- Email: [email protected]
Hydrogen-bond interaction assisted branched copolymer HILIC material for separation and N-glycopeptides enrichment
Talanta 158 (2016) 361–367
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
Hydrogen-bond interaction assisted branched copolymer HILIC material for separation and N-glycopeptides enrichment Wenya Shao a,b, Jianxi Liu a,b, Kaiguang Yang a, Yu Liang a, Yejing Weng a,b, Senwu Li a,b, Zhen Liang a, Lihua Zhang a,n, Yukui Zhang a a National Chromatographic R. & A. Center, Key Laboratory of Separation Science for Analytical Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China b University of Chinese Academy of Sciences, Beijing 100049, China
art ic l e i nf o
a b s t r a c t
Article history: Received 31 January 2016 Received in revised form 9 May 2016 Accepted 13 May 2016 Available online 24 May 2016
Hydrophilic interaction chromatography (HILIC) has attracted increasing attention in recent years due to its efficient application in the separation of polar compounds and the enrichment of glycopeptides. However, HILIC materials are still of weak hydrophilicity and thereby present weak retention and selectivity. In this work, branched copolymer modified hydrophilic material Sil@Poly(THMA-co-MBAAm), with high hydrophilicity and unique “claw-like” polyhydric groups, were prepared by “grafting from” thiol-ene click reaction. Due to the abundant functional groups provided by branched copolymer, the material showed excellent retention for nucleosides, necleobases, acidic compounds, sugars and peptides. Furthermore, Sil@Poly(THMA-co-MBAAm) was also applied for the N-glycosylation sites profiling towards the digests of the mouse brain, and 1997N-glycosylated peptides were identified, corresponding to 686 glycoprotein groups. Due to the assisted hydrogen-bond interaction, the selectivity for glycopeptide enrichment in the real sample reached 94.6%, which was the highest as far as we know. All these results indicated that such hydrogen-bond interaction assisted branched copolymer HILIC material possessed great potential for the separation and large scale glycoproteomics analysis. & 2016 Published by Elsevier B.V.
Keywords: Hydrophilic interaction chromatography Glycopeptide Branched copolymer Grafting from Hydrogen-Bond Interaction
1. Introduction N-glycosylation plays crucial roles in intercellular recognition and communication, protein folding and immune responses, signal transduction, and so forth [1–3]. The analysis based on mass spectrometry (MS) is an efficiency tool for large scale and in-depth glycoprotein/glycopeptides profiling in complex biological samples [4,5]. However, the great challenges are the inherently low abundance and poor ionization efficiency of glycopeptides. Therefore, the selective enrichment of glycopeptides is indispensable for MS identification. A variety of enrichment methods such as lectin affinity [6], hydrazide chemistry [7], boronic acid [8–10] and Hydrophilic interaction chromatography (HILIC) [11] have been developed, among which HILIC has been widely used, with advantage of good compatibility with LC-MS analysis[12] and low bias to different types of glycopeptides. Besides, with the increasing popularity of pharmaceutical analysis, metabolomics and proteomics, HILIC, as a valuable method for the separation of polar substances, has attracted great attention. n
Corresponding author. E-mail address: [email protected] (L. Zhang).
http://dx.doi.org/10.1016/j.talanta.2016.05.034 0039-9140/& 2016 Published by Elsevier B.V.
As the increasing attention, HILIC materials, such as –NH2, – CONH, –CN, and –CH(OH)CH2OH modified particles [13], hydrophilic monoliths [14], have been developed in recent years. However, their hydrophilicity and selectivity limited the application of HILIC in the complex sample; weak interaction and low selectivity are apparent drawback for capturing the glycopeptides. Saccharide-modified silica stationary phase, as a rising star among the HILIC materials, has been successfully applied for the separation of polar compounds [15] and the enrichment of glycopeptides [16,17]. However, such materials were suffered from two limitations: 1) poor hydrophilicity, which would result in high percentage of organic solvent in initial mobile phase, is incompatible with biomolecules especially protein and peptide; 2) low stability of ring glycan structure. Compared with natural saccharides functional groups, N-[Tris(hydroxymethyl)methyl]acrylamide (THMA) with three terminal symmetric hydroxyl would not only possess preferable structure properties but also improve the stability of the glycan structure. In addition, hydrophilicity modification can be achieved by the “grafting from” polymerization. Compared with other polymerization methods, it allows polymer chains to grow from the initiators on the substrate to generate high grafting densities. It should be mentioned that various approaches can be employed for
362
W. Shao et al. / Talanta 158 (2016) 361–367
“grafting from”, e.g. surface-initiated ionic polymerization [18,19], living/controlled free polymerizations such as atom transfer radical polymerization (ATRP) [20,21], reversible-addition fragmentation chain transfer (RAFT) polymerization [22,23] or click chemistry [24,25]. Among them, thiol-ene click chemistry has the characteristics of regiospecificity, stereo-specificity and amenability to variety of readily available starting compounds with enes [26–28]. These characteristics inspired us that “grafting from” might be a solution to the problems on saccharide-modified silica stationary phase. Herein, in this work, a novel hydrophilic stationary phase, branched copolymer modified hydrophilic material Sil@Poly (THMA-co-MBAAm) were fabricated via one-pot thiol-ene click chemistry through “grafting-from” approach with N-[Tris(hydroxymethyl)methyl]acrylamide (THMA) and N,N′-methylenebisacrylamide (MBAAm) as the monomers [29]. Finally, this branched copolymer modified material Sil@Poly(THMA-co-MBAAm) was successfully applied in the separation of polar compounds, sugars, peptides and the analysis of glycosylation sites.
2. Experimental 2.1. Reagents and materials Porous silica particles (diameter 5 mm, pore size 150 Å) were purchased from Fuji Silysia Chemical (Aichi, Japan). Bovine serum albumin (BSA, bovine serum) was obtained from Sino-American Biotec (Luoyang, China). PNGase F was bought from New England Biolabs (Ipswich, MA, US). Dithiothreitol (DTT) and iodoacetamide (IAA) were from Acros (Morris Plains, NJ, US). Acetonitrile (ACN, HPLC grade) was ordered from Merck (Darmstadt, Germany). Water was purified by a Milli-Q system (Millipore, Milford, MA, US). N-[Tris(hydroxymethyl)methyl]acrylamide (THMA, 93%), (3mercaptopropyl)trimethoxysilane (MPS, 95%), N,N′-methylenebisacrylamide (MBAAm, 98%), thymidine, uridine, cytidine, guanosine, cytosine, phenol, benzoic acid, 4-hydroxybenzoic acid, 3,5dinitrobenzoic acid, 3-nitrobenzoic acid, ammonium acetate, acetic acid, trypsin (bovine pancreas), IgG (human serum), and trifluoroacetic acid (TFA) were purchased from Sigma-Aldrich (St. Louis, MO). Venusil XPB C18 particles (5 mm, 100 Å) were ordered from Daiso (Osaka, Japan). 2.2. Synthesis of Sil@Poly(THMA-co-MBAAm) The material was prepared via thiol-ene click copolymerization by the following two steps. At first, the MPS modified silica (SilMPS) was obtained based on the previous report [30]. Silica particles (5 g) were suspended in 100 mL of anhydrous toluene solution, then 1.7 mL of (3-mercaptopropyl)trimethoxysilane was added, followed by refluxing and mechanical agitation for 24 h at 110 °C. After reaction, the Sil-MPS was washed with dry toluene, methanol, acetone, and dried under vacuum overnight. Second, the copolymer grafted particles, Sil@Poly(THMA-co-MBAAm), were prepared as follows: THMA (1.16 g), MBAAm (0.97 g) and AIBN (0.1 g) were added to the suspension of Sil-MPS (3.5 g) in water-methanol (2:3, v/v, 145 mL) and the reaction mixture was bubbled with nitrogen for 15 min to remove oxygen, followed by mechanical agitation at 70 °C for 24 h. After the reaction, excess reagents and oligomers were removed by repeatedly washing with water and ethanol in succession. The obtained Sil@Poly(THMA-coMBAAm) was dried under vacuum at 40 °C for 12 h and stored at room temperature for further usage.
2.3. Characterization Thermogravimetric analysis (TGA) was carried out on a Cahn Thermax 500 instrument (Thermo Scientific, US) from 30 to 980 °C with a heating rate of 10 °C min 1 under flow nitrogen. Fourier transform infrared spectra (FT-IR) were performed on a Bruker Optics HYPERION 3000 (Bruker, Daltonios, Germany) in the reflection mode, and scanned at the range of 500–4500 cm 1. The specific surface area and mesopore size were determined by nitrogen adsorption-desorption measurements (QuadraSorb SI4, Florida, US), while the surface area and the mesopore size were determined from the isotherms by BET method and BJH method, respectively. 2.4. Column packing and chromatographic conditions A slurry of Sil@Poly(THMA-co-MBAAm) in the methanol was prepared by ultrasonication for 2 min, then it was packed into the stainless-steel column (150 mm 4.6 mm i.d.) at a constant pressure of 40 MPa using methanol as the flushing solvent. The HILIC column (10 mm 4.6 mm i.d.) used for glycopeptide enrichment was prepared as well. Separations were performed by the HPLC with a diode array detector (Chromaster HPLC, Hitachi High-Tech, TKY, JP) on 150 mm 4.6 mm i.d. The flow rate was 1.00 mL min 1 with the UV detection at 254 nm, and the column temperature was set at 25 °C. The mobile phase consisted of an aqueous phase with ammonium acetate (NH4Ac) and an organic phase ACN. 2.5. Sample preparation All the small molecules and peptides used for the separation were dissolved in ACN/H2O (85/15, V/V). The concentrations of standard solutions were 60 ng/mL for nucleosides and necleobases separately, 10–50 ng/mL for acidic solutions. IgG and BSA digests were separately prepared according to the following protocol. Proteins were dissolved in denaturing buffer (6 M guanidine hydrochloride), then reduced in 100 mM DTT at 56 °C for 1.5 h, alkylated by 200 mM IAA for 40 min at room temperature in the dark. The solution was diluted ten-fold with 50 mM NH4HCO3 buffer (pH 8.0) and digested with trypsin (enzyme/protein ratio of 1:30, w/w) at 37 °C for 12 h. The tryptic digested peptides were desalted by home-made C18 precolumn, lyophilized and stored at 20 °C before usage. HeLa cells were cultured in a humidified 37 °C incubator with DMEM medium containing 10% fetal bovine serum (FBS) and 5% CO2. Cells were harvested when nearly at 90% confluence and washed with 4 °C phosphate-buffered saline (PBS) for 3 times. Cell pellets were collected and suspended in 6 M guanidine hydrochloride (1% (v/v) protease inhibitor cocktail), ultrasonicated on ice for 120 s in total (10 s intervals every 10 s). Then, the samples were centrifuged at 20,000 rpm at 4 °C for 20 min The resulting supernatants containing the total Hela cell proteins were collected and the protein concentration was determined by the BCA assay. The digestion of Hela cells extracts was the same as that of human IgG. 2.6. Glycopeptides enrichment The enrichment procedure followed the previous protocol [17] with minor modification. The tryptic digested IgG was firstly redissolved in ACN/H2O/TFA (80/20/0.1, v/v/v) (buffer A) and loaded onto the HILIC column (10 mm 4.6 mm i.d.). Then, the column was rinsed with buffer A at 1.0 mL min 1 for 10 min to remove nonglycopeptides. Glycopeptides were eluted with ACN/H2O/TFA (50/50/0.1, v/v/v), lyophilized and redissolved in 20 mM NH4HCO3 buffer (pH 8.0). Subsequently, the glycans were released by adding
W. Shao et al. / Talanta 158 (2016) 361–367
363
PNGase F and incubated at 37 °C for 12 h. Finally, the peptides were desalted and dried for further usage. The tryptic digests from BSA and IgG mixture and Hela cells were treated in the same way. 2.7. Mass spectrometry and data processing The glycopeptides from IgG were analyzed by MALDI-TOF-MS. A 1 mL of sample and 1 mL of DHB matrix solution (10 mg/mL, 0.1% TFA in 60% ACN aqueous solution) were orderly dropped onto a MALDI plate for MS analysis. All experiments were performed on Bruker Ultraflex III MALDI-TOF/TOF MS instrument (Bruker, Daltonios, Germany). The deglycosylated peptides enriched from HeLa cell lysates were analyzed by nanoRPLC-MS/MS on a Q Exactive MS (Thermo Fisher Scientific, USA). The lyophilized sample was dissolved in 70 mL of 0.1% FA solution, and 12 mL of that was loaded onto a homemade capillary separation column (75 mm i.d. 14 cm). Mobile phase A (0.1% FA in H2O) and B (0.1% FA in ACN) were used to establish the 92 min gradient, which went from 2% B to 7% B over 2 min, to 22% B over 60 min, to 35% B over 20 min, and then to 80% B over 10 min, with the flow rate at 300 nL/min. The Q-Exactive was operated in positive ion data dependent mode with 2.5 kV spray voltage and the temperature of ion transfer capillary was set at 275 °C. One full scan MS acquired from m/z 300–1800 was followed by fifteen data dependent MS/MS events. MS1 was performed at the resolution of 70,000, with automatic gain control (AGC) value (1e6), maximum injection time (100 ms). MS2 was performed at the resolution of 17,500 (AGC: 1e5, maximum injection time: 50 ms). The *.raw files produced on Q-Exactive MS were searched against the Swiss-Prot mouse complete proteome sequence database (release 2014_03). The parameters were as follows: enzyme, trypsin; missed cleavages, two; fixed modifications, carboxyamidomethylation (C); variable modifications, oxidation (M) and deamidation (N); peptide tolerance, 7 ppm; MS/MS tolerance, 20 mmu; target FDR (Stric), 0.01; target FDR (Relaxed), 0.05. The N-glycosylation sites were identified if the peptides contained the sequence of N-!P-S/T and had the modification of deamination at N at the same time.
3. Result and discussion 3.1. Preparation and characterization of the Sil@Poly(THMA-coMBAAm) material The Sil@Poly(THMA-co-MBAAm) stationary phase was prepared by “thiol-ene” click copolymerization as shown in Schem. 1. Briefly, MPS was immobilized on the silica surface via Si–O bond. Then, the branched copolymer of silica surface was formed by “thiol-ene”click reaction and “graft from” polymerization (Fig. S1). In this work, MBAAm and THMA were chosen as the co-monomers. Due to the branched polymer grafted surface, this particle possessed the following characteristics (Schem. 1): 1) Hydrophilicity of the material could be enhanced by the abundant of availability amide functional groups on the branched copolymer; 3) The extended “claw-like” polyhydroxy structure would provide stronger and more specific affinity through hydrogen bonding interaction toward glycopeptides. The characterization of the surface morphology of Sil@Poly (THMA-co-MBAAm) was observed by SEM and compared with the raw silica (Fig. S2). The Sil@Poly(THMA-co-MBAAm) particles are well dispersed, smooth and sphere in shape, which are important for the separation. The amount of the grafted copolymer was determined by thermogravimetric analysis (TGA). As shown in Fig. 1, only 1% weight of bare silica particles in the region below 200 °C,
Scheme 1. Synthesis and structure of Sil@Poly(THMA-co-MBAAm).
Fig. 1. TGA analysis of (a) bare silica, (b) Sil-MPS and (c) Sil@Poly(THMA-coMBAAm) particles.
which can be attributed to the loss of physically adsorbed water. Approximately 3% weight loss is observed for Sil-MPS during 200– 700 °C. In contrast, 20% weight loss of Sil@Poly(THMA-co-MBAAm) is found. Therefore, the percentage of graft reaction approximately equals to 17%. It is indicated that high density functional groups were immobilized on the surface of silica particles. The specific surface area and pore size were two most important parameters that determined the loading capacity and diffusion rate of analytes for chromatography stationary phase. The morphologies of the SilPoly(MBAAm-co-AHPS) were further characterized by nitrogen adsorption–desorption analysis. As shown in Fig. 2, for bare silica particles, the pore size is 24 nm by BJH method. After the graft of copolymer, the pore size is reduced to 17 nm with mean pore-size distribution. Deductively, the copolymers were uniformly grown on the surface of the silica particles, due to the high-efficiency thiol-ene reaction. In contrast, the BET specific surface area of Sil@Poly(THMA-co-MBAAm) slightly changed from 157 to 153 m2/ g without obvious decrease, since the stretched copolymers would increase the particle surface. Next, FT-IR was also applied to characterize the stationary phase. As shown in Fig. 3, in the spectrum of Sil@Poly(THMA-coMBAAm), the peak at 1382 cm 1 is the characteristic vibration of – OH in –CH2– of THMA. The peak at 1650 cm 1 can be attributed to C ¼O stretching vibration and the peak at 1531 cm 1 can be assigned to N–H bending vibration bowing to amide group of
364
W. Shao et al. / Talanta 158 (2016) 361–367
Fig. 2. BJH adsorption pore size distribution of (a) bare silica and (b) Sil@Poly(THMA-co-MBAAm) particles.
the silica particles. 3.2. Chromatographic separation performance
Fig. 3. FT-IR spectra of (a) bare silica and (b) Sil@Poly(THMA-co-MBAAm) particles.
copolymer. Bands at 2956 cm 1 and 2882 cm 1 both were assigned to aliphatic C–H stretching. In contrast, there is no corresponding peak can be found in the bare silica particles. These results also demonstrated the successful grafting on the surface of
The chromatographic retention behavior of Sil@Poly(THMA-coMBAAm) for polar compounds was evaluated using thymidine, uridine, cytidine, guanosine and cytosine, which are typically strong polar compounds which have similar skeletal structure. As shown in Fig. 4(a), baseline separation of the five compounds is achieved. Meanwhile, all the retention time is between 3 min to 15 min, indicating the appropriate retention of Sil@Poly(THMA-coMBAAm) packed column. A comparison of Sil@Poly(THMA-coMBAAm) column with the other two HILIC columns (amide and silica) in the separation has been shown in Fig. 4(a). Compared with the other commonly used HILIC stationary phases, 1) Sil@Poly (THMA-co-MBAAm) column was the best in separating necleobases and nucleosides. 2) The retention factors were usually 1.5– 4.5 and 1.5–8.5 respectively for cytosine and cytidine [31], while the Sil@Poly(THMA-co-MBAAm) column exhibited much better retention factors (6.5 and 9.1). To further demonstrate the applicability of the Sil@Poly(THMA-co-MBAAm) stationary phase, the separation of acid compounds (the pKa values listed in Fig. S3) were performed in HILIC mode. As shown in Fig. 4(b), phenol, benzoic acid, 4-hydroxybenzoic acid, 3,5-dinitrobenzoic acid and p-nitrobenzoic acid are well separated within 10 min on the
Fig. 4. (a) A comparison of separation behaviors on Sil@Poly(THMA-co-MBAAm) HILIC column and two different HILIC column. Mobile phase: ACN/H2O (85/15, v/v) containing 15 mM ammonium acetate. Peaks: (1) thymidine, (2) uridine, (3) cytosine, (4) cytidine, (5) guanosine. (b) Separation of acid compounds on Sil@Poly(THMA-coMBAAm) HILIC column Mobile phase: ACN/H2O (86/14, v/v) containing 45 mM ammonium acetate, Peaks: (1) phenol, (2) 3-nitrobenzoic acid, (3) 3,5-dinitrobenzoic acid, (4) benzoic acid, (5) 4-hydroxybenzoic acid. Separation on Sil@Poly(THMA-co-MBAAm) packed column (150 mm 4.6 mm i.d.). Separation condition: T ¼25 °C; flow rate ¼ 1 mL min 1; UV detection, λ ¼254 nm.
W. Shao et al. / Talanta 158 (2016) 361–367
365
Fig. 5. (a) Separation of seven sugars on Sil@Poly(THMA-co-MBAAm) HILIC column. Mobile phase: ACN/H2O (80/20, v/v). Evaporative light scattering detector (ELSD): gas pressure 30 psi, tube temperature 70 °C, gain 10. Peak: (1) methy α-D-mannopyranoside, (2) ribose, (3) mannitol, (4) xylose, (5) N-acetylglucosamine, (6) fructose, (7) maltitol. (b) Separation of four peptides on Sil@Poly(THMA-co-MBAAm) HILIC column Mobile phase: A, H2O; B, ACN; C, 50 mM NaH2PO4; 0–15 min, 15% A, 75% B, 10% C 40% A, 50% B, 10% C;UV detection: 214 nm; Peak: (1) angiotensin Ⅳ, (2) bradykinin, (3) leucine enkephaline, (4) angiotensin Ⅱ.
Fig. 6. (a) Effect of salt concentration on the separation. Mobile phases of ACN/H2O (85/15, v/v) contained different concentration of salt. T ¼25 °C; flow rate ¼1 mL min 1; UV detection, λ¼ 254 nm. Peak: (1) 3-nitrobenzoic acid, (2) thymidine, (3) 4-hydroxybenzoic acid, (4) cytosine. (b) Effect of column temperature on the separation. Mobile phase: ACN/H2O (85/15, v/v) containing 20 mM ammonium acetate; flow rate ¼1 mL min 1. Peak: (1) 3-nitrobenzoic acid, (2) thymidine, (3) 4-hydroxybenzoic acid, (4) cytosine.
Fig. 7. The stability of Sil@Poly(THMA-co-MBAAm) HILIC column. Mobile phase: ACN/H2O (85/15, v/v) containing 15 mM ammonium acetate. Peaks: (1) thymidine, (2) uridine, (3) cytosine, (4) cytidine, (5) guanosine. Other chromatographic conditions were the same as Fig. 4.
Sil@Poly(THMA-co-MBAAm) column. It should be noted that the peaks were sharp and symmetrical. Considering the excellent chromatographic performance of the Sil@Poly(THMA-co-MBAAm) stationary phase, the separation of peptides and sugars were also carried out on this material in HILIC mode. Seven sugars including mono- and disaccharides were mixed as the model compounds. The sugars were efficiently separated with good peak shape with 80% acetonitrile (Fig. 5(a)). Meanwhile, four peptides were well separated with proper retention (Fig. 5(b)). High efficiency separation of nucleosides, necleobases, acidic compounds, sugars and peptides proved that excellent hydrophilicity is achieved through this hydrogen-bond interaction assisted branched copolymer HILIC material. The effects of ACN content, buffer concentration and column temperature were also investigated to reveal the chromatographic characterization. Three polar compounds (thymidine, cytosine and 3-nitrobenzoic acid) were used to evaluated the effects of the percentage of ACN in the mobile phase. As shown in the Fig. S4, retention time of analytes increased with the percentage of ACN in the mobile phase from 65% to 90%, which was consistent with the typical characteristics of HILIC retention. Buffer concentration
366
W. Shao et al. / Talanta 158 (2016) 361–367
Fig. 8. MALDI-TOF mass spectra of enrichment with the Sil@Poly(THMA-co-MBAAm) packed HILIC column (a) IgG digests, (b) the mixture of human IgG and BSA digest (with a mass ratio of 1:10), and (c) the mixture of human IgG and BSA digest (with a mass ratio of 1:100).
could also affect the separation property on HILIC. Retention behavior was investigated by varying the ammonium acetate buffer concentration in the mobile phase from 10 mM to 40 mM by thymidine, cytosine, 4-hydroxybenzoic acid and 3-nitrobenzoic acid. As shown in Fig. 6(a), the retention time of 4-hydroxybenzoic acid and 3-nitrobenzoic acid increased as the buffer concentration increased, while thymidine and cytosine changed little with the increase of buffer concentrations. This result can be ascribed to the electrostatic interactions between the acids and the stationary phase surface. With the increase of the buffer concentration, the electrostatic repulsion interaction was weakened and the retention was enhanced. The effect of the column temperature on separation was additionally investigated. The retention on this Sil@Poly(THMA-co-MBAAm) HILIC column decreases at elevated temperatures (Fig. 6(b)), presumably due to lower viscosity and higher diffusivity in high column temperature. It should be noted that the bandwidths of the peaks of analytes improved as column temperature increased. In conclusion, the percentage of ACN in the mobile phase is the most important factors for the sample retention among the three factors. Stability of the stationary phase was examined using model compounds thymidine, uridine, cytosine, cytidine and guanosine. Fig. 7 shows retention chromatograms for three times injections, among which (a) and (b) were two continuous injections. After working for 72 h in 9 months, this column was repeatedly applied for separation (c). There were no obvious changes with these analytes, indicating that this Sil@Poly(THMA-co-MBAAm) HILIC column was stable. 3.3. Application in the glycopeptide enrichment Based on the hydrophilicity and unique claw-like polyhydroxy of the prepared stationary phase, the Sil@Poly(THMA-co-MBAAm) materials was applied to enrich glycopeptides. IgG, a glycoprotein which contains one N-glycosylated site in each heavy chain (IgG
1 and IgG 2) at the highly conserved asparagine 297 residue in each of the CH2 domains of the Fc region, was digested and utilized as the model to evaluate the performance of the prepared Sil@Poly (THMA-co-MBAAm) column (10 mm 4.6 mm i.d.) [32]. In the MALDI-TOF MS spectrum of the native digest (Fig. S5), the glycopeptides signal was interfered and suppressed by amount of nonglycopeptides. After enrichment with the Sil@Poly(THMA-coMBAAm) column (Fig. 8(a)), 7 clusters of glycopeptides peaks were emerged and the corresponding glycan structure of 17 glycopeptide was determined according to the previous reports [33,34], while the signals of non-glycopeptides disappeared. After treatment with PNGase F with 12 h, the N-linked glycans were cleaved, leaving two deglycosylated peptides at m/z 1158.5 and 1190.5 representing of EEQFN#STFR and EEQYN#STYR (in which N# denote the N-linked glycosylation site) (Fig. S6). The digest mixtures of non-glycoprotein BSA and IgG at a mass ratio of 1:10 and 1:100 were further used to evaluate the enrichment selectivity of this column. As shown in Fig. 8(b) and (c), 14 and 11 glycopeptides were still identified with high S/N, and non-glycopeptides were completely removed. This demonstrated that the glycopeptides could be enriched with high specificity and selectivity. 3.4. Application in N-glycoproteome analyses of mouse brain To further evaluate the enrichment capability of Sil@Poly (THMA-co-MBAAm), the glycopeptides from trypic digests of the proteins extracted from mouse brain were enriched by the materials and then treated with PNGase F, followed by further nanoRPLC-ESI-MS/MS analysis. In total, 2111 unique peptides and 743 nonredundant protein groups were identified, of which 1997 unique N-glycopeptides mapped to 686 N-linked glycoprotein groups (Table S1) were obtained, much more than the results from the previous publications [35–37] based on different glycopeptides enrichment materials, all of which covered less than 1000 unique N-glycopeptides from mouse brain. Furthermore, the selectivity
W. Shao et al. / Talanta 158 (2016) 361–367
was up to 94.6% compared to the previous report 44.2% [38], as far as we know, which was the highest selectivity amount the publications. The 650N-linked glycoproteins (94.8%) were annotated according to the information in the SwissProt Database, and 36Nlinked glycoproteins were newly discovered in this study. The improved glycopeptide enrichment could be contributed to the following characteristics: (1) large amounts of functional groups on the branched copolymer, forming rich hydrophilic layer; (2) the terminal “claw-like” symmetric polyhydroxyl formed hydrogen bonding with the glycopeptide in hydrophilic layer to enhance the binding capacity; (3) the branched peptide backbones with high biological compatibility, flexibility and low steric effect.
4. Conclusions In summary, a novel branched copolymer modified HILIC material Sil@Poly(THMA-co-MBAAm) was facilely synthesized by one-pot “thiol-ene” click reaction. Successful separations of polar compounds, peptides and improved enrichment efficiency of glycopeptides from animal tissue are achieved. We expect this branched copolymer modified HILIC material could not only provide a new method for glycosylation analysis in the complex biological samples, but also inspire branched copolymer strategy available applied to other types of enriched materials and the fabrication of other types of functionalized materials.
Acknowledgement The authors are grateful for the financial supports from the National Natural Science Foundation of China (21235005 and 91543201), National Basic Research Program of China (2012CB910601 and 2013CB911200) and the Creative Research Group Project by NSFC (21321064).
Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.talanta.2016.05. 034.
References [1] D.F. Zielinska, F. Gnad, J.R. Wisniewski, M. Mann, Cell 141 (5) (2010) 897–907. [2] J. Zaia, Nat. Methods 8 (1) (2011) 55–57. [3] T. Liu, W.J. Qian, M.A. Gritsenko, D.G. Camp 2nd, M.E. Monroe, R.J. Moore, R. D. Smith, J. Proteome Res. 4 (6) (2005) 2070–2080. [4] W.R. Alley Jr., B.F. Mann, M.V. Novotny, Chem. Rev. 113 (4) (2013) 2668–2732. [5] H. Zhang, T. Guo, X. Li, A. Datta, J.E. Park, J. Yang, S.K. Lim, J.P. Tam, S.K. Sze, Mol. Cell. Proteom. 9 (4) (2010) 635–647. [6] Z. Dai, J. Zhou, S.J. Qiu, Y.K. Liu, J. Fan, Electrophoresis 30 (17) (2009)
367
2957–2966. [7] J. Huang, H. Qin, Z. Sun, G. Huang, J. Mao, K. Cheng, Z. Zhang, H. Wan, Y. Yao, J. Dong, J. Zhu, F. Wang, M. Ye, H. Zou, Sci. Rep. 5 (2015) 10164. [8] D. Li, Y. Chen, Z. Liu, Chem. Soc. Rev. 44 (22) (2015) 8097–8123. [9] S. Shinde, Z. El-Schich, A. Malakpour, W. Wan, N. Dizeyi, R. Mohammadi, K. Rurack, A. Gjorloff Wingren, B. Sellergren, J. Am. Chem. Soc. 137 (43) (2015) 13908–13912. [10] H.Y. Wang, Z.J. Bie, C.C. Lu, Z. Liu, Chem. Sci. 4 (11) (2013) 4298–4303. [11] B. Jiang, Y. Liang, Q. Wu, H. Jiang, K. Yang, L. Zhang, Z. Liang, X. Peng, Y. Zhang, Nanoscale 6 (11) (2014) 5616–5619. [12] H.R. Jung, S. Sidoli, S. Haldbo, R.R. Sprenger, V. Schwammle, D. Pasini, K. Helin, O.N. Jensen, Anal. Chem. 85 (17) (2013) 8232–8239. [13] M.R. Gama, R.G. da Costa Silva, C.H. Collins, C.B.G. Bottoli, TrAC Trends Anal. Chem. 37 (0) (2012) 48–60. [14] H. Jiang, H. Yuan, Y. Qu, Y. Liang, B. Jiang, Q. Wu, N. Deng, Z. Liang, L. Zhang, Y. Zhang, Talanta 146 (2016) 225–230. [15] Q. Fu, Z.M. Guo, T. Liang, X.L. Zhang, Q. Xu, X.M. Liang, Anal. Methods 2 (3) (2010) 217–224. [16] J. Zhu, Z. Sun, K. Cheng, R. Chen, M. Ye, B. Xu, D. Sun, L. Wang, J. Liu, F. Wang, H. Zou, J. Proteome Res. 13 (3) (2014) 1713–1721. [17] Y. Weng, Z. Sui, H. Jiang, Y. Shan, L. Chen, S. Zhang, L. Zhang, Y. Zhang, Sci. Rep. 5 (2015) 9770. [18] K. Wolski, M. Szuwarzynski, S. Zapotoczny, Chem. Sci. 6 (3) (2015) 1754–1760. [19] A. Kopyshev, N. Lomadze, D. Feldmann, J. Genzer, S. Santer, Polymer 79 (2015) 65–72. [20] Z.X. Dong, H. Wei, J. Mao, D.P. Wang, M.Q. Yang, S.Q. Bo, X.L. Ji, Polymer 53 (10) (2012) 2074–2084. [21] D. Bontempo, N. Tirelli, K. Feldman, G. Masci, V. Crescenzi, J.A. Hubbell, Adv. Mater. 14 (17) (2002) 1239–1241. [22] M. Barsbay, O. Guven, T.P. Davis, C. Barner-Kowollik, L. Barner, Polymer 50 (4) (2009) 973–982. [23] G. Li, H.T. Zheng, Y.X. Wang, H. Wang, Q.B. Dong, R.K. Bai, Polymer 51 (9) (2010) 1940–1946. [24] W.J. Yang, K.G. Neoh, E.T. Kang, S.L.M. Teo, D. Rittschof, Polym. Chem. 4 (10) (2013) 3105–3115. [25] A.S. Nia, S. Rana, D. Dohler, W. Osim, W.H. Binder, Polymer 79 (2015) 21–28. [26] A.S. Goldmann, A. Walther, L. Nebhani, R. Joso, D. Ernst, K. Loos, C. BarnerKowollik, L. Barner, A.H.E. Mü ller, Macromolecules 42 (11) (2009) 3707–3714. [27] Y.R. Xin, J. Sakamoto, A.J. van der Vlies, U. Hasegawa, H. Uyama, Polymer 66 (2015) 52–57. [28] C.E. Hoyle, C.N. Bowman, Angew. Chem. Int. Ed. 49 (9) (2010) 1540–1573. [29] L. Zhang, W. Shao, Y. Liang, K. Yang, Y. Zhang, Patent: 201510862352.2, China, 12, 2015. [30] H. Qiu, M. Zhang, T. Gu, M. Takafuji, H. Ihara, Chem.: Eur. J. 19 (2013) 18004–18010. [31] A. Shen, Z. Guo, L. Yu, L. Cao, X. Liang, Chem. Commun. 47 (15) (2011) 4550–4552. [32] M. Pucic, A. Knezevic, J. Vidic, B. Adamczyk, M. Novokmet, O. Polasek, O. Gornik, S. Supraha-Goreta, M.R. Wormald, I. Redzic, H. Campbell, A. Wright, N.D. Hastie, J.F. Wilson, I. Rudan, M. Wuhrer, P.M. Rudd, D. Josic, G. Lauc, Mol. Cell. Proteom. 10 (10) (2011). [33] L. Yu, X. Li, Z. Guo, X. Zhang, X. Liang, Chem.: Eur. J. 15 (46) (2009) 12618–12626. [34] Y. Wada, P. Azadi, C.E. Costello, A. Dell, R.A. Dwek, H. Geyer, R. Geyer, K. Kakehi, N.G. Karlsson, K. Kato, N. Kawasaki, K.H. Khoo, S. Kim, A. Kondo, E. Lattova, Y. Mechref, E. Miyoshi, K. Nakamura, H. Narimatsu, M.V. Novotny, N.H. Packer, H. Perreault, J. Peter-Katalinic, G. Pohlentz, V.N. Reinhold, P.M. Rudd, A. Suzuki, N. Taniguchi, Glycobiology 17 (4) (2007) 411–422. [35] H.H. Bai, C. Fan, W.J. Zhang, Y.T. Pan, L. Ma, W.T. Ying, J.H. Wang, Y.L. Deng, X. H. Qian, W.J. Qin, Chem. Sci. 6 (7) (2015) 4234–4241. [36] J. Liu, F. Wang, J. Mao, Z. Zhang, Z. Liu, G. Huang, K. Cheng, H. Zou, Anal. Chem. 87 (4) (2015) 2054–2057. [37] H. Kaji, T. Shikanai, A. Sasaki-Sawa, H. Wen, M. Fujita, Y. Suzuki, D. Sugahara, H. Sawaki, Y. Yamauchi, T. Shinkawa, M. Taoka, N. Takahashi, T. Isobe, H. Narimatsu, J. Proteome Res. 11 (9) (2012) 4553–4566. [38] Y. Qu, S. Xia, H. Yuan, Q. Wu, M. Li, L. Zou, L. Zhang, Z. Liang, Y. Zhang, Anal. Chem. 83 (19) (2011) 7457–7463.
Contents lists available at ScienceDirect
Talanta journal homepage: www.elsevier.com/locate/talanta
Hydrogen-bond interaction assisted branched copolymer HILIC material for separation and N-glycopeptides enrichment Wenya Shao a,b, Jianxi Liu a,b, Kaiguang Yang a, Yu Liang a, Yejing Weng a,b, Senwu Li a,b, Zhen Liang a, Lihua Zhang a,n, Yukui Zhang a a National Chromatographic R. & A. Center, Key Laboratory of Separation Science for Analytical Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Sciences, Dalian 116023, China b University of Chinese Academy of Sciences, Beijing 100049, China
art ic l e i nf o
a b s t r a c t
Article history: Received 31 January 2016 Received in revised form 9 May 2016 Accepted 13 May 2016 Available online 24 May 2016
Hydrophilic interaction chromatography (HILIC) has attracted increasing attention in recent years due to its efficient application in the separation of polar compounds and the enrichment of glycopeptides. However, HILIC materials are still of weak hydrophilicity and thereby present weak retention and selectivity. In this work, branched copolymer modified hydrophilic material Sil@Poly(THMA-co-MBAAm), with high hydrophilicity and unique “claw-like” polyhydric groups, were prepared by “grafting from” thiol-ene click reaction. Due to the abundant functional groups provided by branched copolymer, the material showed excellent retention for nucleosides, necleobases, acidic compounds, sugars and peptides. Furthermore, Sil@Poly(THMA-co-MBAAm) was also applied for the N-glycosylation sites profiling towards the digests of the mouse brain, and 1997N-glycosylated peptides were identified, corresponding to 686 glycoprotein groups. Due to the assisted hydrogen-bond interaction, the selectivity for glycopeptide enrichment in the real sample reached 94.6%, which was the highest as far as we know. All these results indicated that such hydrogen-bond interaction assisted branched copolymer HILIC material possessed great potential for the separation and large scale glycoproteomics analysis. & 2016 Published by Elsevier B.V.
Keywords: Hydrophilic interaction chromatography Glycopeptide Branched copolymer Grafting from Hydrogen-Bond Interaction
1. Introduction N-glycosylation plays crucial roles in intercellular recognition and communication, protein folding and immune responses, signal transduction, and so forth [1–3]. The analysis based on mass spectrometry (MS) is an efficiency tool for large scale and in-depth glycoprotein/glycopeptides profiling in complex biological samples [4,5]. However, the great challenges are the inherently low abundance and poor ionization efficiency of glycopeptides. Therefore, the selective enrichment of glycopeptides is indispensable for MS identification. A variety of enrichment methods such as lectin affinity [6], hydrazide chemistry [7], boronic acid [8–10] and Hydrophilic interaction chromatography (HILIC) [11] have been developed, among which HILIC has been widely used, with advantage of good compatibility with LC-MS analysis[12] and low bias to different types of glycopeptides. Besides, with the increasing popularity of pharmaceutical analysis, metabolomics and proteomics, HILIC, as a valuable method for the separation of polar substances, has attracted great attention. n
Corresponding author. E-mail address: [email protected] (L. Zhang).
http://dx.doi.org/10.1016/j.talanta.2016.05.034 0039-9140/& 2016 Published by Elsevier B.V.
As the increasing attention, HILIC materials, such as –NH2, – CONH, –CN, and –CH(OH)CH2OH modified particles [13], hydrophilic monoliths [14], have been developed in recent years. However, their hydrophilicity and selectivity limited the application of HILIC in the complex sample; weak interaction and low selectivity are apparent drawback for capturing the glycopeptides. Saccharide-modified silica stationary phase, as a rising star among the HILIC materials, has been successfully applied for the separation of polar compounds [15] and the enrichment of glycopeptides [16,17]. However, such materials were suffered from two limitations: 1) poor hydrophilicity, which would result in high percentage of organic solvent in initial mobile phase, is incompatible with biomolecules especially protein and peptide; 2) low stability of ring glycan structure. Compared with natural saccharides functional groups, N-[Tris(hydroxymethyl)methyl]acrylamide (THMA) with three terminal symmetric hydroxyl would not only possess preferable structure properties but also improve the stability of the glycan structure. In addition, hydrophilicity modification can be achieved by the “grafting from” polymerization. Compared with other polymerization methods, it allows polymer chains to grow from the initiators on the substrate to generate high grafting densities. It should be mentioned that various approaches can be employed for
362
W. Shao et al. / Talanta 158 (2016) 361–367
“grafting from”, e.g. surface-initiated ionic polymerization [18,19], living/controlled free polymerizations such as atom transfer radical polymerization (ATRP) [20,21], reversible-addition fragmentation chain transfer (RAFT) polymerization [22,23] or click chemistry [24,25]. Among them, thiol-ene click chemistry has the characteristics of regiospecificity, stereo-specificity and amenability to variety of readily available starting compounds with enes [26–28]. These characteristics inspired us that “grafting from” might be a solution to the problems on saccharide-modified silica stationary phase. Herein, in this work, a novel hydrophilic stationary phase, branched copolymer modified hydrophilic material Sil@Poly (THMA-co-MBAAm) were fabricated via one-pot thiol-ene click chemistry through “grafting-from” approach with N-[Tris(hydroxymethyl)methyl]acrylamide (THMA) and N,N′-methylenebisacrylamide (MBAAm) as the monomers [29]. Finally, this branched copolymer modified material Sil@Poly(THMA-co-MBAAm) was successfully applied in the separation of polar compounds, sugars, peptides and the analysis of glycosylation sites.
2. Experimental 2.1. Reagents and materials Porous silica particles (diameter 5 mm, pore size 150 Å) were purchased from Fuji Silysia Chemical (Aichi, Japan). Bovine serum albumin (BSA, bovine serum) was obtained from Sino-American Biotec (Luoyang, China). PNGase F was bought from New England Biolabs (Ipswich, MA, US). Dithiothreitol (DTT) and iodoacetamide (IAA) were from Acros (Morris Plains, NJ, US). Acetonitrile (ACN, HPLC grade) was ordered from Merck (Darmstadt, Germany). Water was purified by a Milli-Q system (Millipore, Milford, MA, US). N-[Tris(hydroxymethyl)methyl]acrylamide (THMA, 93%), (3mercaptopropyl)trimethoxysilane (MPS, 95%), N,N′-methylenebisacrylamide (MBAAm, 98%), thymidine, uridine, cytidine, guanosine, cytosine, phenol, benzoic acid, 4-hydroxybenzoic acid, 3,5dinitrobenzoic acid, 3-nitrobenzoic acid, ammonium acetate, acetic acid, trypsin (bovine pancreas), IgG (human serum), and trifluoroacetic acid (TFA) were purchased from Sigma-Aldrich (St. Louis, MO). Venusil XPB C18 particles (5 mm, 100 Å) were ordered from Daiso (Osaka, Japan). 2.2. Synthesis of Sil@Poly(THMA-co-MBAAm) The material was prepared via thiol-ene click copolymerization by the following two steps. At first, the MPS modified silica (SilMPS) was obtained based on the previous report [30]. Silica particles (5 g) were suspended in 100 mL of anhydrous toluene solution, then 1.7 mL of (3-mercaptopropyl)trimethoxysilane was added, followed by refluxing and mechanical agitation for 24 h at 110 °C. After reaction, the Sil-MPS was washed with dry toluene, methanol, acetone, and dried under vacuum overnight. Second, the copolymer grafted particles, Sil@Poly(THMA-co-MBAAm), were prepared as follows: THMA (1.16 g), MBAAm (0.97 g) and AIBN (0.1 g) were added to the suspension of Sil-MPS (3.5 g) in water-methanol (2:3, v/v, 145 mL) and the reaction mixture was bubbled with nitrogen for 15 min to remove oxygen, followed by mechanical agitation at 70 °C for 24 h. After the reaction, excess reagents and oligomers were removed by repeatedly washing with water and ethanol in succession. The obtained Sil@Poly(THMA-coMBAAm) was dried under vacuum at 40 °C for 12 h and stored at room temperature for further usage.
2.3. Characterization Thermogravimetric analysis (TGA) was carried out on a Cahn Thermax 500 instrument (Thermo Scientific, US) from 30 to 980 °C with a heating rate of 10 °C min 1 under flow nitrogen. Fourier transform infrared spectra (FT-IR) were performed on a Bruker Optics HYPERION 3000 (Bruker, Daltonios, Germany) in the reflection mode, and scanned at the range of 500–4500 cm 1. The specific surface area and mesopore size were determined by nitrogen adsorption-desorption measurements (QuadraSorb SI4, Florida, US), while the surface area and the mesopore size were determined from the isotherms by BET method and BJH method, respectively. 2.4. Column packing and chromatographic conditions A slurry of Sil@Poly(THMA-co-MBAAm) in the methanol was prepared by ultrasonication for 2 min, then it was packed into the stainless-steel column (150 mm 4.6 mm i.d.) at a constant pressure of 40 MPa using methanol as the flushing solvent. The HILIC column (10 mm 4.6 mm i.d.) used for glycopeptide enrichment was prepared as well. Separations were performed by the HPLC with a diode array detector (Chromaster HPLC, Hitachi High-Tech, TKY, JP) on 150 mm 4.6 mm i.d. The flow rate was 1.00 mL min 1 with the UV detection at 254 nm, and the column temperature was set at 25 °C. The mobile phase consisted of an aqueous phase with ammonium acetate (NH4Ac) and an organic phase ACN. 2.5. Sample preparation All the small molecules and peptides used for the separation were dissolved in ACN/H2O (85/15, V/V). The concentrations of standard solutions were 60 ng/mL for nucleosides and necleobases separately, 10–50 ng/mL for acidic solutions. IgG and BSA digests were separately prepared according to the following protocol. Proteins were dissolved in denaturing buffer (6 M guanidine hydrochloride), then reduced in 100 mM DTT at 56 °C for 1.5 h, alkylated by 200 mM IAA for 40 min at room temperature in the dark. The solution was diluted ten-fold with 50 mM NH4HCO3 buffer (pH 8.0) and digested with trypsin (enzyme/protein ratio of 1:30, w/w) at 37 °C for 12 h. The tryptic digested peptides were desalted by home-made C18 precolumn, lyophilized and stored at 20 °C before usage. HeLa cells were cultured in a humidified 37 °C incubator with DMEM medium containing 10% fetal bovine serum (FBS) and 5% CO2. Cells were harvested when nearly at 90% confluence and washed with 4 °C phosphate-buffered saline (PBS) for 3 times. Cell pellets were collected and suspended in 6 M guanidine hydrochloride (1% (v/v) protease inhibitor cocktail), ultrasonicated on ice for 120 s in total (10 s intervals every 10 s). Then, the samples were centrifuged at 20,000 rpm at 4 °C for 20 min The resulting supernatants containing the total Hela cell proteins were collected and the protein concentration was determined by the BCA assay. The digestion of Hela cells extracts was the same as that of human IgG. 2.6. Glycopeptides enrichment The enrichment procedure followed the previous protocol [17] with minor modification. The tryptic digested IgG was firstly redissolved in ACN/H2O/TFA (80/20/0.1, v/v/v) (buffer A) and loaded onto the HILIC column (10 mm 4.6 mm i.d.). Then, the column was rinsed with buffer A at 1.0 mL min 1 for 10 min to remove nonglycopeptides. Glycopeptides were eluted with ACN/H2O/TFA (50/50/0.1, v/v/v), lyophilized and redissolved in 20 mM NH4HCO3 buffer (pH 8.0). Subsequently, the glycans were released by adding
W. Shao et al. / Talanta 158 (2016) 361–367
363
PNGase F and incubated at 37 °C for 12 h. Finally, the peptides were desalted and dried for further usage. The tryptic digests from BSA and IgG mixture and Hela cells were treated in the same way. 2.7. Mass spectrometry and data processing The glycopeptides from IgG were analyzed by MALDI-TOF-MS. A 1 mL of sample and 1 mL of DHB matrix solution (10 mg/mL, 0.1% TFA in 60% ACN aqueous solution) were orderly dropped onto a MALDI plate for MS analysis. All experiments were performed on Bruker Ultraflex III MALDI-TOF/TOF MS instrument (Bruker, Daltonios, Germany). The deglycosylated peptides enriched from HeLa cell lysates were analyzed by nanoRPLC-MS/MS on a Q Exactive MS (Thermo Fisher Scientific, USA). The lyophilized sample was dissolved in 70 mL of 0.1% FA solution, and 12 mL of that was loaded onto a homemade capillary separation column (75 mm i.d. 14 cm). Mobile phase A (0.1% FA in H2O) and B (0.1% FA in ACN) were used to establish the 92 min gradient, which went from 2% B to 7% B over 2 min, to 22% B over 60 min, to 35% B over 20 min, and then to 80% B over 10 min, with the flow rate at 300 nL/min. The Q-Exactive was operated in positive ion data dependent mode with 2.5 kV spray voltage and the temperature of ion transfer capillary was set at 275 °C. One full scan MS acquired from m/z 300–1800 was followed by fifteen data dependent MS/MS events. MS1 was performed at the resolution of 70,000, with automatic gain control (AGC) value (1e6), maximum injection time (100 ms). MS2 was performed at the resolution of 17,500 (AGC: 1e5, maximum injection time: 50 ms). The *.raw files produced on Q-Exactive MS were searched against the Swiss-Prot mouse complete proteome sequence database (release 2014_03). The parameters were as follows: enzyme, trypsin; missed cleavages, two; fixed modifications, carboxyamidomethylation (C); variable modifications, oxidation (M) and deamidation (N); peptide tolerance, 7 ppm; MS/MS tolerance, 20 mmu; target FDR (Stric), 0.01; target FDR (Relaxed), 0.05. The N-glycosylation sites were identified if the peptides contained the sequence of N-!P-S/T and had the modification of deamination at N at the same time.
3. Result and discussion 3.1. Preparation and characterization of the Sil@Poly(THMA-coMBAAm) material The Sil@Poly(THMA-co-MBAAm) stationary phase was prepared by “thiol-ene” click copolymerization as shown in Schem. 1. Briefly, MPS was immobilized on the silica surface via Si–O bond. Then, the branched copolymer of silica surface was formed by “thiol-ene”click reaction and “graft from” polymerization (Fig. S1). In this work, MBAAm and THMA were chosen as the co-monomers. Due to the branched polymer grafted surface, this particle possessed the following characteristics (Schem. 1): 1) Hydrophilicity of the material could be enhanced by the abundant of availability amide functional groups on the branched copolymer; 3) The extended “claw-like” polyhydroxy structure would provide stronger and more specific affinity through hydrogen bonding interaction toward glycopeptides. The characterization of the surface morphology of Sil@Poly (THMA-co-MBAAm) was observed by SEM and compared with the raw silica (Fig. S2). The Sil@Poly(THMA-co-MBAAm) particles are well dispersed, smooth and sphere in shape, which are important for the separation. The amount of the grafted copolymer was determined by thermogravimetric analysis (TGA). As shown in Fig. 1, only 1% weight of bare silica particles in the region below 200 °C,
Scheme 1. Synthesis and structure of Sil@Poly(THMA-co-MBAAm).
Fig. 1. TGA analysis of (a) bare silica, (b) Sil-MPS and (c) Sil@Poly(THMA-coMBAAm) particles.
which can be attributed to the loss of physically adsorbed water. Approximately 3% weight loss is observed for Sil-MPS during 200– 700 °C. In contrast, 20% weight loss of Sil@Poly(THMA-co-MBAAm) is found. Therefore, the percentage of graft reaction approximately equals to 17%. It is indicated that high density functional groups were immobilized on the surface of silica particles. The specific surface area and pore size were two most important parameters that determined the loading capacity and diffusion rate of analytes for chromatography stationary phase. The morphologies of the SilPoly(MBAAm-co-AHPS) were further characterized by nitrogen adsorption–desorption analysis. As shown in Fig. 2, for bare silica particles, the pore size is 24 nm by BJH method. After the graft of copolymer, the pore size is reduced to 17 nm with mean pore-size distribution. Deductively, the copolymers were uniformly grown on the surface of the silica particles, due to the high-efficiency thiol-ene reaction. In contrast, the BET specific surface area of Sil@Poly(THMA-co-MBAAm) slightly changed from 157 to 153 m2/ g without obvious decrease, since the stretched copolymers would increase the particle surface. Next, FT-IR was also applied to characterize the stationary phase. As shown in Fig. 3, in the spectrum of Sil@Poly(THMA-coMBAAm), the peak at 1382 cm 1 is the characteristic vibration of – OH in –CH2– of THMA. The peak at 1650 cm 1 can be attributed to C ¼O stretching vibration and the peak at 1531 cm 1 can be assigned to N–H bending vibration bowing to amide group of
364
W. Shao et al. / Talanta 158 (2016) 361–367
Fig. 2. BJH adsorption pore size distribution of (a) bare silica and (b) Sil@Poly(THMA-co-MBAAm) particles.
the silica particles. 3.2. Chromatographic separation performance
Fig. 3. FT-IR spectra of (a) bare silica and (b) Sil@Poly(THMA-co-MBAAm) particles.
copolymer. Bands at 2956 cm 1 and 2882 cm 1 both were assigned to aliphatic C–H stretching. In contrast, there is no corresponding peak can be found in the bare silica particles. These results also demonstrated the successful grafting on the surface of
The chromatographic retention behavior of Sil@Poly(THMA-coMBAAm) for polar compounds was evaluated using thymidine, uridine, cytidine, guanosine and cytosine, which are typically strong polar compounds which have similar skeletal structure. As shown in Fig. 4(a), baseline separation of the five compounds is achieved. Meanwhile, all the retention time is between 3 min to 15 min, indicating the appropriate retention of Sil@Poly(THMA-coMBAAm) packed column. A comparison of Sil@Poly(THMA-coMBAAm) column with the other two HILIC columns (amide and silica) in the separation has been shown in Fig. 4(a). Compared with the other commonly used HILIC stationary phases, 1) Sil@Poly (THMA-co-MBAAm) column was the best in separating necleobases and nucleosides. 2) The retention factors were usually 1.5– 4.5 and 1.5–8.5 respectively for cytosine and cytidine [31], while the Sil@Poly(THMA-co-MBAAm) column exhibited much better retention factors (6.5 and 9.1). To further demonstrate the applicability of the Sil@Poly(THMA-co-MBAAm) stationary phase, the separation of acid compounds (the pKa values listed in Fig. S3) were performed in HILIC mode. As shown in Fig. 4(b), phenol, benzoic acid, 4-hydroxybenzoic acid, 3,5-dinitrobenzoic acid and p-nitrobenzoic acid are well separated within 10 min on the
Fig. 4. (a) A comparison of separation behaviors on Sil@Poly(THMA-co-MBAAm) HILIC column and two different HILIC column. Mobile phase: ACN/H2O (85/15, v/v) containing 15 mM ammonium acetate. Peaks: (1) thymidine, (2) uridine, (3) cytosine, (4) cytidine, (5) guanosine. (b) Separation of acid compounds on Sil@Poly(THMA-coMBAAm) HILIC column Mobile phase: ACN/H2O (86/14, v/v) containing 45 mM ammonium acetate, Peaks: (1) phenol, (2) 3-nitrobenzoic acid, (3) 3,5-dinitrobenzoic acid, (4) benzoic acid, (5) 4-hydroxybenzoic acid. Separation on Sil@Poly(THMA-co-MBAAm) packed column (150 mm 4.6 mm i.d.). Separation condition: T ¼25 °C; flow rate ¼ 1 mL min 1; UV detection, λ ¼254 nm.
W. Shao et al. / Talanta 158 (2016) 361–367
365
Fig. 5. (a) Separation of seven sugars on Sil@Poly(THMA-co-MBAAm) HILIC column. Mobile phase: ACN/H2O (80/20, v/v). Evaporative light scattering detector (ELSD): gas pressure 30 psi, tube temperature 70 °C, gain 10. Peak: (1) methy α-D-mannopyranoside, (2) ribose, (3) mannitol, (4) xylose, (5) N-acetylglucosamine, (6) fructose, (7) maltitol. (b) Separation of four peptides on Sil@Poly(THMA-co-MBAAm) HILIC column Mobile phase: A, H2O; B, ACN; C, 50 mM NaH2PO4; 0–15 min, 15% A, 75% B, 10% C 40% A, 50% B, 10% C;UV detection: 214 nm; Peak: (1) angiotensin Ⅳ, (2) bradykinin, (3) leucine enkephaline, (4) angiotensin Ⅱ.
Fig. 6. (a) Effect of salt concentration on the separation. Mobile phases of ACN/H2O (85/15, v/v) contained different concentration of salt. T ¼25 °C; flow rate ¼1 mL min 1; UV detection, λ¼ 254 nm. Peak: (1) 3-nitrobenzoic acid, (2) thymidine, (3) 4-hydroxybenzoic acid, (4) cytosine. (b) Effect of column temperature on the separation. Mobile phase: ACN/H2O (85/15, v/v) containing 20 mM ammonium acetate; flow rate ¼1 mL min 1. Peak: (1) 3-nitrobenzoic acid, (2) thymidine, (3) 4-hydroxybenzoic acid, (4) cytosine.
Fig. 7. The stability of Sil@Poly(THMA-co-MBAAm) HILIC column. Mobile phase: ACN/H2O (85/15, v/v) containing 15 mM ammonium acetate. Peaks: (1) thymidine, (2) uridine, (3) cytosine, (4) cytidine, (5) guanosine. Other chromatographic conditions were the same as Fig. 4.
Sil@Poly(THMA-co-MBAAm) column. It should be noted that the peaks were sharp and symmetrical. Considering the excellent chromatographic performance of the Sil@Poly(THMA-co-MBAAm) stationary phase, the separation of peptides and sugars were also carried out on this material in HILIC mode. Seven sugars including mono- and disaccharides were mixed as the model compounds. The sugars were efficiently separated with good peak shape with 80% acetonitrile (Fig. 5(a)). Meanwhile, four peptides were well separated with proper retention (Fig. 5(b)). High efficiency separation of nucleosides, necleobases, acidic compounds, sugars and peptides proved that excellent hydrophilicity is achieved through this hydrogen-bond interaction assisted branched copolymer HILIC material. The effects of ACN content, buffer concentration and column temperature were also investigated to reveal the chromatographic characterization. Three polar compounds (thymidine, cytosine and 3-nitrobenzoic acid) were used to evaluated the effects of the percentage of ACN in the mobile phase. As shown in the Fig. S4, retention time of analytes increased with the percentage of ACN in the mobile phase from 65% to 90%, which was consistent with the typical characteristics of HILIC retention. Buffer concentration
366
W. Shao et al. / Talanta 158 (2016) 361–367
Fig. 8. MALDI-TOF mass spectra of enrichment with the Sil@Poly(THMA-co-MBAAm) packed HILIC column (a) IgG digests, (b) the mixture of human IgG and BSA digest (with a mass ratio of 1:10), and (c) the mixture of human IgG and BSA digest (with a mass ratio of 1:100).
could also affect the separation property on HILIC. Retention behavior was investigated by varying the ammonium acetate buffer concentration in the mobile phase from 10 mM to 40 mM by thymidine, cytosine, 4-hydroxybenzoic acid and 3-nitrobenzoic acid. As shown in Fig. 6(a), the retention time of 4-hydroxybenzoic acid and 3-nitrobenzoic acid increased as the buffer concentration increased, while thymidine and cytosine changed little with the increase of buffer concentrations. This result can be ascribed to the electrostatic interactions between the acids and the stationary phase surface. With the increase of the buffer concentration, the electrostatic repulsion interaction was weakened and the retention was enhanced. The effect of the column temperature on separation was additionally investigated. The retention on this Sil@Poly(THMA-co-MBAAm) HILIC column decreases at elevated temperatures (Fig. 6(b)), presumably due to lower viscosity and higher diffusivity in high column temperature. It should be noted that the bandwidths of the peaks of analytes improved as column temperature increased. In conclusion, the percentage of ACN in the mobile phase is the most important factors for the sample retention among the three factors. Stability of the stationary phase was examined using model compounds thymidine, uridine, cytosine, cytidine and guanosine. Fig. 7 shows retention chromatograms for three times injections, among which (a) and (b) were two continuous injections. After working for 72 h in 9 months, this column was repeatedly applied for separation (c). There were no obvious changes with these analytes, indicating that this Sil@Poly(THMA-co-MBAAm) HILIC column was stable. 3.3. Application in the glycopeptide enrichment Based on the hydrophilicity and unique claw-like polyhydroxy of the prepared stationary phase, the Sil@Poly(THMA-co-MBAAm) materials was applied to enrich glycopeptides. IgG, a glycoprotein which contains one N-glycosylated site in each heavy chain (IgG
1 and IgG 2) at the highly conserved asparagine 297 residue in each of the CH2 domains of the Fc region, was digested and utilized as the model to evaluate the performance of the prepared Sil@Poly (THMA-co-MBAAm) column (10 mm 4.6 mm i.d.) [32]. In the MALDI-TOF MS spectrum of the native digest (Fig. S5), the glycopeptides signal was interfered and suppressed by amount of nonglycopeptides. After enrichment with the Sil@Poly(THMA-coMBAAm) column (Fig. 8(a)), 7 clusters of glycopeptides peaks were emerged and the corresponding glycan structure of 17 glycopeptide was determined according to the previous reports [33,34], while the signals of non-glycopeptides disappeared. After treatment with PNGase F with 12 h, the N-linked glycans were cleaved, leaving two deglycosylated peptides at m/z 1158.5 and 1190.5 representing of EEQFN#STFR and EEQYN#STYR (in which N# denote the N-linked glycosylation site) (Fig. S6). The digest mixtures of non-glycoprotein BSA and IgG at a mass ratio of 1:10 and 1:100 were further used to evaluate the enrichment selectivity of this column. As shown in Fig. 8(b) and (c), 14 and 11 glycopeptides were still identified with high S/N, and non-glycopeptides were completely removed. This demonstrated that the glycopeptides could be enriched with high specificity and selectivity. 3.4. Application in N-glycoproteome analyses of mouse brain To further evaluate the enrichment capability of Sil@Poly (THMA-co-MBAAm), the glycopeptides from trypic digests of the proteins extracted from mouse brain were enriched by the materials and then treated with PNGase F, followed by further nanoRPLC-ESI-MS/MS analysis. In total, 2111 unique peptides and 743 nonredundant protein groups were identified, of which 1997 unique N-glycopeptides mapped to 686 N-linked glycoprotein groups (Table S1) were obtained, much more than the results from the previous publications [35–37] based on different glycopeptides enrichment materials, all of which covered less than 1000 unique N-glycopeptides from mouse brain. Furthermore, the selectivity
W. Shao et al. / Talanta 158 (2016) 361–367
was up to 94.6% compared to the previous report 44.2% [38], as far as we know, which was the highest selectivity amount the publications. The 650N-linked glycoproteins (94.8%) were annotated according to the information in the SwissProt Database, and 36Nlinked glycoproteins were newly discovered in this study. The improved glycopeptide enrichment could be contributed to the following characteristics: (1) large amounts of functional groups on the branched copolymer, forming rich hydrophilic layer; (2) the terminal “claw-like” symmetric polyhydroxyl formed hydrogen bonding with the glycopeptide in hydrophilic layer to enhance the binding capacity; (3) the branched peptide backbones with high biological compatibility, flexibility and low steric effect.
4. Conclusions In summary, a novel branched copolymer modified HILIC material Sil@Poly(THMA-co-MBAAm) was facilely synthesized by one-pot “thiol-ene” click reaction. Successful separations of polar compounds, peptides and improved enrichment efficiency of glycopeptides from animal tissue are achieved. We expect this branched copolymer modified HILIC material could not only provide a new method for glycosylation analysis in the complex biological samples, but also inspire branched copolymer strategy available applied to other types of enriched materials and the fabrication of other types of functionalized materials.
Acknowledgement The authors are grateful for the financial supports from the National Natural Science Foundation of China (21235005 and 91543201), National Basic Research Program of China (2012CB910601 and 2013CB911200) and the Creative Research Group Project by NSFC (21321064).
Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.talanta.2016.05. 034.
References [1] D.F. Zielinska, F. Gnad, J.R. Wisniewski, M. Mann, Cell 141 (5) (2010) 897–907. [2] J. Zaia, Nat. Methods 8 (1) (2011) 55–57. [3] T. Liu, W.J. Qian, M.A. Gritsenko, D.G. Camp 2nd, M.E. Monroe, R.J. Moore, R. D. Smith, J. Proteome Res. 4 (6) (2005) 2070–2080. [4] W.R. Alley Jr., B.F. Mann, M.V. Novotny, Chem. Rev. 113 (4) (2013) 2668–2732. [5] H. Zhang, T. Guo, X. Li, A. Datta, J.E. Park, J. Yang, S.K. Lim, J.P. Tam, S.K. Sze, Mol. Cell. Proteom. 9 (4) (2010) 635–647. [6] Z. Dai, J. Zhou, S.J. Qiu, Y.K. Liu, J. Fan, Electrophoresis 30 (17) (2009)
367
2957–2966. [7] J. Huang, H. Qin, Z. Sun, G. Huang, J. Mao, K. Cheng, Z. Zhang, H. Wan, Y. Yao, J. Dong, J. Zhu, F. Wang, M. Ye, H. Zou, Sci. Rep. 5 (2015) 10164. [8] D. Li, Y. Chen, Z. Liu, Chem. Soc. Rev. 44 (22) (2015) 8097–8123. [9] S. Shinde, Z. El-Schich, A. Malakpour, W. Wan, N. Dizeyi, R. Mohammadi, K. Rurack, A. Gjorloff Wingren, B. Sellergren, J. Am. Chem. Soc. 137 (43) (2015) 13908–13912. [10] H.Y. Wang, Z.J. Bie, C.C. Lu, Z. Liu, Chem. Sci. 4 (11) (2013) 4298–4303. [11] B. Jiang, Y. Liang, Q. Wu, H. Jiang, K. Yang, L. Zhang, Z. Liang, X. Peng, Y. Zhang, Nanoscale 6 (11) (2014) 5616–5619. [12] H.R. Jung, S. Sidoli, S. Haldbo, R.R. Sprenger, V. Schwammle, D. Pasini, K. Helin, O.N. Jensen, Anal. Chem. 85 (17) (2013) 8232–8239. [13] M.R. Gama, R.G. da Costa Silva, C.H. Collins, C.B.G. Bottoli, TrAC Trends Anal. Chem. 37 (0) (2012) 48–60. [14] H. Jiang, H. Yuan, Y. Qu, Y. Liang, B. Jiang, Q. Wu, N. Deng, Z. Liang, L. Zhang, Y. Zhang, Talanta 146 (2016) 225–230. [15] Q. Fu, Z.M. Guo, T. Liang, X.L. Zhang, Q. Xu, X.M. Liang, Anal. Methods 2 (3) (2010) 217–224. [16] J. Zhu, Z. Sun, K. Cheng, R. Chen, M. Ye, B. Xu, D. Sun, L. Wang, J. Liu, F. Wang, H. Zou, J. Proteome Res. 13 (3) (2014) 1713–1721. [17] Y. Weng, Z. Sui, H. Jiang, Y. Shan, L. Chen, S. Zhang, L. Zhang, Y. Zhang, Sci. Rep. 5 (2015) 9770. [18] K. Wolski, M. Szuwarzynski, S. Zapotoczny, Chem. Sci. 6 (3) (2015) 1754–1760. [19] A. Kopyshev, N. Lomadze, D. Feldmann, J. Genzer, S. Santer, Polymer 79 (2015) 65–72. [20] Z.X. Dong, H. Wei, J. Mao, D.P. Wang, M.Q. Yang, S.Q. Bo, X.L. Ji, Polymer 53 (10) (2012) 2074–2084. [21] D. Bontempo, N. Tirelli, K. Feldman, G. Masci, V. Crescenzi, J.A. Hubbell, Adv. Mater. 14 (17) (2002) 1239–1241. [22] M. Barsbay, O. Guven, T.P. Davis, C. Barner-Kowollik, L. Barner, Polymer 50 (4) (2009) 973–982. [23] G. Li, H.T. Zheng, Y.X. Wang, H. Wang, Q.B. Dong, R.K. Bai, Polymer 51 (9) (2010) 1940–1946. [24] W.J. Yang, K.G. Neoh, E.T. Kang, S.L.M. Teo, D. Rittschof, Polym. Chem. 4 (10) (2013) 3105–3115. [25] A.S. Nia, S. Rana, D. Dohler, W. Osim, W.H. Binder, Polymer 79 (2015) 21–28. [26] A.S. Goldmann, A. Walther, L. Nebhani, R. Joso, D. Ernst, K. Loos, C. BarnerKowollik, L. Barner, A.H.E. Mü ller, Macromolecules 42 (11) (2009) 3707–3714. [27] Y.R. Xin, J. Sakamoto, A.J. van der Vlies, U. Hasegawa, H. Uyama, Polymer 66 (2015) 52–57. [28] C.E. Hoyle, C.N. Bowman, Angew. Chem. Int. Ed. 49 (9) (2010) 1540–1573. [29] L. Zhang, W. Shao, Y. Liang, K. Yang, Y. Zhang, Patent: 201510862352.2, China, 12, 2015. [30] H. Qiu, M. Zhang, T. Gu, M. Takafuji, H. Ihara, Chem.: Eur. J. 19 (2013) 18004–18010. [31] A. Shen, Z. Guo, L. Yu, L. Cao, X. Liang, Chem. Commun. 47 (15) (2011) 4550–4552. [32] M. Pucic, A. Knezevic, J. Vidic, B. Adamczyk, M. Novokmet, O. Polasek, O. Gornik, S. Supraha-Goreta, M.R. Wormald, I. Redzic, H. Campbell, A. Wright, N.D. Hastie, J.F. Wilson, I. Rudan, M. Wuhrer, P.M. Rudd, D. Josic, G. Lauc, Mol. Cell. Proteom. 10 (10) (2011). [33] L. Yu, X. Li, Z. Guo, X. Zhang, X. Liang, Chem.: Eur. J. 15 (46) (2009) 12618–12626. [34] Y. Wada, P. Azadi, C.E. Costello, A. Dell, R.A. Dwek, H. Geyer, R. Geyer, K. Kakehi, N.G. Karlsson, K. Kato, N. Kawasaki, K.H. Khoo, S. Kim, A. Kondo, E. Lattova, Y. Mechref, E. Miyoshi, K. Nakamura, H. Narimatsu, M.V. Novotny, N.H. Packer, H. Perreault, J. Peter-Katalinic, G. Pohlentz, V.N. Reinhold, P.M. Rudd, A. Suzuki, N. Taniguchi, Glycobiology 17 (4) (2007) 411–422. [35] H.H. Bai, C. Fan, W.J. Zhang, Y.T. Pan, L. Ma, W.T. Ying, J.H. Wang, Y.L. Deng, X. H. Qian, W.J. Qin, Chem. Sci. 6 (7) (2015) 4234–4241. [36] J. Liu, F. Wang, J. Mao, Z. Zhang, Z. Liu, G. Huang, K. Cheng, H. Zou, Anal. Chem. 87 (4) (2015) 2054–2057. [37] H. Kaji, T. Shikanai, A. Sasaki-Sawa, H. Wen, M. Fujita, Y. Suzuki, D. Sugahara, H. Sawaki, Y. Yamauchi, T. Shinkawa, M. Taoka, N. Takahashi, T. Isobe, H. Narimatsu, J. Proteome Res. 11 (9) (2012) 4553–4566. [38] Y. Qu, S. Xia, H. Yuan, Q. Wu, M. Li, L. Zou, L. Zhang, Z. Liang, Y. Zhang, Anal. Chem. 83 (19) (2011) 7457–7463.