Cross Reactive Material 197 glycoconjugate vaccines contain privileged conjugation sites

Abstract

Production of glycoconjugate vaccines involves the chemical conjugation of glycans to an immunogenic carrier protein such as Cross-Reactive-Material-197 (CRM197). Instead of using glycans from natural sources recent vaccine development has been focusing on the use of synthetically defined minimal epitopes. While the glycan is structurally defined, the attachment sites on the protein are not. Fully characterized conjugates and batch-to-batch comparisons are the key to eventually create completely defined conjugates. A variety of glycoconjugates consisting of CRM197 and synthetic oligosaccharide epitopes was characterised using mass spectrometry techniques. The primary structure was assessed by combining intact protein MALDI-TOF-MS, LC-MALDI-TOF-MS middle-down and LC-ESI-MS bottom-up approaches. The middle-down approach on CNBr cleaved glycopeptides provided almost complete sequence coverage, facilitating rapid batch-to-batch comparisons, resolving glycan loading and identification of side products. Regions close to the N- and C-termini were most efficiently conjugated.

Figure 1. Schematic overview of the three orthogonal approaches employed for the in-depth characterisation of CRM₁₉₇ conjugated with defined synthetic minimal glycoepitope vaccine candidates.

Step 1: The glycan part is synthesised with a spacer carrying an amine group. Step 2: addition of the linker molecule with two leaving groups (LG). Step 3: addition of the glycan linker construct to the carrier protein and subsequent conjugation to a primary amine (e.g. lysine residue).

Figure 2. Conjugation site determination by tandem mass spectrometry.

(A) After Glu-C digestion a doubly charged precursor ion corresponding to glycopeptide ²⁴²KAKQYLEE²⁴⁹ carrying an ST3 epitope was detected and selected for fragmentation by CID and ETD. The red labeled K indicates the site of conjugation identified by tandem MS. (B) CID fragmentation resulted in a prominent Y-ion series indicating the loss of glucuronic acid and glucose, but no significant peptide backbone fragments. (C) ETD fragmentation of the peptide backbone confirmed peptide identity as well as K242 as the site of conjugation. (D) Extracted ion chromatogram of doubly charged ions of unconjugated and conjugated peptide 242–249 showing the separation of isobaric conjugate products by C18 LC [overlay of m/z = 504.40 (unconjugated peptide, black line), 568.85 (peptide with conjugated linker, blue line) and 928.45 (peptide with conjugated glycan, red line)]. Despite the proximity of the two potential conjugation sites, K242 was more effectively conjugated.

Figure 3. MS spectra of the doubly charged precursor ions m/z = 787.90 and m/z = 780.92 as well as their respective CID MS/MS spectra.

Using CID fragmentation even minor amounts of synthesis byproducts containing glucose instead of a glucuronic acid at the second position could be identified after being conjugated to CRM₁₉₇.

Figure 4. Overlay of all LC-MALDI-MS spectra acquired during the LC run of unconjugated CRM₁₉₇.

For each identified peak charge, peptide position and average mass are indicated. The major peaks are all corresponding to the singly charged CNBr peptides without any missed cleavages. All of the eight expected peptides were detected. Doubly charged species were also found for larger peptides as well as low abundant signals corresponding to peptides with missed cleavage sites. The survey view illustrates retention time differences of the various peptides.

Figure 5. Comparison of conjugation efficiency on peptides 231–314 and 460–535 shown for four samples carrying different glycoepitopes.

In the SurveyViewer (retention time vs. m/z) of a (C4) LC-MALDI-MS experiment (left) intensities of the detected signals are plotted vs. their retention times and signal intensities in a gel-like representation, providing a quick overview on the number of conjugates present on each CNBr peptide (right). Spectra obtained for peptide 460–535 at a given retention time. Samples conjugated with the DSA-linker appeared in clusters with varying amounts of linker and glycoconjugates. Linker addition induced a shift of + 128 Da and minor shifts in retention time. A single glycoconjugate addition increased the mass by 848 Da (ST3), 333 Da (GLC), or 991 Da (PS1), respectively. Major differences in loading of free linker and glycans were easily detectable by this approach. ST3 batch-to-batch comparison revealed batch A containing up to one glycan conjugated to peptide 231–314 and up to two conjugated to peptide 460–535, whereas the loading was increased in batch B with up to two glycoconjugates attached to peptide 231–314 and up to three to peptide 460–535, respectively. The lower glycan loading in batch A correlated with stronger signals for linker conjugation. GLC was conjugated via a PNP-linker and showed no detectable free linker additions and therefore no clusters as seen for the DSA-conjugated peptides.

Figure 6. Relative quantitation of the seven lysine containing CNBr peptides derived from two ST3 batches.

Each peptide section was quantified individually. The sum of the area under the curve from all detected signals obtained for a given peptide (with and without modifications) was set to 100 %. The percentages presented in the graph are rounded to the full digit. Differences in glycan and free linker loading could be directly compared for each peptide. The results clearly demonstrated that glycan loading was more effective in batch B over the entire protein sequence. Inversely batch A exhibited higher loading of free linker molecules.

Figure 7. Exemplary MALDI-ISD spectra of CRM₁₉₇ peptide 231–314 acquired from the unconjugated control (top) and an ST3 conjugated batch (bottom) after separation by C8 chromatography.

An m/z + 848 shift of the c_12-ion observed between the top and bottom spectra indicated K242 to be the major modified lysine residue in this peptide. The ISD-spectra also provided extensive amino acid sequence data of the peptide.

Figure 8

(Left): Heatmap showing the lysine residues detected to be conjugated (red label) after proteolytic digestion with trypsin, Glu-C or combinations of trypsin/Glu-C or Glu-C/Asp-N. Grey labelled lysine residues were not observed as being conjugated. Data obtained for four independent conjugates carrying a variety of glycoepitopes (LPG, ST3, PS1 and GLC as represented in each column) showed that several lysine residues were commonly found conjugated. (Right) 3D crystal structure of CRM₁₉₇ dimer (PDB entry: 4AE0). Lysine residues are labelled blue. Lysine residues that were frequently found conjugated with a glycan in all the samples are labelled in red.

Figure 9. Quantitation results from LC-MALDI-MS measurements.

(A) Quantitation of unconjugated peptides from of ST3 batch A, ST3 batch B and PS1 (only peptides containing lysine residues were taken into account). Peptides 15–115 and 460–535 show the lowest amounts of unconjugated peptide, consequently carrying larger amounts of glycoconjugates and/or free linker, indicating that conjugation was more efficient at lysine residues present in these peptides. (B) Quantitation of two batches of GLC (batch a and batch b) conjugated via the PNP-linker. Only glycan conjugates were found to be present on the peptides. Again, peptides 15–115 and 460–535 show the highest loading.