Premium
Capillary zone electrophoresis with a dynamic double coating for analysis of carbohydrate‐deficient transferrin in human serum: Impact of resolution between disialo‐ and trisialotransferrin on reference limits
Author(s) -
Lanz Christian,
Thormann Wolfgang
Publication year - 2003
Publication title -
electrophoresis
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.666
H-Index - 158
eISSN - 1522-2683
pISSN - 0173-0835
DOI - 10.1002/elps.200305685
Subject(s) - carbohydrate deficient transferrin , capillary electrophoresis , transferrin , resolution (logic) , chemistry , chromatography , capillary action , analytical chemistry (journal) , electrophoresis , detection limit , gene isoform , reproducibility , reagent , materials science , biochemistry , alcohol , alcohol consumption , artificial intelligence , computer science , composite material , gene
Capillary electrophoresis with a dynamic double coating formed by charged polymeric reagents represents a very effective tool for the separation of iron‐saturated transferrin (Tf) isoforms and thus the determination of carbohydrate‐deficient transferrin (CDT) in human serum. The resolution between di‐ and trisialo‐Tf is dependent on the applied voltage and capillary temperature. With a 50 μm inside diameter (ID) capillary of about 60 cm total length mounted into the P/ACE MDQ, 28 kV and 40°C, the resolution of the two Tf isoforms is shown to be between 1.0 and 1.4, whereas with reduced voltage and/or temperature, increased resolution at the expense of elongated run times is observed. Best data with complete resolution ( R s ≥ 1.4) are obtained at 20 kV and 30°C. For the determination of CDT in serum, incomplete separation of di‐ and trisialo‐Tf is demonstrated to have an impact on the reference limits. Analysis of the sera of 54 healthy individuals with no or moderate alcohol consumption and using valley‐to‐valley peak integration, the upper (lower) reference limits for CDT in relation to total Tf at the two power levels are 1.33 (0.52) and 1.57 (0.81)%, respectively, representing intervals that are significantly different ( P < 0.001). Furthermore, the reference intervals are shown to be strongly dependent on the peak integration approach used. Valley‐to‐valley peak integration should only be employed for conditions with complete resolution between disialo‐ and trisialo‐Tf.