A high-throughput O-glycopeptide discovery platform for seromic profiling

Abstract

Biomarker microarrays are becoming valuable tools for serological screening of disease-associated autoantibodies. Post-translational modifications (PTMs) such as glycosylation extend the range of protein function, and a variety of glycosylated proteins are known to be altered in disease progression. Here, we have developed a synthetic screening microarray platform for facile display of O-glycosylated peptides (O-PTMs). By introduction of a capping step during chemical solid-phase glycopeptide synthesis, selective enrichment of N-terminal glycopeptide end products was achieved on an amine-reactive hydrogel-coated microarray glass surface, allowing high-throughput display of large numbers of glycopeptides. Utilizing a repertoire of recombinant glycosyltransferases enabled further diversification of the array libraries in situ and display of a new level of potential biomarker candidates for serological screening. As proof-of-concept, we have demonstrated that MUC1 glycopeptides could be assembled and used to detect autoantibodies in vaccine-induced disease-free breast cancer patients and in patients with confirmed disease at time of diagnosis.

Figure 1

Chemical synthesis and on-slide / on-bead capture of glycopeptides. (A) Solid-phase synthesis of 20mer glycopeptides with N-acetylation capping steps. (B) On-slide enrichment of crude product via amine-reactive microarray glass surface. Exemplified Tn-MUC1a glycopeptide printed in ten replicates at two-fold dilutions were detected with biotinylated lectin (VVA) followed by fluorescent labeled Streptavidin-AlexaFluor488 (C) On-bead capture, release (DTT) and MALDI-TOF spectroscopy analysis.

Figure 2

Fine specificities of Tn-MUC1 reactive monoclonal tumor antibodies. Binding of mouse monoclonal antibodies 5E5, 2D9 and VU3C6 followed by detection with Cy3-labeled anti-mouse-IgG. Tn-glycosylated amino acids are displayed in red. Spot-to-spot variations for three replicates of each compound are represented by the error bars.

Figure 3

Epitope mapping of vaccine induced human MUC1 specific antibodies. All sera were diluted 1:20 followed by detection with Cy3-labeled anti-human-IgG (1:1000). Spot-to-spot variations for three replicates of each compound are represented by the error bars. (A) Fine-specificities of selected post-vaccinated patients 1–3. Tn-glycosylated amino acids are displayed in red. (B) Pre-vaccinated patients. (C) Post-vaccinated patients.

Figure 4

Rapid expansion of array and bead glycopeptide libraries by enzymatic glycosylation. (A) Analysis of substrate site-specificity by on-slide glycosylations of immobilized MUC1a glycopeptides (1) unglycosylated MUC1a, (2) mono-Tn-MUC1a and (3) bis-Tn-MUC1a with GalNAc-T2. Sites of incorporation were verified with Tn specific mAbs 5E5 and 2D9. (B) The extent of the glycosylation of peptide (1) from panel (A) was monitored by on-bead glycosylation, and analyzed by MALDI-TOF spectroscopy after DTT cleavage. (C) Comparison of GalNAc-T1, -T2, and -T3 substrate specificities in solution and on array. (D) Sequences of the peptides analysed in Panels A–C.

Figure 5

On-slide enzymatic glycosylation and detection of serum autoantibodies. (A) Image display after on-slide enzymatic glycosylation using core3 β3GlcNAc-T6 (UDP-GlcNAc) (i) and ST6GalNAc-I (CMP-NeuAc) (ii) followed by subsequent staining with sera (1:20) and Cy3 labeled anti-human-IgG. (B) Glycoform specific MUC1 epitope autoantibody reactivity in four different cancer sera. Predominant MUC1 epitope for each sera is depicted as a bold sequence in each serum panel. (C) Detection of cancer sera and healthy sera on a panel of selected MUC1 20mers and MUC1 60mer controls. Spot-to-spot variations for three replicates of each compound are represented by the error bars.