Phenylboronic acid functionalized C3N4 facultative hydrophilic materials for enhanced enrichment of glycopeptides

Author’s Accepted Manuscript Phenylboronic Acid Functionalized C 3N4 Facultative Hydrophilic Materials for Enhanced Enrichment of Glycopeptides Yong Zhang, Hongyu Jing, Tao Wen, Yao Wang, Yang Zhao, Xiangke Wang, Xiaohong Qian, Wantao Ying www.elsevier.com/locate/talanta

PII: DOI: Reference:

S0039-9140(18)30927-5 https://doi.org/10.1016/j.talanta.2018.09.016 TAL19030

To appear in: Talanta Received date: 22 June 2018 Revised date: 3 September 2018 Accepted date: 5 September 2018 Cite this article as: Yong Zhang, Hongyu Jing, Tao Wen, Yao Wang, Yang Zhao, Xiangke Wang, Xiaohong Qian and Wantao Ying, Phenylboronic Acid Functionalized C3N4 Facultative Hydrophilic Materials for Enhanced Enrichment of Glycopeptides, Talanta, https://doi.org/10.1016/j.talanta.2018.09.016 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Phenylboronic Acid Functionalized C3N4 Facultative Hydrophilic Materials for Enhanced Enrichment of Glycopeptides Yong Zhanga1, Hongyu Jinga1, Tao Wenb, Yao Wanga, Yang Zhaoa, Xiangke Wang*b, Xiaohong Qian*a, Wantao Ying*a

a

State Key Laboratory of Proteomics, Beijing Proteome Research Center, National Center for

Protein Sciences (Beijing), Beijing Institute of Lifeomics, Beijing, 102206, China b

College of Environmental Science and Engineering, North China Electric Power University,

Beijing 102206, China

E-mail addresses: [email protected] (X. Wang), [email protected] (X. Qian), [email protected] (W. Ying).

* Corresponding authors.

1

These authors contributed equally to this work. 1

ABSTRACT It is challenging to capture N-glycopeptides with high recovery and high specificity from complicated biosystems. Herein, we present a facile and economical procedure to generate a novel self-assembling 4-Mercaptobenzene boronic acid functionalized and Au-doped Straticulate C3N4 (MASC), with enhanced affinity capability towards glycopeptides. The materials possess low pH value adaptation, high hydrophilicity and stability, good repeatability and recyclability, and provided high selectivity (1:100), low limit of detection (0.33 fmol/μL), high enrichment efficiency (~80%) and high recovery rate (~90%) towards glycopeptides. The materials can capture glycopeptides unbiasedly, as demonstrated by the identification of 37 glycopeptides from IgG and 21 glycopeptides from horseradish peroxidase (HRP). The performance of MASC on human urine and serum glycoproteome analysis was also tested. An average of 1465 glycopeptides from 839 glycoproteins and 1553 glycopeptides from 884 glycoproteins were identified from female and male urine samples in a single mass spectrometry analysis. O-glycopeptides from human urine were also significantly enriched. Additionally, 463 glycopeptides assigned to 209 glycoproteins were identified from 5 μL of human serum. All of these results indicate that MASC presents a good performance and applicability in the field of glycoproteomic research.

Graphical abstract

2

Keywords: Boronic Acid, Hydrophilic, Facultative, Glycosylation

3

1. Introduction Protein glycosylation plays an important role in many physiological activities and disease states, especially in the progression of tumors. To understand the relevance of glycosylation to diseases, it is important to identify glycosites and reveal the glycan heterogeneity at each glycosite (intact glycopeptides) by using mass spectrometry (MS)[1, 2]. However, it is a great challenge to perform the large-scale profiling of intact glycopeptides in complex biosamples because of the signal inhibition from the non-glycopeptide matrix and the relatively low abundance of glycopeptides. Therefore, the enrichment of intact glycopeptides is crucial prior to the MS analysis. Various enrichment reagents or materials for glycopeptides have been developed, such as lectin affinity chromatography boronate affinity chromatography, hydrazide chemistry, hydrophilic interaction chromatography and titanium dioxide chromatography[3, 4]. However, all of these methods have their pros and cons for glycoprotein and/or glycopeptide enrichment. For example, due to the selective binding of the lectin to certain glycan moieties, a lectin can only be applied for the enrichment of a subset of the glycoproteins. Boronic acid-functionalized materials have a unique ability to form reversible, covalent bonds with monosaccharides that featuring vicinal diols[5], but they have not been widely applied in glycoproteomic studies because of their weak binding constants[6]. Hydrazide chemistry is a nonbiased approach and provides a high enrichment efficiency for glycopeptides. However, the glycan structures are destroyed, and it is a challenge to obtain intact glycopeptides. Titanium dioxide-based enrichment has been reported by Larsen et al. The negatively charged sialic acid likely prefers to bind to the TiO2 via multidentate binding[3]. HILIC-based methods selectively enrich glycopeptides by utilizing the hydrophilicity of the glycans, but the selectivity is affected by the interference from hydrophilic peptides. It is worth noting that most of these methods are highly complementary, and combining them can bring about certain enhancements in the depth and coverage of the glycoproteome. Hence, we reason that to reach a satisfactory performance on glycopeptide enrichment, the development of a facile

4

functionalization scheme combining boronate affinity and hydrophilic interaction may be a good choice. Borate materials are useful for the specific capture of cis-diol-containing biomolecules, based on the reversible covalent complex formation/dissociation between boronic acids and cis-diols in alkaline/acidic aqueous solution[7]. However, as in the boronic acid affinity method, an apparent disadvantage is that the enrichment of glycopeptides has to be performed in alkaline media, which can lead to the degradation of labile glycans[8]. An effective solution to this problem is to synthesize novel boronic acids based materials by introducing neighboring groups (e.g. amino)[9]. According to the above strategy, it is possible to prepare boronate-functionalized materials that are functional under neutral conditions or even acidic conditions. Herein, we expect to prepare a new material, that could not only optimize the application environment of boronic acid enrichment, but also maintains good hydrophilicity. The hydrophilic property of the base materials can aggregate glycopeptides to the material surface and increase their local concentration, while the boronic acid is expected to provide the capability of enrichment. To achieve this goal, straticulate carbon nitride (s-C3N4) was chosen as a suitable material. s-C3N4, a typical 2D substrate material, has been used to fabricate various functional materials with excellent enrichment performance for heavy metal ions and organic pollutants in different fields[10]. Ultrathin C3N4 nanosheets have been proven to be hydrophilic. More promisingly, the water suspension of s-C3N4 is stable under both acidic (pH=3) and alkaline (pH=11) environments[11]. Herein, we developed the novel phenylboronic acid functionalized s-C3N4 based facultative materials (MASC) with the aim of achieving the highly specific enrichment of glycopeptides. 2. EXPERIMENTAL SECTION 2.1. Materials and chemicals Horseradish peroxidase (HRP), bovine serum albumin (BSA), human immunoglobulin G (human IgG), dithiothreitol (DTT), iodoacetamide (IAA), formic acid (FA), and trifluoroacetic acid (TFA) were purchased from Sigma (St. Louis, MO, USA). Acetonitrile 5

(ACN), ethanol (EtOH), methanol (MeOH) and acetic acid (HAc) were purchased from Merck (Darmstadt, Germany). Alldeionized water in the experiment was prepared by a Milli-Q system (Millipore, Bedford, MA). Melamine, HAuCl4•4H2O, NaBH4, AgNO3, and 4-mercaptobenzene boronic acid were purchased from Aladdin-Reagent (Shanghai, Peking). 2,5-Dihydroxybenzoic acid (DHB) was obtained from Bruker (Daltonios, Germany). Peptide-N-glycosidase (PNGase F) was purchased from Sigma (St. Louis, MO, USA). Stable-isotope dimethyl-labeling reagents were purchased from Cambridge Isotope Laboratories (Cambridge, MA, USA). Human serum and urine were provided by Beijing First Hospital from healthy volunteers. Acetonitrile (ACN) was purchased from Merck (Darmstadt, Germany). All of the other chemicals were of analytical grade. A Sep-Pack C18 cartridge was purchased from Waters Corporation (Milford, MA). 2.2. Apparatus The morphology of the samples was characterized using a ZEISS EVOLS10 scanning electron microscope at an accelerating voltage of 1 kV. Powder X-ray diffraction (XRD) patterns were recorded with a Philips X’Pert Pro Super X-ray diffractometer equipped with Cu-Kα radiation (λ=1.54178 Å) at a scanning rate of 0.05° min-1. Fourier transform infrared (FTIR) spectra were collected on a Nicolet Magana-IR 750 spectrometer with the KBr pellet technology. The magnetic properties were determined using an MPMS-XL SQUID magnetometer. X-ray photoelectron spectroscopy (XPS) measurements were utilized to analyze the element information with a PHI QuanteraII with Al Ka X-rays (1486.6 eV). The zeta potential values at different pH conditions were measured by a ZETASIZER 3000 HSA system. 2.3. Synthesis of straticulate-C3N4 All reagents were of analytical grade and used without further purification. For the synthesis of straticulate-C3N4 (s-C3N4), melamine was heated at 600 °C for 2 h under air-conditions with a ramp rate of approximately 3 °C/min for both the heating and cooling processes, and the obtained yellow product was the bulk g-C3N4 powder. Approximately 200 mg of bulk g-C3N4 powder was dispersed in 300 mL water and then heated by microwave for 6

approximately 12 h at 95 °C and magnetic stirring at 400 rpm with the microwave actor parameters at 200 W. The initially formed suspension was then centrifuged for 2 min at approximately 1000 rpm to remove the residual unexfoliated g-C3N4. The straticulate-C3N4 (sC3N4) materials exist in suspension. 2.4. Synthesis of MASC In a typical process, an aqueous solution of HAuCl4 (100 μL, 20 mM) was added dropwise to 100 mL of the s-C3N4 suspension (1 mg/mL) and stirred overnight. The deposit was centrifuged from the mixture and washed with 50% methanol through repeated centrifugation, decanting, and resuspension 2 times. Then, 500 μL of fresh NaBH4 (30 mM, 4 °C) was added to the mixture. The color of the solution was changed to light purple. The obtained solid was isolated by centrifugation and washed with 50% methanol until no chloride ions were detected by the AgNO3 test. The centrifuged deposit was dispersed with 50 mL methanol, and then added with 4-mercaptobenzene boronic acid (1 mL, 20 mM) was added, dissolved by the methanol solution, and stirred overnight. The final product was obtained by centrifugation, washing, and drying. 2.5. Tryptic digestion of the standard proteins, human serum and urine The standard glycoproteins (HRP/IgG) were dissolved in 50 mM NH4HCO3 (pH=8.5), and then the protein solution was denatured at 95 °C for 10 min. After cooling down to room temperature, the solution was incubated with trypsin (trypsin : protein=1: 50, w/w) at 37 °C overnight. For the preparation of the human serum digestion, 5 μL human serum was dissolved in 195 μL 50 mM NH4HCO3 (pH=8.5), and then the protein solution was denatured at 95 °C for 10 min. The solution was reduced by 20 mM DTT for 45 min at 56 °C and alkylated by 50 mM IAA in the dark for 1h at room temperature. Trypsin (trypsin : protein=1:50, w/w) was added to proceed with the protein digestion at 37 °C overnight. The reaction was terminated with TFA to a final concentration of 1%, and then the solution was desalted using C18 Sep-Pak cartridges (Waters). Finally, the tryptic digestion was lyophilized and stored at -80 °C for further use. 7

For the preparation of the human urine digestion, midstream urine in the morning was collected into a 50 mL centrifuge tube and centrifuged at 3000 g for 15 min and at 12000 g for 15 min. Afterward, 10 mL supernatant was mixed with 40 mL acetone and the mixture was stored at -20 °C overnight. The precipitate was collected by centrifugation (2000 g for 5min) and dissolved in 1 mL of 8 M UA buffer (8 M urea in 0.1 M Tris-HCl, pH=8.5). Most of the urine proteins can be dissolved in 8M UA via sonication. The solution was centrifuged at 1000 g for 1 min and the obtained supernatant was transferred to an ultrafiltration cartridge with a molecular weight cut off of 3,000 Da. 200 μL of UA buffer was loaded onto the ultrafiltration cartridge. The urine proteins were reduced with 200 μL of 20 mM DTT for 4 h at 37 °C and alkylated with 100 μL of 50 mM iodoacetamide for 30 min at 25 °C in the dark. After washing the ultrafiltration membrane three times using 50 mM NH4HCO3 buffer (pH 8.5), trypsin was added with an enzyme to protein ratio of 1:50, and the solution was incubated overnight at 37 °C. The protein digest was collected via centrifugation at 13,000 g for 15 min and washed three times in 50 mM NH4HCO3 and stored at -80 °C after they were lyophilized. 2.6. Selective enrichment of intact glycopeptides by enrichment materials Briefly, 100 μg materials were added to 1 μg tryptic digests dissolved in 100 μL ACN/H2O/TFA solution (80:20:0.1, v/v/v) and rotated for 2 h at room temperature. Then, the mixture was washed three times for ten minutes each with 500 μL ACN/H 2O/TFA solution (80:20:0.1, v/v/v), followed by centrifugation at 2,000×g for 1 min. The bound glycopeptides were eluted three times with 200 μL of 0.1% TFA by continuously washing at room temperature for 10 min and collected by centrifugation at 14,000×g for 15 min. The obtained glycopeptides were dried by SpeedVac and stored at -20 °C for further use. 2.7. Deglycosylation of N-glycopeptides by PNGase F digestion The enriched intact glycopeptides were dissolved in 50 μL 50 mM NH4HCO3 solution, and incubated with 5 U PNGase F overnight at 37 °C. After desalting, the obtained deglycosylated peptides (containing O-glycopeptides) were analyzed by LC-MS/MS. 2.8. Evaluation of the recovery rate for glycopeptide enrichment 8

Briefly, the tryptic digests of human IgG (10 μg) were respectively labeled with light and heavy dimethyl-labeling reagents. The heavy-tagged human IgG digest was enriched with 1 mg of MASC, and the eluent was dried and blended with the light-tagged human IgG digests. The combined mixture was further enriched by MASC, and the eluent was deglycosylated using PNGase F. The obtained deglycosylated peptides were analyzed by MALDI-TOF MS. The recovery was calculated by the peak intensity ratio of the heavy-tagged to the light-tagged deglycosylated peptides. 2.9. Mass spectrometry analysis All MALDI spectra were taken from a Bruker Ultraflex III MALDI-TOF/TOFMS instrument (Bruker, Daltonios, Germany). 1 μL of elution and 1 μL of DHB (20 mg/mL, 0.1% TFA in 70% ACN solution) matrix solution were dropped onto a MALDI plate for analysis after performing external calibration with a peptide calibration standard (Bruker). The laser intensity was kept constant for all samples. Spectra were acquired in the positive ion reflector mode. For the glycopeptide (deglycosylated peptides) and intact glycopeptide identification, HPLC-MS/MS analysis was performed using a Q Exactive HF Mass Spectrometer (ThermoScientific, USA) equipped with a nanoelectrospray ionization source and an Ultimate 3000 high-performance liquid chromatography system (Thermo, USA). The samples were dissolved in 0.1% FA and separated on a capillary column (150 µm id ×120 mm) packed with C18 (3 µm, 100 Å,) at a flow rate of 600 nL/min. The mobile phase consisted of 0.1% formic acid in water (A) and 0.1% formic acid in ACN (B). Mobile phase A (99.9% water/0.1% FA) and mobile phase B (99.9% ACN/0.1% FA) were used, and the elution gradient used was from 6 to 32% mobile phase B for 78 min. Data acquisition was performed using the data-dependent mode. The Q-Exactive HF was run under positive mode. The MS1 was analyzed with a mass range of 300-2,000 at a resolution of 120,000 at 200 m/z. The automatic gain control (AGC) was set as 3e6, and the maximum injection time (MIT) was 80 ms. The MS2 was analyzed in the data-dependent mode for the 20 most intense ions subjected to fragmentation in the Orbitrap. For each scan, the AGC was set at 5e4, and the MIT was set at 9

80 ms. The dynamic range was set at 12 s to avoid repeated detection of the same ion peaks. 2.10. Data Analysis The data from the MALDI-TOF mass spectrometer were analyzed using flexAnalysis (Version 3.3). The raw data files of the deglycopeptides from the Q Exactive HF mass spectrometer were searched against the human UniProt database (Release 2014-09, 140916 entries) using MaxQuant software (version1.5.3.8). The search parameters were set as follows: fixed modification of cysteine residues (+57.02 Da), variable modification of methionine oxidation (+15.99 Da), protein N-terminal acetylation (+42.01 Da) and asparagine deamidation (+0.98 Da). Two missed cleavages were allowed for trypsin digestion. All other settings were set at the default values, and the protein groups were filtered to 1% FDR based on the number of hits obtained for searches against the forward and decoy databases. Only glycopeptides with N-X-S/T/C (X≠P) were considered as highly reliable results. The raw data files of intact N-glycopeptides from human serum and urine were searched against the human UniProt database (Release 2014-09, 140916 entries) using pGlyco (v2.1.2, Institute of Computing Technology, Chinese Academy of Science.). The search parameters were as follows: the initial precursor and fragment mass tolerances were set at 10 and 20 ppm, respectively, and up to two missed cleavages were allowed for trypsin digestion. Cysteine carbamidomethylation was set as a fixed modification, methionine oxidation, asparagine deamidation, and protein N-terminal acetylation were set as dynamic modifications. Additionally, The glycan database was extracted from GlycomeDB. All other settings were set at the default values. The glycosylation sequon (N-X-S/T/C, X≠P) was modified by changing ‘N’ to ‘J’. The raw data files of intact O-glycopeptides were searched against the human UniProt database using Byonic (v2.10.21, Protein Metrics, Inc.). The search parameters were as follows: the initial precursor and fragment mass tolerances were set at 10 and 20 ppm, respectively, and up to two missed cleavages were allowed for enzyme digestion. Cysteine carbamidomethylation was set as a fixed modification, methionine oxidation, asparagine deamidation, and protein N-terminal acetylation were set as dynamic modifications. 10

Additionally, the “6 most common” O-glycans on Ser/Thr were also specified as customized modifications for all searches. HCD was chosen as the fragmentation type. Three common modifications and 2 rare modifications were allowed to be included per peptide. All other settings were set at the default values, and the protein groups were filtered to 1% FDR. For O-glycopeptides, a Byonic score of 300 was considered a good score.

11

3. Results and Discussion 3.1. Characterization of MASC

Figure 1. The fabrication procedure for the preparation of MASC.

The fabrication procedure for self-assembling phenylboronic acid functionalized straticulate C3N4-based facultative hydrophilic material (MASC) is schematically illustrated in Figure 1. The as-prepared MASC was applied to enrich glycopeptides. The enrichment principle is shown in Figure 2. We reasoned that the hydrophilic property of the amino group in s-C3N4 enables the formation of a hydration shell on the surface of the materials in a water/acetonitrile (ACN) solution, resulting in an increased local concentration of glycopeptides. While the boronic acids bind glycopeptides further with a covalent bond between boronic acids and cis-diols of the glycopeptides. These two kinds of functional groups work together in the hydration shell, and may improve the enrichment efficiency of glycopeptides. When the solvent was replaced with water containing 0.1% TFA, the abundance of H+ transformed the –NH2 to –NH3+, thus disconnecting the covalent bonds between boronic acids and cis-diols, and the glycopeptides can be released[8].

Figure 2. The principle for the enrichment of glycopeptides using MASC.

Graphitic C3N4 (g-C3N4) and s-C3N4 were examined by scanning electron microscopy (SEM) (Figure S1). As shown in Figure S1-a, the g-C3N4 revealed a typical bulk. After microwave treatment in water[12], the structure of the product became thinner and the dispersion apparently better (Figure S1-b). After doping with Au and the further coating of 4-mercaptobenzene boronic acid (MPBA), the product named MASC produced self-assembly flowers (Figure 3). In the synthesis process of MASC, HAuCl4 was adsorbed on the s-C3N4 surface. Then, Au atoms were loaded on the surface of s-C3N4 through NaBH4 reducing. Finally, Au atoms formed Au-S bond with 4-mercaptobenzene boronic acid. As estimated 12

from Figure 3-a, the microstructure of MASC had good dispersity. The amplified images (Figure 3-b, 3-c) clearly show that the flower was assembled by very straticulate pieces and monodispersed. The elevation view (Figure 3-b) showed that the layers were self-assembly alone horizontal direction in the same center, and the layers self-assembled in the same dimension in the lateral view (Figure 3-c).

Figure 3. SEM images of MASC with (a) larger size, (b) elevation view and (c) lateral view.

The X-ray diffraction (XRD) pattern of the MASC shows only one peak of (002) of the g-C3N4 crystal (Figure S2-a), indicating that the flower structure product assembled by the exfoliated nanosheets has a good z-orientation compared with the bulk g-C3N4 powders. The X-ray photoelectron spectroscopy (XPS) of the fresh powder showed that the as-prepared g-C3N4 and MASC are mainly composed of C and N element (Figure S2-b). The molar ratio of N/C of the g-C3N4 was approximately 1.39, which is consistent with the literature[11]. The tiny amount of oxygen in the g-C3N4 can be ascribed to O2 adsorbed on the surface of the synthetic product during the polymerization process, which is a common phenomenon in synthetic g-C3N4 materials[13]. The Au, S, and B peaks can be observed in Figure S2-b. The C1s spectra of the s-C3N4 nanosheets and MASC are shown in Figure 4-a, which can be divided into two main peaks located at approximately 284.6 eV and 288.0 eV[14]. The peak located at 284.6 eV can be ascribed to the signal of the standard reference carbon. In comparison, the peak located at 288.4 eV was identified as the sp2-bonded carbon of s-C3N4. The peak intensity that increased at 288.4 eV for the as-prepared MASC could be attributed to the benzene of the MPBA. From Figure 4-b, the O1s peak of MASC shifted and enhanced significantly

compared

with

s-C3N4,

which

was

attributed

to

the

boronic

of

4-mercaptophenylboronic acid (MPBA). The B1s spectra of the MASC samples can be described as the superposition of two peaks by a Gaussian distribution, located at 192.1 eV and 190.8 eV, respectively (Figure 4-c)[15]. The B1s peaks at 192.1 eV and 190.8 eV are the binding energies of the B-O bond and B-C bond, respectively. The two peaks of S2p were 13

located at 165.1 eV and 163.8 eV, respectively (Figure 4-d). According to the above spectra and analysis, the Au doping and MPBA engrafting were successful. In order to explore the binding mechanism of boronic acid towards glycopeptides, the B1s spectra of MASC before and after IgG binding are compared (Figure S3-a). When boronic acid groups bind glycopeptides, the B1s peak at 192.1 eV shift to high binding energy obviously due to the forming of B-O-C covalent[16]. This is further supported by the shift of the O1s peak to low binding energy after binding glycopeptides (Figure S3-b). Combining with the above analysis, the boronic acid groups of MASC can bind glycopeptides by a covalent bond between boronic acids and cis-diols in our enrichment experiments.

Figure 4. The C1s XPS spectra (a) and the O1s XPS spectra (b) of MASC and s-C3N4; the B1s XPS spectra (c) and S2p XPS spectra (c)of MASC .

As shown in Table S1, the average zeta potential (ζ-potential) values of MASC (200 μg/mL) in 0.1% TFA 80% ACN and 0.1% TFA water solution were 19.77 and 31.77 respectively, demonstrated the good dispersibility and hydrophilicity of MASC, both in water and ACN[17]. The average ζ-potential value of MASC in 0.1% TFA water solution is apparently higher than the value in 0.1% TFA 80% ACN, illustrating that there is a positive electrostatic charge on the surface of MASC. We reasoned that the surface ionization degree of the MASC was relatively low in high ACN solution, whereas the -NH2 was turned into -NH3+ in water, which increased the positive ionization degree of the MASC surface[8]. Through the change of ionization degree, the adsorption and desorption of glycopeptides on the MASC surface were completed. The IR spectrum of s-C3N4 showed nearly identical absorption bands with those of bulk g-C3N4 (Figure S4-a). The broad peaks between 3000 and 3600 cm-1 are contributed by N-H stretching and hydrogen-bonding, and the peaks of s-C3N4 became broadened, and the absorption increases compared with bulk g-C3N4. These changes indicated that there was 14

more amino exposure on the surface of the s-C3N4, which could increase the hydrophilicity of the material. Comparing the FTIR patterns of MASC and IgG-binding MASC (Figure S4-b), the peak at 1308 cm-1 from -B(OH)2 stretching was weakened. Moreover, the broad peaks between 3000 and 3600 cm-1 showed a blueshift. All these changes indicate that boronic acid and amino groups participate in the binding of glycopeptides. In the view of the dispersibility and stability, equal amounts of s-C3N4 and MASC were dispersed into the 2.5 mL 0.1% TFA water solution and 0.1% TFA 80% ACN solution, and it can be seen clearly that no obvious sedimentation was formed after standing still for 2 hours (Figure S5). 3.2. Intact glycopeptide enrichment from standard glycoprotein by MASC The procedure for glycopeptide enrichment is illustrated in Figure S6, which contains enrichment, washing, and elution. IgG and HRP were used to optimize the enrichment conditions and evaluate the universal performance of MASC. The concentration of acetonitrile (ACN) is quite important for the binding of glycopeptides[18, 19]. Therefore, loading buffers that contained 65%, 70%, 75%, 80%, 85% or 90% ACN in H2O were applied for the glycopeptide enrichment. After carefully optimizing and comparing, ACN/H2O/TFA (80/20/0.1, v/v/v) demonstrated better result with more glycopeptides and fewer nonglycopeptides identified (Figure S7). With the optimized protocol, 32/21 glycopeptides could be detected from IgG/HRP, respectively, compared to 5/4 glycopeptides detected before enrichment (Figure 5), demonstrating the excellent performance of MASC. The enrichment repeatability of MASC on glycopeptides was investigated. The spectra of IgG (Figure S8-a) and HRP (Figure S8-b) with three replications were almost the same. The overlap of the identified glycopeptides in three replications was approximately 90% (Figure S9). Finally, up to 37 and 21 glycopeptides could be clearly detected in three replications, for IgG and HRP respectively. Detailed information about the glycopeptides of IgG and HRP could be found in Table S3 and Table S4. To explore whether MASC could be recycled, the materials were regenerated by washing with elution buffer twice and binding buffer thrice before incubated with 10 μg IgG digest. The reusability of MASC was confirmed (4 times at least) because the four repeats provided very 15

consistent spectra (Figure S10). Figure 5. MALDI-TOF spectra of 330 fmol/μL (5 ng/μL) IgG digests (a) before and (b) after enrichment; 30 fmol/μL HRP digest (c) before and (d) after enrichment. Glycopeptide peaks are labeled with ★.

To determine the glycopeptide binding capacity of MASC, different amounts of MASC materials (100, 200, 500, 1000, 2000 μg) were incubated with the same amount of IgG tryptic digests (10 μg) (Figure S11). The peak intensities of m/z 2602.6, 2764.7 and 2926.8 were used as indicators. When 1000 μg of MASC was added, the intensities of the three peaks reached their maxima. Thus, the optimal ratio of peptides to materials should be 1:100 (mass/mass), and the binding capacity was 10 μg per mg (IgG/MASC). To explore the sensitivity of MASC, different concentrations of IgG tryptic digests (33, 3.3, 0.33, 0.033 fmol/μL, 100 μL each) were incubated with the same amount of materials (100 μg) (Figure 6). A large number of glycopeptides were detected after enrichment. Even in the concentration of 0.33 fmol/μL, 10 glycopeptide peaks were identified. The result shows that MASC is suitable for the capture of low-abundance glycopeptides.

Figure 6. MALDI-TOF spectra of N-glycopeptides enriched from different concentrations of human IgG digests using MASC: (a) 33 fmol/μL; (b) 3.3 fmol/μL; (c) 0.33 fmol/μL; (d) 0.033 fmol/μL. Glycopeptides were marked with ★.

To evaluate the recovery rate of MASC for glycopeptides, the dimethyl labeling technique[20, 21] with MALDI-TOF-MS detection was used. The results showed that 95.4±1.3% and 89.2±1.3% recovery rates could be obtained for EEQYN#STYR and (EEQFN#STFR), respectively (Table S2). To evaluate the selectivity of MASC for glycopeptides from complex a matrix, different proportions of tryptic digests of human IgG and BSA (1 : 1, 1 : 10, 1 : 50, 1 : 100, 1 : 200, 16

containing 1000, 100, 20, 10, 5 ng IgG, respectively) were incubated with the same amount of MASC materials (100 μg). Before enrichment, no glycopeptides were detected at a mass ratio of 1:50 (IgG: BSA) because that the signal of the glycopeptides were suppressed by the abundant non-glycopeptides (Figure 7-a). After enrichment, a large number of glycopeptides could be detected, and the non-glycopeptides were almost entirely removed (Figure 7-b, c, d, e, f). The results demonstrated the great capability of MASC for the selective enrichment of glycopeptides from complex biological samples. To test the enrichment efficiency of MASC for glycopeptides, tryptic digests of IgG were detected by LC-MS/MS (Figure S12). Before enrichment, the ratio of glycopeptides spectra to total spectra was 16.1%. After the enrichment, the ratio was increased to 78.7%. In addition, a control experiment was conducted using commercial HILIC beads or s-C3N4 as the adsorbents for the isolation of glycopeptides under the same condition. As shown in Figure S13, 29 and 30 IgG glycopeptides were obtained using HILIC beads and s-C3N4, respectively. Both numbers were lower than that of MASC. It is worth noting that the intensity of the glycopeptides that are enriched using HILIC is about half of that enriched using MASC, which indicates that MASC could provide a higher recovery rate (Figure S13-a). Compared with s-C3N4, MASC also showed better specificity because more nonglycopeptide peaks were observed in Figure S13-b.

Figure 7. MALDI mass spectra of the mixture of tryptic IgG and BSA with the mass ratio of 1:50 before enrichment (a). MALDI mass spectra of the identified glycopeptides enriched by MASC from the tryptic mixture of human IgG and BSA with the molar ratio of (b) 1:1; (c) 1:10; (d) 1:50 ; (e) 1:100 ; (f) 1:200. Glycopeptides were marked with ★.

3.3. Glycopeptides enrichment from human serum by MASC To evaluate the performance and practicability of MASC in the enrichment of glycopeptides from complex biological samples, the N-glycoproteome in human serum was investigated.

17

Human serum is easy to collect and is an attractive clinical material. Many of the protein markers approved for the clinical testing are glycoproteins, including AFP, CA 125, CA 19-9, CA5-3, PSA, OVA1, CEA, and Her 2/neu[22]. After the tryptic digestion, enrichment and deglycosylation, the glycopeptides were analyzed by LC-MS/MS and raw MS files were processed with MaxQuant. Only glycopeptides with N-X-S/T/C (X≠P) were considered as highly reliable results[23]. A total of 463 unique glycopeptides corresponding to 209 glycoproteins from a 5 μL human serum sample were identified in two repeated experiments (detailed information is listed in Table S5). As a control, we also summarized the latest reports on the enrichment of glycopeptides from either standard proteins or biosamples (Table S6). 3.4. The gender-specific N-glycopeptides profiling of human urine Human urine is an easily and noninvasively collected body fluid arousing growing interested in biomarker discovery[24]. Urine contains proteins secreted by cells, many of which are glycoproteins, and have good correlation with pathophysiological statuses[25, 26]. Gender differences also have received attention because they can influence the outcome of all aspects of lifestyle and lead to a differential susceptibility to some diseases or a different pathogenetic mechanism. Here, we compared the composition of the urine glycopeptides (deglycosylated peptides) in males and females using our material and detected gender-based differences. The results showed that a total of 1734 unique glycopeptides corresponding to 953 glycoproteins from male and female urine samples were identified in six repeated experiments (Table S8), which provide by far the highest number of glycoproteins from human urine collection[27]. An average of 1465 glycopeptides from 839 glycoproteins and 1553 glycopeptides from 884 glycoproteins were identified from female and male urine samples in a single mass spectrometry analysis. The result shows that most of the glycopeptides (1612) were identified

in both female and male urine, while only 82 and 40 glycopeptides unique to females and males, respectively (Figure 8-a). Among them, 17 and 29 glycopeptides were detected repeatedly (6 repeats) in female urine and male urine, respectively (Table S7). Urine proteins common to both sexes are likely to reflect kidney function, whereas glycopeptides uniquely identified in the male or female urine could be used to detect gender differences and kidney 18

diseases. Furthermore, we quantitatively compared the glycopeptides in male and female urine (Figure 8-b). A total of 321 glycopeptides are differentially expressed. Among them, 90 glycopeptides are significantly downregulated and 231 are significantly upregulated in male urine. The results showed that urine proteins are differently N-glycosylated in males and females. This indicates that the gender should be considered when performing research on the discovery of biomarkers from urine glycoproteins. With the development of mass spectrometry, some quantitative strategies with different advantages and limitations were developped[28, 29]. Herein, mass spectrometry-based label-free quantitative proteomics was used. Comparing with the classical MS quantitative approaches based on the isotope labeling, label-free quantitative proteomics techniques were faster, cleaner, and simpler[30, 31]. N-glycopeptide peak intensity from electrospray ionization (ESI) which correlates with ion concentration was measured for individual LC-MS/MS runs, and changes in glycoprotein abundance were calculated via a direct comparison between male and female urine using Maxquant software[32]. It is worth noting that, glycopeptide/glycoprotein abundance differences can be infuenced by experimental variations, LC-MS/MS shifts and statistical analysis[33]. Therefore, label-free quantitative experiments need to be more carefully controlled to ensure reproducibility and reliability in large-scale glycoproteomics studies.

Figure 8. Comparison of the N-glycopeptides obtained from male and female urine (a); Volcano plot view of the differential expression of N-glycopeptides between male and female urine (b). Red symbols indicate N-glycopeptides that were significantly upregulated while green symbols indicate N-glycopeptides that were significantly downregulated in male urine samples (p<0.001). N-glycopeptides showing at least 2-fold differences were indicated.

3.5. Simultaneous enrichment of O-glycopeptides One of the challenges in glycoproteome is the enrichment and identification of O-GalNAc type glycopeptides. Further exploration of the MS raw for the N-glycopeptides identification with the Byonic[34] (v2.10.21, Protein Metrics, Inc.) software we identified nearly a thousand 19

O-glycopeptides from a single run of the MASC enriched fraction of the human urine. These promising results indicated that our material can enrich N- and O-glycopeptides simultaneously. However, it should be noted that the HCD mode can’t assign the O-glycosylation site location accurately. This is certainly another direction to work in the next step. 4. Conclusion In summary, the MASC facultative material was synthesized from melamine through a simple, fast and economical strategy using microwave peeling. The as-prepared material bears both enhanced boronic acid functionality and hydrophilic interaction capability between s-C3N4 and glycopeptides. The synthetic MASC demonstrated low pH value adaptation, excellent hydrophilicity and stability, high repeatability and recyclability, good selectivity (1:100), a low limit of detection (0.33 fmol/μL), high enrichment efficiency (~80%) and a high recovery rate (~90%) towards glycopeptide enrichment in standard samples. The material demonstrated impressive performance on both N- and O-glycopeptides from complex biological samples, such as human serum and urine. It could capture more glycopeptides from human serum and urine than other boronic acid materials or reported hydrophilic materials (Table S6). A large number of uniquely or differentially expressed glycopeptides were found between male and female urine using this material. We expect that this material will be widely applicable for N/O-glycopeptides enrichment and for discovering novel disease-specific glycoprotein/ glycopeptide biomarkers.

Acknowledgment We are grateful for financial support from the National Key Program for Basic Research of China (2017YFC0906600, 2016YFA0501300, 2014CBA02001), the National Natural Science Foundation of China (81530021). Beijing Municipal Science and Technology Project (Z161100002616036). The Innovation Foundation of Medicine (BWS14J052, 16CXZ027). Appendix A. Supplementary material

References 20

[1] C.-C. Chen, W.-C. Su, B.-Y. Huang, Y.-J. Chen, H.-C. Tai, R.P. Obena, Interaction modes and approaches to glycopeptide and glycoprotein enrichment, Analyst, 139 (2014) 688-704. [2] P.E. Geyer, N.A. Kulak, G. Pichler, L.M. Holdt, D. Teupser, M. Mann, Plasma proteome profiling to assess human health and disease, Cell systems, 2 (2016) 185-195. [3] G. Palmisano, S.E. Lendal, K. Engholm-Keller, R. Leth-Larsen, B.L. Parker, M.R. Larsen, Selective enrichment of sialic acid-containing glycopeptides using titanium dioxide chromatography with analysis by HILIC and mass spectrometry, Nature protocols, 5 (2010) 1974-1982. [4] B.L. Parker, P. Gupta, S.J. Cordwell, M.R. Larsen, G. Palmisano, Purification and identification of O-Glc N Ac-modified peptides using phosphate-based alkyne CLICK chemistry in combination with titanium dioxide chromatography and mass spectrometry, Journal of proteome research, 10 (2011) 1449-1458. [5] M. Dowlut, D.G. Hall, An improved class of sugar-binding boronic acids, soluble and capable of complexing glycosides in neutral water, Journal of the American Chemical Society, 128 (2006) 4226-4227. [6] S. Jin, Y. Cheng, S. Reid, M. Li, B. Wang, Carbohydrate recognition by boronolectins, small molecules, and lectins, Medicinal research reviews, 30 (2010) 171-257. [7] X. Wang, N. Xia, L. Liu, Boronic Acid-Based Approach for Separation and Immobilization of Glycoproteins and Its Application in Sensing, International Journal of Molecular Sciences, 14 (2013) 20890-20912. [8] L. Ren, Z. Liu, Y. Liu, P. Dou, H.-Y. Chen, Ring-Opening Polymerization with Synergistic Co-monomers: Access to a Boronate-Functionalized Polymeric Monolith for the Specific Capture ofcis-Diol-Containing Biomolecules under Neutral Conditions, Angewandte Chemie International Edition, 48 (2009) 6704-6707. [9] F.L.r. R. Tuytten, E. L. Esmans, W. A.Herrebout, B. J. van der Veken, B. U.W. Maes, E. Witters, R. P., E.D. Newton, Role of Nitrogen Lewis Basicity in Boronate Affinity Chromatography of Nucleosides, Analytical chemistry, 79 (2007) 6662. [10] F.K. Kessler, Y. Zheng, D. Schwarz, C. Merschjann, W. Schnick, X. Wang, M.J. Bojdys, Functional carbon nitride materials — design strategies for electrochemical devices, Nature Reviews Materials, 2 (2017) 17030. [11] X. Zhang, X. Xie, H. Wang, J. Zhang, B. Pan, Y. Xie, Enhanced Photoresponsive Ultrathin Graphitic-Phase C3N4 Nanosheets for Bioimaging, Journal of the American Chemical Society, 135 (2012) 18-21. [12] M. Matsumoto, Y. Saito, C. Park, T. Fukushima, T. Aida, Ultrahigh-throughput exfoliation of graphite into pristine ‘single-layer’graphene using microwaves and molecularly engineered ionic liquids, Nature chemistry, 7 (2015) 730. [13] X. Zhang, H. Wang, H. Wang, Q. Zhang, J. Xie, Y. Tian, J. Wang, Y. Xie, Single-Layered Graphitic-C3N4Quantum Dots for Two-Photon Fluorescence Imaging of Cellular Nucleus, Advanced Materials, 26 (2014) 4438-4443. [14] P. Srinivasu, A. Islam, S.P. Singh, L. Han, M.L. Kantam, S.K. Bhargava, Highly efficient nanoporous graphitic carbon with tunable textural properties for dye-sensitized solar cells, Journal of Materials Chemistry, 22 (2012) 20866. [15] L.E. Ennis, A.P. Hitchcock, Inner shell excitation spectroscopy of transient molecules: HBS, HBO, and H3B3O3, The Journal of Chemical Physics, 111 (1999) 3468-3478. [16] H. Konno, T. Ito, M. Ushiro, K. Fushimi, K. Azumi, High capacitance B/C/N composites for capacitor electrodes synthesized by a simple method, Journal of Power Sources, 195 (2010) 1739-1746. [17] M.R. Larsen, S.S. Jensen, L.A. Jakobsen, N.H. Heegaard, Exploring the sialiome using titanium dioxide chromatography and mass spectrometry, Molecular & cellular proteomics, 6 (2007) 1778-1787. [18] J. Wohlgemuth, M. Karas, T. Eichhorn, R. Hendriks, S. Andrecht, Quantitative site-specific analysis of protein glycosylation by LC-MS using different glycopeptide-enrichment strategies, Analytical biochemistry, 395 (2009) 178-188. [19] G. Xu, W. Zhang, L. Wei, H. Lu, P. Yang, Boronic acid-functionalized detonation nanodiamond for specific enrichment of glycopeptides in glycoproteome analysis, Analyst, 138 (2013) 1876-1885. [20] Z. Xiong, H. Qin, H. Wan, G. Huang, Z. Zhang, J. Dong, L. Zhang, W. Zhang, H. Zou, Layer-by-layer assembly of multilayer polysaccharide coated magnetic nanoparticles for the selective enrichment of glycopeptides, Chemical Communications, 49 (2013) 9284-9286. [21] B. Jiang, Y. Liang, Q. Wu, H. Jiang, K. Yang, L. Zhang, Z. Liang, X. Peng, Y. Zhang, New GO–PEI–Au– 21

L-Cys ZIC-HILIC composites: synthesis and selective enrichment of glycopeptides, Nanoscale, 6 (2014) 5616-5619. [22] J.A. Ludwig, J.N. Weinstein, Biomarkers in Cancer Staging, Prognosis and Treatment Selection, Nature reviews. Cancer, 5 (2005) 845-856. [23] M.N. Melo-Braga, T. Verano-Braga, I.R. León, D. Antonacci, F.C. Nogueira, J.J. Thelen, M.R. Larsen, G. Palmisano, Modulation of protein phosphorylation, N-glycosylation and Lys-acetylation in grape (Vitis vinifera) mesocarp and exocarp owing to Lobesia botrana infection, Molecular & Cellular Proteomics, 11 (2012) 945-956. [24] N. Nagaraj, M. Mann, Quantitative analysis of the intra-and inter-individual variability of the normal urinary proteome, Journal of proteome research, 10 (2011) 637-645. [25] S. Kalantari, A. Jafari, R. Moradpoor, E. Ghasemi, E. Khalkhal, Human Urine Proteomics: Analytical Techniques and Clinical Applications in Renal Diseases, International journal of proteomics, 2015 (2015) 782798. [26] U. Neisius, T. Koeck, H. Mischak, S.H. Rossi, E. Olson, D.M. Carty, J.A. Dymott, A.F. Dominiczak, C. Berry, K.G. Oldroyd, Urine proteomics in the diagnosis of stable angina, BMC cardiovascular disorders, 16 (2016) 70. [27] R. Kawahara, J. Saad, C.B. Angeli, G. Palmisano, Site-specific characterization of N-linked glycosylation in human urinary glycoproteins and endogenous glycopeptides, Glycoconjugate journal, 33 (2016) 937-951. [28] B. Ivanova, M. Spiteller, Quantitative correlations between collision induced dissociation mass spectrometry coupled with electrospray ionization or atmospheric pressure chemical ionization mass spectrometry–Experiment and theory, Journal of Molecular Structure, 1157 (2018) 492-512. [29] B. Ivanova, M. Spiteller, Quantification by Matrix-Assisted Laser Desorption Ionization Mass Spectrometry Using An Approach Based On Stochastic Dynamics. Experimental And Theoretical Correspondences, GRIN Verlag2018. [30] W. Zhu, J.W. Smith, C.-M. Huang, Mass spectrometry-based label-free quantitative proteomics, BioMed Research International, 2010 (2009). [31] V.J. Patel, K. Thalassinos, S.E. Slade, J.B. Connolly, A. Crombie, J.C. Murrell, J.H. Scrivens, A comparison of labeling and label-free mass spectrometry-based proteomics approaches, Journal of proteome research, 8 (2009) 3752-3759. [32] R.D. Voyksner, H. Lee, Investigating the use of an octupole ion guide for ion storage and high‐ pass mass filtering to improve the quantitative performance of electrospray ion trap mass spectrometry, Rapid Communications in Mass Spectrometry, 13 (1999) 1427-1437. [33] M.C. Wiener, J.R. Sachs, E.G. Deyanova, N.A. Yates, Differential mass spectrometry: A label-free LC− MS method for finding significant differences in complex peptide and protein mixtures, Analytical chemistry, 76 (2004) 6085-6096. [34] M. Bern, Y.J. Kil, C. Becker, Byonic: advanced peptide and protein identification software, Current Protocols in Bioinformatics, DOI (2012) 13.20. 11-13.20. 14.

22

23

24

25

26

27

Highlights  A facile and economical procedure to generate phenylboronic acid functionalized C3N4 facultative hydrophilic materials (MPBA-Au@s-C3N4, MASC) was developed.



The new materials possess the properties of low pH value adaptation, high hydrophilicity and stability.



In the enrichment of glycopeptides, the materials show high selectivity (1:100), low limit of detection (0.33 fmol/μL), high enrichment efficiency (~80%) and high recovery rate (~90%).



Compared to previous studies, the materials captured more glycopeptides from human urine and serum.

28