An Enzymatic Method for Glucose

Determination in Body Fluids

Eliot F. Beach and James J. Turner

ENZYMATIC REACTIONS, BECAUSE of their great specificity, are among

the most valuable analytic tools at the disposal of the chemist. The
method for glucose here described represents an adaptation of such
biologic reagents and is a distinct departure from sugar methods
now in common use. It is based upon a unique principle communi-
cated to us by Keston (1) prior to publication (2). Traditional
methods for glucose depend upon its reducing reactions and are gen-
erally based upon the classic work of Benedict (3), Folin and Wu (4),
and Hagedorn and Jensen (5). The conspicuous success of these
procedures and their resistance to fundamental revision indicates
that they will not soon be replaced. This fact does not, however,
minimize our constant need for new alternative procedures that
possess special advantages of their own.
The enzymatic method is based upon beta glucose oxidase or “nota-
tin,” discovered in 1928 by Muller (6) in the molds Aspergillus niger
and Penicihium glaucum. He showed that this ferment promoted the
oxidation of glucose by molecular oxygen to gluconic acid. Later
Franke and Lorenz (7) demonstrated that hydrogen peroxide is pro-
duced simultaneously in this reaction. The hydrogen peroxide formed
in this way is readily detected with an oxygen acceptor like ortho-
tolidine in the presence of peroxidase, an enzyme easily prepared
from horseradish root (8). The great specificity of notatin for glucose
confers a special advantage upon the enzymatic procedure not pos-
sessed by some reduction methods that may respond to some extent
to nonglucose reducing substances present in variable amount in
From the Biochemical Laboratory, Metropolitan Life Insurance Company, New York
10, N. Y.
Received for publication July 29, 1958.

462
Vol. 4, No. 6, 1958 ENZYMATIC METHOD FOR GLUCOSE DETERMINATION 463

urine and in blood filtrates. The enzymatic method will not respond
to any other physiologic constituent except glucose.
Until recently the few published accounts of this new system for
glucose determination came from sources concerned with its com-
mercial development (10, 11, 12) and were lacking, therefore, in de-
tail required for practical laboratory application. Experience in de-
velopment of the method in our own laboratory and successful appli-
cation to a large number of urine specimens led to the belief that a
presentation of the details of this analytical procedure would be of
value. More recently papers by Huggett and Nixon (14) and by
Middleton and Griffiths (15) have described fully the enzymatic
method for blood sugar. Development of the procedure in our own
hands has been somewhat different, inasmuch as it has been applied
both qualitatively and quantitatively to urine as well as to blood
filtrates. The practical handling of this system of glucose determina-
tion is the subject of the present report, together with a comparison
of the results obtained thereby with those yielded by other common
methods. The procedures should prove useful in any well-equipped
laboratory in which “true glucose” estimation is required.

METHODS
GLUCOSE QUALITATIVE TEST FOR URINE SCREENING

Reagents and Materials:

1. Filter paper squares, 1#{189}
inches (Whatman No. 1 chromatog-
raphy paper is suitable).
2. 0.025 per cent glucose solution in saturated benzoic acid. This
is used to check activity of the mixed reagent daily.
3. 2 per cent orthotolidine in 95 per cent alcohol.
4. Citrate-phosphate buffer at pH 5.6. Mix 58.0 ml. of 0.2M Na2
11P04 and 42.0 ml. of 0.1 M citric acid.
5. Glucose oxidase. In 100 cc. of pH 5.6 citrate-phosphate buffer
dissolve 2 Gm. of glucose oxidase.* Activity is preserved at least 4
weeks when held under refrigeration.
6. Peroxidase horseradish root extract. Peroxidase is available
commerciallyf but preparation of high potency extracts is a simple
*Obthinable from Takamine Laboratories, Clifton, N. J., or Sigma Chemical Company,
St. Louis 13, Mo.
tObtainable from Nutritional Biochemicals Corp. Cleveland 28, Ohio, or Worthington
Biochemical Corp. Freehold, N. J.
464 BEACH & TURNER Clinical Chemistry

matter and is the method of choice. Good quality horseradish roots

are procurable throughout the year in New York markets.
Grind 1,000 Gm. of fresh horseradish (about 3 large roots) in a
food chopper using the finest blade. Extract the chopped root with
1,200 ml. of water by mixing successive portions of water and root
in a Waring blendor. Separate the pulp from the extract by straining
through cheesecloth or in a press, after which the extract is cen-
trifuged to remove starch and other particles. Pour off the super-
natant carefully into a 2-liter Erlenmeyer flask, place in a boiling
water bath, and with constant stirrmg bring the contents of the flask
to 80#{176}.
Adjust the bath to 80#{176} by adding cold water and hold the
extract at this temperature for 10 minutes. Peroxidase is stable un-
der these conditions but many other enzymes are inactivated and
much of the protein of the extract is coagulated. Do not exceed 80#{176},
however, because at higher heat peroxidase also is destroyed. Re-
move the coagulated protein by centrifugation and to the clear, yel-
low extract add finely powdered ammonium sulfate at the rate of 65
Gm. per 100 ml. Stir to dissolve the salt completely and let the mix-
ture stand 10 to 15 minutes, during which the active peroxidase, which
has been salted out, rises to the surface of the dense liquid. Remove
as much as possible of the clear, underlying liquid by aspiration
without loss of the salted-out material. This procedure gives the
peroxidase in about a 50 ml. volume of ammonium sulfate solution.
Add to this, 150 ml. of water to dissolve most of the precipitated
protein. Place this solution in cellophane sausage casing (11/8 inch),
tie securely, and dialize against running water until free of sulfate
and glucose. This solution now 200 to 250 ml. in volume is centrifuged
to remove any precipitate. It may be held under refrigeration for at
least 5 weeks with little loss of potency.
7. The combined glucose reagent. Mix 50 ml. of the buffered glu-
cose oxidase solution with 50 ml. of horseradish root peroxidase ex-
tract and add 10 ml. of 2 per cent orthotolidine. If the solution turns
more than very faintly blue it indicates that the horseradish root
extract has not been dialized sufficiently to remove the glucose. Bub-
ble air gently through the mixed solution for about 1 hour and then
centrifuge to clear. Usually reagent prepared in this way will hold
its potency under refrigeration for at least 2 weeks. Its ability to
detect glucose should be checked frequently by testing it with 0.025
per cent glucose solution. If the reagent fails to detect this amount of
glucose it should be discarded.
Vol. 4, No. 6, 958 ENZYMATIC METHOD FOR GLUCOSE DETERMINATION 465

Procedure for Glucose Detection:

The method of glucose detection for urine screening follows. Place
1 drop of the combined glucose reagent in the center of a ifiter paper
square to give a wet reagent spot about #{189} inch in diameter. With a
dropper place a spot of the urine to be tested at the outer edge of the
reagent, taking care not to let the urine flow across the reagent spot
but only to pass into it by capillary attraction in the paper. The ap-
pearance of a blue color in the reagent spot within 4 minutes is an
indication that glucose is present. Such specimens are reserved for
the quantitative evaluation of glucose.
The speed and intensity of the color reaction are related to the
glucose concentration but
the relationship is disturbed by the inter-
ference of several factors. Most significant is the presence in urine of
inhibiting substances that occur in variable amount. Uric acid is the
principal inhibitor. Forthis reason the spot test is always more
sensitive to pure glucose solutions than it is to any urine sample.
When the test is done with pure glucose, a solid blue area is formed
in the reagent spot wherever glucose has penetrated it. With urines,
only a blue band of color is obtained on the edge of the urine spot as
it moves into the reagent area. The inhibitors, fortunately, travel
slower than the glucose and these tend to decolorize the reagent spot
behind the advancing glucose front. it is clear, therefore, why the
reagent spot must never be flooded with urine in the test; otherwise
inhibitor would flood the whole area and might be sufficient to obscure
the glucose test entirely.

QUANTITATIVE METHOD FOR URINE GLUCOSE

Reagents and Materials:

1. Decolorizing charcoal. Some charcoals contain materials that
give high and erratic glucose values and therefore charcoal should be
tested before use. We have found Norite A, technical grade, to be
satisfactory.
2. One per cent orthoanisidine in 95 per cent alcohol.
3. Lloyds reagent, a powdered hydrated aluminum silicate (manu-
factured by Hartman-Leddon Co., Philadelphia).
4. Phosphate-citrate buffer at pH 7.0. Mix 164.7 ml. of 0.2 M
disodium hydrogen phosphate (Na2HPO4) with 35.3 ml. of 0.1 M
citric acid.
5. Glucose oxidase-indicator solution. To 200 ml. of phosphate
466 BEACH & TURNER Clinkal Chems$ry

buffer at pH 7.0, add 250 mg. glucose oxidase. Stir a little to dissolve
the enzyme and then filter. Just before use add 1 per cent orthoani-
sidine at the rate of 2 drops for every 3 ml. of enzyme-buffer mixture.
6. Peroxidase solution. Use horseradish root extract as described
previously under the qualitative test. This solution must be glucose
free and can be tested by running a blank determination as described
below.
7. Standard glucose solution. 0.005 per cent glucose in saturated
benzoic acid.

Procedure:
The method here described is useful on urines containing 0.5 per
cent glucose or less. Urines under examination that exceed this level
will show too intense and rapid color development and must be
diluted further to fall into the correct range of sugar concentration.
Place 1 ml. of urine in a 125-ml. Erlenmeyer flask and add 19 ml. of
water. Add 50 mg. decolorizing charcoal and 1.5 Gm. Lloyds reagent.
It is convenient and sufficiently accurate to use metal measuring
spoons calibrated to deliver the right weight of each adsorbent. Mix
with the diluted urine specimen by shaking the flasks frequently, and
after 6 minutes remove the suspended material by filtration (What-
man No. 40 ifiter paper is suitable). Pipet 1 ml. of this filtrate into a
spectrophotometer tube and add 1 ml. of horseradish root peroxidase
extract. Prepare similarly a blank tube containing 1 ml. of water
and a standard tube containing 1 ml. of 0.005 per cent glucose solution
in lieu of urine filtrate. Because the reaction must be timed carefully
for reading later in the photoelectric colorimeter, the next addition
must be made at measured intervals. To each tube in the series
(standard, unknowns, and blank) add at the proper time 3 ml. of
glucose oxidase-indicator solution and mix thoroughly. All solutions
must be at room temperature. The indicator changes from colorless
to pink as peroxide is liberated from the glucose. The photoelectric
colorimeter is now set to measure at wave length 440 m and the
blank is set so that it gives 100 per cent transmittance.* Then exactly
20 minutes after adding glucose oxidase each tube is read in the in-
strument to determine optical density. To calculate the glucose con-
eThe blank should be tested against water set at 100 per cent transmittance, once near
the time reagent is added and then 30 or more minutes later. Any substantial tendency for
transmittance of the blank to decrease with time suggests that not all glucose has been
removed from the peroxidase solution by dialysis. Such extracts are unsatisfactory for use
without further purification.
Vol. 4, No. 6 958 ENZYMATIC METHOD FOR GLUCOSE DETERMINATION 467

tent of the unknown urine (before dilution with 19 ml. of water) USQ
the formula:
O.D. unknown
X 0.1 = per cent glucose in urine.
O.D. standard

QUANTITATIVE METHOD FOR BLOOD GLUCOSE

Reagents:
1. 0.3 N barium hydroxide solution.
2. 5 per cent zinc sulfate solution.
3. 1 per cent orthoanisidine in 95 per cent alcohol.
4. Peroxidase solution. Horseradish root extract as described for
urine methods.
5. Glucose oxidase-indicator solution. Identical with that de-
scribed for urine quantitative method.
6. Glucose standard. 0.005 per cent glucose in saturated benzoic
acid (5 mg. glucose per 100 ml).

Procedure:
The method may be applied to serum or whole blood only after re-
moval of protein. Tungstic acid filtrates cannot be used, but the
Somogyi zinc filtrates made by using barium hydroxide as the alkali
are free of metallic ions and therefore satisfactory. Blood samples
in which 10 mg. sodium fluoride and 1 mg. mercuric chloride per ml.
have been used as preservative and antocoagulant are suitable.
In a 15-ml. centrifuge tube place 4.5 ml. of water and mix with this
0.3 ml. of blood. Add 0.6 ml. of 0.3 N barium hydroxide solution, mix
and let stand a minute or two until it turns brown. Then add 0.6 ml.
of 5 per cent zinc sulfate with mixing. After 5 minutes, centrifuge and
use 1 ml. of the clear supernatant for analysis in the identical manner
described above for diluted urine except that the step of treating
with charcoal and Lloyds reagent is omitted. The following formula
is used to calculate the glucose concentration in the original blood:
O.D. unknown
x 100 = mg. per cent glucose in blood.
O.D. standard
A NOTE ON THE ASSESSMENT OF PEROXIDASE ACTIVITY IN HORSERADISH
ROOT EXTRACTS

Lyr (16) has described a useful method for determining peroxidase

activity that we have adapted for control of the potency of horse-
468 BEACH & TURNER Clinkal Chemistry

radish root extracts. The procedure is based on the ability of

peroxidase to promote oxidation of ascorbic acid by hydrogen perox-
ide. Ascorbic acid is oxidized preferentially by this system and when
it is completely converted to dehydroascorbic acid the peroxide-
peroxidase system will turn benzidine indicator blue, thus marking
the end point of the ascorbic acid oxidation. The speed of the reaction
is related to peroxidase activity. The adapted procedure of Lyr fol-
lows.
All reagents are brought to 25#{176} and held in a water bath at this tem-
perature; the reaction is also carried out in this bath. In a test tube
place 1 ml. of the extract to be tested and add 1 ml. of a 0.02 per cent
beuzidine solution in 0.1 N acetate buffer pH 5.3 (10.7 Gm. sodium
acetate .31120 plus 2.7 ml. glacial acetic acid made up to 100 ml. with
water). To this mixture add 1 ml. of freshly prepared ascorbic acid
solution containing 50 mg. per 100 ml. Finally add, with immediate
mixing, 1 ml. of 0.3 per cent solution of hydrogen peroxide. At the
moment this is added timing is started with a stop watch. The mix-
ture is placed in the bath and gently agitated by bubbling a slow
stream of air through it. The solution is observed to obtain the exact
time of appearance of the abrupt change from colorless to blue.
Horseradish root extracts suitable for the glucose procedures give
the above described color reaction in from 1 to 2 minutes. Where ex-
tracts are more potent than this they may be used more economically
for glucose detection by further diluting them. Extracts of usable
potency have always been obtained by the method described above.

RESULTS AND DISCUSSION

EVALUATION OF THE QUALITATIVE TEST

The validity of the enzymatic spot test for use in screening urine
samples for glucose has been assessed by comparison with 2 other
methods.
The first, used for many years by life insurance laboratories, was
devised by Benedict (17). Based upon the ability of glucose in alka-
line solution to reduce picrate to orange-colored picramate, it is non-
specific and responds to numerous other reducing substances that are
normally present in urine in variable amounts. The color due to
creatinine is suppressed during the test by adding a few drops of
acetone. The test provides for comparison against 5 permanent
visual color standard tubes ranging from 0.1 to 0.5 per cent glucose.
Higher glucose levels are assessed by dilution to proper range. Long
Vol. 4, No. 6. 1958 ENZYMATIC METHOD FOR GLUCOSE DETERMINATION 469

experience with the procedure has proven that when the 0.3 per cent
level is indicated, “true glucose” is extremely likely to be present in
the urine. However, at this or even higher readings glucose may be
absent. Comparison of this procedure with the enzymatic test on a
selected group of 1,600 urines is presented in Table 1. These urines
were chosen because they fell in the “doubtful” range of the picric
acid method, where reduction is substantial but the presence of true
glucose is uncertain. The results show that as the picrate reduction
level rises the percentage of specimens responding to the enzymatic
test rises and the intensity of those giving a positive reaction also in-
creases. At the 0.25 per cent level indicated by the picrate method
only 14.4 per cent of the samples react to the enzyme test, but at the
0.50 per cent picrate level 84.0 per cent are enzyme positive. This dis-
tribution is consistent with the fact that the enzymatic test is specific
and precise.
The second method used for comparison with the enzyme spot test
was the qualitative copper reduction test of Benedict (18). Table 2
presents a summary of findings in 500 random urine specimens.
Agreement between the 2 tests is very good. In a few instances the
enzymatic reagent gave a slight positive reaction with the Benedict

Table 1. COMPARISON OF’ ENZYMATIC Spor TEST WITH PIcarE REDUCTION METHOD FOR
GLUCOSE DETECTION

Pereentag.
Glucose by Reaction to Enzyme Test Total
Picrate Method Negative Slight Moderate Marked Cases
(Percent age) (Number)

0.25 85.6 7.2 5.9 1.3 1,015

0.30 65.0 13.7 15.6 5.7 263
0.35 49.6 14.9 22.4 13.1 107
0.40 39.7 12.6 29.2 18.5 103
0.45 43.5 4.7 15.3 36.5 85
0.50 16.0 2.5 21.0 60.5 81

Table 2. COMPARISON 01’ ENZYMATIC Sor TEST WITH BsmnICTs COPPER REDUCTION TEST
FOR GLUCOSE DETECTION

Reaction Reaction to Enzyme Test

Benediets (Number of 0 ases) Total
Reagent Negative Slight Definite Oases

Negative 399 9 0 408

Slight 22 9 15 46
Definite 6 9 60 75
470 BEACH & TURNER Clinical Chemistry

reagent negative, suggesting that the enzyme test is of even greater

sensitivity than the copper reagent. In a few instances a positive
Benedict test and a negative enzyme test can be explained on the
basis that the Benedict test is not a specific one and may respond to
reducers other than glucose.
The possibility existed, however, that the enzymatic test might fail
in certain urines that might contain enough inhibitor to be capable of
preventing entirely the reaction with glucose. To examine this point
more fully we have selected 53 urine specimens showing from 0.35
to 1.0 per cent glucose by the picrate test but which were negative to
the enzyme spot test. These urines were reexamined by the picrate
test after exposure to yeast to determine whether the picrate-reduc-
ing material was fermentable. Furthermore the urines were used to
dilute a glucose standard to determine how small an amount of added
true glucose would be still detectable by the enzyme test. The results
of these examinations are presented in Table 3.
The study shows that picrate and enzyme reactions disagree in the
lower range of values where the picrate method is notably unreliable.
Fermentation tests showed that 48 of the 53 urines when treated with
yeast failed to lose any of the picric acid-reducing power. But it
seems unlikely that even in the 5 cases where a decrease occurred with
yeast that the substance in question was really glucose. Thus in the
dilution experiment none of the urines showed inhibition that would
mask as little as 0.03 per cent added glucose and in most instances the
sensitivity was greater than this. It seems unlikely, therefore, that

Table 3. FERMENTATION AND SENsITIVITY S!rutIEs ON TTRINES PoSITIvE TO PICRATE BUT

NEGATIvE TO ENZYME TESTS

Picrate Reaction
Number % Glucose by Decreased Lntt of Sensitivity to Enzyme Test
Oases Picrate Test by Fermentation .OOS..014%* .015-.O3O%*
(Nmnber of Cases)

17 0.35 1 8 9
19 0.40 1 8 11
8 0.45 1 1 7
6 0.50 1 2 4
1 0.70 0 0 1
1 0.80 1 0 1
1 1.25 0 1 0
53 5 20 33

* Percentage of glucose added to the urine that is just detectible by the enzymatic reagent.
Vol. 4. No. 6, 1958 ENZYMATIC METHOD FOR GLUCOSE DETERMINATION 471

the enzymatic test as here described would miss detecting any signifi-
cant level of urinary glucose.
THE QUANTITATIVE METHOD FOR URINE GLUCOSE
The quantitative procedure was studied in regard to the relation-
ship between glucose concentration and optical density measured by
spectrophotometer. Figure 1 shows that color intensity obtained for
various times of reaction up to 60 minutes falls on a series of straight
lines when plotted against glucose concentration.
As will be seen, the color intensity changes with time as the en-
zymatic reaction proceeds. It may be desirable to stop the reaction
and stabilize the color at a specified time. This can be done by adding
to each tube, at the correct moment, 3 drops of 6 N hydrochloric acid.
This shifts the pH to an unfavorable one, stops the reaction, and at
the same time changes the color to much yellower, due to the fact that
the colored product from orthoanisidine is itself an acid base indica-
tor. Thus the optical density at 440 m is decreased by acidification.
This is ifiustrated in Fig. 2, which shows, however, that Beer’s Law is
still obeyed and demonstrates that the acidified solutions are valid for
determining glucose concentration.

20 30 40 30 6070
MO Glucos.
Fig. 1. Relation of color intensity to glucose concentration at various reaction times.
Carried out at room temperature, 22.5#{176}.
Spectrophotometer readings at 440 nit.
472 BEACH & TURNER Clinical Chemistry

30 40 50 60
g Glucos.
Fig. 2. Relation of color intensity to glucose concentration at 60 minute reaction time.
Upper curve at pH 7.0; Lower curve, 3 drops 6 N HCI added to terminate reaction at 60
minutes. Spectrophotometer readings at 440 n1c.

The most critical problem in applying this method to urine is in

the removal of inhibiting substances prior to the enzyme reaction.
Most important to remove is uric acid and for this purpose activated
charcoal is effective. However, we find as did Folin (19) that char-
coal also effectively removes glucose and it may not be used, as recom-
mended by Hugget and Nixon (14), unless the amount be sharply re-
stricted. We have found after many tests that uric acid and other in-
hibitors are best removed by a combination of Lloyds reagent and the
maximum permissible amount of charcoal. These are more useful in
combination than either alone. Table 4 demonstrates that the glucose-
absorbing power of charcoal is appreciable and that only restricted
use is permissible if valid determinations are to result.
Finally, before adopting the combination of activated charcoal and
Lloyds reagent, a test was made of these combined adsorbents to
determine the recovery of glucose added in standard amounts to
glucose-free urines. A typical recovery test is presented in Table 5.
THE QUANTITATIVE METHOD FOR BLOOD GLUCOSE
For an evaluation of the enzymatic method applied to blood sugar a
comparison has been made of values obtained in the same blood with
Vol. 4 No. 6. 195$ ENZYMATIC METHOD FOR GLUCOSE DETERMINATION 473

Table 4. RECovERY OF GLUCOSE As’rza TREATMENT WITH VARYING AMOUNTS OP CHARCOAL

Amount Charcoal Glucose* Found

(mg.) (mg.) Percentage Recorered

None .050 100

25 .050 100
50 .048 96
100 .039 78
200 .032 64
400 .022 44
500 .018 36

*0.050% Glucose was the known concentration of the solutions used for these recovery
experiments.

Table 5. RECOVERY BY THE QUANTITATIVE OXIDASE METHOD OF GLUCOSE ADDED IN VARYING

AMOUNTS ‘10 GLUCOSE-FREE URINE

Known Glucose Concentration Found

Concentration* (Percentage)
(Percentage) Urine A Urine B Urine C

0.010 0.011 0.006 0.012

0.025 0.030 0.019 0.024
0.050 0.056 0.044 0.058
0.10 0.10 0.093 0.10
0.25 0.24 0.26 0.24
0.50 0.54 0.50 0.51

*Glucose was added to the urine to make 1 per cent concentration; this was diluted further
with the glucose-free to yield the series.

2 other commonly used methods; the ferricyanide reduction method

of Folin and Malmros (20) and the Nelson (13) modification of
Somogyi method. They were applied to blood preserved with sodium
fluoride and mercuric chloride for several days after the samples
were drawn. The results are given in Table 6. It appears that the
enzymatic method, which reacts only with glucose, is generally in
good agreement with the Nelson values, and this is consistent with the
fact that alkaline zinc blood filtrates used in the Nelson method con-
tain no reducing substances other than glucose. Higher values are
consistently given by tungstic acid filtrates with the less specific Folin
and Malmros method in all bloods in which the sugar was below 150
mg. per 100 ml. In the bloods with higher sugar concentrations the
Folin method has given lower values than the other methods. This is
due presumably to the fact that the higher intensities of Prussian blue
do not seem to follow Beer’s law. The study demonstrates the reli-
474 BEACH & TURNER Clinical Chemistry

Table 6. BLOOD SUGAR BY THE ENZYMATIC METHOD COMPARED WITH VALUES BY OTHER
PRoCEDURES

Specimen Blood Sugar (mg./100 ml.)

Number Enzymatic Nelson Folin and Malmros

1 47 50 62
2 63 60 74
3 64 59 71
4 64 62 74
5 68 62 77
6 68 69 83
7 72 70 71
8 76 69 82
9 83 86
10 85 81 91
11 86 87 94
12 89 80 94
13 92 80 100
14 97 94 100
15 100 87 112
16 100 94 100
17 101 100 115
18 107 106 118
19 113 129
20 118 123 118
21 128 123 120
22 130 143
23 132 119 125
24 137 141 158
25 165 150 153
26 169 147 157
27 182 187 188
28 207 175 197
29 232 225 221
30 239 214
31 242 242 229
3! 261 236
33 266 260 229
34 285 274 250
35 460 466 400

ability of the enzymatic method for use in routine laboratory deter-

mination of blood glucose.
SUMMARY
1. Details are given for the qualitative detection of glucose in
urine and for its quantitative determination in urine and blood ifi-
trates by an enzymatic procedure.
2. The method, which is highly specific, is based upon the release
Vol. 4, No. 6, 958 ENZYMATIC METHOD FOR GLUCOSE DETERMINATION 475

of hydrogen peroxide from glucose during its oxidation by glucose

oxidase. The peroxide is detected by a suitable indicator in the pres-
ence of horseradish root peroxidase.
3. The results yielded by the new procedures have been compared
with those obtained by other accepted methods.
REFERENCES
1. Keston, A. S., Personal communication, March, 1954.
2. Keston, A. 5., Ab8tr. 12.9th Meeting Am. Chem. Soc., 1956, p. 31 C.
3. Benedict, 5. H., J. Biol. Chem. 9, 57 (1910.11).
4. Folin, 0., and Wu, H., J. Biol. Cheat. 41, 367 (1920).
5. Hagedorn, H. C., and Jensen, B. N., Biochem. Z. 135, 46 (1922).
6. Muller, D., Biochem. Z. 199, 136 (1928).
7. Prazike, W., ard Lorenz, F., Liebig’s Ann. 532, 1 (1937).
8. Elliott, K. A. C., Biochem. .1. 26, 1281 (1932).
9. Keilin, D., and Hartree, E. F., Biochem. J. 42, 221 (1948).
10. Corner, J. P., Anal. Cheat. 28, 1748 (1956).
11. Teller, J. D., Abstr. 130th Meeting Am. C’hem. Soc., 1956, p. 69 C.
12. Free, A. H., Adams, E. C., Kercher, M. L., Free, H. M., and Cook, M. H., COn. Cheat.
3, 163 (1957).
13. Nelson, N., J. Biol. Chem. 153, 375 (1944).
14. Huggett, A. St. G., and Nixon, D. A., Lancet 2, 368 (1957).
15. Middleton, J. E., and Griffiths, w. J., Br’it. Med. J. 2, 1525 (1957).
16. Lyr, H., Biochem. Z. 329, 91 (1957).
17. Benedict, S. R., Proceedings of the Associotion of Life Insurance Medical Directors of
America 10, 115 (1924).
18. Benedict, S. B., J. Bkl. Chern. 5, 485 (1908).
19. Polin, 0., and Berglund, H., 1. Biol. Chem. 51, 209 (1922).
20. Folin. 0., and Malmros, H., J. Biol. Cheat. 83, 115 (1929).