Molecular and domain features of the single-pass type I membrane GP900
The glycoprotein GP900 (1937 aa) is a single-pass membrane protein consisting of an N-terminal signal peptide (SP; positions 1–26 aa) and a single transmembrane domain (TMD; 1863–1887 aa) near the C-terminus (Fig. 1). The presence of an SP and a TMD indicates that GP900 is a type I membrane protein in the secretory pathway, in which GP900 is translated on the endoplasmic reticulum (ER) with N-terminal domain inserted into the ER lumen for initial glycosylation and further glycosylation in Golgi body. The SP is cleaved in the ER after the completion of translation. It predicts that the long N-terminal domain (~ 1837 aa) and very short C-terminal domain (~ 50 aa) are non-cytoplasmic and cytoplasmic, respectively. GP900 as a secretory glycoprotein was also confirmed by its presence in the secretory organelle micronemes in sporozoites and merozoites by IFA and IEM in earlier studies [1–3], and in this study as well (see later sections for detail).
GP900 contains a large number of Thr (n = 503; 26%) and Ser (n = 155; 8%) residues, including two large Thr-rich stretches, that are heavily O-glycosylated, and an average number of Asn residues (n = 89; 4.6%) that are partially N-glycosylated (Fig. 1, marked in green or blue) [5, 6]. GP900 is an intrinsic disorder protein by possessing 10 intrinsic disorder regions across the entire protein (Fig. 1, marked in brown), suggesting its overall structure flexibility. GP900 also possesses 11 Cys residues (Fig. 1, marked in red), all located on the N-terminal half of the protein, suggesting some intramolecular (and potentially intermolecular) cross-links by disulfide bonds to increase the stability of the N-terminal half of the protein.
The transcript of GP900 gene could be detected in oocysts, free sporozoites and intracellular developmental stages (Fig. 2). The transcript levels were relatively high in oocysts and sporozoites when normalized with Cp18S transcript, but varied dramatically in intracellular stages. The lowest levels were observed shortly after invasion (i.e., 3 and 6 hpi; representing early development of meronts), elevated to the highest at 12 hpi (representing the maturation and formation of merozoites in the first generation of merogony), and then steadily declined from 24 hpi to 48 hpi and 72 hpi (representing much less synchronized second and third generations of merogony and gametogenesis).
The C-terminal cytoplasmic domain of GP900 is cleaved in the secretory pathway before reaching to cell surface in C. parvum sporozoites
GP900 was known to be “stored in micronemes prior to appearance on the surface of invasive forms” and discharged to form trails during the gliding of sporozoites [20, 21]. We were interested in confirming the relocation of GP900 from micronemes to cell surface and testing whether this single-pass membrane protein was present on the sporozoite surface as a transmembrane protein in the cytoplasm membrane. Therefore, we raised two antibodies to distinguish the C-terminal cytoplasmic domain (rabbit anti-GP900-C pAb) from the N-terminal non-cytoplasmic domain (mouse anti-GP900-N mAb) (Fig. 1; dark red bars). Both anti-GP900-C pAb (affinity-purified) and anti-GP900-N mAb (prepared in the mouse ascites) recognized a single high molecular weight band above the 250 kDa marker in western blot analysis (Fig. 3). The immunoblotting data was similar to those observed in earlier studies [1–3], confirming the specificity of the two antibodies.
IFA using anti-GP900-C antibody confirmed the protein location in the apical region packed with micronemes in the sporozoites before and after excystation (Fig. 4A, 4B). In this assay, immunolabeling intact oocysts was obstructed by the impermeability of oocyst walls, but resolved by rupturing oocysts with freeze/thaw cycles to allow the access of antibodies to sporozoites . The anti-GP900-N antibody also chiefly labeled the apical region in sporozoites before excystation (as indicated by the presence on one side of the nuclei) (Fig. 4C). However, in excysted sporozoites, anti-GP900-N antibody not only strongly labeled the apical region, but also along the pellicle with weaker but clear signals (Fig. 4D). The labeling of anti-GP900-N antibody on the surface of sporozoites were confirmed by IFA of formaldehyde-fixed, but non-permeabilized sporozoites (Fig. 5A), in which slightly weaker but clear fluorescent signals were observable on the surface of sporozoites, but no signals were detected inside of the sporozoites. For direct comparison, the fluorescent signals from permeabilized sporozoites were detected along the pellicles and inside of the sporozoites that were processed in parallel under the same conditions (Fig. 5B). Furthermore, in the gliding sporozoites, anti-GP900-N antibody again labeled sporozoite pellicles and the gliding trails of some sporozoites (Fig. 5C), whereas anti-GP900-C antibody only labeled the apical region of sporozoites (Fig. 5D).
To give a short summary of the IFA data, GP900 present in the micronemes of the sporozoite apical region contained both C-terminal cytoplasmic and N-terminal non-cytoplasmic domains (i.e., full length GP900, minus signal peptide in theory), whereas the GP900 protein distributed along the sporozoite surface or pellicles contained only the N-terminal non-cytoplasmic domain. The gliding sporozoites also discharged GP900 that lacking the C-terminal cytoplasmic domain. These observations concluded that the C-terminal domain of GP900 was cleaved in the secretory pathway before it reached to the cell surface.
GP900 is present in asexual developmental stages, but absent in sexual stages
During the sporozoite invasion of host cells, anti-GP900-C antibody again labeled apical region in the long sporozoites (early invasion), but became two dense spots in short sporozoites (later invasion) or two semi-circles surrounding the newly formed trophozoites (Fig. 6A). Anti-GP900-N antibody also labeled apical region (more strongly) plus the pellicles in long sporozoites, and became more diffused in short sporozoites and newly formed trophozoites (Fig. 6B). During C. parvum intracellular development, immunostaining by anti-GP900-C antibody was apparent (Fig. 7A, green), showing signals surrounding nuclei in small meronts with one nucleus (Fig. 7A, upper two panels) or strong signals surrounding the large developing meronts with multiple nuclei (Fig. 7A, lower left panel). In mature meronts in which banana-like merozoites were formed, fluorescent signals became concentrated on one side of the nuclei that implied apical distributions in merozoites (Fig. 7A, lower right panel).
Based on co-staining with PVM using a mouse polyclonal antibody against total C. parvum proteins (Fig. 7A, red), the immunostaining by anti-GP900-C antibody was confined to the parasites within the PVM, rather than on the PVM. In IFA using anti-GP900-N antibody (Fig. 7B), uneven signals were observed in mature meronts, implying distributions on the apical region of merozoites, but the signals appeared weak in trophozoites and developing meronts. However, we also observed high background/non-specific signals from host cells using anti-GP900-N antibody that could not be resolved after multiple attempts of changing experimental conditions. In fact, anti-GP900-C antibody also produced relatively weak signals in developing meronts (vs. stronger signal in developed meronts), but it produced low background/non-specific signals in host cells, allowing us to increase exposure time to visualize the specific signals.
While GP900 was observed in all asexual stages including sporozoites and meronts, it was undetectable in the sexual stages of C. parvum. Neither anti-GP900-C nor anti-GP900-N antibodies produced signals in sexual stages, and both antibodies only produced close-to-the-background signals in macrogamonts and microgamonts (vs. clear signals using the anti-total protein antibody) (Fig. 8).
GP900 is released into extracellular space by sporozoites and intracellular parasites
The lower signals from anti-GP900-N antibody in intracellular parasites were also implied by the fact that a significant amount of processed/cleaved GP900 was secreted/discharged to the extracellular space (i.e., into the culture medium) as determined by ELISA (Fig. 9). In this assay, anti-GP900-N antibody (Fig. 9 orange bars), but not anti-GP900-C antibody (green bars), detected secreted GP900 in the medium after incubation with excysted sporozoites (37°C for 1 h), and in the serum-free medium collected from the invasion stage (i.e., 0–3 hpi) and selected intracellular developmental stages (i.e., 0–3, 3–6, 10–12 and 20–24 hpi). Two positive reagent controls showed expected results: 1) in the specimen control, both anti-GP900-N (orange bars) and anti-GP900-C (green bars) antibodies detected GP900 from the supernatant of sporozoite lysate prepared by freeze/thaw; and 2) in the antibody control, the antibody raised against total parasite proteins detected antigens in all samples (gray bars). The OD values were relatively low for the intracellular samples (Fig. 9A), but still notable, particularly when individual datasets were normalized with the OD values obtained using anti-total protein antibody (Fig. 9B). Normalization with antiserum against total parasite proteins was less perfect, as the antiserum might not recognize all C. parvum proteins equally and different parasite stages might also secrete proteins unequally in specified time periods. However, this normalization gave a rough, but sufficient comparison on the relative levels of secreted GP900 between various stages.
Taking together all the ELISA and IFA data, we might build a working model for the biological process of GP900 in sporozoites (Fig. 10): 1) In intact oocysts, this single-pass transmembrane glycoprotein is stored in the micronemes of sporozoites (Fig. 10A-B); 2) it starts to enter the secretory pathway during/after the excystation (Fig. 10C, steps 1), in which the short C-terminal domain is cleaved at the TMD site (step 2) and the secretory vesicles cross the inner membrane complex (IMC) consisting of flatted plates and fused with plasma membrane (PM) (step 3); and 3) the long N-terminal domain is released into the extracellular space (step 4). Because of the general adhesive property for a glycoprotein and the presence of charged amino acids, some discharged GP900 molecules retained on the sporozoites surface. The cleavage is presumably carried out at the TMD by one of the three C. parvum rhomboid peptidases (CpROMs) (Fig. 10D), in which CpROM1 has been recently confirmed to be a micronemal protein as well . In fact, GP900 contained an intramembrane sequence with a high identity to that of known cleavage site of Toxoplasma gondii micronemal proteins (i.e., AA|GG in GP900 vs. IA|GG in TgMIC2 and TgMIC6) (Fig. 10D).