Cross-platform comparison of glycan microarray formats

Abstract

Carbohydrates participate in almost every aspect of biology from protein sorting to modulating cell differentiation and cell-cell interactions. To date, the majority of data gathered on glycan expression has been obtained via analysis with either anti-glycan antibodies or lectins. A detailed understanding of the specificities of these reagents is critical to the analysis of carbohydrates in biological systems. Glycan microarrays are increasingly used to determine the binding specificity of glycan-binding proteins (GBPs). In this study, six different glycan microarray platforms with different modes of glycan presentation were compared using five well-known lectins; concanavalin A, Helix pomatia agglutinin, Maackia amurensis lectin I, Sambucus nigra agglutinin and wheat germ agglutinin. A new method (universal threshold) was developed to facilitate systematic comparisons across distinct array platforms. The strongest binders of each lectin were identified using the universal threshold across all platforms while identification of weaker binders was influenced by platform-specific factors including presentation of determinants, array composition and self-reported thresholding methods. This work compiles a rich dataset for comparative analysis of glycan array platforms and has important implications for the implementation of microarrays in the characterization of GBPs.

Keywords: comparison; glycan microarray; lectin; presentation.

Fig. 1.

Graphic representation of glycan array formats. Array 1-CFG; Array 2-Huflejt; Array 3-Gildersleeve; Array 4-Joshi; Array 5-Reichardt; Array 6-Pieters. Gray surface: Nexterion H^® (NHS) slide, pink surface: epoxide slide, green surface: PamChip^® (maleimide). Representative structures are shown.

Fig. 2.

Venn diagram of determinants for Arrays 1–3. Number of determinants (overlapping and unique) are indicated.

Fig. 3.

Representative images of six glycan arrays for WGA binding. Microarrays are not to scale and, in some cases, are false-colored dependent on the platform. Due to differences in platforms, concentrations of biotinylated WGA used vary: Arrays 1 and 3 (10 μg/mL WGA), Array 2 (100 μg/mL), Array 4 (0.8 μg/mL), Array 5 (1 μg/mL) and Array 6 (5 μg/mL).

Fig. 4.

The process of identifying positive binders using the universal threshold.

Fig. 5.

Binding profile of WGA on Array 6.