| This thesis presents an interdisciplinary research involved in analytical chemistry, chemical biology and so on. Based on biological mass spectrometry (bio-MS), novel methods and technologies were developed for qualitative and quantitative analyses of protein N-glycosylation modification. The developed methods and technologies can be used for accurate and effective protein N-glycosylation profiling, including glycoprotein/glycopeptide, glycosylation site, glycan, and so on.Glycosylation is one of the most common, important, and complicated post-translational modifications of proteins. The oligosaccharide chains of glycoproteins have vital roles in protein conformation, activity, and localization, as well as participate in many biological processes, such as molecular and cellular recognition, signaling, and communication. Gycoproteins can be classified into N-gycoproteins, O-glycoproteins and GPI glycoproteins according to the difference of glycan structure and glycosite, in which the N-gycoproteins are most common and widely presence. More than50%of all proteins in the protein database have been estimated to have at least one glycosylation site, but only10%of them have been reported and most potential glycoproteins have not yet been discovered. Glycosylation investigation is still at the initial stage and lack of effective technologies. On the other hand, with the progress of soft ionization bio-MS, high throughput and systemic proteomics is rapidly advancing. In recent years, bio-MS has also been applied for glycosylation study, including glycosite identification and glycan structure analysis, and corresponding high throughput glycosylation study, called glycoproteomics and glycomics, are also developed. However, glycan synthesis is non-template driven, and thus the structure of glycan is2-dimension and very complicated. At the same time, the ionization efficiency of glycan is low, and glycan has many isomers. Thus, MS-based glycosylation analysis still needs further advancing and progress.Based on bio-MS, this thesis developed several novel technologies and methods for qualitative and quantitative study of N-glycosylation. These strategies can be effectively applied for glycoproteomics and glycomics, to solve the existent problems and bottlenecks in this field. The major contributions of this thesis are as follows:(1) Endoglycosidase (Endo) was applied for large-scale and high throughput MS-based glycosite identification and could be used as a complementary method with traditional peptide-N-glycosidase (PNGase). By these two strategies, a large number of new glycosites were identified from rat liver tissue. (2) Endo mediated incorporation of18O into N-glycans was observed for the first time, and a novel enzymatic glycan reducing end18O labeling GREOL method was developed based on this reaction. This method can be effectively used for MS-based relative glycan quantification, and filled a blank in quantitative glycomics.(3) PNGase mediated N-glycan complete18O labeling was achieved by condition optimization, and novel enzymatic18O4labeling method was developed based on this reaction. Enzymatic18O4labeling can be used for one-pipeline quantitative analysis of glycan, glycopeptide and non-glycopeptide, and realize comprehensive N-glycosylation quantification in single experiment.(4) Sodium borohydride assisted enzymatic glycan18O labeling and double isotope labeling were developed. This method not only improved the stability of18O labeling, but also increased the mass gap between16O/18O labeled glycan to3Da, and thus largely reduced the overlap of isotope envelopes.This thesis consists of four parts which are summarized as follows:Large-scale N-glycosite identification by complementary enzymes:Endo and PNGase:Large-scale glycosite identification is a vital objective for qualitative glycoproteomics. PNGase F is the widest used enzyme for N-glycosite discovery. This enzyme deamidate the Asn residue and simultaneously make a+0.98Da mass difference on the glycosite. Based on this mass label, glycosite can be identified by MS. However, large-scale MS analysis needs a balance between sensitivity and resolution, and thus it is not easy to differentiate0.98Da mass gap during high throughput LC-MS analysis. At the same time, natural conversion between Asn and Asp is normal, which also affects the reliability of identification. Endo is a class of glycosidases that specifically hydrolyzes the glycosidic bond between two proximal GlcNAc in the pentasaccharide core of N-glyoprotein, leaving a terminal GlcNAc still attached to its parent peptide, which provides a mass increase of203.08Da. The accuracy of identification is improved based on the larger203.08Da label, and this method has been tried for glycosite discovery.This thesis utilized Endo H and traditional PNGase F for large-scale analysis of high mannose and hybrid type N-glycosite, and successfully applied Endo H for high throughput N-glycosite identification. By combining Con A lectin affinity enrichment, two enzymes treatment and LC-MS/MS detection, a total of1063unique high mannose and hybrid type glycosites and corresponding560unique glycoproteins were identified from rat liver tissue. Most of these sites were discovered for the first time, providing much novel glycosylation information for rat protein database. This investigation has the following contributions:(1) Endo was applied for large-scale N-glycosite identification at the first time, proved the efficiency of this method for high throughput glycoproteomics;(2) Combining the results of PNGase F and Endo H, a larger dataset of N-glycosites, as well as many new glycosites from rat liver glycoproteins, were obtained;(3) Information concerning glycan subtypes and core-fucosylation on each glycosite was provided by Endo H, and this is valuable for further functional study;(4) The characters and merits of each method were discussed in-depth in this study:the PNGase F method results in a larger dataset, whereas the Endo H method provides more confident assignment;(5) The problems and potential improvements for the Endo H method was discussed in-depth.Endo mediated incorporation of18O into N-glycans for quantitative glycomics:With the development of glycoproteomics, relative quantification of glycosylation has received considerable research attention, which is important for discovering glycosylation changes during diseases and finding potential disease biomarkers. High throughput glycan quantification, called quantitative glycomics, was developed to investigate quantitative changes in structure type, composition, linkage and conformation of glycans form different biological samples. Similar to quantitative proteomics, MS based isotopic label is also a powerful tool for quantitative glycomics, including reducing end labeling, permethylation labeling and metabolic labeling. Moreover, for quantitative proteomics, enzymatic18O labeling is a popular and effective isotopic label technique due to its simplicity:labeling occurs during enzymatic digestion, and all that is required is the presence of18O-water, avoiding extra reagents, additional steps, and side reactions. However, enzymatic18O labeling has never been achieved in quantitative glycomics.When using Endo for further research, we incidentally found that18O is universally incorporated into the N-glycan reducing end in the presence of H218O during enzymatic releasing N-linked glycans by Endo. Based on this reaction, this thesis developed enzymatic glycan reducing-end18O labeling (GREOL) technique, and applied this strategy for glycan quantification. The reaction of Endo stably and completely labeled an18O atom into glycan, and this mechanism was applicable for all subtypes of Endo. After MS analysis and manual deconvolution, the present method provided good linearity with high reproducibility, at least within2orders of magnitude in dynamic range. The detection limit was at pmol level, and thus this strategy can be effective applied for quantitative glycomics. Moreover, this novel strategy also quantitatively discriminated between isomeric hybrid and complex N-glycans in glycoproteins. We also utilized this strategy to quantitatively determine the glycomic changes in sera from healthy individuals and patients diagnosed with hepatocellular carcinoma (HCC). This investigation has the following contributions:(1) Endo mediated incorporation of18O into glycan reducing end was first observed during enzymatic releasing N-glycans in H218O, with the labeling efficiency at100%;(2) Based on this reaction, a novel enzymatic18O labeling technique-GREOL was developed, and successfully applied for MS-based glycan quantification, which filled a blank in quantitative glycomics;(3) The reaction mechanism of enzymatic18O labeling by Endo was discussed in-depth, and a deconvolution method was introduced to avoid partial overlap of glycan pair isotope envelopes, which greatly improved the linearity, reproducibility and sensitivity of GREOL;(4) Quantitative discrimination of isomeric N-glycans was first achieved in this study, providing a novel approach and aspect to investigating glycan changes;(5) Quantitative analysis of HCC associated human serum N-glycans was operated, and increasing biantennary N-glycans, as well as the level of bisecting GlcNAc and NeuNAc residues, was observed in HCC, which can be potential biomarkers for HCC diagnose.PNGase based18O4labeling for protein N-glycosylation quantification:Traditional PNGase cleaves the amide bond between glycan end and glycosylation site, and the reaction mechanism is quite different to glycosidase. PNGase specifically hydrolyzes the covalent bond between the terminal GlcNAc and the Asn residue, and thus Asn is deamidated and transferred to Asp. Then, the released glycosylamines deamidate and form glycans. Due to the wider application of PNGase in glycosylation investigation, we further tried the possibility of PNGase mediated glycan18O labeling. However, results showed that the deamidation was reversible, and thus even in solution without any ammonium, the ammonium hydrolyzed from glycosylamine also inhibited the complete process of deamidation. The remained glycosylamine in solution was hydrolyzed during further purification, and numerous16O from normal16O-water would be labeled onto the formed glycan. This phenomenon consciously affected the label efficiency and negated the application of PNGase mediated18O for glycan quantification. This thesis tried to change the solution’s pH after reaction, and found that, the glycosylamine deamidation could completely process under acid condition. Based on this phenomenon, we performed PNGase digest in H218O prepared buffer, and then adjusted the solution from base to acid. Finally, we realized the complete N-glycan18O labeling by PNGase. After MS analysis and manual deconvolution, this method provides good linearity with high reproducibility, at least within2orders of magnitude in dynamic range. Moreover, by combining PNGase mediated glycan18O labeling, glysite18O labeling and trypsin mediated peptide18O2labeling, a novel18O4labeling method was first developed. By18O4labeling, we can simultaneously quantify glycan, glycosite and glycopeptide in single experiment. We further applied this strategy for quantitative analysis the glycosylation of HCC associated human serum IgG, and observed various HCC associated glycosylation changes.18O4labeling strategy realized one-pipeline quantitative glycoproteomics and glycomics, and would advance the development of the two fields. This investigation has the following contributions:(1) By completing glycosylamine deamidation, PNGase mediated incorporation of18O into glycan reducing end was achieved for the first time, with the labeling efficiency at100%;(2) This labeling method showed high accuracy, linearity and reproducibility for glycan quantification, and could be complementary used with GREOL;(3) Enzymatic18O4labeling method was developed for quantification, and thus one-pipeline glycan18O labeling, glycosite18O labeling and peptide18O2labeling was achieved for comprehensive N-glycosylation quantification;(4) Quantitative analysis of HCC associated human serum IgG N-glycosylation was performed by18O4labeling. No obvious changes were observed at glycopeptide and glycoprotein level, but several changes were found at glycan level:biantennary glycans and bisecting GlcNAc structure increased, while monoantennary glycans decreased. This result suggested that the quantitative changes between protein and glycan are quite different, providing valuable information for further research and potential biomarker discovery.Sodium borohydride assistant enzymatic18O labeling and double isotope labeling:This thesis introduced two novel enzymatic glycan18O labeling methods, but these methods can only label one18O atom into glycan, and form narrow2Da mass gap. This will induce partial overlap of glycan pair isotope envelopes and affect quantification. Although deconvolution can dramatically reduce the negative effect, the calculation step is laborious. Besides, the samples must be analyzed independently before mix, and this will also increase the error. Moreover, the labeled18O will exchange with16O in normal water, and thus some O may be labeled into glycan and reduce the accuracy of quantification.This thesis further developed novel sodium borohydride assisted enzymatic18O labeling. That was, added NaBH4/NaBD4into PNGase16O/18O-buffer after reaction, and then the reducing end of all released glycans were reduced and labeled with a deuterium atom. After this label, the glycan became more stable, and the labeled18O would never exchange with16O again in normal water. Moreover, the mass gap increased to3Da, and the partial overlap of isotope envelopes was largely reduced. We further tested the stability and overlap of this label, which showed good results. By this label reaction, we optimized the enzymatic glycan O labeling, and made the method more suitable for quantitative glycomics. This investigation has the following contributions:(1) By adding NaBH4/NaBD4after enzymatic16O/18O labeling, glycan reducing end was reduced, and the labeled16O/18O became more stable;(2) H/D labeling at glycan end by NaBH4/NaBD4was achieved, and16O+H/18O+D double isotope labeling was developed with the mass gap increased to3Da. By the double labeling, the overlap of isotope envelopes was largely reduced.(3) Fast sodium borohydride treatment method was developed, and the whole experiment just needed several hours and single reagent, maintaining the advantages of enzymatic18O labeling. |
PDF Full Text Request