Programmable Static Droplet Array for the Analysis of CellâCell

Subscriber access provided by Eastern Michigan University | Bruce T. Halle Library

Article

A programmable static droplet array for the analysis of cell-cell communication in a confined microenvironment Si Hyung Jin, Sung Sik Lee, Byungjin Lee, Seong-Geun Jeong, Matthias Peter, and Chang-Soo Lee Anal. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.analchem.7b01462 • Publication Date (Web): 21 Aug 2017 Downloaded from http://pubs.acs.org on August 22, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

Analytical Chemistry is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Analytical Chemistry

A Programmable Static Droplet Array for the Analysis of CellCell Communication in a Confined Microenvironment Si Hyung Jin,†,¶ Sung Sik Lee, ‡,§,¶ , Byungjin Lee,† Seong-Geun Jeong,† Matthias Peter,‡,* and ChangSoo Lee†,* †

Department of Chemical Engineering, Chungnam National University, 99 Daehak-ro, Yuseong-Gu, Daejeon, 34134, Republic of Korea ‡ Institute of Biochemistry, ETH Zürich, Zürich, CH 8093, Switzerland § Scientific Center for Optical and Electron Microscopy (ScopeM), ETH Zürich, Zürich, CH-8093, Switzerland *Phone: +41 44 633 6586. Fax: +41 44 633 1228. E-mail: [email protected]. *Phone: +82 42 821 5896. Fax: +82 42 822 8995. E-mail: [email protected] ABSTRACT: Direct cell-cell communication can occur through various chemical and mechanical signals. However, available cell culture systems lack single cell resolution and are often limited by sensitivity and accuracy. In this study, we present an accurate, efficient and controllable microfluidic device that can be used for in-situ monitoring of natural cell-cell contact and signaling processes in a confined microenvironment. This innovative static droplet array (SDA) enables highly efficient trapping, encapsulation, arraying, storage, and incubation of defined cell populations. For proof-of-principle experiments, we monitored the response of budding yeast to peptide mating pheromones, as it is one of the best understood examples of eukaryotic cell-cell communication. Specifically, we measured the yeast response to varying concentration of synthetic MATα-type mating factor, as well as varying the cell number ratio of MATα and MATa in a confined space. We found clear morphological and doubling-time changes during the mating reaction with a significantly higher accuracy than conventional methods. Further, phenotypic analysis of data generated with the microfluidic static droplet array allowed distinguishing the function of genes in yeast mutants defective for different aspects of pheromone signaling. Taken together, the microfluidic platform provides a valuable research tool to study cell-cell communication and signaling in a controlled microenvironment with the sensitivity and accuracy required for screening and long-term phenotypic analysis.

Cell-cell communication is crucial for organisms to coordinate cellular activity and interactions such as neural transmission, cell polarity, cell-cell fusion, biofilm formation and global inflammatory response.1-5 During cell-cell communication, cells sense extracellular signaling molecules and transduce the information through a biochemical signaling network to elicit an appropriate response. Several studies have shown that the ratio of different populations modulate cell signaling quantitatively and qualitatively, and thereby affect the biological or physical connections with surrounding cells.6-8 However, traditional cell culture systems poorly reproduce cell-cell interactions and the effects of mixed cell populations. Moreover, spatial variations in cell density can develop biochemical signal gradients that can alter cell migration, proliferation and differentiation.6-8 Thus, although considerable progress has been made in understanding the components involved in signal transduction, there is urgent need for experimental systems that accurately control different cell populations in a confined microenvironment to understand cell-cell communication and relevant biological outputs.6-11 Recent developments in microfluidic-based systems overcome many limitations of traditional cell culture.12-17 Microfluidic perfusion systems facilitate injection of medium or stimuli with desired input profile, and allow following the response of single cells using live-cell microscopy.18-22 For example, mi-

crofluidic devices can robustly generate concentration gradients and quantify chemotaxis towards attractants.23-25 However, these devices rely on continuous flow systems and are thus unable to integrate the effects of secreted molecules which are often used as local signals in cell-cell communication. Moreover, these systems are mainly limited to microscopic readouts and fail to capture other single cell measurements. Droplet arrays emerged as a powerful alternative to circumvent these restrictions, as they allow cell culturing in a precisely controlled and confined microenvironment.26-29 Recent advances enable efficient manipulation of microscale volumes, such that droplets can be produced as miniaturized biological systems with defined sizes and composition.26-28 In addition, defined changes can rapidly be introduced by controlled fusion with another droplet of distinct content36-38, thus allowing to monitor cellular responses without washing away or diluting secreted molecules.30-35 Automation of such droplet microfluidics allows generating complex combinational arrays, and individual droplets can be harvested and analyzed at a given time point.26-29 Static droplet arrays (SDA) are capable of generating homogeneous and uniform droplets, and have recently emerged as an attractive platform capable of cell cultivation, sorting and high-throughput screening.26-28 SDA eliminate cross-contamination between adjacent droplets, accurately control the population ratio in a droplet, and allow in situ

ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

analysis of live cells. However, while the power and biological applications of such droplet arrays for cell-cell signaling are apparent, until now current designs suffer from rapid evaporation and are thus unsuited for long-term cell culture studies. Moreover, integrated designs coupled to programmable operation are lacking at present. We are interested in understanding the mating response of budding yeast S. cerevisiae,39,40 a well-studied model system that relies on cell-cell communication at several stages of the complex process. Mating between two haploid cells of opposite mating-type (MATa and MATα) is triggered through secreted pheromones that bind specific G-protein coupled receptors (Ste2 and Ste3, respectively) located at the plasma membrane of the mating partner. MATα-type cells secrete an unmodified amphipathic peptide pheromone (α-factor), while MATa-type cells secrete a farnesylated and carboxymethylated 12 amino acid hydrophobic peptide pheromone (a-factor). Upon binding, pheromones activate a dedicated Mitogen-Activated Protein Kinase (MAPK) signalling pathway, which triggers different cellular responses including changes in gene expression, arrest of the cell cycle prior to DNA replication,41 and initiation of directed polarized growth towards the signal source. After cell-cell contact, cells of opposite mating-type initiate cell- and nuclear fusion, resulting in the formation of a single diploid cell.39 Pheromone signaling has typically been monitored experimentally by measuring the number of diploids formed after mixing a and α-cells, or by following the cellular responses after exposing haploid MATa cells to synthetic α-factor.42,43 While these assays have been successful to identify the signaling components and their intracellular wiring, studying cellular responses that rely on cell-cell interactions such as cell-cell fusion have been difficult to address. Live cell imaging of the mating process has recently been achieved by randomly mixing and immobilizing cells of different mating-type in an open well or beneath an agar pad.44,45 However, the number of actual mating pairs in such systems is often low, and the signal of secreted pheromone molecules is diluted and diffused in the aqueous medium. Recent studies observed that diploid formation strongly depends on cell density and suggested that cells monitor relative mate abundance (sex ratio) within mixed populations.46,47 Thus, cells may adjust their commitment to sexual reproduction in proportion to their estimated chances of successful mating. However, these experiments suffered from technical limitations in controlling the density and population ratio between the two mating types, because the volume of conventional well plates and culture tubes is too large to produce homogeneous solutions. Moreover, cell suspensions preclude live cell imaging to adequately monitor the entire mating process, including cell-cell communication, morphological changes and cell cycle arrest. To circumvent these limitations, we have developed an automated and versatile microfluidic static droplet array (SDA) that can mimic cell-cell communication in a confined and precisely controlled microenvironment. Long-term cell culturing was achieved by implementing a water reservoir channel such that water loss is minimized without diluting secreted molecules. The SDA can be programed to produce within minutes a combinatorial droplet array by sequential repetition of droplet generation, metering, merging, encapsulation of cells, transportation, storage, and addressing. The pipeline is fully auto-

Page 2 of 9

matic and allows studying precise co-cultures of cells in individual droplets by simple merging of two different droplets containing different cell types. As a proof-of-principle application, we have used this SDA platform to study the importance of defined cell-cell interactions for yeast mating. Specifically, we showed that the SDA platform can control the concentration of synthetic α-factor pheromone in each droplet and investigated the concentration effect of pheromone and different mutations on mating related gene expression, cell cycle arrest, polarization, and fusion. In addition, by controlling the population ratio of MATa and MATα cells in confined droplets, we evaluated the contribution of cell density and competition on mating efficiency.

EXPERIMENTAL SECTION Plasmids, Yeast Strains and Liquid Media. All yeast strains are listed in Table S-1, respectively. Standard methods were used for yeast strain construction and molecular biology. Yeast strains are derivatives of BY4741.48 Gene fusions were generated by homologous recombination-based replacement of the endogenous gene, and expressed from their endogenous promoter unless indicated. Yeast cells were cultured overnight in YPD medium (1% w/v yeast extract, 2% w/v bactopeptone, 2% w/v glucose) at 30 °C, and then diluted into fresh YPD or SD medium (Synthetic Defined Media with 2% glucose) until they reached exponential phase (OD600 of 0.1). Synthesized α-factor solution was purchased from American Peptide Company (CA, USA). To quantitatively analyze the doubling time during the pheromone response, we assume cell cycle arrest occurs when yeast cells did not proliferate, and we scored such nonproliferation as 80 hours of doubling times. Microfluidic Device Fabrication. We fabricated the triple layered SDA microfluidic device by multi-layer soft lithography with poly(dimethylsiloxane).10,26,28,49 Three different wafer molds were fabricated by photolithographic processes to create the water reservoir, fluidic, and control layers. The wafer molds for the water reservoir and control layers were fabricated by negative photoresist (SU8 3025, MicroChem). The wafer mold for the fluidic layer was fabricated by using both a negative photoresist (SU8 3025) and a positive photoresist (AZ 9260) to create a rectangular channel and round channel, respectively, both with 15 µm height. The mixture of PDMS elastomer and its curing agent (10:1 ratio) was poured onto the water reservoir mold. It was incubated in a 65 °C oven for 1 hour, and then holes were punched at the end of channels. A middle fluidic layer was spin-coated on the fluidic mold with PDMS mixture (10:1 ratio) to form a 75 µm height, and it was heated for 40 min at 65 °C in an oven. The upper water reservoir layer was aligned with the fluidic layer, and then incubated in a 65 °C oven for 1 hour. The bottom control layer was fabricated with PDMS (20:1 ratio) with a 25 µm height by spin-coating and heating for 40 min at 65 °C oven. Afterwards, it was aligned with a pre-cured water reservoir and fluidic layer assembly, and then incubated in a 65 °C oven for 12 hours. The assembly of triple PDMS layers was peeled off from the wafer mold, and holes were punched to connect the inlet ports to the fluidic and microvalve channels. Finally, the PDMS chip was bonded to a thin cover glass (150 µm) by plasma treatment.


Page 3 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 1. Multi-functional microfluidic droplet array. (A) Schematic diagram of a triple layered microfluidic device. The bottom images show the assembled device, and zoomed images show the droplet generation and droplet array units, respectively. (B) Schematic diagram of a single unit of the hydrodynamic trap. The top layer is the water reservoir channel to prevent the evaporation of droplets. The middle fluidic channel and bottom control microvalve layer are used for controlling droplets precisely. To precisely control a droplet, the hydrodynamic trap has four functional structures, including microwell, capillary path, reaction chamber, and bypass channel. The trap valve is expanded upward by introducing positive pressure in the control layer. The fluctuation valve is expanded downward by introducing a negative pressure in the control layer. (C) The relative volume changes of the stored droplets in the microfluidic droplet array. Droplet shrinkage is prevented by water flowing through the reservoir to compensate the water evaporation from droplets into the bulk PDMS. Representative images are shown above the markers. The experimental data are obtained from 5 repeated experiments. Device Operation. Our SDA device was operated using a lab-made gas perfusion system that control gas by opening and closing of a solenoid valve.49 Nitrogen gas and an air compressor (GAST P104-AA, IDEX Corporation, USA) were used to control the micro-channels and pneumatic valve by applying positive and negative pressure, respectively. The individual pneumatic valves were automatically actuated by a custombuilt program. Additionally, a syringe pump was used to supply water into the water reservoir channel for a long period. Prior to using the device, we filled the FC40 oil (SigmaAldrich, MO, USA) in the pneumatic valve channels under 0.40 MPa. Then, aqueous samples were loaded into corresponding aqueous sample injection ports, and FC40 oil with 30% 1H,1H,2H,2H-perfluoro-1-octanol (Sigma-Aldrich, MO, USA) as a continuous phase was loaded into a main fluidic channel. The experimental pressure was 0.25 MPa for the oil and 0.6 MPa for the aqueous phase. After generating droplets and making a static array, water was loaded into water reservoir channel with a flow rate of 13 µl/ min for maintaining constant droplet volume. The SDA device was mounted onto the microscope, and the temperature during the assays was measured as 21 to 23 °C.

RESULT AND DISCUSSION Basic Design of Microfluidic SDA. The microfluidic static droplet array (SDA) device was designed to allow studying cell-cell interactions and signaling in a confined microenvi-

ronment that allows long-term culturing of defined cell populations. The device is divided into two functional parts (Figure 1A). The droplet generation unit contains four valves for the controlled generation of multi-component aqueous picodroplets and four individual aqueous sample injection ports. The droplet array unit allows immobilizing and analyzing individual droplets and is composed of 40 hydrodynamic traps in series. The device consists of three PDMS layers with a water reservoir layer (top), fluidic layer (middle), and control layer (bottom) (Figure 1B). The fluidic layer contains the main microfluidic channels where cells are encapsulated into an oil droplet and cultivated in a predefined hydrodynamic trap array. The control layer has several individual microvalves for effective control of fluid. The detailed structure of the hydrodynamic trap is depicted in the bottom of Figure 1A and B, and consists of four structures: a “microwell” to trap the droplets, a “capillary path” that is a controllable narrow channel formed by the partial closing of the trap valve, a “reaction chamber” for storing and merging trapped droplets, and a “bypass channel” that is vertically connected to the entrance of the microwell and the exit of the reaction chamber. The main function of the capillary path is to control the flow of fluids within the microchannel and efficiently trap a droplet in the microwell. In this design, one notable advance is the formation of a capillary path in the hydrodynamic trap using the simple operation of microvalve actuation that stabilizes the droplet in the trap. The principle of droplet trapping depends



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 4 of 9

Figure 2. Unit operation of microfluidic droplet array. (A) The process of double sequential droplet loading. Basic principle of double sequential loading of droplets follows hydrodynamic trapping. The black and white arrows indicate a cell containing droplet (Dc) and media droplet (Dm), respectively, in time-lapsed images. (B) Process of controlling chemical composition. The desired chemical composition could be controlled by storing the double sequential loaded droplet (Dc+m) at the microwell (upper images) and merging it with another chemical reagent droplet (Dr) (lower images). The principle of droplet storage is based on the change of pressure balance between the Laplace and hydraulic pressure. The dashed arrows indicate the direction of droplet movement (C) Linear dilution of chemical composition. We could load variously diluted droplets to the microwell from upper to bottom row so that one-pot analysis of multiple concentration conditions could be performed. The enlarged image of the dotted area indicates the encapsulation of yeast cells into the diluted green fluorescent droplets. Scale bar indicates 100 µm. (D) Droplet retrieval. A single droplet is released by simple actuating trap and fluctuation valves. Scale bar indicates 100 µm. on the competition of fluidic resistances in the microstructures of the hydrodynamic trap.10,26-29,50 The immobilization of the droplets at a specific region (microwell) is caused by the hydrodynamic pressure drop due to both the structure of the hydrodynamic trap and the induced capillary path of the microstructure resulting from the microvalving action. This passive trapping mechanism based on hydrodynamics is robust and insensitive to fluctuations that occasionally caused by gas bubbles during the switching of solutions. Conventional droplet based microfluidic systems suffer from droplet shrinkage attributed to gradual evaporation from the aqueous droplet over time. Several approaches have been suggested to prevent droplet shrinkage such as coating with a specific polymer,51 saturating with water,27 and supplying wa ter through thin PDMS membranes.52,53 However, these approaches yield unstable volumes and are limited by complicated procedures. To eliminate this problem, we thus adopted a water reservoir method52,53 which provides continuous water supply from the reservoir channel to the main microfluidic platform (Figure 1B, C). Specifically, the top layer of the SDA comprises a water reservoir channel that fully covers the droplet array. Moreover, the thickness of the middle layer bonded with the top water reservoir layer is approximately 60 µm, which allows free permeation of water without affecting valving action. Thus, after the generated droplets reached the desired composition and volume, water fluid is passed through the water reservoir channel at a flow rate of 13 µl/ min. As shown in Figure 1D, this arrangement indeed prevented droplet shrinkage, while droplets without a water reservoir shrunk within 4 hours. We were able to keep the droplet volume constant for more than 10 hours, implying that the device is able to precisely control and maintain the droplet volume and thus

allow growing and analyzing cells in a confined microenvironment compatible for most signaling processes. Basic Unit Operations of Microfluidic SDA. A combinatorial droplet array provides a critical advantage to the SDA approach, as any droplet can integrate various elements by sequentially repeating the droplet merging, storage, and mixing. The working mechanism of unit operation is illustrated and described in detail in Figure S-1, while Figure 2A and process 1 in Movie S-1 shows the procedure of double sequential loading of droplets for generating the combinatorial SDA (Figure S-1A). Two droplets encapsulating cell (Dc) and media (Dm) (the volume of each droplet = 25 - 27 pL) were sequentially loaded into the microwell and then merged into one large droplet. This merged droplet with a volume of approx. 50 - 55 pL blocks the horizontal flow into the microwell, and as a result subsequent droplets encapsulating cells (Dc) and medium (Dm), cannot be trapped into the previously occupied microwell and thus bypass toward the next empty microwell (Figure 2A). The cell number in a Dc ranges from 0 to 8, and for droplets containing a single cell, the Poisson distribution matches the experimental result as the occurrence of “single cell encapsulation” becomes rare.54 After the droplets are formed in the microwells, they transfer into the reaction chamber for storage (Figure 2B and Figure S-1B and process 2 in Movie S-1), after which a new droplet can be immobilized in the empty microwell (Figure S-1C). Figure 2B shows the addition of droplet containing reagent (Dr) into pre-stored droplet by opening and closing the trap valve for repeating double sequential loading, droplet transferring and storage. To visualize rapid and controlled addition of chemical compounds, droplets containing individual cells were fused with droplets containing a fluorescent dye (Figure


Page 5 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


2B). As a proof of concept, a concentration gradient of fluorescence intensities was produced in each droplet (Figure 2C), demonstrating that the SDA allows to rapidly and accurately generate dose-response curves. An important characteristic of this SDA is the possibility to retrieve individual droplets from the hydrodynamic traps (Figure 2D and process 3 in Movie S-1).10,26-28 As visualized in Figure 2D, trapped droplets can be released from the reaction chamber by turning the trap valve off and switching the fluctuation valve (Figure S-1D). Firstly, the fluctuation valve is switched to “open” to make a concave-shaped bottom in the reaction chamber. When the trap valve is next switched to “close”, droplets are immobilized into reaction chambers (scheme 1 in Figure 2D). By opening the trap valve and closing the fluctuation valve, the droplets in the reaction chamber are sequentially released (scheme 2 and 3 in Figure 2D), which allows removing droplets without proper cell loading (quality control) and analysis of the signaling process at any desired time point. Thanks to the fully automated procedure, this versatile arrangement results in a performance comparable to that of a conventional microwell system. After completion of the experiment, all droplets can be released with no remaining debris in the SDA, which enables reusing the device. The throughput of the microfluidic system is especially important in light of ongoing efforts to develop SDA devices that permit high-throughput screening. Table S-2 thus summarizes the time required for the unit operation and the throughput for one day, which is compatible with small compound screens. Droplet-Based Cell Synthetic Pheromone Response Assay. As a proof-of-principle application of the multifunctional SDA device, we monitored the mating process of budding yeast as a well-known model to study cell-cell communication. Haploid yeast of opposite mating type use reciprocal pheromones (a- and α -factor) and receptors to signal cell cycle arrest, polarization and eventually fuse to form a single diploid cell (Figure 3A).39 To monitor α-factor-induced gene induction, we analyzed genetically-engineered MATa cells harboring a mating-specific reporter based on the FIG1(Factor Induced Gene 1) promoter driving the expression of the quadruple-Venus fluorescent protein.55 Cells were encapsulated in the SDA and exposed to synthetic pheromone in an array of 40 droplets (4 rows by 10 columns) by sequential droplet merging (Figure 3B). Four different concentrations of α-factor (0, 1, 2, 4 µM) were analyzed in each row, which allows 10 identical, replicated experiments of 4 different conditions simultaneously. As expected, no YFP accumulation could be detected in cells in the absence of α-factor (0 µM), and the cells form buds and progress through the cell cycle (Figure 3B and C). However, the intensity of YFP-expression increased with increasing concentrations of α-factor. Concomitantly the cells stop the cell-cycle and form shmoo-like protrusions. Taken together, these data show that the encapsulated cells are alive and efficiently respond to synthetic α-factor in a dose-dependent manner. Visualizing and Quantifying Yeast Mating in Heterogeneous Cell Populations Trapped in a Confined Microenvironment. We next used the SDA platform to monitor mating of controlled yeast populations in a confined microenvironment. Although the basic mechanisms of pheromone signaling are well understood, later processes involving cell-cell interactions and population effects such as the influence of multiple mating partners remain poorly investigated.56 For example, it

Figure 3. One-pot analysis of synthetic α-factor pheromone response of MATa cell in a microfluidic droplet array device. (a) Schematic diagram of pheromone response process of budding yeast in mating. The haploid yeast cell has two opposite mating types (MATa and MATα) and they secrete mating pheromones; a-factor and α-factor, respectively. The dedicated receptors on the surface of MATa or MATα cells recognize the opposite mating pheromones and trigger the pheromone response pathway. The phenotypic features of mating response are cell cycle arrest, cell polarization and cell fusion. (b) MATa cell response to the synthetic α-factor pheromone droplet. The expression of quadruple Venus fluorescent protein (pFig1-qV) and the morphological change of MATa cell response with 4 different synthetic α-factor concentrations are shown. Scale bar is 50 µm for droplet images and 10 µm for enlarged images, respectively. (c) Quantification of mating reporter gene expression (pFig1-qV) in microfluidic droplet array device. The analysis of 4 different concentrations (0, 1, 2, and 4 µM) are executed in a single device. The batch experiment was performed 5 times and quantified by calculating the average value. In addition, average values from 10 droplets in a single SDA device are shown for specific points. is inferred that the total secretion of α- or a-factor by the cell population would affect the mating response in a mixed population. To address these questions, defined ratio’s of MATa and MATα cells were co-encapsulated into 40 droplets and observed during 4 hours (Figure 4). The number of mixed



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 6 of 9

Figure 4. Natural mating in microfluidic droplet array with population regime control. For visual distinction, one of the mating yeast strains is genetically modified to express fluorescence protein. (a) Formation of the cell co-culture in the microfluidic device. We have designed a multifunctional microfluidic co-culture array system for natural mating. For example, we repeat 4 steps of double sequential loading and droplet storage procedures to control the population ratio of MATa and MATα cells (left image). Scale bar indicates 100 µm for device image and 50 µm for enlarged image. (b) Population regime control in microfluidic SDA. Five groups of cell population regimes: single type of cell encapsulation of MATa or MATα; the same number of MATa and MATα cell encapsulation; one mating type is dominantly encapsulated. Scale bar is 50 µm for droplet images and 10 µm for enlarged image. First, within a single droplet, the cell population ratio (r) is calculated by dividing the number of MATa cells by the total number of cells. (r = 0 or 1, negative control with only MATα or MATa encapsulated, respectively) (total cell number per droplet ≤ 30) (c) Natural mating in droplet. The images are taken at 4hours of droplet co-culturing with wild type, ste12∆, and bar1∆ with wild type of partner cell, respectively (d) Quantitative measurement of natural mating events. Doubling time correlates with varying ratios of MATa and MATα cells. The increase of doubling time results from the pheromone-induced cell cycle arrest. The batch experiment was performed 5 times and quantified by calculating the average value. In addition, average values from 10 droplets in a single SDA device are shown for specific points. MATa and MATα cell in droplets were manipulated by repeating of double sequential loading and droplet storing procedures is shown (Figure 4A, “cell loading procedure” of the Supporting Information). The secreted pheromones (a/αfactor) were preserved in the defined droplet (space volume), and morphological markers such as cell-cycle arrest, shmoo formation and cell fusion were scored (Figure 4C and D). The cell population ratio (r) is calculated by dividing the number of MATa cells by the total number of cells in a single droplet. A cell population ratio of 0 or 1 thus means that only MATα or MATa cells are encapsulated, and this set-up was included as a negative control for the pheromone response assays. When the cell population ratio is 0 < r ≤ 0.33, the number of MATα cells is greater than that of MATa cells and indicated as “MATα > MATa”. When the cell population ratio is 0.33 < r ≤ 0.66, the number of MATα cells is similar to that of MATa cells and indicated this as “MATα ≈ MATa”. Finally, when the cell population ratio is 0.66 < r < 1, the number of MATα cells is less than the number of MATa cells and indicated as “MATα < MATa”. Considering cell proliferation and the microscopic readouts, the average total number of cells encapsulated in one droplet was controlled up to 30. For control, ste12∆ cells were also analysed, which are sterile as they lack the transcription

factor Ste12 known to induce expression of multiple pheromone-induced genes. The doubling time of the encapsulated MATa or MATα cells alone was about 2 hours, similar to one of typical budding yeast cells grown at room temperature (Figure 4D and S-3). Interestingly however, when MATa and MATα cells were co-cultured, their doubling time significantly increased with a particular delay of the subordinate cells compared to the mating partner. As expected, this pheromoneinduced cell cycle arrest was not observed with ste12∆ cells. Consistent with these findings, shmoo formation and cell-cell fusion of wild type cells were also increased in a populationdependent manner (Figure S-3), implying that the dominant population of mating cells secrete more pheromone, which influences the behaviour of the entire population. To corroborate these results, we monitored natural mating of wild type and bar1∆ cells. MATa cells secrete the protease Bar1 that specifically degrades α-factor whereas MATα cells lack an equivalent a-factor–specific protease. It is widely thought that Bar1 secretion serves to “desensitize” the pheromone pathway and restore normal cell division if mating is unsuccessful.48,57 Interestingly, recent studies suggest that Bar1 may play a role in determining the location and distance of mating cells, and may be involved in recognizing cell densi-


Page 7 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


ty and the ratio of different mating partners.46,47 However, this postulated effect is difficult to validate experimentally due to the rapid diffusion of α-factor in conventional mating assays.44-47 We thus encapsulated different ratios of wild-type and bar1∆ cells, and compared the doubling time, shmoo formation and cell-cell fusion of the specific cell types (Figure 4C and D). The doubling time of MATa bar1∆ cells increase dramatically when the number of mutant cells is higher than the wild type (wt α < bar1∆ a), which is even apparent when relatively low amounts of α-factor is produced. In contrast to MATα cells, most MATa bar1∆ cells showed pronounced mating projections. These results demonstrate that the absence of the Bar1 protease not only leads to a non-competitive pheromone response but also to the accumulation of α-factor in the confined droplets. Surprisingly however, these results further suggest that MATa bar1∆ cells also produce less a-factor compared to wild type controls. This may indicate that Bar1 not only cleaves α-factor extracellularly, but may also affect the much less studied synthesis and/or secretion of the hydrophobic a-factor. Alternatively, this effect may be indirectly caused by the excessive amount of α-factor in the confined microenvironment. Further work is now required to understand the mechanism of this interesting observation.

CONCLUSION Altogether, we have developed an accurate and controllable SDA device that allows observing natural yeast mating in defined microenvironments. Importantly, the confined environment preserves secreted signaling molecules, which can differentially affect cell-cell interactions and cellular responses. Using this device, we showed that yeast mating is influenced by socio-biological behaviors in population-dependent cellcell interactions. The SDA platform generates miniaturized batch reactors that are ideally suited to capture complex signaling events common in heterogeneous cell populations. The integrated water reservoir is a critical advance that enables long-term cultivation of cells without loss of viability. The reservoir moreover prevents evaporation and the resulting liquid volume changes that would otherwise alter concentrations of soluble factors and confound quantitative analysis. The SDA system is particularly well-suited to study biological problems requiring stable defined microenvironments and high resolution imaging data for quantitative analysis. Indeed, the SDA device is fabricated specifically to allow use of high numerical aperture objective lenses with short working distances. The SDA device is highly versatile and enables highly efficient trapping, merging, arraying, storage and retrieval of individual droplets. It thus allows to efficiently control various microenvironments for culturing and screening by generating combinatorial droplet arrays followed by analytical in situ analysis of various biological outputs at the single cell level. Our SDA platform with the controllable population ratio and preservation of the microenvironment will enable a myriad of other biological applications. By arraying cellencapsulated droplets in predefined hydrodynamic traps, our SDA platform can perform an enormous number of reactions in parallel or simultaneously and has potential for a wide range of applications from fundamental biology to single-cell analysis. Importantly, individual droplets can be fused to add additional reagents or even other cells, further increasing the versatility of the device. Finally, the SDA has improved high-

throughput capabilities by an engineered “world-to-chip” interface. Therefore, the SDA is not limited to cell analysis or biochemical reactions, which significantly expands the utility and feasibility for screening applications where large parameters need to be explored quickly and affordably.

ASSOCIATED CONTENT Supporting Information The Supporting Information is available free of charge on the ACS Publications website. Figure S-1. A schematic diagram regarding the operation of valves. Figure S-2. Enlarged image of Figure 4A. Figure S-3. Droplet-based pheromone response assay. Table S-1. Yeast strains used in this study. Table S-2. Specification of programmable static droplet array. Movie S-1. Demonstration of microfluidic static droplet array.

AUTHOR INFORMATION Corresponding Author *Phone: +41 44 633 6586. E-mail: [email protected]. *Phone: +82 42 821 5896. E-mail: [email protected]

Author Contributions ¶

These authors contributed equally.

Notes The authors declare no competing financial interest.

ACKNOWLEDGMENT We are grateful to Dr. Fabric Caudron for sharing yeast strains, the Peter group and ScopeM for helpful discussions and Dr. Alicia Smith for critical comments on the manuscript. This research was supported by Global Research Laboratory (NRF2015K1A1A2033054) through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT (Information and Communication Technologies) and Future Planning. Work in the Peter-laboratory was supported by an ERC senior award, the Swiss National Science Foundation (SNF), SystemsX.ch and the ETH Zürich.

REFERENCES (1) Chen, E. H.; Grote, E.; Mohler, W.; Vignery, A. FEBS Lett. 2007, 581, 2181-2193. (2) Kempe, S.; Kestler, H.; Lasar, A.; Wirth, T. Nucleic Acids Res. 2005, 33, 5308-5319. (3) Klambt, C.; Jacobs, J. R.; Goodman, C. S. Cell 1991, 64, 801815. (4) Stern, E. L.; Quan, N.; Proescholdt, M. G.; Herkenham, M. J. Neuroimmunol. 2000, 106, 114-129. (5) TessierLavigne, M.; Goodman, C. S. Science 1996, 274, 11231133. (6) Hansen, S. K.; Rainey, P. B.; Haagensen, J. A. J.; Molin, S. Nature 2007, 445, 533-536. (7) Lidstrom, M. E.; Konopka, M. C. Nat. Chem. Biol. 2010, 6, 705-712. (8) Wang, M. Z.; Schaefer, A. L.; Dandekar, A. A.; Greenberg, E. P. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 2187-2191. (9) Frank, T.; Tay, S. Lab Chip 2015, 15, 2192-2200. (10) Jin, S. H.; Jeong, H. H.; Lee, B.; Lee, S. S.; Lee, C. S. Lab Chip 2015, 15, 3677-3686. (11) Ottesen, E. A.; Young, C. R.; Eppley, J. M.; Ryan, J. P.; Chavez, F. P.; Scholin, C. A.; DeLong, E. F. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, E488-E497.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

(12) Conlon, P.; Gelin-Licht, R.; Ganesan, A.; Zhang, J.; Levchenko, A. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, E5896E5905. (13) Ricicova, M.; Hamidi, M.; Quiring, A.; Niemisto, A.; Emberly, E.; Hansen, C. L. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 1140311408. (14) Vyawahare, S.; Zhang, Q. C.; Lau, A.; Austin, R. H. Adv. Drug Delivery Rev. 2014, 69, 217-224. (15) Kim, M. K.; Ingremeau, F.; Zhao, A. S.; Bassler, B. L.; Stone, H. A. Nat. Microbiol. 2016, 1. (16) Darmanis, S.; Sloan, S. A.; Zhang, Y.; Enge, M.; Caneda, C.; Shuer, L. M.; Gephart, M. G. H.; Barres, B. A.; Quake, S. R. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 7285-7290. (17) Jin, S. H.; Jang, S. C.; Lee, B.; Jeong, H. H.; Jeong, S. G.; Lee, S. S.; Kim, K. P.; Lee, C. S. Lab Chip 2016, 16, 1358-1365. (18) Min, S. K.; Lee, B. M.; Hwang, J. H.; Ha, S. H.; Shin, H. S. Korean J. Chem. Eng. 2012, 29, 392-395. (19) Moore, T. I.; Tanaka, H.; Kim, H. J.; Jeon, N. L.; Yi, T. M. Mol. Biol. Cell 2013, 24, 521-534. (20) Taylor, R. J.; Falconnet, D.; Niemisto, A.; Ramsey, S. A.; Prinz, S.; Shmulevich, I.; Galitski, T.; Hansen, C. L. Proc. Natl. Acad. Sci. U. S. A. 2009, 106, 3758-3763. (21) Jo, M. C.; Liu, W.; Gu, L.; Dang, W. W.; Qin, L. D. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 9364-9369. (22) Escalante-Chong, R.; Savir, Y.; Carroll, S. M.; Ingraham, J. B.; Wang, J.; Marx, C. J.; Springer, M. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 1636-1641. (23) Moore, T. I.; Chou, C. S.; Nie, Q.; Jeon, N. L.; Yi, T. M. PLoS One 2008, 3, e3865. (24) Paliwal, S.; Iglesias, P. A.; Campbell, K.; Hilioti, Z.; Groisman, A.; Levchenko, A. Nature 2007, 446, 46-51. (25) Lee, S. S.; Horvath, P.; Pelet, S.; Hegemann, B.; Lee, L. P.; Peter, M. Integr. Biol. 2012, 4, 381-390. (26) Jang, S.; Lee, B.; Jeong, H. H.; Jin, S. H.; Jang, S.; Kim, S. G.; Jung, G. Y.; Lee, C. S. Lab Chip 2016, 16, 1909-1916. (27) Jeong, H. H.; Jin, S. H.; Lee, B. J.; Kim, T.; Lee, C. S. Lab Chip 2015, 15, 889-899. (28) Jeong, H. H.; Lee, B.; Jin, S. H.; Jeong, S. G.; Lee, C. S. Lab Chip 2016, 16, 1698-1707. (29) Sun, M.; Bithi, S. S.; Vanapalli, S. A. Lab Chip 2011, 11, 3949-3952. (30) Agresti, J. J.; Antipov, E.; Abate, A. R.; Ahn, K.; Rowat, A. C.; Baret, J. C.; Marquez, M.; Klibanov, A. M.; Griffiths, A. D.; Weitz, D. A. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 4004-4009. (31) Huang, M. T.; Bai, Y. P.; Sjostrom, S. L.; Hallstrom, B. M.; Liu, Z. H.; Petranovic, D.; Uhlen, M.; Joensson, H. N.; AnderssonSvahn, H.; Nielsen, J. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, E4689-E4696. (32) Terekhov, S. S.; Smirnov, I. V.; Stepanova, A. V.; Bobik, T. V.; Mokrushina, Y. A.; Ponomarenko, N. A.; Belogurov, A. A., Jr.; Rubtsova, M. P.; Kartseva, O. V.; Gomzikova, M. O.; Moskovtsev, A. A.; Bukatin, A. S.; Dubina, M. V.; Kostryukova, E. S.; Babenko, V. V.; Vakhitova, M. T.; Manolov, A. I.; Malakhova, M. V.; Kornienko, M. A.; Tyakht, A. V.; Vanyushkina, A. A.; Ilina, E. N.; Masson, P.; Gabibov, A. G.; Altman, S. Proc. Natl. Acad. Sci. U. S. A. 2017, 114, 2550-2555. (33) Wang, B. L.; Ghaderi, A.; Zhou, H.; Agresti, J.; Weitz, D. A.; Fink, G. R.; Stephanopoulos, G. Nat. Biotechnol. 2014, 32, 473-478. (34) Sjostrom, S. L.; Bai, Y. P.; Huang, M. T.; Liu, Z. H.; Nielsen, J.; Joensson, H. N.; Svahn, H. A. Lab Chip 2014, 14, 806-813. (35) Jeong, H. H.; Issadore, D.; Lee, D. Korean J. Chem. Eng. 2016, 33, 1757-1766. (36) Fallah-Araghi, A.; Baret, J. C.; Ryckelynck, M.; Griffiths, A. D. Lab Chip 2012, 12, 882-891. (37) Abate, A. R.; Hung, T.; Mary, P.; Agresti, J. J.; Weitz, D. A. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 19163-19166. (38) Niu, X. Z.; Gielen, F.; Edel, J. B.; deMello, A. J. Nat. Chem. 2011, 3, 437-442. (39) Bardwell, L. Peptides 2004, 25, 1465-1476. (40) Schrick, K.; Garvik, B.; Hartwell, L. H. Genetics 1997, 147, 19-32.

Page 8 of 9

(41) Gustin, M. C.; Albertyn, J.; Alexander, M.; Davenport, K. Microbiol. Mol. Biol. Rev. 1998, 62, 1264-1300. (42) Nern, A.; Arkowitz, R. A. Nature 1998, 391, 195-198. (43) Elia, L.; Marsh, L. J. Cell Biol. 1998, 142, 1473-1485. (44) Wiget, P.; Shimada, Y.; Butty, A. C.; Bi, E. R.; Peter, M. EMBO J. 2004, 23, 1063-1074. (45) Huh, W.-K.; Falvo, J. V.; Gerke, L. C.; Carroll, A. S.; Howson, R. W.; Weissman, J. S.; O'Shea, E. K. Nature 2003, 425, 686691. (46) Banderas, A.; Koltai, M.; Anders, A.; Sourjik, V. Nat. Commun. 2016, 7, 12590. (47) Diener, C.; Schreiber, G.; Giese, W.; del Rio, G.; Schroder, A.; Klipp, E. PLoS Comput Biol 2014, 10, e1003690. (48) Manney, T. R. J Bacteriol 1983, 155, 291-301. (49) Unger, M. A.; Chou, H. P.; Thorsen, T.; Scherer, A.; Quake, S. R. Science 2000, 288, 113-116. (46) Banderas, A.; Koltai, M.; Anders, A.; Sourjik, V. Nat Commun 2016, 7, 12590. (50) Tan, W. H.; Takeuchi, S. Proc. Natl. Acad. Sci. U. S. A. 2007, 104, 1146-1151. (51) Solvas, X. C. I.; Turek, V.; Prodromakis, T.; Edel, J. B. Lab Chip 2012, 12, 4049-4054. (52) Dewan, A.; Kim, J.; McLean, R. H.; Vanapalli, S. A.; Karim, M. N. Biotechnol. Bioeng. 2012, 109, 2987-2996. (53) Shim, J. U.; Cristobal, G.; Link, D. R.; Thorsen, T.; Jia, Y. W.; Piattelli, K.; Fraden, S. J. Am. Chem. Soc. 2007, 129, 8825-8835. (54) Abate, A. R.; Chen, C. H.; Agresti, J. J.; Weitz, D. A. Lab Chip 2009, 9, 2628-2631. (55) Pelet, S.; Peter, M. In Design and Analysis of Biomolecular Circuits: Engineering Approaches to Systems and Synthetic Biology, Koeppl, H.; Setti, G.; di Bernardo, M.; Densmore, D., Ed.; Springer New York: New York, 2011; pp 369-393. (56) Youk, H.; Lim, W. A. Science 2014, 343, 1242782. (57) Chan, R. K.; Otte, C. A. Mol Cell Biol 1982, 2, 21-29.


Page 9 of 9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


for TOC only


Programmable Static Droplet Array for the Analysis of CellâCell

Recommend Documents

Programmable Static Droplet Array for the Analysis of CellâCell