Microfluidic Two-Dimensional Separation of Proteins Combining

Mar 19, 2015 - E-mail: [email protected]. Abstract. Abstract Image. A two-dimensional separation system is presented combining scanning temperature g...
1 downloads 10 Views 825KB Size
Technical Note pubs.acs.org/ac

Microfluidic Two-Dimensional Separation of Proteins Combining Temperature Gradient Focusing and Sodium Dodecyl SulfatePolyacrylamide Gel Electrophoresis Seyed Mostafa Shameli and Carolyn L. Ren* Department of Mechanical and Mechatronics Engineering, University of Waterloo, 200 University Ave. West, Waterloo, Ontario Canada, N2L 3G1 ABSTRACT: A two-dimensional separation system is presented combining scanning temperature gradient focusing (TGF) and sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) in a PDMS/glass microfluidic chip. Denatured proteins are first focused and separated in a 15 mm long channel via TGF with a temperature range of 16−47 °C and a pressure scanning rate of −0.5 Pa/s and then further separated via SDS-PAGE in a 25 mm long channel. A side channel is designed at the intersection between the two dimensions to continuously inject SDS into the gel, allowing SDS molecules to be compiled within the focused bands. Separation experiments are performed using several fluorescently labeled proteins with single point detection. Experimental results show a dramatic improvement in peak capacity over one-dimensional separation techniques.

T

urations2,14−18 such as 2D micellar electrokinetic chromatography-capillary electrophoresis (MEKC-CE)15 and 2D IEFCE.16 Although these systems provide fast separations with relatively high peak capacities, most of these configurations have complex designs which lead to poor reproducibility and/ or require a mechanism to transfer the peaks between the dimensions as the two separation methods cannot function continuously.2,5 In this work, a simple microchannel network design is presented that integrates scanning temperature gradient focusing (TGF) and SDS-PAGE in a continuous format. Scanning TGF is a focusing based separation technique that utilizes an axial temperature gradient and a varying hydrodynamic bulk flow along the separation channel. Analytes enter the channel sequentially and are focused and separated simultaneously based on their charge to mass ratio as they traverse the temperature gradient.19,20 Compared to other CE techniques such as CE21 and CIEF,22 TGF only requires a very short separation channel and does not need a narrow injection band at the beginning of separation.20 It can also generate highly concentrated peaks up to 104-fold23 and thus is a suitable candidate for 2D separation systems.20 In the current design, denatured proteins are first focused and separated in a 15 mm long separation channel via TGF with a temperature range of 16−47 °C and a pressure scanning rate of −0.5 Pa/s. By gradually reducing the average bulk flow velocity, the focused bands are subsequently transferred toward

wo-dimensional (2D) separation methods are commonly used in proteomic analysis due to their ability to resolve proteins and polypeptides in complex protein mixtures.1−3 A requirement of any successful 2D separation system is orthogonality, which means that the selected dimensions possess different, yet compatible, separation mechanisms. Furthermore, the second dimension should not destroy the resolution achieved by the previous one.1−3 2D polyacrylamide gel electrophoresis (2D-PAGE) is the most common form of such systems. In this method, isoelectric focusing (IEF) is coupled with sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE) for separation of complex protein mixtures based on the differences in their pI and molecular weight.1−4 While 2D-PAGE is widely used in proteomics analysis, it suffers several drawbacks. First, this method is time-consuming and quite labor-intensive due to manual operation, gel preparation, and sample transfer. Second, several steps are required to perform the separation experiments and to transfer the peaks obtained from the IEF into the gel. Third, the results obtained from this method are not quite reproducible between experiments and laboratories2,4 mainly due to the diffusion of the sample peaks inside the gel. The use of microfluidic technology has provided the opportunity to overcome some of these problems by developing miniaturized systems that require less time and sample volume to operate and can provide higher levels of automation.2,4 In recent years, several microfluidic devices have been developed for 2D-PAGE separation of proteins.6−13 A number of efforts have also been made to combine other types of electrophoresis separation techniques in various config© XXXX American Chemical Society

Received: January 29, 2015 Accepted: March 19, 2015

A

DOI: 10.1021/acs.analchem.5b00380 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry

and then 2500 rpm for 60 s to achieve a 20 μm thick microchannel. Then, the wafer was baked at 65 °C for 2 min and 95 °C for 6 min followed by cooling to room temperature gradually. The designed pattern (Figure 1a) was then transferred to the coated SU-8 via UV exposure (at 850 mJ/ cm2, 365 nm) using a previously developed mask (CAD/Art Services). The wafer was baked again at 65 °C for 1 min and 95 °C for 6 min, developed in PGMEA for 4 min, and dried by nitrogen gas. The mixture of PDMS prepolymer base and curing agent with a weight ratio of 10:1 was degassed under vacuum, poured onto the wafer, and then cured at 90 °C for 2 h. Four 1 mm diameter holes were then punched at the reservoir locations. Next, a thin layer of PDMS was spin coated onto a glass substrate at 3000 rpm and baked at 90 °C for 10 min. After cooling down, it was bonded with the PDMS substrate with the channel feature via oxygen plasma (Harrick Plasma, PDC-001) at 29.2 W, 300 mTorr pressure for 20 s ensuring uniform channel surface properties. Right after the plasma bonding, a 2:3:5 mixture of 3-(trimethoxysilyl) propyl methacrylate, glacial acetic acid, and deionized water was used for channel conditioning to ensure covalent bonding with the gel solution.24 The mixture was first thoroughly degassed and sonicated for 5 min and then pumped through the channels for 10 min from reservoir C using a syringe pump (Harvard Apparatus 11 Plus) at 2 μL/min. Following the conditioning, the channels were flushed and purged with air using the syringe pump at 5 μL/min. Next, all the conditioned channels were filled with the unpolymerized acrylamide solution using the syringe pump at 2 μL/min. The separation gel consisted of 6% T, 3.3% C acrylamide/bis-acrylamide (29:1), 1× tris-glycine native buffer, 0.1% (wt/vol) SDS, and 0.2% (wt/vol) VA-086 photoinitiator. Notations %T and %C indicate the percentage of total acrylamide (wt/vol) and cross-linker (w/w), respectively, which can be adjusted by mixing with buffer (e.g., 1× Tris-Glycine) to define the gel pore size.2,25,26 An inverted epi-fluorescence microscope (Olympus, GX-71) equipped with a 100W mercury arc lamp and a CCD camera (Photometrics, CoolSNAP ES) was used to photopolymerize the gel solution along the second dimension channel, as shown in Figure 1a. This microscope setting was also used for the other fluorescence imaging in this study unless stated otherwise. Step-by-step photopolymerization was performed by visualizing the second dimension channel for about 20 min using a 10× objective, while an x−y translation stage was used to slowly move the light beam along the channels. The uncured acrylamide solution was flushed out by DI water injected from reservoir “A” using the syringe pump at a 1 μL/min flow rate for about 10 min. Full polymerization of the gel was ensured via UV exposure (Newport Corporation, 92530) at 365 nm for 5 min. Experimental Setup. Before the experiments, fused-silica capillaries with 200 μm i.d. and 350 μm o.d. were used to connect the sample and buffer vials to the chip through reservoir “A” and “B”, respectively. A counteracting bulk flow was then created by changing the relative pressure head (h) between the two vials using a linear stage.27,28 The buffer vial contained 0.1 M Tris-phenol solution which has a strong temperature dependent ionic strength. Prior to each experiment, the protein samples were denatured without SDS using a previously reported procedure12 and were put in the sample vial. Consequently, reservoir “D” was filled with 1× Tris-glycine buffer and reservoir C was filled with 1× Tris-glycine buffer

the end of the first dimension, where they are continuously introduced into a 25 mm long channel filled with polyacrylamide gel. A side channel is used at the intersection between the two dimensions to continuously inject SDS into the gel, allowing SDS molecules to be mixed within the focused bands.12 The proteins are further separated along the second dimension based on their molecular weight using SDS-PAGE.



EXPERIMENTAL SECTION Materials. Acrylamide/bis-acrylamide (29:1), tris(hydroxymethyl) aminomethane base (Tris), phenol, glycine, sodium dodecyl sulfate (SDS), urea, dithiothreitol (DTT), and iodoacetamide (IAM) were obtained from Sigma-Aldrich (St. Louis, MO, USA). Water-soluble photoinitiator VA-086 were purchased from Wako Chemicals (Richmond, VA). PDMS prepolymer and the curing agent were purchased from Dow Corning (Midland, MI, USA). SU-8 (2025) and propyleneglycol-monoether-acetate (PGMEA) developer were obtained from Microchem (Newton, MA, USA). Alexa Fluor 488 conjugated protein samples (bovine serum albumin (66 kDa), ovalbumin (45 kDa), parvalbumin (12 kDa), and trypsin inhibitor (21 kDa)) were purchased from Molecular Probes (Eugene, OR, USA). Milk proteins containing casein and whey labeled by Cy3B were obtained from AES Life Sciences (Cambridge, ON, Canada). Fused-silica capillaries with 200 μm i.d. and 350 μm o.d. were purchased from Polymicro Technologies (Tucson, AZ, USA). All solutions were made from ultrapure water (Barnstead/Thermolyne, Dubuque, IA, USA). Fabrication of the Chip. Figure 1 shows the schematic of the 2D separation chip which consists of a glass slide coated with a thin PDMS layer and a PDMS substrate with the designed microchannel fabricated using standard soft lithography techniques. Briefly, a master was made on a silicon wafer by first spin-coating SU-8 2025 photoresist at 500 rpm for 20 s

Figure 1. Schematic of the 2D separation chip (a). Top layer is PDMS containing the separation channel. Bottom layer is glass coated with a thin layer of PDMS. The chip consists of a 15 mm channel for TGF separation of analytes followed by a gel filled 25 mm channel for SDSPAGE. (b) Experimental setup for the 2D separation chip. B

DOI: 10.1021/acs.analchem.5b00380 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry

TGF separation experiments. The same procedure was used for preparing other protein mixtures unless stated otherwise. To investigate the efficiency of the proposed 2D separation system, the separation of labeled proteins was recorded for both the first dimension of TGF (at the end of TGF channel right before the cooler element) and the second in the gel at about 8 mm from the gel entrance. The detection point inside the gel was carefully chosen on the basis of previous experimental evaluations so that each protein peak spends enough time inside the second dimension to be thoroughly mixed with SDS molecules while without significant diffusion prior to detection. Images of an area around 0.5 × 0.5 mm at those locations were taken using the microscope with a 10× objective. Electropherograms were then generated by cropping the images and measuring the fluorescent intensity at a smaller area of 10 × 50 μm inside the channels using a Matlab program developed inhouse. Figures 2 and 3 show the separation results for both 1D TGF and 2D separation techniques for Alexa Fluor 488 conjugated

containing 2% (wt/vol) SDS. To apply the temperature gradient along the TGF channel, thermoelectric heater/cooler units (TE-35-0.6-1, TE Technologies) approximately 5 mm in width were glued onto an aluminum holder with a distance of 2 mm from each other. The holder contained a 4 mm wide slit at the middle to enable channel visualization. The chip was then placed on the holder and taped on the heater/cooler configuration.



RESULTS AND DISCUSSION Prior to the experiments, the temperature gradient along the TGF channel was measured using a previously reported twocolor thermometry technique.27 Briefly, fluorescein and Alexa fluor 546 dyes were dissolved in 10 mM Tris-HCl buffer at pH 7.1. Although fluorescein is normally a poor candidate for thermometry measurements because of its weak temperature dependent quantum efficiency, it does possess strong pH dependence in the pH range of 6−7. Concurrently Tris-HCl has a strong pH−temperature dependency which allows the pH dependent fluorescent intensity to be converted into a temperature dependent response. Alexa Fluor 546 which has weak dependence on pH and temperature was used as the background dye to avoid intensity measurement errors. A temperature calibration curve was generated by introducing the solution into a microfluidic chip at different temperatures while the ratio of the fluorescent intensity of the dyes was measured using the microscope. Furthermore, the developed calibration curve was used to quantify the TGF temperature profile by measuring the ratio of the fluorescent intensity of the dyes along the TGF channel. 2.5 V was applied to each of the thermoelectric units to achieve a temperature gradient of 16− 47 °C which was measured experimentally. Experiments started with filling the TGF channel with 0.1 M Tris-phenol buffer and establishing a positive pressure difference of around 600 Pa between the buffer and sample reservoir.28 Electrophoresis separations were then performed by applying a negative high voltage of −2000 V (Labsmith, HVS448−3000) to the sample vial connected to reservoir “A” while keeping reservoir D grounded. The difference in hydraulic head decreased at a predetermined rate to generate a scanning rate of −0.5 Pa/s allowing the samples to be focused and moved along the TGF channel toward the gel entrance where they continuously entered the second dimension channel. A voltage of −1000 V was applied to reservoir C to ensure a high concentration of SDS inside the gel while avoiding the focused peaks to enter the side channel. The average current was measured to be around 100−150 μA during the experiments. Experiments were first performed using an ideal set of Alexa Fluor 488 conjugated proteins (bovine serum albumin (66 kDa), ovalbumin (45 kDa), parvalbumin (12 kDa), and trypsin inhibitor (21 kDa)). These proteins were denatured without SDS using a previously reported procedure.12 Briefly, a total of 0.4 mg of protein sample (0.1 mg per protein) was dissolved into 0.5 mL of DI water, and urea and DTT were thoroughly mixed with the sample solution with a final concentration of 8 M and 100 mM, respectively. The reaction was allowed to proceed in the dark overnight at room temperature. Subsequently, iodoacetamide (IAM) was added to the solution to a concentration of 120 mM, and the solution was placed in the dark for 1 h at room temperature. The resulting solution was diluted 10 times with 0.1 M Tris-phenol solution for use in

Figure 2. Separation of four Alexa Fluor 488 conjugated proteins: (1) bovine serum albumin, (2) ovalbumin, (3) trypsin inhibitor, and (4) parvalbumin using both TGF and the proposed 2D separation method.

proteins and milk proteins containing casein and whey labeled with Cy3B dye, respectively. The separation performance is improved using the proposed 2D separation for both cases which is able to detect peaks that are not observed with 1D TGF separation. To quantify the improvement, peak-to-peak resolution was calculated using the following equation:

Rs =

Δt 0.85(w1 + w1)

(1)

where Δt is the distance between the two adjacent peaks and wi is the width at the half-peak height which is easier to be accurately measured than the baseline width.27 Tables 1 and 2 list the peak-to-peak resolution results for both 1D TGF and 2D separation techniques for the Alexa Fluor conjugated proteins and milk proteins, respectively, confirming that the overall resolution increases when the 2D design is used. The separations were finished within 8 min for both cases which can be further adjusted by varying the TGF scanning C

DOI: 10.1021/acs.analchem.5b00380 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry

Table 3. Chip to Chip Reproducibility of the 2D Separation Experiments for Separation of Four Alexa Fluor 488 Conjugated Proteins ((1) Bovine Serum Albumin, (2) Ovalbumin, (3) Trypsin Inhibitor, and (4) Parvalbumin) and Fluorescently Labeled Milk Proteins ((1) β-Casein, (2) κ-Casein, (3) α-Casein, (4) β-Lactoglobulin, and (5) αLactalbumin)a resolution

sample Alexa Fluor conjugated proteins fluorescently labeled milk proteins a

Figure 3. Separation of fluorescently labeled milk proteins using both TGF and the proposed 2D separation method. Peaks order: (1) βcasein, (2) κ-casein, (3) α-casein, (4) β-lactoglobulin, and (5) αlactalbumin.

resolution parameter

TGF

2D

n/a 0.63 1.1

0.77 0.65 1.87

Table 2. Peak Resolution Obtained from Separation of Fluorescently Labeled Milk Proteins Using Both TGF and 2D Separation Methodsa resolution ID ID ID ID

1−2 2−3 3−4 4−5

TGF

2D

0.77 n/a 1.28 0.77

0.96 0.59 1.99 0.72

chip 1 chip 2 0.77 0.65 1.87 0.97 0.59 2.04 0.73

0.68 0.69 1.61 1.13 0.56 1.77 0.68

chip 3 0.73 0.71 1.70 1.19 0.58 2.19 0.60

6.21 4.47 7.65 10.47 2.54 10.75 9.71

Peaks order is from right to left.

CONCLUSION A PDMS/glass microfluidic chip was developed for continuous two-dimensional separation of proteins, combining temperature gradient focusing (TGF) and sodium dodecyl sulfate (SDS) polyacrylamide gel electrophoresis (PAGE). In this system, denatured proteins are first focused and separated using the scanning TGF method, and the focused bands are further separated on the basis of their size after they are continuously introduced into the polyacrylamide gel which was fabricated using photopolymerization techniques. A side channel is used at the beginning of the second dimension to continuously inject SDS into the gel, allowing defined volumes of SDS to be composed within the focused bands. The temperature gradient along the TGF channel was created using thermoelectric cooler/heater units under the chip surface. The scanning bulk flow was created by adjusting the head difference between the sample and buffer reservoirs using a motorized linear stage. The demonstrated separation resolution for 2D TGF-PAGE is not as high as some of the reported 2D separation methods such as 2D-PAGE, 2D MEKC-CE, and 2D IEF-CE which can be further improved by optimizing the scanning rate. This study focused on exploring the feasibility of 2D TGF-PAGE and therefore did not perform thorough optimization testing. In addition, this method does offer several benefits that other 2D separation methods lack, including (i) its simple design allows for easier and more robust chip fabrication with better reproducibility and allows for reduced reagent consumption which is more environmentally friendly and results in lower cost; (ii) its separation is in a continuous manner eliminating additional actions/steps for sample transfer between dimensions which complicates the operation and often results in low separation resolution; (iii) it allows the focused bands from the first dimension to entirely enter the second dimension

(1) Bovine serum albumin, (2) ovalbumin, (3) trypsin inhibitor, and (4) parvalbumin using both TGF and 2D separation methods. Peaks order is from right to left.

Peak Peak Peak Peak

1−2 2−3 3−4 1−2 2−3 3−4 4−5



a

parameter

ID ID ID ID ID ID ID

for three different chips. As shown in Table 3, the RSDs are less than 8% (n = 3) for Alexa Fluor conjugated proteins and less than 11% for fluorescently labeled milk proteins indicating stable operation of the device and experimental system. Repeatability testing was not able to be conducted because the gel usually became stained which can be neither cleaned nor reused.

Table 1. Peak Resolution Obtained from Separation of Four Alexa Fluor 488 Conjugated Proteinsa

Peak ID 1−2 Peak ID 2−3 Peak ID 3−4

parameter Peak Peak Peak Peak Peak Peak Peak

chip-tochip RSD, %

a Peaks order from right to left: (1) β-casein, (2) κ-casein, (3) α-casein, (4) β-lactoglobulin, and (5) α-lactalbumin.

rate. Lower scanning rates will result in longer separation time with better sensitivity while higher scanning rates will speed up the separation but peaks will be less concentrated. Due to the fact that TGF provides highly concentrated peaks during separation (similar to IEF), the proposed method offers high detection sensitivity comparable to SDS-PAGE. As it can be seen from Figures 2 and 3, although the peak intensity slightly decreases in the second dimension due to the presence of the gel media, the sensitivity is still good and all the peaks are detectable. Reproducibility of the experimental results for the 2D separation system was also examined and is presented in Table 3. Chip-to-chip reproducibility was investigated by comparing the peak-to-peak resolution of the 2D separation experiments D

DOI: 10.1021/acs.analchem.5b00380 Anal. Chem. XXXX, XXX, XXX−XXX

Technical Note

Analytical Chemistry

(27) Shameli, S. M.; Glawdel, T.; Liu, Z.; Ren, C. L. Anal. Chem. 2012, 84, 2968−2973. (28) Shameli, S. M.; Glawdel, T.; Fernand, V. E.; Ren, C. L. Electrophoresis 2012, 33, 2703−2710.

eliminating sample loss; and (iv) it utilizes single point detection resulting in an easier experimental setup. Separations of four Alexa Fluor conjugated proteins and fluorescently labeled milk proteins were performed using the developed 2D configuration, and the results were recorded both during the TGF separation (first dimension) and at about 8 mm from the gel entrance. The results show a dramatic improvement in peak capacity and resolution with the proposed 2D system compared with the 1D TGF separation results.



AUTHOR INFORMATION

Corresponding Author

*Tel: 1-519-888-4567 x 33030. Fax: 1-519-885-5862. E-mail: [email protected]. Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS The authors gratefully acknowledge the support from Natural Science and Engineering Research Council, Canada Foundation for Innovation, and Advanced Electrophoresis Solutions Ltd. through grants to C.L.R.



REFERENCES

(1) Giddings, J. C. J. High Resolut. Chromatogr. 1987, 10, 319−323. (2) Tia, S.; Herr, A. E. Lab Chip 2009, 9, 2524−2536. (3) Xu, X.; Liu, K.; Fan, Z. H. Expert Rev. Proteomics 2012, 9, 135− 147. (4) Garfin, D. E. TrAC 2003, 22, 263−272. (5) Chen, H.; Fan, Z. H. Electrophoresis 2009, 30, 758−765. (6) Becker, H.; Lowack, K.; Manz, A. J. Micromech. Microeng. 1998, 8, 24−28. (7) Yang, C.; Liu, H.; Yang, Q.; Zhang, L.; Zhang, W.; Zhang, Y. Anal. Chem. 2003, 75, 215−218. (8) Chen, X.; Wu, H.; Mao, C.; Whitesides, G. M. Anal. Chem. 2002, 74, 1772−1778. (9) Griebel, A.; Rund, S.; Schonfeld, F.; Dorner, W.; Konrad, R.; Hardt, S. Lab Chip 2004, 4, 18−23. (10) Li, Y.; Buch, J. S.; Rosenberger, F.; DeVoe, D. L.; Lee, C. S. Anal. Chem. 2004, 76, 742−748. (11) Liu, J. K.; Yang, S.; Lee, C. S.; Devoe, D. L. Electrophoresis 2008, 29, 2241−2250. (12) Yang, S.; Liu, J.; Lee, C. S.; DeVoe, D. L. Lab Chip 2009, 9, 592−599. (13) Lu, J. J.; Wang, S.; Li, G.; Wang, W.; Pu, Q.; Liu, S. Anal. Chem. 2012, 84, 7001−7007. (14) Rocklin, R. D.; Ramsey, R. S.; Ramsey, J. M. Anal. Chem. 2000, 72, 5244−5249. (15) Ramsey, J. D.; Jacobson, S. C.; Culbertson, C. T.; Ramsey, J. M. Anal. Chem. 2003, 75, 3758−3764. (16) Herr, A. E.; Molho, J. I.; Drouvalakis, K. A.; Mikkelsen, J. C.; Utz, P. J.; Santiago, J. G.; Kenny, T. W. Anal. Chem. 2003, 75, 1180− 1187. (17) Shadpour, H.; Soper, S. A. Anal. Chem. 2006, 78, 3519−3527. (18) Ki, H. K.; Myeong, H. M. J. Proteome. Res. 2009, 8, 4272−4278. (19) Hoebel, S. J.; Balss, K. M.; Jones, B. J.; Malliaris, C. O.; Munson, M. S.; Vreeland, W. N.; Ross, D. Anal. Chem. 2006, 78, 7186−7190. (20) Shackman, J. G.; Ross, D. Electrophoresis 2007, 28, 556−571. (21) Shultz-Lockyear, L. L.; Colyer, C. L.; Fan, Z. H.; Roy, K. I.; Harrison, D. J. Electrophoresis 1999, 20, 529−538. (22) Ou, J.; Glawdel, T.; Samy, R.; Wang, S.; Liu, Z.; Ren, C. L.; Pawliszyn, J. Anal. Chem. 2008, 80, 7401−7407. (23) Ross, D.; Locascio, L. E. Anal. Chem. 2002, 74, 2556−2564. (24) Herr, A. E.; Singh, A. K. Anal. Chem. 2004, 76, 4727−4733. (25) He, M.; Herr, A. Anal. Chem. 2009, 81, 8177−8184. (26) He, M.; Herr, A. Nat. Protoc. 2010, 5, 1844−1856. E

DOI: 10.1021/acs.analchem.5b00380 Anal. Chem. XXXX, XXX, XXX−XXX