Simon Fraser University scientists have completed a study profiling the N-glycoproteome of the mouse embryonic stem cell E14.Tg2a, a line commonly used in genetic research.
The study provides additional characterization of this line that could prove useful in gene targeting and trapping experiments, Bingyun Sun, an SFU researcher and lead author on the paper, told ProteoMonitor. Beyond that, Sun noted that the work offers a global look at cell surface proteins, a class of molecules that, despite their biological importance, have traditionally been difficult to analyze.
In their study, published this month in PLOS One, the researchers used a Thermo Scientific LTQ linear ion trap instrument to identify a total of 1,182 glycosites across 468 glycoproteins.
Sun and her colleagues chose to use the relatively low-powered LTQ as opposed to a newer machine due to their quick and easy access to this instrument. Key to their work was the glycopeptide enrichment scheme they employed upfront of mass spec analysis, which, Sun said, allowed them to obtain high coverage and specificity despite their machine's limitations.
SFU "certainly has [high-end] mass spec systems…but there was a long queue for those instruments, so I thought maybe we would try an older instrument and see what we could get," Sun said. "In the end we were very happy with our result, which really gives us the confidence that if the front-end sample preparation is very good and selective, the [mass spec] analysis can be done with regular instruments."
The researchers used a hydrazide resin-based enrichment process developed in the lab of Swiss Federal Institute of Technology Zurich researcher Ruedi Aebersold in 2003. In their work, Sun and her team altered the protocol by doing trypsin digestion prior to enrichment rather than before in hopes of improving the solubility of the target membrane and glycoproteins, which is required for the hydrazide enrichment step.
"Many glycoproteins are difficult to dissolve, so you have to first solve this solubility issue," Sun said. To do so by digesting prior to enrichment, however, significantly increases sample complexity, which she feared would limit the effectiveness of the enrichment step.
On the contrary, the researchers found that digestion prior to enrichment actually increased the effectiveness of the enrichment.
"We were kind of puzzled," Sun said. "When we increased our sample complexity we got better selectivity for the glycopeptides. We had expected it to be the other way around."
It turned out, she noted, that by digesting the glycoproteins, the researchers had decreased the effects of steric hindrance that had prevented certain glycosites from binding to the hydrazine resin. Essentially, because full-length proteins have fairly rigid structures, they are only able to bind to the resin at particular points, with some glycosites remaining blocked. Once digested into peptides, however, this steric hindrance is eliminated, meaning that all the protein's glycosites are able to contact the hydrazide resin.
In addition, Sun said, unglycosylated peptides proved less sticky than unglycosylated proteins, making it easier to wash them off, further increasing the technique's selectivity.
Sun noted that more than 95 percent of the peptides they identified included the consensus amino acid sequence indicative of N-glycosylation. The authors also estimated that their method offered sensitivity of detection down into the femtomolar range, enabling identifications of proteins present at levels lower than 100 copies per cell.
"We're pretty confident that using our front-end enrichment [method] we can get a really pure sample that includes a high content of glycopeptides and allows us to go deeper into the cell surface proteome and identify rare receptors and other membrane proteins," she said.
Sun observed that while cell surface proteins are an important class of molecule, particularly as drug targets, they have proven quite challenging to study. She noted that while 25 percent to 35 percent of the human genome is predicted to encode membrane proteins, only two percent of the proteins with known structures are membrane proteins.
Among the PLOS One paper's findings regarding these proteins was the identification of a relationship between glycosylation stoichiometry and the number of transmembrane domains. The researchers discovered that higher levels of glycosylation are associated with lower numbers of TM domains, a relationship that, they noted, was particularly evident when comparing protein receptors – which carried on average four N-glycans per protein and less than two TM domains – and transporters – which carried an average of two glycans and seven TM domains.
"This kind of global information in terms of [glycosylation patterns] of receptors and transporters can help structural biologists understand individual cases for many unknown proteins and what their potential functions could be," Sun said, suggesting the observed higher glycosylation levels on receptors could be important in their roles binding ligands outside the cell to trigger signaling cascades within.
"Glycans are known to participate in protein-protein binding, so the [greater number] of glycosites on receptors could give them … more affinity [for given ligands] or offer protection to other functional sites of the receptor to facilitate its proper function," she said.
Sun added that the SFU researchers plan to follow up on this observed relationship between glycosylation and TM domains. They also plan to perform more similar analyses using high-end mass spectrometers "to see what additional information we can get and the extent to which the advanced instrumentation will help," she said.