| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Department of Clinical Biochemistry and Molecular Pathology, and
2
Cell Biophysics Research Group of the Hungarian Academy of Sciences, Medical and Health Science Center, University of Debrecen, Debrecen H-4012, Hungary.
3
Department of Medical Chemistry, Medical Center,
University of Szeged, Szeged H-6720, Hungary.
a Address correspondence to this author at: Department of Clinical Biochemistry and Molecular Pathology, Medical and Health Science Center, University of Debrecen, PO Box 40, Debrecen H-4012, Hungary. Fax 36-52-417-631; e-mail muszbek{at}jaguar.dote.hu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Methods: In the assay, FXIII was activated by thrombin and Ca2+. Fibrin polymerization was prevented by an inhibitory tetrapeptide. Glycine-ethyl ester and a glutamine residue of a synthetic dodecapeptide served as acyl acceptor and acyl donor transglutaminase substrates, respectively. The amount of ammonia released during the reaction was monitored using glutamate dehydrogenase and NADPH.
Results: The use of a new glutamine substrate and optimization of activator and substrate concentrations increased sensitivity. Substitution of NADPH for NADH and introduction of an appropriate blank eliminated systemic overestimation of FXIII activity. The recovery of FXIII was 96%, the assay was linear up to 470 U/L, the detection limit was 1 U/L, and the imprecision (CV) was <8% even at very low FXIII activities. A reference interval of 108–224 U/L (69–143%) was established. The results correlated well with results obtained by an immunoassay specific for plasma FXIII.
Conclusions: The optimized FXIII assay is a simple, rapid method for the diagnosis of inherited or acquired FXIII deficiencies and increased FXIII concentrations. It can be easily adapted to clinical chemistry analyzers.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-glutamyltransferase; EC 2.3.2.13)
catalyzes an acyl transfer reaction in which the carboxamide group of a
peptide-bound glutamine residue is the acyl donor and an appropriate
primary amine is the acyl acceptor (2)(3)(4). In the first step
of the modified double-displacement reaction, the glutamine residue
forms a thioester with the active site cysteine and ammonia is
released. In the second step, the acyl group is transferred to the acyl
acceptor amine and the amine is linked to the glutamine residue via a
peptide (isopeptide) bond. When the acyl acceptor primary amine is
provided by a peptide-bound lysine residue, the end result of the
transglutaminase action is the cross-linking of two polypeptide chains.
The main physiological functions of FXIII are the cross-linking of
fibrin and the covalent incorporation of
2-plasmin inhibitor
(2-PI), the main physiological inhibitor of
the fibrinolytic enzyme, plasmin, into fibrin polymers (4).
This way FXIII mechanically stabilizes fibrin and protects it from
immediate elimination by the powerful fibrinolytic system. Fibrin
cross-linking occurs through dimerization of chains and the
formation of highly cross-linked -chain polymers.
2-PI provides a single glutamine residue, the
penultimate N-terminal Gln, to the cross-linking reaction and becomes
attached to lysine residues of fibrin chains.
FXIII is essential for maintaining hemostasis (5)(6). Patients with inherited FXIII deficiencies exhibit severe bleeding diathesis and, in most cases, require life-long supplementation therapy. In addition, FXIII is also involved in maintaining pregnancy and in wound healing. Acquired FXIII deficiency may occur in several diseases, including inflammatory bowel diseases and acute leukemia. Increased plasma FXIII activity has been reported in patients with obliterative atherosclerosis (7) and diabetic angiopathy (8), and in chronic leukemia patients with increased megakaryocytic activity (9). Recently, a polymorphism of the FXIII A subunit, a Val-to-Leu transformation at position 34 in the activation peptide (10), has been associated with a protective effect against occlusive vascular diseases (11)(12)(13).
A kinetic FXIII assay based on the continuous monitoring of ammonia released during the transglutaminase reaction was developed in our laboratory (14) and later modified by Fickenscher et al. (15). In this method, ammonia released during the transglutaminase reaction is measured by a glutamate dehydrogenase (GluDH) indicator reaction. A diagnostic method (Berichrom® FXIII) based on this assay has been manufactured and commercialized by Dade Behring. Although the kinetic assay is rapid and easy to adapt to clinical chemistry analyzers, its restricted linearity (16), relatively low sensitivity, and frequent overestimation of FXIII concentrations in patients with severe FXIII deficiency (unpublished observation) prompted us to improve and optimize the assay.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
2-PI
[2-PI(1-12)] and the fibrin polymerization
inhibitor tetrapeptide, Gly-Pro-Arg-Pro-amide (GPRP) (19)
were prepared by the solid-phase method of Merrifield (20)
with the use of Boc protective groups on a peptide synthesizer (AB
431A; PE Biosystems). The final deprotection was achieved by
hydrofluoric acid treatment. The crude product was purified by
preparative HPLC (Shimadzu) and lyophilized. The purities of both
purified peptides were estimated to be >99% by analytical HPLC and
mass spectrometry (Finnigan).
subjects
All procedures were performed in accordance with the Helsinki
Declaration of 1975, as revised in 1996. The reference group of 141
apparently healthy volunteers had a mean age of 29.9 ± 9.4 years
(range, 21–65 years). The group consisted of 61 males (mean age,
30.4 ± 10.1 years) and 80 females (mean age, 29.4 ± 8.9
years) from the same geographical region. FXIII determinations were
also carried out on samples obtained from 200 hospitalized patients,
among them 4 patients with inherited FXIII deficiency. Ethical approval
was obtained from the Ethics Committee of the Medical and Health
Science Center, University of Debrecen.
plasma samples
Blood samples were collected after overnight fasting; 9 volumes of
blood were drawn into Vacutainer Tubes (Becton Dickinson) containing 1
volume of 0.105 mol/L buffered sodium citrate. Platelet-poor plasma was
obtained by centrifugation (2500g for 20 min at 20 °C).
Plasma was removed and, if not analyzed immediately, stored at
-70 °C. To monitor the stability of FXIII in plasma, five plasma
samples were tested after storage at room temperature (22 °C), and
at 4, -20, and -70 °C for up to 6 months. For linearity
determinations, Standard Human Plasma (Dade Behring) was supplemented
with a 1/50th volume of concentrated purified plasma FXIII preparation
of known specific activity. The FXIII activities (U/L) of the
nonsupplemented plasma, the supplemented plasma, and its
dilutions were determined. Dilutions were made with physiological
saline or with FXIII-deficient plasma with no detectable FXIII
activity. In the linear range, the measured activities of the dilutions
of the supplemented plasma corresponded to the activities calculated
from the assayed values and respective volumes of Standard Human Plasma
and purified FXIII. The assigned FXIII% value of Standard Human Plasma
(100%) was used to calculate the FXIII% value of the supplemented
plasma. For interference studies, five human plasmas from healthy
subjects were supplemented with various concentrations of bilirubin (up
to 200 µmol/L) or triglycerides (up to 10 mmol/L).
reagent composition and assay protocol
FXIII activity was measured on a COBAS MIRA Plus analyzer (Roche)
at 37 °C. Routinely, 25 µL of undiluted citrated plasma was mixed
with 250 µL of reagent of the following composition: 20 kU/L
thrombin, 10 mmol/L CaCl2, 5 mg/L polybrene, 2
mmol/L fibrin polymerization inhibitory peptide (GPRP), 0.1 mmol/L
dithiothreitol, 4.4 mmol/L 2-PI(1-12) peptide,
5 mmol/L glycine-ethyl ester (Gly-O-Et), 0.35 mmol/L NADPH, 20 kU/L
GluDH, 0.6 mmol/L ADP, 7 mmol/L -ketoglutarate, and 5.4 g/L bovine
serum albumin in 60 mmol/L HEPES buffer, pH 7.7. The concentrations of
the above components in the final reaction mixture were somewhat less,
91% of reagent concentrations. ADP was required to obtain optimal
stimulation of GluDH above pH 7.0 and to prevent its inhibition by
excess substrates and by inhibitory plasma constituents. Blanks
routinely contained 1 mmol/L iodoacetamide, a transglutaminase
inhibitor. In addition, 20 mmol/L EDTA and 40 kilounits/L
hirudin, inhibitors of FXIII activation, were also used in plasma
blanks.
The absorbance was measured at 340 nm every 25 s up to 12.5 min. The linear interval of the reaction was determined using the KINSEARCH program of the analyzer, and the change in absorbance per minute was calculated according to the manufacturer’s instructions (21). Similar results were obtained when linear regression analysis was performed on measurement points between 5 and 10 min and the slope of the regression line was used for calculations. The values obtained for the blanks were subtracted, and the results were expressed as either U/L or as a percentage of the average normal (the mean of the reference group) plasma FXIII activity.
In experiments to optimize individual assay components,
2-PI(1-12), Gly-O-Et, thrombin,
CaCl2, and GluDH were used in various
concentrations, as indicated below. For determination of the optimal
thrombin concentration, thrombin was blocked at the end of the 5-min
lag period by the addition of 40 kilounits/L hirudin. In experiments to
determine the KM(app) for
2-PI(1-12) and Gly-O-Et, highly purified
plasma FXIII (25 mg/L in 100 mmol/L HEPES buffer, pH 7.7) was
substituted for plasma in the assay mixture. In experiments testing the
effect of ammonia present in the plasma or in reagent components,
NH4Cl solution (up to 400 µmol/L in HEPES
buffer, pH 7.7) was substituted for plasma.
other methods
A new, highly sensitive sandwich ELISA was used to determine the
concentration of FXIII in plasma (22). In the ELISA,
biotinylated monoclonal capture antibody against FXIII-B and
peroxidase-labeled monoclonal detection antibody against FXIII-A
were incubated with plasma dilutions in the wells of a
streptavidin-coated microplate. The detection antibody-FXIII-capture
antibody complex attached to the streptavidin-coated microplate was
quantified by the measurement of peroxidase activity. Only the
tetrameric plasma FXIII reacted in the assay; noncomplexed A or B
subunits did not react. In some experiments, FXIII activity was also
determined by the Berichrom F XIII method. In this case, the
instructions of the manufacturer were followed meticulously. The
FXIII-A Val34Leu polymorphism was identified by a PCR-restriction
endonuclease method based on the amplification-created restriction site
principle (23).
statistical methods
The calibration curves for FXIII activity measurement methods were
analyzed by weighted linear regression, using SPSS 7 for Windows
software. The 1/SD2 values of duplicate
measurements of activities were used as statistical weights. The
Kolmogorov–Smirnov test was used to compare the distribution of plasma
FXIII activities in healthy individuals with the expected
gaussian distribution (STATISTICA software; StatSoft). A
reference interval was established according to the NCCLS guideline
(24). The test proposed by Dixon was used for the detection
of potential outliers (25)(26). The possibility
that separate reference intervals would be required for male and female
reference individuals was tested using the method of Harris and Boyd
(24)(27). The correlation between two methods
was analyzed by Deming regression (28) using CBstat, a
program written and kindly provided by Dr. K. Linnet. The program
yields the straight line parameters with standard errors and the
correlation coefficient, and checks for deviation from linearity. Error
calculations were based on the jackknife principle (29).
Outliers deviating from the fitted line by more than 4 SD of the
residuals were identified and excluded from the fit. Deviation of the
intercept from zero was checked for significance with the Student
t-test. Normalized values were also analyzed with a
Bland-Altman plot (30).
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Activation of FXIII:
 |
 |
 |
Indicator reaction:
 |
 |
Full activation of FXIII by thrombin and Ca2+ occurs during the 5-min lag phase of the reaction. Fibrin polymerization is prevented by the tetrapeptide GPRP. It was also shown in separate experiments that under assay conditions, endogenous or contaminating ammonia up to 400 µmol/L, i.e., up to 10-fold higher than the upper limit of the reference interval for ammonia in plasma, was consumed within 3 min. This was manifested by a sudden drop in the absorbance at 340 nm, which is used to measure the transformation of NADPH to NADP+. In the absence of FXIII, no further change of absorbance occurred after this period.
The formed FXIIIa then cross-links the amine substrate Gly-O-Et to the
substrate glutamine residue of 2-PI(1-12), and
ammonia is released. In the indicator reaction, ammonia is utilized for
glutamate formation, and NADPH, the cofactor in the GluDH reaction, is
transformed to NADP+. NADPH consumption, measured
by the decrease in the absorbance at 340 nm, is directly proportional
to FXIII activity within a wide time window. Measurement of the change
in absorbance between 5 and 10 min is recommended (Fig. 1
).
|
optimization of assay components
The optimal concentrations of activators were established. The
dependence on Ca2+ concentration followed a
bell-shaped curve characteristic of transglutaminases. The maximum was
at 10 mmol/L CaCl2 in the reagent, which
corresponded to 9 mmol/L CaCl2 in the final assay
mixture (Fig. 2A
). When optimizing the thrombin concentration, we paid special
attention to the FXIII-A Val34Leu polymorphism. Recently, highly
purified FXIII from different Val34Leu genotypes was used to
demonstrate that thrombin cleaves the mutant Leu34 allele significantly
faster than the wild-type Val34 allele (23)(31).
However, at full activation, the specific transglutaminase activity of
FXIIIa was the same for the three FXIII Val34Leu genotypes
(23). Fig. 2B
demonstrates that at lower thrombin
concentrations, the activation of FXIII in the plasma of wild-type
individuals occurs to a lesser extent than in individuals homozygous
for the Leu34 variant. The extent of FXIII activation in the plasma of
heterozygous (Val/Leu) individuals was intermediate. At 5 kU/L
thrombin, the difference was still detectable; however, when the
reagent contained 20 kU/L thrombin, full activation of FXIII could be
achieved within 5 min in the plasma of individuals of all three
genotypes. In the reagent for the routine assay, 10 mmol/L
CaCl2 and 20 kU/L thrombin were used.
|
The saturation curve for the acyl donor
2-PI(1-12) substrate and its Lineweaver-Burk
plot are shown in Fig. 3A
. The slight deviation of the highest concentration point from
the regression line and the shape of the Michaelis-Menten plot suggest
that the saturation curve is not fully hyperbolic. From three parallel
measurements, a KM(app) of 5.3 x
10-4 mol/L and a
kcat of 389 s-1
were calculated. The specificity constant,
kcat/KM(app),
was 7.34 x 105
L · mol-1 · s-1.
|
The use of 5 mmol/L Gly-O-Et as the amine substrate was
recommended in a previously described assay (15). FXIIIa can
deamidate peptide-bound substrate Gln residues in the absence of an
amine substrate (1)(14), which explains the
substantial ammonia release at a Gly-O-Et concentration of 0 mmol/L, as
shown in Fig. 3B
. The presence of an amine substrate, in this case
Gly-O-Et, produces a concentration-dependent increase of FXIIIa
activity manifested by increased ammonia production. For the above
reason, no Lineweaver-Burk plot was derived from the data, but the
original saturation curve was fitted to a hyperbolic function plus a
constant to take the residual activity into account. The fit shown in
Fig. 3B
yielded a KM(app) of 4.5
x 10-4 mol/L, and 5 mmol/L Gly-O-Et was indeed
shown to be a saturating concentration. In the assay, 4 mmol/L
2-PI(1-12) and 5 mmol/L Gly-O-Et were used
routinely.
plasma blank
We attempted to obtain an appropriate plasma blank by preventing
the activation of FXIII or by blocking the transglutaminase activity.
These goals were achieved by inhibiting thrombin with hirudin, by
chelating Ca2+, or by adding the transglutaminase
inhibitor and –SH reactant, iodoacetamide. In all three cases,
identical results were obtained; therefore, only the results with
iodoacetamide are presented. We analyzed 341 samples from healthy
individuals and patients for FXIII activity in the presence of
iodoacetamide and found a FXIII-independent decrease of absorbance
corresponding to a FXIII activity of 7.93 ± 3.31 U/L (range,
2.6–24.5 U/L). With 20 mmol/L EDTA or 40 kilounits/L hirudin, values
for plasma blanks were 7.37 ± 2.76 U/L and 7.01 ± 2.02 U/L,
respectively. The combination of these inhibitors of FXIII activation
with each other or with iodoacetamide had no further influence on
values obtained for plasma blanks.
When we did not subtract the values obtained for the blanks,
FXIII activities measured with our assay were overestimated by 5%. In
similar experiments using the Berichrom assay without an appropriate
blank, we observed an 8% overestimation of FXIII activity. Table 1
demonstrates FXIII activity without and with correction using
blanks in four FXIII-deficient patients (samples 1–4) and in two
selected inpatients (samples 5 and 6) who had FXIII activity within the
reference interval. FXIII activity <1% of the average normal (mean of
the reference group) is considered a very severe hemorrhagic diathesis,
whereas FXIII activity of 5–10% in most cases produces only a mild,
if any, bleeding tendency. For both assays, it is clear that clinically
misleading overestimation of FXIII activity in FXIII-deficient patients
would occur if the values obtained for the blanks were not
subtracted. Without correction for the plasma blank, the
Berichrom assay overestimated the FXIII activity in samples obtained
from patients 5 and 6, who had FXIII activities within the reference
interval, by as much as 27% and 22%. The overestimations with our
assay were lower but still significant (14% and 16%).
|
recovery and interference
To investigate the recovery of FXIII in the assay, plasma samples
were supplemented with highly purified FXIII of known activity.
Comparison of the activities of supplemented and nonsupplemented plasma
samples revealed that the recovery of FXIII activity in the assay was
96%.
High plasma ammonia concentrations (>200 µmol/L) led to considerable consumption of NADPH during the lag phase of the reaction, and in this case, we recommend that the sample be diluted and remeasured. Bilirubin up to 200 µmol/L did not interfere with the assay (bias <2%). The addition of 200 µmol/L bilirubin to the plasma produced only a minimal increase in the absorbance at 340 nm (70 milliabsorbance units in the 0.6-cm cuvette of the COBAS MIRA Plus). Bias was negligible up to 7.5 mmol/L triglycerides (<1%). At this concentration, hypertriglyceridemia produced an increase of 900 milliabsorbance units (340 nm, 0.6-cm pathlength) in the sample. At 10 mmol/L triglycerides, a 9.9% negative bias was observed.
linearity, analytical sensitivity, and detection limit
Direct dilutions of FXIII-supplemented plasma in the range of
0–300% of the average normal were used for the determination of
linear range. FXIII activity was measured by the modified assay
described above and, for comparison, by the Berichrom assay (Fig. 4
). Because dilutions with physiological saline or
FXIII-deficient plasma gave similar results, only data obtained with
physiological saline are shown in Fig. 4
. With our optimized assay,
FXIII activity was linear up to 300% (468 U/L; slope ± SD,
1.54 ± 0.03; intercept, 6.20 ± 2.24 U/L;
Sy|x = 6.83; r = 0.999). In
accordance with a previous report (15), the Berichrom assay
was nonlinear above a FXIII activity of 150%; even the point
corresponding to 150% activity was slightly below the regression line.
Among 200 subsequent clinical samples received for FXIII determination
in our laboratory, the activities of 4 (2.1%) were >150% (234 U/L),
but none exceeded 300% (468 U/L).
|
The slopes of the calibration curve for the modified photometric assay and for the Berichrom assay were 1.54 and 0.99, respectively. These results demonstrate that optimization of the assay and the use of new substrate produced a 1.5-fold increase in sensitivity. The detection limit of the optimized photometric assay was estimated to be 1 U/L.
precision and carryover
The within-run and day-to-day imprecision (CV) was
determined on reconstituted lyophilized Standard Human Plasma and
Control Plasma-P (Dade Behring; Table 2
). Because there are no commercially available lyophilized
plasma preparations with FXIII activity <30% of the average normal,
plasma samples were used for the determination of within-run
imprecision in the low FXIII activity range. At FXIII activities >10%
of the average normal, both the within-run and day-to-day CVs were
<5%, and at FXIII activities as low as 3% (5.2 U/L), the within-run
CV was still well below 10%. The carryover in the COBAS MIRA Plus
analyzer was <1%.
|
sample stability
The FXIII activities in five plasma samples were determined before
(mean, 161.3 U/L; range, 81.2–257.6 U/L) and after storage for various
intervals up to 6 months. The plasma samples could be stored at room
temperature for at least 24 h without loss of FXIII activity
(98.6% recovery), and after storage for 3 days, the recovery was still
95.6%. At 4 °C, FXIII remained stable for at least 3 days (98.5%
recovery). With samples stored at -20 and -70 °C, no change of
FXIII activity could be observed during the whole test period (98.5%
and 106.0% recovery after 6 months, respectively).
reference interval
The distribution of FXIII activity in the group of reference
individuals is shown in Fig. 5
. The distribution did not differ significantly from the ideal
gaussian distribution (d = 0.084; P = 0.272).
After log-normal transformation, however, the fit became better (d
= 0.048; P = 0.552). No outliers were detected, and a
reference interval of 108–224 U/L was established by the nonparametric
method. Because laboratory test results of blood coagulation factor
determinations are traditionally expressed as a percentage of the
average normal (mean of the reference group), the reference interval
was also calculated as a percentage. If we accepted the mean FXIII
activity of reference sample group (156.7 U/L) as 100%, the reference
interval was calculated to be 69–143%.
|
Mean FXIII activities of the male and female subgroups were 158.5 ± 32.5 and 155.3 ± 26.6 U/L, respectively. Partitioning of reference values according to gender was not justified by statistical tests. The z value (0.64) was well below the critical z value (2.30). The ratio of SD values for the two subgroups (1.22) was <1.5.
comparison with immunoassay
Because there is no reference method or generally accepted method
for measurement of FXIII activity, we selected a methodologically
well-established FXIII ELISA (22) for comparison. Deming
regression of FXIII activity vs FXIII concentration is shown in Fig. 6A
. The deviation of the intercept from zero was not significant
with the Student t-test (P >0.1). Deviation from
linearity also was not significant. The plot of residuals also
suggested a linear relationship between the results obtained with the
two methods. To bring concentration and activity measurements to a
common platform, we normalized the individual measured values for both
methods by expressing them as percentages of the mean of the reference
group [21.0 mg/L (for concentration) and 156.7 U/L (for activity)
corresponded to 100%]. The difference between the normalized activity
and concentration values vs the means of the normalized values are
shown in Fig. 6B
in a Bland-Altman plot. No bias could be detected.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
It turned out, however, that the modified assay still had problems. A
major problem was the frequent overestimation of FXIII activity in the
plasma of FXIII-deficient patients. Because relatively low
concentrations (5–10%) of FXIII are sufficient to maintain
hemostasis, such overestimation might have serious clinical
consequences. In addition, the monitoring of replacement therapy of
patients with severe FXIII deficiency requires accurate measurement of
FXIII activity in the low activity range. It was suspected that the
overestimation is the result of NADH-consuming reactions that are
independent of the formation of ammonia and that such an overestimation
is also present, but less evident, at higher activities. Indeed, using
appropriate blanks, we found that the Berichrom assay overestimated
FXIII activity by an average of 8%. The replacement of NADPH for NADH
in our optimized assay helped somewhat but did not eliminate the
problem. Data from four FXIII-deficient patients (Table 1
) demonstrated
that correction, using values obtained for blanks, is essential for
clinically relevant results. The need to correct using values obtained
for blanks was not restricted to the low activity range. In two
selected patients with FXIII activity within the reference interval,
the Berichrom assay overestimated FXIII activity by >20% when no
correction was used. The statement that the measurement of a blank is
not necessary with this assay (15) is hardly defensible.
To increase the sensitivity of the assay, a new glutamine substrate was
synthesized and the assay components were optimized. The sequence of
2-PI(1-12) dodecapeptide, NQEQVSPLTLLK,
corresponds to the N-terminal sequence of
2-PI, an excellent natural substrate of FXIIIa
(37). The Gln residue at position 2 serves as the acyl donor
transglutaminase substrate site (38)(39). When
we used this substrate and optimized the concentrations of assay
constituents, the sensitivity of the assay increased 1.5-fold and a
detection limit as low as 1 U/L could be achieved. These changes made
the measurement of low FXIII activities more reliable.
In light of the recent discovery that the Val34Leu polymorphism influences the rate at which activation peptide is released from FXIII-A and thus the thrombin sensitivity of FXIII activation (23)(31), optimization of the thrombin concentration was of particular importance. Thrombin should be used in a concentration sufficient to fully activate all FXIII variants during the lag phase of the reaction, otherwise the activity of the Val34 wild type, which is less attractive for thrombin, would be underestimated. We found that 20 kU/L thrombin was sufficient. In theory, the different activation rates of Val34Leu FXIII genotypes could be the basis for identifying these genotypes by non-molecular genetics methods. However, because the retention times of the Val34 and Leu34 FXIII activation peptides are different in reversed-phase HPLC (23), developing a method based on this principle (40) seemed to be more feasible.
The Berichrom assay has a restricted linearity with a linearity limit close to the upper limit of the reference interval (15)(16). By optimizing the method, we extended the linear range considerably, and very few pathological samples fall outside this extended range.
The reference interval of the optimized assay was determined according to NCCLS guidelines and was found to be 108–224 U/L. This corresponds to 69–143% of average normal FXIII activity. The interval obtained with the optimized assay corresponds well to the intervals (70–130%, 70–140%, and 73–155%) obtained in earlier studies using less rigorous criteria (14)(15)(16) and to the reference interval (67–133%) of the antigenic plasma FXIII assay (22). These results contradict the view that FXIII has a reference interval broader than the reference intervals for most of the clotting factors (41)(42).
In conclusion, we extensively modified the kinetic photometric FXIII assay by introducing a new synthetic glutamine substrate, by optimizing the concentration of the constituents of the assay system, by substituting NADPH for NADH, and by introducing a sample blank. These modifications improved sensitivity and markedly lowered the detection limit. The assay became more reliable, especially in the low FXIII activity range, and the linearity range was greatly extended. Good recovery of FXIII was obtained. The results correlated well with those obtained with an immunoassay specific for plasma FXIII. The reference interval for the assay was established. The assay was adapted to a clinical chemistry analyzer and performed well. It can be widely used for the diagnosis of inherited and acquired FXIII deficiencies and for measuring increased FXIII concentrations.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2-PI, 2-plasmin inhibitor; 2-PI(1-12), synthetic dodecapeptide corresponding to the N-terminal sequence of 2-PI; GluDH, glutamate dehydrogenase; and Gly-O-Et, glycine-ethyl ester. 
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-(-glutamyl) lysine bond formation. Adv Enzymol Relat Areas Mol Biol 1983;54:1-56.[ISI][Medline]
[Order article via Infotrieve]
2-plasmin inhibitor and fibronectin to fibrin by fibrin-stabilizing factor. Biochim Biophys Acta 1981;661:280-286.[Medline]
[Order article via Infotrieve]
2-plasmin inhibitor to fibrin catalyzed by activated fibrin-stabilising factor. J Biol Chem 1982;257:14767-14772.2-Plasmin inhibitor is a substrate for tissue transglutaminase: an in vitro study. Thromb Res 2000;99:399-406.[ISI][Medline]
[Order article via Infotrieve]
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  A. H. Shemirani, E. Szomjak, Z. Csiki, E. Katona, Z. Bereczky, and L. Muszbek Elevated factor XIII level and the risk of peripheral artery disease Haematologica, September 1, 2008; 93(9): 1430 - 1432. [Full Text] [PDF]
|
|
|
|
|
|
Z. Bereczky, E. Balogh, E. Katona, I. Czuriga, I. Edes, and L. Muszbek Elevated factor XIII level and the risk of myocardial infarction in women Haematologica, February 1, 2007; 92(2): 287 - 288. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-W. Wu and Y.-H. Tsai A Rapid Transglutaminase Assay for High-Throughput Screening Applications J Biomol Screen, October 1, 2006; 11(7): 836 - 843. [Abstract] [PDF]
|
|
|
|
 |
|
 R. Ponce, K. Armstrong, K. Andrews, J. Hensler, K. Waggie, J. Heffernan, T. Reynolds, and M. Rogge Safety of Recombinant Human Factor XIII in a Cynomolgus Monkey Model of Extracorporeal Blood Circulation Toxicol Pathol, October 1, 2005; 33(6): 702 - 710. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Mazooz, T. Mehlman, T.-S. Lai, C. S. Greenberg, M. W. Dewhirst, and M. Neeman Development of Magnetic Resonance Imaging Contrast Material for In vivo Mapping of Tissue Transglutaminase Activity Cancer Res., February 15, 2005; 65(4): 1369 - 1375. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK |
) and a FXIII-deficient
plasma sample (FXIII activity, 21 U/L; 
) are shown. Absorbance at
340 nm was measured every 25 s in a COBAS MIRA Plus analyzer. The
interval between the vertical dashed lines represents
the time window recommended for the calculation of FXIII activity.
), heterozygous (Val/Leu; 
), and homozygous (Leu/Leu;
10 kU/L, full activation of FXIII
was achieved in the plasma samples of all three genotypes. The results
are expressed as percentage of maximal activity.
