Development of a Multiplexed Microfluidic Proteomic Reactor and Its

ARTICLE pubs.acs.org/ac

Development of a Multiplexed Microfluidic Proteomic Reactor and Its Application for Studying Protein
Protein Interactions Ruijun Tian,†,^,# Xuyen Dai Hoa,‡ Jean-Philippe Lambert,†,^,# John Paul Pezacki,§,^,|| Teodor Veres,*,‡,z and Daniel Figeys*,†,^,|| †

Ottawa Institute of Systems Biology and ‡ Industrial Materials Institute, National Research Council, Boucherville, QC, Canada J4B 6Y4 Steacie Institute for Molecular Sciences, National Research Council of Canada, 100 Sussex Drive, Ottawa, Canada K1A 0R6 ^ Department of Biochemistry, Microbiology and Immunology, Faculty of Medicine, University of Ottawa, 451 Smyth Road, Ottawa, ON, Canada K1H 8M5 Department of Chemistry, Faculty of Science, University of Ottawa, 10 Maria Curie, Ottawa, ON, Canada K1N 6N5 z Department of Biomedical Engineering, McGill University, Montreal, QC, Canada H3A 2B4

)

§

bS Supporting Information ABSTRACT: Mass spectrometry-based proteomics techniques have been very successful for the identification and study of protein
protein interactions. Typically, immunopurification of protein complexes is conducted, followed by protein separation by gel electrophoresis and in-gel protein digestion, and finally, mass spectrometry is performed to identify the interacting partners. However, the manual processing of the samples is time-consuming and error-prone. Here, we developed a polymerbased microfluidic proteomic reactor aimed at the parallel analysis of minute amounts of protein samples obtained from immunoprecipitation. The design of the proteomic reactor allows for the simultaneous processing of multiple samples on the same devices. Each proteomic reactor on the device consists of SCX beads packed and restricted into a 1 cm microchannel by two integrated pillar frits. The device is fabricated using a combination of low-cost hard cyclic olefin copolymer thermoplastic and elastomeric thermoplastic materials (styrene/(ethylene/butylenes)/styrene) using rapid hot-embossing replication techniques with a polymer-based stamp. Three immunopurified protein samples are simultaneously captured, reduced, alkylated, and digested on the device within 2
3 h instead of the days required for the conventional protein
protein interaction studies. The limit of detection of the microfluidic proteomic reactor was shown to be lower than 2 ng of protein. Furthermore, the application of the microfluidic proteomic reactor was demonstrated for the simultaneous processing of the interactome of the histone variant Htz1 in wildtype yeast and in a swr1Δ yeast strain compared to an untagged control using a novel three-channel microfluidic proteomic reactor.

R

ecent advents in mass spectrometry, high-resolution gel electrophoresis, and other techniques have facilitated the high-throughput studies of the protein content of cells and their post-translational modifications.1,2 One of the most successful examples of proteomics has been the study of protein
protein interactions using mass spectrometry. This is typically done by enrichment of a bait protein and its interacting partners using a universal tag attached to the bait protein. The enriched protein complex is then separated by SDS-PAGE, in-gel digested by proteases, and detected by mass spectrometry. This approach greatly facilitates the identification of protein complexes and their related dynamic pathways.3 The systematic applications of this approach in different organisms, including human and yeast, have been reported.4
6 Although the mapping of protein
protein interactions by immunopurification coupled to mass spectrometry can be performed in high-throughput, it still has serious challenges. In particular, the reliance on gel electrophoresis for the separation of the immunopurified proteins leads to extensive effort for in-gel r 2011 American Chemical Society

digestion. In-solution digestion is another conventional approach for processing a large amount of protein samples. However, this approach is not ideal for digesting immunopurified samples which typically contain less than 10 μg of proteins. Instead, fast digestion approaches have been developed for proteomics study mainly based on immobilized enzymes.7
13 Using these approaches, the digestion of the sample can be done within minutes, which greatly improves the efficiency of the whole workflow. However, most of the immobilized enzyme approaches do not integrate other processing steps such as reduction and alkylation of proteins. Recently, we developed a highefficiency proteomic sample processing platform, termed proteomic reactor (PR), based on strong cation exchange (SCX) packed capillary columns.14
18 The proteomic reactor integrates within one simple device the preconcentration of proteins, buffer Received: January 23, 2011 Accepted: April 26, 2011 Published: April 26, 2011 4095

dx.doi.org/10.1021/ac200194d | Anal. Chem. 2011, 83, 4095–4102

Analytical Chemistry exchanges, reduction, alkylation, and digestion of proteins from proteomic samples. On this device, the processing of proteomic samples is finished within 2
3 h. We and others have extensively used the proteomic reactor for protein
protein interaction studies to complement approaches involving SDS-PAGE and in-gel digestion.19,20 With these recent advances for high-efficiency proteomic sample preparation, further minimization and integration of the whole process will greatly facilitate large-scale mass spectrometry-based protein
protein interaction studies. Fortunately, significant efforts over the past decades were focused on the development of proteomics-on-a-chip, an attempt to incorporate the various components necessary for analytical operations onto a single cost-effective platform.21
25 Fabricated by photolithography, micromachining techniques, imprinting, or embossing, various on-chip components have been described: (1) solidphase extraction (SPE) based on, for example, polymeric membranes, microparticles, or monolith structures; (2) fractionalization based on isoelectric focusing (IEF) and electrophoresis; (3) immunoassay and enzymatic assay; (4) integrated mass spectrometry based on nanoelectrospray and electrospray MS.26
29 Here, we present a novel microfluidic proteomic reactor that features multiplexed channels for parallel high-throughput sample handling with facile and low-cost fabrication. This microfabricated proteomic reactor not only outperforms the conventional proteomic reactors in terms of limit of detection but also allows the simultaneous processing of multiple samples on the same devices. In particular, the microfabricated proteomic reactor is utilized for the simultaneous processing of immunopurified protein partners of the yeast protein Htz1.

’ MATERIALS AND METHODS Microfabrication and Device Assembly. Mold Fabrication: SU-8 molds of the multichannel proteomic devices were first fabricated by standard photolithography on Si wafers. We fabricated a mold consisting of the PR chambers, channel, and frits and a mold for the substrate containing alignment grooves for the input/output glass capillaries on Si wafers. The SU-8 molds were then transferred to a soft polymeric mold30 (MD700 working stamps, Solvay) via an intermediate PDMS transfer mold. PDMS replicas of the SU-8 are fabricated as described by others:31 PDMS and curing agent (Sylgard 184, Dow Corning) were mixed in a 10:1 mass ratio, poured onto the SU-8 mold, degassed, and cured at 80 C for 2 h. The working stamp mold was fabricated by applying MD700 polymer onto the PDMS mold with a photocurable agent and exposed to UV light (40 mJ, 15 min). Rapid Hot-Embossing Process: The device layer was embossed in styrene/(ethylene/butylenes)/styrene (SEBS) thermoplastic elastomers (TPEs). The TPE was previously extruded from pellets into continuous film with 2 mm thickness. The PR features were transferred to the TPE film by hot-embossing (EVG520, EVGroup, Austria) at 140 C with 15 KN for 2 min. The substrate containing the alignment groove for the capillary tubes was replicated in a similar process using Zeonor 1060R wafers. The Zeonor 1060R wafers were previously fabricated by injection molding. Assembly with Glass Capillaries: The embossed hard thermoplastic substrate (Zeonor 1060R) and soft elastomer (SEBS-TPE) layer were then assembled with glass capillary tubes (360 μm o.d., 200 μm i.d.) at the inlet and outlet ports. To ensure a leak-free seal and connection, a small 1 μL volume of chlorobenzene was deposited at the capillary/SEBS/Zeonor interface after the

ARTICLE

insertion of the capillary tubes. The chlorobenzene was wicked by capillary force into the interspace between the glass capillary and the SEBS/Zeonor. It partially dissolved the SEBS and Zeonor to make a leak-tight junction without clogging the glass capillary. The assembled devices were then heated to 85 C for 2 h to bond the SEBS to the Zeonor 1060R and evaporate the remaining solvent. Bead-Bed Formation: Once assembled, the proteomic reactor was loaded with 10 μm strong cation exchanger (SCX) beads (Varian Inc.) suspended in distilled water. Ten microliters of the bead solution was injected into each PR channel with capillary column syringe via the loading port, until the bead bed was completely formed. The loading port was then sealed with a piece of Zeonor 1060R and a small amount of chlorobenzene. Yeast Cell Culture and Immunoprecipitation. The yeast cell culture and immunoprecipitation were performed as previously described with minor modifications.32 Briefly, three yeast strains (untagged control, Htz1-TAP, and Htz1-TAP swr1Δ) previously described32 were grown to an OD600 of 1; they were harvested and lysed in lysis buffer (100 mM HEPES pH 8.0, 20 mM magnesium acetate, 10% glycerol (v/v), 10 mM EGTA, 0.1 mM EDTA þ fresh yeast protease inhibitors cocktail (Sigma-Aldrich, USA)) using a coffee grinder with dry ice and sonication. Then, the lysate was mixed with Nonidet P-40 (NP-40) to a final concentration of 0.4% (v/v) and centrifuged mildly at 1800 rpm for 10 min at 4 C. The immunoprecipitation was performed with rabbit IgGcoated Dynabeads. After 3 h of incubation at 4 C, the beads were washed with 100 mM HEPES, pH 7.4, 20 mM magnesium acetate, 10% (v/v) glycerol, 10 mM EGTA, 0.1 mM EDTA, and 0.5% (v/v) Nonidet P-40 three times, and the purified proteins were eluted in 400 μL of 100 mM glycine, pH 2. The immunoprecipitation success was verified by SDS-PAGE followed by Coomassie blue staining and Western blotting against the TAP tag. Sample Processing Using the Microfluidic Proteomic Reactor. For the evaluation of the chip system performance, a single channel device was used with the standard protein BSA as test sample. All of the solutions were applied onto the chip system using a pressurized vessel. The microfluidic proteomic reactor was operated as follows: the device system was first conditioned with 10 mM citric acid (CA), pH 3. Different amounts of BSA were dissolved in 10 mM CA, pH 3, and mixed with trypsin (Promega, Madison, WI) in the same buffer with a ratio of 5:1 (w/w) and loaded onto the device. After being washed with water, the system was dried by flowing nitrogen gas. For protein reduction, 100 mM DTT in 10 mM ammonium bicarbonate (ABC) was infused into the device and kept in the system for 30 min at room temperature. Then, the device was washed briefly with 10 mM ABC and dried by flowing nitrogen gas. For alkylation and digestion, 10 mM iodoacetamide in 100 mM Tris-HCl, pH 8, was infused on the device and incubated for 1 h at room temperature. Finally, the resulting peptides were eluted with 200 mM ABC, and the solution was adjusted to pH ∼2 by adding 50% (v/v) formic acid. The capillary-based proteomic reactor was performed as previously described with minor modifications.16 The reactor column dimensions were kept at 200 μm i.d.  2 cm for all of the experiments. The same proteomic reactor procedure as described for the microfluidic proteomic reactor was used. Immunoprecipitated samples from three yeast strains (control, Htz1-TAP, and Htz1-TAP swr1Δ) were processed on the threechannel microfluidic proteomic reactor at the same time. As 4096

dx.doi.org/10.1021/ac200194d |Anal. Chem. 2011, 83, 4095–4102

Analytical Chemistry

ARTICLE

Figure 1. (A) Workflow of microfluidic proteomic reactor-based protein
protein interaction study. (B) Microfluidic proteomic reactor design featuring multiplexed channels with glass capillary world-to-chip connection.

shown in Figure 1, the protein samples were mixed with 0.4 μg of trypsin and loaded onto three separated channels from the separate channel entrances. To remove possible carryover in the integrated end, the channels loaded with protein samples were washed with washing buffer (10 mM CA (pH 3), 50 mM NaCl, 20% (v/v) ACN). As the whole proteomic reactor procedure was exactly the same for all three channels, all of the solutions for the following steps were loaded from the common entrance at the other end of the chip. The following proteomic reactor steps for reduction, alkylation, and digestion were as described above, except three times more solutions were loaded onto the system. Online Nano-HPLC-MS/MS Analysis. An HPLC-MS/MS system, which consists of either an Agilent 1100 series capillary HPLC system or Eksigent nanoLC Ultra 1D system and a LTQ linear ion trap mass spectrometer equipped with a nanospray source (Thermo, San Jose, CA, USA), was used for the peptide analysis. The peptide separation was performed on a tip column (75 μm i.d.  7 cm) packed with 3 μm/120 Å ReproSil-Pur C18 resins (Dr. Maisch GmbH, Ammerbuch, Germany) with or without precolumn. The flow rates were 10 μL/min or 400 nL/min during sample loading and 200 nL/min for HPLC-MS/MS during analysis. The buffers used for online analysis were 0.1% (v/v) formic acid water and ACN with 0.1% (v/v) formic acid. The gradient from 5 to 35% (v/v) ACN was either performed in 30 min for BSA or 240 min for immunopurification samples. The spray voltage was set at þ1.8 kV, and the collision energy was set at 35%. The MS and MS/MS spectra were collected in a data-dependent mode with one MS scan followed by 10 MS/MS scans. The acquired MS/MS spectra were converted to mgf files using MM File Conversion Tools (MassMatrix, Chicago, IL, USA). Database searching was performed by Mascot 2.2.02 (Matrix

Science) against 35 sequence database for BSA samples and Saccharomyces cerevisiae NCBI database (6298 sequences, released April 2007) for immunoprecipitation samples. The parameters were set as follows: cysteine residue was set as a static modification of 57.0215 Da, and methionine residue was set as a variable modification of þ15.9949 Da. The precursor and fragment mass tolerances were set at 2.0 and 0.8 Da, respectively. Two miss cleavages were allowed, and the significance threshold was set to 0.05 for BSA samples and 0.01 for immunoprecipitation samples. Mascot score was set at 30, and “bold red peptides” were only accepted.

’ RESULTS AND DISCUSSION Fabrication and Assembly of the Microfluidic Proteomic Reactor. The microfluidic system (Figure 1A) was fabricated in

thermoplastics as a hybrid structure onto which the microfluidic structures were embossed in both a hard thermoplastic material (Zeonor 1060R) and an easily bonded SEBS TPE polymer.33 The fabrication of both microfluidic substrates was accomplished using a high-throughput hot-embossing process.34 Reactor bead beds were then incorporated into the devices, and they were confined by constrictions at both extremities of the channels (Figure 1B). The constrictions consisted of two rows of vertical pillars that were 25 μm in diameter with an interstice of 8 μm between pillars. The device consisted of a hybrid structure with channels embossed in thermoplastic elastomer (TPE) and assembled onto a hard thermoplastic (TP), specifically Zeonor 1060R substrates. The fabrication method with the polymeric mold allowed for the replication of high aspect ratio pillars (AR > 2). The SCX beads were easily introduced and compacted directly into the channels with a manual syringe, and a seal was placed atop the injection port. The bead density was uniform 4097

dx.doi.org/10.1021/ac200194d |Anal. Chem. 2011, 83, 4095–4102

Analytical Chemistry

ARTICLE

Table 1. Performance Comparison between Microfluidic Proteomic Reactor and Conventional Proteomic Reactor for Processing 2 μg of Model Protein BSA and Identified by Nano-HPLC-MS/MS identified

identified unique

sequence

peptides

peptides

coverage (%)

proteomic reactor first second

first

second

first

second

microfluidic

313

293

36

35

59

59

capillary based

239

205

60

54

79

76

Table 2. Limit of Detection for Model Protein BSA Processed by the Microfluidic Proteomic Reactor and Identified by Nano-HPLC-MS/MSa

Figure 2. Microfluidic proteomic reactor processing of model protein BSA. Two micrograms of BSA in 10 mM citric acid, pH 3, was subjected to preconcentration, buffer exchanging, reduction, and alkylation/digestion in the channel packed with 1 cm of SCX beads. Base peak chromatogram represents the detected digested BSA peptides as analyzed by the nano-HPLC-MS/MS system. See Materials and Methods for sample processing and separation conditions. The labeled peaks with / contain peptides belonging to BSA.

across the multiplex channels, as flow rates through the channels were determined to be equivalent. This ensured reproducibility between analyses. Alignment structures were included for incorporation of a glass capillary tube for world-to-chip connection and easy connection to the mass spectrometer (Figure 1B). We have selected a TPE consisting of a block copolymer of SEBS for the fabrication of the devices. Unlike the widely used PDMS in lab-on-chip devices, the TPE can be manipulated using industrially available thermoforming processes (extrusion, injection molding, and hot-embossing). Here, Zeonor COC TP was chosen as a solid substrate, as strong bonding can be achieved between two TPE/TP substrates at low temperature and low applied pressure as the TPE polymer chains can reorganize and reorientate to allow for conformal bonding.33 The assembled devices have been shown to easily resist pressure up to 100 psi. At higher pressure, the seal is often broken at the glass capillary interconnect, which has been addressed by adding a small drop of chlorobenzene to seal the glass/TPE/TP interface. This combination of elastomeric and thermoplastic materials provides a facile fabrication and assembly, critical for eventual mass production. Currently, on a standard microfabricated 6 in. wafer, six devices are replicated per embossing cycle with high fidelity. The use of polymeric molds for the hot-embossing allows for easy deembossing of both the thermoplastic and thermoplastic elastomer substrate. The working stamp molds were resistant to multiple embossing cycles and have been used for more than 100 embossing cycles. Microfluidic Proteomic Reactor Performance. We first compared the performance of the microfluidic proteomic reactor to the conventional proteomic reactor.16 A standard protein, BSA, was processed and analyzed by nano-HPLC-MS/MS system (Figure 2). All of the peptide peaks that originated from BSA were labeled on a base peak chromatogram. More than 30 peaks were detected, and no obvious peak tailing was observed. This result demonstrates that (1) the protein sample is digested

a

BSA

identified

identified unique

sequence coverage

amount

peptides

peptides

(%)

2 ng

7

5

10

20 ng

18

10

20

200 ng 2 μg

72 303

27 36

44 59

The displayed results are the average from two technical replicates.

efficiently and that (2) the proteomic reactor on a polymeric chip does not contaminate the HPLC-ESI-MS/MS with residual chemical from the polymers. As shown in Table 1, around 300 peptides were successfully identified with protein sequence coverage nearing 60%. It is interesting to note that the proteomic reactor on a chip provides comparable results to the conventional capillary-based proteomic reactor under the same operation conditions. However, the microfluidic proteomic reactor system further simplifies the system and reduces possible dead volume and carryover as well as the time it takes to process the sample. By incorporating the whole proteomic reactor setup into one microchannel with integrated pillar frit, external frits and unions are avoided. The same amounts of BSA samples were also processed by the standard in-solution digestion approach. However, no reliable identification was obtained based on the data processing cutoff used in this study. Since the microfluidic proteomic reactor is targeted at proteomics applications with relative limited starting material, especially for the study of protein
protein interactions, we further investigated the limit of detection of the system. As summarized in Table 2, different amounts of BSA samples from 2 μg to 2 ng, which covers three dynamic ranges, were systematically analyzed on the proteomic reactor on a chip. As expected, the number of identified peptides decreases according to the amount of BSA sample. Interestingly, the number of unique peptides and sequence coverage do not decrease dramatically as compared with peptide number. Even for 2 ng of BSA sample, the microfluidic proteomic reactor can still provide 10% of sequence coverage with five unique peptides, which is sufficient for confident protein identification. Our limit of detection with the proteomic reactor on a chip is below 2 ng of proteins, which is over 1000 times below the typical amount of total proteins obtained in a protein
protein study. This low limit of detection will benefit the analysis of lower abundance interaction partners in immunoprecipitated samples. Therefore, the microfluidic proteomic reactor is suitable for the study of protein
protein interactions. 4098

dx.doi.org/10.1021/ac200194d |Anal. Chem. 2011, 83, 4095–4102

Analytical Chemistry

ARTICLE

Figure 3. Htz1 and Htz1 swr1Δ interacted protein complex were purified by TAP tag and detected by SDS-PAGE (4
12% gradient gel)/Coomassie blue staining and Western blotting probed with anti-TAP antibody (A); and processed by three-channel microfluidic proteomic reactor and detected by nano-HPLC-MS/MS (B). The labeled peaks with / contain peptides caused by trypsin autodigestion.

Multiplex Microfluidic Proteomic Reactor and Protein
Protein Interaction Study. A key feature of the microfabricated

proteomic reactor is the possibility of having multiple samples processed on the same device. This is particularly important for high-throughput protein
protein interaction experiments requiring multiple immunoprecipitations. Moreover, such multichannel devices would also outperform lower throughput experiments by allowing all of the control and sample immunopurifications to be simultaneously processed. To this end, we developed a multichannel microfluidic proteomic reactor for the simultaneous processing of immunopurified samples. The proteomic reactor workflow consists of very specific processing steps which permitted the design of a simple and integrated multichannel proteomic reactor. As shown in Figure 1A, three separated channels were integrated into one single chip. Within each channel, a 1 cm bed of SCX beads was restricted into a specific region using two pillar frits. Immunopurified samples can be loaded on an individual channel reactor through the separate

entrances of the channel. Once the proteins are immobilized on the SCX reactors, the channels can be extensively washed with no loss in proteins. With this design, three different protein samples can be loaded onto different channels. On the opposite side of the device, the three SCX bead-packed channels merged together into a single channel. The remaining steps in the processing of the samples can then be performed simultaneously by introducing through that single channel all of the reagents necessary for protein reduction, alkylation, and digestion. This design greatly simplifies the system setup and reduces the operation time. Moreover, the final peptide mixture can then be individually collected from the individual channels for analysis by nanoHPLC-MS/MS. The multiplex microfluidic proteomic reactor workflow was evaluated with standard protein BSA. Two micrograms of BSA protein was loaded onto three channels and processed simultaneously. Two multiplex chips were tested. After the digested peptides were eluted, they were loaded and analyzed 4099

dx.doi.org/10.1021/ac200194d |Anal. Chem. 2011, 83, 4095–4102

Analytical Chemistry

ARTICLE

Figure 4. Integrated protein
protein interaction map of Htz1 and Htz1 swr1Δ as identified by three-channel microfluidic proteomic reactor and nanoHPLC-MS/MS. The interacting proteins are indicated by nodes that are color coded according to their Gene Ontology annotation. Blue lines represent interactions, and the density of the line indicates the strength of the interaction as characterized by spectral counting.

by the LC-MS/MS system separately. On average, 600 peptides were identified per channel with a relative standard deviation of 8% (n = 6 channels), which demonstrates a reasonably low variation between channels. mChIP Coupled with the Three-Channel Microfluidic Proteomic Reactor. Recently, we developed a novel protein
protein interaction approach, termed modified chromatin immunopurification (mChIP), for the specific enrichment and analysis of protein
DNA complexes.32 Typically, in protein interaction studies, individual bait protein and their interaction partners are immunopurified and analyzed by mass spectrometry. Instead, here we combined the mChIP approach and the three-channel microfluidic proteomic reactor to simultaneously perform multiple protein interaction studies allowing bait protein in different background cells and control experiments to be simultaneously performed. We illustrate the performance of this approach by studying the interactions of the histone variant Htz1, which in yeast plays important roles in different biological processes regulated by chromatin.35 The nucleosomes of chromatin are octomers composed of two copies of the heterodimes H2A
H2B and H3
H4. In yeast, the canonical histone H2A can be substituted in specific instances by Htz1 (H2A.Z). The substitution of histone H2A by Htz1 occurs in promoter regions, implicating Htz1 in the regulation of transcription. Interestingly, the exchange of H2A by Htz1 is controlled by a complex called Swr1. Deletion of Swr1 in yeast inhibits the exchange of H2A for Htz1. Therefore, we expect significant changes in the interactomes of Htz1 in wild-type yeast and Htz1 in a swr1Δ yeast strain. Briefly, mChIP experiments were performed for Htz1-TAP expressed in wild-type yeast, Htz1-TAP expressed in a swr1Δ yeast strain, and an untagged control experiment. The TAPtagged bait proteins Htz1 in a wild-type yeast and swr1Δ yeast strain were successfully immunopurified as confirmed by Coomassie blue and Western blotting (Figure 3A). The same set of immunopurified Htz1 from wild-type yeast, swr1Δ yeast strain, and untagged control immunopurification were introduced on the three-channel microfluidic proteomic reactor. The samples in

the individual channel were simultaneously processed, and the obtained tryptic peptides were analyzed by online nanoHPLC-MS/MS. Two biological replicates were performed for the whole process. Figure 3B shows the base peak chromatograms obtained from the three immunoprecipitated samples. The major peaks present in the control sample are from common protein contaminants (e.g., trypsin autodigestion, IgG, and keratins). Htz1 pull down has the highest intensity followed by the Htz1 swr1Δ mutation and the control experiment, which is consistent with the protein intensities observed on the gel. This result demonstrates the efficient sample processing as done by the three-channel microfluidic proteomic reactor. It has to be pointed out that the three samples were loaded and processed at the same time on the chip system within 3 h. The proteins identified in the Htz1 from wild-type yeast, swr1Δ yeast strain, and untagged control immunopurification (minus common contaminants and ribosomal proteins) are listed in the Supporting Information table. Twenty-one unique proteins were identified in Htz1 from wild-type yeast, 11 unique proteins for Htz1 in swr1Δ yeast strain, and four proteins in the untagged control immunopurification. This list of proteins was then used to generate the interactomes of Htz1 from wild-type yeast and swr1Δ yeast strain (Figure 4). As shown in Figure 4, the results were consistent with previous reports as the deletion of swr1 significantly impacts the number of interactors with Htz1. Most of the chromatin remodeling related proteins, such as TAF9 and RIF1, which are key factors for chromatin stability and subsequent transcription, were lost from Htz1 swr1Δ interactome.36,37 However, Nap1, as one of the major interactors, is present in the Htz1 interactome regardless of the swr1Δ deletion. This is consistent with the important role that Nap1 plays in maintaining a soluble pool of Htz1,38 which does not require the presence of Swr1. Furthermore, five proteins were present solely in Htz1 swr1Δ interactome. Of these, Kap114 is one of the major interactors as identified by microfluidic proteomic reactor and mass spectrometry with 16 unique peptides. It has already been shown that Nap1-Htz1 and Kap114 4100

dx.doi.org/10.1021/ac200194d |Anal. Chem. 2011, 83, 4095–4102

Analytical Chemistry form a co-complex in the cytosol and play an important role for importing Htz1 into the nucleus.38 In this study, we developed the first microfluidic chip-based proteomic reactor and demonstrated its potential for processing of limited amount of immunopurified samples. All of the sample processing steps are performed on the microfluidic chip channel packed with SCX beads and completed within 2
3 h. We further exemplified the potential of the microfluidic proteomic reactor by simultaneously processing the interactome of the Htz1 protein in wild-type yeast and in a swr1Δ yeast strain. Overall, these results demonstrate that the three-channel microfluidic proteomic reactor provides a promising platform for highthroughput protein
protein interaction study by mass spectrometry and has the potential to be further parallelized and mass produced for disposable usage.

’ ASSOCIATED CONTENT

bS

Supporting Information. Additional table data. This material is available free of charge via the Internet at http://pubs.acs.org.

’ AUTHOR INFORMATION Corresponding Author

*D.F.: e-mail dfi[email protected], phone (613) 562-5800 ext. 8674, fax (613) 562-5655. T.V.: e-mail [email protected]. gc.ca, phone (450) 641-5232, fax (450) 641-5105. Present Addresses #

The Samuel Lunenfeld Research Institute, Mount Sinai Hospital, 600 University Avenue, Toronto, ON, Canada M5G 1X5.

’ ACKNOWLEDGMENT The first two authors contributed equally to this work. This work was supported by an NSERC-Strategic grant (STPG396508), Genome Canada, the Province of Ontario, and the Genomic and Health Initiative of National Research Council. D. F. would like to acknowledge a Canada Research Chair in Proteomics and Systems Biology. The authors would also like to thanks C.M. Godin and H. Roberge at the IMI-NRC for their support in the fabrication and analysis of the microfabricated proteomic reactor. ’ REFERENCES (1) Aebersold, R.; Mann, M. Nature 2003, 422, 198–207. (2) Nilsson, T.; Mann, M.; Aebersold, R.; Yates, J. R., III; Bairoch, A.; Bergeron, J. J. Nat. Methods 2010, 7, 681–685. (3) Gingras, A. C.; Gstaiger, M.; Raught, B.; Aebersold, R. Nat. Rev. Mol. Cell Biol. 2007, 8, 645–654. (4) Ewing, R. M.; Chu, P.; Elisma, F.; Li, H.; Taylor, P.; Climie, S.; McBroom-Cerajewski, L.; Robinson, M. D.; O’Connor, L.; Li, M.; Taylor, R.; Dharsee, M.; Ho, Y.; Heilbut, A.; Moore, L.; Zhang, S.; Ornatsky, O.; Bukhman, Y. V.; Ethier, M.; Sheng, Y.; Vasilescu, J. Abu-Farha, M.; Lambert, J. P.; Duewel, H. S.; Stewart, I. I.; Kuehl, B.; Hogue, K.; Colwill, K.; Gladwish, K.; Muskat, B.; Kinach, R.; Adams, S. L.; Moran, M. F.; Morin, G. B.; Topaloglou, T.; Figeys, D. Mol. Syst. Biol. 2007, 3, 89. (5) Ho, Y.; Gruhler, A.; Heilbut, A.; Bader, G. D.; Moore, L.; Adams, S. L.; Millar, A.; Taylor, P.; Bennett, K.; Boutilier, K.; Yang, L.; Wolting, C.; Donaldson, I.; Schandorff, S.; Shewnarane, J.; Vo, M.; Taggart, J.; Goudreault, M.; Muskat, B.; Alfarano, C.; Dewar, D.; Lin, Z.; Michalickova, K.; Willems, A. R.; Sassi, H.; Nielsen, P. A.; Rasmussen, K. J.; Andersen, J. R.; Johansen, L. E.; Hansen, L. H.; Jespersen, H.; Podtelejnikov, A.;

ARTICLE

Nielsen, E.; Crawford, J.; Poulsen, V.; Sorensen, B. D.; Matthiesen, J.; Hendrickson, R. C.; Gleeson, F.; Pawson, T.; Moran, M. F.; Durocher, D.; Mann, M.; Hogue, C. W.; Figeys, D.; Tyers, M. Nature 2002, 415, 180–183. (6) Vermeulen, M.; Eberl, H. C.; Matarese, F.; Marks, H.; Denissov, S.; Butter, F.; Lee, K. K.; Olsen, J. V.; Hyman, A. A.; Stunnenberg, H. G.; Mann, M. Cell 2010, 142, 967–980. (7) Calleri, E.; Temporini, C.; Perani, E.; De Palma, A.; Lubda, D.; Mellerio, G.; Sala, A.; Galliano, M.; Caccialanza, G.; Massolini, G. J. Proteome Res. 2005, 4, 481–490. (8) Feng, S.; Ye, M.; Jiang, X.; Jin, W.; Zou, H. J. Proteome Res. 2006, 5, 422–428. (9) Krenkova, J.; Lacher, N. A.; Svec, F. Anal. Chem. 2009, 2004–2012. (10) Liu, Y.; Xue, Y.; Ji, J.; Chen, X.; Kong, J.; Yang, P.; Girault, H. H.; Liu, B. Mol. Cell. Proteomics 2007, 6, 1428–1436. (11) Ma, J.; Liu, J.; Sun, L.; Gao, L.; Liang, Z.; Zhang, L.; Zhang, Y. Anal. Chem. 2009, 81, 6534–6540. (12) Temporini, C.; Perani, E.; Calleri, E.; Dolcini, L.; Lubda, D.; Caccialanza, G.; Massolini, G. Anal. Chem. 2007, 79, 355–363. (13) Hsieh, Y. L.; Wang, H.; Elicone, C.; Mark, J.; Martin, S. A.; Regnier, F. Anal. Chem. 1996, 68, 455–462. (14) Tian, R.; Wang, S.; Elisma, F.; Li, L.; Zhou, H.; Wang, L.; Figeys, D. Mol. Cell. Proteomics 2011, M110.000679 (Epub 2010 June 8). (15) Zhou, H.; Elisma, F.; Denis, N. J.; Wright, T. G.; Tian, R.; Hou, W.; Zou, H.; Figeys, D. J. Proteome Res. 2010, 9, 1279–1288. (16) Ethier, M.; Hou, W.; Duewel, H. S.; Figeys, D. J. Proteome Res. 2006, 5, 2754–2759. (17) Zhou, M.; Veenstra, T. D. Proteomics 2007, 7, 2688–2697. (18) Zhou, H.; Hou, W.; Denis, N. J.; Vasilescu, J.; Zou, H.; Figeys, D. J. Proteome Res. 2009, 8, 556–566. (19) Vasilescu, J.; Zweitzig, D. R.; Denis, N. J.; Smith, J. C.; Ethier, M.; Haines, D. S.; Figeys, D. J. Proteome Res. 2007, 6, 298–305. (20) Chen, C.; Jin, J.; James, D. A.; Adams-Cioaba, M. A.; Park, J. G.; Guo, Y.; Tenaglia, E.; Xu, C.; Gish, G.; Min, J.; Pawson, T. Proc. Natl. Acad. Sci. U.S.A. 2009, 106, 20336–20341. (21) Huikko, K.; Kostiainen, R.; Kotiaho, T. Eur. J. Pharm. Sci. 2003, 20, 149–171. (22) Sedgwick, H.; Caron, F.; Monaghan, P. B.; Kolch, W.; Cooper, J. M. J. R. Soc. Interface 2008, 5, S123–S130. (23) Szita, N.; Polizzi, K.; Jaccard, N.; Baganz, F. Curr. Opin. Biotechnol. 2010, 21, 517–523. (24) Feng, X.; Du, W.; Luo, Q.; Liu, B.-F. Anal. Chim. Acta 2009, 650, 83–97. (25) Henion, J. LCGC North America 2009, 27, 900–915. (26) Lee, J.; Soper, S. A.; Murray, K. K. Anal. Chim. Acta 2009, 649, 180–190. (27) Lion, N.; Rohner, T. C.; Dayon, L.; Arnaud, I. L.; Damoc, E.; Youhnovski, N.; Wu, Z.-Y.; Roussel, C.; Josserand, J.; Jensen, H.; Rossier, J. S.; Przybylski, M.; Girault, H. H. Electrophoresis 2003, 24, 3533–3562. (28) Freire, S. L. S.; Wheeler, A. R. Lab Chip 2006, 6, 1415–1423. (29) le Gac, S., van den Berg, A., Eds. Miniaturization and Mass Spectrometry; RSC Publishing: London, 2009. (30) Glinsner, T.; Veres, T.; Kreindl, G.; Roy, E.; Morton, K.; Wieser, T.; Thanner, C.; Treiblmayr, D.; Miller, R.; Lindner, P. Microelectron. Eng. 2010, 87, 1037–1040. (31) Duffy, D. C.; McDonald, J. C.; Schueller, O. J. A.; Whitesides, G. M. Anal. Chem. 1998, 70, 4974–4984. (32) Lambert, J. P.; Mitchell, L.; Rudner, A.; Baetz, K.; Figeys, D. Mol. Cell. Proteomics 2009, 8, 870–882. (33) Holden, G. Thermoplastic Elastomers; Hanser Publishers: Munich, Germany, 1996. (34) Worgull, M. Hot Embossing: Theory and Technology of Microreplication; William Andrew Elsevier: Oxford, 2009. (35) Wu, W. H.; Wu, C. H.; Ladurner, A.; Mizuguchi, G.; Wei, D.; Xiao, H.; Luk, E.; Ranjan, A.; Wu, C. J. Biol. Chem. 2009, 284, 6200–6207. (36) Krogan, N. J.; Baetz, K.; Keogh, M. C.; Datta, N.; Sawa, C.; Kwok, T. C.; Thompson, N. J.; Davey, M. G.; Pootoolal, J.; Hughes, 4101

dx.doi.org/10.1021/ac200194d |Anal. Chem. 2011, 83, 4095–4102

Analytical Chemistry

ARTICLE

T. R.; Emili, A.; Buratowski, S.; Hieter, P.; Greenblatt, J. F. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 13513–13518. (37) Wu, W. H.; Wu, C. H.; Ladurner, A.; Mizuguchi, G.; Wei, D.; Xiao, H.; Luk, E.; Ranjan, A.; Wu, C. J. Biol. Chem. 2009, 284, 6200–6207. (38) Straube, K.; Blackwell, J. S., Jr.; Pemberton, L. F. Traffic 2010, 11, 185–197.

4102

dx.doi.org/10.1021/ac200194d |Anal. Chem. 2011, 83, 4095–4102