CD4 CELL ISOLATION FCHIP MAGNETOPHO

Macdara Glynn, DBiomedical Diagnostics Institute,

ABSTRACT

With timely diagnosis and correct trliving with HIV/AIDS can consider the diserather than a terminal illness. Still, in regioendemic, rapid diagnosis is a challencomplexity of the instrumentation requinfrastructure in these countries, as well expertise required to carry out the diagnpresents a microfluidic chip based approachquantitative CD4+ cell counting on a cheaportable and instrumentation-free Poindiagnostic device. Flow is driven by fflexible reservoir, and the target cells athrough magnetophoresis. The fluidic test cca. 30 seconds of sample application to the c INTRODUCTION

HIV (human immunodeficiency viruthe subsequent development of the assocknow as AIDS (acquired immunodeficiremains an ongoing disease of pandemic pits identification in the early 1980s. Yetmodern molecular and cellular instrumdiagnostics of the infection, as well as availactive] anti-retroviral therapy ([HA]ART) hinfection to low penetrance in countritechnologies and interventions are reHowever, epidemiological statistics show tincome countries consistently maintain the of HIV (Fig. 1), with Sub-Saharan Africaheavily affected (69% of the 34 million 2011) [1]. Access to rapid and reliable diregions is low due to limited infrastructechnical know-how. Alternatives to the standard” are currently investigated todiagnostics to be deployed in such regions. Care” (PoC) devices aim to be capable remote villages with minimal power and can then be immediately delivered to a patipositive for HIV in these areas.

The target of the HIV virus is the T-hpatient. Progressive loss of these cellssuppression of the immune system and tAIDS. T-helper cells have a specific proteicell surface called CD4, and hence CD4 ckey clinical determinant for diagnosis of Hthe initiation of treatment. While healthy ina CD4 cell count of > 1200 cells μl-1 icommon practice is that HIV positive patitreatment when this number falls belothreshold – currently set by the World Hea

FROM BLOOD USING FINGER-ACTORESIS FOR RAPID HIV/AIDS DIAG

avid J. Kinahan, Danielle Chung, and Jens Duc, School of Physical Sciences, Dublin City Univ

reatment, people ease as a chronic ons were HIV is

nge due to the uired, the poor as the technical

nosis. This paper h allowing semi-ap, rapid, highly nt-of-Care HIV finger-pressing a are immobilized completes within chip.

us) infection and ciated symptoms ency syndrome) proportions since t, application of

mentation to the lability of [highly have brought the es where these

eadily available. that low- to mid-largest viral pool

a being the most global cases in

iagnosis in these cture, costs and expensive “gold allow reliable These “Point-of-of operating at personnel. ART

ient presenting as

helper cells of the s results in the the pathology of in marker on the cell counts are a HIV patients and ndividuals present in whole blood, ients begin ART ow a specified alth Organization

(WHO) at 500 cells μl-1 [2]. The patients with CD4 cell counts atherefore critical. A number of dethat performs at the required levand/or per-test cost can be large, aper-day can also limit the use of the

Figure 1: Global snapshot of the The diameters of the circles inumbers, while the color represennational population living with HIVcontains 50% of all people in the wo

We present here a finger-pressCD4 cell isolation on an autonomoisolating CD4 cells from whole blosample loading. The very simple, and instrument-free microfluidic proresults and thus meets the criteresource-limited regions where Hendemic, and monitoring remains ainvolves the incubation of whole magnetic beads that specifically binT-helper cells. This sample is thedically primed microfluidic chip, antimes past a magnetic captureisolating the CD4 cells to a locationof the blood sample. Critically, theboth the loading of the sample, anare generated by deflecting a fluidithe main chamber via a flexible achieved simply by depressing tfinger of the operator. Following isthey can be semi-quantitatively inspection of the capture chamber.

TUATED ON-GNOSTICS crée ersity, IRELAND

ability to identify HIV around this threshold is evices are on the market vel, but the instruments and the number of tests-e devices [3].

HIV pandemic in 2011. ndicate the population nts the percentage of a

V. The area circled in red orld living with HIV.

s driven magnetophoretic ous microfluidic chip for ood within 30 seconds of

minimum operator-skill ocedure delivers accurate eria for deployment in HIV / AIDS is largely a challenge. The strategy blood with (super)para-

nd to the CD4 epitope of en introduced to a flui-nd passaged a number of chamber, specifically n removed from the bulk e forces needed to drive

nd the fluidic movements ic reservoir connected to membrane. This can be

the membrane with the solation of the CD4 cells,

enumerated by visual

978-1-4799-3509-3/14/$31.00 ©2014 IEEE 256 MEMS 2014, San Francisco, CA, USA, January 26 - 30, 2014

SYSTEM DESIGN The architecture of the chip display

fluidic reservoir (P1) at the base of the chdirectly to the separation channel, a secondais at the top of the separation channel and port is directly below P2. Midway downchannel is the CD4 capture locus. On the the capture locus is a vent to allow any ovduring the cell separation process (Fig. 2).

Figure 2: Finger-press actuated magneisolation chip. i) 3D rendering of chip. ii) with the microfluidic features shown usingSchematic showing features of interest. Dwhen P1 is depressed or released is indicatGreen arrows respectively.

Microchannels (40 μm deep) are fabr[4] and adhered to a glass base. A permplaced perpendicular to the pneumaticallyliquid (Fig. 2 iii), and the chamber is fillbuffer.

For biosafety reasons, we cannot hanblood in our lab. Therefore finger-prick derifrom healthy donors was first depleted of naIt is then spiked with defined concentrationsstained CD4-positive HL60 cells at meconcentrations between 10 and 1000incubated for 3 min with paramagnetic beadCD4 epitope.

SYSTEM OPERATION

To load the chip, the pneumatic chamdepressed to reduce the internal volume of sample is applied to the loading inlet and(Fig. 3 i-ii). The subsequent reconstitutiovolume generates a flow from the samplcapture region, and into the P1 reservoir (Factuation of the second reservoir (P2) c

s the pneumatic hip. P1 connects ary reservoir (P2) the sample input n the separation opposite side to

verflow to occur

etophoretic CD4

Image of a chip g green dye. iii)

Direction of flow ted with Blue and

ricated in PDMS manent magnet is y-driven flow of led with priming

ndle HIV patient ived whole blood ative CD4+ cells. s of fluorescently edically relevant 0 cells μl-1, and ds specific to the

mber (P1) is first f the chip. 4 μl of d P1 is released on of its initial le inlet, past the Fig. 3 iii-iv). The compensates any

shortfall in liquid volume, insuring remains fully primed. The vastmagnetically tagged cells follows tP1; only the magnetically tagged,deflected to the base of the flow(Fig. 4). In order to sequester remcapture chamber, P1 is again presback towards P2 (Fig. 3 v-vi). By rcommences. By repeating these steppass actuation will complete in apQuantification of recovered cellfluorescence and bright-field microchip is amenable to detection usingUV-fluorescence detector.

Figure 3: Operation of the magnetCD4 cells (shown as black circDirection of flow pressure is shownand Green (towards P1) arrows. METHODOLOGY Chip manufacture

The microfluidic chips used infrom polydimethylsiloxane (PDMmixed at a ratio of 10:1 base procedures for making a master arotating magnet in the PDMS haveelsewhere [4-5]. Loading holes anthe PDMS at appropriate locations

that the chamber always t background of non-the laminar flow towards , CD4-positive cells are w-free capture structure

maining CD4 cells to the ssed to drive the sample releasing P1 the cycle re-ps a second time, a dual-

pproximately 30 seconds. ls was carried out by scopy (Figs. 4-5), but the g a handheld LED-based

tophoresis chip to isolate les) from whole blood. n with Blue (towards P2)

n this paper were formed MS; Dow Corning, MI)

and curing agent. The and for securing the co- been described in detail

nd vents were defined in s using a dot punch. The

257

PDMS slab containing the microfluidic feaon a 25 × 60 mm glass coverslip and allo1 min. Finally, the glass / PDMS chip waPMMA base. To prime the microchannelthe chip was placed under vacuum for following which a large drop of priming bubuffered saline [PBS] pH 7.4, 0.1% w/valbumin [BSA], 1 mM EDTA) was immedthe surface of the PDMS, covering both ththe loading chamber, and the vent. Degas primed the channels. Magnets used wecylindrical magnets, with a diameter of 3 mof 6 mm (Supermagnete, Germany). Magnethe molded cavities before the samples wechip. Blood Processing and Cell Culture

Blood was extracted directly from hefinger prick using 1.5 mm sterile lancets (NJ, USA). To prevent coagulation of th60 mM EDTA solution was immediatelsample to result in a final concentration of the whole blood. Blood was isolated anddirectly before experimental use. Native depleted using the T4 Quant Kit (Life TeUSA) according to manufacturer’s instructio

HL60 cells (DSMZ, Braunschweig, cultured in 75 cm2 flasks in RPMI 1640 mun-inactivated fetal bovine serum, 100 Uand 100 μg mL-1 streptomycin. Cultures we37 °C with 5% CO2. Where indicatfluorescently stained with NucBlue Live Technologies) according to the manufacture

Experimental samples composed of blHL60 were treated with Dynabeads CD4 (Life Technologies). Incubations were carriEppendorf tube, and final incubation voluper sample, composed of 140 μL incubation7.4, 0.1% w/v BSA, 1 mM EDTA), 50 μsample, and 10 μL Dynabead CD4. performed at room temperature with rotatioof sample was introduced to the chip as desc RESULTS

Following a dual-pass actuation of spiked with 4 concentrations of HLconcentrations as low as 10 cells μl-1 in wdetected. Furthermore, in the medically relrange of 10-1000 CD4 cells μl-1 of whole bshowed a linear fluorescent signal output cthe concentration of cells. Over this range, Pearson correlation coefficient (RSQ) wa0.97 (Fig. 5-i). This suggests that the CD4 patient blood over diagnostic ranges lie witrange of the chip performance – indicating can lead to high predictability of CD4 coudata-point.

atures was placed wed to bond for as mounted to a s and structures, at least 1 hour,

uffer (phosphate-v bovine serum diately placed on he sample port of driven flow then

ere NdFeB N45 mm and a height ets were placed in ere loaded to the

ealthy donors via (BD Biosciences, e blood sample, y added to the

f 6 mM EDTA in d prepared fresh,

CD4 cells were echnologies, CA, ons. Germany) were

media, with 10% U mL-1 penicillin, ere maintained at ed, cells were Cell Stain (Life

er’s instructions. lood spiked with

magnetic beads ied out in a 2 mL ume was 200 μL n buffer (PBS pH μL experimental Incubation was

on for 3 min. 4 µl cribed.

f blood samples L60 cells, CD4 whole blood were

levant diagnostic blood, the system corresponding to the square of the as calculated as concentration in

thin in the linear that the strategy

unt using a single

Furthermore, when the area of beads / cells is measured using tsoftware ImageJ (NIH, USA), a simconcentration (0.99) is observed (that imaging and quantification caonly bright-field optics.

Figure 4: Results of finger-press puisolation of HL60 cells from CD4Merged DIC / fluorescent imagesimages (right). The edge of the capta yellow dashed line.

f the space occupied with the free image analysis

milar correlation to HL60 (Fig. 5-ii). This suggests an be carried out using

umped magneto-phoretic 4-depleted whole blood. s (left), raw fluorescent ture chamber is shown as

258

Figure 5: Fluorescent analysis of images frintegrated density values (IDV) from iregion-of-interests (ROIs) from the imagesand shown relative to the number of HL6the blood. ii) ROIs were generated on the bshown in Fig. 4. The area of these were number of HL60 cells spiked to the bcalculated on i) and ii) to show linearity of fluorescence was measured at the captuchips containing native- and CD4-deplettreated with DAPI. In spite of the promising results using arethe quantification of the data, a future anamay be more accurate using fluorescencdark-field background. In this case, to keepat a minimum, specific staining using expbased fluorescence markers should be avoidthe chip will exclusively isolate CD4 cellcapture, when run under non-spiked nati

rom Fig. 4. i) The identically sized s were calculated 60 cells spiked to

rightfield images compared to the

blood. RSQ was f the data. iii) UV ure chambers of ted whole blood

ea calculation for alysis instrument

ce imaging on a p the cost-per-test pensive antibody ded. However, as ls using positive ive conditions a

general UV based white blood staallow CD4 specific fluorescent readexpensive antibody binding reagenincubated native whole blood (bothCD4 depleted) with DAPI (1 µg mlCD4 para-magnetic beads, and exachip isolations under UV. A clear in UV signal measured in the canative and CD4-depleted blood (Fig DISCUSSION AND CONCL

To enable rapid, cheap and based on CD4 cell count in a fingeblood, we have developed an autoncapable of delivering a semi-quantinfection. Critically, this chip runs for flow control; and if deployed, wonly a bright-field light source wicamera similar to that on mid-rareadout. Compared to the complexon the market for PoC HIV dpresented here would be notably without a fully quantitative resultcases, a medical practitioner on sitepatient falls into a “treat” or “nodistribution of ART pharmaceutisimplicity of the presented devicnature of the data generated would b ACKNOWLEDGEMENTS

This work was supported by grant No. CF 2011 1317, the EFoundation Ireland under grant No

REFERENCES [1] UNAIDS, Global report: UNA

AIDS epidemic 2012, UNAIGeneva, 2012.

[2] WHO, Consolidated guidelantiretroviral drugs for treatininfection, ISBN: 978 92 4 1505

[3] M. Glynn, D. J. Kinahan and (2013), pp. 2731-2748.

[4] Kirby, D.; Siegrist, J.; KijankNanofluid. 2012, 16, 1–10.

[5] M. Glynn, D. Kirby, D. ChuKijanka and J. Ducrée, JALA10.1177/2211068213504759.

CONTACT M. Glynn, tel: +353 1 7007899, e: mJ. Ducrée, tel: +353 1 7007870, e: je

ain will be sufficient to dout without the need for nts. To this end, we co-h CD4-depleted and non-l-1) concurrently with the amined the resulting on-difference was observed

apture chamber between g. 5-iii).

LUSION simple HIV diagnostics

er-prick volume of whole nomous microfluidic chip titative diagnosis of HIV

without instrumentation would potentially require ith an associated optical ange mobile phones for xity of devices currently

diagnostics, the strategy more affordable, albeit

t. Nevertheless, in most e needs only to know if a o-treat” category for the icals. Hence, given the ce, the semi-quantitative be an acceptable caveat.

Enterprise Ireland under ERDF and the Science 10/CE/B1821.

AIDS report on the global IDS / JC2417E, WHO,

ines on the use of ng and preventing HIV 572 7, Geneva, 2013 J. Ducrée, Lab Chip, 13

ka, G. et al., Microfluid.

ung, D. J. Kinahan, G. , 2013 – in press DOI:

macdara.glynn@dcu.ie ens.ducree@dcu.ie

259