Glycans of plasma ADAMTS13

In this issue of Blood, Verbij et al identified the sites of glycosylation in plasma ADAMTS13 (a disintegrin and metalloproteinase with a thrombospondin type 1 motif, member 13) and determined the composition of the glycan structures at these sites.¹ 

ADAMTS13 is a blood enzyme that controls the multimer size of the hemostatic protein von Willebrand factor.²  After synthesis, ultralarge, hyperreactive von Willebrand factor multimers (up to 20 000 kDa) are secreted into the flowing blood and are immediately cleaved by ADAMTS13 into smaller more quiescent multimers (<10 000 kDa). When ADAMTS13 is deficient, patients suffer from the devastating thrombotic thrombocytopenic purpura (TTP) disorder.³  In TTP patients, ultralarge von Willebrand factor multimers spontaneously bind platelets, and microthrombi are formed that block arterioles and capillaries. This results in severe organ failure, thrombocytopenia, and hemolytic anemia. TTP can be caused by mutations in the ADAMTS13 gene (congenital TTP) or by the development of autoantibodies against ADAMTS13 (acquired TTP).

ADAMTS13 is a multidomain enzyme consisting of 1427 amino acids. Plasma ADAMTS13 is heavily glycosylated (20%) and has an apparent molecular weight of 180 to 190 kDa.²  It is well known that glycosylation plays an important role in many processes such as immune recognition of proteins and protein folding, final structure, secretion, function, and eventual clearance. Only a few studies have investigated the role of glycosylation in ADAMTS13 folding and secretion and in ADAMTS13 function. It has been shown that recombinant ADAMTS13 contains N-linked glycosylation and O-fucosylation sites.^4,5  Both N-linked glycosylation and O-fucosylation seemed to be crucial for proper folding in the heterologous cells and for efficient secretion of recombinant ADAMTS13.^4,5  However, when N-linked glycans were removed from recombinant ADAMTS13, the proteolytic activity of ADAMTS13 was not altered.⁵ 

The glycan profile of recombinant ADAMTS13 might to some extent differ from the glycan structures of plasma ADAMTS13. Hence, to understand the role of glycans in ADAMTS13 biology and pathophysiology, it is crucial to unravel the glycosylation profile of plasma ADAMTS13. Verbij et al used the elegant approach of tandem mass spectrometry with higher-energy collision dissociation and electron transfer dissociation to complete this challenging task. Importantly, they were able to identify or confirm the amino acids that carry ADAMTS13 glycans, and they were also able to unravel the composition of each glycan chain. By using this knowledge, the complete structure of all ADAMTS13 glycosylation chains could be deduced. This work led to 3 categories of glycan structures on plasma ADAMTS13: complex N-linked carbohydrate structures, less complex O-(GalNAc)-linked glycan structures, and simple O-linked fucose and C-linked mannose glycans. Nine of the 10 N-linked glycans are composed of 11 to 13 monosaccharides, including a terminal sialic acid. However, 1 N-linked glycan (8 monosaccharides), situated in the spacer domain, is not sialylated but contains a high mannose structure. The 6 O-(GalNAc)-linked glycans consist of 4 to 7 monosaccharides, again including a terminal sialic acid. Typical for thrombospondin type 1 (TSP) repeats, and with the exception of TSP4, all 7 remaining TSP domains are O-fucosylated with disaccharide structures. Unexpectedly, another O-fucosylation site was identified in the disintegrin domain. Finally, the TSP1, -4, and -7 domains are each C-mannosylated with a single mannose residue.

Knowledge of the glycosylation profile of proteins allows better understanding of the role of glycans in the biology and pathophysiology of proteins. For example, glycans could control the structure of ADAMTS13. The crystal structure of only the disintegrin-like domain/first TSP repeat/cysteine-rich domain/spacer domain fragment of ADAMTS13 is known.⁶  However, it was recently shown that the ADAMTS13 spacer domain interacts with its CUB1-2 domains, suggesting that ADAMTS13 adopts a folded conformation.^7,8  Thus, it will be interesting to unravel whether the glycans in these domains contribute to the stabilization of this overall folded structure of ADAMTS13. In addition, if glycans stabilize the structure of ADAMTS13, then changes in glycosylation patterns, which could occur spontaneously as a consequence of pathological processes,⁹  might lead to different ADAMTS13 conformations. These different conformations might be more prone to proteolysis that renders ADAMTS13 inactive, which could be an explanation of why lower levels of ADAMTS13 activity are detected in certain diseases.¹⁰  In addition, alterations in protein glycosylation may modify or create novel B-cell epitopes.⁹  It has been suggested that changes in glycosylation of ADAMTS13 could expose neo-epitopes which could explain the formation of autoantibodies in acquired TTP. Verbij et al also hypothesized that high-mannose glycans identified in plasma ADAMTS13 contribute to the binding of ADAMTS13 to the mannose receptor on dendritic cells. Whether glycans in ADAMTS13 indeed contribute to all these processes remains to be determined.