O. Letourneur et al., Characterization of Toxoplasma gondii surface antigen I (SAGI) secreted from Pichia pastoris: evidence of hyper O-glycosylation, BIOT APP B, 33, 2001, pp. 35-45
A truncated form of surface antigen I (SAGIt), the immunodominant surface a
ntigen of Toxoplasmo gondii, was expressed in the methylotrophic yeast, Pic
hia pastoris. The truncated protein lacked the C-terminal residues which, i
n native SACI, encompass a glycosylphosphatidylinositol anchorage site. The
single potential N-glycosylation site was mutated and a sequence encoding
a hexahistidine tag was introduced at the C-terminal of the construction to
aid purification by immobilized metal-chelate chromatography. Recombinant
SAG It was efficiently secreted into the culture medium as three protein sp
ecies having molecular masses of 29, 38/45 and 50/60 kDa. This heterogeneit
y was dependent upon the composition of the medium used to grow the yeast t
ransformants. Mass spectrometric analyses, chemical deglycosylation, lectin
recognition and sensitivity to mannosidase treatments showed that SAGIt he
terogeneity was related to the presence of O-linked oligosaccharides contai
ning alpha1-2-, alpha1-3- or alpha1-6-linked mannoses. The glycosylated and
deglycosylated recombinant SAGIt were recognized by monoclonal and human-s
erum-derived antibodies, specific for the native SAGI, which suggested that
the O-glycosylations had no major effect on the protein conformation. Howe
ver, ELISA and Western-blot analysis with human sera showed that the O-carb
ohydrates added by P. pastoris could be recognized as antigenic structures.
As a consequence, purification of the unglycosylated 29 kDa recombinant SA
G It species or deglycosylation is required in order to use recombinant SAG
I as a diagnostic reagent. Moreover, the presence of carbohydrates, not fou
nd on the native protein, suggests that addition of unnatural glycan struct
ures by P. pastoris is a potential drawback that should be considered when
using this expression system.