Nature of Materials in Serum That Interfere inthe Glucose Oxidase by gdf57j


									 CLIN.CHEM.21/1, 119-124 (1975)

 Nature of Materials in Serum That Interfere in the
 Glucose Oxidase-Peroxidase--o- Dianisidine Method
 for Glucose, and Their Mode of Action

 Walter J. Blaedel and James M. Uhi

 Separation           of blood serum on Sephadex                    G- 100 reveals         tomated, have been reported. About 20 references to
 three fractions that interfere with the glucose oxidase-                                  methods reported for clinical use appear in a paper
 peroxidase method for serum glucose when o-dianisi-                                       by Miskiewicz      et al. (3).
dine is used as the chromogen. A low-molecular-weight                                         Attempts     to apply this method directly to plasma
fraction containing primarily uric acid, a fraction con-
                                                                                           or serum samples have not given highly precise re-
taining protein with a molecular weight of about 40 000,
                                                                                           sults, owing to interferences      that are present in vary-
and a fraction of even higher molecular        weight (-.‘
500 000) each interfered with glucose recovery when                                        ing amounts. Plasma has been shown to contain in-
glucose was measured by this procedure. The uric acid                                      hibitors of glucose oxidase methods (4, 5), which are
fraction and the isolated 40 000 molecular weight frac-                                   generally removed by protein precipitation               (4, 6) or
tion interfere by competing with o-dianisidine for hydro-                                 extreme dilution       of serum (5, 7) in manual methods.
gen peroxide in the peroxidase-catalyzed color-forma-                                      In automated       methods, serum is generally          dialyzed
tion step. The high-molecular-weight fraction not only in-                                against a buffer solution to remove the glucose from
terferes with the peroxidase reaction, but also with the                                   the protein serum matrix (8-10), sq that interfering
glucose oxidase reaction itself. These agents cause                                       material     of high-molecular-weight        is excluded.      Be-
values to be low by as much as 20% in the manual de-                                      cause automated         methods are most frequently           used
termination of glucose in normal serum if thejr interfer-
                                                                                          for routine analysis, the interferences         most often dis-
ence is not recognized.
                                                                                          cussed are those caused by substances of low molecu-
AddItional           Keyphrases:        interference by uric acid and high-               lar weight, such as uric acid, ascorbic acid, and biliru-
 molecular-weight           materials            interfering material in urine
                                             #{149}                                        bin. A recent study (3) showed that uric acid was the
                                                                                          only such low-molecular-weight            substance having a
    Keston (1) first conceived an enzymatic         method                                significant    effect on determinations      of glucose in nor-
for determining   glucose in biological fluids, in which                                  mal or slightly above-normal        physiological     concentra-
the glucose oxidase1 reaction is coupled to the peroxi-                                   tion. Interference       by uric acid has been reported by
dase reaction in the presence of a chromogen.        Use of                               many workers and possible mechanisms                 for it have
o- dianisidine  in a quantitative     procedure    was re-                                been proposed (8, 11, 12).
ported by Teller (2) shortly thereafter.       Since then,                                    In early work on an electrochemical         method for tol-
many glucose oxidase methods, both manual and au-                                         lowing the glucose oxidase reaction (13), a high-mo-
                                                                                          lecular-weight      serum fraction     separated with a col-
   Department          of Chemistry,       University     of Wisconsin,      Madison,
                                                                                          umn of Sephadex G-50 was found to inhibit glucose
Wis.    53706.                                                                            oxidase. Separation        of serum by use of Sephade,          G-
   I   Terminology       used:   glucose   oxidase,     l-D-glucose:oxygen       oxido-   100 has revealed two separate high-molecular-weight
reductase,       EC; peroxidase,        donor:hydrogen      peroxide     oxido-
                                                                                          fractions    and a low-molecular-weight          fraction,    each
reductase,       EC;     uricase,    urate:oxygen     oxidoreductase,       EC; and      o- dianisidine,    3,3’-djmethoxybenzidine.                             of which interfered        with the glucose oxidase-peroxi-
   Received       July 25, 1974; accepted        Oct. 21, 1974.                           dase method. Preparations          of each of the fractions

                                                                                                         CLINICAL   CHEMISTRY,    Vol. 21. No. 1, 1975   119
 were tested for their effects on each step of the meth-               was prepared by two successive 250-fold dilutions   of
 od. On the basis of the results obtained from these                   a 30% solution of hydrogen peroxide. All of the above
 tests, a mode of action of each of the interferences   is             chemicals    were reagent grade.
 proposed.                                                                Samples of pooled serum and “Q-Pak Automated
                                                                       Chemistry Control Serum” (Hyland Div. Travenol
 Materials and Methods                                                 Laboratories,    Inc., Costa Mesa, Calif. 92626) were
                                                                       used in the initial chromatographic  separations. The
                                                                       Hyland control serum was used in the preparative
   Visible absorbances at 460 nm were measured with                    separation of the two high-molecular-weight                   frac-
 a Spectronic 88 spectrophotometer (Bausch & Lomb,                     tions, because we found it contained          greater  concen-
 Rochester, N. Y. 14602).                                              trations    of each interference      than did the pooled
   Ultraviolet absorbance studies of uric acid were                    serum.
 performed      with a Cary 14 double-beam                recording        For interference   studies, solutions of the two high-
 spectrophotometer         (Applied     Physics Corp., Pasade-         molecular-weight     fractions were used directly as col-
 na, Calif.).                                                          lected    from    the preparative       separation    outlined
    The oxygen electrode used was a Clark-type                  oxy-   below.
gen sensor, consisting        of a platinum        wire (20 am in
diameter),     heat-sealed     into a glass tube. The end of           Procedures
the glass tube was ground off and polished, exposing
                                                                           Serum was separated by gel-chromatography                    on a
a 20-am cross-section        of the wire, which was covered
                                                                        27.5 cm X 1.7 cm column of Sephadex G-100 (Phar-
with a polypropylene          membrane        (0.001 inch thick).
                                                                        macia Fine Chemicals, Inc., Piscataway,              N. J. 08854).
The internal        reference     electrode     was silver-silver
                                                                       To prove the presence of the various                    interfering
chloride.     The platinum        microelectrode       was held at
                                                                       agents, 3-ml samples of serum were separated                      and
-0.5 V with respect to the reference electrode. Cur-
                                                                       eluted at about 1 ml/min with phosphate buffer (0.1
rents were measured with a Model 414 S picoammet-
                                                                       mol/liter,   pH 6.0). After the first 20 ml was eluted,
er (Keithley        Instruments,        Inc., Cleveland,       Ohio
                                                                       5.0-ml fractions were collected. Each fraction was di-
44139) and recorded on an Omniscribe                chart recorder
                                                                        vided into two 2.0-ml aliquots. To each aliquot, 2.0
(Houston Instruments,          Bellaire, Tex. 77401).
                                                                       ml of peroxidase-o-       dianisidine     reagent solution was
    Rate measurements          were carried out in a 4.5-mi
                                                                       added. To one of the aliquots, 0.1 ml of standard glu-
cylindrical    cavity in a Plexiglas block equipped with
                                                                       cose was added, to the other 0.1 ml of the phosphate
a small magnetic stirrer. Solutions              were added and
                                                                       buffer. To initiate      the reaction, we added 0.1 ml of
withdrawn      by means of hypodermic            syringes through
                                                                       glucose oxidase solution         (5 U/ml). The reaction was
ports, which could be sealed to prevent access of air
                                                                       allowed to proceed for 5.0 mm at 25 #{176}C, was thenand
to the reaction cavity.
                                                                       quenched by adding 0.1 ml of KOH (400 g/liter). Ab-
                                                                       sorbances were read within 10 mm at 460 nm vs. a re-
Reagents                                                               agent blank.
     Glucose oxidase solutions of two different        concen-             We calculated glucose recovery in each fraction by
 trations (70 U/mI and 5 U/mI) were prepared from a                    taking the difference between the absorbances devel-
powder having a glucose oxidase (EC activity                  oped in the two aliquots and dividing by the absorb-
of 110 U/mg (Worthington         Biochemical      Corp., Free-         ance developed in a reaction mixture containing                  only
hold, N. J. 07728).                                                    0.1 ml of glucose standard solution and the enzyme
    A reagent solution containing,      per liter, 100 mg of           reagents.
o- dianisidine   dihydrochloride    and 3000 U of horsera-                 We made preparative         separations     of the two high-
dish peroxidase     (EC; Worthington)        was pre-         molecular-weight      fractions by adding 6.0-mi samples
pared in 0.1 mol/liter      phosphate    buffer (12.814 g of           of serum to the Sephadex G-100 column. The high-
KH2PO4 and 1.02 g of K2HPO4 in 1 liter of solution,                    molecular-weight      fraction was collected between 20.5
adjusted to pH 6.0 with concentrated         KOH).                     and 27.5 ml of eluent, which corresponded                  approxi-
    A 50 U/liter solution of uricase (urate oxidase; EC                mately to a 1.25-fold dilution          of the interfering       pro- was prepared by dissolving 0.0197 g of crude                 tein as compared        to its concentration         in the blood
uricase (17 U/g; Sigma Chemical Co., St. Louis, Mo.                    serum. The 40 000 molecular-weight              fraction was col-
63178) in 5 ml of 0.1 mol/liter     borate buffer (0.6184 g.           lected between 30.0 and 44.0 ml, which approximate-
of H3B03 in 100 ml of solution, adjusted to pH 8.5                     ly corresponded     to a 2.5-fold dilution        as compared to
with concentrated      KOH).                                           its concentration    in the blood serum. A clean separa-
    Glucose standard solutions       (100 mg/100 ml) were              tion was obtained between the two fractions                  by dis-
prepared by dissolving 0.1000 g of anhydrous glucose                   carding some of the eluate between the two fractions.
in 100 ml of water. Two uric acid solutions,            7.0 mg/            The low-molecular-weight           fraction    was prepared
 100 ml and 4.2 mg/100 ml, were prepared by dissolv-                   by dialyzing    3.0 ml of serum vs. 3.0 ml of the phos-
ing 0.0070 g and 0.0042 g of the solid, respectively,        in        phate buffer for 2 h, by using a piece of Spectrapor
100 ml of warmed water. A hydrogen peroxide solu-                      Type 2 dialysis membrane           (Spectrum      Medical Indus-
tion having a concentration       of about 0.2 mmol/liter              tries, Los Angeles, Calif. 90054). Presence of uric acid

120   CLINICAL   CHEMISTRY.    Vol. 21, No. 1, 1975
in the dialysate      was checked        by measuring        its ultravi-
olet absorbance       at 290 nm.                                                       Table 1. Glucose Recovery Experiment’
                                                                                                                     Eluent                          Glucose
   We determined           the effect of interferences             on the           Fraction   no.                volume, ml                       recovery, %
overall glucose    oxidase-peroxidase          system by react-                           1                         20-25                               69
ing a series of solutions containing          a fixed amount of                           2                         25-30                               74
glucose and various amounts of each interfering              frac-                        3                         30-35                               67
tion, with all other variables held constant. Each so-                                    4                         35-40                               65
lution of the series consisted of 1.5 ml of chromogen-                                    5                         40-45                               85
peroxidase    reagent solution, 0.2 ml of standard glu-                                   6                         45-50                              100
cose solution,     interfering      solution,     and sufficient                          7                         50-55                               98
phosphate    buffer to make 4.0 ml (3.0 ml for the uric                                  8                          55-60                              104
                                                                                         9                          60-65                              100
acid studies). The reactions were carried out and re-                                                                                                   57
                                                                                        10                          65-70
coveries calculated as described above.
                                                                                        11                          70-75                               40
    The effect of an interference           on the rate of the glu-                     12                          75-80                               26
cose oxidase reaction alone, independent                    of the cou-                 13                          80-85                               77
pled peroxidasereaction, wasmeasured by noting the
rate of decrease in dissolved               oxygen concentration
                                                                                  ,,3.0 ml of pooled
                                                                                G-100, and fractions
                                                                                                                    was chromatographed
                                                                                                                  used   in glucose     recovery
                                                                                                                                                      on Sephadex
with a Clark-type         oxygen electrode. For a rate mea-
surement,      0.5 ml of standard glucose solution and a
measured quantity           of the interfering          fraction       to be
tested were added to the reaction chamber (see sec-
tion on Apparatus),           which was then filled with the                       Additional          chromatographic                studies       with     Sepha-
phosphate buffer to give a total volume of 4.5 ml. To                           dex G-100 and Sephadex G-200 showed that frac-
start the reaction, we introduced              glucose oxidase (0.2             tions 1-5 (Table 1) contain at least two different  in-
ml of 70 U/ml solution)             into the reaction chamber                   terfering substances. The molecular   weight of inter-
with a syringe, and followed the fall-off                     in current        ference in fractions  3-5 was 40 000 to 45 000, as de-
caused by oxygen consumption.                Initial rates were cal-            termined by the ratio of elution volume to void vol-
culated from the slope of the current-time                      curve be-       ume in several chromatographic      experiments.   The
tween 0 and 3 mm, divided                 by the initial          current,      molecular   weight of the cloudy fractions     is much
which gave the fraction of the oxygen originally                        pres-   higher, greater than 500 000. We made no further at-
ent that was consumed per minute. This procedure                                tempts to identify these two materials more specifi-
corrected for run-to-run          variations      in electrode        sensi-    cally, although they presumably          are proteins.
tivity, and for oxygen originally            present.                              The ultraviolet      absorption    spectrum      of a dialyzed
    The effect of an interference                on the peroxidase              serum sample containing            the low-molecular-weight
step of the method alone, independent                    of the glucose         interfering   material was very similar to that of a 0.1
oxidation     step, was evaluated           by measuring           the ab-      mmol/liter     solution     of uric acid in the phosphate
sorbance developed at 460 nm by 0.5 ml of hydrogen                              buffer. In addition,      treatment    of another serum sam-
peroxide     solution     when combined            with peroxidase-             ple with uricase before dialysis removed ultraviolet
o- dianisidine     solution in the presence of the interfer-                    absorbance     in the dialysate entirely.        Also, after un-
ence. Concentration          of interfering      material and chro-             case treatment       of this dialysate,    interference      in the
mogen concentration            were independently               varied in       glucose determination         was decreased to about a tenth
separate     experiments,         the total         reaction       volume       of that observed for serum dialysate not treated with
being adjusted to 4.0 ml with the phosphate buffer. It                          uricase. Because uric acid evidently            accounts for al-
was not necessary to quench these reactions                              with   most all of the low-molecular-weight            interference    ob-
KOH because, for the small quantities                      of H202 in-          served, we studied the nature of its interference                by
volved, the reaction proceeded to completion                         in less    using solutions of uric acid itself rather than dialys-
than a minute.                                                                  ates.
                                                                                    Urine, when subjected to the same chromatograph-
Results and Discussion                                                          ic procedure,     gave early fractions       (corresponding       to
                                                                                high-molecular-weight          substances)    that did not in-
Isolation of Interferences                                                      terfere at all with glucose recovery.              On the other
   Table 1 shows the glucose recovery data for the                              hand, fractions corresponding          to 10-13 in Table 1 in-
fractions collected in a Sephadex G-100 separation of                           terfered severely and caused glucose recovery values
pooled blood serum. The low recoveries in fractions                             to be only about 10% of the true figure, an effect we
1-5 and 10-13 indicate the presence of two groups of                            ascribed to high concentrations            of uric acid in the
interferences.    Fractions     1 and 2 were cloudy and un-                     urine.
colored, distinctly     different   from fractions 3-5, which
were yellow. Glucose was eluted in fractions                8-11,               Effects of Interferences                 on the Overall Method
overlapping    the low-molecular-weight           interferences                    Figure       1 shows the effect of the presence of in-
in fractions 10-13.                                                             creasing       amounts  of each interfering  substance on

                                                                                                     CLINICAL     CHEMISTRY,          Vol. 21, No. 1, 1975           121
                                                                                                        GLUCOSE OXOASE
                                                                                     GLUCOSC+ 02                                  C   GlUCONATE +

      U                                                                  H202   +0-DIANISIDINE                            t   2 I-lO+OXlDlZED   CHROMOPHORE
      uJ                                                                   #{174}                           #{174}
                                                                         Fig. 2. Possible sites of interference in the glucose oxidase-
      u-i                                                                peroxidase-o-dianisidine            system
                                                                         SItes of interference   are Indicated   by arabic numerals


      0                                                                  Nature of Uric Acid Interference
      0                                                                      The presence   of as much as 1.0 ml of uric acid solu-
      2                                                                  tion (42 mg/liter)    in the reaction chamber had no ef-
                                                                         fect on the rate of glucose oxidase reaction as mea-
                                                                         sured with the oxygen electrode.        Because half this
                                                                         amount of uric acid elicited only 30% recovery in the
                        INTERFERENCE           ADDE      ML              full coupled method, Mode 1 (above) is ruled out. On
                                                                         the other hand, uric acid had a marked effect on the
Fig. 1. Relation between glucose recovery and increasing
amounts of each interfering fraction present in a reaction               peroxidase reaction. Chromogen formation          in an
mixture containing 0.1 ml of glucose solution (1 g/liter)                H2O2-peroxidase-o- dianisidine     system     decreased
A, UricacId (4.2 mg/100 ml). 8, 40000 molecular weight chromatographic
                                                                         with increasing amounts of added uric acid, the rela-
traction. C. Higher molecular weight chromatographic fraction
                                                                         tionship between absorbance and uric acid concen-
                                                                         tration closely resembling the curve of Figure l#{192}.
                                                                             Interference  Mode 2 was ruled out when it was
                                                                         found that the order of adding H202 and peroxidase
                                                                         to a uric acid buffer system made no difference in the
the overall glucose oxidase-peroxidase   reaction. Uric                  final absorbance     reached. If HO2      were destroyed
acid interferes quite strongly: 0.1 ml on the x- axis of                 nonenzymatically     by uric acid, less absorbance would
Figure l#{192} typical of the amounts of uric acid and                   be expected in those reactions        in which H202 and
glucose found in 0.1 ml of normal serum, and this                        uric acid were allowed to mix before addition      of per-
amount causes recovery to be less than 80%, corre-                       oxidase.
sponding to a decrease in the apparent value of more                       To test Mode 3, we increased the amount of perox-
than 20%. The effect of the fraction of 40 000 molecu-                   idase-o- dianisidine reagent in a series of solutions,
lar weight is shown in Figure 1B; 0.25 ml of added in-                   uric acid and H202 being held constant. The result-
terfering        solution,   typical    of the amount that would         ing absorbance increased correspondingly,        as shown
be found in 0.1 ml of normal serum, causes a 77% re-                     in the second column of Table 2, indicating       that uric
covery, again corresponding   to a decrease in the ap-                   acid interferes    by Mode 3. The nonlinear    increase in
parent value of more than 20%. The 500000 molecu-                        absorbance with increasing chromogen concentration
lar weight fraction  gives a small but measurable   ef-                  is expected for a system in which two reactants (chro-
fect (Figure 1C), causing a relative error of about 3%                   mogen and uric acid) compete for a constant amount
in normal serum.                                                         of H202 substrate.
                                                                             Further evidence for Mode 3 was obtained by not-
Possible Modes of Interference
                                                                         ing that the absorbance changes in Table 2 did not
   Five possible ways or modes in which interferences                    occur in the absence of uric acid. We also found that
could affect the glucose oxidase-peroxidase-o               -di-         uric acid was consumed during the interfering          pro-
anisidirie system are listed below and summarized             in         cess, as indicated     by a decreased absorbance of the
Figure 2.                                                                reaction mixture at 290 nm (where o- dianisidine       does
   1. Inhibition   of the glucose oxidase enzyme.                        not absorb) after the reaction. The decrease in uric
   2. Nonenzymatic        destruction      of the H202 pro-              acid concentration       was accompanied     by a corre-
duced in the glucose oxidase reaction.                                   sponding decrease in absorbance at 460 nm (where
   3. Peroxidase-catalyzed         chemical     reaction   with          the oxidized chromophore      absorbs). This could be ex-
H2O2, in competition       with o- dianisidine.                          plained by reaction of uric acid with H202, which in
   4. Inhibition   of the peroxidase enzyme.                             turn causes decreased reaction of H202 with the o-
   5. Destruction     (bleaching)     of the oxidized form of             dianisidine.
o- dianisidine.                                                              Inhibition  of peroxidase (Mode 4) could not be ob-
   In the following sections, experimental          evidence is          served because the peroxidase reaction was so rapid.
presented to elucidate and to test the nature of each                    For the glucose quantitation     procedure used here, the
of the three previously     described interferences.                     peroxidase activity was made sufficiently     high to give

122         CLINICAL   CHEMISTRY,      Vol. 21, No. 1, 1975
                                                                                 creased by increasing      the concentration     of the 40 000
Table 2. Relation Between Amount of Chromogen                                    molecular      weight protein    present     in H202-peroxi-
and Final Absorbance in the Presence of Uric                                     dase-o- dianisidine     systems.
   Acid or 40 000 Molecular Weight Interfering
                   Material’                                                         The third column of Table 2 shows that the ab-
   o.Dianisidine.      Absorbance   with 0.2 ml Absorbance         with 1.0 ml   sonbance developed in a solution containing             a fixed
     peroxidase         of uric acid (7 mg/i00      of 40000 mel wt fraction     amount of the 40 000 molecular            weight protein      in-
     reagent, ml              ml) present                    present
                                                                                 creases with the amount of peroxidase-o-           dianisidine
        0.5                      0.175                       0.091
                                                                                 reagent. No such absorbance increases were observed
        1.0                      0.248                       0.136
        1.5                      0.273                         -
                                                                                 in the absence of the interfering        material.   These re-
        2.0                      0.342                       0.191               sults indicate that the way in which this fraction in-
                                                                                 terferes is the same as that in which uric acid inter-
      Absorbance  at 460 nm developed    by 0.5 ml of H,O, (0.2 mmol/
liter), the noted volumes of interfering   material, and dianisidine-            feres (Mode 3). Similarly,       neither nonenzymatic         de-
peroxidase reagent, plus buffer to make a total volume of 4.0 ml.                struction    of H202 nor bleaching        was observed, and
                                                                                 any inhibition     of peroxidase was not great enough to
                                                                                 be significant.

                                                                                 Nature of the Interference by the Fraction
                                                                                 Containing Protein of 500 000 Molecular Weight
Table 3. Effect of the High-Molecular-Weight
  Inhibitor on the Rate of the Glucose Oxidase                                      When      the     glucoseoxidase reaction    was followed
                    Reaction’                                                    with the oxygen electrode,         increasing amounts of in-
      Vol of           Initial            Initial        Reaction  rate,   %     terference     by this high-molecular-weight          fraction
    inhibitor          slope,            current,          oxygen used
  solution,   ml      pA/mm”                pA’               per mm             caused the rate of the reaction to decrease markedly.
        0                88                486                     18.1          Table 3 summarizes data for three rate measure-
        0                98                536                     18.3          ments     made   in the absence of this high-molecular-
        0                92                511                     18.0          weight inhibitor     (to check the reproducibility       of the
       1.0               74                507                     14.6          measurement       procedure)     and in the presence of two
       2.0               58                519                     11.2          different    concentrations      of the inhibitor.   Table 3
  “Reaction     mixture consists of the indicated   volume of inhibitor          (last column) shows a linearly decreasing reaction
solution,   0.5 ml of glucose (1 mg/mI),     0.2 ml of glucose oxidase           rate with increasing     concentration   of the interfer-
(70 U/mI) and buffer to give a total volume of 4.5 ml.
     From a current-time    curve obtained    with a Clark-type oxygen           ence, indicating   that the high-molecular-weight      ma-
electrode  immersed    in the reaction     mixture.     Initial currents         terial is a glucose oxidase inhibitor operating through
given are corrected for background     currents     of 6-11 pA.                  Mode 1.
                                                                                    The above experiment      does not rule out the possi-
                                                                                 bility that catalase, as an impurity            in   the glucose oxi-
                                                                                 dase preparation,             is the target of the interfering
                                                                                 agent. Activation           of catalase by the high-molecular-
immediate chromogen oxidation             by the H202 pro-                       weight material would appear to cause an inhibition
duced by glucose oxidation.     Thus, uric acid inhibition                       of glucose oxidase. Because cyanide is known to inac-
of peroxidase is not completely       ruled out by the pre-                      tivate    catalase     but not glucose oxidase,            we repeated
ceding experiments-its      inhibition      could be masked                      the experiment          in the presence of 1 mmol of NaCN
by the large excess of peroxidase used.                                          per liter.      We obtained the same results as before,
    Chinh (12) observed that uric acid caused bleach-                            confirming       that the interfering          fraction     actually in-
ing (Mode 5) when 2,2’-azine-diethylbenzthiazoline                               hibits glucose oxidase, and not artifactually                   through
sulfonic acid was used instead of o- dianisidine.        When                    an impurity        in the catalase.
uric acid was added to the H202-peroxidase-o-            dian-                       Some additional          work was done that indicated that
isidine system after color development          was complete,                    the high-molecular-weight                fraction      also interfered
no such bleaching      was observed,          indicating  that                   with the peroxidase             reaction. Two milliliters          of the
Mode 5 was inoperative.                                                          fraction     (corresponding         to a concentration        about 15-
                                                                                 fold greater than that encountered                    in serum), was
Nature of the Interference by the Fraction                                       added to H202-peroxidase-.-o-                 dianisidine       mixture,
Containing Protein of 40 000 Molecular Weight                                    and the resulting            absorbance      was 84% of that ob-
   Experiments    similar  to those for uric acid were                           served in the absence of the interference.                 This degree
performed     by using the chromatographically        sepa-                      of interference,        when combined with that operating
rated fraction containing    the protein of 40 000 molec-                        on the glucose oxidase reaction alone, accounts for all
ular weight, concentrated     by as much as eight-fold   as                      of this interference             by the high-molecular-weight
compared     with its concentration      in normal serum.                        fraction, as measured directly on the overall reaction.
Results were similar to those observed for uric acid.                            The mechanism             of interference       with the peroxidase
   We saw no effect on the rate of the glucose oxidase                           reaction was not resolved further, mainly because the
reaction as followed with the oxygen electrode. Inter-                           turbidity     of the fraction made quantitative                measure-
ference with the peroxidase       reaction alone was in-                         ments difficult.

                                                                                                    CLINICAL    CHEMISTRY,   Vol. 21, No. 1, 1975     123
    This work was supported     in part by the National       Science Foun-              6. Salomon,   L. L., and Johnson,  J. E., Enzymatic    microdetermina-
dation (Grant GP-40694      X). During    1973, J. M. U. was the recipi-                 tion            in blood and urine. Anal. Chem.
                                                                                                of glucose                                   31,453 (1959).
ent of a Du Pont Summer       Fellowship.   Special thanks go to Ronald                  7. Cramp, D. G., New automated       method for measuring      glucose     by
Laessig, Director,  Wisconsin     State Laboratory      of Hygiene,  for pro-            glucose oxidase. J. Clin. Pathol. 20, 910 (1967).
viding the serum samples,     and to Professor     Stuart Updike, Univer-
sity of Wisconsin   School of Medicine,      for helpful discussions      and            8. Hill, J. B., and Kessler, G., An automated       determination     of glu-
for the use of the oxygen electrode.                                                     cose utilizing    a glucose oxidase-peroxidase     system.     J. Lab. Clin.
                                                                                         Med. 57, 970 (1961).
References                                                                               9. Getchell,   G., Kingsley, G. R., and Schaffert,     R. R., An automat-
1. Keston,    A. S., Colorimetric     enzymatic     reagents    for glucose. Ab-         ed direct determination      of glucose by the glucose oxidase-peroxi-
stracts    of Papers,   129th Meeting,        American       Chemical      Society,      dase system. Clin. Chem. 8,430 (1962).
April 1956, p 31C.                                                                       10. Robin, M., and Saifer, A., Determination        of glucose in biologi-
2. Teller, J. D., Direct quantitative        colorimetric      determination        of   cal fluids with an automated      enzymatic procedure.    Clin. Chem. 11,
serum     or plasma   glucose.    Abstracts     of Papers,       130th Meeting,          840 (1965).
American     Chemical   Society, Sept. 1956, p 69C.
                                                                                         11. Caraway,   W., Carbohydrates.    In      Fundamentals       of Clinical
3. Miskiewicz,   S. J., Arnett, B. B., and Simon, G. E., Evaluation                of    Chemistry.   N. W. Tietz, Ed. Saunders,        Philadelphia.   Pa.,   1970, p
a glucose oxidase-peroxidase     method    adapted to a single-channel                   159.
AutoAnalyzer   and SMA 12/60. Clin. Chem. 19, 253 (1973).
4. Saifer, A., and Gerstenfeld,         S., The photometric    microdetermi-             12. Chinh,   N. H., Mechanism     of interference       by uric acid in the
nation of blood glucose with           glucose oxidase.   J. Lab. Clin. Med.             glucose oxidase/peroxidase    method   for serum      glucose. Clin. Chem.
51, 448 (1958).                                                                          20,499 (1974).
5. Kingsley, G. R., and Getchell,   G., Direct ultramicro              glucose oxi-      13. Blaedel,   W. J., and Olson, C., Continuous    analysis by the am-
dase method  for determination    of glucose in biological             fluids. Clin.     perometric   measurement    of reaction   rate. Anal. Chem.     36, 343
Chem. 6, 466 (1960).                                                                     (1964).

124     CLINICAL      CHEMISTRY,         Vol. 21, No. 1, 1975

To top