On-Demand Lensless Single Cell Imaging Activated by Differential

Subscriber access provided by NEW YORK UNIV

Technical Note

On-demand Lensless Single Cell Imaging Activated by Differential Resistive Pulse Sensing Jinhong Guo, Xiwei Huang, and Ye Ai Anal. Chem., Just Accepted Manuscript • Publication Date (Web): 02 Jun 2015 Downloaded from http://pubs.acs.org on June 3, 2015

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 15

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

On-demand Lensless Single Cell Imaging Activated by Differential Resistive Pulse Sensing

Jinhong Guo a, Xiwei Huang b, Ye Ai a*

a

Pillar of Engineering Product Development, Singapore University of Technology and Design, Singapore 487372, Singapore

b

School of Electronics and Information, Hangzhou Dianzi University, Hangzhou 310018, P. R. China

* Corresponding author: Ye Ai. Email: [email protected]; Tel: (+65) 6499 4553

1 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

Page 2 of 15

Abstract We present an on-demand single cell imaging technique activated by differential resistive pulse sensing in a portable system integrating a microfluidic differential coulter counter and a lensless complementary metal-oxide-semiconductor (CMOS) imaging sensor. Dual parametric single cell analysis and on-demand single cell imaging have been demonstrated by microbeads of different sizes and a cell mixture including red blood cells (RBCs) and a tumor cell line HepG2 cells. The on-demand imaging capability could avoid generating useless images without cells and enable selective imaging of single cells within a specific size range.

Keywords Differential resistive pulse sensing; Imaging cytometry; On-chip cell imaging; Single cell analysis


Page 3 of 15

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


1 Introduction Traditional cellular assay based on a large number of cells can only provide population level analyses. Biological cells are intrinsically heterogeneous and population level analysis could be misleading. In recent years, single cell level analysis has emerged as a useful tool to assay heterogeneity within cell populations and identify rare cell populations.1 Coulter counter is the first technique for single cell analysis, which can count cells and characterize the cell size in a continuous flow.2 When a single cell translocates through an aperture with a size comparable to the cell size, the temporal blockage of the cell inside the aperture induces a remarkable change in the electrical resistance that is dependent on the cell size. Combined with microfluidics technology, coulter counter has been implemented as a cost effective and compact platform to analyze various biological cells.3-6 However, this technique can only extract the size information and is very difficult to characterize the cell morphology like cell shape, texture and intracellular structure. In addition, it may not be able to correctly differentiate clumps of smaller cells from larger single cells, which is an important issue in searching rare cell populations. Conventional microscopic imaging technique is able to provide a direct vision of biological cells and thus remains one of the most efficient methods for single cell analysis. However, this imaging technique is typically restricted in well-established modern laboratories due to its high implementation cost and bulky size. The rapidly growing market in on-site healthcare requires portable point-of-care diagnostic tools that can bring imaging-based single cell analysis from modern laboratories to the sites. In recent years, integration of microfluidic channels with lensless complementary metal-oxide-semiconductor (CMOS) imaging sensors has developed onchip cell imaging system in portable devices.7-11 Although the typical size of a single pixel is beyond 1 µm, it is still able to reconstruct images with sub-micron resolution using pixel super-



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 15

resolution algorithms.12-13 Current lensless imaging systems are mainly applied for static cell counting and differentiation.9, 14-15 In a flow-through mode, continuous imaging will generate a huge amount of images with useless information, for example images without biological cells. The huge amount of images will slow down and overload the image processing, which is the major challenge for high-throughput imaging-based single cell analysis. In this work, we integrate the lensless single cell imaging with differential resistive pulse sensing in a portable microfluidic system to address the aforementioned challenges of both techniques for accurate and high-throughput single cell analysis. In such an integrated system, the resistive pulse sensing and imaging-based cytometry can be concurrently performed for single cell analysis. Additionally, the resistive pulse sensing can activate the lensless CMOS sensor, which enables on-demand single cell imaging based on the cell size information.

2 Device and experimental setup 2.1 Device design Fig. 1 shows the schematic structure of the developed on-demand single cell imaging system including a microfluidic device and a lensless CMOS imaging sensor. The microfluidic device has two identical channels sharing the same inlet and outlet reservoirs. The two channels have the same aperture for resistive pulse sensing. The testing sample solution is directly dropped into one of the reservoirs and is then introduced to the microfluidic channel by gravity. Two AgCl electrodes are immersed in the two reservoirs to apply a DC bias across the two channels, which in turn generate electroosmotic flow (EOF) to drive biological cells through the sensing apertures assisted by gravity. At the end of each aperture, a branch microfluidic channel is used to detect the voltage modulation induced by single cell translocation through the aperture. The two branch microfluidic channels are connected to a differential amplifier. When there is no cell


Page 5 of 15

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


translocation through either of the two sensing apertures, the voltages at the two input ports of the differential amplifier are nearly identical, and therefore the measured output voltage, Vout, is zero. Once a single cell is translocating through either of the two sensing aperture, the voltage at the corresponding input port changes accordingly and the other port is used as a reference, the voltage difference between the two input ports is then amplified by the differential gain with a low-pass filter, based on which cells could be counted and the size information could be extracted. In an extremely rare scenario that two exactly identical cells translocate through both apertures at exactly the same time, the output of the differential amplifier would be zero and results in misreading. To reduce the probability of simultaneous cell translocation through both apertures, the sample concentration should be in a reasonable range and the aperture length should be reasonably short. Different from the conventional coulter counter with a single aperture for resistive pulse sensing, the dual-channel configuration can enhance the throughput of single cell analysis. Furthermore, the differential resistive pulse sensing can achieve better signal-to-noise ratio and better signal stability in a long-term monitoring.16 The amplified voltage pulse modulation is collected and digitalized by an ADC chip, and then transferred to a central controller operated by a Field Programmable Gate Array (FPGA, Altera, Cyclone II, USA). A threshold voltage modulation is set in the FPGA controller to detect the translocation event of cells above a certain size through the aperture. Once a voltage pulse modulation higher than the preset threshold value is detected, the FPGA produces a signal to activate the CMOS sensor working in the snapshot mode to capture the cell image. The CMOS sensor (Aptina MT9M032, San Jose, CA, USA) has an imaging area of 3.24 mm (H) × 2.41 mm (V) with a pixel size of 2.2 µm. The frame rate is 60 fps at a resolution of 1280 (H) × 720 (V), which can be further increased by reducing the field-of-view (FOV). In this study, the FOV only



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 15

covers the exit of the two sensing apertures to capture images of single cells when leaving the apertures. To optimize the image contrast, the protective layer and microlens layer on the top of the CMOS imaging sensor are carefully removed and microfluidic channel is directly attached on the CMOS imaging sensor to minimize the distance between single cells and imaging pixels. 2.2 Device fabrication The microfluidic device was fabricated using a soft lithography technique reported previously.17-18 Briefly, the master mold was fabricated by patterning a negative photoresist SU-8 (SU-8 25, Microchem, USA) on a clean glass slide. A polydimethylsiloxane (PDMS) mixture of pre-polymer and curing agent at a weight ratio of 10:1 was casted on the SU-8 master mold after a complete degassing and then cured by polymerization. Cured PDMS was then peered off from the mold and punched in the reservoir locations. The fabricated aperture is 30 µm in length, 30 µm in width, and 30 µm in height. Each microfluidic channel is 2 mm in length, 200 µm in width, and 30 µm in height. The diameter of the reservoirs is 1 mm. The PDMS layer and the bare CMOS imaging sensor were treated by plasma before brought into contact for bonding to form the final chip device. 2.3 Testing sample Two monodisperse particle suspensions of 5 µm and 15.6 µm in diameter polystyrene microbeads were prepared by diluting the original sample in 1X PBS solution to obtain a reasonable concentration (1×104 mL-1). Under this sample concentration and the fabricated aperture length, simultaneous particle translocation through both apertures was hardly observed in the experiments. Bovine serum albumin (BSA) was added in the microbead suspensions to avoid agglomeration and adhesion onto the channel wall. A tumor cell line, HepG2 cells (American Type Culture Collection, MD, USA), were cultured in Dulbecco’s Modified Eagle


Page 7 of 15

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


Medium (DMEM) supplemented with 10% fetal bovine serum (FBS), penicillin (100 µg/mL), 1mM sodium pyruvate, 0.1 mM MEM nonessential amino acids, at 37 °C under a 5% CO2 in a T75 flask. Red blood cells (RBCs) were separated from the whole blood using optiprepRM density gradient medium. After centrifugation at 2000×g for 15 min, RBCs were collected from the bottom of the centrifuge tube. The two different cell samples were washed three times with 1X PBS, and then mixed and suspended in fresh 1X PBS solution.

3 Results and discussion We first used the microbead suspensions to calibrate the developed system integrating resistive pulse sensing and imaging-based cytometry. Fig. 2a shows the measured differential voltage modulation induced by five consecutive translocation events of 5 µm microbeads through the sensing apertures. When there is no particle translocation through either of the two sensing apertures, the measured voltage at the output port of the differential amplifier is nearly zero with a noise of ± 0.01 mV. The upward and downward voltage pulses represent the particle translocation through different sensing apertures. The magnitude of the voltage pulse is from 0.04 to 0.07 mV, which has been successfully applied to trigger the CMOS imaging sensor to take snapshots of single particles. Fig. 2b shows that 15.6 µm microbeads can induce a much larger voltage modulation (a magnitude of ~ 1 mV), which demonstrates that the voltage modulation is highly dependent on the particle size through the sensing aperture. Resistive pulse activated imaging was also demonstrated using the 15.6 µm microbeads. The captured images of 5 µm and 15.6 µm microbeads cover 3 and 7 pixels, respectively. Since each pixel is 2.2 µm, the measured size of 5 µm and 15.6 µm microbeads are 6.6±2.2 µm and 15.4±2.2 µm, respectively. When cells are flowing in a very close distance above the sensing pixel array, the image capturing process is like to digitize the original analog cell images. As long as the size variation



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 15

is larger than half the pixel size, it can be detected. Therefore, the smallest cell size that the system can detect is half the pixel size, which is 1.1 µm in this prototype. Next, we used the cell mixture to demonstrate the capability of the developed system in differentiating different cell types. Fig. 3 shows the measured voltage modulations induced by a series of cell events through the sensing apertures. RBCs and HepG2 cells have an average size of 5 µm and 18 µm, respectively. Therefore, the two different cell types induce quite distinct voltage modulations, based on which 7 RBCs and 4 HepG2 cells were counted and differentiated as shown in Fig. 3. In this experiment, the threshold voltage modulation is set to 0.04 mV for ondemand single cell imaging. Pixel analysis of the captured single cell images indicates that the measured size of RBCs and HepG2 cells are 6.6±2.2 µm (pixel number = 3) and 17.6±2.2 µm (pixel number = 8), respectively. The concurrent resistive pulse sensing and imaging-based cytometry can both differentiate the two different cell types. When the threshold voltage modulation is set above the maximum pulse induced by RBCs, the on-demand imaging system only captures the images of single HepG2 cells, demonstrating the capability of selective single cell imaging based on the cell size.

4 Conclusion In summary, differential pulse resistive sensing and imaging-based cytometry have been integrated together in a portable system for single cell analysis. The differential pulse resistive sensing can actively control the lensless CMOS sensor to implement on-demand single cell imaging. Microbeads of different sizes, as well as a cell mixture including RBCs and HepG2, have been used to validate the on-demand single cell imaging. A significant advantage of this technique is that it can avoid generating useless images without cells and remarkably increase the efficiency of image processing. Furthermore, this technique could enable the ability to


Page 9 of 15

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


selectively capture images of single cells within a specific size range, which could be a potential solution to search rare cell populations from highly heterogeneous biological samples.

Acknowledgments This work was supported by SUTD-MIT International Design Center IDG11300101 (YA).



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 10 of 15

References 1.

Schubert, C., Nature 2011, 480. 133-137.

2.

Sun, T., Morgan, H., Microfluid. Nanofluid. 2010, 8. 423-443.

3.

Sun, J., Kang, Y., Boczko, E. M., Jiang, X., Electroanalysis 2013, 25. 1023-1028.

4.

Guo, J., Lei, W., Ma, X., Xue, P., Chen, Y., Kang, Y., Biomedical Circuits and Systems,

IEEE Transactions on 2014, 8. 35-41. 5.

Guo, J., Chen, L., Ai, Y., Cheng, Y., Li, C. M., Kang, Y., Wang, Z., Electrophoresis

2015, 36. 737-743. 6.

Wang, Y.-N., Kang, Y., Xu, D., Chon, C. H., Barnett, L., Kalams, S. A., Li, D., Li, D.,

Lab Chip 2008, 8. 309-315. 7.

Cui, X., Lee, L. M., Heng, X., Zhong, W., Sternberg, P. W., Psaltis, D., Yang, C., Proc.

Natl. Acad. Sci. U. S. A. 2008, 105. 10670-10675. 8.

Ozcan, A., Demirci, U., Lab Chip 2008, 8. 98-106.

9.

Moon, S., Keles, H. O., Ozcan, A., Khademhosseini, A., Hæggstrom, E., Kuritzkes, D.,

Demirci, U., Biosens. Bioelectron. 2009, 24. 3208-3214. 10.

Li, W., Knoll, T., Sossalla, A., Bueth, H., Thielecke, H., Proc. SPIE 7894, Optical

Fibers, Sensors, and Devices for Biomedical Diagnostics and Treatment XI 2011, 78940Q. 1-12. 11.

Shanmugam, A., Salthouse, C., J Biomed. Opt. 2014, 19. 016002.

12.

Zheng, G., Lee, S. A., Yang, S., Yang, C., Lab Chip 2010, 10. 3125-3129.

13.

Huang, X., Guo, J., Wang, X., Yan, M., Kang, Y., Yu, H., PLoS ONE 2014, 9. e104539.

14.

Su, T.-W., Seo, S., Erlinger, A., Ozcan, A., Biotechnol. Bioeng. 2009, 102. 856-868.

15.

Smith, Z. J., Gao, T., Chu, K., Lane, S. M., Matthews, D. L., Dwyre, D. M., Hood, J.,

Tatsukawa, K., Heifetz, L., Wachsmann-Hogiu, S., Lab Chip 2014, 14. 3029-3036.


Page 11 of 15

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


16.

Wu, X., Kang, Y., Wang, Y.-N., Xu, D., Li, D., Li, D., Electrophoresis 2008, 29. 2754-

2759. 17.

Guo, J., Huang, X., Shi, D., Yu, H., Ai, Y., Li, C. M., Kang, Y., RSC Adv. 2014, 4.

56342-56345. 18.

Guo, J., Chen, L., Huang, X., Li, C. M., Ai, Y., Kang, Y., Electrophoresis 2015, 36. 420-

423.



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 15

White Light

AgCl Electrode

Channel

CMOS Sensor Printed Circuit Board

Cells

(a)

(b)

CMOS Imaging Sensor Microfluidic Channel

AgCl Electrode

AgCl

Differential Amplifier

AgCl

ADC

AgCl Electrode Signal to Activate Imaging Sensor

FPGA Controller (Detection of Pulse and Control of Imaging Sensor) Captured Images and Detected Pulses Storage

(c)

Fig. 1 (a) Schematic structure of the developed on-demand single cell imaging system. The microfluidic differential coulter counter is built on the CMOS imaging sensor with white light illumination. (b) A three-dimensional illustration of the microfluidic device on the CMOS imaging sensor. (c) Working principle of the on-demand single cell imaging system. Two electrodes immersed in the inlet and outlet reservoirs are used to apply an electrical bias across the two sensing apertures. Two branch channels at the end of the sensing apertures are connected to a differential amplifier for the detection of voltage modulation. FPGA is the central microcontroller for signal processing that can actively control the CMOS sensor to take snapshots of single cells based on the voltage modulation. 12 ACS Paragon Plus Environment

Page 13 of 15

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


(a)

(b)

Fig. 2 Measured differential resistive pulses induced by several consecutive 5 µm (a) and 15.6 µm (b) polystyrene microbeads through the sensing apertures. Each upward or downward pulse represents one single particle translocation event. Five and two translocation events are detected in (a) and (b), respectively. Each pulse triggers the CMOS imaging sensor to capture an image of a single microbead.



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 14 of 15

Fig. 3 Measured differential resistive pulses induced by several consecutive cell events, indicating the detection of 11 biological cells (7 RBCs and 4 HepG2 cells) in total. Each upward or downward resistive pulse activates the on-demand imaging of single cells.


Page 15 of 15

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


Graphical abstract


On-Demand Lensless Single Cell Imaging Activated by Differential

Recommend Documents