| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Technical Briefs |
1
Service de Biochimie A, Hôpital Saint-Antoine, AP-HP, 184 rue du Fbg Saint-Antoine, 75571 Paris Cedex 12, France;
2
Laboratoire de Biochimie et Glycobiologie, Faculté de Pharmacie Université René Descartes-Paris V, 75006 Paris, France, and
3
Analis, B-5000, Namur, Belgium;
a author
for correspondence: fax 33 1 49 28 20 77, e-mail nathalie.mario{at}sat.ap-hop-paris.fr
The analysis of human hemoglobins (Hbs) is of medical importance in a number of congenital defects. The hemoglobinopathies are grouped into defective variants of Hb, such as Hb S and >600 other variants, and thalassemias characterized by abnormal expression of the genes for normal globin chains (1)(2)(3). Alkaline electrophoresis, performed historically on cellulose acetate and currently on agarose, combined with citrate agar electrophoresis at acidic pH is still widely used (4)(5)(6)(7), but high-performance cation-exchange chromatography (HPCEC) offers superior resolution, speed, and automation (8)(9)(10). Capillary electrophoresis uses numerous separation principles and shares with HPLC the advantages of high resolution and automation, with on-line detection and direct quantification (11)(12)(13)(14)(15). Capillary isoelectric focusing on coated capillaries can be used to study Hb (16)(17)(18)(19)(20)(21), but is slower than HPCEC (21). The first reported assays based on capillary zone electrophoresis (CZE) gave poor resolution of Hbs or were not quantitative (22)(23)(24). The aim of the present work was to evaluate a rapid CZE assay with dynamic coating of the fused-silica at alkaline pH to detect abnormal Hbs and to quantify Hbs for the diagnosis of thalassemias, and at acidic pH to confirm the identity of Hb variants.
Reagents were obtained from Analis. The alkaline kit A2 contained a hemolyzer, an initiator consisting of a polycation, and an arginine buffer (pH 8.8) containing a polyanion. The acidic kit A1 has been described elsewhere for Hb A1c measurement (25). Adult and newborn samples were collected in EDTA-containing tubes at the maternity unit of the Hôpital Saint-Antoine (AP-HP, Paris, France) and received in the laboratory for Hb analysis. Whole blood samples (20 µL) were lysed by addition of 100 µL of appropriate hemolyzing solution (Analis). Normal and abnormal controls (Lyphochek hemoglobin A2 bi-level; Bio-Rad) were stored and used according to the manufacturer's instructions.
We used a P/ACE 5000 System with an ultraviolet/visible detector (at
415 nm) and System Gold software, Ver. 8.1, from Beckman Instruments.
The Gold software quantifies the data on the basis of corrected peak
areas for velocities, and values are expressed as percentages of total
Hb. The separations were performed on a 25 µm (i.d.) x 24 cm (total
length)/17 cm (length to detector) fused-silica capillary. The
instrument was set up with the anode at the inlet end of the capillary
and the cathode at the outlet. The capillary was thermostated at
26 °C. Before each electrophoresis, the capillary was first
pressure-rinsed with the initiator solution (138 kPa for 3 min), and
then rinsed and filled with the buffer (138 kPa for 1 min). The sample
was injected for 2 s at 3.5 kPa, followed by a 15-s injection with
buffer solution (3.5 kPa). The electrophoresis was performed with a
constant current of 52 µA (
16 kV) for 4 min. Between analyses, the
capillary was rinsed with 0.1 mol/L NaOH and deionized water (138 kPa,
1 min each).
Quantitative data obtained for Hbs A2, F, and S on patient samples by CZE were compared with those measured by HPCEC on a 100 x 5.0 mm polycatA column (Touzart and Matignon), using a method described elsewhere (21). Briefly, the separation was accomplished at a flow rate of 1.5 mL/min by a salt gradient obtained by mixing buffers A (20 mmol/L bis-Tris, 2 mmol/L KCN, pH 5.8) and B (20 mmol/L bis-Tris, 2 mmol/L KCN, 75 mmol/L sodium citrate, pH 5.8). The column was equilibrated with 33% buffer B, 67% buffer A; after injection of the sample, buffer B was increased linearly to 45% at 5 min and to 100% at 9 min, and was decreased to 33% at 11 min for equilibration.
The electropherograms obtained for the abnormal control containing Hbs
A1c, A, A2, F, and S are presented in Fig. 1
A. At alkaline pH, Hbs A2, S, F, and A
were separated, and Hb A1c comigrated with Hb A; at acidic pH, Hbs A1c,
F, A, and S were separated, and Hb A2 comigrated
with Hb A. Some ß-variants that comigrate on alkaline gel
electrophoresis could be differentiated by CZE at alkaline pH, as shown
by the superimposition of the profiles presented in Fig. 1B
. Hbs E and
A2 comigrated, whereas Hbs C and E were
differentiated (Fig. 1B
) as were Hbs C and A2
(Fig. 1B
, top panel); Hb D-Punjab and Hb S were also differentiated
(Fig. 1B
, bottom panel). Moreover, the identities of Hbs S, D, C
,and E could be confirmed by CZE at acidic pH; Hb D comigrated with Hb
A (not shown), whereas Hb S was separated (Fig. 1A
, bottom panel).
However, Hb E comigrated with Hb A (not shown) under acidic conditions,
whereas Hb C was separated (not shown). Hb Hope could be
detected by CZE at alkaline pH and identified at acidic pH (not shown),
whereas it comigrated with Hb A in conventional alkaline gel
electrophoresis.
|
The intra- (n = 20) and interassay (n = 20 days) CVs for Hb A
migration times were of 0.4% and 1.7% for alkaline CZE and 0.2% and
1.3% for acidic CZE. The results of the imprecision study yielded CVs
of 2.1–14% (Table 1
). Measurement of Hb F was linear for values between 0.8% and
80% (not shown), as assessed with mixtures of blood samples with high
and undetectable Hb F concentrations.
|
The values of Hb A2 measured on 45 samples (3 with low Hb A2, 37 from apparently normal individuals, and 5 with minor ß-thalassemia) were significantly higher than by HPCEC (paired t-test, P <0.001), but with a strong correlation (CZE = 1.059 HPCEC + 0.16%; Sy|x= 0.255; r2 = 0.922). The correlation for Hb F was performed on 30 samples (25 samples from adults and 5 samples from newborns). Ten samples from adults contained <0.5% Hb F as measured by HPCEC, whereas these amounts were undetectable by CZE. In the 20 other samples (i.e., 15 samples from adults and 5 samples from newborns), the Hb F concentrations were highly correlated between the two methods (CZE = 0.919 HPCEC + 1.24%; Sy|x = 1.937; r2 = 0.996) and did not differ statistically. The values of Hb S measured by the two methods on 17 samples—14 from patients with the Hb S trait (3 adults and 1 newborn) and 3 from adults with Hb S disease—did not differ statistically (CZE = 0.995 HPCEC + 0.79%; Sy|x = 1.869; r2 = 0.993).
The increased resolution by CZE compared with solid supports was first reported for plasma proteins (26). The CZE assays for Hb analysis at alkaline pH and using uncoated fused-silica capillaries described previously did not provide good separation between Hb F and Hb A (22)(23)(24). Here, we used dynamic coating, which increases the number of negative charges on the capillary wall and creates a high electro-osmotic flow that shortens the analysis duration. At alkaline pH and without ion-pairing agent, the separation of the Hbs is facilitated because they all are negatively charged and migrate in the direction opposite of the flow. Thus, the Hbs A2, S, F, and A were separated in <4 min; in particular Hb F was better separated from Hb A than in alkaline gel electrophoresis. Moreover, some variants that comigrate on agarose gel were separated, e.g., Hb S from Hb D-Punjab, Hb C from Hb E, and Hb Hope from Hb A. This CZE assay at alkaline pH with high electro-osmotic flow seemed to provide as much resolution as some capillary isoelectric focusing assays (16)(18), but not much as some others (17)(19)(20)(21). At acidic pH, as observed with conventional acidic gel electrophoresis, Hbs A2, E, and D all comigrated with Hb A. This assay cannot be used as a first test because it does not allow the detection of some abnormal variants and HbA2. Thus, for the identification of abnormal Hbs, CZE at acidic pH can be used only as a confirmation after CZE at alkaline pH, as in the classical gel approach.
For both CZE assays, the excellent reproducibility of migration times may be partly related to the dynamic coating, which is eliminated after each electrophoretic run and then regenerated before the next. This system avoids both protein adsorption on the uncoated capillary wall and the instability of permanent coatings.
We quantitatively evaluated CZE at alkaline pH for use as a complete qualitative and quantitative assay for the diagnosis of hemoglobinopathies. The imprecision for quantification by alkaline CZE was satisfactory, although the CV reached 14% for low concentrations of Hb F. However, the precise quantification of higher concentrations of Hb F and of high Hb A2 was adapted to the diagnosis of ß-thalassemias. The detection limit of Hb F by CZE at alkaline pH was <0.8%, which is the lowest point of the linearity curve, but >0.5%, which is the lowest concentration detected in adult samples. The linearity of Hb F measurement on a wide range of concentrations, in particular from low concentrations, makes the CZE assay adapted to samples from both adults and newborns. Thus, CZE is more convenient than agarose electrophoresis, which does not allow Hb F measurements below 10%. Quantitative values for Hb A2, Hb F, and Hb S were highly correlated with HPCEC values. The Hb A2 concentrations were substantially higher when measured by CZE, which could be related to the differences for the separation of glycated or otherwise posttranslationally modified Hb A2 forms (21)(27).
In conclusion, CZE at alkaline pH is rapid, precise, and gives high resolution, and is suitable for the screening of hemoglobinopathies, whereas CZE at acidic pH is better suited for confirmation of qualitative abnormalities. The capillary electrophoretic study of Hbs at both alkaline and acidic pH corresponds to an improvement of the historical scheme for the diagnosis of hemoglobinopathies by allowing accurate quantification and complete automation. In particular, the profiles were obtained in 7 min, including rinses, which is competitive to HPCEC on dedicated analyzers. Moreover, capillary electrophoresis systems are robust and could replace HPLC in clinical chemistry or hematology laboratories.
References
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
