ical Chemistry Division, Bureau of Laboratories, Centers for Disease Control, Atlanta, GA, 1980.
7. Byrne MF. New reference method for cholesterol “old fashioned.”
Clin Chem News 6, 16-17 (1980).
8. Cooper GR, Ullman MD, Hazlehurst J, et al. Chemical characteristics and sources of error of an enzymatic cholesterol method.
Clin Chem 25, 1074 (1979). Abstract.
9. Tel EM, Berends GT. Incomplete hydrolysisof cholesterol esters
during enzymatic cholesterol determination as evidenced by aqueous cholesteryl ester solutions: Comparison of six enzymatic procedures with the Liebermann-Burchard method. J Clin Chem Clin
Biochem 18, 595-601
(1980).
10. Folch J, Lees M, Sloane-Stanley GH. A simple method for the
CLIN. CHEM. 29/6, 1080-1082
isolation and purification of total lipids from animal tissues. J Biol
Chem 226, 497-509 (1957).
11. Siedel J, Schlumberger H, Klose S, et al. Improved reagent for
the enzymatic determination of serum cholesterol. J Clin Chem
Clin Bioc/zem 19, 838 (1981). Abstract.
12. Erb W, Boehler E. Dunnschichtchromatographische
Untersuchungen von Stuhl-Lipiden. J Clin Chem Clin Biochem 6, 379-383
(1968).
13. Halpaap H. D.C.-Fertigplatten mit Konzentrierungszone: Bestimmung von Lipiden. Kontakte (Merck) 1/78, pp 32-34 (1978).
14. Ludewigs M. Test report Monotest#{174}
Cholesterin, High Performance. Boehringer Mannheim GmbH, Printed Matter no. 4194
(1978).
(1983)
Glycosylated Hemoglobin Measured by Affinity Chromatography: MicroSample Collection and Room-Temperature Storage
Randle R. Little,1Jack D. England,2 Hsiao-Mei Wiedmeyer,2 and David E. Goldstein1’2
Under proper conditions, whole blood can be stored at room
temperature for as long as 21 days before measurement of
glycosylated
hemoglobin
by affinity chromatography.
Whole
blood (anticoagulated with EDTA or heparin) was placed in
capillarytubes, which were then sealed at both ends and
stored at room temperature. Just before assay, whole blood
was nnsed from the tubes and diluted 10-fold with water.
Samples of each patient’s blood were assayed as wholeblood hemolysates by affinity chromatography
after zero,
seven, 14, and 21 days of storage. Values for glycosylated
hemoglobin
did not change over 21 days of storage and
values for each storage day correlated well (r = 0.97, p <
.0001) with hemoglobin Aic measured in fresh erythrocyte
hemolysates by “high-performance”
liquid ion-exchange
chromatography.
Addftlonal Keyphrases: screening
diabetes
Glycosylated hemoglobin is now widely measured as an
index of long-term glycemic control in diabetic patients (1).
We recently reported the use of m-aminophenylboronic
acid
affinity chromatography to measure glycosylated hemoglobin (2). This method has certain important advantages for
the clinical laboratory over previously described techniques
in that it is relatively
insensitive
to changes in temperature
and is not affected by labile hemoglobin fractions (2). In
another report (Little et al., pp 1112-1114,
this issue) we
showed excellent stability of affinity chromatography
values
for whole blood stored for up to eight days at room tempera-
‘Department
of Pathology and 2Department of Child Health,
University of Missouri-Columbia, Health Sciences Center, Columbia, MO 65212.
Received Dec. 27, 1982; accepted Mar. 7, 1983.
1080
CLINICAL CHEMISTRY, Vol. 29, No. 6, 1983
tare,but with increases in values thereafter. In that study,
erythrocytes
from stored whole blood were washed and
incubated before assay. In this report, we show that whole
blood hemolysates prepared before assay, without washing
or incubation of erythrocytes,
show stability for at least 21
days. We also give data on a micro-scale method for blood
sample collection that is well suited to this affinity method.
Materials and Methods
We obtained blood specimens from 34 diabetic and nondiabetic subjects by venipuncture and collection in EDTAcontaining tubes. Part of each specimen was transferred to
four Caraway capillary tubes (300-350 L capacity), which
were then sealed with putty sealant (Critoseal; American
Scientific, St. Louis, MO 63043)on one end and Parafllm on
both ends. Three of the capillary tubes were then stored at
room temperature
(20-24 #{176}C)
until assayed seven, 14, and
21 days later. The fourth capillary tube from each subject
was assayed by affinity chromatography on the same day
the blood was collected. Nine of the 34 specimens collected in
EDTA were also collected in heparinized tubes.
Five additional specimens of capillary blood were collected via finger-stick directly into standard heparinized microhematocrit
capillary tubes. These capillary tubes were then
sealed at both ends with plastic caps (Critocaps, American
Scientific).
To determine the effects of glucose on affinity chromatography results before and after storage of whole blood, some
samples were supplemented
with glucose to give final
plasma glucose concentrations
ranging from 0.93 to 13.6 g/
L. An aliquot of each sample was then assayed and the
remaining samples were stored at room temperature in
capillary tubes until assay seven and 14 days after collection.
Just before assay, whole blood was washed from the
capillary tubes with water and finally diluted about 10-fold.
Glycosylated hemoglobin was measured with the Glycosylated Hemoglobin Clinovative Diagnostic Kit (Pierce Chemical Co., Rockford, IL 61105). Fifty microliters of the diluted
hemolysate was applied directly onto each column, and each
sample was assayed in duplicate according to instructions
provided in the kit. For quality control, pooled hemolysates
stored at -70 #{176}C
from diabetic and non-diabetic subjects
were assayed with each batch of samples.
For comparison with affinity chromatography results, we
used an aliquot of each fresh whole-blood sample to prepare
an erythrocyte hemolysate for analysis for hemoglobin A1
by a semi-automated “high-performance”
liquid-chromatographic ion-exchange method (3). These hemolysates were
stored at -70 #{176}C
until analysis.
Results
-
Unity line
Regresslsline
y.
1.03x
-
0.50
.
- 0.99
z
0
16
U
14
C.,
12
Glucose
concn, g/L
0.93
6.72
Days of storage
0
7
14
11.65
11.06
10.97
11.16
11.21
11.21
11.01
11.40
11.53
13.24a
11.59
11.06
12.76a
12.56a
14.24a
7.30
9.48
13.60
a >2 SD above the mean of 10 repetitive values measured on day 0.
Glycosylated hemoglobin as measured by affinity chromatography on each of the storage days correlated strongly
with hemoglobin A1 measured in fresh hemolysates by ionexchange chromatography,
an established method (y =
1.53x
0.996, all r values = 0.97, p < .0001).
Results by the affinity chromatography
method do not
appear to be affected directly by variations in glucose
concentration in the range we studied. However, after
storage at room temperature for 14 days, blood samples with
plasma glucose values greater than 6.72 g/L showed significant increases in glycosylated hemoglobin (Table 1).
-
Values for glycosylated hemoglobin in diabetic and nondiabetic samples stored as whole blood in capillary tubes
showed no appreciable change in glycosylated hemoglobin
after 21 days at room temperature. The mean glycosylated
hemoglobin values (± SEM) of 34 whole-blood samples were
13.45 ± 0.59, 13.62 ± 0.60, 13.30 ± 0.58, and 13.22 ± 0.61
on days zero, seven, 14, and 21, respectively. None of the
differences is significant. Figure 1 shows the relationship
between
individual
glycosylated hemoglobin values obtained on day 21 vs day 0. The regression line (y = 1.03x
0.50) falls very close to the unity line and individual
variability around these lines is relatively small. The average coefficient of variation between stored whole-blood samples assayed on different days was 3%. The inter-assay CVs
for the pooled diabetic and non-diabetic hemolysates were
3.05% and 2.59%, respectively. There were also no significant differences in glycosylated hemoglobin values between
paired EDTA-treated and heparinized blood samples stored
for 21 days.
18
Table 1. Effect of Glucose on Glycosylated
Hemoglobin after Storage at Room Temperature
I
S
10
10
12
14
16
PERCENT
GLYCOSYI.ATED
I*M%IOBIN
day 0
(Affinity Chromatography I
Fig. 1. Correlation between glycosylated
hemoglobinvalues measured
for whole-blood hemolysates after 20-24 0C storage of whole blood for
21 days and 0 days
Blood collectedby venipuncture in EDTA (0), in heparin (S), and by fingerstick
(x) into heparinizedmicrohematocrit tubes
Discussion
Many techniques are available for measuring glycosylated hemoglobin. For the clinical laboratory, the choice of a
method can be very difficult. Important
considerations in
selecting an appropriate method include its accuracy, equipment costs, extent of interference by labile fractions, and
temperature and pH sensitivity. Other important considerations include convenience of sample collection, sample
storage stability, and assay speed (4).
We and others have shown that sample storage conditions
and sample preparation before assay are critical for most
assay techniques. For ion-exchange and electrophoretic
methods, it is necessary to remove the labile fraction (proA1) which co-chromatographs or co-migrates with the stable glycosylated hemoglobin (3, 5). Ion-exchange and electrophoretic results are also affected by artifactual increases
in the hemoglobin A10+bfraction during storage of whole
blood or hemolysates at room temperature (3, 6). Analysis
for glycosylated hemoglobin by the thiobarbituric acid colorimetric method requires that glucose be completely removed
before assay (4).
Affinity chromatography is technically simple and is not
significantly affected by the above factors. Our results show
that it is feasible to measure glycosylated hemoglobin by
affinity chromatography in whole blood collected in capillary tubes and stored at room temperature for long intervals
without removing glucose or labile glycosylated hemoglobin
before assay. As little as 10 L of whole blood collected by
finger stick suffices for assay in duplicate, and either EDTA
or heparin can be used as an anticoagulant. The method is
easily adaptable to large-volume batch assays, the average
assay time being about 6 mm per duplicate sample. Results
are not affected appreciably by sample glucose concentration
unless it exceeds 6.7 g/L. One important technical consideration is that samples not be allowed to dry during storage to
avoid in vitro protein glycosylation with spuriously increased test results (7). Furthermore,
for storage at room
temperature, stability is greatest when hemolysates are
prepared directly from whole blood, without washing the
erythrocytes
or incubation with saline before assay.
With this assay method, the ease of sample collection and
the storage stability of samples have important clinical
applications aside from the direct advantages for the clinical
CLINICAL CHEMISTRY, Vol. 29, No. 6,
1983
1081
laboratory.
This method
may be particularly
well suited
to
large diabetic screening studies and as an adjunct to routine
diabetes care where patients are carrying out home monitoring of capillary blood glucose. In this situation, patients
could collect whole-blood samples at home and send them
directly to the central laboratory.
This study was supported in part by USPHS Research Grant HL13632.Our thanks to Mrs. Betty Payne for her technical assistance
in the preparation of this manuscript.
References
1. Trivelli
LA, Ranney HH, Lai HT. Hemoglobin components in
with diabetes mellitus. NEng1J Med 284,353-357
(1971).
patients
2. Klenk DC, Hermanson GT, Krohn RI, et al. Determinationof
glycosylated hemoglobin by affinity chromatography:
Comparison
with colorimetric and ion-exchange methods, and effectsof common
interferences. Gun Chem 28, 2088-2094 (1982).
3. Goldstein DE, Peth SB, England JD, et al. Effects of acute
changes in blood glucose on HbA,0. Diabetes 29, 623-628 (1980).
4. Goldstein DE, Parker KM, England JD, et al. Clinical application of glycosylated hemoglobin measurements. Diabetes 31, 70-78
(1982).
5. Bunn HF. Evaluation of glycosylated hemoglobin in diabetic
patients. Diabetes 30, 613-618 (1981).
6. Simon M, Hoover JC. Effect of sample instability on glycohemoglobin (HbA,) measured by cation-exchange chromatography. Clin
Chem 28, 195-198 (1982).
7. Goldstein DE, Wiedmeyer HM, England JD, et al. Glycosylated
protein in whole blood spotted on filter paper. Clin Chem 28, 386
(1982).
CLIN. CHEM. 29/6, 1082-1084 (1983)
Liquid-Chromatographic Study of Fluorescent Materials in Uremic Fluids
James S. Swan,1 Edith Y. Kragten,2 and Hans Veening3
Using reversed-phase
“high-performance”
liquid chromatog-
raphy with fluorescence detection, we separated and identified some naturally fluorescent compounds in uremic serum
and hemodialysate from patients with chronic renal disease.
Several of the naturally fluorescent compounds were identified as indole derivatives by co-chromatography with authentic standards. In one specific case, the identity was confirmed
by an enzymic peak-shift method. Compounds identified
included indican, kynurenic acid, tryptophan, and 5-hydroxyindole-3-acetic acid. Comparison of normal and uremic Serum showed that the fluorescent materials are present in
significantly greater concentrations in samples from uremic
patients.
AddItional Keyphrases: chromatography, reversed-phase
hemodialysis patients
chronic renal disease - indole derivatives
enzymic peak-shift method
Recent Letters to this journal have discussed the presence
of unidentified
naturally
fluorescent compounds in the
serum, urine, hemodialysate, and hemofiltrate of patients
with chronic renal failure (1-4). Vladutiu et al. (1) reported
finding one or more fluorescent substances (maximum emission near 340 nm) in the serum of patients with chronic
renal failure but not in the serum of those with acute renal
failure. They concluded that the substances were not vitamins or anabolic
steroids
and could be bound to a serum
protein. Schwertner et al. (2) found a fluorescent compound
(excitation maximum 322 nm, emission maximum 415 nm)
Department of Chemistry, Bucknell University, Lewisburg, PA
17837.
‘Present address: Scientific Systems, Inc., State College, PA
16801.
2Present address: Laboratory for Analytical Chemistry, University of Amsterdam, Amsterdam, The Netherlands.
3Address correspondence to this author.
Presented at the Pittsburgh Conference on Analytical Chemistry
and Applied Spectroscopy, March 1983.
Received Jan 17, 1983; accepted Mar. 7, 1983.
1082 CLINICAL CHEMISTRY, Vol. 29, No. 6, 1983
not only in serum, but also in hemofiltrate,
hemodialysate,
and urine of patients with chronic renal disease. The
substance was 80-fold more concentrated in uremic serum
than in normal serum. They also determined (by gel filtration) that the relative molecular mass (Mr) of the substance
was probably <1000, and that it was not a drug or drug
metabolite. Digenis et al. (4), finding a linear correlation
between the relative fluorescent intensity and the creatinine concentration in serum from patients with chronic
renal failure, concluded that deterioration of renal function
is accompanied by an increase in the unidentified fluorescent substances.
We have separated and identified some naturally fluorescent compounds in uremic serum and hemodialysate, by use
of reversed-phase “high-performance”
liquid chromatography (HPLC) with fluorescence detection. We also used a
unique
enzymic
peak-shift
identification
procedure.
Materials and Methods
Materials
Apparatus.
The mobile phase was pumped through the
apparatus with an SP8700 solvent-delivery
system (SpectraPhysics, Santa Clara, CA 95051). Samples were injected via
a syringe loading valve fitted with a 100-FL sample loop
(Model 7120; Rheodyne Inc., Berkeley, CA 94710) onto a 3.9
x 300 mm stainless-steel
column containing
7 m Zorbax
ODS (Du Pont Co., Wilmington, DE 19898) packing material. The compounds of interest were detected with an Aminco
filter fluorometer
(American
Instrument
Co., Silver Spring,
MD 20910) equipped with a G.E. germicidal lamp (Model
J4-7126), a Wratten 2C emission filter (peak wavelength
405 nm, approximate bandpass 10 nm), and an Aminco J47469 excitation filter (peak wavelength
295.4 nm, bandwidth at 50% of maximum transmission 9.2 nm). All chromatograms, retention times, and peak areas were recorded
with an SP4 100 computing
integrator
(Spectra-Physics).
Water was purified by passage through a Milli-R04
system, followed by a final cleanup through a Milli-Q
system (Millipore Corp., Bedford, MA 01730). Proteins (M