Acquired thrombotic thrombocytopenic purpura (TTP) is a life-threatening disorder that results from the development of auto-antibodies against ADAMTS13 disrupting the binding of ADAMTS13 to von Willebrand factor and thereby preventing the proteinase activity and/or increasing the clearance from the circulation. Previous research from our department identified 9 O-linked glycosylation, 6 O-fucosylation and 2 C-mannosylation sites on plasma derived ADAMTS13. One of the N-linked glycosylation sites (N1354) is close to one of the previously identified peptides preferentially presented on HLA-DRB1*0301 and HLA-DRB1*1501 (ASYILIRD amino acid A1355-D1362) and also close to the HLA-DRB1*1101 peptide (FINVAPHAR amino acid F1328-R1336) suggesting a possible role for the glycosylation in the onset of acquired TTP.
To study the glycosylation and glycan trees ADAMTS13 purified from cryosupernatant was reduced with dithiothreitol, alkylated with iodoacetamide and subsequently processed into peptides overnight with either trypsin or chymotrypsin. The peptides were then purified using ZIC-HILIC proteatips and finally analyzed by tandem mass spectrometry employing both higher-energy collision dissociation (HCD) and electron transfer dissociation (ETD). The data files were analyzed using the BYONIC software package as well as manually.
Using this approach we identified the glycan structure on 10 N-linked glycosylation. Nine out of 10 glycans contained complex carbohydrate structures terminating in sialic acid. The glycans at these N-linked sites were identified both with or without a fucose on the primary GlcNAc. We were unable to identify a GalNAc residue in the glycan linked to N614 in the spacer domain. This suggest that the glycan on N614 consist primarily of high mannose structures. Binding of ADAMTS13 to the mannose receptor on dendritic cells is most likely facilitated by the high mannose glycan on N614.
Furthermore we identified 6 O-linked glycosylation sites either on a serine of a threonine. One O-linked glycan is located in the spacer domain, 2 were found in the thrombospondin type 1 repeat-6 (TSP6), another one was found in TSP8 and 1 O-linked glycosylation site was found in both of the CUB domains. Four out of 6 O-glycans contained terminal sialic acid of which 2 also contained a fucose attached to the GlcNAc. Several O-glycans contained a terminal galactose residue; one O-glycan in TSP6 terminated in both a GlcNAc and a GalNAc residue.
O-fucosylation is a common post-translational modification of thrombospondin type 1 repeats. We identified 9 O-fucosylation sites in the TSP repeats. Seven out of 9 sites adhered to the consensus sequence previously defined for O-fucosylation. TSP1 and 2 contained an additional O-fucosylation site at residues T407 and S724; these sites did not match the consensus sequence for O-fucosylation. Interestingly, two additional O-fucosylation sites were identified in cysteine rich and spacer domain at residue S553 and S698. All these residues were predicted to contain a glucose-fucose modification. Next to these glucose-fucose modifications we also identified 2 fucose modification in both of the CUB domains at residues S1170 and T1344. These results show that ADAMTS13 is extensively modified by O-fucosylation.
Evidence for C-mannosylation of 8 different tryptophans was obtained. In accordance with previous findings the W387 or W390 (TSP1) and W884 (TSP4) were found to be C-mannosylated. We also found C-mannosylated tryptophans at position and W730 (TSP2) and W1081 (TSP8). Four additional C-mannosylated tryptophans were detected at position W208 (metallo proteinase domain), W1307 (CUB1 domain) and W1379 and W1406 (CUB2 domain). These results show that C-mannosylation is a common post translational modification in ADAMTS13 that is also found outside the TSP domains.
Taken together these findings highlight the extensive post translational modification of ADAMTS13 by diverse carbohydrate structures. We anticipate that our findings might be relevant for the clearance and/or immune recognition of ADAMTS13.
No relevant conflicts of interest to declare.
Asterisk with author names denotes non-ASH members.