Dehui Sun

Traditional phenotype-based screening for β-globin variant and β-thalassemia using hematological parameters is time-consuming with low-resolution detection. Development of a MALDI–TOF–MS assay using alternative markers is needed. We constructed a MALDI–TOF–MS-based approach for identifying various β-globin disorders and classifying thalassemia major (TM) and thalassemia intermedia (TI) patients using 901 training samples with known HBB/HBA genotypes. We then validated the accuracy of population screening and clinical classification in 2 separate cohorts consisting of 16 172 participants and 201 β-thalassemia patients. Traditional methods were used as controls. Genetic tests were considered the gold standard for testing positive specimens. We established a prediction model for identifying different forms of β-globin disorders in a single MALDI–TOF–MS test based on δ- to β-globin, γ- to α-globin, γ- to β-globin ratios, and/or the abnormal globin-chain patterns. Our validation study yielded comparable results of clinical specificity (99.89% vs 99.71%), and accuracy (99.78% vs 99.16%) between the new assay and traditional methods but higher clinical sensitivity for the new method (97.52% vs 88.01%). The new assay identified 22 additional abnormal hemoglobins in 69 individuals including 9 novel ones, and accurately screened for 9 carriers of deletional hereditary persistence of fetal hemoglobin or δβ-thalassemia. TM and TI were well classified in 178 samples out of 201 β-thalassemia patients. MALDI–TOF–MS is a highly accurate, predictive tool that could be suitable for large-scale screening and clinical classification of β-globin disorders.

Glycosylation is a post-translational modification essential for maintaining the structure and function of proteins. Abnormal N-glycan patterns have been found in various diseases compared to healthy controls. A decrease in terminal galactosylated N-glycans of serum IgG in rheumatoid arthritis (RA) and osteoarthritis (OA) may be involved in their immunopathogenesis. However, how glycan patterns differ between RA and OA remains unclear. Here, we identified 15 glycan forms of serum IgG from RA and OA using MALDI-TOF MS. We found that IgG galactosylation represented a suitable candidate for differentiating RA from healthy controls (AUC > 0.9). Then, we performed binary logistic regression to screen out three bisecting N-acetylglucosamine (GlcNAc) glycoforms for distinguishing between OA and RA. Combined ROC analysis of the selected glycans yielded an AUC of 0.81 between OA and RA and an AUC of 0.79 between OA and RF/ACPA negative RA. Similar results were found in the validation set. In conclusion, our analysis demonstrates that RA and OA are distinguished on the basis of their different IgG glycan patterns, which thus serve as suitable candidates as biomarkers for reliably identifying clinical conditions such as RA and OA.