Assessment of acyl groups and reaction conditions in the competition between perhydrolysis and hydrolysis of acyl resorufins for developing an indicator reaction for fluorometric analysis of hydrogen peroxide

Article abstract of DOI:10.1248/cpb.50.169

Perhydrolysis of acetyl resorufin (AR) was reported previously to work as a fluorometric indicator reaction for glucose determination using only glucose oxidase. However, hydrolysis of AR in blank solution rendered the working concentration range of this

Full text of DOI:10.1248/cpb.50.169

February 2002
Chem. Pharm. Bull. 50(2) 169—174 (2002)
169
Assessment of Acyl Groups and Reaction Conditions in the Competition
between Perhydrolysis and Hydrolysis of Acyl Resorufins for Developing
an Indicator Reaction for Fluorometric Analysis of Hydrogen Peroxide
Hatsuo MAEDA,* Shinya MATSU-URA, Mari NISHIDA, Yuji YAMAUCHI, and Hidenobu OHMORI
Graduate School of Pharmaceutical Sciences, Osaka University, 1–6 Yamada-oka, Suita, Osaka 565–0871, Japan.
Received July 24, 2001; accepted November 9, 2001
Perhydrolysis of acetyl resorufin (AR) was reported previously to work as a fluorometric indicator reaction
for glucose determination using only glucose oxidase. However, hydrolysis of AR in blank solution rendered the
working concentration range of this method less than two orders of magnitude. To exclude or at least signifi-
cantly reduce this interference, acyl groups and reaction conditions in the competition between perhydrolysis
and hydrolysis of various acyl resorufins were assessed. Fluorometric evaluation of reactions in the presence or
absence of H2O2 in phosphate buffer (pH 7.5, 100 mM)–CH3CN at 25 °C demonstrated that in tert-butylacetyl,
isobutyryl, cyclohexanecarbonyl and pivaloyl resorufins (TBAR, IBR, CHR and PVR, respectively) among 10
acyl resorufins examined here, the competitive situation was shifted in a much more favorable way to perhydroly-
sis than in AR, although fluorometric responses due to their H2O2-dependent deacylation were suppressed in
comparison with AR. Examination of the effects of pH, components and concentrations of buffers as well as reac-
tion temperature established reaction conditions that not only allowed perhydrolysis of each of these four com-
pounds to prevail over hydrolysis more effectively, but also improved the H2O2-based fluorometric responses.
Thus, perhydrolysis of TBAR, IBR, CHR and PVR in phosphate buffer (pH 8.0, 20 mM)–CH3CN at 25 °C worked
effectively as fluorometric indicator reactions for H2O2 analysis, affording a calibration curve over a concentra-
tion range of three orders of magnitude. Taking sensitivity, reproducibility and the response for blank solution
into consideration, PVR seemed to be the best choice as a fluorochromogen for H2O2 determination under these
conditions. For H2O2 analysis at lower pH, perhydrolysis of IBR in phosphate buffer (pH 7.5, 20 mM)–CH3CN
was shown to effectively function as an indicator reaction. Applicability of the fluorometric methods with PVR
and IBR to blood glucose determination was also discussed, comparing with Trinder’s method with phenol, 4-
aminoantipyrine and peroxidase (POD).
Key words acyl resorufin; perhydrolysis; hydrolysis; fluorometric analysis; hydrogen peroxide; deacylation
Recently, we found that acetyl resorufin (AR) is trans- ity of esters to hydrolysis can be controlled by changing
formed to fluorescent resorufin through deacetylation by steric factors of acyl groups. To our knowledge, there have
H2O2 (perhydrolysis), and hence perhydrolysis of AR can be been no reports of the effects of acyl groups on perhydrolysis
used as a fluorometric indicator reaction for glucose determi- of esters, although perhydrolysis of esters or amides has been
nation using only glucose oxidase (GOD).1) In the fluoromet- recognized as a useful tool for generation of peroxyacids in
ric method with GOD and AR, glucose analysis is performed chemical bleaching processes.4,5) Since the molecular sizes of
without interference by ascorbic acid, uric acid or bilirubin. HOOϪ and HOϪ are quite similar, steric effects in perhydrol-
These compounds interfere with colorimetric determination ysis and hydrolysis must be in the same order of magnitude.
of serum components with indicator reactions based on Accordingly, it seemed doubtful whether judicious choice of
H2O2-dependent oxidative coupling of two chromogens in acyl groups in AR could shift the competition between per-
the presence of peroxidase (POD), which are represented by hydrolysis and hydrolysis in a manner favorable towards
so called Trinder’s reactions of 4-aminoantipyrine with phe- H2O2-based deacetylation. However, perhydrolysis of aryl ac-
nolic or anilinic derivatives.2) Thus, it was demonstrated that etates is known to be much faster than their hydrolysis, i.e.,
the fluorometric method with GOD and AR offers a more re- the nucleophilic reactivity of HOOϪ toward aryl esters is
liable and accurate tool for determination of blood glucose markedly higher than that of HOϪ,6—9) which is referred to as
than the POD-dependent colorimetric method with 4- the a-effect: nucleophilicity is enhanced when the atom adja-
aminoantipyrine and phenol.3) However, there is a problem to cent to a nucleophilic site bears a lone pair of electrons.10—12)
be resolved in utilizing perhydrolysis of AR as a general flu- AR is a type of aryl acetate, and hence the a-effect is likely
orometric indicator reaction for determination of H2O2: al- to be operative in the H2O2-dependent deacetylation of AR,
though AR remains intact in CH3CN at room temperature for which allows the fluorometric determination of glucose with
more than 6 months,3) AR undergoes spontaneous hydrolysis the generation of resorufin as an indicator reaction even
when its CH3CN solution is mixed with pH 7.4 phosphate though AR undergoes hydrolysis. Taking the a-effect into
buffer (blank solution), which prevents the fluorometric consideration, it was expected that establishing reaction con-
method from being employed for the measurement of glu- ditions for the a-effect to be significantly appreciated as well
cose over a concentration range of more than two orders of as replacing the acetyl group in AR with acyl groups resis-
magnitude.
tant to hydrolysis will exclude or at least significantly reduce
The key to success in improving the measurement range in the interference by hydrolysis in fluorometric analysis of
the fluorometric method seems to be molecular design con- H2O2. Thus, we assessed acyl groups and reaction conditions
ferring the ability to resist hydrolysis of AR. The susceptibil- in the competition between perhydrolysis and hydrolysis of
To whom correspondence should be addressed. e-mail: h-maeda@phs.osaka-u.ac.jp
© 2002 Pharmaceutical Society of Japan

170
Vol. 50, No. 2
2.48 (s, 2H), 1.15 (s, 9H); 13C-NMR: d 185.98, 169.55, 153.29, 149.08,
147.97, 144.11, 134.91, 134.56, 130.95, 130.89, 119.19, 109.58, 107.02,
47.68, 31.26, 29.60 (3C); Anal. Calcd for C18H17NO4: C 69.44; H 5.50; N
4.50. Found: C 69.39; H 5.54; N 4.52.
various acyl resorufins. Here, we describe the results of these
experiments, demonstrating that perhydrolyses of pivaloyl re-
sorufin (PVR) in phosphate buffer (pH 8.0)–CH3CN and
isobutyryl resorufin (IBR) in phosphate buffer (pH 7.5)–
CH3CN are promising indicator reactions for fluorometric
analysis of H2O2 over wide working concentration ranges.
Existing indicator reactions for spectrophotometric deter-
mination of H2O2 are generally classified into two categories:
oxidative formation of dyes in the presence of POD2,13—21)
and complex formation with metals associated with batho-
chromic shift.22—26) Although the former has been found a
wide variety of clinical application, the interference men-
tioned above must be always taken into consideration. The
latter was shown to overcome such interference, and yet
sensitivity and accuracy of detection methods based on
bathochromic shift are inevitably low. As perhydrolysis of
AR was found to proceed without the influences of ascorbic
acid, uric acid or bilirubin, and also as generation or con-
Isobutyryl Resorufin (IBR): Yield 43%; mp 165 °C (from EtOAc); IR
1
(KBr): 1755, 1623 cmϪ1; H-NMR: d 7.80 (d, 1H, Jϭ8.41 Hz), 7.43 (d, 1H,
Jϭ9.89 Hz), 7.14—7.09 (m, 2H), 6.86 (dd, 1H, Jϭ9.89, 1.98 Hz), 6.33 (d,
1H, Jϭ1.98 Hz), 2.85 (sept, 1H, Jϭ6.93 Hz), 1.34 (d, 6H, Jϭ6.93 Hz); 13C-
NMR: d 185.90, 174.38, 153.61, 149.08, 147.99, 144.14, 134.89, 134.55,
130.95, 130.88, 119.03, 109.44, 107.03, 34.24, 18.78 (2C); Anal. Calcd for
C16H13NO4: C 67.84; H 4.62; N 4.95. Found: C 68.00; H 4.72; N 4.95.
Cyclohexanecarbonyl Resorufin (CHR): Yield 71%; mp 184 °C (from
EtOAc); IR (KBr): 1751, 1624 cmϪ1 1H-NMR: d 7.79 (dd, 1H, Jϭ8.41,
;
0.50 Hz), 7.43 (d, 1H, Jϭ9.89 Hz), 7.14—7.09 (m, 2H), 6.86 (dd, 1H,
Jϭ9.89, 1.98 Hz), 6.33 (d, 1H, Jϭ1.98 Hz), 2.61 (tt, 1H, Jϭ11.05, 3.63 Hz),
2.16—2.02 (m, 2H), 1.87—1.29 (m, 8H); 13C-NMR: d 185.90, 173.31,
153.66, 149.09, 147.95, 144.14, 134.88, 134.55, 130.92, 130.85, 119.09,
109.48, 107.02, 43.15, 28.82 (2C), 25.60, 25.24 (2C); Anal. Calcd for
C19H17NO4: C 70.58; H 5.30; N 4.33. Found: C 70.72; H 5.37; N 4.28.
Pivaloyl Resorufin (PVR): Yield 86%; mp 211 °C (from EtOAc); IR
1
(KBr): 1756, 1628 cmϪ1; H-NMR: d 7.80 (d, 1H, Jϭ8.73 Hz), 7.44 (d, 1H,
Jϭ9.73 Hz), 7.13—7.08 (m, 2H), 6.86 (dd, 1H, Jϭ9.73, 2.14 Hz), 6.33 (d,
sumption of resorufin has been shown to be a highly sensitive 1H, Jϭ2.14 Hz), 1.39 (s, 9H); 13C-NMR: d 185.88, 175.89, 153.88, 149.07,
fluorometric indicator reaction,27—38) it was expected that flu-
orometry based on perhydrolysis of acyl resorufins would
offer a highly sensitive and accurate method for H2O2 deter-
147.95, 144.13, 134.87, 134.55, 130.91, 130.84, 119.04, 109.44, 107.01,
39.28, 26.99 (3C); Anal. Calcd for C17H15NO4: C 68.68; H 5.09; N 4.71.
Found: C 68.81; H 5.19; N 4.73.
(1-Adamantanecarbonyl) Resorufin (ADR): Yield 64%; mp 256 °C (from
benzene); IR (KBr): 1752, 1627 cmϪ1; 1H-NMR: d 7.79 (d, 1H, Jϭ8.41 Hz),
7.43 (d, 1H, Jϭ9.89 Hz), 7.11—7.06 (m, 2H), 6.86 (dd, 1H, Jϭ9.89,
1.98 Hz), 6.32 (d, 1H, Jϭ1.98 Hz), 2.16—1.98 (m, 9H), 1.84—1.74 (m, 6H);
13C-NMR: d 185.92, 174.98, 154.00, 149.12, 147.91, 144.15, 134.87,
134.55, 130.89, 130.82, 119.14, 109.51, 107.01, 41.27, 38.64 (3C), 36.32
(3C), 27.81 (3C); Anal. Calcd for C23H21NO4: C 73.58; H 5.64; N 3.73.
Found: C 73.70; H 5.74; N 3.69.
mination, provided that the contribution of spontaneous hy-
drolysis to generation of resorufin can be significantly re-
duced. Therefore, the present study was performed to de-
velop a totally new class of indicator reactions for fluoromet-
ric analysis of H2O2.
Experimental
Benzoyl Resorufin (BR): Yield 72%; mp 261—263 °C (from benzene); IR
1
A fluorescence spectrophotometer (FP-750, JASCO) equipped with a (KBr): 1735, 1627 cmϪ1; H-NMR: d 8.21 (d, 2H, Jϭ7.99 Hz), 7.86 (d, 1H,
Peltier thermostatted single cell holder (ETC-272, JASCO) was used. All Jϭ8.41 Hz), 7.72—7.66 (m, 1H), 7.58—7.52 (m, 2H), 7.45 (d, 1H,
melting points were measured on a Yanako MP-S3 micro-melting point ap- Jϭ9.73 Hz), 7.30—7.25 (m, 2H), 6.87 (dd, 1H, Jϭ9.73, 1.98 Hz), 6.35 (d,
paratus, and are given uncorrected. Infrared (IR) spectra were taken on a 1H, Jϭ1.98 Hz); 13C-NMR: d 185.98, 164.06, 153.54, 149.08, 148.09,
JASCO VALOR-III spectrometer. H- and 13C-NMR spectra were obtained 144.19, 134.95, 134.58, 133.99, 131.10, 130.98, 130.12 (2C), 128.56 (2C),
1
in CDCl3 at 270 and 67.8 MHz on a JEOL EX-270 spectrometer. 13C-NMR 128.40, 119.24, 109.72, 107.09; Anal. Calcd for C19H11NO4: C 71.92; H
spectral data were obtained under off-resonance conditions. Chemical shifts 3.50; N 4.42. Found: C 71.92; H 3.65; N 4.34.
are given in ppm relative to tetramethylsilane (TMS) as an internal standard
(d 0.00).
(4-Methoxybenzoyl) Resorufin (MOBR): Yield 93%; mp 233 °C (from
benzene); IR (KBr): 1732, 1628 cmϪ1; 1H-NMR: d 8.16 (d, 2H, Jϭ8.90 Hz),
Materials GOD from Aspergillus niger (EC 1.1.3.4, 238 U/mg) was 7.84 (d, 1H, Jϭ8.41 Hz), 7.45 (d, 1H, Jϭ9.89 Hz), 7.28—7.23 (m, 2H), 7.01
used as supplied from Wako Pure Chemical Industries, Ltd. All other chemi- (d, 2H, Jϭ8.90 Hz), 6.87 (dd, 1H, Jϭ9.89, 2.14 Hz), 6.35 (d, 1H,
cals were of reagent grade and were used as received. Stock solutions of acyl Jϭ2.14 Hz), 3.92 (s, 3H); 13C-NMR: d 186.00, 164.12, 163.74, 153.80,
resorufins were prepared in HPLC-grade CH3CN (Wako Pure Chemical In-
149.13, 147.95, 144.19, 134.90, 134.59, 132.32 (2C), 130.99, 130.92,
dustries, Ltd.). Solutions of H2O2 were prepared daily in deionized and dis- 120.59, 119.36, 113.89 (2C), 109.74, 107.04, 55.51; Anal. Calcd for
tilled water, phosphate (Na2HPO4ϩNaH2PO4; pH 7.0, 7.5 or 8.0; 100, 50, C20H13NO4: C 69.16; H 3.77; N 4.03. Found: C 69.02; H 3.93; N 3.96.
20 or 10 mM), HEPES (2-[4-(2-hydroxyethyl)-1-piperazinyl]ethansulfonic
acidϩNaCl adjusted with aqueous 0.1 M NaOH; pH 8.0 or 8.5; 50 mM) or
borate (H3BO3ϩKCl adjusted with aqueous 0.1 M NaOH; pH 8.0 or 8.5;
(2-Furoyl) Resorufin (FUR): Yield 77%; mp 255—258 °C (from ben-
1
zene); IR (KBr): 1747, 1623 cmϪ1; H-NMR: d 7.84 (d, 1H, Jϭ8.41 Hz),
7.73—7.72 (m, 1H), 7.46—7.43 (m, 2H), 7.29—7.24 (m, 2H), 6.87 (dd, 1H,
50 mM) buffer. AR was prepared and purified as reported previously.1) Other Jϭ9.89, 1.32 Hz), 6.64—6.63 (m, 1H), 6.34 (d, 1H, Jϭ1.32 Hz); 13C-NMR:
acyl resorufins were synthesized as follows: To a suspension of resorufin d 185.97, 155.59, 152.70, 149.03, 148.20, 147.59, 144.16, 142.90, 134.99,
sodium salt (2.0 g, 8.5 mmol) in pyridine (20 ml) at Ϫ40 °C was added the
134.58, 131.17, 131.02, 120.32, 119.02, 112.31, 109.58, 107.13; Anal. Calcd
corresponding acid chloride (10.2 mmol) under a nitrogen atmosphere, and for C17H9NO5: C 66.45; H 2.95; N 4.56. Found: C 66.41; H 3.16; N 4.57.
the resulting mixture was stirred at Ϫ40—Ϫ10 °C for 6 h. The mixture was
poured into CH2Cl2 (400 ml), and the organic layer was washed with 1 M aq.
HCl (400 ml), 10% aq. NaHCO3 (400 ml), and brine (400 ml) and dried over
MgSO4. After concentration under reduced pressure, the residue was sub-
Assessment of Acyl Groups and Reaction Conditions All fluoromet-
ric measurements were carried out at excitation and emission wavelengths of
572 and 589 nm, respectively. A CH3CN solution of acyl resorufin (0.1 mM,
1.0 ml), blank buffer (1.0 ml) and H2O2 solution (1.2 or 0 mM, 1.0 ml) in the
jected to column chromatography with CH2Cl2–acetone (50 : 1—20 : 1) as same buffer were added in this order to a cuvette (10ϫ10ϫ45 mm) in the
the eluent, and the obtained product was further purified by recrystallization.
cell holder with stirring at 500 rpm and the specified temperature. The fluo-
Isovaleryl Resorufin (IVR): Yield 48%; mp 154 °C (from EtOAc); IR rometric measurement of the mixture was started 30 s after addition of H2O2
1
(KBr): 1762, 1625 cmϪ1; H-NMR: d 7.79 (dd, 1H, Jϭ8.66, 0.50 Hz), 7.43 solution, and the reaction was followed for 300 s.
(d, 1H, Jϭ9.73 Hz), 7.14—7.09 (m, 2H), 6.86 (dd, 1H, Jϭ9.89, 1.98 Hz),
6.33 (d, 1H, Jϭ1.98 Hz), 2.49 (d, 2H, Jϭ7.09 Hz), 2.33—2.18 (m, 1H), 1.08
Calibration Curves The same procedure as described above was used
except that a CH3CN solution of acyl resorufin (0.1 mM, 0.8 ml), blank buffer
(d, 6H, Jϭ6.60 Hz); 13C-NMR: d 185.95, 170.35, 153.34, 149.06, 147.97, (2.0 ml) and aqueous H2O2 solution (0.2 ml) were added to the cuvette in this
144.11, 134.90, 134.55, 130.94, 130.89, 119.11, 109.52, 107.02, 43.18,
25.79, 22.33 (2C); Anal. Calcd for C17H15NO4: C 68.68; H 5.09; N 4.71.
Found: C 68.62; H 5.14; N 4.68.
order.
Glucose Determination In a microtube (1.5 ml), H2O (150 ml), aqueous
ZnSO4 (62 mM, 600 ml), aqueous Ba(OH)2 (30 mM, 600 ml) and aqueous glu-
tert-Butylacetyl Resorufin (TBAR): Yield 45%; mp 168 °C (from EtOAc); cose solution (0—300 mg/dl, 150 ml) or plasma (150 ml) were added, and the
1
IR (KBr): 1759, 1625 cmϪ1; H-NMR: d 7.80 (d, 1H, Jϭ8.41 Hz), 7.43 (d, resulting mixture was centrifuged at 4 °C and 10000 rpm for 10 min. The ob-
1H, Jϭ9.89 Hz), 7.14—7.09 (m, 2H), 6.86 (d, 1H, Jϭ9.89 Hz), 6.33 (s, 1H), tained supernatants were used as glucose standard solution or plasma sam-

February 2002
171
ple. The fluorometric detection of glucose was performed as follows. A
CH3CN solution of PVR or IBR (0.1 mM, 0.8 ml), a GOD (0.25 mg/ml) solu-
tion in phosphate buffer (20 mM, pH 8.0 or 7.5, 2.0 ml) and glucose standard
solution or plasma sample (0.2 ml) were added in this order to the cuvette in
the cell holder with stirring at 500 rpm and 25 °C. The fluorometric response
of the mixture was measured 300 s after addition of glucose standard solu-
tion or plasma sample. Glucose determination by Trinder’s method was car-
ried out with the same glucose standard solution or plasma in the previously
reported manner.3)
Fig. 1. Chemical Structures and Abbreviations of Acyl Resorufins Used in
This Study
Results and Discussion
The acyl resorufins examined in this study are shown in
Fig. 1. Generation of resorufin by deacylation in the presence
or absence of H2O2 was fluorometrically followed from 0.5 to
5.5 min after starting the reactions. Typical traces for AR and
PVR are shown in Fig. 2. The effects of acyl groups and re-
action conditions on the competition between perhydrolysis
and hydrolysis were assessed with the following measures
obtained from the fluorometric traces: increments in fluores-
cence intensity during the 5 min measurements of perhydrol-
ysis and hydrolysis (DI and DI0, respectively) and their ratio
(DI/DI0). Values of DI and DI0 were informative for estimat-
ing the extent of perhydrolysis and hydrolysis of acyl re-
sorufins. Especially, DI and DI/DI0 values served to assess
acyl groups and reaction conditions in the competition.
Acyl groups were first screened. Deacylation of acyl re-
sorufins (0.1 mmol) were carried out in phosphate buffer (pH
7.5, 100 mM, 2.0 ml)–CH3CN (1.0 ml) in the presence or ab-
sence of H2O2 (1.2 mmol) at 25 °C. The results are summa-
rized in Table 1. Judging from DI0 values, FUR was hy-
drolyzed as easily as AR and the other acyl resorufins were
less susceptible to hydrolysis than AR. Retardation of hydrol-
ysis of these acyl resorufins was always accompanied by de-
creases in fluorometric responses by H2O2-dependent deacy-
lation (cf. DI values). However, comparison of DI/DI0 values
suggested that replacing the acetyl group with sterically hin-
dered acyl groups such as tert-butylacetyl, cyclohexanecar-
bonyl and pivaloyl exerted suppressive effects on hydrolysis
much more strongly than on perhydrolysis. Thus, TBAR,
CHR and PVR were chosen for further examination to de-
velop a perhydrolysis-based indicator reaction for H2O2
analysis. In addition, IBR was also chosen as a candidate for
our purpose because IBR exhibited much larger DI than
TBAR, CHR and PVR, although its DI/DI0 was lower than
those for other three compounds.
Fig. 2. Fluorometric Traces Obtained for Generation of Resorufin from
AR (a, b) and PVR (c, d) (0.1 mmol Each) in the Presence (a, c) or Absence
(b, d) of H2O2 (1.2 mmol) in Phosphate Buffer (pH 7.5, 100 mM, 2.0 ml)–
CH3CN (1.0 ml) at 25 °C
Table 1. Increments in Fluorescence Intensity during Perhydrolysis (DI) or
Hydrolysis (DI0) of Acyl Resorufin (0.1 mmol) in the Presence or Absence of
H2O2 (1.2 mmol), Respectively, in Phosphate Buffer (pH 7.5, 100 mM,
2.0 ml)–CH3CN (1.0 ml) at 25 °Ca)
Acyl resorufin
DI
DI0
DI/DI0
AR
IVR
261.5
43.8
18.0
116.8
55.4
38.1
10.1
40.4
13.2
140.8
63.1
4.5
0.9
13.6
5.4
2.8
4.2
5.0
3.2
54.8
4.1
9.8
20.0
8.6
10.3
13.6
2.4
8.1
4.1
2.6
TBAR
IBR
CHR
PVR
ADR
BR
MOBR
FUR
The dependence of the competition between perhydrolysis
and hydrolysis of TBAR, IBR, CHR and PVR on pH and
buffer components was examined, where phosphate (pH 7.0,
7.5 and 8.0, 100 mM), HEPES (pH 8.0 and 8.5, 50 mM) and
borate (pH 8.0 and 8.5, 50 mM) buffers were used. Deacyla-
a) The values of DI or DI0 were obtained as changes between the fluorometric re-
sponses observed 0.5 and 5.5 min after starting the reactions in the presence or absence
of H2O2, respectively.
tion of these acyl resorufins (0.1 mmol) in the presence or ab- In borate buffer (pH 8.5)–CH3CN, DI0 values were increased
sence of H2O2 (1.2 mmol) was carried out at 25 °C in buffer to a greater extent than DI, probably because the interaction9)
(2.0 ml)–CH3CN (1.0 ml). Figure 3 compares DI/DI0 and DI of H2O2 with B(OH)3 inhibits perhydrolysis of acyl re-
obtained in deacylation of these acyl resorufins. Similar sorufins. Examination of DI/DI0 values demonstrated that
trends in these measures were recognized regardless of acyl perhydrolysis of these acyl resorufins was favored over hy-
groups. When the reactions were performed in phosphate or drolysis when pH 8.0 phosphate or pH 8.5 HEPES buffer was
HEPES buffer–CH3CN, larger values of DI as well as DI/DI0 used. However, as can be seen in Fig. 3, the effectiveness of
were obtained with increasing pH of the buffer. The pH-de- buffer components to increase DI values in buffer (pH
pendent increments in DI/DI0 were due to greater increase in 8.0)–CH3CN was borateϾphosphateϾHEPES. The values of
DI rather than decrease in DI0 by more basic media. For bo- DI observed in HEPES buffer (pH 8.5)–CH3CN were less
rate buffer, larger values of DI/DI0 were obtained at pH 8.0 than 70% of those in phosphate buffer (pH 8.0)–CH3CN.
rather than pH 8.5, although DI was larger at pH 8.5 than 8.0. Thus, from the standpoint of sensitivity, it was concluded

172
Vol. 50, No. 2
Fig. 5. Effects of Reaction Temperature on DI/DI0 and DI Obtained in
Deacylation of TBAR, IBR, CHR and PVR (0.1 mmol Each) in the Presence
or Absence of H2O2 (1.2 mmol) in Phosphate Buffer (pH 8.0, 20 mM,
2.0 ml)–CH3CN (1.0 ml)
Fig. 3. Effects of pH and Components of Buffer on DI/DI0 and DI Ob-
tained in Deacylation of TBAR, IBR, CHR and PVR (0.1 mmol Each) in the
Presence or Absence of H2O2 (1.2 mmol) in Buffer (2.0 ml)–CH3CN (1.0 ml)
at 25 °C
The concentrations of phosphate, HEPES and borate buffers were 100, 50 and
50 mM, respectively.
perhydrolysis became larger, while hydrolysis was sup-
pressed to a greater extent, leading to a larger DI/DI0; and
[from 20 to 10 mM] DI was almost constant, while DI0 de-
creased further. Thus, when the concentration of phosphate
buffer was changed from 100 to 10 mM, DI reached almost
the maximum value in phosphate buffer (20 mM)–CH3CN,
while DI0 continued to decrease. The competition between
perhydrolysis and hydrolysis was shifted in the most favor-
able manner for perhydrolysis of these four acyl resorufins as
an indicator reaction for H2O2 analysis, when the fluoromet-
ric measurements were conducted in phosphate buffer (pH
8.0, 20 mM)–CH3CN. It should be emphasized here that em-
ploying these conditions considerably restored the loss of
H2O2-dependent fluorometric responses induced by replacing
the acetyl group in AR with tert-butylacetyl, isobutyryl, cy-
clohexanecarbonyl and pivaloyl groups when perhydrolysis
was carried out in phosphate buffer (pH 7.5, 100 mM)–
CH3CN.
Reaction temperature also exerted an influence on both
perhydrolysis and hydrolysis of TBAR, IBR, CHR and PVR.
As deacylation of these acyl resorufins was carried out at a
lower reaction temperature, both DI and DI0 became smaller.
However, the suppressive effects on hydrolysis were about
twofold stronger than on perhydrolysis, leading to larger val-
ues of DI/DI0 at lower reaction temperature as shown in Fig.
5. With regard to DI/DI0, the best reaction temperature was
15 °C. However, consideration of DI values as well as experi-
mental convenience for analytical application indicated that
fluorometric analysis of H2O2 would be effectively achieved
with perhydrolysis of these acyl resorufins at 25 °C as an in-
dicator reaction.
The relationships between H2O2 concentration and fluores-
cence intensity of resorufin generated through perhydrolysis
of TBAR, IBR, CHR or PVR were examined. Perhydrolysis
was performed by adding aqueous H2O2 solution (0.2 ml) to
a mixture of a CH3CN solution of each acyl resorufin
(0.1 mM, 0.8 ml) and phosphate buffer (pH 8.0, 20 mM,
Fig. 4. Effects of the Concentration of pH 8.0 Phosphate Buffer on DI/DI0
and DI Obtained in Deacylation of TBAR, IBR, CHR and PVR (0.1 mmol
Each) in the Presence or Absence of H2O2 (1.2 mmol) in the Buffer
(2.0 ml)–CH3CN (1.0 ml) at 25 °C
that phosphate buffer (pH 8.0)–CH3CN is the solvent of
choice for perhydrolysis of acyl resorufins.
Both perhydrolysis and hydrolysis of TBAR, IBR, CHR
and PVR in phosphate buffer (pH 8.0)–CH3CN at 25 °C were
affected by the buffer concentration. As shown in Fig. 4,
DI/DI0 values became larger as a lower concentration buffer
was employed. The concentration effects were smaller when
acyl groups bore less substitutions at a-carbons as in the
case of TBAR. These observations stemmed from the follow-
ing: [from 100 to 20 mM] DI informative for the extent of

February 2002
173
Table 2. Calibration Curves for H2O2 Obtained by Fluorometry with Per- Table 3. Comparison between Fluorometry with Perhydrolysis of PVR at
hydrolysis of TBAR, IBR, CHR or PVR as an Indicator Reaction
pH 8.0 (Method I) or IBR at pH 7.5 (Method II) and Colorimetry with Phe-
nol, 4-Aminoantipyrine and POD (Trinder’s Method) as a Tool for Glucose
Determination
Data for calibration curves
Conditionsa)
Slope
(a.u./nmol)
Interception
(a.u.)
Linear range
(nmol)
Method I
Method II
Trinder’s Method
Calibration curve 300—50a,b) mg/dl 300—50a,b) mg/dl 300—50a,b) mg/dl
TBAR/pH 8.0
IBR/pH 8.0
CHR/pH 8.0
PVR/pH 8.0
IBR/pH 7.5
0.070
0.356
0.184
0.128
0.129
5.80
21.12
10.88
4.99
1200—1.2
1200—0.6
1200—1.2
1200—0.6
1200—1.2
(r)
(1.000)
10 mg/dl
(11.5%)
(1.000)
5 mg/dl
(18.5%)
(1.000)
20 mg/dl
(19.6%)
Detection limitb)
(RSD, nϭ3)
Glucose level obtained
for plasma
9.21
87.5 mg/dl
(1.2%)
91.2 mg/dl
(1.9%)
83.2 mg/dl
(1.2%)
(RSD, nϭ6)
Effectsc) of ascorbic acid
0.2 mM
a) Fluorometric measurements were carried out in 20 mM phosphate buffer at 25 °C.
99.0%
102.6%
97.2%
99.4%
101.1%
93.5%
2.0 ml) at 25 °C. Calibration curves were obtained by plotting
H2O2 concentrations vs. absolute fluorescence intensity ob-
served 5.5 min after commencement of perhydrolysis. The
results are summarized in Table 2. The lowest detection limit
was determined as the H2O2 concentration allowing its fluo-
rescence intensity to always exceed that of blank solution
when the measurements were repeated three times.
1.0 mM
a) Determination limit. b) The limits of determination and detection were deter-
mined as the glucose concentrations giving spectrophotometric responses with RSD not
exceeding 10 and 30%, respectively. c) Shown as relative values between responses
observed for 100 mg/dl glucose in the presence and absence of ascorbic acid.
cordingly, the sensitivity in the present fluorometry was
Perhydrolysis of each of TBAR, IBR, CHR and PVR ex- lower than in PO-CL methods. However, highly sensitive and
hibited a good linear relationship with H2O2 concentration reproducible detection of H2O2 by the CL methods seem to
over the specified range with three orders of magnitude, with depend on how an oxalate ester and a fluorophore in organic
correlation coefficients (r) being more than 0.999. Relative solvent such as ethyl acetate, chloroform or CH3CN are ef-
standard deviations (RSDs) (nϭ3) of fluorescence responses fectively mixed with H2O2 in water. This is at least in part
obtained with these acyl resorufins were less than 2.5% over why PO-CL methods have been used mainly in a flow sys-
the specified concentration ranges: using perhydrolysis of tem. In addition, the solubility and stability of generally used
TBAR, IBR, CHR and PVR, RSD values of fluorescence in- oxalates esters such as bis(2,4,6-trichlorophenyl) oxalate in a
tensity for the lowest H2O2 concentrations were 2.2, 1.7, 1.3 water-rich system seem much poorer than those of PVR and
and 1.1%, and those for blank solution were 2.0, 1.1, 1.8 and IBR. Thus, the present fluorometry is believed to work as a
0.2%, respectively. Although perhydrolysis of these four acyl complementary tool useful for determination of H2O2 in a
resorufins will work as fluorometric indicator reactions, the static and water-rich system, although the sensitivity must be
method with PVR under these conditions seemed to provide improved somehow.
the best tool for H2O2 analysis, taking sensitivity, repro-
ducibility and the response for blank solution into considera- at pH 8.0 (Method I) and that of IBR at pH 7.5 (Method II)
tion. as a method of glucose determination with GOD was evalu-
Possibility of the fluorometry with perhydrolysis of PVR
Comparison of the slope values indicated that the method ated by comparison with colorimetry using phenol, 4-
with IBR exhibited the highest sensitivity, although its inter- aminoantipyrine and POD as an indicator reaction system
cept indicated that the stability of IBR in blank solution was (Trinder’s method). The results are summarized in Table 3.
the worst. The possibility of perhydrolysis of IBR as an indi- Determination and detection limits were determined as the
cator reaction at pH lower than 8.0 was examined. Perhydrol- glucose concentrations giving spectrophotometric responses
ysis was conducted under essentially the same conditions ex- with RSD values not exceeding 10 and 30%, respectively.
cept that pH 7.5, 20 mM phosphate buffer was used. Under The determination limits were the same in these three meth-
these conditions, the sensitivity was decreased by more than ods, providing linear calibration graphs up to 300 mg/dl.
40% of that with pH 8.0 buffer, and yet the method with IBR Both Methods I and II gave slightly lower detection limits
provided a satisfactory calibration curve with a correlation than Trinder’s method. Although a deproteinization proce-
coefficient of 1.000 where RSDs of fluorescence responses dure was required as in the case of AR,3) the present methods
for the lowest H2O2 concentration and for blank solution were successfully applied for blood glucose determination
were 0.3 and 0.8%, respectively. Thus, perhydrolysis of IBR with satisfactory reproducibility. The determined glucose
may function well as an indicator reaction for H2O2 analysis levels for the same plasma sample were comparable to each
at pH 7.5.
other, but slightly higher than that obtained by Trinder’s
The present fluorometric methods with PVR and IBR fea- method. It was expected that the methods with PVR and IBR
ture that the indicator reactions proceed at a neutral pH re- were not adversely affected by ascorbic acid, uric acid or
gion without recourse of POD, which seemed to eliminate in- bilirubin similarly to the fluorometry with AR.1,3) To confirm
terference by reducing biological compounds such as ascor- this anticipation, the effects of ascorbic acid were examined.
bic acid as described below. Under similar neutral and POD- As also shown in Table 3, ascorbic acid even at 1.0 mM much
independent conditions, peroxyoxalate chemiluminescence higher than in normal plasma had only negligible influence
(PO-CL) methods are known to be useful for determination on the fluorometric responses observed for a 100 mg/dl glu-
of H2O2.39—47) Detection limits of PO-CL methods were re- cose solution by Methods I and II. In Trinder’s method, glu-
ported to range from 10Ϫ10 to 10Ϫ12 mol levels, depending on cose at the same concentration was successfully detected in
oxalate esters, fluorophores and reaction systems used. Ac- the presence of ascorbic acid at 0.2 mM, but was underesti-

174
Vol. 50, No. 2
17) Schubert F., Wang F., Rinneberg H., Mikrochim. Acta, 121, 237—247
(1995).
18) Mizoguchi M., Ishiyama M., Shiga M., Sasamoto K., Bunseki Kagaku,
45, 111—124 (1996).
19) Zhu Q.-Z., Zheng X.-Y., Xu J.-G., Liu F.-H., Li Q.-G., Microchem. J.,
57, 332—338 (1997).
20) Mizoguchi M., Ishiyama M., Shiga M., Sasamoto K., Anal. Commun.,
35, 71—73 (1998).
21) Kakuragawa A., Taniai T., Okutani T., Anal. Chim. Acta, 374, 191—
200 (1998).
22) Matsubara C., Ishii K., Takamura K., Microchem. J., 26, 242—249
(1981).
23) Matsubara C., Kudo K., Kawashita T., Takamura K., Anal. Chem., 57,
1107—1109 (1985).
24) Warren L. B., J. Inst. Brew., 97, 457—462 (1991).
25) Harms D., Than R., Pinkernell U., Schmidt M., Krebs B., Karst U., An-
alyst, 123, 2323—2327 (1998).
26) Harms D., Meyer J., Westerheide L., Krebs B., Karst U., Anal. Chim.
Acta, 401, 83—90 (1999).
27) Herrmann R., Chimia, 45, 317—318 (1991).
28) Prough R. A., Burke M. D., Mayer R. T., “Methods in Enzymology,”
ed. by Wood W. A., Academic Press, New York, 1978, Vol. 52, pp.
372—376.
29) Brotea F. P., Thibert R. J., Microchem. J., 37, 368—376 (1988).
30) Brotea G. P., Draisey T. F., Thibert R. J., Microchem. J., 39, 1—9
(1989).
31) Tokutake S., Kasai K., Tomikura T., Yamaji N., Kato M., Chem.
Pharm. Bull., 38, 3466—3470 (1990).
mated by more than 5% for the glucose solution containing
1.0 mM ascorbic acid.
In conclusion, the results described here demonstrated that
judicious choice of acyl groups and reaction conditions can
shift the competition between perhydrolysis and hydrolysis of
acyl resorufins in a manner favorable for perhydrolysis to
work as a fluorometric indicator reaction for H2O2 analysis
with a desirable working range. Furthermore, the fluorometry
with perhydrolysis of PVR and IBR chosen as the best indi-
cator reactions for determination of H2O2 at pH 8.0 and 7.5,
respectively, were shown to be applied for determination of
blood glucose with only GOD. Thus, acyl resorufins such as
PVR and IBR are believed to work not only as novel fluo-
rochromogens useful for clinical chemistry but also as new
types of fluorescent probes for detecting H2O2 released from
or generated in cells. However, for clinical or cell-biological
uses, H2O or non-toxic dimethylsulfoxide is generally pre-
ferred to toxic CH3CN as a solvent for a reagent solution.
Accordingly, water-soluble acyl resorufins must be designed
or a detection system with acyl resorufins in dimethylsulfox-
ide at as small volumes as possible should be developed. Fur-
ther studies along these lines are currently underway in our
laboratory.
32) Simpson D. J., Unkefer C. J., Whaley T. W., Marrone B. L., J. Org.
Chem., 56, 5391—5396 (1991).
33) Kasai K., Yamaji N., Anal. Sci., 8, 161—164 (1992).
34) Hadd A. G., Raymond D. E., Halliwell J. W., Jacobson S. C., Ramsey
J. M., Anal. Chem., 69, 3407—3412 (1997).
Acknowledgement This study was supported in part by Mitsubishi
Chemical Corporation Fund and by a Grant-in-Aid for Scientific Research
(12877353) from the Ministry of Education, Science, Sports and Culture,
Japan.
35) Tortorella M. D., Arner E. C., Inflamm. Res., 46, S122—S123 (1997).
36) Zhou M., Diwu Z., Panchuk-Voloshina N., Haugland R. P., Anal.
Biochem., 253, 162—168 (1997).
37) O’neill R. B., Dillon S. A., Morgan P. M., Biochem. Soc. Trans., 26,
S84 (1998).
References
1) Maeda H., Matsu-ura S., Senba T., Yamauchi Y., Ohmori H., Chem.
Pharm. Bull., 49, 294—298 (2001).
2) Michal G., Möllering H., Siedel J., “Methods of Enzymatic Analysis,”
ed. by Bergmeyer H. U., 3rd ed., VCH Publishers, Weinheim, 1983, 38) Candeias L. P., MacFarlane D. P. S., McWhinnie S. L. W., Maidwell N.
Vol. 1, pp. 197—232.
3) Matsu-ura S., Yamauchi Y., Ohmori H., Maeda H., Bunseki Kagaku,
50, 475—479 (2001).
L., Roeschlaub C. A., Sammes P. G., Whittlesy R., J. Chem. Soc.
Perkin Trans. 2, 1998, 2333—2334.
39) Williams III, D. C., Huff G. F., Seitz W. R., Anal. Chem., 48, 1003—
1006 (1976).
40) Sherman P. A., Holzbecher J., Ryan D. E., Anal. Chim. Acta, 97, 21—
27 (1978).
41) Scott G., Seitz W. R., Ambrose J., Anal. Chim. Acta, 115, 221—228
(1980).
42) Gübitz G., van Zoonen P., Gooijer C., Velthorst N. H., Frei R. W.,
Anal. Chem., 57, 2071—2074 (1985).
43) Abdel-Latif M. S., Guilbault G. G., Anal. Chem., 60, 2671—2674
(1988).
4) James A. P., Mackirdy I. S., Chem. Ind. (London), 1990, 641—645.
5) Grime K., Clauss A., Chem. Ind. (London), 1990, 647—653.
6) Jencks W. P., Carriuolo J., J. Am. Chem. Soc., 82, 1778—1786 (1960).
7) Jencks W. P., Gilchrist M., J. Am. Chem. Soc., 90, 2622—2637 (1968).
8) Gilbert B. C., Holmes R. G. G., Marshall R. D. R., Norman R. O. C., J.
Chem. Res. (M), 1977, 1949—1973.
9) Davis D. M., Deary M., J. Chem. Res. (M), 1988, 2720—2736.
10) Klopman G., Tsuda K., Louis J. B., Davis R. E., Tetrahedron, 26,
4549—4554 (1970).
11) Grekov A. P., Veselov Y. Ya., Rus. Chem. Rev. (Engl. Transl.), 47,
631—648 (1978).
44) Nakashima K., Maki K., Kawaguchi S., Akiyama S., Tsukamoto Y.,
Imai K., Anal. Sci., 7, 709—713 (1991).
12) Hoz S., Nuncel E., Isr. J. Chem., 26, 313—319 (1985).
13) Capaldi D. J., Taylor K. E., Anal. Biochem., 129, 329—336 (1983).
14) Madsen B. C., Kromis M. K., Anal. Chem., 56, 2849—2850 (1984).
15) Reljíc R., Ries M., Aníc N., Ries B., Clin. Chem., 38, 522—525
(1992).
45) Nakashima K., Hayashida N., Kawaguchi S., Akiyama S., Tsukamoto
Y., Imai K., Anal. Sci., 7, 715—718 (1991).
46) Katayama M., Takeuchi H., Taniguchi H., Anal. Lett., 24, 1005—1015
(1991).
47) Hasebe T., Hasegawa E., Kawashima T., Anal. Sci., 12, 881—885
(1996).
16) Mizoguchi M., Shiga M., Sasamoto K., Chem. Pharm. Bull., 41,
620—623 (1993).