Clinical Chemistry 46: 644-649, 2000;
(Clinical Chemistry. 2000;46:644-649.)
© 2000 American Association for Clinical Chemistry, Inc.
Routine 
-Amylase Assay Using Protected 4-Nitrophenyl-1,4--D-maltoheptaoside and a Novel -Glucosidase
Klaus Lorentza,1
1
Institut für Klinische Chemie, Medizinische Universität Lübeck, D-23538 Lübeck, Germany.
a Address for correspondence: Hugo-Kauffmann-Strasse 7, D-83209 Prien, Germany. Fax 49-8051-969032.
 |
Abstract
|
|---|
Background: In contrast to numerous methods for measuring
-amylase activity, the approved IFCC reference method offers
an invariable time-independent constant product pattern, thus avoiding
possibly changing stoichiometric calculations. However, reference
methods do not lend themselves to routine use, so that such methods
need to be modified.
Methods: Ethylidene-blocked 4-nitrophenylmaltoheptaoside (EPS-G7)
is degraded to glucose and 4-nitrophenol in a coupled assay with a
bacterial -glucosidase under the following measurement conditions:
3.5 mmol/L EPS-G7, 7.1 kU/L -glucosidase, 70 mmol/L sodium chloride,
1 mmol/L calcium chloride, 50 mmol/L HEPES, pH 7.15, at 37 °C. The
increase of absorbance is continuously monitored for 3 min at
405 nm after a lag phase of 2 min.
Results: Catalytic concentrations up to 15-fold higher than the
upper reference limit can be determined without dilution. Precision
studies in manual performance show CVs of 1.4–2.6%
(within-run) and 1.9–2.8% (day-to-day). There was no interference
from 100 mmol/L glucose, 30 mmol/L triacylglycerols, 610 µmol/L
bilirubin, and 2.95 g/L hemoglobin. The method closely correlates with
other chromogenic assays. The preliminary 0.95 reference interval for
adults, not dependent on age and sex, is 33.6–96.2 U/L.
Conclusion: The procedure is a robust adaptation of the reference
method to routine use at 37 °C with increased sensitivity, fewer
interferences, and reduced cost.
|
Introduction
|
|---|
Despite of its lack of diagnostic sensitivity and specificity, the
measurement of -amylase (1,4--D-glucan
glucanohydrolase; EC 3.2.1.1) is the most widely used test for
diagnosing acute pancreatitis (1). Consequently,
several methods have emerged that use different substrates and produce
diverse results (2). Hence, to enable primary
standardization based on a reference method, the IFCC has
approved a reference measurement procedure that applies
4,6-ethylidene(G7)-1[4-nitrophenyl(G1)]-1,4--D-maltoheptaoside
(EPS-G7)1
and a novel -glucosidase
(-D-glucoside glucohydrolase; EC 3.2.1.20),
which uniformly degrades all reaction products to 4-nitrophenol and
glucose (3). However, reference methods are optimized for
manual performance at a reaction temperature of 30 °C, and they are
not compromised by economic considerations. Hence, the present report
describes an adaptation for use in service laboratories that retains
the analytical qualities of the original method.
|
Materials and Methods
|
|---|
instruments
We followed the reaction rates with Eppendorf spectral line
photometers 1101M and 6114S (Netheler & Hinz), equipped with a
recorder, in thermostated 10-mm light path cuvettes. An Eppendorf EPOS
5060 analyzer was used for method comparisons, the establishment of
reference intervals, and as a spectrometer for the determination of
molar absorption coefficients at 405 nm. The pH 531 pH meter
(WTW) with a glass electrode (405-S7; Ingold) was calibrated with
buffer solutions (related to standard reference materials of the
National Institute of Standards) from Merck.
specimens and reagents
We used human salivary (4) and pancreatic amylase
[Certified Reference Material 476 from the National Institute of
Biological Standards and Control, Potters Bar, United Kingdom] and
sera containing amylase concentrations fivefold higher than the upper
reference limit. Extensively dialyzed [against 2.5 mmol/L
piperazine-N,N'-bis(2-ethanesulfonic acid), pH 7.0]
enzyme preparations were used for effector studies. For the
establishment of reference intervals, we prospectively selected sera
from 80 apparently healthy adults (divided into five groups with eight
males and females, each representing a decade between 20 and 69 years
of age). All individuals fulfilled the following inclusion criteria:
protein concentration, 58–76 g/L with usual electropherogram; glucose
(fasting) <6.1 mmol/L; creatinine <95 µmol/L; and
-glutamyltransferase <20 U/L (25 °C). Method comparisons were
performed with sera of various amylase concentrations taken at random
using the IFCC method {(5) substrate
1-[2-chloro-4-nitrophenyl(G1)]--D-maltotrioside
(CNP-G3)} and the test kits -Amylase II-A®
{(6); substrate,
6-benzyl(G5)-1[4-nitrophenyl(G1)]--D-maltopentaoside;
auxiliary enzymes, -glucosidase and glucoamylase; Wako} and
-Amylase EPS® [(7); substrate,
EPS-G7; auxiliary enzyme, -glucosidase from yeast; Boehringer].
4-Nitrophenol was crystallized three times from water to meet the
purity criteria given by Bowers et al. (8). We obtained
EPS-G7 and bacterial -glucosidase (Toyobo Co., Osaka, Japan) as
generous gifts from Boehringer (Mannheim, Germany). Reagents for
zwitterionic buffers came from Boehringer and Serva. All other
chemicals were of analytical reagent-grade quality and supplied by
Merck.
procedures
Experiments were done in triplicate at 37.0 ± 0.05 °C,
and all absorbance data were collected at 405 nm. Concentrations always
refer to assay conditions, if not otherwise indicated. We calculated
molar absorption coefficients by measuring solutions of 100 µmol/L
4-nitrophenol and used pH-specific absorbances to correct all data from
photometric readings. Kinetic constants of pancreatic and salivary
amylase were determined using Eadie–Hofstee linear transformation
plots. Catalytic concentrations were measured with two enzyme
concentrations, representing single and fivefold activity, except in
multivariate examinations, which we conducted as two central composite
three-factor and five-level response-surface experiments (pH 6.3, 6.6,
7.1, 7.6, and 7.9; chloride, 5, 33, 50, 77, and 95 mmol/L;
substrate, 0.5, 1.8, 3.7, 5.6, and 6.9, and 0.5, 2.9, 6.5, 10.1, and
12.5 mmol/L), each with 16 sets of solutions (9). Table 1
presents the procedure together with the assay conditions and
calculation of results for the proposed method.
We derived the reference limits from central 0.95 interfractile
intervals (10) and estimated sex- and age-related
differences by nonparametric tests (Mann–Whitney U-test and
Kruskal–Wallis test, respectively). Following previous recommendations
on quality control (11) in method comparisons, we accepted
only means of duplicates with CVs 
5% for statistical computations by
ordinary linear regression because all correlation coefficients were
0.975 (11)(12).
|
Results
|
|---|
absorptivity of 4-nitrophenol
In contrast to other observations (13), the effect of
zwitterionic buffers with a pK between 6.6 and 7.4
(37 °C) on molar absorptivity at pH 7.15 was identical within the
limits of experimental error, showing a reduction of ~3% in
phosphate (Table 2
). Because chromophore ionization increases with temperature and
decreases by addition of salts and proteins, the molar absorption
coefficient at 37 °C was ~8% higher than at 30 °C, so that
405 (m2/mol) was
calculated to 1200.5 ± 14 (mean ± SD; n = 8) for
protein-free samples and to 1120 ± 15 for sera (with 67 g/L
protein) under the conditions of the assay.
View this table:
[in this window]
[in a new window]
|
Table 2. Effect of buffers on molar absorptivity of 4-nitrophenol
and pancreatic -amylase activity at pH 7.15 and
37 °C.1
|
|
assay conditions
Buffer and pH.
Apart from
N-(2-acetimido)-2-iminodiacetic acid complexing calcium and
bis(2-hydroxyethyl)amino-tris(hydroxymethyl)methane, which
inhibits the indicator enzyme (like all amino-type buffers), equal
reaction rates of pancreatic amylase were measured in zwitterionic
buffers and collidine; however, the rates were 10% lower in phosphate
(Table 2
). We chose HEPES, which is available worldwide in the highest
quality, but 2-(N-morpholino)propanesulfonic acid or
N,N'-bis(2-hydroxyethyl)-2-aminoethanesulfonic acid
may be used as well. The asymmetric pH profiles of Fig. 1
show activity optima at pH 6.9–7.0 for human pancreatic
amylase (HPA) and at pH 6.7–6.8 for human salivary amylase (HSA).
Therefore, at pH 7.15, the pancreatic isoenzyme retained 98% of its
maximal activity, whereas the activity was only 96% at pH 6.75.
Because at pH values above the optimum, the increasing absorptivity of
4-nitrophenol compensates for the loss of activity, we selected pH 7.15
to make the assay more robust against inevitable changes of pH rather
than pH 6.9–7.0, the activity maximum. Moreover, the influence
of HSA at pH 7.15 was reduced.
Substrate and effector concentrations.
Substrate and chloride
dependency exactly followed Michaelis-Menten kinetics, which allowed
the calculation (n = 5; mean ± SD) of the following
Km values: substrate, 0.117 ±
0.07 mmol/L (HPA) and 0.182 ± 0.09 mmol/L (HSA); chloride,
4.0 ± 0.4 mmol/L (HPA) and 6.8 ± 0.6 mmol/L (HSA). At 3.5
mmol/L EPS-G7 and 72 mmol/L chloride, 95% of the maximal velocity was
attained for HPA and ~92% was attained for HSA. In brief, as with
the pH value above the maximum, the pancreatic isoenzyme is slightly
favored by the selected conditions. Calcium is not essential, but its
addition protects amylase against complexing agents, e.g., EDTA or
citrate in specimens; it may be replaced by magnesium in the reagent as
well.
Alhough the isoactivity contour plots calculated from the narrower
substrate range (0.5–6.9 mmol/L; see Materials and Methods)
did not always represent maximal activity (Fig. 2
), we preferred them to EPS-G7 concentrations covering 0.5–12.5
mmol/L because of the more precise localization of the substrate
plateaus. These multivariate experiments showed the expected strong
influence of substrate concentration on reaction velocity and largely
confirmed the selected conditions, in particular the choice of pH 7.15.
On the other hand, Pareto charts (not shown) did not reveal strong
interactions of pH with substrate or chloride, accounting for a
difference in optimum pH between uni- and multivariate
investigation.
View larger version (37K):
[in this window]
[in a new window]
|
Figure 2. Two-variable contour plots showing the fractional activity
of HPA (I) and HSA (II) activity at
37 °C and 7.1 kU/L -glucosidase, under conditions representing
the center points of the factorial design.
(I-A and II-A), pH 7.1 (range, 6.3–7.9);
(I-B and II-B), 3.7 mmol/L EPS-G7 (range,
0.5–6.9 mmol/L); (I-C and II-C), 50
mmol/L chloride (range, 5–95 mmol/L).
|
|
Lag phase and reaction products.
As known from the reference
method (3), between 25 and 40 °C, the temperature
coefficient of Q10 of the auxiliary enzyme (2.12) clearly
excels the respective temperature coefficients of pancreatic (1.63) and
salivary (1.61) amylase. Thus, -glucosidase activities 6.2 kU/L
(37 °C) always kept the lag phase below 120 s. Higher
concentrations up to 10 kU/L did not significantly reduce this time. We
selected 7.1 kU/L only to prevent lag time extensions, which occurred
when the storage temperature of the reagent was inadequate. Under these
conditions, 5 kU/L HPA degraded 250 nmol of substrate (corresponding to
0.25 mmol/L) quantitatively to 4-nitrophenol within 120 s.
analytical variables and method comparison
Reagent stability, range of linearity, and
interferences.
From repeated measurements of pooled sera with
enzyme activities of 49.3, 118, and 398 U/L, and allowing ± 3%
deviation from activity at day 1, the storage at 5 °C without
additives was limited to 3 weeks for the substrate and 5 weeks for the
enzyme solution. Because of the shorter lag and measurement times, the
linearity of results by far exceeded that of the reference method
(3): sera could be assayed without dilution up to a
A405 of 0.540/min (1500 U/L or
15-fold higher than the upper reference limit).
Using the cited 100% ± 3% recovery criterion and the above-mentioned
serum pool, we observed no interference by 100 mmol/L glucose, 30
mmol/L triacylglycerols, 1 mmol/L ascorbic acid, 1 mmol/L sodium
heparinate, 610 µmol/L bilirubin, and 2.95 g/L hemoglobin.
Precision, sensitivity, and detection limit.
Table 3
summarizes the results of manual performance in precision
testing with pooled sera of normal, borderline high, and
above-normal catalytic concentration. We also used these sera to
compare the sensitivity of some current -amylase tests, as presented
in Table 4
. The data, obtained by transforming the photometric signals of
different sample volume fractions to the same ratio, reflect the
increasing reaction velocity with temperature, the higher absorbance of
2-chloro-4-nitrophenol at the test pH, and the complete release of
4-nitrophenol by bacterial -glucosidase compared with the
established EPS-G7 test, which uses glucosidase from yeast.
Additionally, the small SDs for the relative sensitivity, calculated
from three different activities, demonstrate the wide range of
linearity in all methods compared.
The lower limit of detection was 1.43 U/L (calculated from the mean of
10 reagent blanks + 3 SD), and the resulting limit of quantification
(at a tolerated CV of 10%) was 4.8 U/L. Both thresholds were
~1.4-fold higher than the respective values of the reference method
(3), which roughly corresponds to the ratio of their
catalytic concentrations in comparison studies.
The relationships between the proposed method and four other assays
over a wide range of catalytic concentrations are shown in Table 5
. Similarly close correlations were found only if specimens with
-amylase activities within the reference range were used, as could
be expected from the small SDs in the sensitivity studies (Table 4
). In
consequence, it seems important to note that the conversion of values
between these methods is feasible with only a small degree of
uncertainty.
preliminary reference values
Reference values for males did not significantly deviate from
those for females, nor were there any significant differences among the
age groups [P 0.90 (ß) in all two-tailed tests].
Therefore, from the total of 80 healthy adults, a preliminary 0.95
reference interval of 33.9–96.2 U/L (median, 56.6 U/L) was
established.
|
Discussion
|
|---|
All continuously monitoring -amylase assays that release a
chromophore have excellent linearity, wide dynamic ranges extending
across three decades, high sensitivity and precision, negligible
interference by matrix constituents, sufficient stability of reagents,
and ease of handling. The direct test using CNP-G3 (14) has
undoubtedly improved analytical standards and equals those of the
presented method. However, only EPS-G7 together with bacterial
-glucosidase provides an unambiguous reaction stoichiometry that is
unaffected by the change of external conditions. Moreover, in contrast
to CNP-G3 (15), the proposed procedure allows a specific
determination of pancreatic isoenzyme by optimal action of antibodies
against the salivary isoenzyme during the preincubation period
(16).
Proceeding from these advantages of the reference method
(3), the routine measurement procedure decreases the
necessary sample volume, the time of measurement, most interferences by
matrix constituents, and by 30%, the cost of the test. Likewise, the
altered reaction conditions, e.g., performing the assay at 37 °C
rather than 30 °C, increase the sensitivity of the test and
extend the range of linear response. These advantages, also tested
under automated conditions, make the assay more robust and economic
without any loss of analytic qualities, as evidenced by the results of
our evaluation studies.
In conclusion, the proposed method seems suitable for reliable
-amylase determinations in the clinical laboratory.
|
Acknowledgments
|
|---|
I thank Ragnhild Albrecht, Barbara Gütschow, and Sandra Rohlf
for excellent technical assistance, and Drs. Berding and Herb (Roche
Diagnostics GmbH) for valuable help in performing response-surface
experiments.
|
Footnotes
|
|---|
1 Nonstandard abbreviations: EPS-G7, 4,6-ethylidene(G7)-1[4-nitrophenyl(G1)]-1,4--D-maltoheptaoside; CNP-G3, 1-[2-chloro-4-nitrophenyl(G1)]--D-maltotrioside; HPA, human pancreatic amylase; and HSA, human salivary amylase. 
|
References
|
|---|
-
Tietz NW. Support of the diagnosis of pancreatitis by enzyme tests—old problems, new techniques. Clin Chim Acta 1997;257:85-98.
[ISI][Medline]
[Order article via Infotrieve]
-
Foo AY. Amylase measurement—which method?. Ann Clin Biochem 1995;32:239-243.
-
Lorentz K. Approved recommendation on IFCC methods for the measurement of catalytic concentration of enzymes. Part 9. IFCC method for -amylase (1,4--D-glucan glucanohydrolase, EC 3.2. 1.1). Clin Chem Lab Med 1998;36:185-203.
[ISI][Medline]
[Order article via Infotrieve]
-
Lorentz K. Properties of human -amylases from urine, pancreas, and saliva. Enzyme 1982;28:233-241.
[ISI][Medline]
[Order article via Infotrieve]
-
Lorentz K. IFCC methods for the measurement of catalytic concentration of enzymes. Part 9. IFCC method for -amylase [1,4--D-glucan glucanohydrolase, EC 3.2. 1.1]. Clin Chim Acta 1999;281:S5-S39.
[ISI][Medline]
[Order article via Infotrieve]
-
Satomura S, Sakata Y, Omichi K, Ikenaka T. -Amylase assay with use of a benzyl derivative of p-nitrophenyl -maltopentaoside, BG5P. Clin Chim Acta 1988;174:315-324.
[ISI][Medline]
[Order article via Infotrieve]
-
Rauscher E, von Bülow S, Hägele EO, Neumann U, Schaich E. Ethylidene protected substrate for the assay of human -amylase. Fresenius Z Anal Chem 1986;324:304-305.
-
Bowers GN, JR McComb, RB Christensen, RG Schaffer R. High-purity 4-nitrophenol: purification, characterization, and specifications for use as a spectrophotometric reference material. Clin Chem 1980;26:724-729.
[Abstract/Free Full Text]
-
Rautela GS, Snee RD, Miller WK. Response-surface co-optimization of reaction conditions in clinical chemical methods. Clin Chem 1979;25:1954-1964.
[Abstract/Free Full Text]
-
Dybkaer R, Solberg HE. Approved recommendation (1987) on the theory of reference values. Part 6. Presentation of observed values related to reference values. J Clin Chem Clin Biochem 1987;25:657-662.
-
Stöckl D, Dewitte K, Thienpont M. Validity of linear regression in method comparison studies: is it limited by the statistical model or the quality of the analytical input data?. Clin Chem 1998;44:2340-2346.
[Abstract/Free Full Text]
-
. National Committee for Clinical Laboratory Standards. Method comparison and bias estimation using patients’ samples; tentative guideline 1995:1-36 NCCLS NCCLS Publication EP9-A. Villanova, PA. .
-
Dupuy G, Hilaire G, Aubry C. Rapid determination of -amylase activity by use of a new chromogenic substrate. Clin Chem 1987;33:524-528.
[Abstract/Free Full Text]
-
Lorentz K, Gütschow B, Renner F. Evaluation of a direct -amylase assay using 2-chloro-4-nitrophenyl--D-maltotrioside. Clin Chem Lab Med 1999;37:1053-1062.
[ISI][Medline]
[Order article via Infotrieve]
-
Gella F-J, Gubern G, Vidal R, Canalias F. Determination of total and pancreatic -amylase in human serum with 2-chloro-4-nitrophenyl--D-maltotrioside as substrate. Clin Chim Acta 1997;259:147-160.
[ISI][Medline]
[Order article via Infotrieve]
-
. Roche Diagnostics Inc. P-Amyl, instructions for use 1999 Roche Diagnostics Indianapolis, IN. .
The following articles in journals at HighWire Press have cited this article:
|
|
|
|
 
M. H. Rivera, A. Lopez-Munguia, X. Soberon, and G. Saab-Rincon
{alpha}-Amylase from Bacillus licheniformis mutants near to the catalytic site: effects on hydrolytic and transglycosylation activity
Protein Eng. Des. Sel.,
July 1, 2003;
16(7):
505 - 514.
[Abstract]
[Full Text]
[PDF]
|
|
|
|
|
|
|
R. Siebers
Data Inconsistencies in Abstracts of Articles in Clinical Chemistry
Clin. Chem.,
January 1, 2001;
47(1):
149 - 149.
[Full Text]
[PDF]
|
|
|
Top
Abstract
Introduction
) and HPA (•) activity with pH.