Studying O-GlcNAc glycosylation in the brain
Many lines of scientific evidence implicate reversible O-linked N-acetylglucosamine (O-GlcNAc) modifications in brain function. For example, the enzymes responsible for adding or removing the posttranslational modification are present at high levels in neurons. In addition, several proteins important for neuronal function are glycosylated with O-GlcNAc.
Current methods for monitoring O-GlcNAc levels are not sensitive and often do not detect all of the proteins that have this modification. So, Linda Hsieh-Wilson and co-workers at Howard Hughes Medical Institute, the California Institute of Technology, the Genomics Institute of the Novartis Research Foundation, the University of Wisconsin Madison, and the David Geffen School of Medicine at the University of California Los Angeles developed a strategy to detect and quantify O-GlcNAc modifications. With the new method, called quantitative isotopic and chemoenzymatic tagging (QUIC-Tag), they studied O-GlcNAc glycosylation dynamics on several neuronal proteins. They also discovered that an excitatory stimulus can change the O-GlcNAc glycosylation profile of the brain proteome in vivo.
QUIC-Tag combines a tagging method that the researchers previously described with an isotopic labeling procedure. O-GlcNAc proteins are labeled with a ketogalactose-biotin tag, which produces characteristic fragments upon collisionally activated dissociation (CAD). Next, the proteins are digested, and stable isotopes are introduced onto peptide N-terminal amines and ∈-amino groups of lysines by treatment with heavy or light formaldehyde/sodium cyanoborohydride. Hsieh-Wilson and colleagues then combine isotopically labeled and tagged peptides from two cellular states and enrich for O-GlcNAc peptides with avidin chromatography. The relative levels of O-GlcNAc glycosylation are assessed by MS. MS/MS analyses verify that a peptide is modified with O-GlcNAc. To identify the protein and pinpoint the glycosylation sites, the researchers perform CAD and electron transfer dissociation.
To study O-GlcNAc dynamics, Hsieh-Wilson and colleagues applied the method to lysates of cultured rat neurons that were untreated or treated with an inhibitor of an enzyme that removes the modification. A total of eight proteins that underwent reversible O-GlcNAc glycosylation were identified. This group included some proteins that were not previously known to be O-GlcNAc glycosylated. Finally, the investigators studied in vivo glycosylation by injecting rats with kainic acid, which generates an excitatory stimulus in the brain. The cerebral cortices of rats were dissected at several time points, and QUIC-Tag was performed on the samples. Four proteins with increased O-GlcNAc levels were identified. According to the researchers, this study is the first to demonstrate that an extracellular stimulus other than glucose can alter O-GlcNAc dynamics. (Nat. Chem. Biol. 2007, 3, 339–348)
Phosphopeptide validation strategy
To be confident of a protein identification in a shotgun proteomics experiment, most researchers would say that ≥2 peptides of that protein should be identified. To validate the presence of a particular peptide, such as a phosphopeptide, with MS, however, a researcher generally has only one shot because that peptide shows up as a single peak. So, John Eriksson and colleagues at the University of Turku and Åbo Akademi University (both in Finland) developed a new strategy to validate phosphopeptide identifications in proteomics experiments. Compared with current validation methods, the method is fast, and it is applicable to high-throughput proteomics experiments.
In the method, tryptic phosphopeptides are enriched with TiO2 chromatography. Half of the sample is dephosphorylated with alkaline phosphatase treatment, whereas the other half is left alone. The samples are analyzed independently by LC/MS/MS with a Q-TOF instrument. Then, the peptide elution profiles and the mass spectra of the phosphorylated and dephosphorylated samples are compared. Essentially, researchers can observe a peptide twice with MS in this strategy, with and without a phosphate group. The locations of phosphorylation sites can be pinpointed with MS/MS analyses.
Eriksson and colleagues applied the method to the study of phosphoproteins involved in differentiation in mouse myoblasts. A total of 297 possible phosphopeptides were isolated. Of those, 135 phosphopeptides had a dephosphorylated partner in the other sample. The Mascot search engine predicted that the phosphopeptides should contain 138 phosphorylation sites. The locations of most of these were pinpointed, but 22 sites could not be localized with certainty. The researchers say that they currently are developing a software tool to automate the analysis of data from these types of experiments. (Mol. Cell. Proteomics 2007, DOI 10.1074/mcp.M600480-MCP200)
A little heat helps the analysis of membrane proteins
Heat is known to enhance LC peak selectivity and resolution, but can it also improve the recovery of hydrophobic peptides? As Christine Wu and colleagues at the University of Colorado School of Medicine discovered, the answer is yes.
To study the role of heat in a shotgun proteomics experiment on membrane proteins, Wu and colleagues analyzed two types of samples. The first was a tryptic digest of a HeLa plasma membrane fraction that contained integral membrane proteins and some soluble proteins. The other sample was treated with high pH, proteinase K, and cyanogen bromide (hppK–CNBr) to enrich for membrane-embedded protein domains. The temperature of the LC system was adjusted with a block column heater. The samples were run on the LC system at 20, 30, 40, 50, and 60 °C.
Peak shape was not greatly improved for either sample at elevated temperature, possibly because the samples were so complex. However, protein and peptide identifications increased dramatically as the column was heated to 60 °C. When GRAVY scores were plotted, the investigators observed that higher temperatures allowed the identification of more hydrophobic peptides than lower temperatures. Proteins with multiple transmembrane domains were more highly represented in the hppK–CNBr sample than in the tryptic digest, and this effect was enhanced with heat. The proteins identified at 60 °C included unique proteins as well as most of those identified at lower temperatures, so the researchers suggest adding heat to LC columns when hydrophobic peptides are analyzed. (Anal. Chem. 2007, 79, 4613–4620)
Targeted glycoproteomics and glycan site mapping
Glycosylation is a difficult posttranslational modification to identify and map. Glycans placed onto proteins can be branched and complex, and they can vary in the linkages by which they are attached to proteins. Therefore, Chi-Huey Wong, Benjamin Cravatt, and colleagues at Scripps Research Institute and Academia Sinica (Taiwan) developed a glycoproteomics and glycan mapping strategy, called GIDmap, that enriches for specific types of glycoproteins. With the method, researchers also can identify the glycoproteins in a sample and determine the exact sites of glycosylation.
With GIDmap, a specific type of glycoprotein is modified by the addition of an alkynyl sugar derivative to growing cells. The resulting alkynyl glycans are biotin-tagged with click chemistry, and the tagged proteins are recovered with immobilized streptavidin. Proteins are digested with trypsin, and the liberated nonglycosylated peptides are analyzed with LC/MS/MS to identify the enriched proteins. The researchers retrieve those peptides still bound to the beads by treating them with an enzyme that cleaves at the glycan–peptide bond. This process adds 1 Da to the peptide’s mass for the identification of glycan sites.
The investigators applied the method to the study of sialylated glycoproteins in prostate cancer cells. An alkynyl sugar derivative of N-acetylmannosamine (ManNAc) called ManNAcyne was add- ed to the cells to specifically label sialylated glycoproteins. The method selectively enriched for N-linked glyco- proteins. A total of 219 unique N-glycosylated peptides that belong to 108 nonredundant proteins were identified. Most of the mapped glycosylation sites were novel. Of those that had been described previously, ~26% are involved in cancer processes. The researchers say that, unlike other glycosylation enrichment methods, the GIDmap method does not require complex separations and does not destroy carbohydrate structures. Therefore, other analyses could be conducted on the isolated glycans after GIDmap is performed. (J. Am. Chem. Soc. 2007, 129, 7266–7267)
Detection of peanut allergens
You might want to think twice before sending your child to school with a batch of homemade peanut butter cookies to share. Peanut allergies are becoming more common in the U.S. , and often children don’t know whether they are allergic. Although some allergic individuals merely experience itchy skin, others suffer much more serious consequences, such as nausea, dizziness, or loss of consciousness. In rare cases, fatal reactions can occur. Obvious triggers, such as peanut butter in a labeled jar, are easy to avoid, but sometimes the presence of peanuts is more obscure. For example, food that does not intentionally include the legume as an ingredient can become contaminated with allergens if it’s processed by machines that are also used to make peanut-based products. Because accidental contamination is a major health issue, Hubert Chassaigne and co-workers at the European Commission’s Directorate-General Joint Research Centre (Belgium) sought new methods to detect the three main peanut allergens. With proteomics techniques, they discovered several sequence tags that appear to indicate unambiguously the presence of peanut allergens.
Current detection methods typically are not specific for peanut allergens. Epitopes detected with ELISAs often are unknown, and cross-reactions with other food components can result in false positives. PCR analyses can detect DNA encoding for an allergen, but the presence of a snippet of DNA does not mean that the corresponding protein also is present.
To overcome these challenges, Chassaigne and co-workers applied proteomics techniques. Peanut extracts were digested with trypsin and analyzed by LC/MS/MS. The researchers discovered that five candidate peptides from the three major peanut allergens could serve as sequence tags. These peptides are specific for peanuts, have a sequence that overlaps with epitopes that are recognized by the human immune system, and can be detected in raw and roasted peanuts with the LC/MS/MS method. In addition, several peptides are specific to raw, mild-roasted, or strong-roasted peanuts, so these can be used to determine the degree of processing. (J. Agric. Food Chem. 2007, 55, 4461–4473)
Protein modifications on a population scale
Little is known about the incidence of protein modifications in a large human population. So, Dobrin Nedelkov and co-workers at Intrinsic Bioprobes determined the prevalence of modifications of five plasma proteins in 1000 human subjects living in four states within the U.S. In the study, the researchers identified 27 modifications, including oxidations, truncations, and point mutations.
Affinity pipettes were prepared with antibodies against beta 2 microglobin, cystatin C, retinol-binding protein, transferrin, or transthyretin for MS immunoassays. A total of 20 posttranslational modifications and 7 point mutations were detected. The most commonly observed variants were those that were oxidized or that had single amino acids missing from one of the ends. Point mutations and extensive truncations were less frequently observed within the studied population. Interestingly, the samples obtained from subjects in California contained fewer modified proteins than the others. Nedelkov and co-workers say that the reason for this difference is unclear. Gender variations were observed for two proteins. Most of the subjects who had a carbohydrate-deficient form of transferrin were male. This variant is a biomarker for alcoholism, and the investigators point out that this finding may reflect the higher prevalence of alcoholism among U.S. men. In addition, cystatin point mutations were found only in males, but the significance of this result is unknown. The researchers say that large-scale population studies such as this one could help determine the composition of the normal proteome. (Mol. Cell. Proteomics 2007, DOI 10.1074/mcp.M700023-MCP200)
Return to Top
Copyright © American Chemical Society, 1155 Sixteenth Street N.W., Washington, DC 20036
