Development of a Lab on a Chip Flow Cytometer

Portable and Affordable Flowcytometer for Point of Care Diagnostics

in Rural Areas

A. Mohan

, N. Marshkole

, A. P. Nair

, A. Bharadwaj

, A. Prabhakar

and T. Saiyed

Indian Institute of Technology Madras, Chennai, India

Center for Cellular and Molecular Platform, Bangalore, India

Keywords: Flow Cytometer, Forward Scatter, Side Scatter, Hydrodynamic Focusing, Flow Control, Lensed Fiber.

Abstract: We have developed a lab prototype of a microfluidic flow analyzer, which is capable of quick and efficient

analysis of biological samples. Low cost and portability makes it suitable for point of care diagnostics in

rural area of developing countries. A significant size reduction has been achieved by choosing a

microfluidic flow and pumping system, micro-electronic components, integrated circuits boards, and fiber

optics. A two dimensional microfluidic chip fabricated with nanolithography technique integrates the

fluidics and optics into a single platform. Forward scatters (FSC), side scatter (SSC) and fluorescence (FL)

are measured from polystyrene beads as well as from different live cells. Overall dimension achieved for the

final prototype is 39 cm x 22 cm x 10 cm.

1 INTRODUCTION

Among the existing methods for biological fluid

analysis, flow cytometry is always preferred for its

functionality and capability of detailed analysis

(Shapiro 1995). It is a non destructive way of

knowing the physical and biochemical properties of

biological samples based on optical detection. In

addition to applications in biomedical research field

for immunology (Chattopadhyay et al. 2006; Mohan

et al. 2015), single cell analysis (Chattopadhyay and

Roederer 2012; De Rosa 2012; Telford et al. 2012)

and molecular biology (Nunez 2001; Chattopadhyay

et al. 2006), it is also being availed in a clinical

environment (Glencross et al. 2002). As far as

haematological diseases are concerned, it is very

important to check the status and progression of

disease in a regular basis. Taking the case of rural

areas in developing counties, many of them are

struggling with communicable and non

communicable diseases of severe nature like AIDS,

leukaemia etc. Lack of proper diagnostics and

treatment facilities is one of the major reasons for

unsolved health issues. This has been a motivation to

researchers in the field of biomedical devices for the

past few decades (Martinez et al. 2008; Martinez et

al. 2010). There have been many efforts to develop a

portable and cost effective flow cytometer, which

can be used in primary health care centres in rural

area (Tung et al. 2004; Wang et al. 2004;

Chattopadhyay et al. 2006; Mao et al. 2009; Mao et

al. 2012). Numerous studies have proved that the

transportation and handling of blood sample can

significantly affect test results while on spot tests

can improve reliability. In addition to rural

communities in need of testing for HIV health

monitoring, many urban localities throughout India,

will benefit from access to a low-cost and more

instrument.

Conventional models of a flow cytometer are as

expensive as $10,000 and work only in centralised

facilities of major health care centres (Shapiro

1995). Bulkiness and complexity of the instrument

demands proper maintenance and well trained

expertise to operate the machine. Qualified people

have to go through special training for using the

machine, analyzing the data and make reports. In

addition to that, fixing any functional failure and

troubleshooting requires access to technicians. All

these add a considerable cost to the maintenance.

Our approach for designing a portable and low

cost flow cytometer involved understanding the

major factors that cause the bulkiness and

complexity of the conventional system, and methods

Mohan A., Marshkole N., P. Nair A., Bharadwaj A., Prabhakar A. and Saiyed T.

Development of a Lab on a Chip Flow Cytometer - Portable and Affordable Flowcytometer for Point of Care Diagnostics in Rural Areas.

DOI: 10.5220/0006175301790185

In Proceedings of the 10th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2017), pages 179-185

ISBN: 978-989-758-216-5

179

to reduce them. A conventional flow cytometer

design encompasses three major disciplines of

technology, namely are fluidics, optics and

electronics. Fluidics deals with guiding the sample

inside the machine once it is loaded. Around 2-5 mL

of the sample fed into the inlet of the machine is

sucked in by fluidic pumps through the respective

tubing. This sample is then subject to a 3

dimensional fluid focusing technique, called

hydrodynamic focusing (HDF). A guiding fluid will

flow around the sample keep the flow focused to a

certain diameter. Pumping of the sheath fluid is

controlled such that the sample stream makes a

single file flow of cells or particles down the flow.

This flow then undergoes optical interrogation with

one or more lasers. Collection of various signals

from the sample is achieved with the detectors kept

at different angles with respect to the incident laser.

Most of the incident light gets scattered in the

forward direction, without much interaction with the

internal structures of the cell. This signal collected

between 0

and 20

gives information regarding the

cell size and is called forward scatter (FSC). Internal

complexity information of the cell is given by the

scatter at larger angles 45

-90

, called side scatter

(SSC) (De Rosa et al. 2001). While the FSC and

SSC data together described the physical and

structural properties of cells, we can label the cells

with specific fluorescent tags and identify them by

detecting the emitted fluorescence (FL). Many of

them have multicolour detection for which a bunch

of additional lasers and detectors are used. Although

the functionality is high, bulkiness, complexity and

cost is also high because of the design. The optical

unit working in free space domain consumes

considerable amount of space and it also requires

several components like lenses, prisms etc. to guide

the light in a desired way. Such a system would also

be highly sensitive to dust and other disturbances in

a primary health care (PHC). Use of several detector

units such as photomultiplier tubes (PMT), and

Avalanche photo diodes (APD) with high gain,

again increases both the complexity as well as power

consumption.

In this work, what we describe is a miniature

flow analyser as a lab-on-chip; combining principles

of optics, flow cytometry, microfluidics device

fabrication, and nano electronics to allow rapid cell

analysis and quantification. After minimal

preparation, the sample flows through a microfluidic

device in which a chosen laser detects the presence

of specific biomarkers e.g. CD4 in case of HIV

immune health monitoring. Data is collected and

processed on a small electronics board. The device

gives the user a clear, concrete measure of the level

of these biomarkers in patient samples. This

technology can be adapted to cell culture assays,

detection of water contamination, platelet and other

blood cell counts and oncology tests. The proposed

device will need an investment of only USD 5000,

uses a smaller sample volume, is highly portable and

is cheaper and easier to maintain and upgrade.

2 DESIGN AND FABRICATION

Reduction of size and cost being the major concerns

of the design, we have carefully chosen each

component for fluidics, optics and electronics.

Microfluidic pumps and chip have played a role in

reducing the physical volume considerably.

Microfluidics is generally chosen for the fact that its

channel dimensions are well suited to the

dimensions of cells being analyzed (El-Ali et al.

2006; Young and Beebe 2010; Thompson et al.

2014). The heart of the design is a microfluidic chip

with rectangular flow channels and grooves for

optical fiber insertion. Devices were fabricated with

the commonly used PDMS (Poly dimethyl Siloxane)

(Tung et al. 2004) as well as PMMA (Polymethyl

methacrylate). PDMS based devices were fabricated

using the photolithographic technique on silicon

wafer. Mixture of PDMS and cross linking agent in

10:1 ratio was poured onto the Silicon master and

baked in a hot oven at 70

ºC

for 6 hours. The mould

(Figure 1a) was peeled off from the silicon and

bonded with onto a glass slide after plasma

treatment. Inlets and outlets for tubing were punched

at the respective positions before the bonding.

Sample fluid and the sheath fluid are fed into their

respective inlets through silicon tubing. Once the

flow is started from the respective micro pump,

(Dolomite) the fluids will get into the main fluidic

channel of width 130 µm. Flow rates are controlled

such that the sample stream attains a desired width

depending on the size of the particle being

investigated. The sheath fluid guides the sample

sideways by developing a laminar flow. Required

volume of samples and reagents are tested to be 100

to 500 µl in such a device. The microfluidic chip

also has grooves meant for inserting optical fibers

for light illumination and detection. These groves

are located at a distance of 2 mm from the fluidic

junction (x) in a direction normal to the main flow

channel. Particles in the flow will undergo the

optical interrogation at specific locations in this

manner. Customised optical fibers with tapered/ len

sed tips (Lase Optics Inc.) serve to illuminate and

BIODEVICES 2017 - 10th International Conference on Biomedical Electronics and Devices

180

(a) (b)

Figure 1: (a) Microfluidic device fabricated with PDMS (b) Micro flow analyser setup design describing the functions of

each fiber channels. The fluidic channel and fiber grooves are sown as thick lines.

collect light. The lensed single mode fiber helps to

focus the light to 15-20 µm dia. at a distance of 150

µm, which is the distance from the fiber groove

inner end to the centre of the fluidic channel as

shown in Figure 2. Fiber at the collection end has a

hemispherical lens with a higher numerical aperture,

to increase the collection efficiency. Detection fibers

are chosen to be multimode, since it is required to

maximize the amount of light being collected.

Losses due to presence of an air gap were eliminated

by filling the fiber groove with an index matching

fluid (IMF) of refractive index 1.4. The complexity

with free space optics and components was thus

avoided by availing of lensed fibers and self-guided

fiber grooves.

Figure 2: Top view of the microfluidic chip depicting

optical interrogation of particles in flow channel.

Figure 1b, shows the complete design of the

instrument including three lasers and four detectors.

First two pairs of fiber grooves FR1 and FR2 are

meant for collecting the FSC signal at +50 and -50

respectively. A diode laser at 1550 nm with fiber

coupled output was used as a common source for

FR1 and FR2. Output power of 2 mW is split into

nearly equal powers by using a 50:50 fiber splitter

and fed to the tapered fibers that were fit into the

grooves. A cell passing through the fluidic channel

gets interrogated at point 1 and 2, and the FSC is

collected by separate fiber coupled InGaAs p-i-n by

correlating PIN1 and PIN2 outputs during the flow.

A feedback system connected to the micro pump is

designed to monitor and control the flow rates. Point

3 in figure 1b is the junction where FL and SSC are

being collected. Our choice laser and detector for

fluorescence excitation were according to the

biomarker being used. In this work we have used a

650 nm source at 2 mW, and a single photon

counting module (SPCM - Perkin Elmer) for

detection of fluorescence signal at 5

. SSC is also

detected at the same point in the device using a 1310

nm, 180 mW laser source. Fiber grooves F3a and

F3b are used to make 90

between the source and

detector. We have used an avalanche photodiode

(APD), as the scattered signal is too weak to be

detected with a PIN diode. The design is such that

one can customize the combination of lasers and

detectors to be used for a given experiment. Desired

combinations of F2a, F2b, F3a and F3b can be used

for measuring different parameters of interest.

3 EXPERIMENT AND RESULTS

3.1 Flow Control and Manipulation

Having established hydrodynamic focusing in the

microfluidic chip, and interrogated cells at different

locations down the channel, it was important for us

to know the extent of correlation between the events

different sizes (Sigma Aldrich). Original samples of

1 million/mL concentration were fed to the sample

inlet and DI water was used as the sheath fluid.

Sample and sheath were flowed at the rate of 10

µL/min and 30 µL/min respectively. The lag time of

the particle, as it travels from FR1 to FR2 was

observed for different flow rates. A reduction of lag

time was found as the flow rate as increased as

Development of a Lab on a Chip Flow Cytometer - Portable and Affordable Flowcytometer for Point of Care Diagnostics in Rural Areas

181

Figure 3: (a) Experimentally observed traverse time (time delay) between FR1 and FR2, as a function of flow rate. (b)

Intensity versus time plot showing the correlation of events occurring in FR1and FR2 channel. 5 pair of events are

identified and marked.

Figure 4: (a) SSC versus FSC scatter plot of different size beads. (b) Scatter plot for cells of different sizes. Area of FSC

signal is plotted against the area of SSC signal in both cases.

shown in Figure 3a. Observing the trend between the

two parameters also helped in validating the concept

of flow control using the velocity measurement and

feedback system. FSC events detected at both the

channels are found to be reasonably correlated.

Figure 3b shows an instance of flow with

polystyrene beads. We have FR1 and FR2 locations

separated by a distance of 2 mm, and the

corresponding time delay observed in the data is

around 1ms. Time delay measured from this data

allows one to estimate the flow rate of the sample.

Correlation measure between FR1 and FR2 yield a

reliable metric for use as flow control.

3.2 Size based Isolation of Beads and

Cells

The primary testing of the FSC and SSC detection

was carried out with commercially available

polystyrene beads of 2 µm, 10 µm and 30 µm

standard sizes. Therefore, we validated the detection

of scattering from different sized live cells such as

yeast (3-5 µm), U937 (~10 µm) and BV2 (~15 µm).

Figure 4a shows the isolation of different sized

beads plotting the area of FSC signal with that of

SSC signal. We can see that the beads are

reasonably separated in terms of FSC amplitude. A

similar plot of different types of cells is shown in

figure 4(b). We could see a considerable difference

in the FSC area of the yeast cells from the other two

cell groups which are not properly resolved. The

sample stream is flowing under a two dimensional

(2D) flow focusing condition in which the sample

fluid is getting focused only in the lateral directions.

This lateral focusing also allows the particles to

distribute along the height, 150 µm of the flow

channel. Hence it never achieves a complete single

file flow in a 2D device. Studies have shown to have

improved results when we use 3D flow focusing

methods (Rosenauer et al. 2011; Shivhare et al.

2016) (Mao et al. 2009).

BIODEVICES 2017 - 10th International Conference on Biomedical Electronics and Devices

182

3.3 Cell Labelling and Detection of

Fluorescence

Our setup for fluorescence detection system consists

of a 650 nm laser source which can yield 30 mW of

power. Collection of the signal was made at 135

with respect to the source direction. The choice of

laser wavelength and filter (670 nm long pass) was

optimized for APC - Cy5.5 dye, which has a

maximum excitation at 650 nm and emission peak at

Figure 5: Spectrums of Cy 5 fluorophore and APC - Cy

5.5 used for beads and cells respectively.

680 nm. Figure 5 shows the spectra of Cy5 and APC

– Cy 5.5. Fluorescent beads (Bangs Laboratory)

were customized to have Cy5 dye which has

excitation and emission at 630 nm 670 nm

respectively. Figure 6 is showing the capability of

Figure 6: Histogram of fluorescent beads detected over the

noise. The fluorescent signal intensity is well separated

from the noise histogram.

the system to separates out the intensity of

fluorescent beads from the noise. Having detected

the fluorescence signal from beads, we moved on to

measure the FSC and FL signals simultaneously

from live cells.

3.3.1 Sample Preparation – Staining of

Peripheral Blood Mononuclear Cells

with CD4- AP - Cy 5.5

We prepared a both stained and unstained samples

of PBMC (Himedia laboratory) with a concentration

of 1 million /100 µl. The sample was spun down at

1500 rpm and 4

ºC

for 3 minutes. Blocking buffer

(1% BSA, 0.01 % Azide and 1X PBS) was added to

it after removal of media. One million cells were

counted, washed and kept for 10 minutes of room

temperature incubation on adding 2.5 µg of true

stain. It was taken and washed again in blocking

buffer. 15 µl of CD4 APC Cy 5.5 were then added

into 100 µl of reaction volume. After doing one

more incubation for 20 minutes in ice, the cells were

washed in buffer (1X PBS, 0.01 % Azide). Same

count of cells was taken for unstained sample after

doing the washing. Both the samples were tested in

Guava analyzer and the micro flow analyser

simultaneously. Figure 7 shows the comparison of

counts obtained in Guava and our micro flow

analyser. Out of 10000 events in the Guava, 7500

were giving positive staining results, making the

Figure 7: Comparison of detection efficiency of MFA and

Guava with CY5 stained PBMC cells.

ratio 1.33. We have compared this number to the

ratio of FSC events to FL events detected in micro

flow analyzer. And it came to be as close as 1.4.

Results with PBMC were found to be encouraging to

continue the experiments with human blood, since

the system has proved it’s the ability to pick up the

staining levels of PBMCs. In the following section

we showcase the hardware of lab prototype which

includes the optics, fluidics and electronics parts of

the system.

3.4 Product Packaging

We have been able to encase the entire setup of our

micro flow analyzer in the form of a portable

instrument (Figure 6(d)) along with user friendly

software. Lasers with infrared wavelengths being

Development of a Lab on a Chip Flow Cytometer - Portable and Affordable Flowcytometer for Point of Care Diagnostics in Rural Areas

183

available at a lower cost, we decided to use them for

FSC and SSC detection. The fluorescence

wavelengths being dependent on the sample

(fluorophore) being used, the current version was

designed for APC Cy 5.5, and we used a 650 nm

laser for excitation. To test any other sample, the

excitation wavelength and filters can be changed. A

wavelength tunable laser will also serve this purpose

for testing variety of samples. All the lasers and

detectors were integrated onto a single board as

shown in figure 8(b). Use of optical fibers to guide

light has helped in saving physical space. All the

sources and detectors were spliced with lensed fibers

and were guided right till the point of interrogation.

The specially designed grooves for fibers in the

microfluidic chip kept the fiber in place well aligned

to the microfluidic channel. Microfluidic piezo

pumps (Dolomite) greatly contributed to the

reduction of sample volume as well as the

instrument size. Electrical signals from the detectors

were converted by a 10 bit analog to digital

converter (ADC) and digital data was transferred to

the 16 bit micro-controller. Data communication

with the computer was achieved over a UART

interface. We have demonstrated a compact

packaging method for the instrument which is

enclosed with a black acrylic box as shown in Figure

8(d). The components were arranged in two stacks

supported by acrylic plates and metallic posts. The

bottom plate consists of the electronic board with

lasers and detectors, microcontroller board and

power supply. Top plate has the microfluidic chip,

micro pump and controllers etc. (Figure 8(c)).

Optical fibers spliced to the detectors of bottom

plate are brought to the top plate through given slots.

The whole package has been finally covered with a

compatible lid of same material. Containers for

sample, sheath and wash fluids were kept in the

small outer box attached to the package. One can

load the sample at this point so that it will be taken

in through the tubing connected with the micro

pump. Figure 8 shows different parts of the final

package. Connecting ports for power supply and

USB connectivity are also given at one side of the

box. The software for real time data collection and

analysis has been designed on Python in which user

is allowed to control the lasers and detectors from a

graphical user interface.

Figure 8: (a) Microfluidic system of the Micro flow analyser, comprising the microfluidic chip, piezo pump and tubing. (b)

Bottom plate of the instrument consisting of integrated electronic board of lasers and detectors, microcontroller board, and

power supply unit. (c) Top plate with pump controllers, SPCM and microfluidic system. (d) Complete final look of the

instrument packaged in a black acrylic enclosure of size 39 cm x 22 cm x 10 cm.

BIODEVICES 2017 - 10th International Conference on Biomedical Electronics and Devices

184

4 CONCLUSIONS

We have developed a laboratory scale prototype,

which is in the final stages of optimization. We have

successfully tested this prototype for microfluidic,

optics and electronics integration along with flow

rates for beads and cells testing. We have also

optimized the fluorescence testing with beads and

PBMC cells with the final integration. This novel

design and integration has been patented in US and

South Africa (Saiyed et al. 2016). Currently we are

working on alternative methods for 3D flow

focusing since presently employed methods requires

complex fabrication techniques.

ACKNOWLEDGEMENTS

The authors would like to thank Feroz Musthafa and

Asish Kumar Sen for their assistance in

microfluidics and the Center for Nano science and

Engineering, Indian Institute of Science, for use of

their device fabrication facility. We would also

thank BIRAC (Biotechnology Industry Research

Assistance Council, India) for the funding under

their BIPP scheme.

REFERENCES

Chattopadhyay PK, Price D a, Harper TF, et al (2006)

Quantum dot semiconductor nanocrystals for

immunophenotyping by polychromatic flow

cytometry. Nat Med 12:972–977. doi:

10.1038/nm1371.

Chattopadhyay PK, Roederer M (2012) Cytometry:

Today’s technology and tomorrow’s horizons.

Methods 57:251–258. doi:

10.1016/j.ymeth.2012.02.009.

De Rosa SC (2012) Vaccine applications of flow

cytometry. Methods 57:383–391. doi:

10.1016/j.ymeth.2012.01.001.

De Rosa SC, Herzenberg L a, Herzenberg L a, Roederer

M (2001) 11-color, 13-parameter flow cytometry:

identification of human naive T cells by phenotype,

function, and T-cell receptor diversity. Nat Med

7:245–248. doi: 10.1038/84701.

El-Ali J, Sorger PK, Jensen KF (2006) Cells on chips.

Nature 442:403–411. doi: 10.1038/nature05063.

Glencross D, Scott LE, Jani I V., et al (2002) CD45-

assisted PanLeucogating for accurate, cost-effective

dual-platform CD4+ T-cell enumeration. Clin Cytom

50:69–77. doi: 10.1002/cyto.10068.

Mao X, Lin SCS, Huang TJ (2009) High-throughput on-

chip flow cytometry system using “microfluidic

drifting” based three-dimensional (3D) hydrodynamic

focusing. TRANSDUCERS 2009 - 15th Int Conf

Solid-State Sensors, Actuators Microsystems 425–

428. doi: 10.1109/SENSOR.2009.5285473.

Mao X, Nawaz AA, Lin SCS, et al (2012) An integrated,

multiparametric flow cytometry chip using

“microfluidic drifting” based three-dimensional

hydrodynamic focusing. Biomicrofluidics. doi:

10.1063/1.3701566.

Martinez AW, Phillips ST, Carrilho E, et al (2008) Simple

telemedicine for developing regions: Camera phones

and paper-based microfluidic devices for real-time,

off-site diagnosis. Anal Chem 80:3699–3707. doi:

10.1021/ac800112r.

Martinez AW, Phillips ST, Whitesides GM, Carrilho E

(2010) Diagnostics for the developing world:

microfluidic paper-based analytical devices. Anal

Chem 82:3–10. doi: 10.1021/ac9013989.

Mohan A, Bharadwaj A, Marshkole N, et al (2015) Opto-

fluidic flow analysis for monitoring of immunity

levels. Int Conf Opt Photonics 2015 9654:96540W.

doi: 10.1117/12.2182901.

Nunez R (2001) DNA measurement and cell cycle

analysis by flow cytometry. Curr Issues Mol Biol

3:67–70.

Rosenauer M, Buchegger W, Finoulst I, et al (2011)

Miniaturized flow cytometer with 3D hydrodynamic

particle focusing and integrated optical elements

applying silicon photodiodes. Microfluid Nanofluidics

10:761–771. doi: 10.1007/s10404-010-0707-z.

Saiyed T, Mondal S, Prabhakar A, Krishnamurthy H

(2016) Microfluidic-based flow analyzer.

Shapiro HM (1995) Practical Flow Cytometry. Cytometry

19:376–376. doi: 10.1002/cyto.990190414.

Shivhare PK, Bhadra A, Sajeesh P, et al (2016)

Hydrodynamic focusing and interdistance control of

particle-laden flow for microflow cytometry.

Microfluid Nanofluidics. doi: 10.1007/s10404-016-

1752-z.

Telford WG, Hawley T, Subach F, et al (2012) Flow

cytometry of fluorescent proteins. Methods 57:318–

330. doi: 10.1016/j.ymeth.2012.01.003.

Thompson a M, Paguirigan a L, Kreutz JE, et al (2014)

Microfluidics for single-cell genetic analysis. Lab

Chip 14:3135–42. doi: 10.1039/c4lc00175c.

Tung YC, Zhang M, Lin CT, et al (2004) PDMS-based

opto-fluidic micro flow cytometer with two-color,

multi-angle fluorescence detection capability using

PIN photodiodes. Sensors Actuators, B Chem 98:356–

367. doi: 10.1016/j.snb.2003.10.010.

Wang Z, El-Ali J, Engelund M, et al (2004) Measurements

of scattered light on a microchip flow cytometer with

integrated polymer based optical elements. Lab Chip

4:372–377. doi: 10.1039/b400663a.

Young EWK, Beebe DJ (2010) Fundamentals of

microfluidic cell culture in controlled

microenvironments. Chem Soc Rev 39:1036–1048.

doi: 10.1039/b909900j.

Development of a Lab on a Chip Flow Cytometer - Portable and Affordable Flowcytometer for Point of Care Diagnostics in Rural Areas

185