Clinical Chemistry 43: 2390-2396, 1997;
(Clinical Chemistry. 1997;43:2390-2396.)
© 1997 American Association for Clinical Chemistry, Inc.
Glycohemoglobin filter assay for doctors' offices based on boronic acid affinity principle
Frank Frantzen1,a,
Kjersti Grimsrud2,
Dag-Erik Heggli2,
Arne Ludvig Faaren1,
Trond Løvli2 and
Erling Sundrehagen2
1
Axis Nord, P.O. Box 6073, N-8018 Mørkved, Norway.
2
Axis Biochemicals ASA, P.O. Box 2123,
Grünerløkka, N-0505 Oslo, Norway.
a Author for correspondence. Fax 47-75516185; e-mail frank.frantzin{at}axisbio.no
 |
Abstract
|
|---|
We present a new filter assay for the determination of glycohemoglobin
as a unique application of the boronic acid affinity principle. With
the use of a water-soluble blue-colored boronic acid derivative and a
specific precipitation method for hemoglobin, total hemoglobin
including bound boronic acid is precipitated and collected on a filter
strip before quantification. Hemoglobin and boronic acid are quantified
by a dual-wavelength reflectometric measurement, and the result is
reported directly as percent glycohemoglobin. The test is simple,
quick, and designed as a doctors' office test for the monitoring and
management of diabetes. The imprecision of the assay is <4% over the
range 3–18% Hb A1c, and the method is linear up to at
least 20% Hb A1c. Comparisons with four well-established
glycohemoglobin methods yielded correlation coefficients ranging from
0.94 to 0.99, with slopes from 0.94 to 1.01.
|
Introduction
|
|---|
Boronic acids are known to form cyclic boronate esters with
1,2-cis-diol groups, and immobilized phenylboronic acids
have widely been used as affinity matrices for the separation of
diol-containing substances of biological interest (1)(2)(3).
In the clinical laboratory, this principle has been important in the
determination of glycohemoglobin
(GHb),1
first as a manual system with the use of affinity columns
(3) and later in automated systems such as the Primus
system (Gamidor, Abingdon, UK, affinity columns) and the Abbott Vision
(4) and Abbott IMx systems (5) (Abbott
Labs.). The last two are performed with 3-aminophenylboronic
acid-derivatized agarose beads and a polyanionic boronic acid
conjugate, respectively.
We have described the use of soluble, highly colored boronic acid
derivatives as probes for GHb (6). We now report the
further development of this assay principle and present a new
semihomogeneous assay for the determination of GHb. The assay is a
unique application of the boronic acid affinity methodology and is
designed as a doctors' office test performed with filter-strips and
reflectometric readings. Only the stable GHb fraction binds the
blue-colored boronic acid derivative used, so no preincubation or other
pretreatment steps are necessary to remove the labile GHb fraction.
Clinical data, the influence of different assay variables, and
correlation studies are presented to show the performance of the assay.
|
Materials and Methods
|
|---|
Apparatus.
All reflectance data were measured with a small
prototype reflectometer (Nycomed Pharma AS) measuring reflectance (%R)
at 620 and 470 nm. Measurements at these wavelengths were used to
quantitate the blue-colored boronic acid conjugate and hemoglobin (Hb),
respectively. The instrument automatically performed Kubelka–Munk
transformations (7) to linearize the recorded reflectance
data. The data were reported as K/S values according to Eq. 1
.
 | (1) |
Reagents.
ZnCl2, formamide,
NaN3, MgCl2, NaCl, and HEPES were purchased
from Sigma Chemical Co. Glycinamide-HCl was from Fluka and Triton
X-100 from Pierce Europe. The blue-colored boronic acid derivative used
in the assay, a xylene
cyanol-1,3-diaminopropanol-4-carboxyphenylboronic acid conjugate
(XC-DAPOL-CPBA), was synthesized at Axis Nord (Bodø, Norway). All
chemicals used were of analytical grade.
The GHb assay solution contained buffer salt, Zn(II) ions for the
precipitation of Hb, and the blue-dyed boronic acid conjugate to bind
the glycated residues of Hb: 50 mmol/L glycinamide, 10 mmol/L
ZnCl2, 23 mmol/L MgCl2, 200 mmol/L NaCl,
and 0.5 g/L NaN3 adjusted to pH 8.1. To this was added
XC-DAPOL-CPBA, formamide, and Triton X-100 at 0.19 mmol/L, 42 mL/L, and
1 mL/L, respectively.
Wash solution was 100 mmol/L HEPES and 0.5 g/L NaN3
adjusted to pH 8.1, with Triton X-100 added to 1 mL/L.
Assay principle.
The presented GHb assay utilizes the ability
of boronic acid to bind to the stable glycated residues of GHb because
of their specificity for cis-diols. A specially designed
nonimmobilized, water-soluble, highly colored boronic acid derivative
(XC-DAPOL-CPBA) with absorption maximum at 618 nm is used as a probe
for the glycated residues of Hb. Binding of this compound to a glycated
residue in Hb is shown in Fig. 1
. However, with the potential of also reacting with free
carbohydrates and other posttransitionally nonenzymatically glycated
proteins present in the sample, a separation system is necessary to
obtain specificity. This is achieved with the use of a specific
zinc(II)-precipitation method selectively precipitating Hb in the
sample (8). The whole assay is performed in a simple
two-step procedure as follows. Step 1: lysis of erythrocytes and
selective precipitation of both glycated and nonglycated Hb in the
presence of the colored boronic acid binding to the glycosyl moieties
of GHb. Step 2: precipitated glycated and nonglycated Hb isolated by
filtration, excess dye-boronic acid removed by washing, and the
percentage GHb (%GHb) quantitated from a dual-wavelength
reflectometric reading measuring both Hb (470 nm) and the colored
boronic acid (620 nm).
View larger version (13K):
[in this window]
[in a new window]
|
Figure 1. Binding of the soluble blue-colored boronic acid
derivative XC-DAPOL-CPBA to the glycated residues of Hb.
|
|
On the basis of the boronic acid affinity principle the presented assay
measures the same population of GHb species as other boronic affinity
methods do. Analytical advantages with this assay principle, i.e., low
susceptibility to variations in temperature (9)(10)(11) and pH
(12)(13), are maintained.
Assay procedure.
The following standard procedure was used in
all experiments: 3.5 µL of whole-blood sample was added to 200 µL
of GHb assay solution, vortex-mixed for 2–3 s, and left to stand for 3
min. The solution was mixed again, and 16 µL of solution containing
precipitated Hb was applied on the filter, followed by 16 µL of
washing solution. After drying for 3 min, the reflectometric signal
(%R) was measured at 470 nm and 620 nm. The K/S values reported from
the instrument provided the basis for the calculation of the glycation
of the sample (see Calculations and Calibration of the
method).
Filter device.
The filter device used was an in-house
prototype composed of an upper polypropylene strip (0.6 mm thick) with
wells for sample application (diameter: 4 mm). A 10-mm disc of glass
fiber filter (Gelman Sciences, A/E-filter, P/N:61635) was placed
underneath the well to collect the Hb precipitate, and an additional
glass fiber disc (17 mm) was used as a pad (Gelman ITLC-SA membrane,
P/N:51432). The upper and lower part of the device was spaced by 0.5 mm
with the use of polypropylene strips to keep the filters from being
compressed. Although this was the standard filter used, a continual
search for better filters has pointed out that more applicable
top-filters are found among asymmetric polysulfone membranes (0.2
µm/10 µm, MemTec Corp./Amicon). This latter type of filter was used
during the reproducibility studies.
Calculations.
The K/S values from the instrument were used to
calculate the K/S ratio (Eq. 2 ), which was the variable used to
establish the calibration curve for the %Hb A1c or %GHb
values.
 | (2) |
Calibration of the method.
Because of the lack of certified
standards, calibration of the method to read %Hb A1c
values (or %GHb) was performed on the basis of the correlation of the
K/S ratio against the Abbott IMx "standardized" %Hb
A1c values (or %GHb with the same method)
(5). Any Hb A1c method can in principle be
used for this calibration, but we chose the Abbott method, also
utilizing the boronic acid affinity principle, because this made it
possible to calibrate the present method to read both %GHb values and
"standardized" %Hb A1c values. Notably, only one
calibration curve needs to be established to calibrate the present
method, i.e., the relationship between the K/S ratio vs glycation of
Hb. The K/S ratio for a certain glycation is a constant value, not
varying according to the reflectance signal of Hb. Still, because of
the reading process itself, this is true only as long as the extremes
are avoided during the reflectance measurements, i.e., too high or too
low amounts of precipitated Hb present on the filter surface. The
zinc(II)-precipitation system applied in the assay (8) is
designed to precipitate enough Hb to avoid deviations during the
reading process. This is confirmed by the tolerance for variation in
blood sample hematocrit values (see Results and Discussion).
The relationship between the K/S ratio obtained with the presented
method and Abbott standardized %Hb A1c values was
established by analyzing 49 clinical samples with GHb values ranging
from 3.5% to 13% with both methods (Results and
Discussion, Fig. 3
). With the use of the regression equation
defining this relationship, the glycation of the sample was determined
by converting K/S ratios to %Hb A1c (or %GHb).
View larger version (13K):
[in this window]
[in a new window]
|
Figure 3. Calibration curve.
K/S ratio for a series of 49 clinical samples is plotted as a function
of standardized %Hb A1c values obtained by the Abbott
IMx method. The resulting dose–response curve obtained after linear
regression analysis is given by y =
0.0297x - 0.0147 (r = 0.979,
Sy|x = 0.0187).
|
|
Precision and linearity studies.
Reproducibility of the
present method was determined by analyzing two clinical samples (GHb
4.4% and 12%, as determined by the Abbott IMx method) and one bovine
blood sample (GHb <1% determined by Pierce GlycoTest II). The samples
were analyzed in replicates of 10 each per run over 3 days. A limited
time span was used for these studies to reduce possible errors that
result from sample and instrument instability. Because the present
assay is based on reflectance measurements of Hb, oxidation of the
sample and formation of methemoglobin will introduce spectrum changes
that might affect the results. The prototype instrument was not turned
off during testing, and the same instrument calibration was used
throughout the whole experiment. The day-to-day variance was calculated
according to a one-stage nested design and ANOVA.
The linearity of the present method was demonstrated by analyzing a
bovine control sample (GHb 22%, determined by boronic affinity
minicolumns, see Controls) and dilutions of this sample with
the use of nonglycated blood from the same specimen as diluent (GHb
<1%).
Interference studies.
The possible interference of endogenous
glucose, glycoalbumin, and fructosamine was tested by adding these
compounds in various amounts to whole-blood samples. The effect of
plasma fructosamine was tested by addition of washed erythrocytes to
plasma containing various amounts of fructosamine. The effect of the
labile GHb fraction (Schiff base, pre-GHb) was evaluated after
incubation of three whole-blood samples with glucose (14 g/L) for
3 h at 37 °C. The samples used were assayed in triplicate
before and after incubation with glucose.
Blood samples.
Whole blood was collected in evacuated
blood-collecting tubes containing EDTA. To test the possible
interference from citrate and oxalate used as anticoagulants, a few
samples were collected in both EDTA tubes and tubes containing citrate
(0.0129 mol/L sodium citrate/mL of blood) or oxalate (potassium oxalate
2 g/L blood).
For the correlation studies, samples assayed for GHb with the Bio-Rad
Diamat system (HPLC, ion-exchange), Boehringer Mannheim
Tinaquant-method (immunological), and the Abbott IMx Hb
A1c assay (boronic acid affinity) were collected from
different clinical laboratories running these methods. The specimens
were assayed for GHb by all methods within 1 week after sampling.
GHb controls.
GHb controls were produced from bovine blood by
reductive glycation (14). The control values were assigned
by the Pierce GlycoTest II and the Abbott IMx Hb A1c
method.
|
Results and Discussion
|
|---|
We have previously presented dye–boronic acid derivatives as
probes for GHb (6). Compared with these derivatives, the
XC-DAPOL-CPBA molecule contains the following important characteristics
to improve its feasibility. Firstly, the absorption maximum of the
chromophoric residue is located above 600 nm
(Amax = 618 nm with molar absorptivity 
70 000
L mol-1 cm-1). This wavelength area is not
interfered with by the absorption bands of Hb, an important limitation
of the earlier derivatives.
Second, to facilitate increased water solubility, the molecule contains
a sulfonic acid group and a hydrophilic spacer between the chromophore
and the phenylboronic acid. Lastly, but not least, the
3-aminophenylboronic acid residue, routinely used in boronic affinity
matrices (3)(15)(16)(17) and in earlier
dye–phenylboronic acid derivatives (6), is replaced by a
4-carboxyphenylboronic acid residue. The crucial importance of this
configuration is seen by a highly increased chemical stability of the
compound and lowered pKa of the boronic acid residue
itself. This facilitates esterification of the probe with the glycated
residues of hemoglobin at a pH appropriate for the selective
precipitation method applied. Additionally, increased acid strength of
the boronic acid also contributes to increased solubility of the
conjugate.
The reflectance characteristics of the blue-colored boronic acid
derivative compared with those of Hb are shown in Fig. 2
. The visible absorption maximum is close to 620 nm, the
wavelength used for the reflectometric determinations. At this
wavelength spectroscopic interference from Hb itself is negligible.
Similarly, because Hb can be measured at 470 nm without interference
from the probe, a simple two wavelength measurement is sufficient to
quantitate the chromophores present in the Hb precipitate. For the
relationship (Fig. 3
) between calculated K/S ratio (y) and Abbott IMx
standardized Hb A1c values (x) on 49 samples,
the regression equation was: y =
0.0297x - 0.0147 (r = 0.979,
Sy|x = 0.0187). This relationship was used to
calibrate the presented method transforming the measured K/S ratios to
%Hb A1c values. Samples with the same Hb
concentration but different extents of glycation would all give the
same K/S (470 nm) reading (assuming equal precipitation and application
of sample on the filter), but the K/S (620 nm) readings would vary
between the samples and increase in proportion to the glycation of Hb
in the sample. Thus, the resulting K/S ratio increases. Because the
boronic acid compound XC-DAPOL-CPBA is a probe of glycated residues,
the K/S ratio calculated from Eq. 2
ideally depicts the molar ratio
between Hb and the glycated residues present. However, because of the
heterogeneity of the GHb molecules related to the exact number of
glycated residues per molecule, possible unspecific binding of the
probe, signal background, and measuring errors, the recorded K/S ratio
of a series of samples has to be related to the amount of glycation in
the same samples to establish the calibration curve. Although samples
assayed with the Abbott IMx method were used for calibration of the
present method, other well-documented high-precision methods for
assaying GHb (HPLC, ion-exchange/boronic affinity) could be used
beneficially. Studies to find the optimal calibration of the present
method are currently taking place.
View larger version (18K):
[in this window]
[in a new window]
|
Figure 2. Spectrum characteristics of the blue-dyed boronic acid
derivative XC-DAPOL-CPBA compared with Hb.
The spectra were recorded by reflectometric measurements (%R) over the
wavelength area 400–700 nm with a spectrum resolution of 10 nm. The
recorded %R values were linearized with Eq. 1
before they were plotted
as K/S values. Although the illustrated spectra were recorded on pure
compounds, a theoretical Hb precipitate showing a spectrum
corresponding to the sum of the two illustrated spectra (K/S ratio
= 1.069) would, from the regression line given in Fig. 3
, represent a
sample containing 36.5% Hb A1c.
|
|
Precision.
As seen in Table 1
the imprecision of the assay is acceptable. Only small
differences were observed between inter- and intraassay variation and
between the different samples tested. The small difference between
inter- and intraassay imprecision occurs because only single
measurements are performed in the assay and the whole assay procedure
is completed for each sample.
The design of the filter device used in the assay and the compatibility
with the reader have been crucial to the analytical results. Although
glass fiber filters normally were used in the filter devices,
background readings of the blue color were sometimes observed.
Typically, this was seen when little Hb was applied on the filter,
i.e., analyzing samples with low hematocrits. This phenomenon can be
corrected for both optically and electronically, but this facility was
not available in the instrument used in this study. Still, varying the
amount of Hb applied on the filter by as much as ±20% with the use of
samples with normal hematocrit values did not affect the quantitative
determination. This property makes it possible to perform the assay
without precision pipetting of the whole-blood sample.
Assay procedure/conditions.
As already mentioned, the design
of the filter device is important in achieving good assay performance.
The filter has to trap the Hb precipitate efficiently and still have
good flow properties combined with low nonspecific binding of
XC-DAPOL-CPBA. Fig. 4
shows the results after optimization experiments mapping the
slope of the dose–response curve (defined as the difference in K/S
value between a high and a low GHb sample) as a function of blood
sample volume and sample application/washing volume. This was done to
find the most stable and reliable assay conditions with the filter
device used. Focusing on ease of performance of the assay, identical
volumes were chosen for both sample application on the filter and the
washing step. As can be seen from the figure, the slope of the
dose–response curve approaches maximum as the blood sample volume is
minimized. The highest slope was found with the use of as little as 1
µL of sample, combined with a sample application/washing volume on
the filter of ~12 µL. However, these assay conditions makes
precision pipetting necessary because of high sensitivity toward the
blood sample volume. Additionally, high sensitivity against the
hematocrit value of the blood sample also affects the reproducibility
and general performance of the assay. Although a 3.5-µL blood sample
and 16-µL application/washing volume were used in the standard
procedure, only small variations in assay performance were observed
when the sample volume was varied between 3.5 and 5 µL and the
application/washing volume between 15 and 18 µL. This area of optimal
assay conditions is found on the ridge seen in Fig. 4
.
View larger version (147K):
[in this window]
[in a new window]
|
Figure 4. Optimization of assay volumes.
A three-dimensional contour plot of the slope of the dose–response
curve shown as a function of blood sample volume (x axis)
and the volume used for application of sample (s) and washing (w) on
the filter (y axis). Slope is defined as the difference in
K/S ratio (x1000) between two blood samples containing 4.7% and
12.8% Hb A1c, respectively. The response surface (slope)
was determined after simplex optimization with blood sample volume and
s and w as variables. The volumes s and w were identical and varied in
parallel during the experiment. GHb assay solution (200 µL) was used
in all experiments.
|
|
Linearity and detection limit.
Dilution of in vitro glycated
bovine samples with high glycation showed a highly linear relationship
between observed and calculated values. Use of a sample with 22% GHb
(as determined with boronic acid minicolumns) resulted in a observed
linear correlation between the recorded %Hb A1c values
(y) and the calculated %GHb values (x) of:
y = 0.92x + 0.81 (r = 0.998,
Sy|x = 0.521). Calibrating the presented method
against %GHb resulted in the relation: y =
1.00x + 0.001 (r = 0.996,
Sy|x = 0.523).
Because human blood samples with close to zero Hb A1c
are hard to obtain and difficult to prepare, native bovine whole-blood
samples (GHb <1%, as determined with boronic acid affinity columns)
were used as zero samples to determine the detection limit of the
method. This resulted in a recorded Hb A1c amount of 1.0%
and a CV of 5% (see Precision), corresponding to a
detection limit of 0.15% GHb (three SDs of the zero sample)—well
below physiological GHb in humans.
Effect of pH and temperature.
Variations in important assay
variables like pH and temperature are known to affect the analytical
performance of GHb assays, especially with the use of ion-exchange
methods (9)(10). Analytical results after the
temperature was varied between 4 and 34 °C are shown in Table 2
. As seen, only negligible effects were observed in spite of the
range tested. Similarly, varying the pH of the buffer by ±0.2 unit did
not affect the analytical results. These findings agree well with the
limited temperature and pH susceptibility known to characterize
boronate affinity methods (6)(9)(10)(11)(12)(13).
Interference studies.
The results after glucose (up to 100
mmol/L) and plasma fructosamine (up to 700 µmol/L) were added to
whole-blood samples with normal and above-normal concentrations of GHb
are shown in Table 3
and Table 4
, respectively.
Notably, all samples showed <4% variation in the %Hb
A1c values recorded across all interferents and
concentrations tested. Similarly, no effects were seen after adding
glycoalbumin (glycation 15%) up to a concentration of 50 g/L to the
samples. These results illustrate that free glucose and available
cis-diols in the form of fructosamine-containing plasma or pure
glycoalbumin (accounting for 80–90% of all glycated plasma proteins
(18)(19)) do not affect the analytical
performance of the method presented. These results are first of all
explained by the specificity of the precipitation method applied
(8), and secondly by aspects related to the
concentration-dependent dye–phenylboronic acid–cis-diols equilibrium
reaction.
Incubation of whole-blood samples with glucose to produce pre-GHb did
not induce substantial changes in the recorded %Hb
A1c values compared with nonincubated samples (data
not shown). This result is in accordance with other reports showing
that the preglycated, labile Hb fractions do not interfere with GHb
methods relying on the boronic acid affinity esterification
(9)(10)(12). This eliminates the
need for pretreatment of samples to remove the preglycated fraction,
thereby simplifying the assay procedure. Citrate and oxalate used as
additives in sample collection tubes showed no significant effect.
Agreement with other methods.
The present method was compared
(Fig. 5
) with three major methods used in the determination of GHb. Of
these, two were boronic acid affinity methods. Linear regression data
are presented in Table 5
. The present method correlates well with all four GHb methods
(r = 0.94–0.99), and slopes of the regression equation
were 1.00 to 1.01. These results are in agreement with the linear
relationship demonstrated between boronic affinity and ion-exchange
methods (20)(21). As Table 5
shows, the
present method reports lower glycation of the samples than the boronic
acid affinity minicolumns do. This is expected because minicolumns
measure total GHbs and not exclusively the Hb A1c fraction.
Notably, calibrating the present method to read %GHb values resulted
in a slope of the correlation curve close to unity.
View larger version (14K):
[in this window]
[in a new window]
|
Figure 5. Correlation with other GHb methods: Pierce GlycoTest II
method (n = 17) (A), Boehringer Mannheim Tinaquant
method (n = 57) (B), and Bio-Rad Diamat (n = 34)
(C).
The depicted data are all based on singleton measurements. Linear
regression data are shown in Table 5
.
|
|
Although the variability of the results around the regression line
seems to be a bit larger with the Boehringer Mannheim Tinaquant Hb
A1c method (Fig. 5B
, Table 5
) than with the methods
from Pierce (GlycoTestII) and Bio-Rad (Diamat), the differences seen
between the compared methods can be explained by assay precision and
the different assay principles used. All methods measure GHb; however,
different populations of the various glycated residues of Hb can be
determined depending on whether ion-exchange, immunological, or boronic
acid affinity methods are used. More precise GHb methods than the
Abbott IMx method will also be evaluated for calibration of the present
method.
In conclusion, the analytical method presented represents a novel
application of the boronic acid affinity principle for the
determination of GHb. The method shows good analytical performance and
correlates well with other accepted methods in the determination of
GHb. The test is fast and easy to perform (5 min on filter strips) and
relies on a nonexpensive instrument reporting the Hb
A1c result directly based on simple and reliable
reflectometric readings. These characteristics, combined with the ease
of execution, makes the assay well suited for the doctors' office for
the monitoring of diabetes mellitus.
|
Footnotes
|
|---|
1 Nonstandard abbreviations: GHb, glycohemoglobin; Hb, hemoglobin; XC-DAPOL-CPBA, xylene cyanol-1,3-diaminopropanol-4-carboxyphenylboronic acid conjugate. 
|
References
|
|---|
-
Weith HL, Wiebers JL, Gilham PT. Synthesis of cellulose derivatives containing the dihydroxyboryl group and a study of their capacity to form specific complexes with sugars and nucleic acid components. Biochemistry 1970;9:4396-4401.
[Medline]
[Order article via Infotrieve]
-
Maestas RR, Prieto JR, Kuehn GD, Hageman JH. Polyacrylamide-boronate beads saturated with biomolecules: a new general support for affinity chromatography of enzymes. J Chromatogr 1980;189:225-231.
-
Mallia AK, Hermanson GT, Krohn RI, Fujimoto EK, Smith PK. Preparation and use of a boronic acid affinity support for separation and quantitation of glycosylated hemoglobins. Anal Lett 1981;14:649-661.
-
Fiechtner M, Ramp J, England B, Knudson MA, Little RR, England JD, et al. Affinity binding assay of glycohemoglobin by two-dimensional centrifugation referenced to hemoglobin A1c. Clin Chem 1992;38:2372-2379.
[Abstract/Free Full Text]
-
Wilson DH, Bogacz JP, Forsythe CM, Turk PJ, Lane TL, Gates RC, Brandt DR. Fully automated assay of glycohemoglobin with the Abbott IMx® analyzer: novel approaches for separation and detection. Clin Chem 1993;39:2090-2097.
[Abstract]
-
Frantzen F, Grimsrud K, Heggli D-E, Sundrehagen E. Soluble highly colored phenyl boronic acids and their use in glycohemoglobin quantification. Clin Chim Acta 1997;263:207-224.
[ISI][Medline]
[Order article via Infotrieve]
-
Kubelka P. New contributions to the optics of intensely light scattering materials. J Opt Soc Am 1948;38:448-457.
-
Frantzen F, Grimsrud K, Heggli D-E, Sundrehagen E. Selective precipitation of human hemoglobin by organic solvents and metal cations. Hemoglobin 1997;21:155-172.
[ISI][Medline]
[Order article via Infotrieve]
-
Miedema K, Casparie T. Glycosylated haemoglobins: biochemical evaluation and clinical utility. Ann Clin Biochem 1984;21:2-15.
-
Mayer TK, Freedman ZR. Protein glycosylation in diabetes mellitus: a review of laboratory measurements and of their clinical utility. Clin Chim Acta 1983;127:147-184.
[ISI][Medline]
[Order article via Infotrieve]
-
Oremek G, Seiffert UB, Schmid G. Determination of glycated hemoglobin by affinity chromatography. Clin Chim Acta 1987;168:81-86.
[ISI][Medline]
[Order article via Infotrieve]
-
Middle FA, Bannister A, Bellingham J, Dean PDG. Separation of glycosylated haemoglobins using immobilized phenylboronic acid. Biochem J 1983;209:771-779.
[ISI][Medline]
[Order article via Infotrieve]
-
Hall P, Cook JGH, Gould BJ. An inexpensive, rapid and precise affinity chromatography method for the measurement of glycosylated haemoglobins. Ann Clin Biochem 1983;20:129-135.
[ISI][Medline]
[Order article via Infotrieve]
-
Frantzen F, Sundrehagen E. Haemoglobin standards,
International Patent Application PCT/96/00975, 1996..
-
Frantzen F, Grimsrud K, Heggli D-E, Sundrehagen E. Protein–boronic acid conjugates and their binding to low-molecular-mass cis-diols and glycated hemoglobin. J Chromatogr B Biomed Appl 1995;670:37-45.
[ISI][Medline]
[Order article via Infotrieve]
-
Bissè E, Wieland H, Coupling of
m-aminophenylboronic acid to s-triazine-activated
Sephacryl: use in the affinity chromatography of glycated hemoglobins.
J Chromatogr B Biomed Appl 1992;575:223..
-
Koyama T, Terauchi K. Synthesis and application of boronic acid-immobilized porous polymer particles: a novel packing for high-performance liquid chromatography. J Chromatogr B Biomed Appl 1996;679:31-40.
[ISI][Medline]
[Order article via Infotrieve]
-
Schleicher E, Wieland OH. Protein glycation: measurement and clinical relevance. J Clin Chem Clin Biochem 1989;27:577-587.
[ISI][Medline]
[Order article via Infotrieve]
-
Austin GE, Mullins RH, Morin LG. Non-enzymatic glycation of individual plasma proteins in normoglycemic and hyperglycemic patients. Clin Chem 1987;33:2220-2224.
[Abstract/Free Full Text]
-
Little RR, England JD, Wiedmeyer HM, McKenzie EM, Mitra R, Erhary JB, et al. Interlaboratory standardization of glycated hemoglobin determinations. Clin Chem 1986;32:358-360.
[Abstract/Free Full Text]
-
Flückiger R, Woodtli T, Berger W. Quantitation of glycosylated hemoglobin by boronate affinity chromatography. Diabetes 1984;33:73-76.
[Abstract]
Top
Abstract
Introduction
