Integrated Device for Online Sample Buffer Exchange, Protein

Feb 12, 2010 - Therefore, an online sample buffer exchange device is indispen- sable to establish RPLC ... Graduate School of Chinese Academy of Scien...
0 downloads 0 Views 780KB Size
Anal. Chem. 2010, 82, 2574–2579

Integrated Device for Online Sample Buffer Exchange, Protein Enrichment, and Digestion Liangliang Sun,†,‡ Junfeng Ma,†,‡ Xiaoqiang Qiao,†,‡ Yu Liang,†,‡ Guijie Zhu,† Yichu Shan,† Zhen Liang,† Lihua Zhang,*,† and Yukui Zhang† National Chromatographic R. & A. Center, Key Laboratory of Separation Science for Analytical Chemistry, Dalian Institute of Chemical Physics, Chinese Academy of Science, Dalian 116023, China, and Graduate School of Chinese Academy of Sciences, Beijing 100039, China An integrated sample treatment device, composed of a membrane interface and a monolithic hybrid silica based immobilized enzymatic reactor (IMER), was developed for the simultaneous sample buffer exchange, protein enrichment, and online digestion, by which for the sample buffer, the acetonitrile content was reduced to ∼1/10 of the initial one, and the pH value was adjusted from ∼3.0 to ∼8.0, compatible for online trypsin digestion. Furthermore, the signal intensity of myoglobin digests was improved by over 10 times. Such an integrated device was successfully applied to the online treatment of three protein eluates obtained by reverse-phase liquid chromatography (RPLC) separation, followed by further protein digest analysis with microreverse-phase liquid chromatography-electrospray ionization-tandem mass spectrometry (µRPLC-ESI-MS/MS). The experimental results showed that the performance of such an integrated sample treatment device was comparable to that of the traditional offline sample treatment method, including lyophilization and in-solution digestion. However, the consumed time was reduced to 1/192. All these results demonstrate that such an integrated sample treatment device could be further online coupled with protein separation, peptide separation, and identification, to achieve high-throughput proteome analysis. Because of the superior analysis throughput and detection sensitivity, the bottom-up or shotgun approach is widely used in proteome study, by which proteomes are first digested into peptides, followed by multidimensional high-performance liquid chromatography (HPLC) separation, tandem mass spectrometry (MS/MS) identification, and data mining.1-8 However, not only * To whom correspondence should be addressed. E-mail: [email protected]. Phone: +86-411-84379720. Fax: +86-411-84379560. † Dalian Institute of Chemical Physics, Chinese Academy of Science. ‡ Graduate School of Chinese Academy of Sciences. (1) Aebersold, R.; Mann, M. Nature 2003, 422, 198. (2) Chait, B. T. Science 2006, 314, 65–66. (3) Link, A. J.; Eng, J.; Schieltz, D. M.; Carmack, E.; Mize, G. J.; Morris, D. R.; Garvik, B. M.; Yates, J. R., 3rd. Nat. Biotechnol. 1999, 17, 676–682. (4) Yates, J. R., 3rd; Carmack, E.; Hays, L.; Link, A. J.; Eng, J. K. Methods Mol. Biol. 1999, 112, 553–569. (5) Wang, F. J.; Dong, J.; Ye, M. L.; Jiang, X. G.; Wu, R. A.; Zou, H. F. J. Proteome Res. 2008, 7, 306–310. (6) Wang, F. J.; Dong, J.; Jiang, X. G.; Ye, M. L.; Zou, H. F. Anal. Chem. 2007, 79, 6599–6606.

2574

Analytical Chemistry, Vol. 82, No. 6, March 15, 2010

is the simultaneous separation of thousands peptides extremely challenging for HPLC, but also peptides from high-abundance proteins might undermine the detection of those from lowabundance ones. Therefore, protein fractionation before digestion is essential to obtain detailed information about protein sequence and post-translational modification. At present, several techniques have been developed for protein fractionation, such as two-dimensional electrophoresis (2-DE),9,10 in solution-isoelectric focusing (IEF),11 and multidimensional HPLC.12-15 However, protein fractionation, digestion, and peptide separation are usually performed offline, resulting in unavoidable sample loss and rather long analysis time. With the development of immobilized enzymatic reactor (IMER),16-20 online protein fractionation and digestion, followed by peptide separation and identification, becomes possible. Although RPLC is effective for protein prefractionation, the eluates are usually of high organic modifier concentration and low pH value, incompatible with online protein digestion by IMER. Therefore, an online sample buffer exchange device is indispensable to establish RPLC based integrated proteome analysis platforms. Schriemer et al.21,22 demonstrated online tryptic digestion in conjunction with protein separation by RPLC via a tee joint. (7) Motoyama, A.; Venable, J. D.; Ruse, C. I.; Yates, J. R. Anal. Chem. 2006, 78, 5109–5118. (8) Bailey, A. O.; Miller, T. M.; Dong, M. Q.; Vande Velde, C.; Cleveland, D. W.; Yates, J. R. Anal. Chem. 2007, 79 (16), 6410–6418. (9) Klose, J.; Kobalz, U. Electrophoresis 1995, 16, 1034–1059. (10) Gazzana, G.; Borlak, J. J. Proteome Res. 2007, 6 (8), 3143–3151. (11) Myung, J. K.; Lubec, G. J. Proteome Res. 2006, 5, 1267–1275. (12) Millea, K. M.; Krull, I. S.; Cohen, S. A.; Gebler, J. C.; Berger, S. J. J. Proteome Res. 2006, 5, 135–146. (13) Gao, M. X.; Zhang, J.; Deng, C. H.; Yang, P. Y.; Zhang, X. M. J. Proteome Res. 2006, 5, 2853–2860. (14) Assiddiq, B. F.; Snijders, A. P. L.; Chong, P. K.; Wright, P. C.; Dickman, M. J. J. Proteome Res. 2008, 7, 2253–2261. (15) Dowell, J. A.; Frost, D. C.; Zhang, J.; Li, L. J. Anal. Chem. 2008, 80, 6715– 6723. (16) Duan, J. C.; Liang, Z.; Yang, C.; Zhang, J.; Zhang, L. H.; Zhang, W. B.; Zhang, Y. K. Proteomics 2006, 6, 412–419. (17) Ma, J. F.; Liang, Z.; Qiao, X. Q.; Deng, Q. L.; Tao, D. Y.; Zhang, L. H.; Zhang, Y. K. Anal. Chem. 2008, 80, 2949–2956. (18) Ye, M. L.; Hu, S.; Schoenherr, R. M.; Dovichi, N. J. Electrophoresis 2004, 25, 1319–1326. (19) Palm, A. K.; Novotny, M. V. Rapid Commun. Mass Spectrom. 2004, 18, 1374–1382. (20) Deng, Y. H.; Deng, C. H.; Qi, D. W.; Liu, C.; Liu, J.; Zhang, X. M.; Zhao, D. Y. Adv. Mater. 2009, 21, 1–6. (21) Slysz, G. W.; Schriemer, D. C. Anal. Chem. 2005, 77, 1572–1579. (22) Slysz, G. W.; Lewis, D. F.; Schriemer, D. C. J. Proteome Res. 2006, 5, 1959– 1966. 10.1021/ac902835p  2010 American Chemical Society Published on Web 02/12/2010

Through the dynamic titration of protein eluates, the compatibility of eluates with IMER was improved. However, the sample dilution was unavoidable, resulting in decreased detection sensitivity for protein digests. Membrane materials have been widely applied in bioanalysis, especially for large molecules analysis, for desalting,23,24 concentration,25-27 or buffer exchange in 2D-CE.28 However, to the best of our knowledge, its application in the simultaneous buffer exchange and protein concentration for RPLC eluates to achieve online hyphenation with IMER for digestion has not been reported. In this technical note, an integrated sample treatment device, composed of a hollow fiber membrane interface and a monolithic hybrid silica based IMER, was developed. With the dialysis capacity of the membrane, the sample buffer could be online exchanged to be compatible with tryptic digestion. Simultaneously, with the backpressure generated by IMER during continuous injection, proteins could be online concentrated by buffer splitting and sample stacking, followed by online digestion with IMER with fast speed. Such an integrated sample treatment device was successfully applied into the analysis of proteins in three RPLC eluates and showed promise in high-throughput protein analysis. EXPERIMENTAL SECTION Reagents and Materials. Trypsin (bovine pancreas), cytochrome c (bovine heart), and myoglobin (equine skeletal muscle) were ordered from Sigma-Aldrich (St. Louis, MO). Bovine serum albumin (BSA, bovine serum) was obtained from Sino-American Biotec (Luoyang, China). Tetraethoxysilane (TEOS, 95%) and 3-aminopropyltriethoxysilane (APTES, 99%) were purchased from Acros Organics (Geel, Belgium) and dried by solid-phase extraction (SPE) columns packed with anhydrous silica (Sipore Chrom Corp., Dalian, China). Dithiothreitol (DTT) and iodoacetamide (IAA) were from Acros (Morris Plains, NJ). Urea was purchased from Shenyang Chemical Reagents (Shenyang, China). Acetonitrile (HPLC grade) was ordered from Merck (Darmstadt, Germany). Deionized water purified by a Milli-Q system (Millipore, Milford, MA) was used in all experiments. Fused-silica capillaries (100 µm i.d. × 375 µm o.d.) were obtained from Sino Sumtech (Handan, China). The hollow cellulose acetate fiber membrane (molecular weight cutoff (MWCO), 3 000 Da, 200 µm i.d. × 220 µm o.d.) was taken from a capillary dialyzer (GFS Plus 12, Gambro Dialysatoren GmbH, Hechingen, Germany). The reversed phase column (C8, 4.6 mm i.d. × 250 mm, 5 µm, 300 Å) for protein separation was ordered from Sipore Chrom Corp. (Dalian, China). Instrumentation. A precise syringe pump (Baoding Longer Precision Pump Company, Baoding, China) was used to push samples through the integrated sample treatment device. A paradigm GM4 µHPLC system (Michrom Bioresources Inc., Auburn, CA) coupled with a LCQDUO quadrupole ion trap mass (23) Xiang, F.; Lin, Y. H.; Wen, J.; Matson, D. W.; Smith, R. D. Anal. Chem. 1999, 71, 1485–1490. (24) Sun, L. L.; Duan, J. C.; Tao, D. Y.; Liang, Z.; Zhang, W. B.; Zhang, L. H.; Zhang, Y. K. Rapid Commun. Mass Spectrom. 2008, 22, 2391–2397. (25) Zhang, R.; Hjerte´n, S. Anal. Chem. 1997, 69, 1585–1592. (26) Song, S.; Singh, A. K.; Kirby, B. J. Anal. Chem. 2004, 76, 4589–4592. (27) Zhou, Y.; Shen, H. L.; Yi, T.; Wen, D. W.; Pang, N. N.; Liao, J.; Liu, H. W. Anal. Chem. 2008, 80, 8920–8929. (28) Yang, C.; Liu, H. C.; Yang, Q.; Zhang, L. Y.; Zhang, W. B.; Zhang, Y. K. Anal. Chem. 2003, 75, 215–218.

Figure 1. Schematic of an integrated sample treatment device for buffer exchange, protein concentration, and online digestion: (a) centrifugal tube, (b) hollow fiber membrane, (c) capillaries coated with linear polyacrylamide, (d) IMER, (e) Teflon tube.

spectrometer (LCQ-IT MS, Thermo Fisher, San Jose, CA) was used for protein identification. A SpeedVac (Thermo Fisher, San Jose, CA) was used to lyophilize samples. Sample Preparation. Cytochrome c, myoglobin, and BSA were, respectively, dissolved in 8 M urea and then denatured for 1 h at 56 °C. Subsequently, they were reduced in 8 mM DTT for 1 h at 56 °C. After cooled to room temperature, the cysteines were alkylated in 20 mM IAA for 40 min in the dark. Myoglobin, 0.1 mg/mL and 7.7 µg/mL, dissolved in 50 mM NH4HCO3 containing 0.8 M urea, and 50% acetonitrile (ACN) containing 0.8 M urea and 0.1% TFA, respectively, were prepared to evaluate the interface performance. Unless specified otherwise, the percentage in this tecnical note represented volume ratio. After the proteins were denatured and alkylated, a mixture of cytochrome c (0.5 mg/mL), myoglobin (0.25 mg/ mL), and BSA (0.25 mg/mL) was prepared for RPLC separation. Construction of Integrated Device. As shown in Figure 1, the integrated sample treatment device for buffer exchange, protein concentration, and digestion was composed of a membrane interface and an IMER, combined by a Teflon tube. The membrane interface prepared was referred to in our previous work.24 In brief, first one end of each connecting capillary (100 µm i.d. × 375 µm o.d., coated with linear polyacrylamide) was etched by 40% HF until the capillary could be inserted into the 7 cm-long hollow fiber membrane. Then they were glued together via epoxy glue, threaded through a centrifugal tube (4 mL), and fixed with two pieces of Teflon tubes. Finally, two holes were drilled on the two opposite sides of the centrifugal tube to introduce exchange buffer. Unless specified otherwise, the exchange buffer was 50 mM NH4HCO3/NH3 · H2O (pH 8.5). An IMER (100 µm i.d. × 375 µm o.d.) was prepared by the immobilization of trypsin on hybrid silica monolith according to our previous work.17 Briefly, TEOS and APTES were polymerized to form the monolithic matrix. Then, the support was activated with glutaraldehyde, followed by trypsin immobilization via covalent bonding. In the following experiments, unless specified otherwise, the IMER length was 5.3 cm, and protein digestion was carried out at room temperature. Buffer Exchange by Membrane Interface. A series of 20 mM NH4HCO3 solutions containing ACN with concentration ranging from 0 to 80% was prepared and filled into a 20 cmlong capillary to measure the current after 5 kV was applied, to obtain the curve representing the relationship between electric current and ACN concentration. Analytical Chemistry, Vol. 82, No. 6, March 15, 2010

2575

To evaluate the ACN removal capacity, pure ACN was pumped through the membrane interface at a flow rate of 5 µL/min. Simultaneously, deionized water was applied as exchange buffer to generate force convection, with flow rate varied from 1.3 to 3 mL/min. The eluent was collected and dissolved in 20 mM NH4HCO3 for electric current measurement. To evaluate the pH adjustment capacity, 0.1% TFA was pumped through the membrane interface at the flow rate of 5 µL/min, and 50 mM NH4HCO3/NH3 · H2O (pH 8.5) was chosen as the exchange buffer, with a flow rate of 2 mL/min. The pH value of the eluent was measured with the precise test paper. To further evaluate the compatibility of the exchanged sample with online IMER digestion, the following three samples were pumped through the IMER, and the measured flow rate of the eluates was 360 nL/min. (I) Myoglobin (0.1 mg/mL) dissolved in 50% ACN, 0.1% TFA, and 0.8 M urea (pH ∼3.0); (II) the same sample as in part I but treated with the membrane interface for buffer exchange, with the flow rate of sample introduction and exchange buffer as 5 µL/min and 2 mL/min; (III) myoglobin (0.1 mg/mL) dissolved in 50 mM NH4HCO3 and 0.8 M urea (pH ∼8.0). The digests were collected and further analyzed by microreverse-phase liquid chromatography-electrospray ionization-tandem mass spectrometry (µRPLC-ESI-MS/MS). Protein Online Concentration and Digestion. Myoglobin (7.7 µg/mL) dissolved in 50 mM NH4HCO3 and 0.8 M urea (pH ∼8.0) was pumped into the integrated device at a flow rate of 5 µL/min for 30 min, and the flow rate of exchange buffer was 3 mL/min. Because of the backpressure generated by the IMER, the flow rate of the eluates was 360 nL/min. The collected protein digests were further analyzed by µRPLC-ESIMS/MS. For comparison, myoglobin (7.7 µg/mL) dissolved in 50 mM NH4HCO3 and 0.8 M urea (pH ∼8.0) was pumped through the IMER directly with the same flow rate, followed by offline µRPLC-ESI-MS/MS analysis. Online Buffer Exchange, Protein Enrichment, and Digestion. In order to investigate the reproducibility of the integrated device for online buffer exchange, protein enrichment, and digestion, 7.7 µg/mL denatured and alkylated myoglobin, dissolved in 50% ACN containing 0.8 M urea and 0.1% TFA, was chosen as the sample. The sample was injected into the integrated device with the inlet flow rate of 5 µL/min, and the flow rate of peptide eluates was about 360 nL/min. The flow rate of the exchange buffer was 2 mL/min. The sample treatment process was performed three times. The collected digested products were analyzed by µRPLC-ESI-MS/MS. The mixture of cytochrome c (0.5 mg/mL), myoglobin (0.25 mg/mL), and BSA (0.25 mg/mL) was injected (20 µL) and separated into three fractions with a C8 column. The collected fraction volume of cytochrome c, myoglobin, and BSA was, respectively, 2, 2, and 1 mL. Therefore, if the sample loss was neglected, the concentration of cytochrome c, myoglobin, and BSA in the collected fractions was calculated to be 5, 2.5, and 5 µg/ mL, respectively. For comparison, such three RPLC fractions were, respectively, treated with two different methods. (I) The samples were injected into the integrated device with the inlet flow rate of 5 µL/min, and the flow rate of the exchange buffer was 2 mL/min. Because of the existence of backpressure generated by the IMER, the flow 2576

Analytical Chemistry, Vol. 82, No. 6, March 15, 2010

rate of peptide eluates was about 0.5 µL/min. (II) With the traditional method, the fractions were lyophilized and then, respectively, resuspended in 200 µL, 200 µL, and 100 µL of 50 mM NH4HCO3, 1/10 of the original eluates volume. The concentration of proteins in the resuspension solutions was thus calculated to be 50 µg/mL (cytochrome c), 25 µg/mL (myoglobin), and 50 µg/mL (BSA), respectively. Then, in-solution digestion was performed by adding trypsin into the resuspension solutions with a substrate-to-enzyme ratio of 30:1 (w/w), followed by incubation at 37 °C for 12 h. Finally, 5 µL of formic acid was added into the solutions to terminate the reaction. All the protein digests treated by the two different methods were further analyzed by µRPLC-ESI-MS/MS under the same experimental conditions. µRPLC-ESI-MS/MS Experiments and Data Analysis. The samples were separated by µHPLC with a homemade C18 column (5 µm, 200 Å pore, 300 µm i.d. × 5 cm) at the flow rate of 5 µL/min, using buffer A (2% ACN containing 0.1% formic acid) and buffer B (98% ACN containing 0.1% formic acid). The gradient was kept at 0% B for 5 min, to 40% B in 40 min, to 80% B in 0.1 min, and then 80% B for 10 min. The sample injection volume was 2 µL. The LCQ instrument was operated at positive ion mode. A spray voltage of 2 kV was employed, and the heated capillary temperature was set to 150 °C. Total ion current chromatograms and mass spectra covering the mass range from m/z 400 to 2000 were recorded with Xcalibur software (version 1.4). MS/MS spectra were acquired at data-dependent acquisition mode with two precursor ions selected from one MS scan. Precursor selection was based on parent ion intensities, and the normalized collision energy for MS/MS scanning was 35%. Protein identification was performed using BioWorks software 3.1 with SEQUEST. MS/MS raw data obtained from µRPLC-ESIMS/MS analysis was used for database search, respectively, from bovine.fasta for cytochrome c and BSA and from equine.fasta for myoglobin. Cysteine residues were searched as a static modification of 57.0215 Da. Peptides were searched using fully tryptic cleavage constraints, and up to two internal cleavages sites were allowed for tryptic digestion. The mass tolerances were 2 Da for parent masses and 1 Da for fragment masses. The peptides were considered as positive identification if Xcorr was higher than 1.9 for a singly charged peptide, 2.2 for a doubly charged peptide, and 3.75 for a triply charged peptide. RESULT AND DISCUSSION Evaluation of Buffer Exchange. With pure ACN as the sample and water as the exchange buffer, the ACN removal capacity of the membrane interface was evaluated. In our experiments, the residual ACN concentration after membrane treatment was calculated according to the curve reflecting the relationship between electric current and ACN concentration, as shown in Figure 2a. Since the removal of ACN was based on the molecular diffusion between sample and exchange buffer, the higher the exchange buffer flow rate, the less the residual ACN content, which was proven by our experimental results shown in Figure 2b. In addition, when the counter flow rate was higher than 2 mL/min, the electric current increased to 7.2 µA, and the residual ACN concentration was obviously decreased to about 12%, compatible with IMER.17 Therefore, in the following experiments, the counter flow rate was set as over 2 mL/min.

Table 1. Database Searching Results of the Digests of Myoglobin with Different Sample Treatment Methodsa position

peptide sequence

MH+

1-17 16-32 31-43 31-48 47-57 63-78 63-79 79-97 118-134 133-146 139-146 145-153

-.GLSDGEWQQVLNVWGK.V K.VEADIAGHGQEVLIR.L R.LFTGHPETLEK.F R.LFTGHPETLEKFDKFK.H K.HLKTEAEMK.A K.HGTVVLTALGGILK.K K.HGTVVLTALGGILKK.K K.GHHEAELKPLAQSHATK.H K.HPGDFGADAQGAMTK.A K.ALELFRNDIAAK.Y R.NDIAAK.Y K.YKELGFQG.peptides matched sequence coverage, %

1817.00 1607.79 1272.43 1938.22 1087.28 1379.67 1507.85 1855.05 1503.62 1361.57 631.70 942.05

A × × ×

B × × ×

C ×

× × × × × × × × × × 9 11 2 64.71 77.12 17.65

× × × × ×

a Digests of 0.1 mg/mL myoglobin dissolved in 50 mM NH4HCO3 and 0.8 M urea directly by the IMER (A); digests of 0.1 mg/mL myoglobin dissolved in 50% ACN containing 0.8 M urea and 0.1% TFA by the IMER with (B) and without (C) buffer exchange by the membrane interface.

Figure 2. Relationship between electric current and ACN concentration (a) and the effect of exchange buffer flow rate on residual ACN concentration after treated by the membrane interface (b).

In addition, the pH adjustment capacity of the membrane interface was evaluated with 0.1% TFA (pH ∼3.0) as the sample. Our experimental results show that with 50 mM NH4HCO3/ NH3 · H2O (pH 8.5) as the exchange buffer, with the flow rate of 2 mL/min, the sample pH could be increased to about 8.0 in 30 s and maintained at pH 8.0 ± 0.5 in 50 min (as shown in Suppl. Figure 1 in the Supporting Information), which meant that by the membrane interface, the pH value of the sample buffer could be effectively exchanged and kept stable within 1 h, compatible with IMER for protein digestion. To further confirm the compatibility of the exchanged sample buffer with IMER, three kinds of myoglobin samples (0.1 mg/ mL) were online digested by IMER, followed by analysis with µRPLC-ESI-MS/MS. As listed in Table 1, for myoglobin dissolved in 50 mM NH4HCO3 and 0.8 M urea (A), 9 matched peptides were identified, with sequence coverage of 64.71% (base peak chromatogram shown in Suppl. Figure 2A in the Supporting Information). For myoglobin dissolved in 50% ACN containing 0.8 M urea and 0.1% TFA, after buffer exchange by the membrane interface (B), 11 matched peptides were identified, with a sequence coverage of 77.12% (base peak chromatogram shown in Suppl. Figure 2B in the Supporting Information), comparable to that with IMER compatible sample buffer. However, for the

direct analysis of myoglobin dissolved in 50% ACN containing 0.8 M urea and 0.1% TFA, without buffer exchange by the membrane interface (C), only 2 matched peptides were identified, with sequence coverage of 17.65%. Furthermore, the peak of myoglobin was also evident in the base peak chromatogram, showing that the digestion was not complete (as illustrated in Suppl. Figure 2C in the Supporting Information). All these results demonstrate that with the membrane interface, samples dissolved in buffer with high ACN concentration and low pH could be effectively adjusted to that compatible with online protein digestion by IMER. Evaluation of Online Protein Enrichment. To evaluate the online protein enrichment capacity of the integrated sample treatment device, myoglobin samples (7.7 µg/mL) were, respectively, digested by IMER without and with online enrichment by the device, followed by analysis with µRPLC-ESI-MS/MS. From the base peak chromatograms shown in Figure 3, it could be seen that after online enrichment, the peak intensity of 2 peptides identified in both cases (marked with an *) was improved by 11.5 times (Suppl. Table 1 in the Supporting Information). Besides, an additional 6 peptides (marked with “M”) were identified after enrichment, resulting in the sequence coverage of myoglobin improved from 18.95% to 60.78% (Suppl. Table 1 in the Supporting Information). These results demonstrate that by the integrated sample treatment device, online protein concentration and digestion could be performed simultaneously, with improved identification capacity for low-concentration proteins. The principle of protein enrichment could be explained by the law of mass conservation. The hollow fiber membrane in the integrated device could be viewed as a “three ports”. V0 and C0 are the flow rate and concentration of proteins introduced into the membrane (before splitting); V1 and C1 are the flow rate and concentration of proteins that flow into the exchange buffer via small pores on the membrane (splitting out); V2 and C2 are the flow rate and concentration of proteins through the membrane. On the basis of the law of mass conservation, eq 1 could be obtained. V0C0 ) V1C1 + V2C2 Analytical Chemistry, Vol. 82, No. 6, March 15, 2010

(1) 2577

Figure 3. Base peak chromatograms of the digests of myoglobin by IMER without (A) and with (B) online concentration by the integrated sample treatment device. Experimental conditions were illustrated in the Experimental Section. Sample, myoglobin (7.7 µg/mL) dissolved in 50 mM NH4HCO3 and 0.8 M urea.

Figure 4. Base peak chromatograms of the digests of myoglobin treated by the integrated device in three consecutive runs. Experimental conditions were illustrated in the Experimental Section. Sample, myoglobin (7.7 µg/mL) dissolved in 50% ACN containing 0.8 M urea and 0.1% TFA.

With the backpressure generated by IMER, at the interface, small molecules, such as ACN and TFA, could be driven out of the membrane to exchange with those in the exchange buffer. However, 2578

Analytical Chemistry, Vol. 82, No. 6, March 15, 2010

since the cutoff molecular weight of the membrane is 3000 Da, proteins could not be split out of the membrane and C1 should be close to zero. Therefore, eq 1 could be simplified to eq 2.

Table 2. Comparison of Database Searching Results of the Digests of Protein Eluates By RPLC Separation after Treated by the Integrated Device (A) and Traditional Method (B)

matched peptide number sequence coverage (%)

A B A B

cytochrome c

BSA

myoglobin

3 4 34.62 35.58

15 12 32.29 22.4

2 2 18.95 17.65

C2 V0 ) C0 V2

(2)

In our experiment, V0 and V2 were 5 µL/min and 360 nL/min, respectively. Therefore, according to eq 2, proteins could be concentrated by over 10 times, in accordance with that for peptide signals identified. Reproducibility for Analysis of Proteins. Denatured and alkylated myoglobin, 7.7 µg/mL, dissolved in 50% ACN containing 0.8 M urea and 0.1% TFA, was treated with the integrated device three times, and the collected digested products were analyzed by µRPLC-ESI-MS/MS. From the base peak chromatograms shown in Figure 4, it could be seen that the peptides profiling in three consecutive runs were of good reproducibility, and 3, 3, and 4 peptides (marked with “M”) were, respectively, identified, with the sequence coverage as 26.14%, 30.07%, and 39.23%. The results demonstrate that the reproducibility of the developed integrated device for protein treatment is good. Analysis of HPLC Eluates. Cytochrome c, BSA, and myoglobin separated by a C8 column (as shown in Suppl. Figure 3 in the Supporting Information) were collected into three fractions and directly applied to the integrated sample treatment device for online buffer exchange, concentration, and digestion, followed by µRPLC-ESI-MS/MS analysis (with base peak chromatograms shown in Suppl. Figure 4A in the Supporting Information). As shown in Table 2, within 5 min online sample treatment, all proteins were efficiently identified by µRPLC-ESI-MS/MS, with sequence coverages, respectively, of 34.62%, 32.29%, and 18.95%. In contrast, the three proteins fractions were also treated by the traditional sample treatment method, including lyophilization

and in-solution digestion, followed by µRPLC-ESI-MS/MS analysis (with base peak chromatograms shown in Suppl. Figure 4B in the Supporting Information). As shown in Table 2, even with sample buffer volume decreased to 1/10 during lyophilization compared to that before treatment, resulting in 10 times enrichment on protein concentration, the obtained sequence coverage of proteins was just comparable to that obtained for online treated proteins. However, the total sample treatment time was 960 min, which was 192 times longer than that required for the integrated device (5 min). All the above-mentioned results demonstrate that such an integrated sample treatment device could be successfully applied for the high-throughput analysis of HPLC eluates. CONCLUSIONS An integrated sample treatment device for online sample buffer exchange, protein enrichment, and digestion was developed, by which the incompatible protein buffer for digestion could be exchanged effectively with a membrane interface, and proteins could be online concentrated and digested with an IMER within 5 min. Compared with the traditional offline method, the total sample treatment time was shortened by near 200 times. All these results demonstrate that such an integrated device could be further online coupled with RPLC based protein fractionation, and µRPLC-ESI-MS/MS for peptide separation and identification to achieve high-resolution, high-sensitivity, and high-throughput proteome analysis. ACKNOWLEDGMENT The authors are grateful for the financial support from the National Nature Science Foundation (Grants 20935004 and 20775080), National Basic Research Program of China (Grant 2007CB914100), and Knowledge Innovation Program of Chinese Academy of Sciences (Grant KJCX2YW.H09). SUPPORTING INFORMATION AVAILABLE Additional information as noted in text. This material is available free of charge via the Internet at http://pubs.acs.org. Received for review December 14, 2009. Accepted January 30, 2010. AC902835P

Analytical Chemistry, Vol. 82, No. 6, March 15, 2010

2579