| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Articles |
1
Service de Biochimie A, Hôpital Saint-Antoine, AP-HP, and
2
Laboratoire de Biochimie et Glycobiologie, UFR Pharmacie Université René Descartes-Paris V, Paris, France.
a Address for correspondence: 184 rue du Faubourg Saint-Antoine, 75571 Paris Cedex 12 France. Fax 33 1 49 28 20 77;
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, ß, 
, 
), to each of which is bound a heme group. Hb in
normal adult human blood is >96% Hb A
(2ß2), ~2–3% Hb A2
(22), and <1% Hb F
(22). Normal newborn blood contains Hb F
as the major constituent (60–80%) and the rest is Hb A
(1). Analysis of the composition of human blood is of
major clinical interest in several congenital defects associated with
abnormal Hb content. These hemoglobinopathies are grouped into
defective variants of Hbs (such as Hb S and >600 other variants) and
thalassemias, which are characterized by abnormal proportions of normal
globin chains (2)(3)(4). Various electrophoretic approaches, including cellulose acetate electrophoresis at alkaline pH, citrate agar electrophoresis at acidic pH, and isoelectric focusing (IEF) in polyacrylamide or agarose gels, as well as HPLC, immunological assays, structure analysis, and genotypic methods are used to investigate hemoglobinopathies (5)(6). The electrophoretic methods, still widely used, are declining, given their difficulty of automation and inaccurate quantification of minor Hb constituents. HPLC is now considered to be a sensitive, specific, and reproducible alternative to electrophoresis and its use has been significantly expanded, especially with the development of rapid and well-resolving methods, some of them being applied to fully automated analyzers (7)(8)(9). Moreover, newborn screening programs currently developed in various areas of Europe and the US need a universal method to detect and quantify Hb variants in small amounts of samples. In this field IEF emerges as a remarkable resolving method that allows unequivocal identification of a large number of Hb variants; HPLC with its automation and its quantitative power is starting to be used for such purposes (10)(11)(12).
Capillary electrophoresis is a new versatile analytical electrophoresis technique that uses numerous separation principles, including capillary isoelectric focusing (CIEF) (13)(14)(15)(16)(17)(18). Capillary zone electrophoresis of intact Hbs and of globin chains, as well as the characterization of tryptic digests, has been reported by different groups (19)(20)(21)(22)(23), and several have recently shown that Hb separation can also be achieved by CIEF with various capillaries (21)(24)(25)(26). The aim of the present work was to develop and compare, for both resolution and quantification capabilities, two methods suitable for complete routine analysis of Hbs—a CIEF assay and a HPLC assay.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
origin and preparation of samples
Samples from adults and newborns were collected at the maternity
ward of Hôpital Saint-Antoine and sent to the laboratory for
routine Hb analysis. Blood was drawn by venipuncture into
EDTA-containing tubes. The samples used for this study were also
analyzed by conventional methods for Hb analysis, i.e., alkaline and
acidic electrophoreses, microcolumn ion-exchange chromatography for Hb
A2, and HPLC for Hb F. Blood sampling was in
accordance with the ethical standards of our hospital's committee.
Whole-blood samples (30 µL) were lysed by addition of 1 mL of
hemolyzing solution (KCN, 5 mmol/L), after which the hemolysate was
centrifuged for 3 min at 12 000g. The clear supernatant was
used directly for CIEF or diluted twofold for HPLC.
Normal and abnormal controls (Lyphochek Hemoglobin A2 Bi-Level; Bio-Rad) were stored and used according to the manufacturer's instructions. The pathological control containing Hb A2, S, F, and A (pI 7.42, 7.21, 7.15, and 7.10, respectively) was included in each series of analyses.
cief
Instrumentation.
We used a P/ACE 5000 System with a
UV/visible detector set at 415 nm and System Gold software (vers. 8.1)
from Beckman Instruments. The separations were performed on 50 µm
(i.d.) x 37 cm (total length) [30 cm (length to detector)] neutral
(polyacrylamide)-coated capillaries from Beckman.
Separation program.
The instrument was set up with the
anode at the inlet end of the capillary (anolyte:
H3PO4, 20 mmol/L) and the cathode at the outlet
(catholyte: NaOH, 20 mmol/L). The capillary was thermostated at
25 °C. The ampholyte solution consisted of a mixture of Pharmalyte
pH 6–8 and Pharmalyte pH 7–9 (3:1, by vol), at a final concentration
of 20 mL/L in a 4 g/L methylcellulose solution. Before each run, the
capillary was pressure-rinsed and filled with the ampholyte solution
(138 kPa, 3 min), then backflushed with the catholyte (138 kPa, 0.1
min) to limit the focusing area to the capillary section anodal to the
detection window. The sample was loaded by low-pressure injection (3.5
kPa, 15 s), focused 3 min at 30 kV, and then mobilized by low
pressure while maintaining a high voltage (25 kV). Between each
analysis the capillary was rinsed with deionized water (138 kPa, 2
min). The mobility of each Hb fraction in CIEF was expressed as a
retention time, as for HPLC.
hplc
Instrumentation.
We used the integrated HPLC System Gold
from Beckman, consisting of a Model 126 pump gradient, a Model 166
UV/visible detector set at 418 nm, a Model 507E autosampler, and System
Gold software (vers. 8.1). The system was equipped with a 100 x
4.0 mm column packed with a weak cation-exchanger, porous (100-nm pore
size) 5-µm microparticulate polyaspartic acid–silica (Poly Cat A)
purchased from Touzart & Matignon (Les Ulis, France).
Gradient program.
Separation of Hbs was accomplished by
a salt gradient obtained by mixing buffers A (Bis-Tris 20 mmol/L, KCN 2
mmol/L, pH 5.8) and B (Bis-Tris 20 mmol/L, KCN 2 mmol/L, sodium citrate
75 mmol/L, pH 5.8). The flow rate was 1.5 mL/min. The column was
equilibrated with B:A (33:67 by vol). After injection of the sample,
the proportion of B was increased linearly to B:A (45:55 by vol) at 5
min and to 100:0 at 9 min; finally, the mobile phase was returned to
33:67 at 11 min, for reequilibration.
data analysis
The System Gold software quantifies the data on the basis of peak
areas; values are expressed as a percentage of total Hb. Because of the
selective absorption of heme at 415–418 nm, all major and minor peaks
detected were attributed to Hb.
The differences between groups of data were analyzed by Student's paired t-test with a two-tail probability distribution. The correlations between variables were calculated by linear regression analysis.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
). The resolution in CIEF (Rs) can be defined as the difference
in isoelectric point of adjacent peaks and, in a linear pH gradient,
can be calculated with the formula: Rs =
(t2 -
t1)/
t, where
t2 and t1 are the
migration times of two adjacent peaks, and t is the peak
width (27). The gradient was linear between pH 7.0 and
7.5; based on the separation of Hb F from Hb A, therefore, the
resolution was 0.02 pH unit.
|
The analysis by HPLC and CIEF of some ß-variants that comigrate with
Hb A2 in alkaline electrophoresis are
presented in Figs. 2
and
3. With HPLC, Hb A2 and Hb E were completely
separated from Hb C, but Hb C-Harlem coeluted with Hb C (not shown) and
Hb E with Hb A2. With CIEF, Hb A2, Hb E, and Hb
C-Harlem were partially resolved from Hb C, but Hb E always comigrated
with Hb A2. Fig. 4
compares the separations by CIEF of Hb S and Hb D-Punjab, which
comigrate in alkaline electrophoresis; as shown by the superimposition
of the CIEF profiles, these hemoglobins could be differentiated.
However, there was no difference between the hemoglobins of group D,
such as Hb D-Punjab and Hb D-Korle Bu (not shown).
|
|
reproducibility of retention times
The reproducibility of retention times was evaluated for both
methods with a normal adult sample stored at -20 °C. Intraassay
reproducibility was calculated from the retention times of Hb A
obtained in 20 successive runs made in 1 day and interassay
reproducibility from the retention times obtained in 10 different days
with use of different reagent preparations. The within-day CV (n =
20) was 1.2% for HPLC and 1.0% for CIEF. The between-day CV (n =
8) was 3.3% for HPLC and 4.9% for CIEF.
quantification of hb variants
For each method, imprecision was evaluated by assaying various
samples: one with normal Hb A2, a second with increased
values of Hb A2 and Hb F, and a third containing Hb S.
Intraassay variability was determined by analyzing each specimen 20
times in a series and interassay by analyzing each one, stored at
-20 °C, on 10 different days. The results yielded CVs between 1%
and 12%, as reported in Table 1
. Intraassay CVs always were less than corresponding interassay
CVs, more particularly for HPLC. All but low proportions of Hb
A2 by CIEF were analytically suitable.
|
Table 2
summarizes results of correlation between CIEF and HPLC for
quantification of Hb A2, Hb F, and Hb S. The values of
Hb A2 measured in 43 samples (35 from apparently normal
individuals and 8 with minor ß-thalassemia) were significantly
(P <0.001) lower by CIEF than by HPLC but with a strong
correlation. In the 35 apparently normal individuals, the mean Hb
A2 value was 2.0% (range 1.6–2.7%) by HPLC and 1.8%
(range 1.3–2.4%) by CIEF. In the 8 patients with minor
ß-thalassemia, Hb A2 was 4.9% (range 3.7–6.6%) of
total Hb by HPLC and 4.1% (range 3.4–5.8%) by CIEF. The values of Hb
F measured in 24 samples (16 adults and 8 newborns) did not
statistically differ between both methods and were highly correlated
between them. The values of Hb S measured in 20 samples (15 adults, 3
newborns with Hb S trait, and 2 adults with Hb S disease) were
significantly (P <0.001) higher by CIEF than by HPLC but
still correlated strongly.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Several CIEF configurations for Hb analysis utilizing various capillaries and pH gradients have been described; among these, two-step methods have emerged, first on fused-silica capillaries (21) and then on coated capillaries, for better control of electroosmotic flow and improved resolution (25)(26). Many of the electropherograms presented in those studies showed poor resolution and contained several unidentified minor peaks. An enhanced resolution was obtained in narrow pH gradients but depended on the capillary used. The best resolution was observed by Hempe and Craver (25), using a dimethylpolysiloxane-coated capillary (DB-1); in our experience, however, this capillary gave extremely high variability in retention times. A neutral capillary was developed in a one-step CIEF configuration and in a wide range of pH values, initially to separate monoclonal antibodies (30).
To improve Hb analysis by CIEF, we developed a new CIEF assay, keeping the physico-chemical conditions of the two-step CIEF assay described for use with DB-1 but instead using the neutral capillary applied to CIEF of monoclonal antibodies. In our configuration, the resolution did not reach that obtained with DB-1 CIEF. However, it was better than that obtained with a linear acrylamino-coated capillary, which did not allow baseline resolution between Hb F and Hb A, as reported by Conti et al. (26). The satisfactory reproducibility on retention times at this resolution allowed unequivocal identification of common Hb variants. However, we observed a progressive decrease in the retention times, leading to a loss of resolution after ~200 runs; the instability of the coating or of the inner treatment of the capillary was presumably responsible, as has been reported for other capillaries with use of narrow pH gradients (31). Thus, the capillary must be changed when baseline resolution between Hb F and A is not obtained.
The variability on retention times was less in our HPLC assay with the Poly Cat A column than in our CIEF assay, perhaps related to greater stability of the former. The resolving powers of the CIEF and HPLC assays differ slightly, sometimes in complementary ways, because they involve different modes of separation. For example, (a) CIEF allowed separation of Hb C and Hb C-Harlem, whereas HPLC did not—although this separation was effected in another HPLC assay, which seemed to allow more resolution but with longer run times (8); (b) HPLC could separate Hb C from Hb A2, whereas CIEF did not allow baseline resolution between these Hbs, so HPLC is unique in distinguishing homozygous Hb C disease from double heterozygous inheritance of Hb C/ßo-thalassemia, in which Hb A2 is increased.
Both assays were able to separate and quantify in a single run normal Hb forms and common pathological variants, i.e., Hb A, Hb F, Hb A2, Hb S, Hb C, Hb E, and even Hb A1c. Both also appeared to be more efficient than classical gel electrophoresis and more convenient and accurate than gel IEF. The imprecision of quantification estimated from peak area was satisfactory for both methods, although CV reached 12% for low amounts of Hb A2 on CIEF. The greater between-day variabilities may be linked to a progressive degradation with time of the column or, more specifically, of the capillary.
Quantitative values for Hb A2, Hb F, and Hb S were
highly and linearly correlated between both methods. The lower Hb
A2 values obtained with CIEF may be related to the
separation of glycated Hb A2 from the nonglycated Hb
A2, as previously described (25). Both CIEF
and HPLC allowed the diagnosis of ß-thalassemia by an accurate
quantification in a single run of both Hb F and Hb A2,
without overlap between Hb A2 values for normal individuals
and for ß-thalassemia patients. For this differentiation, both assays
were more convenient than classical methods, which need two
experiments, i.e., alkali denaturation for Hb F and microcolumn
ion-exchange for Hb A2. Nevertheless, the reference range
for Hb A2 by CIEF needs to be established, particularly the
lower limit for normal values, because the diagnosis of minor
-thalassemia is characterized by a low Hb A2 value. The
mean values of Hb F given by both methods were identical, and both
allowed accurate quantification of Hb F at low values (±1%), as
observed in adults (normal or with minor ß-thalassemia), and at high
values, as found in newborns (
90% in premature infants).
Both CIEF and HPLC require a relatively high investment in materials, but this can be widely compensated for by savings in technical time, as realized by the complete automation, and by the fact that either method can be used as a unique assay for complete Hb analysis. The HPLC columns are more expensive than the CIEF capillaries but have a longer lifespan. Unlike HPLC, CIEF uses very low volumes of buffer, which reduces both reagent and waste-disposal costs. The duration of analysis, ~15 min for HPLC and 20 min for CIEF, allows a relatively high daily throughput, adapted to primary use in medium-size laboratories.
In conclusion, HPLC on Poly Cat A columns as well as our new CIEF assay on a neutral capillary are suitable for routine investigation on hemoglobinopathies. Each or both can be performed as a unique assay for identification and quantification of major and minor Hb fractions in small sample volumes and can also be used for the diagnosis of hemoglobinopathies in newborns, e.g., in screening programs. These two methods, which offer common qualities with different separation principles, could become complementary but also competitive for Hb analysis.
|
|
Acknowledgments |
|---|
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-thalassemia with hemoglobinopathies. Hematol Oncol Clin North Am 1991;5:453-473.
[ISI][Medline]
[Order article via Infotrieve]
-thalassemias. J Chromatogr 1993;652:119-129.
The following articles in journals at HighWire Press have cited this article:
|
|
 |
|
  L. B. Thomas, S. J. Agosti, M. A. Man, and S. M. Mastorides Screening for Hemoglobinopathies During Routine Hemoglobin A1c Testing Using the Tosoh G7 Glycohemoglobin Analyzer Ann. Clin. Lab. Sci., January 1, 2007; 37(3): 251 - 255. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Joutovsky, J. Hadzi-Nesic, and M. A. Nardi HPLC Retention Time as a Diagnostic Tool for Hemoglobin Variants and Hemoglobinopathies: A Study of 60000 Samples in a Clinical Diagnostic Laboratory Clin. Chem., October 1, 2004; 50(10): 1736 - 1747. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. M. Clarke and T. N. Higgins Laboratory Investigation of Hemoglobinopathies and Thalassemias: Review and Update Clin. Chem., August 1, 2000; 46(8): 1284 - 1290. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Cotton, C. Lin, B. Fontaine, B. Gulbis, J. Janssens, and F. Vertongen Evaluation of a Capillary Electrophoresis Method for Routine Determination of Hemoglobins A2 and F Clin. Chem., February 1, 1999; 45(2): 237 - 243. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Mario, B. Baudin, A. Bruneel, J. Janssens, and M. Vaubourdolle Capillary Zone Electrophoresis for the Diagnosis of Congenital Hemoglobinopathies Clin. Chem., February 1, 1999; 45(2): 285 - 288. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
