 : 600 (e.g. 10~1 blood into 6 ml buffer).
(2) Exactly 3.0 ml of the suspension were pipetted into a cuvette.
* California Corporation for Biochemical Research, Los Angeles, California.
t This is now available from the Aldrich Chemical Co.. 2369 No. 29th. Milwaukee 10, Wisconsin.
$ Strip charts printed in absorbance units are available from Minneapolis-Honeywell Corporation,
Chart 5871.
4 Red cell counts were performed by the clinical laboratory.
F
I
G
.

1
.

P
h
o
t
o
g
r
a
p
h

o
f

s
t
r
i
p

c
h
a
r
t

r
e
c
o
r
d

o
f

t
w
o

i
d
e
n
t
i
c
a
l

a
s
s
a
y
s
.

A
t

t
h
e

a
r
r
o
w
,

p
h
y
s
o
s
t
i
g
m
i
n
e

s
a
l
i
c
y
l
a
t
e

(
f
i
n
a
l

c
o
n
c
e
n
t
r
a
t
i
o
n
,

3

x

I
O
-

M
)

w
a
s

a
d
d
e
d

t
o

a

t
h
i
r
d

r
e
p
l
i
c
a
t
e
.

f
a
c
i
n
g

p
a
g
e

9
0

A new and rapid calorimetric determination of acetylcholinesterase activity
91
(3) If only red cell cholinesterase was desired, 10 ~1 of O-1 per cent quinidine sulfate
was added to inhibit plasma esterase. 8 If total esterase was desired, this step was
omitted. See comment on quinidine in the discussion section.
(4) Of the DTNB reagent, 25 ~1 were added. The cuvette was placed in the photo-
meter.
(5) The slit of the photometer was adjusted so that the absorbance (at 412 mp) of
the suspension in the cuvette was zero.
(6) Of the substrate, 20 ~1 were added to this cuvette. Changes in absorbance at
412 rnp were recorded for at least 6 min.
(7) Calculations :
moles substrate hydrolyzed/min per R.B.C. = (4.41) (lo-14) RqC.
. . .
where 4.41 (lo-14) = factor for dilution, extinction coefficient, and changes in units;
zzz g!!!!!!!6 x -._L__ x
1
,
lo6 mm3/1. lo6 R.B.C.
A4 = change in absorbance/min;
R.B.C. = red cell count (in millions per mm3).
0 ther tissues
We have also assayed rat lung, liver, stomach, heart, muscle and brain by this
procedure. The steps we used were these:
(1) The tissue was homogenized (approximately 20 mg of tissue per ml of phosphate
buffer (pH 8.0, 0.1 M)) in a Potter-Elvehjem homogenizer. For muscular tissue,
considerable mincing was necessary before homogenizing.
(2) A O-4-ml aliquot of this homogenate was added to a cuvette containing 2.6 ml
of phosphate buffer (pH 8.0, O-1 M).
(3) Of the DTNB reagent, 100 ~1 were added to the photocell. The absorbance was
measured at 412 rnp; when this had stopped increasing, the photometer slit was opened
so that the absorbance was set to zero.
(4) Of the substrate, 20 ~1 were added. Changes in absorbance were recorded and
the change in absorbance per min. was calculated.
(5) The rates were calculated as follows:
AA 1
R = 1.36 (104) (4oo1312o)c,
= 5.74 (10-d) A+
0
where R = rate, in moles substrate hydrolyzed per min per g of tissue;
AA = change in absorbance per min;
Co = original concentration of tissue (mg/ml).
Reproducibility
RESULTS AND DISCUSSION
The results of repeated assays under various conditions are shown in Table 1.
The standard deviations are about 4 per cent of the mean rates, a variability that could
be accounted for as pipetting variation.
92 GEORGE L. ELLMAN et al.
The rate of hydrolysis of acetylthiocholine by bovine erythrocyte esterase as a
function of enzyme concentration is shown in Fig. 2. The unit of enzyme activity is
that described by Ammon,B and is the amount which hydrolysed 0.01 mg of acetyl-
choline bromide in 1 min in 3 ml of 04092 M acetylcholine bromide. The observed
rate was a linear function of enzyme concentration. The data were plotted in terms of
the final concentration of enzyme.
TABLE 1. REPRODUCIBILITY OF RATE MEASUREMENTS
Substrate ) Enzyme
(moles/l.) 1 (units)
5.0 x 10-4 0.033
5.0 x lo-* 0.083
2.5 x 10-4 , 0.083
Mean
rate7
owO2
0.1003
0.0842
I
Standard
deviation / Replicates
.-
0~0012 / 2
0.0038 9
0~0041 i 4
* Bovine erythrocyte esterase.
t Rate is given in change of absorbance/min. Other detailsaregivenunder Procedures.
0.15
-; O.IC
2
P
h
g 0.01
0.0 I
I-
i-
l-
T
/
/
.
L.- I
0.01 0.05 0.10
Enzyme concentration, units/ml
FIG. 2. Rate of thiocholine production as a function of enzyme concentration.
The rate of hydrolysis of acetylthiocholine by bovine erythrocyte cholinesterase as a
function of substrate concentration is shown in Fig. 3. The substrate was hydrolysed
at an appreciable rate without enzyme above 5 x 1O-3 M. The rate plotted for the
enzyme is thus the net rate after non-enzymic hydrolysis rate is subtracted from the
observed rate. As expected for the true enzyme, substrate inhibition was observed,
and the data in this experiment were analysed by a reaction sequence that includes this
factor. The method of Wright and SabinelO
was used to analyze these data. The results
of the analysis, as well as inhibition constants (also evaluated by the method of Wright
and Sabine), are shown in Table 2. The agreement with literature values is close enough
that estimates of enzyme rates determined with acetylthiocholine may be considered
good estimates of the rate for acetylcholine hydrolysis.
A new and rapid calorimetric determination of acetylcholinesterase activity 93
Twelve samples of human blood were assayed by our technique and the ferric
hydroxamate method.* The comparison is shown in Fig. 4. It is apparent that the
observed rates were directly proportional. Thus, the two assays were clearly measuring
the same activity. We used quinidine sulfate (2 x 1O-5 M, final concentration) to
inhibit the plasma esterase.* Using the above technique with samples obtained from
eighty-seven people, we obtained as the mean and standard deviation (1.08 f. O-16) x
lo-l5 moles substrate hydrolysed per min per red blood cell. Each assay took about
5 min. It is possible to determine at least fifty samples per day.
0.1 6
2 O. 2
2
9
0.06
p
0
= 0.04
0
_
IO- IO IO Id
Acetylthiocholine concentration. m/l
FIG. 3. Rate of thiocholine production as a function of substrate concentration; 0, enzymic
x , non-enzymic.
It should be noted that the concentration of quinidine used will not necessarily
inhibit the non-specific esterase of all species or of all tissues (see e.g.5). It may,
in fact, not be possible to inhibit all the non-specific esterase without inhibiting some of
the acetylcholinesterase. It may be necessary, in the study of a given tissue to determine
the appropriate concentration of a selective inhibitor (e.g. DFP), in order to use the
method for determination of specific acetylcholinesterase.
TABLE 2. KINETIC CONSTANTS OF ERYTHROCYTE CHOLINESTERASE
Constant
1 Acetylthiocholine i Acetylcholine 1 Ref.
Michaelis (KM)
Inhibition (KI)
Substrate
Decamethonium
Physostigmine
Quinidine
1.4 (10-4)
i 29 (lo-)
6.1 (1O-6)
2.0 (10-h)
1.5 (10-Z)
2.5 (10-b)
6.1 (1O-8)
9.6 (lo-3
10
10
12
13
10
The activity of various rat tissues on acetyl- and butyryl-thiocholine was measured
and results are presented in Table 3. These results were obtained in order to establish
the applicability of the method, and are not intended to represent a study of these
enzymes in these tissues. Since most of these tissues contain non-specific esterases
(possibly thiolesterases), the non-cholinesterase hydrolysis might best be determined
by incorporating the proper amount of physostigmine in an extra blank to be
subtracted.ll
94 GEORGE L. ELLMAN efnl.
The method described here is, in principle, similar to that of Gal and Roth. How-
ever, we feel, that our method has several advantages: (a) it is dependent on changes in
the visible region of the spectrum, so that unusual changes in the absorbance can be
checked imm~iately, e.g. appearance of turbidity, spills on the photocell windows,
etc. ; (b) since the extinction coefficient of the nitrothiobenzoate ion is 13,600, as
against 5140 for acetylthiocholine chloride, there is at least a 2.5fold potential
increase in sensitivity available; (c) measurements of the appearance of products are
usually more sensitive than disappearance of substrate; (d) homogenates of tissue do
not require any special handling, i.e. precipitation of protein before readings, etc.;
(e) the reagents required are commercially available. For these reasons, we believe that
our method is more convenient than that of Gal and Roth.
TABLE 3. AVERAGE RATES OF HYDROLYSIS OF ACETYL- AND BUTYRYL-THIOCHOLINE
BY RAT TISSUES
1 Rates (moles/l. per min x 10e6 per g of tissue)
Tissue
/
Acetyl
/
Butyryl
Lung
Liver
Muscle (thigh)
Kidney
Brain (whole)
1.63 0.54
1.07 0.65
1.82 0.12
0.28 0.21
1031 0.31
The rates given are average values obtained from tissues of four
Sprague-Dawley rats.
Rote, AA/min
FIG. 4. Comparison of rates of R.B.C. cholinesterase obtained from twelve human R.B.C. samples
when the present method and the Sabines method were used.
The hydroxamate method* determines the residual substrate after a period of
incubation. Hence, only one time period can be measured per analysis, and the
execution of kinetic studies with this method becomes difficult. In addition, like the
Gal-Roth technique, the disappearance of substrate is being measured. The acidimetric
A new and rapid coiorimetric determination of acetylcholinesterase activity 9.5
and gasometric methods, while capable of being used to study kinetics, are inherently
less sensitive, often requiring several hundred milligrams of tissue to be useful.
Ack~~~Zedgements-We gratefully a~knowIed~ support from the U.S. Public Health Service,
National Institutes of Health (Grant CY3861) and State of California Mental Hygiene Commission
(Grant 60-l-1302). We wish to thank Dr. J. C. Sabine for her valuable advice and assistance.
REFERENCES
1. S. L. BONTING and R. M. FEA~HERSTONE, Arch. B&hem. B~op~ys. 61, 89 (1956).
2. S. HESTRIN, J. Biol. Chem. 180, 249 (1949).
3. M. JONES, R. M. FEATHERSTONE and S. L. BONTING, J. Pharmacol. 116, 114 (1955).
4. G. B. KOELLE, J. Pharmacoi. 100, 158 (1951).
5. G. L. ELLMAN, Arch. B~ochem. Biophys. 82,70 (1959).
6. I . I. A. TABACHXIK, Biochim. Biophys. Acta 21,580 (1956).
7. E. M. GAL and E. ROTH, Clin. Chim. Acta 2,316 (1957).
8. J. C. SABINE, Blood 10, 1132 (1955).
9. R. G. AMMON, P&g. Arch. ges. Physiol. 233, 486 (1933).
10. C. I. WRIGHT and J. C. SABINE, J. PhurmucoZ. 93, 230 (1948).
11. G. B. KOELLE, J. Pharmacol. 114, 167 (1955).
12. F. BERGMANN, I. B. WILSON and D. NACHMANSOHN, Biochim. Biophys. Acta 6, 217 (1950).
13. D. NACHMAN~~HN and I. M. WILSON, Advanc. Enzymol. 12,259 (1951).