Serum N-Glycans: A New Diagnostic Biomarker for Light Chain Multiple Myeloma

Abstract

The aim of this study was to evaluate the diagnostic and differential diagnostic power of serum N-glycans for light chain multiple myeloma (LCMM). A total of 167 cases of subjects, including 42 LCMM, 42 IgG myeloma, 41 IgA myeloma, and 42 healthy controls were recruited in this study. DNA sequencer-assisted fluorophore-assisted capillary electrophoresis (DSA-FACE) was applied to determine the quantitive abundance of serum N-glycans. The core fucosylated, bisecting and sialylated modifications were analyzed in serum of LCMM patients (n=20) and healthy controls (n=20) randomly selected from the same cohort by lectin blot. Moreover, serum sialic acid (SA) level was measured by enzymatic method. We found two N-glycan structures (NG1A2F, Peak3; NA2FB, Peak7) showed the optimum diagnostic efficacy with area under the ROC curve (AUC) 0.939 and 0.940 between LCMM and healthy control. The sensitivity and specificity of Peak3 were 88.1% and 92.9%, while Peak7 were 92.9% and 97.6%, respectively. The abundance of Peak3 could differentiate LCMM from IgG myeloma with AUC 0.899, sensitivity 100% and specificity 76.2%, and Peak7 could be used to differentiate LCMM from IgA myeloma with AUC 0.922, sensitivity 92.9% and specificity 82.9%. Serum SA level was significantly higher in patients with LCMM than that in healthy controls. Moreover, the decreased core fucosylation, bisecting and increased sialylation characters of serum glycoproteins were observed in patients with LCMM. We concluded that serum N-glycan could provide a simple, reliable and non-invasive biomarker for LCMM diagnosis and abnormal glycosylation might imply a new potential therapeutic target in LCMM.

Fig 1. A representative N-glycan profiling of total serum glycoproteins.

At least 12 peaks can be identified by DNA sequencer-assisted fluorophore-assisted capillary electrophoresis (DSA-FACE) technology in four groups. The structure of the N-glycan peaks and the symbols indicated the corresponding glycans are displayed below the panel.

Fig 2. Peak abundances for the N-glycans which had the obvious overall changes among control subjects, moderate and severe LCMM subjects.

Peak3, Peak4 and Peak6 decreased, whereas Peak5 increased associated with the severity of LCMM disease. Error bars are 95% confidence interval (95% CI) of the means.

Fig 3. Peak3 and Peak7 could be used as diagnosis and differential diagnosis markers for LCMM.

(A and B): showed the changes of two biantennary glycans (NG1A2F, Peak 3; and NA2FB, Peak 7) among 4 groups and error bars represent 95% confidence intervals (95% CI) of the means. Peak3 and Peak7 were decreased in LCMM group compared with the other 3 groups, except that Peak 3 was elevated in patients with LCMM relative to patients with IgA MM. (C): ROC curves for Peak 3 and Peak7 in patients with LCMM versus healthy controls with the area under the curve (AUC) of 0.939 and 0.940, respectively. (D): ROC curve for Peak3 in patients with LCMM versus IgG MM and the AUC was 0.899. (E): ROC curve for Peak7 in patients with LCMM versus IgA MM with the AUC of 0.922.

Fig 4. The abundance of core fucosylation and bisecting modification in N-glycan and glycoprotein levels.

The abundance of core fucosylation and bisecting modification in N-glycan and glycoprotein levels were illustrated using DSA-FACE and serum lectin blot. (A) and (B): The abundance of total core fucosylation residues (the sum of Peak1, 2, 3, 4, 6, 7 and Peak10) from DSA-FACE decreased in LCMM patients (n = 42) than that in healthy controls (n = 42), and decreased according to the severity of the disease [moderate LCMM (n = 14) and severe LCMM (n = 28)]. (C) and (D): LCA binding core-fucosylated glycoprotein level detected by serum lectin blot was decreased in LCMM patients (n = 20) than that in healthy controls (n = 20). The level of serum core-fucosylated glycoprotein was significantly lower in severe LCMM (n = 13) but not in moderate LCMM (n = 7). The vertical axis indicates the ratio of fluorescence intensity of fucosylated proteins to total proteins stained by coomassie blue (CBB). (E) and (F): The abundance of total bisecting residues (the sum of Peak2 and Peak7) from DSA-FACE deceased in LCMM patients than that in healthy controls, and decreased associated with the severity of disease. (G) and (H): PHA-E binding bisecting glycoprotein level detected by lectin blot was decreased in LCMM patients and decreased associated with the severity of disease. The abundance on vertical axis was represented as mean ± standard deviation.

Fig 5. The levels of serum salic acid (SA) and sialylated glycoproteins.

The increased levels of serum salic acid (SA) and sialylated glycoproteins in LCMM patients were clarified by enzymatic method and using lectin blot, respectively. (A): The horizontal axis represents the experimental groups: control (n = 42), LCMM (n = 42). The vertical axis indicates the serum SA level (mg/dl), and error bars are mean ± standard deviations. (B) and (C): Lectin blots of serum glycoproteins were probed with MAL II and SNA representing the α-2,3-linked and α-2,6-linked sialylated proteins. The horizontal axis represents the experimental groups: control (n = 20) and LCMM (n = 20). The vertical axis indicates the fluorescence intensity ratio of MAL II or SNA binding proteins to total proteins stained by coomassie blue (CBB), and error bars are mean ± standard deviations. MAL II binding α-2,3-linked sialylated protein increased in LCMM patients, whereas there was no significant change in SNA binding α-2,6-linked sialylated proteins between LCMM patients and controls.