Development of a Microfluidic Device - Essay Example

?Development of a microfluidic device for efficient protein extraction Eman Alzahrani* Department * E-mail address: Sample pre-treatment is a required step in proteomics to remove interferences and concentrate the samples. A novel method to fabricate a continuous porous silica monolith inside a glass microchip and modify it with Octadecyl (C18) ligand was successfully developed. The performance of the microfluidic device was evaluated for preconcentration and extraction of proteins using six different molecular weight proteins at a concentration of 30 µM. The results show that the silica monolith was highly permeable and a high extraction recovery of the proteins was achieved (90.89 - 99.71%). This newly microfluidic device for protein extraction may find an application in the area of proteomic research. Keywords: Microfluidic device; Sol-gel; Silica monolith; Protein extraction; Octadecyl (C18) 1. Introduction It is becoming increasingly important in the development of new medicines to use important a microfluidic tool for identifying proteins implicated in disease pathways. As the search for novel molecules to tackle diseases increases, the need to identify proteins on biological targets also increases. Efficient extraction of proteins is the most critical step for proteomics by removing the interfering materials and improving the detection sensitivity (Ahn & Wang, 2008). The recently invented silica monolithic materials are highly permeable to liquid flow and have high mass transport compared with the packed beds. Moreover, the monolithic stationary phase does not need frits, which can cause air bubbles to form and the proteins can be adsorbed into the frits and remain trapped (Cabrera et al., 2002 ). Fabrication silica monolith inside the microfluidic devices can decrease the volume of the sample and the reagents, and reduce the time of the analysis (Girault et al., 2004). Bienvenue et al. (2006) have observed that the negative aspect of the sol-gel monolith in microfluidic device is the fact that it shrinks while the monolith is formed. They further explain that this is can then cause the creation of an opening between the silica network and the microchip wall resulting in reduced surface area for protein adsoption. The aim of this contribution is to investigate the fabrication of a simple microfluidic device contained in a crack-free silica monolith to decrease sample handling, reduce contamination, be truly portable, and decrease analysis time. Moreover, its aim is to modify the surface of the silica monolith to Octadecyl silica (ODS) to use it for pre-concentration and extraction of proteins. 2. Materials and methods 2.1. Chemicals and materials Poly (ethylene oxide) (PEO) MW=10,000 Da, trimethylchlorosilane, tetramethylorthosilicate 99 % (TMOS), chlorodimethyloctadecylsilane 95 %, 2,6-lutidine 99 %, NaCl, and trizma base were purchased from Sigma Aldrich (Poole, UK) and used as received without any further purification. Bovine pancreas insulin, bovine heart cytochrome C, chicken egg white lysozyme, ?-lactoglobulin from milk bovine, haemoglobin from human, and bovine serum albumin (BSA) were purchased from the same. Nitric acid, ammonia, toluene, HPLC grade acetonitrile (ACN), and trifluoroacetic acid (TFA) was obtained from Fisher Scientific UK Ltd. (Loughborough, UK). MicroTight Adapter was purchased from Kinesis (Cambs, UK). Poly (ether ether ketone) (PEEK) tubing was purchased from Anachem (Luton, UK). 2.2. Instrumentation Baby bee syringe pump from Bioanalytical System Inc. (West Lafayette, USA). The instrument used for detection was HPLC-UV detection: 785A UV/Visible Detector from Perkin Elmer (California, USA). The reversed-phase analytical column was Symmetry C8 column, 4.6 mm ? 250 mm packed with silica particles (size 5 µm) from Thermo Fisher Scientific (Loughborough, UK). Scanning electron microscope (SEM) (EVO 60. Manufacturer: Carl Zeiss Ltd. (Welwyn Garden City, UK). SEMPREP 2 Sputter Coater from Nanotechnology Ltd. (Sandy, UK). 2.3. Fabrication of the silica-based monolith The monolithic silica rod was fabricated following previously reported procedures (Nakanishi, 1997) with some modifications in the fabrication conditions. The porous silica rod was fabricated by adding 0.282 g of poly (ethylene oxide) (PEO) to 2.537 ml of 1 M nitric acid solution and 0.291 ml of distilled water. The resultant mixture was stirred for 20 minutes while immersed in an ice bath. Then, 2.256 ml of tetramethylorthosilicate (TMOS) was added to the mixture and mixed for 30 minutes. The resulting solution was poured down inside a 2 ml disposable plastic syringe which acted as a mould. The end of the syringe was sealed and the syringe was left for 24 hours at 40 °C to give a white solid rod. The silica monolith rod was soaked in a water bath for 2 hours at room temperature. Then, it was treated with the basic environment and introduced by thermal decomposition of 1 M aqueous ammonia solution at elevated temperature (85 °C) for 24 hours to form mesopores. Then, the rod was washed with distilled water. The monolithic silica rod was placed in the oven for 24 hours at 40 °C, followed by placing the rod in the oven for another 24 hours at 100 °C. For heat treatment, the rod was placed in the oven at 500 °C for 2 hours. After fabrication of the silica rod, it was cut using a blade and was placed inside the chamber on the base plate of the glass microchip. 2.4. Fabrication of the microfluidic extraction device The microchip was made of glass and fabricated using photolithographic and wet etching techniques (Broadwell et al., 2001). It consisted of two plates i.e. the top and base plates, as can be seen in figure 1. The dimensions of the base plate were 16.5 mm length, 14 mm width and a thickness of 3 mm, while the thickness of the top plate was 1 mm, all other dimensions being the same. Both plates contained an access hole of 1.5 mm, which was created using traditional glass drilling techniques. The base plate was placed in the chamber that was used to place the silica monolith and the depth of the extraction chamber was etched to that of 100 µm and -- µm width. After placing the silica monolith inside the chamber, the two plates were thermally bonded by placing them on quartz plate inside a muffle furnace with a block of quartz (mass 70 g) being placed on top. The oven temperature was then maintained at 575 °C for 3 hours. Figure 1. Schematic diagram of the design of the glass microchip used for extraction of protein, the access holes in the top and base plates were 1.5 mm for attachment of PEEK tubing, the base plate contained the extraction chamber – width and 100 µm depth. 2.5. Derivatisation of the silica-based monolith with C18 After fabrication of the microchip, the silica monolith was chemically modified with C18 using silanisation reagent, chlorodimethyloctadecylsilane in 10 ml toluene and 10 drops 2,6-lutidine, by a method similar to the one previously described by (Chuanhui et al., 2006 ) with some modifications. Deriving the silica-based monolith was done by continuous feeding using a syringe pump at a flow rate of 20 ?l min-1 for 6 hours at 80 °C. The end capping procedure was done using 1 g trimethylchlorosilane in 10 ml toluene for another 6 hours in order to block unreacted silanol moieties. After this derivatisation, the monolith was flushed with toluene and then with methanol using a syringe pump for 2 hours prior to use. The morphology of the monolith before and after derivatisation was characterized by scanning electron microscopy (SEM). The sample was coated with a thin layer of gold-platinum (with a thickness of around 2 nm) using a SEMPREP 2 Sputter Coater. Silanes for the chemical modification were handled with caution for safety since they may be cancerous. 2.6. Using the monolithic silica microchip for extraction Hydrodynamic pumping was used to inject all solutions by connecting the syringe to the microchip and, the fluidic connection was done using poly-ether ether ketone (PEEK) tubing. The tubing was fixed into the ports that were drilled into the microfluidic device using epoxy resin. In addition, a Micro Tight adapter was utilized to connect the PEEK tubing to the plastic syringe which was connected to syringe pump in order to control the flow rates. The standard proteins used in extraction were insulin, cytochrome C, lysozyme, ?-lactoglobulin, haemoglobin and bovine serum albumin. They were dissolved individually in 50 mM trizma base solution (pH 7.0) containing 10 mM NaCl and then the sample was adjusted with 0.1 % TFA The concentration of proteins was 30 µM. All experiments were carried out at an ambient/optimum temperature around 23 ?C and the flow rate for all injections was 10 µl min-1 for all steps except for loading the sample (5 µl min-1). The proteins were extracted individually to calculate the recovery of each protein. The sorbent was cleaned with 100 µl ACN (0.1 % TFA) solution and then equilibrated with 100 µl ACN. After the sample application (800 µl), the monolith was washed with 50 µl ACN (0.1 % TFA) solution. Finally, the sample was eluted using 500 µl 60 % ACN (0.1 % TFA) solution and dispensed into eppendorf tube. Aliquot of the eluent was injected directly into the HPLC-UV detector to study the spectra of the proteins and compare them with the spectra of the proteins after extraction to calculate the efficiency of extraction. The mobile phase was acetonitrile-water in the presence of 0.1 % trifluoroacetic acid (TFA) under the isocratic condition and the wavelength was adjusted to 280 nm and the injection volume was 20 µl. The recovery was calculated as given by Furuno et al. (2004) using the following equation: Extraction recovery (%) = (Ielute/Itotal) ? 100 (where Ielute is the amount eluted form the microchip, and Itotal is the amount of protein introduced to the microchip) 3. Results and discussion 3.1. Preparation of the bare silica monolith inside the glass microchip The purpose of this work was to develop a microfluidic device for protein extraction to decrease the volume of the sample and reagents, and allow for fast time analysis. This work involved the fabrication of an effective monolith to use it as a solid phase extraction inside the microchip. The reason for choosing the silica monolith was because it is mechanically stable i.e. they are not affected by temperature and/or organic solvents. In addition, they contain bimodal pores which are macropores that can reduce the back-pressure, and mesopores that can increase the surface area and this leads to give a good level of interaction with the analyte and also maximise the load-capacity of the column. However, fabrication of the sol-gel monolith directly inside the microchip was not easy; this is because the location of the monolith inside the extraction chamber could not be defined since the preparation of the silica monolith depends on using thermal initiation. In addition, the preparation of the sol gel skeleton is associated with shrinkage in the silica rod (Hosoya et al., 2000). To overcome this problem, it was decided that the silica monolith rod be fabricated, cut to the desired disc size and then placed in the base plate of the microchip. Finally, the two plates had to be bound since the silica monolith is stable under very high temperature. The monolithic silica rod was prepared by a sol-gel method. The composition of the starting mixtures were an alkoxy silicon derivative which was tetramethylorthosilicate (TMOS) which had undergone hydrolytic polymerisation reaction in the presence of water-soluble organic polymer, which was poly (ethylene oxide) (PEO). Nitric acid solution was used as a catalyst to start the hydrolysis and condensation reactions. The disposable plastic syringe was used as a mould for preparing the monolithic silica rod. It was observed that the monolith was white, and crack-free. After formation of the network structure of silica skeletons, the internal pore structure of the monolith was tailored to form mesopores on the skeletons by using the basic environment that can increase the surface area of the monolith by converting the micropores, which have low surface area, to mesopores (high surface area) within the monolithic silica skeletons. Consequently, the surface area can increase the binding of proteins to the monolith. This pore-tailoring process involved treating the wet gel with 1 M aqueous ammonia solution at 85 °C. After the thermal decomposition step, the monolith was calcinated at 500 ?C to decompose the organic residues and remaining polymer in the rod with no serous deformation of gel specimens. Figure 2 (a) shows the appearance of the synthesis silica monolith rod and figure 2 (b) displays a photograph of the fabricated microchip contained in the silica monolith without being affected by the thermal bonding of the two plates. Figure 2. (a) The appearance of the porous silica monolith rod, the sol-gel precursor was 0.282 g PEO, 2.537 ml of 1 M HNO3 and 0.291 ml of distilled H2O. (b) Photograph of fabricated glass microchip after thermal bonding step at 575 °C for 3 h. 3.2. Derivatisation the silica monolith with C18 The surface of the silica rod was chemically modified with C18 to make the sorbent hydrophobic. The reason for modifying the silica monolith after placing the bare silica monolith inside the microchip was to avoid destroying the alkyl chain groups when using very high temperatures (575 ?C) for binding the two plates. The derivatisation was by continuous feeding of the porous silica monolith inside the microchip with chlorodimethyloctadecylsilane to produce an octadecyl-silica (C18-silica) using a syringe pump at a flow rate of 20 ?l min-1 for 6 hours at 80 °C. After derivatisation, the end capping was carried out after bonding using trimethylchlorosilane. This was done to decrease the adsorption of proteins since the silanol groups can interact with proteins especially at high pH, by blocking the unreacted silanol groups. The structural morphology of the monolithic silica was examined by scanning electron microscopy (SEM). The SEM micrographs of silica-based monolith before and after modification with C18 can be observed in Figure 3. In general, SEM photographs indicate that the monolithic silica has high homogeneity and a spongy structure as characterized by through-pores penetrating several layers of these skeletons and a network structure of skeletons of different sizes. It can be seen that the shape of the through-pores in the modified silica were relatively round compared with the non-modified silica monoliths. This confirms that the derivatisation of the surface of the silica monolith was successfully done. Figure 3. The SEM micrographs showing the main structure of (a) non modified silica monolith, prepared at 40 °C for 24 hours (b) after modification with C18. The derivatisation reagent was 1 g chlorodimethyloctadecylsilane in 10 ml toluene and 10 drops 2,6-lutidine, and the end capping reagent was 1 g trimethylchlorosilane in 10 ml toluene. 3.3. Extraction proteins using the modified monolithic silica microchip As previously iterated the objective of this study was to evaluate the glass microchip contained in the octadecyl (C18) silica monolith in protein extraction. Six proteins were used to evaluate the performance of the microchip. For extraction, the modified silica inside the glass microchip was cleaned with 100 µl acetonitrile (ACN) to remove impurities and then it was equilibrated with 100 µl ACN (0.1 % TFA) to ensure optimum binding to proteins. Although the permeability of the silica monolith is high and high flow rate (30 µl min-1) could be used without leakage, all solutions were injected using the syringe pump at flow rate 10 µl min-1 for all steps through the SPE column except applying the sample to make sure that the sorbent was clean and conditioned well. After equilibrium, the sorbent, 800 µl protein solution was applied at low flow rate (5 µl min-1) to obtain good percolation between the analyte and the sorbent. Figure 4 shows the appearance of the silica monolith inside the chamber before and after binding of cytochrome C (12,327 Da, red colour), figure 4 (b), and haemoglobin (64,500 Da, brown color), figure 4 (c). This confirmed that the proteins were bound to the monolith and concentrated as is apparent from the change in the colour of the sorbent from white to red for cytochrome C and brown for haemoglobin. Figure 4. (a) The octadecylated silica monolith inside the glass microchip before extraction (white) (b) binding cytochrome C (red colour) on the monolith (c) binding hemoglobin (brown colour) on the modified silica with C18 monolith. The sample was injected using syringe pump at flow rate 5 µl min-1. Concentration of proteins was 30 µM. Before eluting the analyte, the sorbent was washed with 50 µl ACN (0.1 % TFA) so that the residual matrix components which were weakly bound to the sorbent were rinsed from the sorbent. It was found that this volume of the solvent did not decrease the recovery of the analytes and was found by checking the wash solution for breakthrough of the proteins. Finally, the proteins were concentrated and purified using a slightly acidic aqueous-organic solvent, which was a 500 µl 60 % ACN (0.1 % TFA) solution and the eluent was dispensed into the eppendorf tube then injected into the HPLC-UV instrument to study the extraction recoveries (ER) of the proteins, which were calculated from the UV spectrum by comparing the peak area of extracted sample to non-processed sample solution. Figure 5 shows a huge difference in the peak areas of lysozyme between the direct injection of sample (without extraction), and after extraction using the glass microchip. This means lysozyme was concentrated after extraction since the concentration of analyte is proportional to the peak area. In addition, the sensitivity of the detection was enhanced. Figure 5. The comparison of the peak area of lysozyme (a) without purification, and (b) with purification by octadecylated silica monolithic glass microchip. Experimental condition: the mobile phase was composed of ACN and distilled water (50:50) containing 0.1 % TFA, flow rate 0.2 ml min-1. The separation column was Symmetry C8, 4.6 mm ? 250 mm packed with silica (size 5 µm). Between consecutive analyses, a needle for the automated injector was washed with 70 % aqueous methanol. Detection: UV at 280 nm, injected sample volume: 20 µl The recoveries of proteins purified with a C18 phase were calculated, as can be seen in Table1. Each value represents the mean of five measurement carried out. In general, it was found that the extraction recoveries of proteins were high, ranging from 90.89 to 99.71 %. The relative standard deviation (RSD) for peak area counts were calculated to be 2.03- 6.10 for octadecylated silica monolithic microchip. Table 1. The extraction recovery of proteins purified with C18-bonded monolith glass microchip. The extraction recovery was calculated by the following equation: Extraction recovery (%) = (Ielute/Itotal) ? 100. 4. Conclusions In this work, a novel microfluidic device has been successfully developed that contained crack-free monolithic silica modified with octadecyl (C18) ligand in order to speed the time of the analysis, reduce the volume of the analyte and the reagents. It was evaluated for performing solid phase extraction (SPE) for preconcentration of proteins. The modified monolith inside the chamber had low flow resistance, was stable over time, and gave reproducible extraction efficiencies for proteins. This study shows that the effecient extraction of proteins with different molecular weights and isoelectric points was acheived. Work is currently in progress to establish a fully integrated device for protein extraction and separation. References 1. Ahn, N. and Wang, A. (2008). Proteomics and genomics: perspectives on drug and targetdiscovery. Current Opinion in Chemical Biology, 12(1): 1-3. 2. Bienvenue, J., Ferrance, J., Giordano, B., Hassan, B., Kwok, Y., Landers, J., Norris, P., and Wu, Q. (2006). Microchip-based macroporous silica sol-gel monolith for efficient isolation of DNA from clinical samples. Analytical Chemistry, 78(16): 5704-5710. 3. Broadwell, I., Fletcher, P., Haswell, S., McCreedy, T., and Zhang, X. (2001). Quantitative 3-dimensional profiling of channel networks within transparent ‘lab-on-a-chip’ microreactors using adigital imaging method. Lab on a Chip, 1(1): 66–71. 4. Cabrera, K., Leinweber, F., Lubda, D., and Tallarek, U. (2002). Characterization of silica-based monoliths with bimodal pore size distribution. Analytical Chemistry, 74(11):2470-2477 5. Chuanhui, X., Hanfa, Z., Mingliang, Y., Xiaogang, J., and Wenhai, J. (2006). Octadecylated silica monolith capillary column with integrated nanoelectrospray ionization emitter for highly efficient proteome analysis. Molecular and Cellular Proteomics, 5(3): 454-461. 6. Furuno, M., Ishizuka, N., Minakuchi, H., Miyazaki, S., Morisato, K., Nakanishi, K., Shintani, Y. (2004) Development of a monolithic silica extraction tip for the analysis of proteins. Journal of Chromatography A, 1043(1): 19-25. 7. Girault, H., Lion, N., Reymond, F., and Rossier, J. (2004).Why the move to microfluidics for protein analysis. Current Opinion in Chemical Biology, 15(1): 31-37. 8. Hosoya, K., Ishizuka, N., Minakuchi, H., Nagayama, H., Nakanishi, K., Soga, N., and Tanaka, N. (2000). Performance of a monolithic silica column in a capillary under pressure-driven and electrodriven conditions. Analytical Chemistry, 72(6): 1275-1280. 9. Nakanishi, K. (1997). Pore structure control of silica gels based on phase separation. Journal Porous Materials, 4(2): 67-112. Read More