Ovarian cancer (OC) has the highest mortality rate of all the gynecological cancers. One way to improve patient prognosis is to track the effectiveness of a course of treatment. The current blood glycoprotein biomarker, CA125, is unreliable for this purpose [1], requiring the discovery of other molecular biomarkers.
Abnormal ascites fluid, a hallmark of late-stage OC is routinely removed over the course of treatment and presents a potentially unique source of new biomarkers. This work aimed to assess the protein N- and O-glycosylation differences between total cells from ascites and three distinct cell sub-populations isolated according to marker status: cancer (CD45-/EpCAM+), immune (CD45+) and marker-negative (CD45-/EpCAM-) cells from three chemo-naïve patients to determine baseline before treatment. Using porous graphitised carbon – liquid chromatography and electrospray ionisation tandem mass spectrometry (PGC-LC-ESI-MS/MS) based glycomics, a difference in the N-glycan signatures between the isolated cell populations was observed. This included a relatively higher expression of oligomannosidic structures in cancer cells, and an opposing increase in sialylated complex-type structures in immune cells. For example, using PGC-LC separation, three sialylated isomers with the composition (Hex)3 (HexNAc)3 (Fucose)1 (NeuAc)3 + (Man)3(GlcNAc)2 (m/z 1512.62-) were found to delineate immune and marker-negative cells from cancer cells. Furthermore, the marker-negative population of cells had a distinct glycomic signature including bisecting GlcNAc motifs and branched, highly sialylated complex structures.
The O-glycomic signature of cells from OC ascites fluid included a high degree of sialylation accounting for up to 90% of glycans in each cell type and the presence of sulphated glycans and fucosylated glycans including two unusual O-fucose structures [2]. Interestingly, the non-sialylated O-fucose linked glycan with the composition (Hex)1 (HexNAc)1 (Fucose)1 (m/z 530.21-) was present in all three cell types however the sialylated form with the composition (Hex)1 (HexNAc)1 (Fucose)1 (NeuAc)1 (m/z 821.31-) was only detected in the cancer cells. Additionally, like the N-glycans, isomer-specific distribution trends were observed for the O-glycans, with for example, four isomers with the composition (Hex)2 (HexNAc)2 (NeuAc)2 (m/z 665.22-) differentially expressed across the cell types. These results highlight the importance of resolving the structural detail of glycan isoforms using powerful separating techniques of PGC-LC. In summary, the differences in glycomic signatures found in this study, between cell types found in ovarian cancer ascites, lays the groundwork for tracking the changes in protein glycomic biomarkers in sorted ascites cells over a treatment course, and ultimately could have clinical use in tracking OC patient response.