Microfluidic Actuation via 3D-Printed Molds toward Multiplex

Jul 19, 2019 - Multiplexed analysis of biochemical analytes such as proteins, enzymes, and immune products using a microfluidic device has the potenti...
0 downloads 0 Views 4MB Size
Subscriber access provided by University of Rochester | River Campus & Miner Libraries

Article

Microfluidic actuation via 3D-printed molds towards multiplex biosensing of cell apoptosis Bac Van Dang, Amin Hassanzadeh-Barforoushi, Maira Syed, Danting Yang, Sung-Jin Kim, Robert A. Taylor, Guo-Jun Liu, Guozhen Liu, and Tracie Barber ACS Sens., Just Accepted Manuscript • DOI: 10.1021/acssensors.9b01057 • Publication Date (Web): 19 Jul 2019 Downloaded from pubs.acs.org on July 20, 2019

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 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 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.

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 39 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

ACS Sensors

Microfluidic actuation via 3D-printed molds towards multiplex biosensing of cell apoptosis Bac Van Dang1 ‡, Amin Hassanzadeh-Barforoushi1,2 ‡, Maira Shakeel Syed1, Danting Yang3,4, Sung-Jin Kim5, Robert Taylor1, Guo-Jun Liu6,7, Guozhen Liu3*, Tracie Barber1*

1School

of Mechanical and Manufacturing Engineering, Faculty of Engineering, The University of New South Wales, Sydney, NSW 2052, Australia

2Cancer

Division, Garvan Institute of Medical Research/the Kinghorn Cancer Centre, Sydney, NSW 2010, Australia

3Graduate

School of Biomedical Engineering, ARC Centre of Excellence in Nanoscale

BioPhotonics, Australian Centre for NanoMedicine, Faculty of Engineering, The University of New South Wales, Sydney, NSW 2052, Australia 4Department

of Preventative Medicine, Zhejiang Provincial Key Laboratory of Pathological

and Physiological Technology, Medical School of Ningbo University, Ningbo, Zhejiang 315211, China 5Department

of Mechanical Engineering, Konkuk University, Seoul, 05029, Republic of Korea

1

ACS Paragon Plus Environment

ACS Sensors 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

6Australian

Nuclear Science and Technology Organisation, New Illawarra Road, Lucas Heights, NSW 2234, Australia.

7Discipline

of Medical Imaging & Radiation Sciences, Faculty of Medicine and Health, Brain

and Mind Centre, University of Sydney, 94 Mallett Street, Camperdown, NSW 2050, Australia.

KEYWORDS: Microfluidic valve, multiplexed microfluidics, multiplexed biosensing, 3D printing, cell apoptosis, colorimetric assay

2

ACS Paragon Plus Environment

Page 2 of 39

Page 3 of 39 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

ACS Sensors

ABSTRACT Multiplexed analysis of biochemical analytes such as proteins, enzymes and immune products using a microfluidic device has the potential to cut assay time, reduce sample volume, realize high-throughput, and decrease experimental error without compromising sensitivity. Despite these huge benefits, the need for expensive specialized equipment, and the complex photolithography fabrication process for the multiplexed devices has, to date, prevented widespread adoption of microfluidic systems. Here, we present a simple method to fabricate a new microfluidic-based multiplexed biosensing device by taking advantage of 3Dprinting. The device is an integration of normally closed (NC) microfluidic valving units which offer superior operational flexibility by using PDMS membrane (E~1-2 MPa) and require minimized energy input (1-5 kPa). To systematically engineer the device, we first report on the geometrical and operational analysis of a single 3D-printed valving unit. Based on the characterization, we introduce a full prototype multiplexed chip comprised of several microfluidic valves. The prototype offers—for the first time in a 3D-printed microfluidic device—the capability of on-demand performing of both a sequential and a parallel biochemical assay. As a proof of concept, our device has been used to simultaneously measure the apoptotic activity of 5 different members of the caspase protease enzyme family. In summary, the 3D-printed valving system showcased in this study overcomes traditional bottlenecks of microfabrication, enabling a new class of sophisticated liquid manipulation required in performing multiplexed sensing for biochemical assays.

3

ACS Paragon Plus Environment

ACS Sensors 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

In the last decade, performing sensitive, high throughput biochemical assays have become a reality through development of case-specific microfluidic devices and integrated systems1,2. Recent influential examples of such developments include biomimetic models,3,4 biomolecular analysis,5 drug discovery,6,7 and single cell analysis.8,9 At the core of these systems is an advanced liquid-handling mechanism, such as mixing,10 pumping,11 and droplet manipulation.12 Thus, the key factor that leads to successful operation of these microfluidicbased assays is the ability to precisely deliver liquids on demand. A simple-to-fabricate liquid-handling mechanism would enable much wider adoption of microfluidic systems for countless chemical and biochemical assay applications. As a leading liquid-handling approach, microfluidic valves have been designed to connect or isolate liquid-containing microchannels and microchambers upon actuation, through passive,13 electrokinetic,14 pinch15 and phase-changing16 methods. Membrane-based microvalves (categorized into normally open17 (NO) and normally closed18,19 (NC) valves) are particularly attractive for assay applications because they offer simplicity in operation via the deflection of a thin membrane2. In addition to their operational simplicity, NC microvalves are becoming increasingly important in sophisticated microfluidic circuits since it offer low operation pressure20, and can be tailored to a wide range of actuating pressures21,22 and frequencies.23,24 In contrast to these advantages, the microvalve-based system’s design and fabrication process has historically been complex, time-consuming, and costly, preventing its wider applications. In the conventional process, two or more photomasks and photoresist layers are required along with a highly experienced-based alignment procedure between UVexposure/chemical development steps in a clean room environment.25

4

ACS Paragon Plus Environment

Page 4 of 39

Page 5 of 39 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

ACS Sensors

During the last few years, 3D printing has emerged as a rapid and cost-effective tool for fabricating microfluidic systems with increasing resolution and quality.26,27 With respect to the fabrication of microvalve-based devices, one-step fabrication of the whole valving system has been reported recently.28,29 For example, Folch’s group fabricated, via stereolithography (SL), a microvalve entirely from WaterShed plastic with a membrane thickness of 115 µm.28 A later design from the Folch group used a 3D-printed Quake-style valve based on poly(ethylene diacrylate) (PEG-DA-258) with a thinner, 10-25µm-thick, membrane to simplify the valve’s architecture and piping for large-scale arrays.30 In these designs, however, the range of the valve’s operating pressures and actuation frequencies were limited by the elasticity of the printing materials (for PEG-DA-258, E ∼ 130 MPa). In another attempt, Nordin’s group reported a new procedure incorporating a modified resin for printing durable microvalves with TM of ~50 µm for a valve with a 1.08 mm diameter using a digital light processing-stereolithographic (DLP-SLA) 3D printer.29 However, this valve also required relatively high control pressure (4-12 psi) and draining of the PEG-DA-258 precursor after printing proved to be a challenge. In a different approach, 3D printers were utilized recently to create master molds that can be used for soft lithographic fabrication of NO valves using PDMS.31,32 However, 3D printing for fabricating NC microvalves has not yet been reported. Due to the zero distance of membrane-to-valve-seat in the initial state, current methods of direct 3D-printing, which require removal of supporting material, are not able to print this type of microvalve. In this paper, we introduce the use of 3D-printed molds for fabricating NC valves and present it as a new way for fabricating microvalve-based multiplexed microfluidic devices. With this method, the complications associated with either photolithography or one-step 3D printing of the whole microfluidic system are eliminated. The NC valves fabricated with this 5

ACS Paragon Plus Environment

ACS Sensors 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

method use PDMS membranes and therefore can be operated at significantly lower operating pressures (e.g < 5 kPa as compared to 25-220 kPa reported in previous 3D-printed valve studies).28,29,32 Based on the characterization of a single valving unit, we introduce a multiplex microfluidic chip with the capability of performing both sequential and parallel logics required for high-throughput biochemical assays. Most of the previous microfluidic systems were introduced with only one of the sequential33,34 or parallel11,35 fluidic manipulations and yet demand sophisticated computer control and complex peripheral equipment. CD microfluidic systems have been introduced with the capability of performing sequential and parallel assays; however, they do not offer the operational flexibility provided by valve-based systems and their fabrication is relied on the complex lithographic processes.36 The device presented here also outperforms other non-microfluidic based multiplexed sensing methods37,38 by increasing the assay throughput as well as eliminating the complications associated with assay design and setup. As a proof of concept, our multiplex microfluidic device has been used to measure the apoptotic activity of different members of the caspase protease enzymes family, that are playing essential roles in programmed cell death. This is achieved through controlling the sequential loading and capturing of colorimetric sensors in microfluidic channels isolated at the valves' close state and running the apoptotic samples over the coated channels connected at the valves' opening state. This study aims to: (1) analyze and fabricate NC-valve-integrated multiplexed microfluidic chip based on 3D printing; (2) perform on the chip both sequential and parallel logics, which required for high-throughput biochemical assays.

EXPERIMENTAL SECTION 3D- printed Molding for Microvalve Fabrication 6

ACS Paragon Plus Environment

Page 6 of 39

Page 7 of 39 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

ACS Sensors

Figure 1. Schematic illustration of fabricating valve-based microfluidic devices via 3Dprinted molds. (A) (i) Individual molds were 3D-printed for the valve top and bottom layers, (ii) Soft lithography on 3D-printed molds and casted PDMS layers; a thin PDMS membrane was fabricated using spin coater at a certain speed; (iii) Schematic of the assembled normallyclosed microfluidic actuation unit. (B) Top view and Side view of the valve. (C) Crosssectional view demonstrating the valve operation mechanism. As shown in Fig. 1A, we suggest here a simple additive manufacturing technique which employs 3D printers for the fabrication of NC membrane microvalving system as small as 100 µm with the aim to replace the previous laborious photolithographic fabrication of these valves18 while keeping the high flexibility advantage of a PDMS membrane in the valve structure. Single valve -top and -bottom layers features are made by 3D-printed molds (Fig. 1A(i)). PDMS pre-polymer is casted on the molds (soft-lithography process) to create the top and bottom PDMS layers. Subsequently, the PDMS slabs are bonded together with a thin 7

ACS Paragon Plus Environment

ACS Sensors 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 8 of 39

PDMS membrane as a middle layer (Fig. 1A(ii)). The final product includes the main fluid and secondary fluid/ gas control channels (as shown in Fig. 1A(iii) and 1B). 3D-printed Valve Operation and Design Considerations The structure of the valve follows the initial membrane valve designs proposed by Grover et al.18 As shown in Fig. 1C, the 3D-printed NC elastomeric valve, consists of three layers: The top layer (gray color) with pressures 𝑃I and 𝑃Oin the inlet and outlets respectively, the bottom (green color) layer with pressure 𝑃C and a middle layer of PDMS membrane (pink color). The valve corresponds to an electronic transistor that inputs a variable resistance to the fluidic stream.39 The key factor in designing valve-based microfluidic system is determination of the geometry of each single valve and its membrane. To do this, a numerical parametric study was performed by solving for the valve's deflection under three different geometrical parameters of valve size (SV), valve-seat size (WS) and membrane thickness (TM) (Table 1). As can be seen in Fig. S1B, decreasing the PC and TM and increasing the SV can be used to achieve higher membrane deflection. Depending on the maximum deflection, the bottom control layer was designed to make sure the valve can be fully opened (see supplementary eqn. (S2)). Table 1: Geometrical dimensions of the 3D-printed valves. Valve size- SV (mm2)

Valve-seat size- WS (µm)

Membrane thickness- TM (µm)

1×1

300

30

2×2

600

70

4×4

900

110

Fabrication Procedure

8

ACS Paragon Plus Environment

Page 9 of 39 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

ACS Sensors

The CAD models of the molds for both top and bottom layers were designed using SolidWorks 2017 and 3D-printed with ProJet® 3500 HD Max, 3D printer (3D systems, SC, USA). This 3D printer is based on the multi-jet manufacturing (MJM) technology that enables it to print small models with high surface quality with a printing resolution of 750×750×1600 dpi and 16μm per layer. Based on this resolution, channels and valves with sizes as small as 100 µm were successfully and reliably fabricated. The fabrication procedure is illustrated in Fig. 2A and detailed in Supplementary section II. The final products are shown in Fig. 2B(i) for the actuation unit and in Fig. 2B(ii) for the multiplex microfluidic chip, a system of multiple valves.

Figure 2. 3D printing process and experiment set up. (A) Fabrication steps for the valving system using 3D-printed molds. (B) Image showing the single valve as well as a multiplexed

9

ACS Paragon Plus Environment

ACS Sensors 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

biosensing device made of PDMS from 3D-printed molds. (C) Schematic of the experimental setup. (D) Microscopic pictures of valve opening and closing states.

Fluidic Circuit Setup As shown in Fig. 2C, the valve control channels were connected to the microfluidic pressure & vacuum pump (AF1-dual, Elveflow, Paris, France) (pressure range from -700 mbar to 1bar). Actual controlling pressure is measured by disposable pressure transducers (range 0-36 kPa, Utah Medical Products Inc, Utah, USA) which were controlled by a National Instruments NI USB-9174 Digital I/O device connected to a power supply at constant voltage 10V. LabVIEW software (National Instruments, TX, USA) was used to handle the measurement endpoints. For the single-valve characterization, the inlet of the top layer was supplied by hydrostatic pressure of an inlet well held at a certain height Hi while the outlet is connected to an outletwell at the zero surface. To show opening and closing states, the inlet of the flow channel is connected to a bottle containing blue food coloring dye (as shown in Fig 2D). The flow rate was measured by weighing the outlet liquid by an electronic balance. All the readings from the balance were recorded at an interval of 3 minutes. Rhodamin B was used as the working fluid to measure the change in the valve's opening ratio with the applied PC. The images were taken using an Olympus IX73 epifluorescent inverted microscope with ImageJ software (NIH, Bethesda, USA) used for post-processing. Cell Apoptosis The Jurkat cells used in this study were 82% identical to Jurkat Clone E6–1 (ATCC: TIB152), determined using short tandem repeat (STR) profiling.40 The Jurkat cells were cultured in the Dulbecco's Modified Eagle's Medium (DMEM), supplemented with 10% v/v fetal calf serum (FCS) (Invitrogen, Carlsbad, CA, USA), 2 mM glutamine, and 1% penicillin and 10

ACS Paragon Plus Environment

Page 10 of 39

Page 11 of 39 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

ACS Sensors

streptomycin in 5% CO2 at 37°C before and during induction of apoptosis. The induced apoptosis by UV and 𝛾-radiation was performed according to Gottlieb et al.41 Syljuâsen & McBride42 and the manufacturer BioVision’s instruction for Caspase-3/CPP32 Colorimetric Assay kit. The complete procedure for inducing cell apoptosis is reported in supplementary section III. Protocol for PDMS Surface Modification The surface of the sequential microfluidic device was functionalized to accommodate the pnitroaniline (pNA) substrate (caspase- 1, -2, -3, -6, -8 colorimetric assay kit, Biovision, Milpitas, USA) on the channels for subsequent cleavage of the substrate with caspase present in the sample. Initially, the PDMS surface was cleaned by quickly running piranha etch solution in the channel (H2SO4:H2O2, 3:1, v/v), followed by a generous wash with DI water. Experimental error due to overexposure with this strong acidic solution can be seen at Supplementary section IV. The device was then immersed in 20 mM 3-aminopropyl triethoxysilane (APTES) solution for 35 min (5 min in sonication and 30 min to rest in the APTES solution) to form -NH2 group on the surface. Next, channels were rinsed 5 times with ethanol and dried in vacuum oven at 80 °C for 1.5 h. Activated pNA substrate by 1-ethyl3-(3dimethylaminopropyl) carbodiimide hydrochloride (EDC) solution was immobilized onto the surface of channels by incubating with -NH2 modified surface of the channels for 1 h at room temperature. After the surface modification of pNA substrates in different channels, 100 µg of caspase protein solution and DTT containing reaction buffer were injected into the device and incubated for 2 h at room temperature before measuring absorbance using SpectraMax iD5 microplate reader (Molecular Devices, San Jose, USA). Quantification of Colorimetric Signal in Microfluidic Device by Plate Reader

11

ACS Paragon Plus Environment

ACS Sensors 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 12 of 39

The microfluidic device was positioned on the head of a 96 well plate to be able to identify the location of each channel in the microplate reader. The microplate reader was set up at 400 and 405 nm absorbance wavelength and the data were extracted for the wells overlapping with the target channels of the microfluidic device. The quantitative value of the absorbance of the caspase solution over a well area (S) after subtraction of background reading (XS) is calculated from the reading value from channels based on the below equation (details of the derivation are in Supplementary section V and Fig. S4 and S5): 𝑋S =

(𝑅 ― 𝐴0𝑥0)𝜋𝑟2 𝐴

― 𝑋0

(1)

where x0 is absorbance of background PDMS per unit area, X0 (𝑋0 = 𝑆𝑥0) is absorbance of background PDMS (achieved by measurement), A0 is the area of background PDMS, A is the area of designed channel covered inside the well area; 𝑆 = 𝜋𝑟2, is the total area of one well of the 96-well plate, r is radius of a single well (= 3.2 mm) and R is the reading absorbance value with the channels inside the well (achieved by measurement). With a certain channel height, XS represents the absorbance value of caspases when the whole well area is covered by caspase and there is no effect of background to the reading.

RESULTS AND DISCUSSION 3D-printed Single Microvalve Characterization Despite the presence of reports on the fluidic characterization of NC valves, none of them have presented the characteristics of NC valves made out of 3D printed molds and with the size and resolution offered by 3D printing technology43. In order to systematically design a 3D-printed functional multiplex fluidic circuit, it is crucial to evaluate the performance of a single NC valve under different control pressures (PC). Here, we characterize the 3D-printed

12

ACS Paragon Plus Environment

Page 13 of 39 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

ACS Sensors

valve’s fluidic operation by fabricating and testing them with geometrical variations of SV, WS, and TM (as shown in Table 1).

Geometrical characterization To investigate the flow rate through the valve, the inlet pressure was kept at PI=3 kPa while the control pressure PC was decreased, changing the valve's state from fully closed to the fully opened condition (Fig. 3A(i)). As can be seen, when valve-seat size is small (e.g. when WS=300 µm or 600 µm), the valves are opened earlier at a similar control pressure (PC~2 kPa). Only with a bigger valve-seat size (e.g. WS=900 µm), the valve is closed more tightly and further reduction in control pressure (PC=1.7 kPa) is required to open the valve. When PC is reduced to a certain value (PC ~0.5 kPa) the valve is in the "fully open" state and all valves show equal values for the flow rate. Fig. 3A(ii) demonstrates the effect of valve size on flow rate of a 3D-pinted single valve. Interestingly, the value of the opening threshold pressure (PthO) shows a significant segregation between the valves with different valve sizes. The biggest valve size (4 mm×4 mm) opens first when PC is decreased to ~2.5 kPa (PI-PC= 3-2.5= 0.5 kPa), whereas, the smallest valve (1 mm×1 mm) remains in the closed state until PC is reduced to just about 1 kPa (PI-PC =3-1= 2 kPa). Finally, when the valve is fully open (PC = -1.5kPa), the smaller valve shows lower flow rates due to its higher valve resistance (Rv) (See eqn. (S4) in Supplementary). Furthermore, figure 3A(iii) shows the variation of flow rate with different membrane thicknesses. Valves of different thicknesses are in the fully closed condition at the similar control pressure (PC~3 kPa). Once the PC decreases down to a critical value (~1 kPa), the membrane suddenly deflects. Notably, we observed the sudden jump-up of the flow rate with 13

ACS Paragon Plus Environment

ACS Sensors 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

the thicker membrane due to the corresponding sudden decrease of Rv, expediting the valve's turning into the fully open state.

Figure 3. 3D printed valving unit characterization. (A) Variation of flow rate through the valving unit with control pressure under different valve geometries: (i) Valve-seat sizes (WS) of 300μm, 600μm, and 900μm, when valve size (SV)=2×2(mm2)and membrane thickness TM=30µm; (ii) SV=1×1, 2×2, 4×4(mm2), when WS=600μm and TM=30µm; (iii) Membrane thicknesses of 30μm, 70μm, and 110μm when SV=2×2(mm2) and WS=600μm (n=3 for each experiment). The dashed line shows the control pressure PC=0 relative to atmospheric pressure. (B) Variation of opening fraction of the valve unit with different control pressure: PC are 4 and -3 kPa for closing and opening states. (i) Experimental microscopic picture of valve in open and closed states; (ii) Corresponded intensity value. (C) Variation of PthO with

14

ACS Paragon Plus Environment

Page 14 of 39

Page 15 of 39 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

ACS Sensors

coating and non-coating cell culture media conditions, (i) Setup for PthO test and inlet pressure value, (ii) PthO for different valve sizes. All tests are with WS=600 μm, SV= 2mm×2mm (except C (ii)), and TM=30μm.

Opening fraction Opening fraction which is proportional to the distance between the membrane and the valve-seat determines the range of control pressure and ensures accurate fabrication. We evaluated this factor based on the intensity of Rhodamine B when the solution passing through the valve at different control pressures (Fig. 3B(i) and Movie S1). As shown in Fig. 3B(ii), the fluorescent intensity increases proportionally with control pressure from the minimum value of 18.8 (a.u.) at PC=4 kPa to the maximum value of 37 (a.u.) at PC=-3 kPa corresponding to valve at fully closed and fully open states respectively (as could also be inferred from eqn. (S2)).

Opening threshold pressure (PthO) Lower PthO reduces the valve's pressure input which consequently leads to an easier valve control and saving the operational energy. To assess the valve of this critical parameter, the inlet of a single valve was connected to a syringe pump on the top-layer when the valve was maintained at the closed state (see Fig. 3C(i)). The fluid was pumped continuously into the inlet until the valve gets open. The value of PthO was calculated based on the supplementary section I and eqn. (S1). As shown in Fig. 3C(i), a sudden decrease in inlet pressure was observed in a single valve as it opens. In order to reduce PthO, a strategy was adopted to coat the valve's inner space with culture medium (DMEM, Sigma Aldrich, USA). Measurement of PthO for different valve sizes confirmed that the smallest valve (1 mm×1 mm) has the biggest PthO (see Fig. 3C(ii)).

3D-printed Multiplexed Chip 15

ACS Paragon Plus Environment

ACS Sensors 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

The dependence of the opening threshold pressure on the valve size (Fig. 3A(ii) and Fig. 3C) can be used as a guide for the design of a more complex system which allow sequential control of series of valves. Here, we demonstrate a novel design which enables multiplexed control of different liquid samples by incorporating microvalves via 3D-printed molds (Fig. S2). The device integrates 3 types of valve sizes which have been characterized above: 1×1, 2×2 and 4×4 mm2 with PthO ranges from ~1 to ~2.3 kPa (Fig. 4B). As shown in Fig. 4A, the sequential device consists of 5 main channels (with inlets and outlets shown in green, purple, yellow, pink and red, and named A, B, C, D, E, respectively). These channels are equipped with valves of different sizes integrated at their inlets and outlets which allow the channels to be coated or filled with 5 different solutions or chemicals. The other channel (Channel X in blue color) is specifically designed for the purpose of filling the reaction solution and is able to run through all 5 counterpart channels. In the first stage, the 6 food color wells are kept at a constant height and connected to the inlets of 5 channels and channel X, respectively. Connections to the air pumps are connected to dual pumps 1 and 2. Application of positive and negative pressures to these channels enables controlling the valves opening and closing modes and subsequently leads to filling each channel with the color samples on demand. Figure 4C demonstrates the process by which solutions A, B, C, D, E are sequentially injected into the channels. The pumps were adjusted by considering the PthO of the valves (see Fig. 4B). Initially, the valves connected to channel X were closed by applying PC=4 kPa via pump 1. Next, the air channel controlling the 5 main channels was activated. In the first step, valve 4×4 was opened by applying PC =2.3 kPa via pump 2, which let solution A to flow through and fill channel 1 (see Fig. 4. C(i)). As shown in Fig. 4C(ii), when PC is decreased to 2.1 kPa, channel 2 and 3 were filled with B and C due to the opening of valve 2×2. In the next step, by adjusting PC to 1.6 kPa, valve 1×1 gets open, leading to the release 16

ACS Paragon Plus Environment

Page 16 of 39

Page 17 of 39 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

ACS Sensors

of samples D and E into channels 4 and 5 (see Fig. 4C(iii)). Finally, pressure was applied to close all the valves connected to the 5 sample channels while pump 1 was released to open the remaining valves and letting reaction solution X (shown in blue) running through the device (see Fig. 4C(iv) and (v)). The reaction solution X, therefore, is able to run through all 5 main channel compartments and react with the samples in each of those channels (Movie S2). Using this approach, sequential logic controls (X+A, X+A+B, ..., X+A+B+C+D+E) can be obtained. In another way, if X was inserted from inlets of individual channels A, B, C, D, E after coating time, the parallel logics (X+A, X+B, ..., X+E) can be achieved along with the sequential opening of valves connected to these channels. Similar parallel results can also be achieved if we use additional channels connect to original channel X to the outlets of each individual channel A, B, C, ..., E with suitable pre-defined hydraulic resistances. Having the capability of opening all of the valves on demand due to their different control pressures, we can: (1) Selectively coat arbitrary channels instead of coating all channels at the same time, (2) Apply different incubation time for different channels, (3) Achieve different combination of reaction solution X with solution A, B, C, D using only a single air-control line. The capability of using a single air-control line eliminates the need for several solenoid valves and DAQ controllers in a more complex system incorporating much higher number of valves. One limitation of our device is that Ptho is affected by valve size due to 3D printing resolution and the limited range of valve size’s difference, which consequently affects the packing density of the multiplexed device. Nevertheless, the different Ptho can also be achieved by controlling coating material as presented in Fig. 3C. The printing resolution limitation is shared by most of the existing commercial 3D printers and the development of 17

ACS Paragon Plus Environment

ACS Sensors 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

improved 3D printing materials and technologies44,45 can lead to a more compact multiplexed system.

Figure 4. Proof-of-concept of a 3D-printed sequential valving system for multiplexed assay using food colors. (A) Schematic of device in final coating process. (B) Opening threshold pressure of the valves operated as single valve and in the sequential device. (C) Operation of the sequential valving system: 5 main channels are first filled sequentially with A, B, C, D, E solutions (shown in green, violet and yellow, pink and red, respectively) before the X solution (in blue) run through all channels. The operation is due to the different control pressures introduced by the pumps. Multiplex biosensing of cell apoptosis Apoptosis is a cellular process that causes cell death in a controlled fashion. It plays an essential role during the tissue growth, immune surveillance and fetal development. During 18

ACS Paragon Plus Environment

Page 18 of 39

Page 19 of 39 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

ACS Sensors

apoptosis, intracellular compartments such as nucleus, mitochondria and plasma membrane are broken down by a family of protease enzymes known as caspases. Therefore, the ability to determine caspase activities in cells stimulated with various apoptotic events provides critical information regarding the molecular pathways regulating those events and may lead to development of novel caspase targeting drugs. Members of the caspase family have unique roles including apoptotic signal initiation (caspase 2, 8, 9, and 10), proteolytic operation (caspase 3, 6, and 7) and inflammatory signaling (caspase 1, 4, 5, 11, and 12).46 Despite technological advances, previous caspase activity assays involve sophisticated processes such as design of molecular probes and cell transfection,47 and are designed for detecting only a particular member of the caspase family.48 Depending on the study, the caspase activity assay can be reported as the number of apoptotic cells with a fluorescent tag inside the cells,49 the number of apoptotic cells found through flow cytometry50 or through measurement of extracellular secretions.51 The sequential valving device developed in this study enables simultaneous measurement of the activity of different caspases present in an apoptotic sample through a simple colorimetric assay. Fig. 5A demonstrates the process of surface functionalization of the channels in the microfluidic device. The process starts with oxidation of PDMS surface and generation of hydroxyl groups using pirhana-etch solution followed by silanization of the surface using treatment with APTES solution. The APTES modified surface is then able to react with a mixture of EDC (as a crosslinker) and the caspase substrate DEVD-pNA. Application of caspase sample in the next step cleaves the substrate and releases chromophore pNA. The pNA absorbance can then be quantified using microplate reader.52 In order to evaluate different methods of inducing cell apoptosis, an initial experiment was conducted in a 96 well plate comparing the caspase activities in Jurkat cells treated with 19

ACS Paragon Plus Environment

ACS Sensors 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

different induction methods (Fig. 5B(i)). The same mount 100 µg protein was loaded in all conditions. The result shows the absorbance values of caspase 1, 3 and 6 induced by UV and 𝛾-radiation were markedly higher than those in control. Since 𝛾-radiation yielded low protein concentrations; UV treatment was used for producing apoptotic sample in the subsequent experiments in the multiplexed chip. A simple experiment is conducted in both 96 well plate and the channel with the same amount of reagent to confirm the ability of measuring absorbance of microfluidic device in the microplate reader (see Fig. 5B(ii)). The result shows the comparable absorbance values as measuring absorbance in PDMS device. Fig. 5C(i) presents the process of mounting the microfluidic sequential device in the microplate reader. As can be seen here, three reaction channels overlap with the area of a single well of a 96 well-plate. By absorbance value of each measurement was then presented by subtracting the background signal as well as calculating the share of the reaction channels in each image (see eqn. (1)). Figure 5C(ii) presents the absorbance of released pNA molecules in each channel. Each of the five channels were coated with a unique pNA substrate that can only be cleaved if the corresponding caspase exists in the reaction sample. The more from a particular caspase present in the apoptotic sample, the higher would be the absorbance measured on the channel coated with the corresponding pNA substrate. The apoptotic sample contains higher amounts of caspase 3 and caspase 6 with the highest amount recorded for caspase 6 (n=3, p