Article pubs.acs.org/ac
Integrated Microfluidic Platform for Multiple Processes from Microalgal Culture to Lipid Extraction Hyun Seok Lim,† Jaoon Y. H. Kim,† Ho Seok Kwak, and Sang Jun Sim* Department of Chemical and Biological Engineering, Korea University, Seoul, 136-713, Republic of Korea ABSTRACT: For economically viable biofuel production from microalgae, it is necessary to develop efficient analytical platforms for quantitative evaluation of different lipid productivities of numerous microalgal species. Currently, microalgal culture, lipid accumulation, and lipid extraction depend on conventional benchtop methods requiring laborious and time-consuming processes. A poly(dimethylsiloxane) (PDMS)-based integrated microfluidic platform was developed to perform multiple steps in sample preparation on a single device for efficient and quantitative analysis of lipid from various microalgal strains. To achieve this goal, a simple microchannel with a micropillar array was integrated to connect the cell chamber and output reservoir, which act as a filtration unit that enables medium change and solvent extraction by fluid injection using a syringe pump. Multiple processes of cell culture, lipid accumulation, and lipid extraction were successfully accomplished using a single device without time-consuming and laborintensive steps. Various conditions of solvent volume and temperature were investigated to optimize lipid extraction yield in the microfluidic device. The lipid extraction efficiency in the microfluidic system was higher than that in bulk using the same solvent. The lipid extraction efficiency achieved using less toxic aqueous isopropanol on the integrated device was 113.6% of that obtained with the conventional Bligh−Dyer method. Finally, lipid productivities of different microalgal strains grown in the microfluidic device were analyzed and compared. These results demonstrate that this simple integrated microfluidic platform can be applied as an alternative to conventional benchtop methods for efficient sample preparation in microalgal lipid analysis.
T
rapid and efficient analysis of performance of numerous microalgal species. During the past decades, a great deal of research using microfluidic systems has been performed in the field of chemical and biological analysis, which revolutionizes biological and chemical analysis by transferring established conventional technologies from benchtop scale to miniaturized microscale to provide portability and rapid analysis.5,6 Microfluidics has a lot of advantages over conventional benchtop methods, such as high sensitivity, enhanced mass and heat transfer, reduced runtime and cost, lower consumption of reagent due to small volume, large surface-to-volume ratio, etc.7,8 Microfluidics can provide useful platforms for biochemical analysis requiring multiple steps by allowing integration of a range of unit operations, including cell culture, sample preparation, separation, and detection of target analytes in a consecutive manner on a single device.6 Integration of multiple functions in microfluidics can reduce error and contamination problems caused by multiple steps during sample manipulation by performing the entire preparation in an enclosed system.9,10 Microfluidic approaches have been rapidly developed and extensively applied for detection and analysis of biomolecules, especially nucleic acids (DNA, RNA) and proteins from cell lysates.11−13 However, sample preparation steps (sampling,
he depletion of fossil fuels and global warming caused by increased energy consumption have triggered worldwide efforts on the development of renewable energy sources. Microalgae-derived biodiesel has recently gained a lot of interest as an attractive alternative to fossil fuels due to their sustainability. Microalgae have many advantages over terrestrial crops, such as rapid growth, high photosynthetic efficiency, high oil productivity per unit area, and noncompetitiveness for arable land.1 Nevertheless, there are still many obstacles to economic viability of microalgae-derived biodiesel due to low productivity. It is estimated that there are more than 50,000 microalgal species, and 30,000 species have been identified.2 Although tremendous efforts have been paid to screening several thousands of microalgal species for their lipid contents during the few past decades, most studies have focused on less than 20 species for biodiesel production.3 Therefore, it is necessary to develop efficient analytical platforms for quantitative evaluation of various lipid productivities of numerous microalgal species and rapid selection of strains with high performance for economically viable biofuel production using selected strains. However, quantitative analysis of microalgal lipid currently depends on conventional benchtop methods for microalgal culture, lipid accumulation, and lipid extraction.4 These methods include time-consuming and labor- and energy-intensive steps, such as cell harvesting, medium change, resuspension, dewatering, lipid extraction by organic solvent, and phase separation, which are obstacles to © 2014 American Chemical Society
Received: February 20, 2014 Accepted: August 4, 2014 Published: August 4, 2014 8585
dx.doi.org/10.1021/ac502324c | Anal. Chem. 2014, 86, 8585−8592
Analytical Chemistry
Article
rapid and efficient analysis of microalgal lipid productivity on a single device. To date, there has been one report on the microfluidic sample preparation in the field of lipid analysis, which performs cell lysis and solid phase lipid extraction using silica beads.18 In the case of microalgal lipid analysis, cell culture and lipid accumulation, including cell harvesting and medium change, are required for lipid extraction. Thus, it is highly necessary to incorporate these multiple functions onto a device to promote process efficiency.
sample pretreatment) have been less developed and remain a major hurdle for realization of μ-TAS compared to other analytical steps, separation, and detection.9,14−16 Despite increasing interest in microalgae-derived biodiesels as mentioned, development of efficient sample preparation platforms for quantitative lipid analysis has been limited in microfluidic systems. Fluorescence-based detection approaches using lipophilic fluorescent dyes (Nile red, Bodipy) have been mainly applied to analysis of microalgal lipid in microfluidics, but these methods are qualitative in nature.4,17 Recently, an integrated microfluidic approach, enabling the part of sample preparation including cell lysis and lipid extraction to be performed on a glass chip packed with silica beads, was developed for the analysis of glycerophospholipids from bacterial cells.18 For efficient preparation of lipid samples from microalgae, it is highly required to incorporate overall processes of cell culture, lipid induction using nutrient-deprived medium, and lipid extraction into a single microfluidic device.19,20 Here we describe a poly(dimethylsiloxane) (PDMS)-based integrated microfluidic platform to perform efficiently multiple steps in sample preparation from microalgal culture to lipid extraction on a single device for quantitative microalgal lipid analysis using gas chromatography (GC). PDMS offers numerous benefits for biological analysis due to its nontoxic, transparent, gas-permeable properties,21,22 which are essential for microalgal culture requiring light and carbon dioxide. Besides, it has advantages of ease of fabrication and low-cost compared to other materials, such as glass and silicon.23 Although many kinds of hard materials, including glass and silicon, have been used for cell lysis and separation of biomolecules, fabrication of these materials is expensive, timeconsuming, and labor-intensive.24,25 Silica and magnetic beads are often used as packing materials in microfluidic channels for efficient separation of biomolecules from cell lysates.18,26−28 However, integration of these materials into microfluidic devices increases the cost and complexity of a device and fabrication process.25 To perform multiple processes on a single device, we integrated a microchannel filled with a micropillar array, which connects the cell chamber and output reservoir, for filtration of cells during medium change and of cell debris during lipid extraction. For the extraction of lipid from microalgae, solvent extraction methods using chloroform− methanol, such as the Folch and Bligh−Dyer methods, are perceived as the gold standards due to the highest extraction efficiency.29 However, these methods have severe environmental and health risks caused by the toxicity of chloroform, classified as a probable carcinogen by the U.S. EPA.30 Nonpolar solvents are also unsuitable for microalgal lipid extraction because of the high polar lipid and water contents in microalgae.31 Due to these drawbacks, alcohols have been regarded as alternatives for lipid extraction because they can be used for wet biomass and extract polar lipids.32 Different from nonpolar solvents that significantly swell PDMS, alcohols are compatible with PDMS without swelling.23 Thus, we used aqueous alcohols for extraction of lipids from microalgae using this microfluidic platform. By combination of polar solvents and PDMS-based integrated devices, we successfully accomplished multiple processes for lipid extraction on a single chip and reduced multiple sample transfer steps by simple injection of medium and solvent through a microfluidic channel. In the present study, we aimed at the development of a simple, affordable, and easy-to-use microfluidic platform that enables performing multiple steps in sample preparation for
■
EXPERIMENTAL SECTION Design and Fabrication of Microfluidic Device. Designs of microfluidic structures were generated using AutoCAD software (Autodesk, USA) and printed on transparency photomask film (Han & All Technology, Korea). Two masters were fabricated using SU-8 negative photoresist, SU-8 50 and 5 (Microchem, USA) for the top and bottom layers of PDMS respectively, on silicon wafers by standard photolithography. The microfluidic device was composed of two layers of PDMS. The bottom layer (15 mm in thickness) contains a cell chamber and output reservoir with a diameter of 8 mm and 6 mm, respectively. The top layer (5 mm in thickness) contains a microchannel (20 mm in length, 5 mm in width) filled with an array of square micropillars of 5 μm in height, 10 μm by 10 μm in cross-section, and 10 μm of interspace to filter microalgal cells and cell debris during lipid extraction. Each layer was fabricated by pouring PDMS prepolymer (10:1 mixture of 184 Sylgard base and curing agent, Dow Corning) onto SU-8 masters and curing at 80 °C in an oven. The bottom layer was bonded to a glass slide via oxygen plasma treatment, and the top layer was bonded to the bottom layer to connect the cell chamber and output reservoir with a microchannel. Swelling of PDMS (degree of swelling (%)) in various solvents was measured and calculated as described previously.33 Strains and Culture Condition. Chlamydomonas reinhardtii CC503 (cw92 mt+), CC124 (mt− agg-1 nit1 nit2), and CC4348 (BAF-J5; cw15 arg7-7 nit1 nit2 sta6-1::ARG7) were obtained from the Chlamydomonas Resource Center at University of Minnesota. Cells were grown in a 250 mL flask containing TAP medium34 at 23 °C in a shaking incubator at an agitation rate of 140 rpm under the light intensity of 50 μmol photons m2 s−1 prior to inoculation into a microfluidic system. For lipid accumulation, cells grown to an exponential phase for 4 days were harvested by centrifugation, resuspended, and incubated in TAP-N medium for 4 days under continuous light (50 μmol photons m2 s−1) at 23 °C in a shaking incubator for lipid accumulation. Lipid Extraction in Bulk. Lipids were extracted from liquid cultures grown in a flask under nitrogen-deprived conditions for 4 days. A modified Bligh−Dyer method with repeated extraction step (1−5 times) was used to compare efficiency of lipid extraction.35 Briefly, 3.9 mL of chloroform/methanol (1:2 v/v) was added to 0.5 mL of liquid culture in a 15 mL Teflon screw-capped glass tube and the mixture was intermittently vortexed at room temperature for 30 min. Then, 1.3 mL of chloroform and 1.8 mL of water were added. After vortexing for 1 min, the mixture was centrifuged for 10 min at 2000 rpm. The lower phase was collected using a Pasteur pipet, and chloroform was evaporated under a stream of nitrogen. Lipids were also extracted using two kinds of aqueous alcohols. To extract lipids from 0.5 mL culture, 5 mL of 70% ethanol (w/w) or 70% isopropanol (w/w) was added and incubated at 60 °C for 30 min with intermittent mixing. 8586
dx.doi.org/10.1021/ac502324c | Anal. Chem. 2014, 86, 8585−8592
Analytical Chemistry
■
Article
RESULTS AND DISCUSSION Fabrication of Microfluidic Platform. We designed a simple two layered PDMS device (Figure 1). The bottom layer
The mixture was centrifuged for 10 min at 2000 rpm. The liquid phase was collected and evaporated. Total lipid extract was dissolved in 0.1−1 mL of hexane for further analysis. Cells were disrupted by addition of acid washed glass beads (1 g/mL culture) and vigorous vortexing for 10 min before lipid extraction to achieve maximum lipid extraction using the Bligh−Dyer method. The measurement of dry cell weight was performed before lipid extraction as described previously.36 Microalgal Culture and Lipid Extraction in Microfluidic System. Cells grown to exponential phase in TAP medium were diluted to OD800 of 0.1, and diluted culture was injected into the cell chamber in the integrated microfluidic device using a 10 mL syringe mounted on a syringe pump (PHD 2000, Harvard Apparatus). The device was incubated at 23 °C under continuous light of 50 μmol photons m2 s−1. Cell growth was monitored by measuring the optical density of the cell chamber at 800 nm in a microplate reader (Infinite 200 Pro, Tecan). After 4 days, 5 mL of TAP-N medium was flowed into the inlet at a flow rate of 200 μL/min to replace the TAP medium in cell chamber for induction of lipid accumulation. To monitor lipid accumulation, 1 mL of TAP-N medium containing Nile red (1 μg/mL) was injected through the inlet to stain the lipid body. The fluorescence intensity of lipid bodies stained with Nile red was measured using a plate reader at an excitation wavelength of 530 nm and an emission wavelength of 580 nm. Lipids were extracted by flowing aqueous alcohols (70% (w/w) isopropanol and 70% (w/w) ethanol) at a flow rate of 200 μL/min into a microfluidic device placed on a hot plate to maintain temperature during lipid extraction. The temperature in the microfluidic chamber was measured using a digital thermometer with a K-type thermocouple probe (Tecpel, Taiwan). After extraction, aqueous alcohol was evaporated at 60 °C under a stream of nitrogen for further thin layer chromatography (TLC) analysis of neutral lipids or GC analysis of the fatty acid methyl ester (FAME) as described above. Lipid Analysis. TLC was performed on silica gel matrix aluminum plates (Sigma-Aldrich). Soybean oil, glyceryl trioleate, oleic acid, 1,3-diolein, and rac-glycerol 1-monooleate (Sigma-Aldrich) were used as standards, and lipids were separated with a solvent mixture of hexane/diethyl ether/acetic acid (80:30:1). Lipids separated on a TLC plate were stained by exposure to iodine vapor. The amount of triacylglycerol (TAG) was quantified by densitometry using ImageJ software. FAME was prepared by acid-catalyzed transesterification of total lipid. Two milliliters of methanolic sulfuric acid (3%, v/v) was added to a 15 mL screw-capped glass tube containing total lipid extract in 1 mL of hexane spiked with 1 mg of pentadecanoic acid as an internal standard. The mixture was vortexed and heated at 95 °C for 1.5 h. After cooling, 2 mL of water and hexane were added and FAME was separated by collecting the organic phase. The extracted FAME was analyzed using a gas chromatograph (Agilent 7890A) equipped with a flame ionization detector and a Stabilwax column (Restek, USA) with following conditions: injection volume, 1 μL; split ratio, 1:50; injector temp, 250 °C; detector temp, 270 °C; oven temp, hold at 200 °C for 5 min, increase to 230 °C at 20 °C/ min, and hold at 230 °C for 13 min. Lipid class separation was performed by silica gel column chromatography, and neutral lipid, phospholipid, and glycolipid were successively eluted using chloroform, acetone/methanol (9:1, v/v), and methanol, respectively. Each lipid class was dried and weighed.
Figure 1. Design of the microfluidic device and operation scheme for microalgal culture, lipid accumulation, and extraction. (A) Injection of seed culture (0.5 mL) into the microfluidic chamber and incubation for 4 days. (B) Induction of lipid accumulation by a supply of nitrogen deficient media. (C) Lipid extraction by injection of solvent (70% isopropanol, 70% ethanol).
has a cell chamber of 8 mm in diameter and 15 mm in height to contain 500 μL of culture broth, the volume required to meet detection limits of TLC and GC with a flame ionization detector (GC-FID), the most common analytical tool for lipid analysis. The top layer has a microchannel filled with an array of square micropillars (height of 5 μm, side length of 10 μm, and interspace of 10 μm) to filter microalgal cells (C. reinhardtii), 10 μm in diameter (Figure 1C). Because the size of the microalgal cell varies according to the species and growth phase, the height and interspace can be adjusted to filter microalgal cells with different sizes. The cell chamber was filled from bottom to top to perform efficiently medium change and solvent extraction. This simple structure facilitated filtration of cell and cell debris at a low cost during the process of cell seeding into the cell chamber, medium change for lipid accumulation, and lipid extraction to remove sample transfer steps, cell harvesting using centrifugation, and resuspension. Solvent Compatibility of Microfluidic Device. Although the Folch or Bligh−Dyer method using chloroform−methanol is perceived as a standard for lipid extraction, alcohols have advantages as solvents for lipid extraction from microalgae. Alcohols can extract lipids from wet biomass and have less 8587
dx.doi.org/10.1021/ac502324c | Anal. Chem. 2014, 86, 8585−8592
Analytical Chemistry
Article
Figure 2. Microalgal culture and lipid accumulation in the microfluidic system. (A) Photo image of device after injection of seed culture. (B) Photo image of device after cell growth for 4 days. (C) Photo image of device after lipid accumulation under nitrogen-deprived conditions. (D) Microscope image of cells in a microfluidic cell chamber. (E) Microscope image of microchannel with a micropillar array during injection of seed culture. (F) Cell growth curve in a microfluidic device (filled circle) and a flask (empty circle). Nitrogen deficient media were injected after 4 days for induction of lipid accumulation. (G) Fluorescence intensity profile of cells accumulating lipid stained with Nile red (microfluidic device: filled circle; flask: empty circle). Values represent the mean of triplicate experiments. Error bars indicate standard deviation (SD).
toxicity compared to nonpolar organic solvents.37 It has been reported that polar organic solvents (ethanol, 1-propanol, etc.) can be used with PDMS without swelling differently compared to nonpolar organic solvents.23,33 However, isopropanol has a higher swelling ratio than other alcohols;38,39 thus, we compared the degrees of swelling of 70% isopropanol and 70% ethanol to those of pure solvents. We observed that 70% ethanol and 70% isopropanol showed reduced degrees of swelling, 1.35% and 5.36%, each corresponding to 30% of ethanol (4.5%) and 72% of isopropanol (7.4%), respectively. There was no noticeable deformation of the microchannels in PDMS devices that were soaked in 70% aqueous isopropanol and ethanol for more than several hours. Microalgal Culture and Lipid Accumulation in Microfluidic System. In this study, we used C. reinhardtii, as a model organism of microalgae, for microalgal culture, lipid accumulation, and lipid extraction in the integrated microfluidic device. For the culture in the microfluidic device, seed culture was injected into the cell chamber using a syringe pump (Figure 2A). After injection, growth was monitored by measuring the absorbance of the cell chamber (Figure 2B). We observed that cells stay in the cell chamber without moving to the output reservoir during the process due to filtration of the micropillar array (Figure 2D) while medium flowed through the microchannel filled with micropillar (Figure 2E). Due to the
transparency of PDMS, we easily monitored cell density by measuring the absorbance in the microplate reader system without sampling. Cells entered into the stationary phase after 4 days (Figure 2F); then nitrogen-deprived medium (TAP-N) was injected into the inlet to replace the original medium (TAP) in the cell chamber for induction of lipid accumulation. Lipid accumulation was observed by measuring the fluorescence intensity of Nile red for 5 days after nitrogen starvation. Fluorescence intensity gradually increased and reached the plateau in 4 days (Figure 2G). We also observed cells turn yellowish green (Figure 2C) compared to those under nitrogen-replete conditions (Figure 2B), which indicates lipid accumulation. We observed that the growth rate in the device is faster than that in the bulk and the oil accumulation rate in the microdevice is similar to that in the bulk (Figure 2F,G). Effect of Temperature and Solvent Volume on Lipid Extraction. As the efficiency of the lipid extraction depends on the lipids solubility affected by temperature and solvent volume, we investigated the effect of temperature and volume of 70% isopropanol and 70% ethanol on the yield of lipid extraction in microfluidics. During the solvent extraction, we measured the temperature in the microfluidic chamber and confirmed that the temperature was stably maintained (Figure 3J). We compared the amount of lipid extract (TAG) obtained from the microfluidic device using 5 mL of each solvent at three 8588
dx.doi.org/10.1021/ac502324c | Anal. Chem. 2014, 86, 8585−8592
Analytical Chemistry
Article
Figure 3. Lipid extraction in a microfluidic device and quantification of TAG. TLC images of lipid extracted at three different temperatures using (A) 70% ethanol (w/w) and (B) 70% isopropanol (w/w). S1: soybean oil, S2: lipids standard (triolein, oleic acid, diolein, monoolein; top to bottom), BD: lipid extract using the Bligh−Dyer method. (C) Measurement of TAG obtained at three different temperatures (black bar: ethanol; gray bar: isopropanol). (D and E) TLC images lipid extracted using four different volumes of (D) 70% ethanol and (E) 70% isopropanol. (F) Measurement of TAG obtained using four different volumes of solvents (black bar: ethanol; gray bar: isopropanol). (G) Microscope image of the micropillar array after solvent extraction. (H) Photo image of lipid extract obtained from the microfluidic device. (I) Cell debris included in lipid extracts obtained using microfluidic devices with (left) and without (right) a micropillar array. (J) Temperature in the microfluidic chamber during solvent extraction. Values represent the mean of triplicate experiments. Error bars indicate standard deviation (SD).
different temperatures (50, 60, 70 °C). Although the highest amount of TAG was extracted at 70 °C due to the temperature dependence of lipid solubility in alcohols (Figure 3A, B),32 the improvement of TAG amount (18.6% in 70% ethanol, 9.9% in 70% isopropanol) by increasing temperature from 60 to 70 °C was not significant compared to that (82% in 70% ethanol, 49% in 70% isopropanol) from 50 to 60 °C (Figure 3C). We also compared lipid extracts obtained from the microfluidic device using four different volumes (1, 2, 5, 10 mL) of each solvent at 60 °C (Figure 3D, E). The highest amount of TAG was extracted when we used 10 mL of each solvent, but the improvement of TAG yield (12% in 70% ethanol, 13.5% in 70% isopropanol) by increasing solvent volume from 5 to 10 mL was not significant compared to that (320% in 70% ethanol, 309% in 70% isopropanol) from 2 to 5 mL (Figure 3F). Thus, further experiments for lipid extraction were performed using 5 mL of solvent at 60 °C because it is more efficient in terms of energy and solvent consumption. We found 70% isopropanol extracts more lipids than 70% ethanol under the same conditions due to the relatively high solubility of lipids in isopropanol compared to ethanol.32 During the extraction process, we also observed cell debris was filtered by the micropillar array after solvent extraction and attached to the surface of the microchannel as its aggregated form due to the inherent hydrophobicity of PDMS40 (Figure 3G). We could obtain clear lipid extract from the integrated microfluidic device without further filtration (Figure 3H). We also extracted lipids using a device without a micropillar array in the microchannel on the top layer. We observed that lipid extract from the device
without a miropillar array contains large amounts of cell debris from solvent extraction of 0.5 mL culture, which support that the micropillar array functions as a filter during solvent extraction (Figure 3I). Because alcohols can precipitate nonlipid biomass and proteins,37 this property of alcohols is also beneficial for removal of cell debris from lipid extract and improvement of purity of the lipid extract. Efficiency of Lipid Extraction in Microfluidic System. We evaluated the efficiency of lipid extraction in the microfluidic device by comparing it to that extracted in the bulk phase. For lipid extraction in bulk, cell growth and lipid accumulation were performed in a flask under the same conditions (temperature, light intensity, period) as those in the microfluidic culture. Lipids were extracted using 5 mL of three different solvents (chloroform−methanol (1:2, v/v), 70% ethanol (w/w), and 70% isopropanol (w/w)) in glass tubes from 0.5 mL of liquid culture (CC503 strain) incubated for 4 days under nitrogen-deprived conditions. We performed repeated extraction of lipids using the Bligh−Dyer method from one to five times and compared the amount of lipids (total FAME) by gas chromatography. We observed there is little difference in the amounts of lipids obtained from four and five times repeated extraction (Figure 2). Therefore, we presented the lipid amount obtained from five times repeated extraction as the reference (100%). We also evaluate whether cell disruption can improve lipid extraction yield before lipid extraction using the Bligh−Dyer method. However, there was no difference in lipid extraction yield between the two methods: with and without cell disruption (Figure 4B). We thought that 8589
dx.doi.org/10.1021/ac502324c | Anal. Chem. 2014, 86, 8585−8592
Analytical Chemistry
Article
Figure 4. Comparison of lipid extraction efficiency using different solvents in bulk and microfluidic systems. (A) Relative amount of lipid extract from repeated extractions using the Bligh−Dyer method. (B) Relative amount of lipid extract using the Bligh−Dyer method with and without cell disruption. (C) Lipid extraction efficiency from five times repeated extraction using the Bligh−Dyer method was applied as the reference (100%). ** indicates significant difference (p < 0.01, t test) (D) Fatty acid composition of total lipid obtained using different methods. Values represent the mean of triplicate experiments. Error bars indicate standard deviation (SD). EtOH: ethanol; IPA: isopropanol.
analyzed the composition of the lipid class obtained using the Bligh−Dyer method and isopropanol extraction. We found that lipid extract from isopropanol extraction contains more phospholipid (17%) than that from the Bligh−Dyer method (13.6%). However, the overall contents of neutral lipid (B-D: 39.7%, isopropanol: 39.1%) and polar lipid (B-D: 60.3%, isopropanol: 60.9%) were similar (Table 1). We further
the amount of solvent (5 mL) and the extraction time are sufficient to extract lipids completely from a relatively small volume of cell culture (0.5 mL) without further cell disruption. For lipid extraction from 0.5 mL culture in a microfluidic device, 5 mL of two different solvents (70% ethanol (w/w), 70% isopropanol (w/w)) were used. For the extraction using aqueous alcohols, temperature was maintained at 60 °C for bulk and microfluidic conditions. We compared the extraction yields of lipid (amount of total FAME per mL culture) obtained by the bulk method and the microfluidic method. We obtained 258 μg of total FAME per mL culture using the Bligh−Dyer method with five times repeated extraction and applied this as reference (100%). We achieved 76% and 79.9% of lipid yield using 70% ethanol and 70% isopropanol in bulk, respectively, compared to the Bligh−Dyer method. We achieved higher lipid extraction yields (100.9% in 70% ethanol, 113.6% in 70% isopropanol compared to the Bligh−Dyer method) using 70% alcohols in the microfluidic device than using the same solvent in bulk (Figure 4A). Notably, the lipid extraction yield using 70% isopropanol in microfluidics corresponds to 113.6% of that obtained using the Bligh−Dyer method. Thus, we were able to extract lipid using our microfluidic platform without reduction of yield compared to the Bligh−Dyer method in bulk. We surmised this improvement of lipid extraction yield in the microfluidic device is attributed to enhanced heat and mass transfer and high pressure at the entrance of the micropillar array generated by injection of solvent and flow resistance. We
Table 1. Comparison of Lipid Class Composition of Lipid Extract from the Bligh−Dyer Method and Isopropanol Extraction extraction method lipid class
B-D method (%)
70% isopropanol (%)
neutral lipid glycolipid phospholipid
39.72 ± 0.30 46.71 ± 0.14 13.57 ± 0.16
39.10 ± 0.05 43.51 ± 0.06 17.38 ± 0.01
examined the fatty acid composition in the total lipid extract obtained using bulk and microfluidic methods. Although there were differences in the fatty acid compositions of the lipid extracts due to growth environment and solvent properties, we confirmed the major fatty acids from the Bligh−Dyer method and alcohol extraction are 16:0, 16:1, 18:0, 18:1, 18:2, and 18:3, as previously reported (Figure 4D).41 Fatty acid formation and composition depends on the growth conditions of microalgae, including culture environment. Although we cultured micro8590
dx.doi.org/10.1021/ac502324c | Anal. Chem. 2014, 86, 8585−8592
Analytical Chemistry
Article
algal cells in the microfluidic device at the same temperature and light intensity as in the flask (bulk condition), cell growth was not the exactly same. We thought these different culture environments affected the fatty acid compositions of cells under microfluidic and bulk conditions. However, when we consider the overall composition of fatty acid in the lipid extract from the B-D method and isopropanol extraction in the microfluidic device, the difference of C16:0 is not significant (C16:0, 48.2 ± 1.3% (B-D method) and 41.2 ± 2.2% (IPA in device)). Comparison of Strain Performance for Lipid Production. We examined whether our integrated microfluidic platform can be applied to evaluation of strain performance for lipid production. We investigated three strains of C. reinhardtii (CC503, CC4348, and CC124) in terms of lipid production yield. While CC503 and CC4348 are cell-wall-less mutants, CC124 has an intact cell wall composed of hydroxyproline-rich glycoproteins that can be a barrier to solvent penetration.42 CC4348 is a starchless mutant strain in which ADP-glucose pyrophosphorylase is disrupted43 and can accumulate more lipid than other strains such as CC124 and CC503 under nitrogen-deprived conditions.44 Therefore, these strains with different properties were used to verify the ability of our integrated microfluidic system to perform lipid extraction. To compare the lipid production yield, cell growth and lipid accumulation were performed under the same conditions. Total lipids were extracted using 5 mL of 70% aqueous isopropanol from the integrated microfluidic device containing 0.5 mL culture of each strain that accumulated lipid for 4 days under nitrogen-deprived conditions. The lipid yields obtained from the cultures of CC503 and CC124 were 15.2 ± 1.2 and 16.4 ± 1.0 (% DCW), whereas the yield from the starchless mutant, CC4348, was 34.9 ± 1.8 (% DCW) (Figure 5A). This result is consistent with previous reports on improved lipid content in sta6 mutant compared to other strains.41,44 This result also indicates this microfluidic system using aqueous isopropanol can extract lipid from various strains regardless of the presence of a cell wall when we compared the lipid production yields from different strains in this study with those in a previous study.45 We further investigated whether there were differences in lipid synthesis among strains by comparing the fatty acid compositions of total lipids (Figure 5B). Although there were minor differences among strains, we confirmed major fatty acids of each strain are 16:0, 16:1, 18:0, 18:1, 18:2, and 18:3, as previously reported.41,45
Figure 5. Evaluation of lipid productivities of microalgal strains using 70% isopropanol (w/w) in a microfluidic system. (A) Lipid production yield of three strains of C. reinhardtii (CC124, CC4348, and CC503). (B) Fatty acid composition of total lipid obtained from three strains of C. reinhardtii. Values represent the mean of triplicate experiments. Error bars indicate standard deviation (SD).
we were able to analyze and compare different lipid productivities of various strains of C. reinhardtii. Therefore, the integrated microfluidic device could provide a solution to developing fully integrated μ-TAS for microalgal lipid analysis. Due to simple design, inexpensive material, and ease-of-use of the presented system, it can be practically used with broad accessibility and portability as an alternative to the conventional benchtop method used for microalgal sample preparation for lipid analysis.
■
■
CONCLUSION Integration of multiple functions into a microfluidic system is critical to achieve the true μ-TAS enabling rapid and efficient analysis, but many sample preparations are still performed offchip. In the present study, for the first time (to our knowledge), we developed a simple and easy-to-use microfluidic platform to conduct multiple operations from cell culture to lipid extraction on a chip for efficient sample preparation for microalgal lipid analysis. By integration of the micropillar array, we successfully accomplished cell culture, lipid accumulation, and lipid extraction on a chip by simple injection of fluids without centrifugation and sample transfer steps, which are energy- and labor-intensive. Lipid extraction efficiencies achieved using aqueous alcohols in microfluidic system were higher than those in bulk due to the advantages provided by the microfluidic environment. The lipid extraction efficiency achieved using 70% aqueous isopropanol in the microfluidic system was 113.6% of that obtained with the Bligh−Dyer method. Using the device,
AUTHOR INFORMATION
Corresponding Author
*E-mail:
[email protected]. Author Contributions †
H.S.L. and J.Y.H.K. contributed equally to this work.
Notes
The authors declare no competing financial interest. All authors have given approval to the final version of the manuscript.
■
ACKNOWLEDGMENTS This work was supported by a Korea CCS R&D Center grant (2011-0031997) (the main project that supported this work) and a National Research Foundation of Korea (NRF) grant (NRF-2013R1A2A1A01015644/2010-0027955) funded by the Korea government (Ministry of Science, ICT & Future Planning). 8591
dx.doi.org/10.1021/ac502324c | Anal. Chem. 2014, 86, 8585−8592
Analytical Chemistry
■
Article
(36) Yoo, J. J.; Choi, S. P.; Kim, J. Y.; Chang, W. S.; Sim, S. J. Bioprocess Biosyst. Eng. 2013, 36, 729−736. (37) Yao, L.; Lee, S.; Wang, T.; Gerde, J. A. J. Am. Oil Chem. Soc. 2013, 90, 571−578. (38) Wang, Y.; Balowski, J.; Phillips, C.; Phillips, R.; Sims, C. E.; Allbritton, N. L. Lab Chip 2011, 11, 3089−3097. (39) Park, C. S.; Han, Y.; Joo, K. I.; Lee, Y. W.; Kang, S. W.; Kim, H. R. Opt. Express 2010, 18, 24753−24761. (40) Davies, R. T.; Kim, D.; Park, J. J. Micromech. Microeng. 2012, 22, 055003. (41) Hejazi, M. A.; Wijffels, R. H. Trends Biotechnol. 2004, 22, 189− 194. (42) Zabawinski, C.; Van Den Koornhuyse, N.; D’Hulst, C.; Schlichting, R.; Giersch, C.; Delrue, B.; Lacroix, J. M.; Preiss, J.; Ball, S. J. Bacteriol. 2001, 183, 1069−1077. (43) Li, Y.; Han, D.; Hu, G.; Dauvillee, D.; Sommerfeld, M.; Ball, S.; Hu, Q. Metab. Eng. 2010, 12, 387−391. (44) Work, V. H.; Radakovits, R.; Jinkerson, R. E.; Meuser, J. E.; Elliott, L. G.; Vinyard, D. J.; Laurens, L. M.; Dismukes, G. C.; Posewitz, M. C. Eukaryot. Cell 2010, 9, 1251−1261. (45) Wang, Z. T.; Ullrich, N.; Joo, S.; Waffenschmidt, S.; Goodenough, U. Eukaryot. Cell 2009, 8, 1856−1868.
REFERENCES
(1) Chisti, Y. J. Biotechnol. 2013, 167, 201−214. (2) Richmond, A. Hankbook of Microaglal Culture: Biotechnology and Applied Phycology, 1st Ed.; Blackwell Science: Oxford, 2004. (3) Larkum, A. W. D.; Ross, I. L.; Kruse, O.; Hankamer, B. Trends Biotechnol. 2012, 30, 198−205. (4) Holcomb, R. E.; Mason, L. J.; Reardon, K. F.; Cropek, D. M.; Henry, C. S. Anal. Bioanal. Chem. 2011, 400, 245−253. (5) Roman, G. T.; Kennedy, R. T. J. Chromatogr., A 2007, 1168, 170−188. (6) Yeo, L.; Chang, H.; Chan, P. P. Y.; Friend, J. R. Small 2011, 7, 12−48. (7) Manz, A.; Graber, N.; Widmer, H. M. Sens. Actuator B 1990, 1, 244−248. (8) Thaitrong, N.; Charlermroj, R.; Himananto, O.; Seepiban, C.; Karoonuthaisiri, N. PLoS One 2013, 8, e83231. (9) Kim, J.; Johnson, M.; Hill, P.; Gale, B. K. Integr. Biol. 2009, 1, 574−586. (10) El-Ali, J.; Sorger, P. K.; Jensen, K. F. Nature 442, 403−411. (11) Oblath, E. A.; Henley, W. H.; Alarie, J. P.; Ramsey, J. M. Lab Chip 2013, 13, 1325−1332. (12) Tsaloglou, M. N.; Laouenan, F.; Loukas, C. M.; Monsalve, L. G.; Thanner, C.; Morgan, H.; Ruano-López, J. M.; Mowlem, M. C. Analyst 2013, 138, 593−602. (13) Hughes, A. J.; Herr, A. E. Proc. Natl. Acad. Sci. U.S.A. 2012, 109, 21450−21455. (14) Siegrist, J.; Gorkin, R.; Bastien, M.; Stewart, G.; Peytavi, R.; Kido, H.; Bergeron, M.; Madou, M. Lab Chip 2010, 10, 363−371. (15) Rios, A.; Zougagh, M. Trends Anal. Chem. 2013, 43, 174−188. (16) Culbertson, C. T.; Mickleburgh, T. G.; Stewart-James, S. A.; Sellens, K. A.; Pressnall, M. Anal. Chem. 2014, 86, 95−118. (17) Erickson, R. A.; Jimenez, R. Lab Chip 2013, 13, 2893−2901. (18) Sun, T.; Pawlowski, S.; Johnson, M. E. Anal. Chem. 2011, 83, 6628−6634. (19) Han, A.; Hou, H.; Li, L.; Kim, H. S.; Figueiredo, P. Trends Biotechnol. 2013, 31, 225−232. (20) Kim, H. S.; Weiss, T. L.; Thapa, H. R.; Devarenne, T. P.; Han, A. Lab Chip 2014, 14, 1415−1425. (21) Vollmer, A. P.; Probstein, R. F.; Gilbert, R.; Thorsen, T. Lab Chip 2005, 5, 1059−1066. (22) Regehr, K. J.; Domenech, M.; Koepsel, J. T.; Carver, K. C.; Ellison-Zelski, S. J.; Murphy, W. L.; Schuler, L. A.; Alarid, E. T.; Beebe, D. J. Lab Chip 2009, 9, 2132−2139. (23) Lee, J. N.; Park, C.; Whitesides, G. M. Anal. Chem. 2003, 75, 6544−6554. (24) Graß, B.; Neyer, A.; Jöhnck, M.; Siepe, D.; Eisenbeiß, F.; Weber, G.; Hergenröder, R. Sens. Actuators, B 2001, 72, 249−258. (25) Reinholt, S. J.; Behrent, A.; Greene, C.; Kalfe, A.; Baeumner, A. J. Anal. Chem. 2014, 86, 849−856. (26) Bhattacharyya, A.; Klapperich, C. M. Anal. Chem. 2006, 78, 788−792. (27) Hagan, K. A.; Bienvenue, J. M.; Moskaluk, C. A.; Landers, J. P. Anal. Chem. 2008, 80, 8453−8460. (28) Berry, S. M.; Alarid, E. T.; Beebe, D. J. Lab Chip 2011, 11, 1747−1753. (29) Sheng, J.; Vannela, R.; Rittmann, B. E. Bioresour. Technol. 2011, 102, 1697−1703. (30) Andersen, M. E.; Meek, M. E.; Boorman, G. A.; Brusick, D. J.; Cohen, S. M.; Dragan, Y. P.; Frederick, C. B.; Goodman, J. I.; Hard, G. C.; O’Flaherty, E. J.; Robinson, D. E. Toxicol. Sci. 2000, 53, 159−172. (31) Wang, G.; Wang, T. J. Am. Oil Chem. Soc. 2012, 89, 335−345. (32) Johnson, L. A.; Lusas, E. W. J. Am. Oil Chem. Soc. 1983, 60, 229−242. (33) Honda, T.; Miyazaki, M.; Nakamura, H.; Maeda, H. Lab Chip 2005, 5, 812−818. (34) Gorman, D. S.; Levine, R. P. Proc. Natl. Acad. Sci. U.S.A. 1965, 54, 1665−1669. (35) Bligh, E. G.; Dyer, W. J. Can. J. Biochem. Physiol. 1959, 37, 911− 917. 8592
dx.doi.org/10.1021/ac502324c | Anal. Chem. 2014, 86, 8585−8592
