Chemical Transfection of Cells in Picoliter ... - Semantic Scholar

Oct 3, 2011 - Department of Chemical Engineering, Virginia Tech, Blacksburg, Virginia 24061, United States. §. Department of Agricultural and ... bas...
0 downloads 11 Views 2MB Size
TECHNICAL NOTE pubs.acs.org/ac

Chemical Transfection of Cells in Picoliter Aqueous Droplets in Fluorocarbon Oil Fangyuan Chen,†,‡ Yihong Zhan,§ Tao Geng,§ Hongzhen Lian,† Peisheng Xu,z and Chang Lu*,‡ †

Department of Chemistry, Nanjing University, Nanjing 210093, P.R. China Department of Chemical Engineering, Virginia Tech, Blacksburg, Virginia 24061, United States § Department of Agricultural and Biological Engineering, Purdue University, West Lafayette, Indiana 47907, United States z Department of Pharmaceutical and Biomedical Sciences, University of South Carolina, Columbia, South Carolina 29208, United States ‡

bS Supporting Information ABSTRACT: The manipulation of cells inside water-in-oil droplets is essential for high-throughput screening of cell-based assays using droplet microfluidics. Cell transfection inside droplets is a critical step involved in functional genomics studies that examine in situ functions of genes using the droplet platform. Conventional water-in-hydrocarbon oil droplets are not compatible with chemical transfection due to its damage to cell viability and extraction of organic transfection reagents from the aqueous phase. In this work, we studied chemical transfection of cells encapsulated in picoliter droplets in fluorocarbon oil. The use of fluorocarbon oil permitted high cell viability and little loss of the transfection reagent into the oil phase. We varied the incubation time inside droplets, the DNA concentration, and the droplet size. After optimization, we were able to achieve similar transfection efficiency in droplets to that in the bulk solution. Interestingly, the transfection efficiency increased with smaller droplets, suggesting effects from either the microscale confinement or the surfaceto-volume ratio.

R

ecent years have witnessed explosion of interests in dropletbased microfluidic technology.1 5 Droplet microfluidics with water-in-oil droplets generated at rates of hundreds to thousands per second provides a high-throughput platform for chemical and biological screenings.6,7 The tiny volume of these droplets (in the range of pico- to nanoliters) creates confined microscale compartments for observation of unique biology,8 screening optimal conditions,9,10 and analysis of chemical/biological molecules of extremely low quantities.11 15 The manipulation and analysis of cells in droplets is also an intensively studied field.16 The encapsulation of individual cells is often desired for single cell screening.17 Such encapsulation has mostly been implemented by producing droplets from cell suspension of low density with the eventual cell occupancy following Poisson statistics.18 Interesting efforts have been made on the use of cell alignment by hydrodynamics for high-efficiency single cell encapsulation.19 Furthermore, in contrast to mineral oil and other hydrocarbon oils used in early reports, the newly demonstrated fluorocarbon oil/surfactant system proved to be highly compatible with in-droplet cell survival and growth due to its gas permeability.18,20,21 All these works paved the way to wide application of droplet microfluidics to cell manipulation and screening. Delivery of genes into cells is an important first step for studying the functions of genes. Functional genomic studies often demand systematically expressing or silencing genes r 2011 American Chemical Society

corresponding to a large fraction of the genome, and such screening allows the identification of genes that are required for a particular phenotype to occur. Functional studies of genes in cells guarantee that the proteins are synthesized in situ with proper post-translational protein folding and glycosylation. Droplet microfluidics has been applied to produce high-quality cationic lipid/DNA complexes used in gene delivery (with cell transfection conducted in bulk medium).22,23 However, there have been very few reports on gene delivery or cell transfection inside droplets. Chemical transfection has been demonstrated in fairly large droplets (∼150 nL) on a digital microfluidic platform that does not require interface between the oil and aqueous phases.24 We used electroporation to deliver genes into cells encapsulated in aqueous droplets in hydrocarbon oil.25 Cells were released from the droplets and transferred to bulk media immediately after electroporation. In comparison, chemical transfection requires long-time incubation (several hours) in droplets. Such operation is not practical using hydrocarbon oil because its low permeability to gases and strong extraction of organic transfection reagents would lead to massive cell death and very low transfection.

Received: August 28, 2011 Accepted: October 3, 2011 Published: October 03, 2011 8816

dx.doi.org/10.1021/ac2022794 | Anal. Chem. 2011, 83, 8816–8820

Analytical Chemistry

TECHNICAL NOTE

Scheme 1. Synthesis of PEG-PFPE-PEG Triblock Copolymer Surfactant

In this report, we investigated chemical transfection of cells in picoliter aqueous droplets in fluorocarbon oil. Droplets encapsulating DNA, transfection reagent (PolyFect), and cells were generated at a frequency of several thousand Hz by flow focusing. We have examined the effects of various parameters such as the incubation time in the droplets, the DNA concentration, and the droplet size on the cell viability and transfection efficiency. The fluorocarbon oil/ surfactant system yielded high cell viability (87% after 6 h incubation in the droplets). Overall, the transfection efficiency in droplets was similar to that in bulk medium. Interestingly, we found that the transfection efficiency increased with smaller droplet size.

’ EXPERIMENTAL SECTION Device Fabrication and Operation. The microfluidic chip was fabricated in polydimethylsiloxane (PDMS) (GE Silicones RTV 615, MG Chemicals) using soft lithography.26 SU-8 2025 (MicroChem) was spun at 500 rpm for 10 s and then at 2000 rpm for 30 s to create a 52 μm thick layer on a 3 in. silicon wafer. The photoresist was then exposed to UV light and developed to produce the SU-8/silicon wafer master. Prepolymer PDMS was poured onto the master at a 10:1 ratio, degassed, and cured in an oven at 80 °C for 1 h. The PDMS piece was then cut and peeled from the master, and access holes for inlets and outlets were punched with a flat needle. Glass slides were cleaned in a basic solution (H2O/30% NH4OH/27% H2O2 = 5:1:1, v/v) at 75 °C for 3 h before they were rinsed by DI water and blown dry. Precleaned glass slide and PDMS were treated in a plasma cleaner (Harrick) and then brought into contact to form irreversible bonding. The device was baked at 80 °C for 1 h for further strengthening of the bonding. Before use, Aquapel (PPG Industries) was used to coat the channel before it was blown out of the channel by air. Aquapel treatment was important for generating a continuous oil phase that wet the surface. Water-inoil droplets were generated by coflowing an oil stream Fluorinert FC-40 (3M) containing 5.0% (w/w) perfluorinated polyether polyethylene glycol (PFPE-PEG) block-copolymer surfactant (synthesis described below) together with culture medium containing cells, DNA, and transfection reagent (PolyFect). The oil was prefiltered by a 0.2 μm filter (VWR). Flows were driven at

constant volumetric flow rates by two syringe pumps (Fusion 400, Chemyx). Surfactant Synthesis. Krytox 157FSH (PFPE with carboxylic acid functionality) was purchased from Dupont. HFE-7100 (>99.5% methoxy-nonafluorobutane) was purchased from 3M. All other chemicals were purchased from Sigma and used without further purification. The triblock copolymer PEG-PFPE-PEG was synthesized according to a method (Scheme 1) modified from the literature.27 Briefly, Krytox 157FSH was reacted with thionyl chloride at the molar ratio of 1:10 under reflux overnight with nitrogen purge. The excess thionyl chloride and solvent were removed by rotary evaporation. The resulting intermediate was dissolved in the mixed solvent of HFE-7100 and benzotrifluoride, followed by addition into the polyethylene glycol (PEG, MW: 400 Da) solution under stirring with the existence of triethylamine (TEA). The reaction was kept under reflux for 24 h. The reaction product was purified by filtration and washed with chloroform and DI water twice. The final product, PEGPFPE-PEG, was dried under high vacuum. Plasmid Preparation. We used pEGFP-C1 plasmid (4.7 Kb, Clontech) encoding enhanced green fluorescent protein (EGFP) as a model vector for observing transfection.28 The plasmid was propagated in E. coli cells and purified using QIAfilter Plasmid Giga Kit (Qiagen) according to the protocol from the manufacturer. The plasmid DNA was dissolved in Tris-EDTA buffer and stored at 20 °C. The DNA concentration was detected by UV absorbance at 260 nm. The OD 260/280 nm ratio was between 1.8 and 2.0. Cell Culture and Transfection. Chinese hamster ovary (CHO-K1) cells were cultured in F-12 Nutrient medium (Fisher Scientific) supplemented with 10% (v/v) fetal bovine serum (Sigma) and 100 mg/mL streptomycin (Sigma) at 37 °C with 5% CO2. Cells were subcultured every 2 days at a ratio of 1:10. PolyFect (a cationic dendrimer from Qiagen) was used to deliver plasmid DNA into cells. PolyFect dissolved in the culture medium with no serum was mixed with DNA solution at a ratio suggested by the manufacturer’s protocol. The PolyFect/DNA mixture was incubated for 5 min at room temperature to allow PolyFect/DNA complex to form. In the droplet transfection experiments, we mixed 50 μL of newly formed PolyFect/DNA complexes with 150 μL of CHO cell sample (107 cells/mL). Such a mixture was immediately 8817

dx.doi.org/10.1021/ac2022794 |Anal. Chem. 2011, 83, 8816–8820

Analytical Chemistry

TECHNICAL NOTE

Figure 1. Layout of the microfluidic device for generating microscale droplets that encapsulate cells, DNA, and transfection reagent. The device has filtration structures to remove solid debris. In the inset image, droplets of 44 μm diameter were generated at a frequency of ∼6000 Hz by having a flow rate of 16 μL/min for the aqueous phase from the central channel and a total flow rate of 64 μL/min for the oil phase from the two side channels.

loaded into a syringe and used in the droplet production by flow focusing. Roughly 100 μL of droplets was produced and collected into a centrifuge tube. Culture medium (100 μL) was then added to cover the droplet layer to prevent evaporation and coalescence during incubation in a cell incubator (37 °C, 5% CO2). After incubation of a certain time, 400 μL of culture medium was added to the emulsion before the mixture was vortexed to release cells into the aqueous phase. The bulk of the aqueous phase was then aspirated and transferred into 96 well plate for culture of an additional 40 h. We captured both phase contrast and fluorescent images of cells using an inverted fluorescence microscope (IX-71, Olympus) equipped with a 20 dry objective and a CCD camera (ORCA-R2, Hamamatsu) for examination of the transfection. Cell viability after droplet transfection was measured immediately after the cells were released from the droplets. Cells were centrifuged at 300g to settle down to the bottom of the well plate. The upper solution was aspirated, and PBS was added to rinse cells. Cells were then stained by 1 μg/mL of propidium iodide (PI, Invitrogen) in PBS. After incubating cells in the dark for 20 min at room temperature, we examined the PI exclusion by fluorescence and phase contrast imaging. The transfection efficiency (defined by the percentage of fluorescent cells among live cells) and the cell viability (determined by the percentage of nonfluorescent cells among the entire cell population after PI staining) were calculated on the basis of results from two trials with ∼1000 cells examined in each trial. As a comparison to the droplet transfection, 150 μL of cell sample (107 cells/mL) was mixed with 50 μL of PolyFect/DNA complexes and cultured in the 96 well plate for a certain duration. Cells were then washed by the culture medium to remove the uninternalized DNA and cultured for an additional 40 h.

’ RESULTS AND DISCUSSION Figure 1 shows the device design for generating aqueous droplets in fluorocarbon oil by flow focusing.29 The aqueous

Figure 2. Incubation of microscale droplets at 37 °C with 5% CO2 in an incubator. The images show droplets of 44 μm initial diameter incubated for various times up to 6 h.

phase (containing DNA, the transfection reagent PolyFect, and cells) came into the device from the center stream while fluorocarbon oil (FC-40) containing a perfluorinated polyether polyethylene glycol (PFPE-PEG) block-copolymer surfactant20 (5%, w/w) came in from the two side streams. PolyFect (from Qiagen) is a cationic activated dendrimer that forms a complex with nucleic acids for cell uptake by nonspecific endocytosis. The mixing of DNA, PolyFect, and cells was conducted immediately before the droplet production in order to make sure that no transfection occurred in the bulk solution. The PFPE-PEG block copolymer was synthesized following a published protocol with minor modifications (detailed in the Experimental Section).27 In our experiments, the droplet size was varied in the range of 44 to 26 μm in the diameter by maintaining the flow rate of the aqueous phase at 16 μL/min and varying the oil phase flow between 64 and 150 μL/min (shown in Supporting Information Figure S1). 8818

dx.doi.org/10.1021/ac2022794 |Anal. Chem. 2011, 83, 8816–8820

Analytical Chemistry

Figure 3. Variation of the cell viability (A) and transfection efficiency (B) with the incubation time in the droplets. Droplets of 44 μm diameter contained DNA of 2.3 μg/mL, associated PolyFect and cells. The viability was measured immediately after cell release from droplets by PI exclusion. The transfection efficiency was determined by the percentage of fluorescent cells among live cells, examined at ∼40 h after cells were released from the droplets by vortex.

The cell concentration in the aqueous phase was ∼7.5  106 cells/mL. When the droplet diameter was ∼44 μm, this cell concentration yielded single cell encapsulation in roughly 24% of the droplets and double occupancy in ∼4% of the droplets, with the rest of the droplets being empty. In our experiments, with the fixed cell concentration, when the droplet size decreased, the droplet occupancy of cells became lower. Figure 2 shows that the droplets were very stable during the first 3 h of incubation (37 °C, 5% CO2 in a cell incubator). Some level of coalesce was observed around 4 h. More unevenness in the droplet size was exhibited at 5 and 6 h (11 and 22% RSD in the droplet diameter, respectively), presumably due to transfer of liquid among droplets. The evaporation from the droplets appeared to be little given the small change in the average droplet diameter after 6 h (from 44 to 42 μm). We extracted cells from droplets at various times during incubation to examine their viability and transfection. Encapsulated cells were transferred from droplets into bulk culture medium for viability examination (tested immediately after the incubation in the droplets) and transfection efficiency determination (conducted after 40 h of additional culture in bulk medium). The dramatic dilution by the bulk culture medium effectively terminated transfection when

TECHNICAL NOTE

Figure 4. Comparison of transfection efficiency in droplets and in well plates with various DNA concentrations (1.1 3 μg/mL). (A) Transfection after a 6 h incubation in droplets of 44 μm diameter. Cells were released from droplets after 6 h of incubation and transferred into culture medium for a 40 h additional culture before the transfection efficiency was determined. (B) Transfection in 96 well plates by incubating cells with DNA and PolyFect for 6 h. The cells were then washed to remove uninternalized DNA and cultured for an additional 40 h before the transfection efficiency was determined.

cells exited droplets. The fluorocarbon oil is permeable to gases20 and permits high cell viability after incubation in the droplets. Figure 3A shows that the cell viability was between 87% (after 6 h incubation) and 92% (after 1 h incubation) with declination over longer incubation. The transfection efficiency increased with longer incubation time in the droplets but such increase plateaued after 3 h (Figure 3B). We also varied the DNA concentration between 1.1 and 3.0 μg/mL (while keeping PolyFect amount proportional to that of DNA) (Figure 4). We tested the transfection both in droplets (Figure 4A) and in bulk culture medium (Figure 4B). In both cases, higher DNA concentration improved the transfection efficiency until it reached a plateau. The transfection in the bulk culture medium appeared to be more efficient than in droplets (∼44 μm diameter) at low concentrations (