Figure 1. Multiplexed microfluidic device for simultaneous chemotaxis assays.

(a,b) Continuous flow through a microfluidic junction (a) stratifies chemostimulus, cell, and buffer solutions, demonstrated here with fluorescein, DI water, and DI water, respectively. Upon halting the flow (b) diffusion establishes a chemical gradient across the channel, which is repeated at each observation channel in the MCD (d, red and orange boxes). Scale bars, 0.1 mm. (c) Assembly of the MCD showing the PDMS dilution layer (blue) and cell injection layer (red) microchannels mounted on a glass slide (grey; Materials and methods). (d) Scaled drawing of the dilution layer, which receives chemical (pressure, $p_{in,1}$) and buffer ($p_{in,2}$) solutions. Initial chemical concentration (C₀) is sequentially diluted 10-fold to each of four additional concentrations ($C_{1-4}$), plus a control solution ($C_5=0$). These six chemostimulus solutions are merged separately with additional cell ($A_i$) and buffer ($B_i$) solutions from the cell injection layer (e) for chemotaxis assays in respective observation channels (dashed black box, corresponding to c and f). (e) Scaled drawing of the cell injection layer which injects a cell suspension ($p_{in,3}$) and buffer solution ($p_{in,4}=p_{in,3}$) into the dilution layer ($A_i,B_i$; Materials and methods). Scale bars d,e, 2 mm. (f) Photograph of the completed MCD with dyed water to visualize the chemical (yellow), cell (red), and buffer (blue) fluid streams in the channel network. Scale bar, 5 mm. (g) Stratified chemical (C₀), cell, and buffer solutions in the first observation region (d, red box). (h) Dilution of the chemical (C₀) by the buffer prior to mixing in the first micromixer (Stroock et al., 2002) to produce concentration C₁ (d, purple box). (i), Stratified chemical solution after initial dilution (C₁, green) in the second observation region (d, orange box). Scale bars g-i, 0.2 mm. (j) Measured chemical concentrations (see Materials and methods) generated from the dilution microchannels (d) for various driving pressures $p_{in,1}=p_{in,2}=[100,150,200]\mathrm{mbar}$ (square, circle, and triangle, respectively). (k) Measured evolution of the chemical gradient (b) produced in the C₀ observation region (Figure 1—figure supplement 3; Materials and methods) by the MCD shows the chemical diffusion across the channel with increasing time t.

Figure 1—figure supplement 1. Micromixer geometry and mixing performance.

(a,b) The herringbone micromixer (Stroock et al., 2002) - used to incorporate chemical and buffer during serial dilution - consists of a main rectangular channel ($W=200\mu \mathrm{m}$, $H=90\mu \mathrm{m}$) with herringbone ridges ($X=50\mu \mathrm{m}$, $d=30\mu \mathrm{m}$), which enhance mixing by generating transverse flow. Each herringbone half cycle ($\approx 700\mu \mathrm{m}$ long) consists of six ridges (pitch, θ = 45°; a). The distance, q, from the sidewall alternates every half cycle between 2 W/3 and W/3. (c) The degree of mixing (DOM; Materials and methods) (Stroock et al., 2002) was quantified for a long test channel having 29 herringbone cycles for two different flow rates: $22\mu \mathrm{l}{\min }^{-1}$ (blue triangles) and $222\mu \mathrm{l}{\min }^{-1}$ (red circles) using fluorescein dye (Petrášek and Schwille, 2008) ($D=436\mu {\mathrm{m}}^2{\mathrm{s}}^{-1}$). Distance indicates the downstream position from the point where the dye and water solutions meet. The solution is considered mixed when $\text{DOM}\le 0.05$ (90% complete mixing, dashed line), which is achieved after 9 cycles for both flow rates. The final micromixer design comprised 26 herringbone cycles (see Figure 1—figure supplements 2 and 4) to ensure complete mixing across a range of potential chemostimulants.

Figure 1—figure supplement 2. Hydraulic circuit design of MCD dilution layer and cell injection layer.

(a) Circuit representation of the dilution layer (Sugiura et al., 2010) of the MCD (Figure 1d). The two inlets ($p_{in,1}$, $p_{in,2}$) receive the base chemical solution (C₀) and the buffer solution ($C=0$), respectively, and the outlet is open to atmospheric pressure ($p_{out}=0\mathrm{mbar}$). The hydraulic network was designed such that a prescribed inlet pressure $p_{in,1}=p_{in,2}=100\mathrm{mbar}$ from a pressure controller produces outlet flow rates of $Q_{out}=20\mu \mathrm{l}{\min }^{-1}$. The resistances of the bridge channels ($R_B$), micromixer channels ($R_M$), and observation channels ($R_{4,4}$) were predetermined (Materials and methods). The remaining flow rates, pressures, and hydraulic resistances were determined by solving a linear set of equations for the hydraulic circuit (Kirby, 2010), and the resulting resistances determine the individual channel geometries (Appendix 1). (b) Circuit representation of the cell injection layer of the MCD (Figure 1e). The cell injection layer network was designed to flow the cell ($p_{in,3}$) and buffer ($p_{in,4}$) solutions into the dilution layer to ensure the cell solution would be centered in the observation region. To do so, the applied inlet pressures were designed to be $p_{in,3}=p_{in,4}=50\mathrm{mbar}$ and flow rates in each microchannel for the cell ($Q_{CIL,1}$) and buffer ($Q_{CIL,2}$) were $5\mu \mathrm{l}{\min }^{-1}$ and, $20\mu \mathrm{l}{\min }^{-1}$ respectively. The resulting channel resistances and dimensions were determined using the Hagen-Poiseuille law (Kirby, 2010; Oh et al., 2012; Materials and methods), and accounted for the resistance of the observation region ($R_{4,4}$; Appendix 1). The labels in a and b corresponds to that in Figure 1d and e.

Figure 1—figure supplement 3. Validation of chemostimulus distribution and gradient evolutionin in MCD observation regions.

(a,b) The initial symmetry of the stratified chemical, cell, and buffer solutions flowing in the observation channel was verified by measuring the widths of the three fluid streams through fluorescent dye visualization (Materials and methods). (a) A fluorescein solution was flowed through both inlets of the dilution layer (chemical and buffer inlets; Figure 1d) and also the buffer solution inlet of the cell injection layer (Figure 1e). Solid red and dashed blue lines indicate measured channel wall and stream interface locations, respectively, for a driving pressure $p_{in,1}=p_{in,2}=200\mathrm{mbar}$ and $p_{in,3}=p_{in,4}=140\mathrm{mbar}$ (Figure 1—figure supplement 2). Scale bar, 100 μm. The normalized measured widths,$w_i^{\prime }=w_i/W$, of the chemical, cell, and buffer streams illustrate their high degree of symmetry across all observationchannels ($C_{0-5}$); where w_i are the fluid stream widths (a; dashed blue lines) and $W$ is the total width of the intensity profile (a; solid red lines). The bar plots and error bars indicate the mean and standard deviation of the normalized stream widths, investigated for three different driving pressures of the dilution layer and cell injection layer: 100/70 mbar, 150/105mbar, and 200/140 mbar ($p_{in,1-2}/p_{in,3-4}$). The applied inlet pressure of the cell injection layer was tuned (Materials and methods) until symmetric flow was achieved in the observation channels. The mean percent difference of the measured widths of the chemical, cell, and buffer solution relative to the theoretical widths (Materials and methods) was only 2.6%, 3.0%, and 1.8%, respectively. (c,d) The uniformity of the chemostimulus gradient evolution across the various observation regions ($C_{0-5}$) was determined through fluorescent dye (fluorescein) visualization of stop-flow experiments (Materials and methods). Each observation region started with the same initial (fixed) concentration ($C_i=C_0$) by injecting the same dye concentration into both the chemical and buffer inlets of the dilution layer (see also Figure 2—figure supplement 1b). The fluorescence intensity enables visualization of the chemical concentration profile evolution across each channel over the course of approximately 9 min (starting from $t=0$ post flow). Measurements were performed for two different driving pressures to ensure that the resulting gradients were insensitive to flow rate: 100/70 mbar (c) and 200/140 mbar (d).

Figure 1—figure supplement 4. Two-layer photolithography and soft lithography microfabrication of the MCD.

(a–c) Herringbone micromixers (Stroock et al., 2002) were fabricated in the dilution layer of the MCD using multilayer photolithography (subsection of full photomasks shown; see also Figure 1). The first photolithography layer produced the rectangular cross-section main channel (a; 200 μm wide), and a second photolithography layer formed the herringbone ridges on top (Anderson et al., 2000) (b; see also Figure 1—figure supplement 1). Complementary alignment markers were used to align the herringbone ridges (b, hollow cross and circle) with the main channel (a, cross). (c) Overlay of the herringbone micromixer channel components from a and b (see also Figure 1—figure supplement 1). (d,e) The dilution layer (d, blue) and cell injection layer (e, red) PDMS microchannels for the MCD were cast individually from SU-8 molds via soft lithography (McDonald et al., 2000; Materials and methods). The dilution layer microchannel was first plasma bonded (McDonald et al., 2000) to a standard double-wide glass slide (d; grey; 75 mm × 50 mm×1 mm). Subsequently, the cell injection layer was aligned and plasma bonded (Eddings et al., 2008; McDonald et al., 2000) on top of the dilution layer (e). (f,g) Side and top-down view of the assembled MCD, respectively, showing the locations of the four inlets, single outlet, and observation regions (dashed box; see also Figure 1).