Biosensors and Bioelectronics 73 (2015) 93–99
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Sequential phosphorylation analysis using dye-tethered peptides and microfluidic isoelectric focusing electrophoresis Hoseok Choi a,b,1, Nakchul Choi c,1, Butaek Lim a,b,1, Tae-Wuk Kim a,b, Simon Song c,d,n, Young-Pil Kim a,b,d,nn a
Department of Life Science, Hanyang University, Seoul 133-791, Republic of Korea Research Institute for Natural Sciences, Hanyang University, Seoul 133-791, Republic of Korea c Department of Mechanical Convergence Engineering, Hanyang University, Seoul 133-791, Republic of Korea d Institute of Nano Science and Technology, Hanyang University, Seoul 133-791, Republic of Korea b
art ic l e i nf o
a b s t r a c t
Article history: Received 28 February 2015 Received in revised form 11 May 2015 Accepted 21 May 2015 Available online 25 May 2015
We report a simple method for analyzing sequential phosphorylation by protein kinases using fluorescent peptide substrates and microfluidic isoelectric focusing (μIEF) electrophoresis. When a dye-labeled peptide substrate was sequentially phosphorylated by two consecutive protein kinases (mitogenactivated protein kinase (MAPK) and glycogen synthase kinase 3 (GSK3)), its differently phosphorylated forms were easily separated and visualized by fluorescent focusing zones in the μIEF channel based on a change in the isoelectric point (pI) by phosphorylation. As a result, ratiometric and quantitative analysis of the fluorescent focusing regions shifted by phosphorylation enabled the analysis of phosphorylation efficiency and the relevant inhibition of protein kinases (MAPK and GSK3) with high simplicity and selectivity. Furthermore, the GSK3 activity in the cell lysates was elucidated by μIEF electrophoresis in combination with immunoprecipitation. Our results suggest that this method has great potential for analyzing the sequential phosphorylation of multiple protein kinases that are implicated in cellular signaling pathways. & 2015 Elsevier B.V. All rights reserved.
Keywords: Sequential phosphorylation Microfluidic Isoelectric focusing Protein kinase Inhibition assay Immunoprecipitation
1. Introduction Phosphorylation cascades involving multiple protein kinases play a central role in signal transduction, protein regulation, and metabolism in living cells (Cohen, 2002; Hunter, 1995; Schenk and Snaar-Jagalska, 1999; Walsh and MacDonald, 2011). The analysis of these phosphorylation cascades, therefore, will provide new insights into their physiological functions in many biological events. While general methods for probing protein kinase activity have adopted [gamma-32P]ATP-labeled substrate in electrophoresis or non-radioisotopic immunoblotting to differentiate between phosphorylated and unphosphorylated substrates (Hastie et al., 2006; Peck, 2006; Yamamoto et al., 2006; Zhu et al., 2000), they n Corresponding author at: Department of Mechanical Convergence Engineering, Hanyang University, Seoul 133-791, Republic of Korea. Tel.: þ 82 2 2220 0423. nn Corresponding author at: Department of Life Science and Research Institute for Natural Sciences, Hanyang University, Seoul 133-791, Republic of Korea. Tel.: þ 82 2 2220 2560. E-mail addresses: [email protected] (S. Song), [email protected] (Y.-P. Kim). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.bios.2015.05.047 0956-5663/& 2015 Elsevier B.V. All rights reserved.
are not only based on long processing times, but they are limited to the rapid identification of a cascade event involving multiple kinases or the determination of the quantitative amount of phosphorylation. Fluorogenic or fluorescence resonance energy transfer-based methods using peptide substrates have been alternatively used for monitoring various phosphorylation events, which are based on the fluorescent change from substrate phosphorylation using dye-labeled antibodies or chemicals (Ghadiali et al., 2010; Harvey et al., 2008; Kupcho et al., 2003; Riddle et al., 2006; Shiosaki et al., 2013), but these are mostly used for single kinase reactions. In addition, mass spectrometric analysis has been implemented as a standard means for identifying the degree of phosphorylation (Kim et al., 2007; Mann et al., 2002; Salih, 2005), which a time-consuming and expensive process. To circumvent these issues, increasing attention has been paid to isoelectric focusing (IEF) technique because it is capable of easily separating diverse phosphorylated proteins according to their isoelectric point (pI) based on the negatively accumulating increase in net charge by sequential addition of phosphate group to the protein (Anderson and Peck, 2008). Although capillary-type IEF and two-dimensional IEF gel electrophoresis have been reported to be useful for separating phosphorylated proteins (Kinoshita et al., 2009; O'Neill et al., 2006), fewer efforts have been
94
H. Choi et al. / Biosensors and Bioelectronics 73 (2015) 93–99
made to analyze phosphorylation cascade, especially in a rapid and quantitative manner. Given a common substrate for multiple kinases, the IEF technique would allow for the phosphorylation cascades of the substrate to be easily resolved during multiple kinase reactions. Here we report a simple method for analyzing sequential phosphorylation using fluorescent peptide substrates and microfluidic isoelectric focusing (μIEF) electrophoresis. Despite recent advances in μIEF techniques for biological analyses (Das et al., 2007; Sommer and Hatch, 2009; Wen et al., 2010), many attempts have focused on the easy separation of complex biomolecules according to the average pIs. To the best of our knowledge, there have been no attempts in integrating the μIEF technique for the simultaneous detection of cascade phosphorylation. To demonstrate the sequential phosphorylation event, we chose two consecutive protein kinases: p42 mitogen-activated protein kinase (MAPK), also known as extracellular signal-regulated kinase 2 (ERK2), and glycogen synthase kinase 3 (GSK3). Since MAPK is considered to be one of the priming kinases for GSK3 in a process by which target substrates are sequentially phosphorylated (EldarFinkelman, 2002), we examined the phosphorylation cascade using an annotated peptide substrate from heat shock transcription factor 1 (HSF1), which is consecutively controlled by two protein kinases (Chu et al., 1996; He et al., 1998). Upon sequential phosphorylation, a dye-labeled peptide substrate gave rise to a mono- or di-phosphorylated form, allowing for easy separation and rapid detection in a polydimethylsiloxane (PDMS) μIEF channel equipped with an imaging analyzer. Moreover, the activity and inhibition of the protein kinases were quantitatively investigated by comparing two or three separated lines of the peptides on the μIEF. In contrast to conventional IEF electrophoresis, a microfluidic-based format enabled more reliable analyses with rapid assay time and much smaller reaction volumes (a few microliters).
2. Materials and methods 2.1. Materials P42 mitogen-activated protein kinase (MAPK) and glycogen synthase kinase 3 (GSK3) was purchased from New England Biolabs. Anti-human GSK3 beta polyclonal antibody, a protein G magnetic bead and a C18 pipette-tip column were purchased from Thermo Fisher Scientific Inc. (USA). RIPA buffer was purchased from Cell Signaling Technology. Methylcellulose (viscosity 400 cP, MC), lithium chloride (99%, LiCl), adenosine 5′-triphosphate (ATP) disodium salt hydrate (99%, 5’-ATP-Na2), α-cyano-4-hydroxycinnamic acid (98%, CHCA) and trifluoroacetic acid (99%, TFA) were purchased from Sigma-Aldrich. Acetonitrile (ACN, HPLC grade) was purchased from Merck. Ampholyte (Pharmalyte, broad range pH 3–10) was purchased from GE Healthcare Life Sciences. FR180204 was purchased from Santa Cruz Biotechnology Inc. (USA). Polydimethylsiloxane (PDMS) was purchased from Dow Corning. Isopropyl alcohol (99.5%, IPA) was purchased from Daejung. SU-8, a photoresist (PR), and SU-8 developer (PGMEA) was purchased from MicroChemicals. Tetramethyl-6-carboxyrhodamine (TAMRA)-labeled peptides (T-Pep, TAMRAKEEPPSPPQSPR; T-Pep(p), TAMRA-KEEPPSPPQpSPR; T-Pep(pp), TAMRA-KEEPPpSPPQpSPR) were synthesized from Peptron, Inc. (Korea). All other chemicals were of analytical grade and were used as received. 2.2. Fabrication of a microfluidic isoelectric focusing (μIEF) device A
microfluidic
chip
was
fabricated
by
bonding
Scheme 1. (A) Schematic of IEF electrophoresis in a PDMS microfluidic device equipped with fluorescent microscopy and imaging analysis. (B) Illustration of sequential phosphorylation of TAMRA-labeled peptide substrate (T-Pep) by two consecutive protein kinases (MAPK and GSK3), leading to the reduction in its pI value.
polydimethylsiloxane (PDMS) substrate on a flat PDMS substrate. A PDMS substrate was prepared by using a standard molding technique and soft lithography. To make a PDMS mold, a PR material was uniformly spin-coated at a thickness of 20 μm on a 4 in. Si wafer that was sequentially cleaned with acetone, IPA, and deionized (DI) water. The coated wafer was soft-baked for 3.5 min at 95 °C. Then it was exposed to 365 nm UV light (MDA-400M, Midas) through a film-type photo mask having a pattern of straight channel. In order to enhance the crosslinking of PR, a postexposure-baking of the wafer was performed for 4.5 min at 95 °C. The wafer was developed for 3.5 min in a solution of PGMEA. The wafer was washed with IPA to remove unexposed PR and dried with nitrogen gas. To fabricate the PDMS substrate, a mixture of PDMS (DC-184A) and curing agent (DC-184B) at a ratio of 10 to 1 was poured into the fabricated mold and degassed for 30 min and cured for 2 h at 70 °C in a vacuum oven (JSVO-60T, JSR). This PDMS substrate was bonded with a flat PDMS substrate after exposure to oxygen plasma (Plasma prep 300, Nanotech). Finally, the PDMS channel surface was coated with MC to make it hydrophilic for the purpose of reducing electroosmotic flow (EOF). The microfluidic chip had a straight channel with a length of 3 cm, a width of 300 μm and a height of 20 μm as shown in Scheme 1A. The reservoirs were attached on the inlet and outlet for the injection of electrolytes. 2.3. μIEF-based analysis of sequential phosphorylation by protein kinases Typically, protein kinase reactions were performed in a tube prior to μIEF analysis. For the single kinase reaction, appropriate TAMRA-labeled peptide substrate (1 μL at 100 μM), MAPK or GSK3 (1 μL at 25 U), ATP (0.4 μL at 10 mM), and 10 kinase reaction buffer (2 μL) were mixed at a final volume of 20 μL in standard reaction buffer (20 mM Tris–HCl, pH 7.4). For the sequential protein kinase reaction, the equivalent mixture (each 1 μL at 25 U) of MAPK and GSK3 was added to the same reaction solution. All kinase reactions were typically run for 90 min at 30 °C, followed by a thermal inactivation process for 20 min at 60 °C in order to load the aliquot into the PDMS microfluidic channel under the same condition. As a control, three peptides (T-Pep, T-Pep(p), and T-Pep (pp)) were mixed at a 1:1:1 molar ratio (each 1 μL at 100 μM) at a
H. Choi et al. / Biosensors and Bioelectronics 73 (2015) 93–99
final volume of 20 μL in reaction buffer (20 mM Tris–HCl, pH 7.4). A typical 15 μL of each reaction was initially mixed with ampholyte (15 μL at 12% w/w DI-water) and MC (15 μL at 1.2% w/w DIwater) and subsequently a 2 μL aliquot of the mixture solution was loaded into the microfluidic channel through an anodic reservoir prior to electrophoresis. The final loading concentrations of peptide, ampholyte, and MC were 1.7 μM, 4% w/w DI-water, and 0.4% w/w DI-water, respectively. The anodic and cathodic reservoirs were then placed with 50 mM H3PO4 in 0.4% w/w MC as the anolyte and 50 mM NaOH in 0.4% w/w MC as the catholyte, respectively. Platinum wire electrodes were connected to the reservoirs, and focusing was carried out at constant electric potentials up to 100 V cm 1 using a power supply (PS350, SRS). After focusing was complete for 10 min, the fluorescent image of the channel was acquired from an inverted microscope (IX71, Olympus) equipped with a CCD camera (DP71, Olympus). For the kinase inhibition assay in the μIEF channel, a stock solution of LiCl or FR180204 was serially diluted to different concentrations in the reaction buffer, and an equal volume (2 μL) of each inhibitor concentration was then mixed with a solution containing peptide substrate, protein kinase, and ATP dissolved in the reaction buffer. At a 20 μL reaction, the final concentrations of peptide, protein kinase, and ATP were 5 μM, 1.25 U, and 200 μM, respectively. After incubation for 90 min at 30 °C, the reactant was followed by a thermal inactivation process for 20 min at 60 °C. Each 15 μL of reaction samples with different concentrations of inhibitor was initially mixed with ampholyte (15 μL at 12%) and MC (15 μL at 1.2%) and then a 2 μL aliquot of the mixture solution was loaded into the microfluidic channel for IEF electrophoresis. 2.4. Determination of phosphorylation efficiency and IC50
95
0.5% TFA/50% ACN. Elution of the target peptide from the C18 tip was conducted directly on a standard stainless steel MALDI target by dispensing about 0.7 μL of the matrix solution containing 1 mg of CHCA in 0.5% TFA/50% ACN. Mass spectrometric analysis of the peptides was performed using an Axima-CFR (Shimadzu). 2.6.
μIEF-based analysis of the protein kinase activity in cell lysates
MCF-7 cells grown in the 75 cm2-culture flask were washed three times with phosphate-buffered saline (PBS, pH 7.4), and collected in micro-centrifuge tubes by scraping in 2 mL of RIPA buffer (25 mM Tris–HCl, pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate and 0.1% sodium dodecyl sulfate). Cells were lysed for 10 min in an ice bath with intermittent vortexing. The cytosolic fraction was obtained by centrifuging the cell lysate at 13,000 rpm for 20 min and the supernatant was collected in new tubes. For immunoprecipitation (IP), protein A/G magnetic beads (25 μL) were coated with polyclonal anti-GSK3 antibody (5 μg) for 12 h at 4 °C, and then incubated with 100 μL of the supernatant (with or without spiking GSK3) in binding buffer (50 mM Tris–HCl, pH7.5, NaCl 150 mM NaCl, 0.2% NP-40 and 0.05 mg mL 1 BSA) for 2 h at 4 °C on a gently rotating device. Magnetic beads were washed three times in washing buffer (50 mM Tris–HCl, pH7.5, 100 mM NaCl and 0.1% NP-40) using a magnetic stand and resuspended in kinase reaction buffer. For the kinase reaction, the immunoprecipitated magnetic beads (10 μL) were further incubated with T-Pep(p) (1 μL at 100 μM) for 90 min at 30 °C, followed by a thermal inactivation process for 20 min at 60 °C. Each 10 μL of the reactants under different IP conditions was initially mixed with ampholyte (10 μL at 12%) and MC (10 μL at 1.2%) and then a 2 μL aliquot of the mixture solution was loaded into the microfluidic channel for IEF electrophoresis.
The phosphorylation efficiency of the kinase reaction was calculated using the following equation;
EP (%) =
IZone (p) IZone + IZone (p)
3. Results and discussion
× 100
(1)
where EP is the phosphorylation efficiency (%), IZone and IZone(p) are the fluorescent intensities of the control focusing zone in the microfluidic channel before the kinase reaction and the phosphorylated focusing zone after the kinase reaction, respectively. The fluorescent intensity of the focusing zones was determined by imaging analysis using a Matlab software (ver 2012, The MathWorks, Inc.) For the IC50, the EP was plotted as a function of inhibitor concentration, and then fitted to a 4-parameter logistic equation (dose–response model for ligand binding) by the nonlinear regression procedure of Origin Pro (ver 8.0, OriginLab Corporation) using the following equation:
EP (%, for inhibition) =
EPmax − EPmin + EPmin 1 + (C /IC50 )n
(2)
where EPmax and EPmin are the maximum and minimum phosphorylation efficiencies, respectively, at a given kinase-substrate reaction, IC50 is the median inhibitory concentration, C is the inhibitor concentration, and n is the slope factor. 2.5. Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) analysis A C18 pipette-tip was used to concentrate and desalt the peptide substrates with or without the kinase reaction according to the manufacturer's specifications. The C18 tip was wetted in 0.5% TFA/50% ACN and activated in 0.5% TFA/distilled water. A 10 μL of the sample reactant was adsorbed onto the C18 tip and rinsed in
3.1. Fabrication and detection principle Scheme 1A illustrates a simple microfluidic channel in a PDMS for IEF-based separation of the sequentially phosphorylated peptide. While 0.4% methylcellulose (MC) was used with the anodic or cathodic electrolyte to reduce electroosmosis and peak drift in IEF by applying a dynamic coating of MC to channel walls, as reported earlier (Cui et al., 2005), fluorescent peptide substrates were mixed with 4% carrier ampholytes and 0.4% MC, and poured into the PDMS microfluidic channel through an anodic electrolyte reservoir under a high electric field, leading to migration and focusing into discrete zones in areas where the local pH was equal to the pI of the analyte. The resulting fluorescent zones were then imaged by the fluorescence microscope equipped with a CCD camera. As depicted in Scheme 1B, we employed a model peptide substrate labeled with TAMRA dye at its N-terminus (TAMRAKEEPPSPPQSPR, T-Pep). The synthetic peptide substrate (T-Pep) was then be sequentially phosphorylated by protein kinases (MAPK and GSK3) to its diphosphorylated form (TAMRAKEEPPpSPPQpSPR, T-Pep(pp)) through its monophosphorylated form (TAMRA-KEEPPSPPQpSPR, T-Pep(p)), which is based on the reported kinase signaling pathway of HSF1 possessing the peptide sequence (Chu et al., 1996). It is noteworthy that GSK3 is constitutively regulated by other protein kinases, which is central to multiple signaling pathways in association with a variety of extracellular stimuli (Doble and Woodgett, 2003); therefore, sequential phosphorylation by GSK3 activity and its coupled kinases (MAPK here) is very crucial to understand its multifunctional role. Inspired by this observation, the zone position by sequential phosphorylation of T-Pep was investigated in the μIEF channel to
96
H. Choi et al. / Biosensors and Bioelectronics 73 (2015) 93–99
Fig. 1. Fluorescent IEF images of peptide substrates obtained in a microfluidic PDMS channel under the different reaction conditions: (i) mixture of synthetic T-Pep, T-Pep(p), and T-Pep(pp) at a 1:1:1 molar ratio as a control, (ii) T-Pep þ MAPK, (iii) T-Pep þGSK3, (iv) T-Pep þ MAPKþ GSK3, (v) T-Pep(p)þ GSK3, and (vi) T-Pep (p) þGSK3 without ATP. The protein kinase reaction was performed by incubating the appropriate peptide with protein kinase(s) for 90 min at 30 °C in reaction buffer including 100 μM ATP, 10 mM MgCl2, 0.1 mM EDTA, 2 mM DTT and 0.01% Brij 35 in 20 mM Tris–HCl (pH 7.4). For IEF electrophoresis, 4% ampholyte and 0.2% MC was loaded into the μIEF channel with each sample. Unphosphorylated and phosphorylated peptides were displayed above the microfluidic channel, which were matched with fluorescent focusing zones based on their corresponding pI.
ascertain the significant change in its pI. The expected pI values of the modified peptides were calculated from the mean pKa values for the phosphorylated amino acids in Table S1. In the case of the traditional IEF electrophoresis kit that is generally useful for the two-dimensional separation of target proteins of large size, the relatively smaller peptides used were not discrete in their shift in zone position when three synthetic peptides (T-Pep, T-Pep(p), and T-Pep(pp)) were mixed in a molar ratio of 1:1:1 and loaded into the IEF gel strip (Fig. S1). In addition, the running time of IEF gel electrophoresis should last about 2 h with a high current (around 50 μA). Therefore, we envisioned that μIEF-based analysis of small peptide substrates could provide a new method for discerning the phosphorylation cascade by protein kinases. 3.2. Sequential phosphorylation analysis using
μIEF
To examine this possibility, different protein kinase reactions were conducted with fluorescent peptide substrate and the resultant solution was examined in the μIEF channel (Fig. 1). Unlike Fig. S1, when a mixture of three synthetic peptides (T-Pep, T-Pep (p), and T-Pep(pp)) at a molar ratio of 1:1:1 was loaded into the μIEF channel as a control, they were distinctly separated with similar fluorescent band widths (Fig. 1(i)), where the peptide with the relatively lower pI value was focused at a region closer to the anodic electrode. As a result, a positive linear relationship was observed between the expected pI values (3.8, 4.5, and 6.1) of the three peptides and the respective distances from the anodic electrode to their averaging focusing zones in the μIEF channel (5.8, 8.9, and 12.4 mm, respectively) (Fig. S2). This result indicates that the designated μIEF in the PDMS is sufficient for detecting the monophosphorylation and diphosphorylation of fluorescent peptide substrate. Significantly, when T-Pep was treated with MAPK, a strong shift in zone position to the positive electrode was observed at the same location as that of the synthetic T-Pep(p) (Fig. 1(ii)), whereas GSK3 alone did not change the zone position of T-Pep (Fig. 1(iii)). Remarkably, only when T-Pep was reacted to the mixture of MAPK and GSK3, its fluorescent shift occurred in the far left zone (equal to the location of T-Pep(pp)) (Fig. 1(iv)). In contrast to that of T-Pep in Fig. 1(iii), the zone position of T-Pep(p) was shifted left after the GSK3 reaction (Fig. 1(v)) and was not after the ATP-deficient GSK3 reaction (Fig. 1(vi)). Taken together, this result indicates that the sequential phosphorylation by two protein kinases was easily monitored in the μIEF channel, where MAPK functioned as a priming kinase to first phosphorylate the substrate and GSK3 was responsible for additional phosphorylation after the
primed phosphorylation. It was reported that MAPK can predominantly phosphorylate the second serine residue of the peptide substrate (KEEPPSPPQSPR), which originated from HSF1 (Chu et al., 1996), and it should also be noted that the unique structure of GSK3 allows a positively charged pocket adjacent to its active site to bind a priming phosphate group located four residues toward the C-terminal to the site of GSK3 phosphorylation (S/ TXXXpS/pT) (Dajani et al., 2001; ter Haar et al., 2001). Therefore, the used peptide was proven to be a common substrate for the two protein kinases. This sequential phosphorylation of used peptide substrate was also verified using MALDI-MS, where two strong molecular ion signals shifted by a mass equivalent to that of HPO3 (80 Da) or 2 HPO3 (160 Da) were observed at m/z 1848.97 ([MHþ HPO3] þ ) or m/z 1928.88 ([MH þ2(HPO3)] þ ), where the T-Pep (M, m/z 1767.8) was treated with single kinase (MAPK) or dual kinases (MAPK and GSK3), respectively (Fig. S3). Most importantly, the fluorescence intensity at the focusing zones in the microfluidic channel easily permitted the gathering of quantitative data, thereby allowing for the rapid analysis of the degree of phosphorylation (Fig. S4). According to Eq. (1) and Fig. S4, the phosphorylation efficiencies were calculated to be 76%, 0%, 78%, 81%, and 0% for (ii), (iii), (iv), (v), and (vi) in Fig. 1, respectively. To further examine the dependency of GSK3 on MAPK activity, the isoelectric mobility of T-pep was investigated with various concentrations of MAPK (Fig. 2). As the MAPK concentration increased with a fixed concentration of GSK3, the fluorescence intensity in the diphosphorylated region (T-Pep(pp)) increased with a reduction to that of the original T-Pep. The MAPK-dependent phosphorylation efficiency (EP) was calculated by the intensity ratio of T-Pep(pp) over the summation of T-Pep and T-Pep(pp), according to Eq. (1). The maximal phosphorylated efficiency was not attainable to 100% in Fig. 2(vi), which may be caused by inherently limited efficiency of MAPK for this peptide substrate, as shown in Fig. 1(ii). Nonetheless, this result clearly indicates that MAPK is a priming kinase for the sequential phosphorylation of the target substrate and μIEF is suitable for analyzing primingdependent phosphorylation by consecutive kinases in a quantitative manner. 3.3. Inhibition assay for protein kinase activity using
μIEF
To explore the potential applications of μIEF, we investigated the possibility of conducting an inhibitor assay for the dual protein kinases in the μIEF channel (Fig. 3). To quantify the inhibition effect, the half-maximal inhibitory concentration (IC50) was calculated according to Eq. (2) in Section 2.4. When lithium chloride (LiCl) or FR180204 was used as an inhibitor of GSK3 (Fig. 3A) or MAPK (Fig. 3B), respectively, and each inhibitor was loaded into the microfluidic channel after being added to the reaction solution including kinase and the appropriate peptide substrate (T-Pep
Fig. 2. μIEF-based quantification of the sequential phosphorylation of T-Pep by MAPK-dependent GSK3. The final concentration of GSK3 was fixed at 1.25 U, whereas the concentration of MAPK was varied at 0, 0.25, 0.50, 0,75, 1.50, and 2.25 U (from (i) to (vi)). The phosphorylation efficiency (EP) was calculated from two focusing regions at T-Pep and T-Pep(pp) and represented as a function of the MAPK concentration on the right side.
H. Choi et al. / Biosensors and Bioelectronics 73 (2015) 93–99
97
Fig. 3. μIEF-based inhibition assay of protein kinase activity as a function of inhibitor concentration: (A) GSK3 inhibition assay using lithium chloride (LiCl, 0–400 mM) in the presence of T-Pep(p) and (B) MAPK inhibition assay using FR180204 (0–8 μM) in the presence of T-Pep. The final concentrations of GSK3 and MAPK were 1.25 U in 20 μL of reaction volume. The EP and IC50 were calculated according to Eqs. (1) and (2).
(p) for GSK3 and T-Pep for MAPK), a dose-dependent decrease in EP was significantly observed for both protein kinases over the different range of inhibitor concentrations. The IC50 value of LiCl or FR180204 was determined to be 30.5 mM or 1.3 μM, respectively, both of which were much higher than what was described in a previous report, showing that LiCl inhibited GSK3 with an IC50 of 2.5 mM (Ryves and Harwood, 2001) and FR180204 inhibited MAPK with an IC50 of 0.33 μM (Ohori et al., 2005). It is important to note that lithium (Li þ ) is a competitive inhibitor of GSK3 with respect to magnesium (Mg2 þ ), rather than ATP (Ryves and Harwood, 2001), whereas FR180204 is known to selectively inhibit ERK1 and ERK2 by structural specificity as a competitive inhibitor of ATP (Ohori et al., 2005). Considering the fact that our assay employed relatively higher concentations of Mg2 þ and ATP and lower concentrations of protein kinases than what has been reported in previous methods, the discrepancy in IC50 values was probably due to different competitors, such as Mg2 þ and ATP, as well as different assay format (surface assay vs. solution assay) and reaction time. Despite the distinct roles of the tested inhibitors, we suggest that our μIEF approach is effective at quantifying the inhibition of protein kinase activity. Moreover, this method also enabled us to determine the selectivity of the inhibitor during sequential phosphorylation, which was conducted by adding each inhibitor to the mixture of the two kinases (Fig. 4). Compared to the fluorescent zone when T-Pep was treated with two kinases in
Fig. 4. μIEF-based detection of inhibitor specificity of two protein kinases. T-Pep was reacted with two protein kinases (MAPKþ GSK3) in the absence (i) or presence (ii–iii) of inhibitor (LiCl or FR180204). T-pep(p) was reacted with a single kinase in the presence of FR180204 (iv). The final concentrations of MAPK, GSK3, LiCl, and FR180204 in 20 μL of reaction volume were 1.25 U, 1.25 U, 200 mM, and 8 μM, respectively.
the absence of inhibitors (Fig. 4(i)), the fluorescent zone of T-Pep in the presence of LiCl led to no change in IEF mobility (Fig. 4(ii)), suggesting that LiCi also affected MAPK activity and functioned as a broad inhibitor against both MAPK and GSK3. Similarly, the fluorescent zone in the presence of FR180204 was observed in the T-Pep region due to the inhibition of MAPK (Fig. 4(iii)), but the activity of GSK3 was not affected by this treatment (Fig. 4(iv)), suggesting that FR180204 impeded MAPK activity but not GSK3 activity. This result suggests that the specificity and selectivity of inhibitors are likely to be screened through the μIEF-based sequential phosphorylation analysis. 3.4. Phosphorylation analysis in real sample using immunoprecipitation-combined μIEF To verify the applicability of this method in real samples, we attempted to find evidence of GSK3 activity in cell lysates by the μIEF method combined with immunoprecipitation (IP) (Fig. 5). Despite the results of a previous report on the endogenous activity of GSK3 in MCF-7 cell lysate (Medunjanin et al., 2005), no GSK3 activity in a used amount of cell lysates was detected by IP-combined MALDI-MS using either T-Pep(p) or other phosphate-priming peptides (data not shown). In order to perform a proof-ofconcept experiment for the detection of GSK3 activity in a complex sample by IP-combined μIEF, active GSK3 was exogenously added to soluble cell lysates. Next, IP was initially performed using antiGSK3 antibody-coated magnetic beads to specifically capture the active GSK3 in the cell lysates (Fig. 5A(i)), and the GSK3-captured beads were then washed several times; this process was necessary to remove many interferants that may adversely affect the mobility of the peptide substrate in the μIEF channel. Once the synthetic T-pep(p) that had been incubated with the GSK3-attached beads was transferred into the μIEF channel (Fig. 5A(ii) and A(iii)), a strong fluorescent intensity was observed at the pI zone of T-pep (pp) with a marginal intensity at the T-pep(p) zone, meaning that T-pep(p) was phosphorylated into T-pep(pp) by GSK3 in the cell lysates (Fig. 5B(i)). In contrast, there was no change in the zone position of T-Pep(p) when using the control cell lysate without active GSK3 (Fig. 5B(ii)) or using bare magnetic beads without
98
H. Choi et al. / Biosensors and Bioelectronics 73 (2015) 93–99
fluorescent focusing zones in the μIEF channel, where the priming phosphorylation and priming-dependent phosphorylation were sequentially determined in a highly quantitative manner. In addition, this method resulted in detecting the inhibitory effect of protein kinases with high selectivity as well as assaying the protein kinase activity in cell lysates in combination with IP. Owing to its simplicity and general applicability with low sample volumes, we anticipate that our approach opens up new ways to elucidate the sequential phosphorylation of many target substrates as well as investigate the physiological roles of protein kinases in a highthroughput manner.
Acknowledgment This work was supported by Mid-career Researcher Program (Nos. 2013R1A2A2A03015161 and 2013R1A2A2A01014234) and Nano Material Technology Development Program (No. 2012M3A7B4035286) through the National Research Foundation (NRF) funded by the Ministry of Science, ICT, and Future Planning (MSIP). This work was also supported by Basic Science Research Program (No. 2012R1A6A1029029) through the NRF funded by the Ministry of Education.
Appendix A. Supplementary material Fig. 5. (A) Schematic of immunoprecipitation (IP)-combined μIEF for the detection of GSK3 activity in cell lysates. Anti-GSK3 monoclonal antibody on Protein A/G magnetic beads was used to capture GSK3 in MCF-7 cell lysates and peptide substrate (T-Pep(p)) was then reacted to the GSK3-captured beads. (B) Fluorescent μIEF images of T-Pep(p) under the different reaction conditions: i) immunoprecipitated GSK3 for positive cell lysates, ii) immunoprecipitated GSK3 for negative cell lysates, and iii) immunoprecipitated GSK3 without anti-GSK3 antibody.
anti-GSK3 antibody (Fig. 5B(iii)). This evidence suggests that this method would be very suitable for the in vitro analysis of GSK3 activity in cell lysates. Based on these results, our μIEF electrophoresis has many advantages over conventional IEF gel electrophoresis or traditional kinase assays. These include faster assay time, better reproducibility, more quantitative analysis, and smaller sample consumption for the detection of protein kinase reactions and their inhibition. In particular, diverse degrees of phosphorylation for peptide substrates was reliably separated in μIEF, allowing for the straightforward analysis of sequential phosphorylation in a quantitative manner, that otherwise might not be possible using a traditional immunoblotting method. Moreover, as a result of being coupled to multiple microfluidic functionalities, the strengths of this method include multiplexibility, functional integration and automation (Chovan and Guttman, 2002; Reyes et al., 2002). Given the advancements in microfluidic devices directly connected with IP (Wu et al., 2012) and/or single cell analysis (Hughes et al., 2014; Junkin and Tay, 2014; Song et al., 2014), this μIEF method is anticipated to elucidate phosphorylation cascades in association with multiple protein kinases and can be applied for the analyses of other enzyme activities and relevant modifications.
4. Conclusions We demonstrated sequential phosphorylation analysis through microfluidic separation based on the pI of dye-labeled peptide substrates. Despite the subtle modification of peptide substrates by two protein kinases (MAPK and GSK3), the degree of phosphorylation of the peptide was remarkably discriminated by
Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2015.05.047.
References Anderson, J.C., Peck, S.C., 2008. Plant J. 55, 881–885. Chovan, T., Guttman, A., 2002. Trends Biotechnol. 20, 116–122. Chu, B.Y., Soncin, F., Price, B.D., Stevenson, M.A., Calderwood, S.K., 1996. J. Biol. Chem. 271, 30847–30857. Cohen, P., 2002. Nat. Cell Biol. 4, E127–E130. Cui, H., Horiuchi, K., Dutta, P., Ivory, C.F., 2005. Anal. Chem. 77, 1303–1309. Dajani, R., Fraser, E., Roe, S.M., Young, N., Good, V., Dale, T.C., Pearl, L.H., 2001. Cell 105, 721–732. Das, C., Zhang, J., Denslow, N.D., Fan, Z.H., 2007. Lab Chip 7, 1806–1812. Doble, B.W., Woodgett, J.R., 2003. J. Cell Sci. 116, 1175–1186. Eldar-Finkelman, H., 2002. Trends Mol. Med. 8, 126–132. Ghadiali, J.E., Cohen, B.E., Stevens, M.M., 2010. ACS Nano 4, 4915–4919. Harvey, C.D., Ehrhardt, A.G., Cellurale, C., Zhong, H.N., Yasuda, R., Davis, R.J., Svoboda, K., 2008. Proc. Natl. Acad. Sci. USA 105, 19264–19269. Hastie, C.J., McLauchlan, H.J., Cohen, P., 2006. Nat. Protoc. 1, 968–971. He, B., Meng, Y.H., Mivechi, N.F., 1998. Mol. Cell. Biol. 18, 6624–6633. Hughes, A.J., Spelke, D.P., Xu, Z.C., Kang, C.C., Schaffer, D.V., Herr, A.E., 2014. Nat. Methods 11, 749–755. Hunter, T., 1995. Cell 80, 225–236. Junkin, M., Tay, S., 2014. Lab Chip 14, 1246–1260. Kim, Y.P., Oh, E., Oh, Y.H., Moon, D.W., Lee, T.G., Kim, H.S., 2007. Angew. Chem. Int. Ed. 46, 6816–6819. Kinoshita, E., Kinoshita-Kikuta, E., Matsubara, M., Aoki, Y., Ohie, S., Mouri, Y., Koike, T., 2009. Electrophoresis 30, 550–559. Kupcho, K., Somberg, R., Bulleit, B., Goueli, S.A., 2003. Anal. Biochem. 317, 210–217. Mann, M., Ong, S.E., Gronborg, M., Steen, H., Jensen, O.N., Pandey, A., 2002. Trends Biotechnol. 20, 261–268. Medunjanin, S., Hermani, A., De Servi, B., Grisouard, J., Rincke, G., Mayer, D., 2005. J. Biol. Chem. 280, 33006–33014. O'Neill, R.A., Bhamidipati, A., Bi, X.H., Deb-Basu, D., Cahill, L., Ferrante, J., Gentalen, E., Glazer, M., Gossett, J., Hacker, K., Kirby, C., Knittle, J., Loder, R., Mastroieni, C., MacLaren, M., Mills, T., Nguyen, U., Parker, N., Rice, A., Roach, D., Suich, D., Voehringer, D., Voss, K., Yang, J., Yang, T., Vander Horn, P.B., 2006. Proc. Natl. Acad. Sci. USA 103, 16153–16158. Ohori, M., Kinoshita, T., Okubo, M., Sato, K., Yamazaki, A., Arakawa, H., Nishimura, S., Inamura, N., Nakajima, H., Neya, M., Miyake, H., Fujii, T., 2005. Biochem. Biophys. Res. Commun. 336, 357–363. Peck, S.C., 2006. Plant J. 45, 512–522. Reyes, D.R., Iossifidis, D., Auroux, P.A., Manz, A., 2002. Anal. Chem. 74, 2623–2636. Riddle, S.M., Vedvik, K.L., Hanson, G.T., Vogel, K.W., 2006. Anal. Biochem. 356, 108–116. Ryves, W.J., Harwood, A.J., 2001. Biochem. Biophys. Res. Commun. 280, 720–725.
H. Choi et al. / Biosensors and Bioelectronics 73 (2015) 93–99
Salih, E., 2005. Mass Spectrom. Rev. 24, 828–846. Schenk, P.W., Snaar-Jagalska, B.E., 1999. Biochim. Biophys. Acta-Mol. Cell Res. 1449, 1–24. Shiosaki, S., Nobori, T., Mori, T., Toita, R., Nakamura, Y., Kim, C.W., Yamamoto, T., Niidome, T., Katayama, Y., 2013. Chem. Commun. 49, 5592–5594. Sommer, G.J., Hatch, A.V., 2009. Electrophoresis 30, 742–757. Song, K., Jeong, H.H., Jin, S.H., Park, J.S., Lee, C.S., 2014. Biochip J. 8, 227–233. ter Haar, E., Coll, J.T., Austen, D.A., Hsiao, H.M., Swenson, L., Jain, J., 2001. Nat. Struct. Mol. Biol. 8, 593–596. Walsh, M.P., MacDonald, J.A., 2011. Arch. Biochem. Biophys. 510, 73–75.
99
Wen, J., Wilker, E.W., Yaffe, M.B., Jensen, K.F., 2010. Anal. Chem. 82, 1253–1260. Wu, A.R., Kawahara, T.L.A., Rapicavoli, N.A., van Riggelen, J., Shroff, E.H., Xu, L.W., Felsher, D.W., Chang, H.Y., Quake, S.R., 2012. Lab Chip 12, 2190–2198. Yamamoto, T., Matsuzaki, H., Kamada, S., Ono, Y., Kikkawa, U., 2006. Nat. Protoc. 1, 2791–2795. Zhu, H., Klemic, J.F., Chang, S., Bertone, P., Casamayor, A., Klemic, K.G., Smith, D., Gerstein, M., Reed, M.A., Snyder, M., 2000. Nat. Genet. 26, 283–289.
Contents lists available at ScienceDirect
Biosensors and Bioelectronics journal homepage: www.elsevier.com/locate/bios
Sequential phosphorylation analysis using dye-tethered peptides and microfluidic isoelectric focusing electrophoresis Hoseok Choi a,b,1, Nakchul Choi c,1, Butaek Lim a,b,1, Tae-Wuk Kim a,b, Simon Song c,d,n, Young-Pil Kim a,b,d,nn a
Department of Life Science, Hanyang University, Seoul 133-791, Republic of Korea Research Institute for Natural Sciences, Hanyang University, Seoul 133-791, Republic of Korea c Department of Mechanical Convergence Engineering, Hanyang University, Seoul 133-791, Republic of Korea d Institute of Nano Science and Technology, Hanyang University, Seoul 133-791, Republic of Korea b
art ic l e i nf o
a b s t r a c t
Article history: Received 28 February 2015 Received in revised form 11 May 2015 Accepted 21 May 2015 Available online 25 May 2015
We report a simple method for analyzing sequential phosphorylation by protein kinases using fluorescent peptide substrates and microfluidic isoelectric focusing (μIEF) electrophoresis. When a dye-labeled peptide substrate was sequentially phosphorylated by two consecutive protein kinases (mitogenactivated protein kinase (MAPK) and glycogen synthase kinase 3 (GSK3)), its differently phosphorylated forms were easily separated and visualized by fluorescent focusing zones in the μIEF channel based on a change in the isoelectric point (pI) by phosphorylation. As a result, ratiometric and quantitative analysis of the fluorescent focusing regions shifted by phosphorylation enabled the analysis of phosphorylation efficiency and the relevant inhibition of protein kinases (MAPK and GSK3) with high simplicity and selectivity. Furthermore, the GSK3 activity in the cell lysates was elucidated by μIEF electrophoresis in combination with immunoprecipitation. Our results suggest that this method has great potential for analyzing the sequential phosphorylation of multiple protein kinases that are implicated in cellular signaling pathways. & 2015 Elsevier B.V. All rights reserved.
Keywords: Sequential phosphorylation Microfluidic Isoelectric focusing Protein kinase Inhibition assay Immunoprecipitation
1. Introduction Phosphorylation cascades involving multiple protein kinases play a central role in signal transduction, protein regulation, and metabolism in living cells (Cohen, 2002; Hunter, 1995; Schenk and Snaar-Jagalska, 1999; Walsh and MacDonald, 2011). The analysis of these phosphorylation cascades, therefore, will provide new insights into their physiological functions in many biological events. While general methods for probing protein kinase activity have adopted [gamma-32P]ATP-labeled substrate in electrophoresis or non-radioisotopic immunoblotting to differentiate between phosphorylated and unphosphorylated substrates (Hastie et al., 2006; Peck, 2006; Yamamoto et al., 2006; Zhu et al., 2000), they n Corresponding author at: Department of Mechanical Convergence Engineering, Hanyang University, Seoul 133-791, Republic of Korea. Tel.: þ 82 2 2220 0423. nn Corresponding author at: Department of Life Science and Research Institute for Natural Sciences, Hanyang University, Seoul 133-791, Republic of Korea. Tel.: þ 82 2 2220 2560. E-mail addresses: [email protected] (S. Song), [email protected] (Y.-P. Kim). 1 These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.bios.2015.05.047 0956-5663/& 2015 Elsevier B.V. All rights reserved.
are not only based on long processing times, but they are limited to the rapid identification of a cascade event involving multiple kinases or the determination of the quantitative amount of phosphorylation. Fluorogenic or fluorescence resonance energy transfer-based methods using peptide substrates have been alternatively used for monitoring various phosphorylation events, which are based on the fluorescent change from substrate phosphorylation using dye-labeled antibodies or chemicals (Ghadiali et al., 2010; Harvey et al., 2008; Kupcho et al., 2003; Riddle et al., 2006; Shiosaki et al., 2013), but these are mostly used for single kinase reactions. In addition, mass spectrometric analysis has been implemented as a standard means for identifying the degree of phosphorylation (Kim et al., 2007; Mann et al., 2002; Salih, 2005), which a time-consuming and expensive process. To circumvent these issues, increasing attention has been paid to isoelectric focusing (IEF) technique because it is capable of easily separating diverse phosphorylated proteins according to their isoelectric point (pI) based on the negatively accumulating increase in net charge by sequential addition of phosphate group to the protein (Anderson and Peck, 2008). Although capillary-type IEF and two-dimensional IEF gel electrophoresis have been reported to be useful for separating phosphorylated proteins (Kinoshita et al., 2009; O'Neill et al., 2006), fewer efforts have been
94
H. Choi et al. / Biosensors and Bioelectronics 73 (2015) 93–99
made to analyze phosphorylation cascade, especially in a rapid and quantitative manner. Given a common substrate for multiple kinases, the IEF technique would allow for the phosphorylation cascades of the substrate to be easily resolved during multiple kinase reactions. Here we report a simple method for analyzing sequential phosphorylation using fluorescent peptide substrates and microfluidic isoelectric focusing (μIEF) electrophoresis. Despite recent advances in μIEF techniques for biological analyses (Das et al., 2007; Sommer and Hatch, 2009; Wen et al., 2010), many attempts have focused on the easy separation of complex biomolecules according to the average pIs. To the best of our knowledge, there have been no attempts in integrating the μIEF technique for the simultaneous detection of cascade phosphorylation. To demonstrate the sequential phosphorylation event, we chose two consecutive protein kinases: p42 mitogen-activated protein kinase (MAPK), also known as extracellular signal-regulated kinase 2 (ERK2), and glycogen synthase kinase 3 (GSK3). Since MAPK is considered to be one of the priming kinases for GSK3 in a process by which target substrates are sequentially phosphorylated (EldarFinkelman, 2002), we examined the phosphorylation cascade using an annotated peptide substrate from heat shock transcription factor 1 (HSF1), which is consecutively controlled by two protein kinases (Chu et al., 1996; He et al., 1998). Upon sequential phosphorylation, a dye-labeled peptide substrate gave rise to a mono- or di-phosphorylated form, allowing for easy separation and rapid detection in a polydimethylsiloxane (PDMS) μIEF channel equipped with an imaging analyzer. Moreover, the activity and inhibition of the protein kinases were quantitatively investigated by comparing two or three separated lines of the peptides on the μIEF. In contrast to conventional IEF electrophoresis, a microfluidic-based format enabled more reliable analyses with rapid assay time and much smaller reaction volumes (a few microliters).
2. Materials and methods 2.1. Materials P42 mitogen-activated protein kinase (MAPK) and glycogen synthase kinase 3 (GSK3) was purchased from New England Biolabs. Anti-human GSK3 beta polyclonal antibody, a protein G magnetic bead and a C18 pipette-tip column were purchased from Thermo Fisher Scientific Inc. (USA). RIPA buffer was purchased from Cell Signaling Technology. Methylcellulose (viscosity 400 cP, MC), lithium chloride (99%, LiCl), adenosine 5′-triphosphate (ATP) disodium salt hydrate (99%, 5’-ATP-Na2), α-cyano-4-hydroxycinnamic acid (98%, CHCA) and trifluoroacetic acid (99%, TFA) were purchased from Sigma-Aldrich. Acetonitrile (ACN, HPLC grade) was purchased from Merck. Ampholyte (Pharmalyte, broad range pH 3–10) was purchased from GE Healthcare Life Sciences. FR180204 was purchased from Santa Cruz Biotechnology Inc. (USA). Polydimethylsiloxane (PDMS) was purchased from Dow Corning. Isopropyl alcohol (99.5%, IPA) was purchased from Daejung. SU-8, a photoresist (PR), and SU-8 developer (PGMEA) was purchased from MicroChemicals. Tetramethyl-6-carboxyrhodamine (TAMRA)-labeled peptides (T-Pep, TAMRAKEEPPSPPQSPR; T-Pep(p), TAMRA-KEEPPSPPQpSPR; T-Pep(pp), TAMRA-KEEPPpSPPQpSPR) were synthesized from Peptron, Inc. (Korea). All other chemicals were of analytical grade and were used as received. 2.2. Fabrication of a microfluidic isoelectric focusing (μIEF) device A
microfluidic
chip
was
fabricated
by
bonding
Scheme 1. (A) Schematic of IEF electrophoresis in a PDMS microfluidic device equipped with fluorescent microscopy and imaging analysis. (B) Illustration of sequential phosphorylation of TAMRA-labeled peptide substrate (T-Pep) by two consecutive protein kinases (MAPK and GSK3), leading to the reduction in its pI value.
polydimethylsiloxane (PDMS) substrate on a flat PDMS substrate. A PDMS substrate was prepared by using a standard molding technique and soft lithography. To make a PDMS mold, a PR material was uniformly spin-coated at a thickness of 20 μm on a 4 in. Si wafer that was sequentially cleaned with acetone, IPA, and deionized (DI) water. The coated wafer was soft-baked for 3.5 min at 95 °C. Then it was exposed to 365 nm UV light (MDA-400M, Midas) through a film-type photo mask having a pattern of straight channel. In order to enhance the crosslinking of PR, a postexposure-baking of the wafer was performed for 4.5 min at 95 °C. The wafer was developed for 3.5 min in a solution of PGMEA. The wafer was washed with IPA to remove unexposed PR and dried with nitrogen gas. To fabricate the PDMS substrate, a mixture of PDMS (DC-184A) and curing agent (DC-184B) at a ratio of 10 to 1 was poured into the fabricated mold and degassed for 30 min and cured for 2 h at 70 °C in a vacuum oven (JSVO-60T, JSR). This PDMS substrate was bonded with a flat PDMS substrate after exposure to oxygen plasma (Plasma prep 300, Nanotech). Finally, the PDMS channel surface was coated with MC to make it hydrophilic for the purpose of reducing electroosmotic flow (EOF). The microfluidic chip had a straight channel with a length of 3 cm, a width of 300 μm and a height of 20 μm as shown in Scheme 1A. The reservoirs were attached on the inlet and outlet for the injection of electrolytes. 2.3. μIEF-based analysis of sequential phosphorylation by protein kinases Typically, protein kinase reactions were performed in a tube prior to μIEF analysis. For the single kinase reaction, appropriate TAMRA-labeled peptide substrate (1 μL at 100 μM), MAPK or GSK3 (1 μL at 25 U), ATP (0.4 μL at 10 mM), and 10 kinase reaction buffer (2 μL) were mixed at a final volume of 20 μL in standard reaction buffer (20 mM Tris–HCl, pH 7.4). For the sequential protein kinase reaction, the equivalent mixture (each 1 μL at 25 U) of MAPK and GSK3 was added to the same reaction solution. All kinase reactions were typically run for 90 min at 30 °C, followed by a thermal inactivation process for 20 min at 60 °C in order to load the aliquot into the PDMS microfluidic channel under the same condition. As a control, three peptides (T-Pep, T-Pep(p), and T-Pep (pp)) were mixed at a 1:1:1 molar ratio (each 1 μL at 100 μM) at a
H. Choi et al. / Biosensors and Bioelectronics 73 (2015) 93–99
final volume of 20 μL in reaction buffer (20 mM Tris–HCl, pH 7.4). A typical 15 μL of each reaction was initially mixed with ampholyte (15 μL at 12% w/w DI-water) and MC (15 μL at 1.2% w/w DIwater) and subsequently a 2 μL aliquot of the mixture solution was loaded into the microfluidic channel through an anodic reservoir prior to electrophoresis. The final loading concentrations of peptide, ampholyte, and MC were 1.7 μM, 4% w/w DI-water, and 0.4% w/w DI-water, respectively. The anodic and cathodic reservoirs were then placed with 50 mM H3PO4 in 0.4% w/w MC as the anolyte and 50 mM NaOH in 0.4% w/w MC as the catholyte, respectively. Platinum wire electrodes were connected to the reservoirs, and focusing was carried out at constant electric potentials up to 100 V cm 1 using a power supply (PS350, SRS). After focusing was complete for 10 min, the fluorescent image of the channel was acquired from an inverted microscope (IX71, Olympus) equipped with a CCD camera (DP71, Olympus). For the kinase inhibition assay in the μIEF channel, a stock solution of LiCl or FR180204 was serially diluted to different concentrations in the reaction buffer, and an equal volume (2 μL) of each inhibitor concentration was then mixed with a solution containing peptide substrate, protein kinase, and ATP dissolved in the reaction buffer. At a 20 μL reaction, the final concentrations of peptide, protein kinase, and ATP were 5 μM, 1.25 U, and 200 μM, respectively. After incubation for 90 min at 30 °C, the reactant was followed by a thermal inactivation process for 20 min at 60 °C. Each 15 μL of reaction samples with different concentrations of inhibitor was initially mixed with ampholyte (15 μL at 12%) and MC (15 μL at 1.2%) and then a 2 μL aliquot of the mixture solution was loaded into the microfluidic channel for IEF electrophoresis. 2.4. Determination of phosphorylation efficiency and IC50
95
0.5% TFA/50% ACN. Elution of the target peptide from the C18 tip was conducted directly on a standard stainless steel MALDI target by dispensing about 0.7 μL of the matrix solution containing 1 mg of CHCA in 0.5% TFA/50% ACN. Mass spectrometric analysis of the peptides was performed using an Axima-CFR (Shimadzu). 2.6.
μIEF-based analysis of the protein kinase activity in cell lysates
MCF-7 cells grown in the 75 cm2-culture flask were washed three times with phosphate-buffered saline (PBS, pH 7.4), and collected in micro-centrifuge tubes by scraping in 2 mL of RIPA buffer (25 mM Tris–HCl, pH 7.6, 150 mM NaCl, 1% NP-40, 1% sodium deoxycholate and 0.1% sodium dodecyl sulfate). Cells were lysed for 10 min in an ice bath with intermittent vortexing. The cytosolic fraction was obtained by centrifuging the cell lysate at 13,000 rpm for 20 min and the supernatant was collected in new tubes. For immunoprecipitation (IP), protein A/G magnetic beads (25 μL) were coated with polyclonal anti-GSK3 antibody (5 μg) for 12 h at 4 °C, and then incubated with 100 μL of the supernatant (with or without spiking GSK3) in binding buffer (50 mM Tris–HCl, pH7.5, NaCl 150 mM NaCl, 0.2% NP-40 and 0.05 mg mL 1 BSA) for 2 h at 4 °C on a gently rotating device. Magnetic beads were washed three times in washing buffer (50 mM Tris–HCl, pH7.5, 100 mM NaCl and 0.1% NP-40) using a magnetic stand and resuspended in kinase reaction buffer. For the kinase reaction, the immunoprecipitated magnetic beads (10 μL) were further incubated with T-Pep(p) (1 μL at 100 μM) for 90 min at 30 °C, followed by a thermal inactivation process for 20 min at 60 °C. Each 10 μL of the reactants under different IP conditions was initially mixed with ampholyte (10 μL at 12%) and MC (10 μL at 1.2%) and then a 2 μL aliquot of the mixture solution was loaded into the microfluidic channel for IEF electrophoresis.
The phosphorylation efficiency of the kinase reaction was calculated using the following equation;
EP (%) =
IZone (p) IZone + IZone (p)
3. Results and discussion
× 100
(1)
where EP is the phosphorylation efficiency (%), IZone and IZone(p) are the fluorescent intensities of the control focusing zone in the microfluidic channel before the kinase reaction and the phosphorylated focusing zone after the kinase reaction, respectively. The fluorescent intensity of the focusing zones was determined by imaging analysis using a Matlab software (ver 2012, The MathWorks, Inc.) For the IC50, the EP was plotted as a function of inhibitor concentration, and then fitted to a 4-parameter logistic equation (dose–response model for ligand binding) by the nonlinear regression procedure of Origin Pro (ver 8.0, OriginLab Corporation) using the following equation:
EP (%, for inhibition) =
EPmax − EPmin + EPmin 1 + (C /IC50 )n
(2)
where EPmax and EPmin are the maximum and minimum phosphorylation efficiencies, respectively, at a given kinase-substrate reaction, IC50 is the median inhibitory concentration, C is the inhibitor concentration, and n is the slope factor. 2.5. Matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS) analysis A C18 pipette-tip was used to concentrate and desalt the peptide substrates with or without the kinase reaction according to the manufacturer's specifications. The C18 tip was wetted in 0.5% TFA/50% ACN and activated in 0.5% TFA/distilled water. A 10 μL of the sample reactant was adsorbed onto the C18 tip and rinsed in
3.1. Fabrication and detection principle Scheme 1A illustrates a simple microfluidic channel in a PDMS for IEF-based separation of the sequentially phosphorylated peptide. While 0.4% methylcellulose (MC) was used with the anodic or cathodic electrolyte to reduce electroosmosis and peak drift in IEF by applying a dynamic coating of MC to channel walls, as reported earlier (Cui et al., 2005), fluorescent peptide substrates were mixed with 4% carrier ampholytes and 0.4% MC, and poured into the PDMS microfluidic channel through an anodic electrolyte reservoir under a high electric field, leading to migration and focusing into discrete zones in areas where the local pH was equal to the pI of the analyte. The resulting fluorescent zones were then imaged by the fluorescence microscope equipped with a CCD camera. As depicted in Scheme 1B, we employed a model peptide substrate labeled with TAMRA dye at its N-terminus (TAMRAKEEPPSPPQSPR, T-Pep). The synthetic peptide substrate (T-Pep) was then be sequentially phosphorylated by protein kinases (MAPK and GSK3) to its diphosphorylated form (TAMRAKEEPPpSPPQpSPR, T-Pep(pp)) through its monophosphorylated form (TAMRA-KEEPPSPPQpSPR, T-Pep(p)), which is based on the reported kinase signaling pathway of HSF1 possessing the peptide sequence (Chu et al., 1996). It is noteworthy that GSK3 is constitutively regulated by other protein kinases, which is central to multiple signaling pathways in association with a variety of extracellular stimuli (Doble and Woodgett, 2003); therefore, sequential phosphorylation by GSK3 activity and its coupled kinases (MAPK here) is very crucial to understand its multifunctional role. Inspired by this observation, the zone position by sequential phosphorylation of T-Pep was investigated in the μIEF channel to
96
H. Choi et al. / Biosensors and Bioelectronics 73 (2015) 93–99
Fig. 1. Fluorescent IEF images of peptide substrates obtained in a microfluidic PDMS channel under the different reaction conditions: (i) mixture of synthetic T-Pep, T-Pep(p), and T-Pep(pp) at a 1:1:1 molar ratio as a control, (ii) T-Pep þ MAPK, (iii) T-Pep þGSK3, (iv) T-Pep þ MAPKþ GSK3, (v) T-Pep(p)þ GSK3, and (vi) T-Pep (p) þGSK3 without ATP. The protein kinase reaction was performed by incubating the appropriate peptide with protein kinase(s) for 90 min at 30 °C in reaction buffer including 100 μM ATP, 10 mM MgCl2, 0.1 mM EDTA, 2 mM DTT and 0.01% Brij 35 in 20 mM Tris–HCl (pH 7.4). For IEF electrophoresis, 4% ampholyte and 0.2% MC was loaded into the μIEF channel with each sample. Unphosphorylated and phosphorylated peptides were displayed above the microfluidic channel, which were matched with fluorescent focusing zones based on their corresponding pI.
ascertain the significant change in its pI. The expected pI values of the modified peptides were calculated from the mean pKa values for the phosphorylated amino acids in Table S1. In the case of the traditional IEF electrophoresis kit that is generally useful for the two-dimensional separation of target proteins of large size, the relatively smaller peptides used were not discrete in their shift in zone position when three synthetic peptides (T-Pep, T-Pep(p), and T-Pep(pp)) were mixed in a molar ratio of 1:1:1 and loaded into the IEF gel strip (Fig. S1). In addition, the running time of IEF gel electrophoresis should last about 2 h with a high current (around 50 μA). Therefore, we envisioned that μIEF-based analysis of small peptide substrates could provide a new method for discerning the phosphorylation cascade by protein kinases. 3.2. Sequential phosphorylation analysis using
μIEF
To examine this possibility, different protein kinase reactions were conducted with fluorescent peptide substrate and the resultant solution was examined in the μIEF channel (Fig. 1). Unlike Fig. S1, when a mixture of three synthetic peptides (T-Pep, T-Pep (p), and T-Pep(pp)) at a molar ratio of 1:1:1 was loaded into the μIEF channel as a control, they were distinctly separated with similar fluorescent band widths (Fig. 1(i)), where the peptide with the relatively lower pI value was focused at a region closer to the anodic electrode. As a result, a positive linear relationship was observed between the expected pI values (3.8, 4.5, and 6.1) of the three peptides and the respective distances from the anodic electrode to their averaging focusing zones in the μIEF channel (5.8, 8.9, and 12.4 mm, respectively) (Fig. S2). This result indicates that the designated μIEF in the PDMS is sufficient for detecting the monophosphorylation and diphosphorylation of fluorescent peptide substrate. Significantly, when T-Pep was treated with MAPK, a strong shift in zone position to the positive electrode was observed at the same location as that of the synthetic T-Pep(p) (Fig. 1(ii)), whereas GSK3 alone did not change the zone position of T-Pep (Fig. 1(iii)). Remarkably, only when T-Pep was reacted to the mixture of MAPK and GSK3, its fluorescent shift occurred in the far left zone (equal to the location of T-Pep(pp)) (Fig. 1(iv)). In contrast to that of T-Pep in Fig. 1(iii), the zone position of T-Pep(p) was shifted left after the GSK3 reaction (Fig. 1(v)) and was not after the ATP-deficient GSK3 reaction (Fig. 1(vi)). Taken together, this result indicates that the sequential phosphorylation by two protein kinases was easily monitored in the μIEF channel, where MAPK functioned as a priming kinase to first phosphorylate the substrate and GSK3 was responsible for additional phosphorylation after the
primed phosphorylation. It was reported that MAPK can predominantly phosphorylate the second serine residue of the peptide substrate (KEEPPSPPQSPR), which originated from HSF1 (Chu et al., 1996), and it should also be noted that the unique structure of GSK3 allows a positively charged pocket adjacent to its active site to bind a priming phosphate group located four residues toward the C-terminal to the site of GSK3 phosphorylation (S/ TXXXpS/pT) (Dajani et al., 2001; ter Haar et al., 2001). Therefore, the used peptide was proven to be a common substrate for the two protein kinases. This sequential phosphorylation of used peptide substrate was also verified using MALDI-MS, where two strong molecular ion signals shifted by a mass equivalent to that of HPO3 (80 Da) or 2 HPO3 (160 Da) were observed at m/z 1848.97 ([MHþ HPO3] þ ) or m/z 1928.88 ([MH þ2(HPO3)] þ ), where the T-Pep (M, m/z 1767.8) was treated with single kinase (MAPK) or dual kinases (MAPK and GSK3), respectively (Fig. S3). Most importantly, the fluorescence intensity at the focusing zones in the microfluidic channel easily permitted the gathering of quantitative data, thereby allowing for the rapid analysis of the degree of phosphorylation (Fig. S4). According to Eq. (1) and Fig. S4, the phosphorylation efficiencies were calculated to be 76%, 0%, 78%, 81%, and 0% for (ii), (iii), (iv), (v), and (vi) in Fig. 1, respectively. To further examine the dependency of GSK3 on MAPK activity, the isoelectric mobility of T-pep was investigated with various concentrations of MAPK (Fig. 2). As the MAPK concentration increased with a fixed concentration of GSK3, the fluorescence intensity in the diphosphorylated region (T-Pep(pp)) increased with a reduction to that of the original T-Pep. The MAPK-dependent phosphorylation efficiency (EP) was calculated by the intensity ratio of T-Pep(pp) over the summation of T-Pep and T-Pep(pp), according to Eq. (1). The maximal phosphorylated efficiency was not attainable to 100% in Fig. 2(vi), which may be caused by inherently limited efficiency of MAPK for this peptide substrate, as shown in Fig. 1(ii). Nonetheless, this result clearly indicates that MAPK is a priming kinase for the sequential phosphorylation of the target substrate and μIEF is suitable for analyzing primingdependent phosphorylation by consecutive kinases in a quantitative manner. 3.3. Inhibition assay for protein kinase activity using
μIEF
To explore the potential applications of μIEF, we investigated the possibility of conducting an inhibitor assay for the dual protein kinases in the μIEF channel (Fig. 3). To quantify the inhibition effect, the half-maximal inhibitory concentration (IC50) was calculated according to Eq. (2) in Section 2.4. When lithium chloride (LiCl) or FR180204 was used as an inhibitor of GSK3 (Fig. 3A) or MAPK (Fig. 3B), respectively, and each inhibitor was loaded into the microfluidic channel after being added to the reaction solution including kinase and the appropriate peptide substrate (T-Pep
Fig. 2. μIEF-based quantification of the sequential phosphorylation of T-Pep by MAPK-dependent GSK3. The final concentration of GSK3 was fixed at 1.25 U, whereas the concentration of MAPK was varied at 0, 0.25, 0.50, 0,75, 1.50, and 2.25 U (from (i) to (vi)). The phosphorylation efficiency (EP) was calculated from two focusing regions at T-Pep and T-Pep(pp) and represented as a function of the MAPK concentration on the right side.
H. Choi et al. / Biosensors and Bioelectronics 73 (2015) 93–99
97
Fig. 3. μIEF-based inhibition assay of protein kinase activity as a function of inhibitor concentration: (A) GSK3 inhibition assay using lithium chloride (LiCl, 0–400 mM) in the presence of T-Pep(p) and (B) MAPK inhibition assay using FR180204 (0–8 μM) in the presence of T-Pep. The final concentrations of GSK3 and MAPK were 1.25 U in 20 μL of reaction volume. The EP and IC50 were calculated according to Eqs. (1) and (2).
(p) for GSK3 and T-Pep for MAPK), a dose-dependent decrease in EP was significantly observed for both protein kinases over the different range of inhibitor concentrations. The IC50 value of LiCl or FR180204 was determined to be 30.5 mM or 1.3 μM, respectively, both of which were much higher than what was described in a previous report, showing that LiCl inhibited GSK3 with an IC50 of 2.5 mM (Ryves and Harwood, 2001) and FR180204 inhibited MAPK with an IC50 of 0.33 μM (Ohori et al., 2005). It is important to note that lithium (Li þ ) is a competitive inhibitor of GSK3 with respect to magnesium (Mg2 þ ), rather than ATP (Ryves and Harwood, 2001), whereas FR180204 is known to selectively inhibit ERK1 and ERK2 by structural specificity as a competitive inhibitor of ATP (Ohori et al., 2005). Considering the fact that our assay employed relatively higher concentations of Mg2 þ and ATP and lower concentrations of protein kinases than what has been reported in previous methods, the discrepancy in IC50 values was probably due to different competitors, such as Mg2 þ and ATP, as well as different assay format (surface assay vs. solution assay) and reaction time. Despite the distinct roles of the tested inhibitors, we suggest that our μIEF approach is effective at quantifying the inhibition of protein kinase activity. Moreover, this method also enabled us to determine the selectivity of the inhibitor during sequential phosphorylation, which was conducted by adding each inhibitor to the mixture of the two kinases (Fig. 4). Compared to the fluorescent zone when T-Pep was treated with two kinases in
Fig. 4. μIEF-based detection of inhibitor specificity of two protein kinases. T-Pep was reacted with two protein kinases (MAPKþ GSK3) in the absence (i) or presence (ii–iii) of inhibitor (LiCl or FR180204). T-pep(p) was reacted with a single kinase in the presence of FR180204 (iv). The final concentrations of MAPK, GSK3, LiCl, and FR180204 in 20 μL of reaction volume were 1.25 U, 1.25 U, 200 mM, and 8 μM, respectively.
the absence of inhibitors (Fig. 4(i)), the fluorescent zone of T-Pep in the presence of LiCl led to no change in IEF mobility (Fig. 4(ii)), suggesting that LiCi also affected MAPK activity and functioned as a broad inhibitor against both MAPK and GSK3. Similarly, the fluorescent zone in the presence of FR180204 was observed in the T-Pep region due to the inhibition of MAPK (Fig. 4(iii)), but the activity of GSK3 was not affected by this treatment (Fig. 4(iv)), suggesting that FR180204 impeded MAPK activity but not GSK3 activity. This result suggests that the specificity and selectivity of inhibitors are likely to be screened through the μIEF-based sequential phosphorylation analysis. 3.4. Phosphorylation analysis in real sample using immunoprecipitation-combined μIEF To verify the applicability of this method in real samples, we attempted to find evidence of GSK3 activity in cell lysates by the μIEF method combined with immunoprecipitation (IP) (Fig. 5). Despite the results of a previous report on the endogenous activity of GSK3 in MCF-7 cell lysate (Medunjanin et al., 2005), no GSK3 activity in a used amount of cell lysates was detected by IP-combined MALDI-MS using either T-Pep(p) or other phosphate-priming peptides (data not shown). In order to perform a proof-ofconcept experiment for the detection of GSK3 activity in a complex sample by IP-combined μIEF, active GSK3 was exogenously added to soluble cell lysates. Next, IP was initially performed using antiGSK3 antibody-coated magnetic beads to specifically capture the active GSK3 in the cell lysates (Fig. 5A(i)), and the GSK3-captured beads were then washed several times; this process was necessary to remove many interferants that may adversely affect the mobility of the peptide substrate in the μIEF channel. Once the synthetic T-pep(p) that had been incubated with the GSK3-attached beads was transferred into the μIEF channel (Fig. 5A(ii) and A(iii)), a strong fluorescent intensity was observed at the pI zone of T-pep (pp) with a marginal intensity at the T-pep(p) zone, meaning that T-pep(p) was phosphorylated into T-pep(pp) by GSK3 in the cell lysates (Fig. 5B(i)). In contrast, there was no change in the zone position of T-Pep(p) when using the control cell lysate without active GSK3 (Fig. 5B(ii)) or using bare magnetic beads without
98
H. Choi et al. / Biosensors and Bioelectronics 73 (2015) 93–99
fluorescent focusing zones in the μIEF channel, where the priming phosphorylation and priming-dependent phosphorylation were sequentially determined in a highly quantitative manner. In addition, this method resulted in detecting the inhibitory effect of protein kinases with high selectivity as well as assaying the protein kinase activity in cell lysates in combination with IP. Owing to its simplicity and general applicability with low sample volumes, we anticipate that our approach opens up new ways to elucidate the sequential phosphorylation of many target substrates as well as investigate the physiological roles of protein kinases in a highthroughput manner.
Acknowledgment This work was supported by Mid-career Researcher Program (Nos. 2013R1A2A2A03015161 and 2013R1A2A2A01014234) and Nano Material Technology Development Program (No. 2012M3A7B4035286) through the National Research Foundation (NRF) funded by the Ministry of Science, ICT, and Future Planning (MSIP). This work was also supported by Basic Science Research Program (No. 2012R1A6A1029029) through the NRF funded by the Ministry of Education.
Appendix A. Supplementary material Fig. 5. (A) Schematic of immunoprecipitation (IP)-combined μIEF for the detection of GSK3 activity in cell lysates. Anti-GSK3 monoclonal antibody on Protein A/G magnetic beads was used to capture GSK3 in MCF-7 cell lysates and peptide substrate (T-Pep(p)) was then reacted to the GSK3-captured beads. (B) Fluorescent μIEF images of T-Pep(p) under the different reaction conditions: i) immunoprecipitated GSK3 for positive cell lysates, ii) immunoprecipitated GSK3 for negative cell lysates, and iii) immunoprecipitated GSK3 without anti-GSK3 antibody.
anti-GSK3 antibody (Fig. 5B(iii)). This evidence suggests that this method would be very suitable for the in vitro analysis of GSK3 activity in cell lysates. Based on these results, our μIEF electrophoresis has many advantages over conventional IEF gel electrophoresis or traditional kinase assays. These include faster assay time, better reproducibility, more quantitative analysis, and smaller sample consumption for the detection of protein kinase reactions and their inhibition. In particular, diverse degrees of phosphorylation for peptide substrates was reliably separated in μIEF, allowing for the straightforward analysis of sequential phosphorylation in a quantitative manner, that otherwise might not be possible using a traditional immunoblotting method. Moreover, as a result of being coupled to multiple microfluidic functionalities, the strengths of this method include multiplexibility, functional integration and automation (Chovan and Guttman, 2002; Reyes et al., 2002). Given the advancements in microfluidic devices directly connected with IP (Wu et al., 2012) and/or single cell analysis (Hughes et al., 2014; Junkin and Tay, 2014; Song et al., 2014), this μIEF method is anticipated to elucidate phosphorylation cascades in association with multiple protein kinases and can be applied for the analyses of other enzyme activities and relevant modifications.
4. Conclusions We demonstrated sequential phosphorylation analysis through microfluidic separation based on the pI of dye-labeled peptide substrates. Despite the subtle modification of peptide substrates by two protein kinases (MAPK and GSK3), the degree of phosphorylation of the peptide was remarkably discriminated by
Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.bios.2015.05.047.
References Anderson, J.C., Peck, S.C., 2008. Plant J. 55, 881–885. Chovan, T., Guttman, A., 2002. Trends Biotechnol. 20, 116–122. Chu, B.Y., Soncin, F., Price, B.D., Stevenson, M.A., Calderwood, S.K., 1996. J. Biol. Chem. 271, 30847–30857. Cohen, P., 2002. Nat. Cell Biol. 4, E127–E130. Cui, H., Horiuchi, K., Dutta, P., Ivory, C.F., 2005. Anal. Chem. 77, 1303–1309. Dajani, R., Fraser, E., Roe, S.M., Young, N., Good, V., Dale, T.C., Pearl, L.H., 2001. Cell 105, 721–732. Das, C., Zhang, J., Denslow, N.D., Fan, Z.H., 2007. Lab Chip 7, 1806–1812. Doble, B.W., Woodgett, J.R., 2003. J. Cell Sci. 116, 1175–1186. Eldar-Finkelman, H., 2002. Trends Mol. Med. 8, 126–132. Ghadiali, J.E., Cohen, B.E., Stevens, M.M., 2010. ACS Nano 4, 4915–4919. Harvey, C.D., Ehrhardt, A.G., Cellurale, C., Zhong, H.N., Yasuda, R., Davis, R.J., Svoboda, K., 2008. Proc. Natl. Acad. Sci. USA 105, 19264–19269. Hastie, C.J., McLauchlan, H.J., Cohen, P., 2006. Nat. Protoc. 1, 968–971. He, B., Meng, Y.H., Mivechi, N.F., 1998. Mol. Cell. Biol. 18, 6624–6633. Hughes, A.J., Spelke, D.P., Xu, Z.C., Kang, C.C., Schaffer, D.V., Herr, A.E., 2014. Nat. Methods 11, 749–755. Hunter, T., 1995. Cell 80, 225–236. Junkin, M., Tay, S., 2014. Lab Chip 14, 1246–1260. Kim, Y.P., Oh, E., Oh, Y.H., Moon, D.W., Lee, T.G., Kim, H.S., 2007. Angew. Chem. Int. Ed. 46, 6816–6819. Kinoshita, E., Kinoshita-Kikuta, E., Matsubara, M., Aoki, Y., Ohie, S., Mouri, Y., Koike, T., 2009. Electrophoresis 30, 550–559. Kupcho, K., Somberg, R., Bulleit, B., Goueli, S.A., 2003. Anal. Biochem. 317, 210–217. Mann, M., Ong, S.E., Gronborg, M., Steen, H., Jensen, O.N., Pandey, A., 2002. Trends Biotechnol. 20, 261–268. Medunjanin, S., Hermani, A., De Servi, B., Grisouard, J., Rincke, G., Mayer, D., 2005. J. Biol. Chem. 280, 33006–33014. O'Neill, R.A., Bhamidipati, A., Bi, X.H., Deb-Basu, D., Cahill, L., Ferrante, J., Gentalen, E., Glazer, M., Gossett, J., Hacker, K., Kirby, C., Knittle, J., Loder, R., Mastroieni, C., MacLaren, M., Mills, T., Nguyen, U., Parker, N., Rice, A., Roach, D., Suich, D., Voehringer, D., Voss, K., Yang, J., Yang, T., Vander Horn, P.B., 2006. Proc. Natl. Acad. Sci. USA 103, 16153–16158. Ohori, M., Kinoshita, T., Okubo, M., Sato, K., Yamazaki, A., Arakawa, H., Nishimura, S., Inamura, N., Nakajima, H., Neya, M., Miyake, H., Fujii, T., 2005. Biochem. Biophys. Res. Commun. 336, 357–363. Peck, S.C., 2006. Plant J. 45, 512–522. Reyes, D.R., Iossifidis, D., Auroux, P.A., Manz, A., 2002. Anal. Chem. 74, 2623–2636. Riddle, S.M., Vedvik, K.L., Hanson, G.T., Vogel, K.W., 2006. Anal. Biochem. 356, 108–116. Ryves, W.J., Harwood, A.J., 2001. Biochem. Biophys. Res. Commun. 280, 720–725.
H. Choi et al. / Biosensors and Bioelectronics 73 (2015) 93–99
Salih, E., 2005. Mass Spectrom. Rev. 24, 828–846. Schenk, P.W., Snaar-Jagalska, B.E., 1999. Biochim. Biophys. Acta-Mol. Cell Res. 1449, 1–24. Shiosaki, S., Nobori, T., Mori, T., Toita, R., Nakamura, Y., Kim, C.W., Yamamoto, T., Niidome, T., Katayama, Y., 2013. Chem. Commun. 49, 5592–5594. Sommer, G.J., Hatch, A.V., 2009. Electrophoresis 30, 742–757. Song, K., Jeong, H.H., Jin, S.H., Park, J.S., Lee, C.S., 2014. Biochip J. 8, 227–233. ter Haar, E., Coll, J.T., Austen, D.A., Hsiao, H.M., Swenson, L., Jain, J., 2001. Nat. Struct. Mol. Biol. 8, 593–596. Walsh, M.P., MacDonald, J.A., 2011. Arch. Biochem. Biophys. 510, 73–75.
99
Wen, J., Wilker, E.W., Yaffe, M.B., Jensen, K.F., 2010. Anal. Chem. 82, 1253–1260. Wu, A.R., Kawahara, T.L.A., Rapicavoli, N.A., van Riggelen, J., Shroff, E.H., Xu, L.W., Felsher, D.W., Chang, H.Y., Quake, S.R., 2012. Lab Chip 12, 2190–2198. Yamamoto, T., Matsuzaki, H., Kamada, S., Ono, Y., Kikkawa, U., 2006. Nat. Protoc. 1, 2791–2795. Zhu, H., Klemic, J.F., Chang, S., Bertone, P., Casamayor, A., Klemic, K.G., Smith, D., Gerstein, M., Reed, M.A., Snyder, M., 2000. Nat. Genet. 26, 283–289.
