Letter pubs.acs.org/ac High-Sensitivity N‑Glycoproteomic Analysis of Mouse Brain Tissue by Protein Extraction with a Mild Detergent of N‑Dodecyl β‑D-Maltoside Jing Liu,†,‡ Fangjun Wang,*,† Jiawei Mao,†,‡ Zhang Zhang,†,‡ Zheyi Liu,†,‡ Guang Huang,†,‡ Kai Cheng,†,‡ and Hanfa Zou*,† † Key Laboratory of Separation Sciences for Analytical Chemistry, National Chromatographic R&A Center, Dalian Institute of Chemical Physics, Chinese Academy of Sciences (CAS), Dalian 116023, China ‡ University of Chinese Academy of Sciences, Beijing 100049, China S Supporting Information * ABSTRACT: N-dodecyl β-D-maltoside (DDM), a mild detergent with the ability to maintain the enzyme activity and solubilize hydrophobic proteins without changing their structures, was applied for N-glycoproteomic analysis of minute protein sample from mouse brain tissue. After combining with the capillary-based glycoproteomic reactor, 281 N-glycosylation sites were successfully characterized from 50 μg of mouse brain tissue, which was 110% higher at least than those obtained by conventional strategies. P buffers with urea and conventional detergents (SDS, Triton X100, and RapiGest SF) for minute protein sample extraction. Furthermore, as DDM is compatible with trypsin digestion, the tedious detergent removal steps were avoided and the protein sample can be online purified by a reversed phase trap column before liquid chromatography coupled with tandem mass spectrometry (LC−MS/MS) analysis. Combining this new protein sample extraction strategy with our previous reported capillary-based N-glycoproteomic reactor,18 we have characterized 281 N-glycosylation sites corresponding to 170 Nglycoproteins from only 50 μg of mouse brain tissue, which is about 2 times the numbers identified by using conventional sample preparation strategies. We first optimized the concentration of DDM within the lysis buffer. Mouse brain tissues (about 5−10 mg for each) were lysed using the lysis buffers with 0, 0.1%, 0.25%, 0.5%, 1%, and 2% DDM at the ratio of 10 μL buffer/1 mg tissue, respectively. The protein concentration of each lysate was 5.0 ± 0.2, 6.5 ± 1.0, 7.2 ± 1.0, 7.5 ± 1.1, 9.0 ± 0.5, and 8.8 ± 0.6 mg/ mL, respectively (Supporting Information, Figure S1 and Table S1). The protein concentration first increased along with the increase of DDM concentration and reached to a maximum after the DDM concentration was ≥1%. Therefore, lysis buffer with 1% DDM was utilized in our following experiments. Then, we investigated the performance of protein extraction by using rotein extraction, including the lysis of cells or tissues, release of proteins from different cell compartments, and solubilization of proteins into extraction buffer, is the first step of protein sample preparation.1,2 Usually, the transmembrane proteins are too hydrophobic to extract into the aqueous solution, and strong detergents and chaotropes are necessary to disrupt the membranes and increase the protein extraction efficiency.3−5 Strong detergents, such as sodium dodecyl sulfate (SDS),6 are most widely utilized in the extraction of proteins with high hydrophobicity, such as the glycosylated transmembrane proteins. However, the strong detergents would significantly decrease the efficiency of enzyme digestion,7,8 chromatography separation, and mass spectrometry detection. Thus, removing the detergent by protein precipitation with organic solvents, dialysis, gel filtration, or the recently developed filter-aided sample preparation (FASP)6 is necessary before protein digestion and MS detection. Unfortunately, the process of detergent removal usually induces significant protein sample loss due to the buffer exchange and redissolving steps, which compromises the sensitivity, reliability, and accuracy of proteomic analysis especially for the post translational modifications (PTMs) analysis of a minute amount of protein samples.3,9 N-Dodecyl β-D-maltoside (DDM) is a nonionic detergent composed of a lauryl hydrophobic chain and a maltose hydrophilic part. It is a mild detergent with the ability to maintain the enzyme activity,10 purify membrane proteins,11,12 and solubilize hydrophobic proteins without changing their structures.13−17 Here, we demonstrated the lysis buffer with DDM and urea exhibited comparable performance to the © 2015 American Chemical Society Received: December 18, 2014 Accepted: February 3, 2015 Published: February 3, 2015 2054 DOI: 10.1021/ac504700t Anal. Chem. 2015, 87, 2054−2057 Letter Analytical Chemistry seven different lysis buffers. Briefly, TEAB buffer, TEAB buffer containing 8 M urea, TEAB buffer containing 8 M urea and 1% detergent (Triton X-100, RapiGest SF or SDS), TEAB buffer containing 4% SDS,1,6 and TEAB buffer containing 8 M urea and 1% DDM were utilized. Finally, the protein concentration of each lysate was 2.6 ± 0.3 (TEAB buffer), 4.9 ± 0.2 (8 M urea buffer), 8.6 ± 1.0 (8 M urea buffer with 1% Triton X-100), 9.0 ± 1.2 (8 M urea buffer with 1% RapiGest SF), 9.1 ± 0.7 (8 M urea buffer with 1% SDS), 8.8 ± 0.3 (4% SDS buffer), and 8.9 ± 0.4 (8 M urea buffer with 1% DDM) mg/mL, respectively (Figure 1 and Supporting Information, Table S2). Obviously, Table 1. Sequence Coverages and Numbers of Unique Peptides Identified from 1 μg of BSA Digests with Different Concentration of DDM DDM concn (mg/mL) sequence coverage (%) unique peptides 0 0.1 0.2 0.5 1 2 5 10 77.3 78.3 77.3 78.7 76.4 78.4 74.0 74.6 55 52 51 52 53 51 50 52 the relative standard deviation (RSD) for peptides retention time were all <2% (Supporting Information, Table S3). Further, it was observed the DDM was eluted from the LC column only at a high concentration (80%) of ACN (Supporting Information, Figure S2). This can be attributed to the DDM that is retained onto the head of the C18 LC column during the peptide separation process as the main peptides separation range is 0−35% ACN in the RP binary gradient separation. On the other hand, the peak intensity of DDM was also relatively low even if 10 μg DDM was loaded, due to it is nonionic detergent and is hard to be ionized. Then, identical protein samples extracted from HeLa cells were also digested in the presence of DDM with a different concentration (Supporting Information, Experimental Section), and similar results were also obtained among different conditions (Supporting Information, Table S4, Figures S3 and S4). Moreover, the good compatibility of DDM with other widely used enzymes, such as Glu-C, Lys-C (Supporting Information, Table S5) and PNGase F (Supporting Information, Figure S5), was also demonstrated. Therefore, the above results demonstrated that DDM has little influence on the enzyme activity, LC separation, and MS detection and is compatible with proteomic analysis without any detergent removal steps in sample preparation procedures. We also compared the performance of proteomic analysis by using the DDM strategy as described above with different protein sample preparation strategies using different detergent (Triton X-100, RapiGest SF, and SDS) followed with relevant detergent removing steps (precipitation, acidification, and FASP) (Supporting Information, Experimental Section). Finally, it was observed that the sample preparation with DDM extraction exhibited better performance than those methods (Precipitation, RapiGest, and FASP) with conventional detergent (Triton X-100, RapiGest SF, and SDS) extraction followed with relevant detergent removal procedures (precipitation, acidification, and FASP) in proteomic analysis of 15 μg of mouse brain tissue, and 340 (32%), 126 (10%), and 178 (15%) more proteins were identified (Supporting Information, Tables S6 and S7). Protein glycosylation is one of the most important posttranslational modifications and plays a crucial role in a series of physiological and pathological processes, such as cell−cell recognition and communication, protein folding, stabilization, and translocation.19,20 Aberrant N-glycosylation on some important proteins is highly associated with various types of cancers.21,22 N-glycosylated proteins are also important for disease diagnosis, prognosis, and therapeutic response to drugs.23 Various methods have been developed for in-depth characterization of the N-glycoproteome, such as hydrazide Figure 1. Protein extraction efficiency by using lysis buffers containing different detergents for mouse brain tissues and the protein recovery after detergent removal. (Precipitation method was used for 1% triton X-100/8 M urea buffer and 1% SDS/8 M urea buffer; FASP method was used for 4% SDS buffer). the protein extraction performance of the buffer with DDM and urea was comparable to the buffers containing urea and other commonly used detergents (Triton X-100, RapiGest SF, and SDS), and all of them were much better than urea buffer or merely TEAB buffer without detergent. Furthermore, we also quantitatively investigated the protein loss during the detergent removal by precipitation and FASP for the protein samples extracted from 1 mg of tissue. After detergent removal, the protein concentration of each lysate was 4.8 ± 0.4 (1% Triton X-100/8 M urea buffer with precipitation), 3.7 ± 0.4 (1% SDS/ 8 M urea buffer with precipitation), and 5.1 ± 0.1 (4% SDS buffer with FASP) mg/mL (Figure 1 and Supporting Information, Table S2), which demonstrated about 40−60% protein samples were lost during the detergents removal steps. Then, we investigated whether DDM is compatible with trypsin activity and MS detection. The standard BSA proteins were digested with trypsin and analyzed by LC−MS/MS in the presence of a different concentration of DDM (Supporting Information, Experimental Section). After database searching, it was observed that the numbers of unique peptides identified and protein sequence coverage were almost not changed along with the increase of DDM concentration from 0 to 1% (equivalent to 10 mg/mL) (Table 1). The extracted-ion current (XIC) of 14 random selected BSA peptides were checked and marked within the base peak chromatograms with different DDM concentrations (Figure 2). Obviously, the peak intensity, peptides elution profiles, and LC separation windows of these tryptic BSA peptides were all highly similar to each other and 2055 DOI: 10.1021/ac504700t Anal. Chem. 2015, 87, 2054−2057 Letter Analytical Chemistry Figure 2. Base peak chromatograms of LC−MS/MS analysis of 1 μg of BSA digest with the presence of 0 (A), 0.1 (B), 1.0 (C), and 10 (D) mg/mL DDM. (Peptide peaks labeled by “∗”: (1) HLVDEPQNLIK, (2) KVPQVSTPTLVEVSR, (3) RHPEYAVSVLLR, (4) RPCFSALTPDETYVPK, (5) KQTALVELLK, (6) SLHTLFGDELCK, (7) LVNELTEFAK, (8) SHCIAEVEKDAIPENLPPLTADFAEDKDVCK, (9) LFTFHADICTLPDTEK, (10) QTALVELLK, (11) LGEYGFQNALIVR, (12) TVMENFVAFVDK, (13) MPCTEDYLSLILNR, and (14) DAFLGSFLYEYSR). Table 2. Numbers of Identified N-Glycosylation Sites and N-Glycoproteins by Using Different Protein Sample Preparation Strategies experiment 1 experiment 2 total sample preparation glyco-proteins glycosylation sites glyco-proteins glycosylation sites glycosylation sites glyco-proteins urea extraction precipitation method DDM extraction 72 91 188 55 66 123 111 96 194 78 68 125 134 130 281 95 88 170 chemistry for human plasma and urine,24,25 filter aided sample preparation for mouse plasma and other organs,23 and so on. However, usually large amount of staring material is required for these methods, which limits their application for analysis of minute amount of samples, such as clinical samples. As most of the N-glycosylated proteins are hydrophobic transmembrane proteins, we further investigated if the protein extraction buffer with DDM can improve the performance of N-glycoproteomic analysis for minute protein samples from mouse brain tissue. The reversed phase-hydrophilic (RP-HILIC) capillary-based glycoproteomic reactor18 was utilized for N-glycoproteomic sample preparation in this study. Finally, 281 N-glycosylation sites from 170 glycoproteins were characterized from protein samples extracted from 50 μg of mouse brain tissue (Table 2 and Supporting Information, Table S8). Comparing with the sample preparation without any detergent and with strong detergent followed with precipitation purification, the numbers of identified N-glycosylation sites and glycoproteins were increased 110%, 116% and 77%, 93%, respectively (Table 2), and the sample preparation with DDM extraction covered most parts of the glycoproteins identified by the other two methods (69% and 84%, as shown in Figure 3). As the hydrophobic part of the RapiGest SF will precipitate and block the capillary column during acidification for detergent removal and part of the protein will be retained onto the filter membrane and cannot be loaded onto the glycoproteomic reactors for FASP strategy, the RapiGest SF and FASP strategies were not compatible with the RP-HILIC capillary-based glycoproteomic Figure 3. Venn diagram showing the overlap of the glycoproteins identified by three different sample preparation strategies. (In the parentheses were the average Gravy values of proteins in each region). reactor. After the transmembrane helices prediction (TMHMM Server v. 2.0, http://www.cbs.dtu.dk/services/TMHMM), 117 glycoproteins (69%) obtained by protein extraction with DDM were predicted to have at least one transmembrane helices, which was much better than the other two sample preparation methods with 53 (56%) and 58 (66%) transmembrane glycoproteins (Supporting Information, Table S9). Gene Ontology (GO) analysis was further performed to identify the cellular location of the identified glycoproteins. For the DDM strategy, 119 proteins (70%) were membrane proteins, which was also better than the urea (61 (64%) membrane proteins) and precipitation (63 (72%) membrane proteins) strategies, respectively. These results were consistent with the protein transmembrane analysis and further indicated the better 2056 DOI: 10.1021/ac504700t Anal. Chem. 2015, 87, 2054−2057 Analytical Chemistry ■ performance of sample preparation method based on DDM on the analysis of membrane glycoproteins. After calculation of Gravy value of the glycoproteins identified by different sample preparation approaches, it was observed the glycoproteins identified by DDM extraction only were the most hydrophobic ones with an average Gravy value of −0.21 (Figure 3). This could be attributed to the low hydrophobic protein extraction efficiency for protein extraction without any detergent and the significant sample loss during precipitation and the redissolving process for protein precipitation. Therefore, the advantage of DDM in protein extraction for N-glycoproteomic analysis of minute protein sample is more obvious than in common proteomic analysis due to most of the glycoproteins are hydrophobic transmembrane proteins, which are hard to be redissolved into protein digestion buffer after detergent removal by protein precipitation. Protein precipitation or ultrafiltration is unavoidable to eliminate the interference of detergents on enzyme digestion in conventional protein sample preparation approaches. However, sample loss is unavoidable during this detergent removal procedure, and the purified protein pellet is also hard to fully redissolve into the aqueous solution without detergent. This type of protein sample loss greatly compromises the performance of proteomic analysis, especially for the PTMs analyses of minute protein samples. DDM, a mild detergent, compatible with trypsin digestion and LC−MS/MS analysis, was successfully applied in the protein sample extraction in this study. We observed the protein extraction performance with 1% DDM and 8 M urea was comparable to those with 4% SDS or 8 M urea with conventional detergents (Triton X-100, RapiGest SF). After combining with the capillary-based glycoproteomic reactor, 281 N-glycosylation sites were successfully characterized from 50 μg of mouse brain tissue, which was 110% and 116% higher than those methods with urea extraction and strong detergent extraction followed with protein precipitation, respectively. Thus, we believe DDM has great potential in PTMs analysis of minute amount of protein samples. ■ REFERENCES (1) Maurer, M.; Mueller, A. C.; Wagner, C.; Huber, M. L.; Rudashevskaya, E. L.; Wagner, S. N.; Bennett, K. L. J. Proteome Res. 2013, 12, 1040−1048. (2) Switzar, L.; van Angeren, J.; Pinkse, M.; Kool, J.; Niessen, W. M. A. Proteomics 2013, 13, 2980−2983. (3) Ning, Z. B.; Seebun, D.; Hawley, B.; Chiang, C. K.; Figeys, D. J. Proteome Res. 2013, 12, 1512−1519. (4) Sun, L. L.; Tao, D. Y.; Han, B.; Ma, J. F.; Zhu, G. J.; Liang, Z.; Shan, Y. C.; Zhang, L. H.; Zhang, Y. K. Anal. Bioanal. Chem. 2011, 399, 3387−3397. (5) Chen, E. I.; McClatchy, D.; Park, S. K.; Yates, J. R. Anal. Chem. 2008, 80, 8694−8701. (6) Wiśniewski, J. R.; Zougman, A.; Nagaraj, N.; Mann, M. Nat. Meth. 2009, 6, 359−362. (7) Yu, Y. Q.; Gilar, M.; Lee, P. J.; Bouvier, E. S. P.; Gebler, J. C. Anal. Chem. 2003, 75, 6023−6028. (8) Chen, E. I.; Cociorva, D.; Norris, J. L.; Yates, J. R. J. Proteome Res. 2007, 6, 2529−2538. (9) Wang, N.; Xu, M.; Wang, P.; Li, L. Anal. Chem. 2010, 82, 2262− 2271. (10) Rosevear, P.; Vanaken, T.; Baxter, J.; Fergusonmiller, S. Biochemistry 1980, 19, 4108−4115. (11) Loo, T. W.; Clarke, D. M. J. Biol. Chem. 1995, 270, 21449− 21452. (12) Hazarda, A.; Montemagno, C. Arch. Biochem. Biophys. 2002, 407, 117−124. (13) Whitelegge, J. P.; le Coutre, J.; Lee, J. C.; Engel, C. K.; Prive, G. G.; Faull, K. F.; Kaback, H. R. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 10695−10698. (14) le Coutre, J.; Whitelegge, J. P.; Gross, A.; Turk, E.; Wright, E. M.; Kaback, H. R.; Faull, K. F. Biochemistry 2000, 39, 4237−4242. (15) Morgner, N.; Kleinschroth, T.; Barth, H. D.; Ludwig, B.; Brutschy, B. J. Am. Soc. Mass Spectrom. 2007, 18, 1429−1438. (16) Aivaliotis, M.; Karas, M.; Tsiotis, G. J. Proteome Res. 2007, 6, 1048−1058. (17) Lee, S. C.; Bennett, B. C.; Hong, W. X.; Fu, Y.; Baker, K. A.; Marcoux, J.; Robinson, C. V.; Ward, A. B.; Halpert, J. R.; Stevens, R. C.; Stout, C. D.; Yeager, M. J.; Zhang, Q. H. Proc. Natl. Acad. Sci. U.S.A. 2013, 110, E1203−E1211. (18) Liu, J.; Wang, F. J.; Lin, H.; Zhu, J.; Bian, Y. Y.; Cheng, K.; Zou, H. F. Anal. Chem. 2013, 85, 2847−2852. (19) Helenius, A.; Aebi, M. Science 2001, 291 (5512), 2364−2369. (20) Lowe, J. B. Cell 2001, 104 (6), 809−812. (21) Kobata, A.; Amano, J. Immunol. J. Cell Biol. 2005, 83 (4), 429− 439. (22) Woods, R. J.; Edge, C. J.; Dwek, R. A. Nat. Struct. Biol. 1994, 1 (8), 499−501. (23) Zielinska, D. F.; Gnad, F.; Wisniewski, J. R.; Mann, M. Cell 2010, 141 (5), 897−907. (24) Ishihara, T.; Fukuda, I.; Morita, A.; Takinami, Y.; Okamoto, H.; Nishimura, S. I.; Numata, Y. J. Proteomics 2011, 74 (10), 2159−2168. (25) Halim, A.; Nilsson, J.; Ruetschi, U.; Hesse, C.; Larson, G. Mol. Cell. Proteomics 2012, 11 (4), M111.013649. ASSOCIATED CONTENT S Supporting Information * Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. ■ Letter AUTHOR INFORMATION Corresponding Authors *Phone: +86-411-84379576. Fax: +86-411-84379620. E-mail: [email protected]. *Phone: +86-411-84379610. Fax: +86-411-84379620. E-mail: [email protected]. Notes The authors declare no competing financial interest. ■ ACKNOWLEDGMENTS The authors greatly appreciate Dr. X. Liang and Dr. Z. Guo for providing the HILIC materials as a gift. Financial support is gratefully acknowledged for the China State Key Basic Research Program Grant (Grants 2013CB911203 and 2012CB910601), the financial support from the Creative Research Group Project by NSFC (Grant 21321064), and the National Natural Science Foundation of China (Grants 21235006, 21305139, and 81161120540). 2057 DOI: 10.1021/ac504700t Anal. Chem. 2015, 87, 2054−2057
© Copyright 2026 Paperzz