The bittersweet study of glycomics

Carbohydrates are important in many biological processes and often implicated in disease states but studying them is far from simple. Luckily, new analytical techniques are being applied to help unravel the secrets of the glycomeMuch attention has been given to the “omic” approaches to biology, starting with genomics, the entire genetic complement of the cell, and continuing with transcriptomics and proteomics. While these studies have required the application of research resources on a previously unprecedented scale, they are in some ways the simplest of the systems biology approaches. DNA, RNA and proteins are all linear structures with a relatively small number of building blocks. On the other hand, glycomics, the study of the glycan content of a cell, tissue, or organism at any point in time, involves a very large number of monosaccharide building blocks that can take numerous isomeric forms. To add to the complexity, glycans can take a myriad of branched structures that can change with the physiological state of the organism, and they can bind to proteins as well as lipids.

The challenges posed by the study of glycomics are even more significant when the importance of the function of glycan structures is considered. They are involved in cellular signaling pathways and modulation of cell function, including cell fate and other cell-cell interactions. They also participate in immune response, inflammation and cancer progression, helping to transmit the signals that trigger unchecked cell growth. Glycan structures have also been implicated in the development of Parkinson's, Alzheimer's and infectious diseases like AIDS and herpes.  Many of these functions involve glycans bound to proteins. The fields of glycomics and glycoproteomics are therefore essential to many areas of life science research, including cell biology, biochemistry, and medicine. Many analytical techniques have been applied to unravel the complexities of glycan and glycoprotein structures, including chromatographic, electromigratory, or mass spectrometric methods. The end goal of all of these approaches is to detect and quantify all types of glycans, separate all isomers and assign a particular structure including the overall topology of the molecule and all linkages. While each approach has its advantages and disadvantages,  only the combination of shape-based chromatography and MS can address the complexity and diversity of protein- or lipid-bound oligosaccharide structures1.

[caption id="attachment_33368" align="aligncenter" width="440" caption="Figure 1: Relative abundances of high-mannose glycans, undecorated complex/hybrid (C/H) glycans, fucosylated (but non-sialylated) C/H glycans, sialylated (but non-fucosylated) C/H glycans, and finally, fucosylated and sialylated C/H glycans found in the P (poor prognosis) and G (good prognosis) patient groups. Asterisks denote statistically significant differences (p < 0.05) between patient groups which could potentially be used to predict outcomes."][/caption]

The high shape-discrimination capability of porous graphitized carbon (PGC) toward isomeric and isobaric glycan structures makes it ideal as a liquid chromatography matrix to provide the first step in determining the composition and structure of a glycan, including isomers. The second step is best provided by Q-TOF MS, due to its exquisite accurate mass capability, which enables structure elucidation of a complex mixture of glycans. This two-step approach has been combined with microfluidics to provide a nano-LC/Q-TOF MS system that can separate the glycan building blocks and their isomers and fully characterise the glycan structures with MS and MS/MS, in a single “chip” that interfaces with the Q-TOF MS.  This chip-based system provides high sensitivity, wide instrumental dynamic range, minimal ion suppression, and low sample consumption, making it ideal for the analysis of complex glycan structures.

This unique system can deliver results in minutes, rather than days, and is run in automated mode. The researcher simply places the sample in an autosampler, and all of the analysis steps are performed automatically. Software and a glycan database are then used to identify and determine the glycan structures, isomers, quantities, and ratios. This is the only analytical approach that can separate every single glycan isomer.

The HPLC Chip-based nano-LC/Q-TOF MS system has been used in several instances to characterise both free and protein-bound glycans. For example, this approach has been used for the characterisation of N-linked glycans from glycoproteins, delivering results in 10 to 30 minutes, in automated mode. The glycoprotein is deglycosylated on the chip using an on-line PNGase F reactor. The resulting glycans are then concentrated on a PGC enrichment column and glycans are separated on a PGC analytical column. The chip reproducibly separates and quantitates all of the common N-linked glycans, making it a useful Process Analytical Technology (PAT) method for assuring batch-to-batch reproducibility of the glycan composition of a therapeutic protein.

Glycosylation is dynamic and has been shown to play a role and in many diseases including cancer. Potential biomarkers for several types of cancer and diseases have been identified using glycan compositional profiling of human serum with mass spectrometry. Unfortunately, composition analysis alone cannot fully describe glycan stereo- and positional isomeric diversity2. A formidable analytical challenge is posed by the extensive structural heterogeneity of glycans. HPLC Chip-based nano-LC/Q-TOF MS has been used to develop a method to identify and quantify isomeric native glycans2.

[caption id="attachment_33369" align="aligncenter" width="412" caption="Figure 2: Overlaid chromatograms and associated structural assignments of glycopeptides from bovine Ribonuclease B. Inset, relative abundances of each glycoform of bovine Ribonuclease B"][/caption]

The utility of this method for structure-specific biomarker discovery was determined through the analysis of serum samples from prostate cancer patients. More than 300 N-glycan species were identified, including isomers, comprising more than 100 N-glycan compositions.The patients were divided into two groups with different prognoses. Individual glycan abundances were determined and compared for patients with poor prognoses (P group) and good prognoses (G group). Individual N-glycan compounds were sorted into several classes, including high-mannose, undecorated complex/hybrid, fucosylated (but nonsialylated) complex/hybrid, sialylated (but nonfucosylated) complex/hybrid and fucosylated and sialylated complex/hybrid glycans. Relative abundances were calculated versus the total ion abundance of all N-glycans in a particular nano-LC run. Fucosylated, non-sialylated complex/hybrid glycans were significantly more abundant in the G group than in the P group. In contrast, sialylated, non-fucosylated complex/hybrid glycans were significantly more abundant in the P group (Figure 1). These differences in glycosylation may be useful as biomarkers of prognosis, and they may help identify glycan synthesis pathways affected by the disease.

This HPLC chip-based nano-LC/Q-TOF MS system has also enabled an approach that can simultaneously provide information on both the proteomic and glycomic aspects of a biological system3. A method has been developed using this technology that identifies isomer-specific differentiation at specific sites on glycoproteins. The method utilises Pronase E, a mixture of proteolytic enzymes, to non-specifically hydrolyze all peptide bonds and isolate single glycosylation sites, while reducing signal suppression by breaking down non-glycosylated peptides into amino acids. High mass accuracy data from Q-TOF MS is used to assign glycopeptide compositions, and MS/MS data is used to confirm glycopeptide assignments and elucidate the structure of chromatographically separated isomers, over a dynamic range of five orders of magnitude.

[caption id="attachment_33371" align="aligncenter" width="396" caption="Figure 3a: Extracted compound chromatogram (ECC) of glycopeptides eluted with the 20% ACN fraction of the SPE fractionation step. Figure 3b: ECC of glycopeptides eluted with the 40% ACN fraction of the SPE fractionation step. Glycosites corresponding to each assigned glycopeptide are highlighted in bolder font to distinguish them"][/caption]

This method provided extensive site-specific, isomer-specific profiling of N- and O-linked glycosylation in several proteins, including Ribonuclease B (Figure 2), describing glycan heterogeneity in a site-specific and quantitative manner. The extra information provided by site-specific analysis could prove invaluable to the discovery of new disease biomarkers. Tracking changes in glycoprotein heterogeneity during a biological process will help provide a new understanding of the role of the glycoproteome3.

This HPLC Chip-based nano-LC/TOFMS method for site-specific characterisation of glycosylation has been further refined by tuning the collision-induced dissociation (CID) energies for MS/MS analysis of the glycopeptides4. Glycosylation introduces an analytical complexity that requires extremely high-quality MS/MS spectra for correct glycopeptide analysis in complex samples. Tuning CID energies, combined with the high mass accuracy of the QTOF MS, enabled the assignment of a unique composition  to approximately 80% of all putative glycopeptides in three standard glycoproteins containing a variety of high mannose, hybrid, and complex type N-glycosylations4.

The success of the application of the HPLC Chip-based nano-LC/Q-TOF MS technology in characterising site-specific glycosylation on individual glycoproteins led to its application to glycoprotein mixtures. As proof of concept, bovine lactoferrin, kappa casein, and bovine fetuin were digested using the Pronase cocktail of proteases immobilised to agarose beads, followed by enrichment and separation on a graphitized carbon cartridge (GCC). Each GCC fraction was analysed by chromatographic separation and Q-TOF MS and MS/MS on the chip. The mass list of the glycopeptide precursor ions from the MS/MS analysis was analysed with the authors’ in-house software “GP finder” for rapid glycopeptide assignment. In total, 233 glycopeptides corresponding to 18 glycosites were observed and determined in a single mixture, identified based on composition and including isomers 5. The glycopeptides were a mixture of primarily sialylated O-linked glycopeptides, along with N-linked glycopeptides containing high mannose, complex and hybrid glycans. Detailed glycan microheterogeneity information was obtained. Pronase digestion assures no large peptides in the mixtures, so that the chromatogram corresponds primarily to glycopeptides. Figure 3A is representative of the extracted compound chromatogram (ECC) of glycopeptides eluted with the 20% ACN during the GCC fractionation step.

Short peptide tags (3 to 5 amino acids) connected to high mannose type glycans (GlcNAc2Man59) eluted ahead of longer peptide-containing glycopeptides (6 to 7 amino acids). The analysis identified these N-linked high mannose-containing glycopeptides as originating from lactoferrin (represented in Figure 3A with peaks shaded in pink). In addition, glycopeptides corresponding to kappa casein (peaks shaded in green) and bovine fetuin (peaks shaded in blue) were also observed. The glycopeptides observed in the 40% ACN fraction (Figure 3B) predominantly corresponded to O-linked glycopeptides from kappa casein (peaks shaded in green) and bovine fetuin (peaks shaded in blue). Thus, this approach presents a platform to simultaneously characterise N- and O-glycosites in a glycoprotein mixture with extensive site heterogeneity.

Unlocking the secrets of the extremely complicated structure of the glycome can bring sizable rewards, including a better understanding of disease, biomarkers for disease states, highly effective biotherapeutics, and a more complete knowledge of biological processes. The step forward in glycan analysis technology represented by HPLC Chip-based nano-LC/Q-TOF MS has and will continue to greatly expand our knowledge of complex glycan and glycoprotein structures and the vital roles that they play.

The author wishes to acknowledge Professor Carlito Lebrilla at University California-Davis and Professor Hyun-Joo An at Chungnam National University for their collaboration with Agilent Technologies in the development of these applications of the HPLC-Chip based nano-LC/Q-TOF MS technology.

Author: Rudi Grimm, Director of Science and Technology, and Life Science Business Development Manager Asia Pacific, Agilent Technologies, Inc.