Rapid determination of transferrin isoforms by immunoaffinity liquid chromatography and electrospray mass spectrometry

Background: Congenital disorders of glycosylation (CDG) are autosomal reces sive disorders that produce increased serum carbohydrate-deficient transfer rin (CDT) isoforms. Methods to resolve CDT from fully glycosylated tansferr in (Trf) have been based on a neutral shift in the isoelectric focusing (IE F) pattern or on a reduction in the negative charge, allowing resolution by anion-exehange chromatography. Our purpose was to develop a method of reso lution and relative quantification of Trf isoforms using online immunoaffin ity liquid chromatography-mass spectrometry (LC-MS). Methods: Serum (25 muL) was diluted with 100 CLL Of wafer before applicatio n to an immunoaffinity column that sequestered Trf isoforms. Trf isoforms w ere eluted from the immunoaffinity column, concentrated on a C-4 column, el uted from the C-4 column, and introduced into the mass spectrometer. Analys is of the Trf isoforms was entirely automated and completed in <10 min per sample. Results: The LC-MS method demonstrated that the major abnormal Trf isoforms in CDG lack one complete oligosaccharide structure (mono-oligosaccharide) or both oligosaccharide structures (a-oligosaccharide), but not the sialic acids, as presumed on the basis of IEF methods. Calculation of relative rat ios among three possible species (mono-/di-oligosaccharide and a-/dioligosa ccharide) is reproducible [mean intra- and interassay CVs were 9.3% (n = 10 ) and 10% (n = 5), respectively]. A reference range for patients <18 years was determined by analysis of 209 samples (for mono-/di-oligosaccharide, th e median was 0.041 and the range was 0.018-0.083; for a-/di-oligosaccharide , the median was 0.007 and the range was 0.002-0.036). Comparison of data o btained with an affinity chromatography-IEF method and the LC-MS method dem onstrated equivalence in the interpreted results (n = 170). Conclusions: Advantages of the LC-MS method include improved sensitivity, m inimal sample preparation, and an analysis time of <10 min. The method was automated, which allowed high throughput, with >100 samples analyzed in a s ingle day. Moreover, the nature of the oligosaccharide defect in CDG is acc urately reflected by mass resolution, and subtle oligosaccharide truncation s may also be detected by this method. (C) 2001 American Association for Cl inical Chemistry.