Subscriber access provided by TULANE UNIVERSITY
Article
An Integrated Pipeline of Isotopic Labeling and Selective Enriching for Quantitative Analysis of N-Glycome by Mass Spectrometry Lijun Yang, Xiaoxian Du, Ye Peng, Yan Cai, Lei Wei, Ying Zhang, and Haojie Lu Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.8b04525 • Publication Date (Web): 17 Dec 2018 Downloaded from http://pubs.acs.org on December 18, 2018
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 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 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.
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 10 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
Lijun Yanga, Xiaoxian Dua, Ye Pengb, Yan Caib, Lei Weia, Ying Zhang*a,b and Haojie Lu*a,b a
Shanghai Cancer Center and Department of Chemistry, Fudan University, Shanghai, 200032, P. R. China. Institutes of Biomedical Sciences and Key Laboratory of Glycoconjugates Research Ministry of Public Health, Fudan University, Shanghai 200032, P. R. China. b
ABSTRACT: Quantitative N-glycomics can reveal abnormal expression of N-glycan in diseases. However, mass spectrometry (MS)-based N-glycome quantitative analysis is still technically challenging. To achieve the quantitation of N-glycome with high accuracy and sensitivity, it is required to efficiently label the Nglycans with isotopic tags and selectively enrich N-glycans to avoid suppression from other substances. Herein, we developed an integrated pipeline that combines isotopically fluorous tag labeling and fluorous solid-phase extraction to quantitatively analyze the N-glycome by MS. In this strategy, the N-glycans were labeled with light- and heavy aminoxy-functionalized fluorous tags (PFBHA and PFBHA-d2) through the oxime click reaction. Then, through the fluorous solid-phase extraction (FSPE), the fluorous tag labeled N-glycan could be purified from contaminants like salts and proteins for the following quantitative analysis by mass spectrometry. This new approach enables the selective purification (molar ratio of glycan to protein at 1:100), and the accurate (R2>0.99) and reproducible (CV95%), making the quantitation reliable. Second, the fluorous tag enables the selective purification of the labeled N-glycans through fluorous affinity interaction with high specificity for the following MS analysis. Third, isotopic PFBHA labeling enables the accurate quantitation of glycans with high reproducibility (CV0.99) within 2 orders of magnitude. Moreover, the PFBHA labeling provides unique diagnostic ions (D and [D221] ions) for identifying the composition of the 6-antenna and distinguishing the N-glycan isomers.
Page 2 of 10
lution was added, followed by incubation at 37 °C overnight. To release glycans from serum, the pooled serum was quantified by Bicinchoninic acid (BCA) at 562 nm and then equal amounts of serum (containing about 400 μg proteins) were reduced with 10 mM DTT followed by alkylation with 20 mM IAA. After that, low molecular compounds were removed by ultrafiltration (MWCO, 3 kDa). The collected proteins were methylamidated to avoid the loss of sialic acid during MALDI analysis. The methylamidation was performed as reported with minor modification.25 Briefly, the proteins were dissolved in 100 L of DMSO solution containing 5 M methylamine hydrochloride, and then 100 L of PyAOP (250 mM in 30% N-methylmorpholine/DMSO) solution was added. After that, the solution was shaken at room temperature for 2 h. After reaction, the solution was diluted to reduce the DMSO to less than 5% and the proteins were collected by ultrafiltration (MWCO, 3 kDa). After the proteins were redissolved in 25 mM ABC buffer (pH 8.0), PNGase F solution was added and then the solution was incubated at 37 °C overnight and then stored at -20 °C for future use. PFBHA Derivatization. The DP7 was dissolved at concentration of 1.15 mg/mL (1 mM) in distilled water. PFBHA was dissolved at concentration of 24.96 mg/mL (100 mM) in MeOH. 1 L of DP7 was lyophilized before labeling, and 5 L of PFBHA and 15 μL of MeOH were added and the solution was incubated at 60 °C for 2 h. At last, the labeled glycans were enriched by FSPE or subjected to MS analysis. Fluorous Solid Phase Extraction. 15 mg of nGF was first washed with MeOH and then with 10% MeOH. The glycans labeled with PFBHA were redissolved in 200 L of 10% MeOH and added into the nGF materials, and then incubated for 60 min. After binding, the nGF materials were washed once with 200 L of 10% ACN and twice with 200 L of 10% MeOH to remove nonspecifically bounded proteins and salts. After that, the glycans were eluted with 300 L of MeOH via 20 min incubation. The solution phase was then collected through centrifugation and then lyophilized. At last, the collected glycans were redissolved in distilled water prior to MS analysis. MALDI Sample Preparation and Analyses. One microliter of sample was loaded onto a MALDI target and mixed with 1 L of DHB solution and dried in the air prior to MS analysis. The MALDI-TOF MS and MS/MS spectra were performed on 5800 Proteomics Analyzer (Applied Biosystems, Framingham, MA, USA) equipped with a Nd:YAG laser (355 nm), an acceleration voltage of 20 kV and a repetition rate of 400 Hz. The mass spectrometer was operated in the positive, negative reflection and linear mode according to different samples. In addition, MS/MS spectra were interpreted manually with the aid of the GlycoWorkbench software. External mass calibration was performed using peptides from myoglobin digests.
Materials and Chemicals. Trifluoroacetic acid (TFA), (7azabenzotriazol-1-yloxy) tripyrrolidinophosphonium hexafluorophosphate (PyAOP), methylamine hydrochloride, 4methylmorpholine, dithiothreitol (DTT), iodoacetamide (IAA), o-(2, 3, 4, 5, 6-pentafluorobenzyl)hydroxylamine hydrochloride (PFBHA), 2, 5-dihydroxybenzoic acid (DHB), bicinchoninic acid kit for protein determination, ribonuclease B from bovine pancreas (RNase B), asialofetuin from fetal calf serum (ASF), fetuin from fetal bovine serum, albumin from chicken egg white (OVA), bovine serum albumin (BSA) and ammonium bicarbonate (ABC) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Deuterium labeled o-(2, 3, 4, 5, 6pentafluorobenzyl)hydroxylamine hydrochloride-d2 (PFBHAd2) was obtained from Puen Scientific instrument Co., Ltd (Guangzhou, China). Nanographite fluoride (nGF) was obtained from Xianfeng Nanotech (Nanjing, China). Maltoheptaose (DP7, 95%) was obtained from Hayashibara Biochemical Laboratories (Okayama, Japan). A1 glycan was purchased from Ludger, Ltd (Oxford, UK). Peptide N-glycosidase F (PNGase F, 500 U/μL) was obtained from New England Biolabs (Ipswich, MA, USA). Centrifugal filters with MWCO of 3 kDa were obtained from Millipore (Bedford, MA, USA). HPLC-grade methanol (MeOH) and acetonitrile (ACN) were obtained from Merck (Darmstadt, Germany). Distilled water was purified by a Milli-Q system (Milford, MA, USA). Serum from all participants including healthy individuals (n=3) and hepatocellular carcinoma (n=3) were obtained from Fudan University Shanghai Cancer Center. The research was handled in accordance with ethical and legal standards. Sample Preparation. To release N-glycans from standard glycoproteins, the solution of glycoproteins was first denatured at 100 °C water bath for 10 min and then PNGase F so-
PFBHA Derivatization Combined with FSPE. As shown in Scheme 1, N-glycans are labeled with PFBHA or PFBHA-d2 at the reducing end through oxime click chemistry. Then the fluorinated glycans are selectively enriched by nanographite fluorides through fluorous affinity interaction, while the unlabeled substances (i.e., proteins, salts, and other contaminants) are removed during the washing step. First, glycan DP7 was used as a model to optimize the reaction conditions. As shown
2 ACS Paragon Plus Environment
Page 3 of 10 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
Scheme 1. (a) PFBHA derivatization of N-glycans through oxime click chemistry and (b) Workflow of enriching the fluorinated glycan by FSPE for MS analysis.
Figure 2. MALDI-TOF mass spectra of a mixture of PFBHA fluorinated DP7 and BSA (1:100, molar ratio). (a, b) directly analyzed without enrichment and (c, d) analyzed after FSPE treatment. (a, c) were obtained from positive reflection mode, and (b, d) were obtained from positive linear mode. [M] denotes the signal of BSA. The low abundance and low ionization efficiency of Nglycans usually leads to poor detection of glycans in MS. Therefore, it is necessary to selectively enrich glycans from complex biological mixtures to increase the detection sensitivity before MS analysis. As shown in Figure S-2, PFBHA labeled N-glycans can be selectively enriched via fluorous affinity interaction. Before enrichment, the S/N ratio of fluorinated DP7 was extremely low (S/N 41.34), Figure S-2a. After FSPE treatment, the S/N of fluorinated DP7 increased to 629.39, Figure S-2b. Even when the starting concentration of glycans was 25 fmol/L, the fluorinated DP7 could also be enriched and clearly observed after enrichment (Figure S-3). To confirm that the glycans were enriched through fluorous affinity interaction, equal amounts of native DP7 and fluorinated DP7 were mixed and enriched by FSPE. Before FSPE treatment, both native DP7 and fluorinated DP7 were observed in the spectrum (Figure S-4a). After FSPE treatment, only fluorinated DP7 was detected in the elution fraction (Figure S-4c). Besides, none of fluorinated DP7 was observed in flowthrough of loading solution (Figure S-4b) showing high binding affinity of this method. These results confirmed that PFBHA fluorinated glycans were selectively enriched through fluorous affinity interaction, rather than the interaction with the native glycan moiety. It should be noted that, to check whether the unreacted PFBHA compete with the fluorinated glycan to bind to the nGF materials, fluorinated DP7 sample containing the unreacted PFBHA was captured by different amount of nGF. The flowthrough of loading solution was analyzed by MS to confirm whether all of the fluorinated DP7 can be captured by the nGF materials with the interference of excess unreacted PFBHA. And native DP7 was added as an internal standard to quantify the remaining fluorinated DP7. As shown in Figure S-5, when the amount of nGF materials was increased to 15 mg, none of fluorinated DP7 could be observed in the
Figure 1. MALDI-TOF mass spectra of (a) PFBHA fluorinated DP7, (b) PFBHA fluorinated A1 glycan (methylamidated), (c) PFBHA-d2 fluorinated DP7, and (d) PFBHA-d2 fluorinated A1 glycan (methylamidated). in Figure S-1, when the reaction was performed in MeOH solution at 60 °C for 2 h with the concentration of PFBHA at 25 mM, the labeling efficiency reached almost 100%. After being derivatized with PFBHA, the fluorinated DP7 appeared at m/z 1370.38 ([DP7 + PFBHA + Na]+) and m/z 1386.49 ([DP7 + PFBHA + K]+), while the native DP7 signal (m/z at 1175.28, [DP7 + Na]+) disappeared, indicating the high labeling effi- ciency (Figure 1a). The presence of fluorinated DP6 here was due to the impurities of the original sample. The labeling efficiency was same for PFBHA-d2 derivatization (Figure 1c). The labeling is also applicable to sialic glycans with high labeling efficiency (Figure 1b and 1d). For sialylated glycans, the A1 glycan was first methylamidated (+13.01 Da per reactive site) to avoid the loss of sialic acid during MS analysis. It can be seen that the A1 glycan was efficiently fluorinated, without the detection of native A1 (methylamidated, m/z at 1967.71). Therefore, PFBHA derivatization was applicable to both neutral and sialylated glycans with high labeling efficiency.
3 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
Page 4 of 10
Figure 3. MALDI-TOF/TOF tandem mass spectra of PFBHA fluorinated N-glycans (a) Man5GlcNAc2 and (b) Man6GlcNAc2 from RNB, (c) Gal2Man3GlcNAc4 and (d) Gal3Man3GlcNAc5 from ASF, (e) Man5GlcNAc4 and (f) Man3GlcNAc3 from OVA, (g) Gal2Man3GlcNAc4Fuc from IgG, and (h) Gal3Man3GlcNAc5NeuAc3 from fetuin (methylamidated). [M’] denotes the mass of sodium-adduct of native N-glycan. All the precursor ions chosen for MS2 experiments were sodium-adducts and the fragments were noted according to the Domon and Costello nomenclature.
4 ACS Paragon Plus Environment
Page 5 of 10 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
flowthrough after loading, indicating completed adsorbtion of fluorinated glycans . As co-existing proteins, peptides, salts and other small molecules would hamper the detection of N-glycans in MS, enrichment of N-glycans from complex biological samples is a prerequisite before MS analyses. Figure S-6 shows that the PFBHA fluorinated DP7 was efficiently purified from different types of salt solutions. Before enrichment, the signal of fluorinated DP7 was severely suppressed by salts including a high concentration of chaotropes (2 M urea), inorganic salts (1 M DTT), saturated NH4HCO3 (2.6 M) and saturated NaCl (6.2 M). However, after enrichment, the quality of mass spectra were dramatically improved. To further test the ability of purifying glycan from protein mixtures, a mixture of PFBHA fluorinated DP7 and BSA (1:100 molar ratios) was tested. The signal of fluorinated DP7 was poor (S/N 44.3) due to the presence of BSA (Figure 2a). However, glycans were well detected after FSPE treatment (Figure 2c). In addition, from the mass spectra of untreated sample and elute fraction (linear mode), it can be concluded that the BSA was well removed (Figure 2b and 2d). Overall, PFBHA derivatization combined with FSPE showed good performance in purifying the glycans from different types of salts and mixtures of proteins. Structural Elucidation of Fluorinated Glycan. Tandem mass spectrometry (MSn) is demonstrated as a useful tool in glycan identification and structural characterization. We then investigated how the PFBHA labeling affected the fragmentation patterns of the glycan. For fragmentation of a linear glycan DP7 (precursor ion [DP7 + Na]+), only several B/Y ions (cleavages at glycosidic bonds) were detected, which provided limited structural information (Figure S-7a). After derivatization with PFBHA, except for B/Y ions, more complete [C-2] fragments (2 Da lower than C ions) were observed, so that more composition and sequence information of the glycan can be obtained. Furthermore, several cross-ring ion fragments were also obtained, allowing for the analysis of linkage and branching information of the glycan (Figure S-7b). To further verify the superiority of the PFBHA fluorinated tags labeling in terms of the structural analysis of glycans, the fragmentation of different kinds of PFBHA fluorinated N-glycans were studied. As shown in Figure 3, for high-mannose, hybrid, complex types and bisecting N-glycans, fragments of [C-2] ions were observed along with B/Y ions, which were helpful for the composition and sequence analysis of the glycan. Surprisingly, antenna-specific D and [D-221] ions (D-GlcNAc) were observed, which were often observed in negative ion mode and used as diagnostic ions.34,35 The D ion is generated from the loss of the chitobiose core and loss of the residues of the 3-antenna. Accordingly, the D ion only contains the 6antenna and the branching mannose residue, which can be used to identify the composition of the 6-antenna and distinguish the isomeric configuration. For example, there is the same mass of D ions at m/z 671 in Figure 3a and 3b, indicating that three mannose residues were located in the 6-antenna for both glycans, so the other two mannoses of Man6 were in the 3-antenna.36 Similarly, glycan Gal2Man3GlcNAc4 and Gal3Man3GlcNAc5 produced the same mass of D ions at m/z 712 (Figure 3c with 3d), indicating the same composition of 6-antenna (Gal-GlcNAc-Man) of these two glycans, and further confirming that the third antenna was located in the 3antenna but not on 6-antenna for the glycan of Gal3Man3GlcNAc5.37 This was also the same for methylamidated sialylated N-glycan from fetuin with the D ions at m/z
Figure 4. MALDI-TOF mass spectra of the mixture of PFBHA and PFBHA-d2 labeled DP7 with equal molar ratio of (a) directly analyzed, (b) analyzed after store at -20 ℃ for one month, and (c) dynamic range and accuracy of the glycan quantitation with PFBHA and PFBHA-d2 labeling (n= 6). 1015.94, indicating that the composition of 6-antenna is NeuAc-Gal-GlcNAc-Man (Figure 3h).37 For bisecting N-glycans, the [D-221] ion can be used as a diagnostic ion to confirm the presence of a bisecting GlcNAc residue, as shown in Figure 3e and 3f.38 The fragmentation patterns of glycan changed greatly after PFBHA derivatization. We assumed that the change of fragmentation patterns might be attributed to the unique electronegativity of fluorine on PFBHA, which caused the change of electron transfer and led to the formation of [C2], D and [D-221] ions. To further demonstrate the ability of PFBHA derivatization in the structural elucidation of the isomeric N-glycan (Figure 3c and 3e), we chose a pair of isomeric N-glycans with the m/z 1663.58 from ASF and OVA. After PFBHA fluorination, m/z of their sodium-adducts is 1858.59. As shown in Figure 3c, the consecutive fragments of [C-2] ions were detected, and
5 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
Page 6 of 10
Figure 5. MALDI-TOF mass spectra of an equimolar mixture of PFBHA and PFBHA-d2 labeled N-glycans from (a) RNase B, (d) ASF, (g) OVA, and (j) fetuin (methylamidated). (b, c) were the enlarged spectra of RNase B (a), (e, f) were the enlarged spectra of ASF (d), (h, i) were the enlarged spectra of OVA (g), and (k, l) were the enlarged spectra of fetuin (j). “L” denotes PFBHA labeled glycans and “H” denotes PFBHA-d2 labeled glycans. (Supporting method, Table S-1 and Table S-2).23,39 After deconvolution, the calculated ratio was 1.05, which was in agreement with the theoretical ratio. Even after being stored at -20 °C for one month, the ratio of light/ heavy-labeled DP7 did not change, indicating the good stability of the PFBHA labeled complex (Figure 4b). The light/heavy-labeled DP7 mixed at different ratios (1:10, 1:5, 1:2, 2:1, 5:1, and 10:1) were also investigated. As shown in Figure 4c, this quantitation method has a good reproducibility and accuracy (R2=0.9997, CV3) from human serum were observed and their proposed structures were presented. The successful detection of N-glycans from serum was attributed to efficient glycan derivatization and efficient removal of proteins and salts through FSPE treatment. This result indicated that PFBHA fluorination combined with FSPE treatment was suitable for highly complex sample analysis. Later, to further investigate the enrichment combining with the quantitation of N-glycan in complex biological samples, the Nglycans from healthy human serum and HCC serum were respectively fluorinated with light/heavy tags, and mixed and put through FSPE treatment. To provide an internal standard, equimolar of native DP7 was added into both samples to correct the ratio. In addition, the same amount of N-glycans from the healthy control group were respectively fluorinated with light/heavy tags and then subjected to FSPE treatment. Most of the PFBHA/PFBHA-d2 ratios of N-glycans from same control serum were between 0.8 and 1.3, so HCC/normal ratios of >1.3 or 0.99, CV
