N-Glycopeptide Reduction with Exoglycosidases Enables Accurate

Nov 3, 2016 - Characterization of Site-Specific N‑Glycosylation ... together to enable the identification of peptide sequence, glycosylation site an...
0 downloads 3 Views 819KB Size
Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)

Article

N-glycopeptide Reduction with Exoglycosidases Enables Accurate Characterization of Site-specific N-glycosylation Rui Chen, Kai Cheng, Zhibin Ning, and Daniel Figeys Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.6b03531 • Publication Date (Web): 03 Nov 2016 Downloaded from http://pubs.acs.org on November 8, 2016

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

N-glycopeptide Reduction with Exoglycosidases Enables Accurate Characterization of Site-specific N-glycosylation

Rui Chen#, Kai Cheng#, Zhibin Ning, Daniel Figeys* Ottawa Institute of Systems Biology, Department of Biochemistry, Microbiology and Immunology and Department of Chemistry and Biomolecular Sciences, University of Ottawa, 451 Smyth Road, Ottawa, ON K1H 8M5 Correspondence should be addressed to Dr. Daniel Figeys ([email protected]). # Those authors contributed equally

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Abstract We report a new approach called NGlycoReduction that generates an oligomannosylated N-glycopeptidome by enzymatically removing sugars outside the N-glycopeptide pentasaccharide core with exoglycosidases. This approach is based on our discovery that the fragmentation of glycopeptides is glycan-structure dependent and glycans with core mannose structures overwhelmingly lead to the generation of Y1 ions when subjected to MS/MS in mass spectrometry. Oligomannosylated glycopeptidome produced by NGlycoReduction can be mixed with the intact N-glycopeptidome and analyzed by HPLC-ESI-MS together to enable the identification of peptide sequence, glycosylation site and the structure of intact glycopeptides. The glycan structure of intact glycopeptides can be identified from MS/MS spectra of their own and their peptide sequences were identified by the MS3 spectra of the oligomannosylated glycopeptides with the same Y1 ion. Both mass tolerance and difference in retention time were further used to increase the confidence in the Y1 ion alignment. This approach has the advantage of low cost and ease of processing and can be expanded to other samples, especially for characterizing site-specific N-glycosylation involving complex N-glycans. In this study, simultaneous analysis of the combined oligomannosylated N-glycopeptidome and the native glycopeptidome leads to the identification of 609 N-glycopeptides from the secretome and lysates of Huh7 cells.

ACS Paragon Plus Environment

Page 2 of 21

Page 3 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Introduction N-glycosylation plays a vital role in biological processes, and its miss-regulation has been implicated in diseases1. Changes in glycosylation site occupancy and the relative abundance of different glycans at each site are found in cognitive glycosylation disorders2, cancers and inflammatory diseases3,4. It is important to identify the sites of glycosylation on proteins as well as the relative abundance of the different glycans at each site. Unfortunately, mass spectrometric characterization of site-specific glycosylation has been hindered by the enormous macro and micro heterogeneities of glycan structures, sites of attachments, and relative occupancies5. Although glycan complexity can be reduced using endoglycosidases such as PNGase F and Endo H to cleave off the glycans and identify N-glycosylation sites, the link between glycan structural heterogeneity and the protein attachment site is then lost, and the ability to understand the functions of the different glycosylated forms of proteins is also reduced. A few methods and bioinformatic tools have been developed to reconnect the deglycosylated peptides and intact glycopeptides by analyzing both deglycosylated peptides and intact glycopeptides. Those methods utilized the feature of Y1 ion6 fragment of intact glycopeptides, which contains just the peptide backbone and one acetyl glucosamine. With the high accuracy MS/MS spectra, molecular weight of peptide backbone could be calculated from the m/z and charge state of Y1 ion and matched to the deglycosylated peptides identified separately with database search. Although the accuracy in this method could be improved by adding retention time as additional criterion, the false detection rate is difficult to estimate with current target-decoy approach and the existence of isobaric peptides from the list of deglycosylated peptides.

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Data-dependent MS3 acquisition was an alternative method that enabled the identification of glycan composition by MS/MS and peptide sequences by further fragmentation of Y1 ions7,8. As the peptides sequence was identified from the same intact glycopeptide, it provided the most accurate way to identify the peptide sequence of an intact glycopeptide. However, the successful identification of an intact glycopeptide was dependent on the relative abundance of Y1 ions. Unfortunately, Y1 ions are often among the less intense fragments in the MS/MS spectra of N-glycopeptides compared with O-glycopeptides. Therefore, the chance of selecting Y1 ions for MS3 analysis is lower and most of the MS3 acquisition were wasted on the fragments rather than Y1 ion. This method would be more suitable for O-glycopeptides analysis as O-glycopeptides have short glycan chains. In previous attempts to identify intact glycopeptides, fewer intact N-glycopeptides than Oglycopeptides have been identified by data-dependent MS3 acquisition. We present a novel approach named NGlyoReduction, which trims N-glycopeptides with complex glycans into the pentasaccharides core structure by a combination of exoglycosidases, which boosted the relative abundance of Y1 ion during MS/MS fragmentation. Unlike deglycosylation with endoglycosidases, the reduction of glycans of glycopeptides generated an oligomannosylated glycopeptidome which share similar retention time and ionization efficiency with intact glycopeptides. The oligammnosyalted glycopeptidome could be mixed with the intact glycopeptidome and analyzed together by LC-MS. The glycan structure of intact glycopeptides can be identified from their own MS/MS spectra and their peptide sequences were identified by the MS3 spectra of the oligomannosylated glycopeptides with the same Y1 ion by an in-house developed

ACS Paragon Plus Environment

Page 4 of 21

Page 5 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

algorithm. Application of this approach to human cell line sample resulted accurate identification 609 N-glycopeptides from the secretome and lysates of Huh7 cells. Experimental Section Sample Preparation Model glycoproteins (fetuin and RNaseB) were purchased from Sigma (St.Louis, MO). Proteins were denatured by the addition of 20 mM dithiothreitol, followed by heating at 56 °C for 45 min, and then alkylated by adding 40 mM iodoacetamide at room temperature in the dark for 30 min. Fetuin was digested with trypsin, and RNaseB with chymotrypsin. Proteases were added to samples (1:50 w/w) and incubated at 37 °C for 18 h. Protein digests were dried via Speed Vac and reconstituted in 80 % acetonitrile with 0.1 % trifluoroacetic acid. A total of 20 µg of protein digest was used for each enrichment and glycopeptide enrichment was performed as described previously with 20 mg of Poly hydroxyethyl sorbents packed in gel-loading tips9. Glycopeptides enriched from fetuin were cleaved by α 2-3,6 neuraminidase (New England Biolabs), β1-4 galactosidase (QA bio) and β-glucosaminidase (QA bio). All exoglycosidase cleavages were performed according to the manufacturers’ protocols using the supplied buffers. The hepatocyte-derived cellular carcinoma cell line (Huh7) was cultured in DMEM supplemented with 10 % fetal calf serum and 1 mM sodium pyruvate. When the cells reached 80 % confluence, the medium was replaced with serum-free DMEM containing 1 mM sodium pyruvate. Media were collected after 24 h and were concentrated by ultrafiltration using Amicon Ultra-15 centrifugal filter units (3 kDa cutoff, Millipore). Cells were rinsed three times with cold PBS and suspended in lysis buffer (20 mM Tris-

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

HCl, 1 % Triton X-100, pH 7.4) containing a protease inhibitor cocktail. The homogenates were sonicated and then centrifuged at 16 000 g to pellet the debris. For exoglycosidase-cocktail cleavage, protein samples were incubated with a mixture of the aforementioned exoglycosidases for 48 h in 50 mM sodium phosphate buffer (pH 5.5) and then digested with trypsin. A total of 200 µg of digested peptides was used for glycopeptide enrichment. Protein samples without exoglycosidase cleavage were digested directly, and glycopeptides were also enriched using HILIC SPE. LC-MS/MS Analysis For fragmentation studies, glycopeptides were analyzed on an HPLC-MS/MS system consisting of an Agilent 1100 and an LTQ-Orbitrap XL mass spectrometer (Thermo, Bremen, Germany). Peptides were separated using a 60-min gradient. Peptide separation was performed on tip columns (75 µm i.d.×10 cm) packed with 3 µm/120 Å ReproSil-Pur C18 resin (Dr. Maisch GmbH, Ammerbuch, Germany). Glycopeptides were analyzed by collision induced dissociation (CID) at 35 % energy in data-dependent mode, and each full MS scan by the Orbitrap was followed by 5 MS/MS scans on the LTQ. The resolution was set at 60,000 at m/z=400 for survey scans, and the scan range was set from m/z=600 to 2000. For complex sample analysis, glycopeptides were injected into an LC-MS/MS system consisting of an Eksigent NanoLC 400 liquid chromatograph and an LTQ-Orbitrap Elite mass spectrometer (Thermo, Bremen, Germany). Peptides were separated using a 240min gradient. Automatic gain control was set at 1×106 for full MS and 10,000 for MS/MS. The dynamic exclusion function was set as follows: repeat count 1, repeat duration 30 s,

ACS Paragon Plus Environment

Page 6 of 21

Page 7 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

and exclusion duration 90 s. For full MS scans, the resolution was set at 60,000 at m/z=400, and mass was set from m/z=800 to 2000. Glycopeptides were fragmented by CID at 35 % collision energy in data-dependent mode, and each full MS scan by the Orbitrap was followed by 10 MS/MS scans in CID mode. Only top-ranked fragments from MS/MS were sent for data-dependent MS3 acquisition with the same CID settings. Database search and glycopeptide identification Procedures for the identification of intact glycopeptides were integrated into an in-housedeveloped, java-based program with the following steps: Extract MS3 spectra of the Y1 ion: Y1 ions from the cleaved glycopeptides were verified based on the charge state and m/z of the fragments in the MS/MS spectra. The Y1 ion must be the highest ranked fragment and should be followed by a series of fragments representing the pentasaccharide structure, i.e., fragments with mass differences of 203.0799 (HexNAc), 365.131 (HexNAc+Hex), 527.183 (HexNAc+2Hex) and 689.235 (HexNAc+3Hex). At least three ions from the pentasaccharide structures should be matched in the MS2 spectra. Once a Y1 ion was verified, the MS3 spectrum was then extracted for database searching. Database search and FDR calculation: For N-glycosite identification from cleaved glycopeptides, extracted MS3 spectra were searched against the Uniprot human database using Mascot (2.3.0). Enzyme specificity was set as KR/P, and a maximum of two missed cleavages was allowed; cysteine residue was set as a static modification of 57.0215 Da; methionine oxidation and acetyl glucosamine attached to asparagine were set as a variable modification of 15.9949 Da and 203.079 Da, respectively. The automatic decoy database search strategy embedded in Mascot was utilized for Y1 ion identification. After the search was completed, the results were directly extracted from the .dat files. Spectra

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

matches with ion scores above the homology or identity threshold (significance threshold = 0.05) were reserved. The N-glycosylation motif NX[S/T] (X = any amino acid except proline) was also required in the peptide sequence. The false discovery rate was calculated by dividing the number of glycopeptide matches in the decoy database (FP) with the total matches of glycopeptides in the target database (TP) (FDR = FP/TP×100 %). Identification of glycopeptides from MS2 spectra: Y-type ions from each spectrum were determined from the mass differences between two adjacent fragments, and the molecular weights of four types of monosaccharides (HexNAc (203.079), Hex (162.0528), NeuAc (291.0954) and Fucose (146.0579)) were used in this step. Spectra with less than four Y ions were discarded. For each MS/MS spectrum, if the sequence of a glycopeptide had been identified by a database search from the following MS3 spectrum, the molecular weight of the peptide backbone part and the glycan part were then calculated, and the glycan composition was then identified by searching a glycan database, as described previously9. If no glycan composition could be matched from the database, a de novo strategy was then used to identify the glycan composition10 , and some incorrect/uncommon compositions were filtered with the rules described in this reference. For those spectra from intact glycopeptides, the peptide sequences could not be identified by searching their MS3 spectra. Instead, adjacent glycopeptides with known sequences were selected as candidates, and MS2 spectrum recognition was performed to determine if any fragments could be matched to sequence-known Y1 ions by the m/z and charge state. Glycopeptide identification by restricting the retention time difference. We defined the glycopeptides identified by the MS2/MS3 combined strategy as “type 1”

ACS Paragon Plus Environment

Page 8 of 21

Page 9 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

identifications and glycopeptides identified solely by MS2 as “type 2” identifications. All “type 2” identifications were further validated by retention the time difference, which was defined as “the closest distance between the retention times of a “type 2” glycopeptide and a “type 1” glycopeptide with the same peptide sequence”. For glycopeptides with the same peptide sequence but different glycan compositions, glycopeptides with smaller glycans typically elute later, whereas those with larger glycans elute earlier. As a result, glycopeptides with the smallest glycans were used as references for the group of glycopeptides that share the same peptide sequence. The criterion for the retention time difference was dependent on sample complexity and the separation ability of the LC system. In this experiment, the retention time tolerance for glycopeptides that earlier eluted was found to be 5 min, whereas that for later-eluting glycopeptides was 3 min. All “type 2” glycopeptide identifications that were out of the range were eliminated. All raw data and matched spectra could be accessed at iProX (IPX00072501). Results and Discussion Fragmentation Study Reveals Relative Abundance of Y1 Ion is Glycan-dependent We report that the relative abundance of Y1 ions in MS/MS of intact N-glycopeptides is glycan dependent, as a higher abundance of Y1 ions were detected from glycopeptides with high-mannose glycans compared with those with complex glycans (Figure S1) regardless of the CID parameters (Figure S2). Moreover, we discovered that following exoglycosidase treatment (Figure S3), the relative abundance of Y1 ions in MS/MS increased only after all sugars outside the mannose core structure were cleaved off, and any residual acetyl glucosamines attached to the mannose core significantly lowered the

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

intensity of the Y1 ions (Figure S4). By contrast, MS/MS analysis of high-mannose-type N-glycopeptides showed that increasing the number of mannose residues did not decrease the yield of Y1 ions in MS/MS (Figure S5), which was also observed from previous study11 and the fragmentation pattern of glycopeptides with various glyan type was consistent with the pattern found with time-of-flight time mass spectrometry12. These results suggest that glycosidic bonds involving β-linked acetyl glucosamine were a key factor limiting the generation of Y1 ions. Fragments generated from Gal (β14)GlcNAc were consistently the most abundant ions in the MS/MS spectra, despite the presence of sialic acid (Figure S6), which is consistent with the relative intensities of glycan oxonium ions (Figure S7). These results demonstrated that fragmentations outside the pentasaccharide core structure can drastically affect the intensity of the Y1 ions. Therefore, for glycopeptides with complex N-glycans, removing the sugars outside the core structure increases the chance of identifying peptide sequences from Y1 ions generated by glycopeptides. NGlycoReduction Generates Oligomannosylated N-glycopeptidome We developed a new approach for intact glycopeptide analysis named NGlycoReduction, in which the N-glycoproteome is enzymatically transformed into an oligomannosylated N-glycopeptidome to favor peptide sequencing (Figure 1, a). Briefly, we reduced all complex N-glycans attached to proteins to their core mannose structures using a set of exoglycosidases, which also reduces the complex microheterogeneity of N-glycosylation. Glycopeptides with core mannose structures were isolated by tryptic digestion and HILIC enrichment (Figure S8). The completeness of this process was monitored by gel staining (Figure S9), and mass spectrometric analysis revealed that greater than 90 % of fetuin

ACS Paragon Plus Environment

Page 10 of 21

Page 11 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

glycopeptides were cleaved to the core mannose structure (Figure S10). We further applied this approach to the secretome of Huh7 cells, and the secreted glycoproteins with complex N-glycans were also cleaved to their core structures (Figure 1, b). Although there is no exoglycosidase available to cleave alpha 1-6 fucose linked to N-acetyl glucosamine, core-fucosylation did not affect the generation of Y1 ions in MS/MS, and the Y1 ions remained the most abundant fragments from both glycoforms (Figure S11). From the processed secretome samples, Y1 ions became the most abundant ions in nearly 50 % of the MS/MS spectra of glycopeptides, whereas only 6 % of Y1 ions were the most abundant fragments in MS/MS spectra of the native glycoproteome (Figure 1, c), most of which were from O-glycopeptides or N-glycopeptides with high-mannose-type N-glycans (Figure S12). Thus, cleavage is necessary to characterize N-glycopeptides with complex N-glycans, and cleavage significantly improved the quality of the MS3 spectra of the glycopeptides (Figure 1, d), which increased the chance of identifying peptide sequences (Figure 1, e). This result also suggested that most secreted glycoproteins mainly have complex/hybrid N-glycans, as the majority of N-glycosites were only identified from exoglycosidase-cleaved glycoforms. Intact N-glycopeptidome Characterized with NGlycoReduction We tested whether the oligomannosylated N-glycopeptidome could be used to characterize the microheterogeneity of native N-glycosylation. Because Y1 ions from the cleaved glycoforms of fetuin glycopeptides have the same m/z and charge state as Y1 ions from intact glycopeptides (Figure S13), the peptide sequences of the intact glycopeptides can be identified from the MS3 spectra of the cleaved glycoforms by analyzing a combined sample. For complex samples, the retention times of glycoforms

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

can be added to the m/z and charge state as criteria to identify Y1 ions from different glycoforms with the same peptide sequence. From the N-glycosites identified from the pure oligomannosylated N-glycopeptidome, fewer than 2 % of isobaric peptides (∆MW≤ 1Da) eluted within the same time window (Figure S14), and they were separated by differences in glycan composition (Figure S15). Consequently, simultaneous analysis of the combined oligomannosylated N-glycopeptidome and the intact glycoproteome by data-dependent MS3 acquisition would enable the identification of both peptide sequence and glycan composition (Figure 2, a). We investigated the effect of various monosaccharides on glycopeptide retention time. Only sialic acid altered the hydrophobicity of the glycopeptides and significantly affected their retention time (Figure S16). Alternatively, the oligomannosylated N-glycopeptidome was also combined with a desialylated glycoproteome to ensure that all glycoforms with the same peptide sequence would elute in a narrow time window (Figure S17). The presence of sialylation can be confirmed by the increased peak intensity of the asialoglycopeptide after desialylation (Figure S18). A program was written to automatically identify the glycan composition and peptide sequences of intact glycopeptides (Figure S19), and the effect of setting different retention times was evaluated to set the optimal criterion (Figure S20). After filtering, the elution order of different glycoforms with the same peptide sequence was consistent with the order found from different glycoforms of fetuin glycopeptides (Figure 2, b). These results demonstrated that combining the oligomannosylated Nglycoproteome generated by NGlycoReduction with the native glycopeptidome is an accurate approach for characterizing the microheterogeneity of N-glycosylation. The improvement was significant compared with MS analysis of the native glycopeptidome

ACS Paragon Plus Environment

Page 12 of 21

Page 13 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

alone (Figure 2, c, Supplementary Table 1). We expanded this approach to analyze the Nglycosylation of Huh7 cell lysates, and 151 N-glycosites were identified (Supplementary Table 2). Although some overlap was observed in the intact N-glycopeptides from the secretome and the total lysate (Figure 3, a); different patterns in the type of N-glycans were also observed (Figure 3, b). N-glycopeptides from the cell lysates are mainly associated with high-mannose-type N-glycans, suggesting that high-mannose-type Nglycans are predominant among intracellular glycoproteins. This result is consistent with previous studies based on either glycoproteomics9 or glycomics13 approaches and demonstrates the differences between site-specific N-glycans from intracellular compartments and secreted glycoproteins. These findings also support our hypothesis that the generation of Y1 ions was favored in oligomannosylated glycopeptides regardless of the number of mannose residues, which highlighted the necessity of NGlycoReduction for samples enriched with complex N-glycans, such as cell-surface or secreted glycoproteins14,15. Overall, we demonstrated that the generation of Y1 ions from glycopeptides is greatly influenced by the glycan structure. Reducing the glycan structure to a core mannose structure greatly improved the generation of Y1 ions in MS/MS of glycopeptides with complex glycans and their identification by MS3 and spiking the reduced Nglycopeptidome into the native glycopeptidome enabled accurate identification of native site-specific N-glycosylation. Although less N-glycopeptides were identified from this study than our previous publication by using the deglycosyaltion-correlation strategy, this novel method demonstrated enhanced accuracy as the reduced N-glycopeptide and its original format would elute at nearly the same time during chromatography separation. In

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

most of the deglycosyaltion-correlation strategies, the deglycosylated and intact glycopeptides were analyzed separately, which brought uncertainity in the retention time correlation. Moreover, the shift in retention time brought by deglycosyaltion and the transformation of asparagine to aspatic acid were not predictable either, which was worsened with complex samples. The other and the most important criterion in the deglycosyaltion-correaltion method is the molecular weight. However, isobaric matches could not be excluded as there is no evidence of peptide spectra, which could be detected by MS3 acquisition in the new approach. The performance of this approach could be further increased by adding fractionation step as mixing two glycopeptidome together doubled complexity of sample. The depth in the characterization of site-specific Nglycosylation might be further increased by combination with other MS compatible separation approaches such as hydrophilic interaction chromatography16 and capillary electrophoresis11 to increase the separation of different isotopic glycofroms with various linkage. Engineered mammalian cells that generate less-complex glycoproteins have been reported previously by genetic engineering techniques17-19; however, oligomannosylated glycopeptides have not been reported using this genetic approach. Although it is possible to generate the oligomannosylated peptidtime by knocking down a set of glycosyltransferases, the impact of such a drastic manipulation on the cells, as well as on the secretion of glycoproteins, remains unclear. Besides, this technique can not be accessed by everyone in glycoprotein analysis field and the cost is much higher than just utilizing a set of exoglycosidases. Instead, NGlycoReduction can be readily applied to any sample for glycoproteomic studies using a single-step exoglycosidase cleavage.

ACS Paragon Plus Environment

Page 14 of 21

Page 15 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Conclusion We demonstrated that the generation of Y1 ions from glycopeptides is greatly influenced by the glycan structure. Reducing the glycan structure to a core mannose structure greatly improved the generation of Y1 ions in MS/MS of glycopeptides with complex glycans and their identification by MS3 and spiking the reduced N-glycopeptidome into the native glycopeptidome enabled accurate identification of native site-specific N-glycosylation. The developed approach resulted accurate identification of 609 N-glycopeptides from Huh7 cells. Supporting Information Supplementary Figures (Figure S1-S20) and Supplementary Tables, Table S1: List of identified N-glycopeptides from the secretome of Huh7 cells and Table S2: List of Nglycopeptides identified from Huh7 cell lysate

Acknowledgments D. Figeys acknowledges a Canada Research Chair in Proteomics and Systems Biology and funding from the Natural Sciences and Engineering Research Council of Canada (NSERC).

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

References (1)Ohtsubo, K.; Marth J.D. Cell 2006,126, 855-867 (2)Freeze, H.H. J. Biol. Chem. 2013,288, 6936-6945 (3)Hakomori,S. Cancer Res. 1996, 56, 5309-5318 (4)Dube, D.H.; Bertozzi C.R. Nat. Rev. Drug Discov. 2005, 4,477-488 (5)Dell, A.; Morris H.R. Science 2001, 291, 2351-2356 (6)Domon, B.; Costello C.E. Glycoconjugate J. 1988, 5,397-409 (7)Woo, C.M.; Iavarone A.T.;Spiciarich D.R.; Palaniappan K.K.; Bertozzi C.R. Nat. Methods 2015, 12,561-567 (8)Nilsson, J., Rüetschi U.; Halim A.; Hesse C.; Carlsohn E.; Brinkmalm G.; Larson G. Nat. Methods 2009, 6,809-811 (9)Cheng, K., Chen R.; Seebun D.; Ye M.; Figeys D.; Zou H. J.Proteomics 2014, 110,145-154 (10)Joenvaara, S.; Ritamo, I.; Peltoniemi, H.;Renkonen, R. Glycobiology 2008,18, 339349 (11) Koller A.; Khandurina J.; Li J.; Kreps J.;, Schieltz D.; Guttman A. Electrophoresis 2004, 25, 2003–2009 (12) Hinneburg H.; Stavenhagen K.; Schweiger-Hufnagel U,;Pengelley S .; Jabs W,; Seeberger P.H.; Silva D.V.; Wuhrer M.;Kolarich D. J. Am. Soc. Mass Spectrom. 2016, 27,507-519 (13) Fujitani, N.;Furukawa J.; Araki K.; Fujioka T.; Takegawa Y.; Piao J.; Nishioka T.; Tamura T.; Nikaido T.; Ito M.; Nakamura Y.;Shinohara Y. Proc. Natl. Acad.Sci. U. S. A. 2013,110, 2105-2110 (14) Lau, K.S.; Partridge E.A.; Grigorian A,; Silvescu C.I.; Reinhold V.N.; Demetriou M.; Dennis J.W. Cell 2007, 129,123-134 (15)Roth, J. Chem. Rev. 2002, 102,285-303 (16) Zauner G.; Koeleman C.A.; Deelder A.M.; Wuhrer M. J. Sep. Sci. 2010 33,903-910 (17) Meuris L.; Santens F.;Elson G.; Festjens N.; Boone M.; Dos Santos A.; Devos S.; Rousseau F.; Plets E.; Houthuys E.;Malinge P.; Magistrelli G.; Cons L.; Chatel L.; Devreese B.; Callewaert N. Nat. Biotechnol. 2014, 32,485-489 (18) Yang Z.; Wang S.; Halim A.; Schulz M.A.; Frodin M.; Rahman S.H.; VesterChristensen M.B.; Behrens C.; Kristensen C.;Vakhrushev S.Y.; Bennett E.P.; Wandall H.H.; Clausen H. Nat. Biotechnol. 2015, 33, 842-844 (19) Steentoft C.; Vakhrushev S.Y.; Vester-Christensen M.B.; Schjoldager K.T.; Kong Y.; Bennett E.P.; Mandel U.; Wandall H.; Levery S.B.;Clausen H. Nat. Methods 2011, 8, 977-982

ACS Paragon Plus Environment

Page 16 of 21

Page 17 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure Legends Figure 1 Identification of N-glycosylation sites by NGlycoReduction and data-dependent MS3 acquisition. a, comparison of N-glycosite identification from intact glycopeptides and N-glycopeptides cleaved by an exoglycosidase cocktail. After the removal of monosaccharides outside the pentasaccharide core, the Y1 ion became the most abundant fragment, and only one MS3 scan was required to identify the peptide sequence and localization of the N-glycosides. Conversely, the relative abundance of Y1 ions was lower in the MS/MS spectra of intact glycopeptides. Several MS3 or even MS4 scans were necessary to select the Y1 ion, which resulted in less efficient identification of Nglycopeptides; b, MS2 spectra of glycopeptides from fibronectin after exoglycosidase cleavage (upper) and its intact glycoform (lower); c, comparison of the rank of Y1 ions from the MS/MS spectra of an intact N-glycopeptidome and a cleaved N-glycopeptidome from the secretome of Huh7 cells; d, improved quality of MS/MS spectra of Nglycopeptides with NGlycoReduction. More MS3 spectra were extracted from the raw files of a cleaved N-glycopeptidome than from an intact N-glycopeptidome from the Huh7 cell secretome; e, comparison of MS3 spectra that enabled the identification of Nglycosites from the raw files of an intact N-glycopeptidome and a cleaved Nglycopeptidome from the secretome of Huh7 cells. Figure 2 Characterizing the microheterogeneity of N-glycosylation by mixing the cleaved N-glycopeptidome with the intact N-glycopeptidome. a, scheme of the mixing strategy. The cleaved glycopeptides were added to the desialylated glycopeptides or intact glycopeptides and analyzed by LC-MS with data-dependent MS3 acquisition. Glycopeptides bearing the same peptide backbone were grouped by their retention times and the m/z of the Y1 ion. The peptide sequence was identified from the MS3 spectra of the cleaved glycopeptides; b, distribution of the retention times of different glycoforms bearing the same peptide sequences. Glycoforms whose peptide sequences were identified by MS3 are labeled in pink, whereas glycoforms identified by Y1 ion matching are labeled in blue. Cleaved glycoforms were found to elute later than desialylated glycoforms and earlier than glycoforms containing sialic acid. The elution order is consistent with that observed for fetuin glycopeptides. c, a significantly higher number of spectral matches are obtained with unique glycopeptides and unique peptide sequences from the combined analysis of the cleaved glycopeptidome with native glycopeptides than by analyzing the native glycopeptidome alone. The number of total intact Nglycopeptides identified from the secretome increased 2-fold using the developed approach; Figure 3 Comparison of glycopeptides identified by NGlycoReduction from the lysate and secretome of Hhu7 cells. a,overlap between the peptide sequences and intact Nglycopeptides identified from the secretome and lysates of Huh7 cells; b, hierarchical clustering of intact N-glycopeptides from the secretome and lysates of Huh7 cells by spectral counts reveals different patterns of intracellular and extracellular N-glycosylation.

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 1

ACS Paragon Plus Environment

Page 18 of 21

Page 19 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

Figure 2

ACS Paragon Plus Environment

Analytical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 3

ACS Paragon Plus Environment

Page 20 of 21

Page 21 of 21

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

For Table of Contents Only

ACS Paragon Plus Environment