O-Glycosylation of thrombospondin type 1 repeats : identification and characterisation of β1,3-glucosyltransferase

Rare types of protein glycosylation often occur in a domain-specific manner required for important
biological functions (Okajima et al. 2008). O-linked fucosylation, where the fucose is linked directly to
the hydroxyl groups of serine or threonine residues, has so far only been reported on three distinct
protein modules: epidermal growth factor (EGF)-like domains; thrombospondin type 1 repeats (TSRs);
and on the protease inhibitor PMP-C. The function of O-fucosylation is well studied on EGFs, where
changes in the O-fucose glycan on Notch EGFs alter Notch signalling (reviewed in Okajima et al.
2008).
TSRs are modified by the unusual Glc-β1,3-Fuc-O- disaccharide (Hofsteenge et al. 2001;
Gonzalez de Peredo et al. 2002). Protein O-fucosyltransferase 2 (POFUT2) initiates the fucosylation
of properly folded TSRs. Deletion of POFUT2 in C. elegans causes an early dorsal migration
phenotype of the anterior distal tip cell (Canevascini et. al, manuscript in preparation). Recently it was
shown that Pofut2 null embryos die by mid-gastrulation (Du et al. 2007). Further, O-fucosylation of
TSRs in ADAMTS13 (Ricketts et al. 2007) and ADAMTS-like 1/punctin (Wang et al. 2007) is required
for their secretion in mammalian cells. Since these experiments abolished the fucosylated glycan,
Glc-β1,3-Fuc-O-, it remains to be determined what role the glucose plays. Therefore, the β1,3-
glucosyltransferase (β3Glc-T) that catalyses the last step in the formation of this disaccharide, needed
to be identified and characterised.
In order to identify the β3Glc-T a relevant enzyme substrate in the form of properly folded
TSR-fucose needed to be produced. I describe a reproducible purification strategy for the production
of a large amount of highly pure, correctly folded TSR and fucosylated TSR domain (Chapter 6.1).
Using this TSR-fucose substrate in a specific radiochemical assay, I was able to identify the β3Glc-T
as a member of the glycosyltransferase family GT31, which includes Fringe, the enzyme that modifies
O-fucose in EGF repeats. The cloned β3Glc-T specifically catalyses the transfer of glucose from
UDP-glucose to TSR-fucose to yield Glc-β1,3-Fuc-O- (Chapter 6.2). There is no reactivity towards
non-fucosylated substrates or fucosylated EGF. The β3Glc-T protein consists of an N- and C-terminal
domain of approximately equal size that share internal sequence homology. Within the C-terminal
domain putative catalytic centre, several key residues were identified (Asp-349, Asp-351, Asp-421) for β3Glc-T activity (Chapters 6.2 and 6.3). In addition, the removal of either of the two N-glycosylation
sites in β3Glc-T resulted in diminished activity (Chapter 6.3). The β3Glc-T N-terminal domain does not
possess intrinsic β3Glc-T activity but affects its activity. Expression of the β3Glc-T C-terminal domain
requires the co-expression of the β3Glc-T N-terminal domain. I suggest that the β3Glc-T N-terminal
domain may function as a stabiliser or internal chaperone (Chapter 6.4). Moreover, it was recently
shown that mutations in the gene encoding β3Glc-T cause the severe developmental disease, Peters
Plus syndrome (PPS) (Lesnik Oberstein et al. 2006). Constructs encoding these mutated β3Glc-T
sequences express truncated proteins lacking the catalytic domain (Chapter 6.4). This raises the
possibility that the truncated β3Glc-T proteins may play a role in PPS.
The work presented in this thesis provides the basis for further studies on the role of β3Glc-T
in regulating TSR biological function, through glycosylation, in health and disease.