cancer research surface-immobilized aptamers for...

Tum

Surand

YuanAndre

Abst

Intro

Earticallybeen scells, iical fointerasortinical anbasedefficie

AuthorResearUnivers5AnnetNeurolTexas;Commiand UnUnivers

Note:Resear

Current

CurrentGwang

Corres500 SoPhone:

doi: 10

©2010

www.a

Published OnlineFirst November 9, 2010; DOI: 10.1158/0008-5472.CAN-10-0568 Published OnlineFirst November 9, 2010; DOI: 10.1158/0008-5472.CAN-10-0568 Published OnlineFirst November 9, 2010; DOI: 10.1158/0008-5472.CAN-10-0568

Canceresearch

or and Stem Cell Biology

face-Immobilized Aptamers for Cancer Cell Isolation

R

Microscopic Cytology

Wan1,2, Young-tae Kim1,2, Na Li3, Steve K. Cho4,5, Robert Bachoo4,5,6,
w D. Ellington3, and Samir M. Iqbal2,7,8
ractExp

tissuewe repcommdiffusegicallyplatfoincludenginein Ink4both t

on affinincy and

s' Affiliatioch and Teacity of Texate G. Strauogy, Univer7Departmettee of Bioeiversity oity of Texas

Supplemench Online (h

address fo

address foju Institute o

ponding Auuth Cooper817-272-02

.1158/0008-

American A

acrjourna

DownloDownloDownlo

osing rare but highly malignant tumor cells that migrate from the primary tumor mass into adjacent(s) or circulate in the bloodstream is critical for early detection and effective intervention(s). Here,ort on an aptamer-based strategy directed against epidermal growth factor receptor (EGFR), the moston oncogene in glioblastoma (GBM), to detect these deadly tumor cells. GBMs are characterized byinfiltration into normal brain regions, and the inability to detect GBM cells renders the disease sur-incurable with a median survival of just 14.2 months. To test the sensitivity and specificity of our

rm, anti-EGFR RNA aptamers were immobilized on chemically modified glass surfaces. Cells testeded primary human GBM cells expressing high levels of the wild-type EGFR, as well as geneticallyered murine glioma cells overexpressing the most common EGFR mutant (EGFRvIII lacking exons 2–7)a/Arf-deficient astrocytes. We found that surfaces functionalized with anti-EGFR aptamers could capturehe human and murine GBM cells with high sensitivity and specificity. Our findings show how noveler substrates could be used to determine whether surgical resection margins are free of tumor cells,
aptam
or more widely for detecting tumor cells circulating in peripheral blood to improve early detection and/ormonitoring residual disease after treatment. Cancer Res; 70(22); 9371–80. ©2010 AACR.

proaclevelsare colink anheteroThe

greathave bcompaof bein

duction

ly detection of cancer and its metastasis can drama-change treatment and improve prognosis. There haveeveral approaches reported for the detection of tumorncluding microfabricated devices that rely on mechan-rces (1, 2), dielectrophoresis (3), microscale opticalctions (4), immunohistochemistry (5, 6), magnetic cellg (7, 8), and flow cytometry (9). In contrast to mechan-d electrical sorting techniques, detection and sorting

ty interactions is expected to yield highergreater specificity (10). Affinity-based ap-

and caicallysite-spare mvide smersas in aly receto sorresultsgrowthumaengineThe ato isoquentall hugrowtductiosion, c

ns: 1Department of Bioengineering, 2Nanotechnologyhing Facility, 3Institute for Cell and Molecular Biology,s at Arlington, Arlington, Texas; 4Internal Medicine,ss Center for Neuro-Oncology, and 6Department ofsity of Texas Southwestern Medical Center, Dallas,nt of Electrical Engineering and 8Joint Graduatengineering Program, University of Texas at Arlingtonf Texas Southwestern Medical Center at Dallas,at Arlington, Arlington, Texas

tary data for this article are available at Cancerttp://cancerres.aacrjournals.org/).

r N. Li: AM Biotechnologies, LLC, Houston, TX 77034.

r S.K. Cho: School of General Studies, GIST College,f Science and Technology, Gwangju, Korea.

thor: Samir M. Iqbal, University of Texas at Arlington,Street, M.S. 19072, Room 217, Arlington, TX 76019.28; Fax: 817-272-7458; E-mail: [email protected].

5472.CAN-10-0568

ssociation for Cancer Research.

ls.org

Research. on May 17, 2018cancerres.aacrjournals.org aded from Research.

on May 17, 2018cancerres.aacrjournals.org aded from Research.

on May 17, 2018cancerres.aacrjournals.org aded from

hes that rely on antibodies are often subject to highof off-target cross-reactivity (5, 11, 12). Moreover, therensiderable technical challenges to reproducibly cross-tibodies to the surfaces of miniaturized devices due togeneity of conjugation and surface denaturation.re is increasing recognition that aptamers may haveutility in cancer diagnosis and therapeutics. Aptamerseen shown to have affinities and specificities that arerable with those of antibodies, but have the advantageg highly stable at a variety of salt and ionic conditionsn be reversibly denatured (13, 14). These can be chem-synthesized, site-specifically labeled, and thereforeecifically immobilized. Moreover, because aptamersuch more hydrophilic than antibodies, they may pro-urface passivation against nonspecific binding. Apta-have been used in cell labeling studies (15, 16) as wellctivating cell signaling pathways (17–20). However, on-ntly, aptamers have been used in lab-on-chip devicest, isolate, and detect tumor cells (21). Here, we reportfrom an RNA aptamer substrate to isolate epidermalh factor receptor (EGFR)–overexpressing primaryn glioblastoma (hGBM) cells, as well as geneticallyered mouse glioma cells that photocopy human glioma.pproach provides a strong cytologic analysis modalitylate and identify cancer cells. EGFR is the most fre-ly overexpressed receptor tyrosine kinase oncogene inman malignancies that is activated on binding varioush factors and, in consequence, initiates a signal trans-
n cascade that promotes cell migration, adhesion, inva-ell proliferation, angiogenesis, and antiapoptosis (22).
9371

. © 2010 American Association for Cancer. © 2010 American Association for Cancer. © 2010 American Association for Cancer

http://cancerres.aacrjournals.org/



The ovtumor(23–25100,00million(27). Tfrom(exonssmallQuanthumanseveraHere, waptampressiEGFRvapprosubpoand sp

Mate

Allotherw

AptamThe

select(R&Drandorine mhumanIgG1 (affinittional,sequestructwith pThe

aptamanti-EGUGCCAUGAAUGUCCGCAAUGUAGand mGGCGCATTThe

a dsDtechnonol prGTP,10 houDNase

rifiedwas vexciseovernpelletaptam(Thermthe Dsequeed approbeits 5′

PrepafunctThe

tions

Wan et al.

Cance9372

Published OnlineFirst November 9, 2010; DOI: 10.1158/0008-5472.CAN-10-0568

erexpression of EGFR has been associated with severals, and EGFR is an attractive target for cancer therapy). The expression levels of EGFR vary from 40,000 to0 proteins per cell in normal cells (26) and up tos of proteins per cell in some tissue culture cell lineshe most common mutant of EGFR, EGFRvIII, resultsa deletion of the extracellular amino acids 6 to 2732–7), and this variant is expressed in glioma, non–cell lung carcinomas, and breast carcinomas (28).itative fluorescence-activated cell sorting analysis ofglioma biopsy-derived cells has shown that there are

l hundred thousand EGFRvIII receptors per cell (28).e present evidence that our surface-immobilized EGFRers can be used to capture both primary hGBM cells ex-ng the endogenous wild-type EGFR as well as mutantIII. The data presented here provide a proof-of-principleach for identifying and isolating uniquely malignant
pulations of tumor cells with a high degree of sensitivity into 4
(H2O2

ing wiglass swatertemperinsed30 minylform1 mmsubstrdichloDNAend wwith 1of 5 μalloweEach sand dv/v). Tping u6-ami5 houDMF,was cthree1 μmo1× anEDTAtion aer andcontrothe sa7.5) w1 weewas aEGFRand inblocke

ecificity.

rials and Methods

chemicals were obtained from Sigma-Aldrich unlessise noted.

er preparationanti-EGFR RNA aptamer was isolated by iterativelying binding species against purified human EGFRSystems) from a pool that spanned a 62-nucleotidem region (29). The EGFR protein was purified from mu-yeloma cells and contained the extracellular domain ofEGFR (Leu25-Ser645) fused to the Fc domain of human

Pro100-Lys330) via a peptide linker (IEGRMD). A high-y (Kd = 2.4 nmol/L) anti-EGFR aptamer and a nonfunc-scrambled counterpart were extended with a capturence. The capture sequence did not disrupt aptamerures but was used as a hybridization handle for bindingrobes immobilized on surface.sequences for the extended anti-EGFR aptamer, mutanter, and relevant capture oligonucleotides were as follows:GFR aptamer (5′-GGCGCUCCGACCUUAGUCUCU-CGCUAUAAUGCACGGAUUUAAUCGCCGUAGAAAAG-UCAAAGCCGGAACCGUGUAGCACAGCAGAGAAUUA-CCCGCCAUGACCAG-3′), mutant aptamer (5′-GGCGC-ACCUUAGUCUCUGUUCCCACAUCAUGCACAAGGA-UCUGUGCAUCCAAGGAGGAGUUCUCGGAACCGU-CACAGCAGAGAAUUAAAUGCCCGCCAUGACCAG-3′),odified probe oligonucleotide (5′-amine-CTGGTCAT-GGCATTTAATTC-3′ or 5′-6FAM-CTGGTCATGGCGGG-TAATTC-3′). The capture sequence is underlined.anti-EGFR aptamer was prepared by transcribingNA template using Durascribe kits (Epicentre Bio-logies). The DNA template was PCR amplified, etha-ecipitated, and mixed with reaction buffer, DTT, ATP,2′ F-CTP, 2′ F-UTP, and a mutant T7 polymerase for
rs at 37°C. The DNA template was then degraded withtreatment for 30 minutes at 37°C. Aptamer was pu-
20 misolutio

r Res; 70(22) November 15, 2010

Research. on May 17, 2018cancerres.aacrjournals.org Downloaded from

on an 8% denaturing PAGE. The band for the aptamerisualized by UV shadowing, and the aptamer wasd and eluted in 0.3 mol/L sodium acetate (pH 5.2)ight at 37°C followed by ethanol precipitation. Thewas dissolved in water, and the concentration ofer was measured on a NanoDrop spectrophotometero Scientific). The aptamer was modified by extendingNA template at its 3′ end with a 24-nucleotidence tag, and then hybridizing the transcribed, extend-tamer with a cDNA oligonucleotide (referred to asoligonucleotide) labeled with 6-FAM or an amine atend.

ration of anti-EGFR aptamer/antibodyionalized substratesattachment method was adapted from earlier descrip-(30, 31). The glass slides, used as substrates, were cut× 4 mm2 pieces and cleaned in piranha solution

/H2SO4 in a 1:3 ratio) for 10 minutes at 90°C. After rins-th deionized (DI) water and drying in nitrogen flow, theubstrates were immersed in a 19:1 (v/v) methanol/DIsolution containing 3% APTMS for 30 minutes at roomrature. The silanized substrates were then sequentiallywith methanol and DI water and cured at 120°C forutes. The substrates were then immersed in a dimeth-amide (DMF) solution containing 10% pyridine andol/L phenyldiisothiocyanate (PDITC) for 2 hours. Eachate was then washed sequentially with DMF and 1,2-roethane and dried under a stream of nitrogen. Theprobes with an amine group modification at the 5′ere prepared at 30 μmol/L concentration in DI water% (v/v) N,N-diisopropylethylamine (DIPEA). A volumeL of DNA solution was placed on each substrate andd to incubate in a humidity chamber at 37°C overnight.ubstrate was then sequentially washed with methanoliethyl pyrocarbonate (DEPC)–treated DI water (0.02%,he functionalized surface was then deactivated by cap-nreacted PDITC moieties by immersion in 50 mmol/Lno-1-hexanol and 150 mmol/L DIPEA in DMF forrs. Each substrate was then sequentially rinsed withmethanol, and DEPC-treated DI water. The incubatorleaned with RNase-free and DEPC-treated DI watertimes. A volume of 5 μL of anti-EGFR RNA aptamer atl/L concentration was placed on each substrate innealing buffer [10 mmol/L Tris (pH 8.0), 1 mmol/L(pH 8.0), 100 mmol/L NaCl]. After 2 hours of hybridiza-t 37°C, substrates were washed with 1× annealing buff-DEPC-treated DI water for 5 minutes. The negativel devices were hybridized with mutant aptamer usingme protocol. The substrates were placed in 1× PBS (pHith 5 mmol/L magnesium chloride and kept at −20°C fork or used immediately. The EGFR antibody attachmentdapted from previous reports (32, 33). A 100 μg/mLantibody solution was placed on the glass substratescubated at 37°C for 1 hour. Then, the substrates wered with bovine serum albumin (10 mg/mL) solution for
nutes, washed thoroughly with PBS, and placed in PBSn.
Cancer Research

. © 2010 American Association for Cancer


Genetof EG(Ink4Em

from Ineurorus exnicityBachoweremonotein fo

IsolathGB

the UnTX) waveragmediucells wGBMsmors.GBMmay hand fomisedbeenmotheonly Cof thisresecttype Ewith plabeletec) anBoth Cically20 ngfibrob(Invitrand 1(HyCl60-mmclonalshownreferretransdprotein

MeninRat

from pfrom tfor 30trypsincells wmediugrow f

Resu

AptamTo

the acaptuand fiThe D(Alphaand Dto 70°tempefibrobameteprobeunderwith1× PBfluorecells,Mutacontroas thimageand erespec

TumoantibIn a

super(withcells.subst90 mition o10 mitime oTheredifferThe bThe s30 micellsaptamThe eexactl

Monitthe tufunctTo

substropticatumordescri

Surface Immobilized Aptamers for Cancer Cell Isolation and Microscopic Cytology

www.a


ic engineering, isolation, and characterizationFR-overexpressed mouse-derived tumor cellsa/Arf−/− EGFRvIII neural stem cells)bryonic (E13.5) neural stem cells (NSC) were isolatednk4a/Arf−/− embryo brain, maintained under standardsphere culture conditions, and infected with a retrovi-pressing the mutant EGFRvIII receptor. The tumorige-of these cells has been extensively characterized byo and colleagues (34). The Ink4a/Arf−/− EGFRvIII NSCsalso stably transduced with a lentivirus expressingmeric-cherry (referred to as m-cherry) fluorescent pro-r live cell imaging and identification.

ion and characterization of hGBM cellsM samples were obtained from consenting patients ativersity of Texas Southwestern Medical Center (Dallas,ith the approval of the Institutional Review Board. One, specimens >50 mm3 were placed into ice-cold HBSSm immediately on removal from the brain. Red bloodere removed using lymphocyte-M (Cedarlane Labs).are highly cellular and molecularly heterogeneous tu-Recent studies suggest that a minor cell population ofcells identified by a cell surface glycoprotein, CD133,ave an inexhaustible ability to self-renew, proliferate,rm a tumor when implanted into an immunocompro-host. GBM tumor cells that express CD133 have alsoreported to be highly resistant to radiation and che-rapy. While acknowledging the controversy of whetherD133+ cells are truly tumorigenic, for the purposesstudy, we used CD133+ cells isolated from surgically

ed samples that were also found to overexpress wild-GFR. The hGBM tumor tissue was gently dissociatedapain and dispase (both 2%), triturated, and then

d with a CD133/2 (293C3)–PE antibody (Miltenyi Bio-d sorted with FACSCalibur machine (BD Biosciences).D133+ and CD133− cells were suspended in a chem-

defined serum-free DMEM/F-12 medium, consisting of/mL of mouse EGF (PeproTech), 20 ng/mL of basiclast growth factor (PeproTech), 1× B27 supplementogen), 1× Insulin-Transferrin-Selenium-X (Invitrogen),00 units/mL penicillin–100 μg/mL streptomycinone), and plated at a density of 3 × 106 live cells/plate. Both CD133+ and CD133− fractions underwentexpansion and formed orthotopic tumors (data not). For all the experiments, we used the CD133+ fraction,d to as hGBM cells. The hGBM cells were also stablyuced with a lentivirus expressing m-cherry fluorescent.

ge-derived primary fibroblast-derived primary meningeal fibroblasts were obtainedostnatal day 3 rat pups. Briefly, meninges were peeledhe cerebral cortices and then processed by incubationminutes in 0.5% collagenase and 20 minutes in 0.06%/EDTA, and then triturated. Following trituration, theere plated in T-75 tissue culture flasks in DMEM/F-12
m containing 10% fetal bovine serum and allowed toor 1 week to confluence.
on thehumid

acrjournals.org


lts

er binding to cultured tumor cellsshow the selective binding of aptamer to tumor cells,nti-EGFR aptamer, annealed with 6-FAM–modifiedre oligonucleotide, was incubated with tumor cellsbroblasts, and interaction was measured as follows:NA probe labeled with 6-FAM was used as receivedDNA). Equal amounts of anti-EGFR RNA aptamer

NA capture probe were annealed by heating samplesC for 10 minutes and then slowly cooling to roomrature. Both mouse-derived tumor cells and primarylasts were seeded into separate PDMS wells (8-mm di-r) and cultured for 48 hours. The RNA/DNA capturewas incubated with cells at 37°C for 30 minutes5% CO2. After incubation, the cells were washed

1× PBS three times and stored in freshly sterilizedS for differential interference contrast (DIC) andscence imaging. DIC data were used to image theand fluorescence imaging was focused on aptamers.nt aptamer was also applied into the cells as al, and all experimental procedures were the sameose for the anti-EGFR aptamer. The fluorescences were taken using appropriate filters. The excitationmission wavelength of 6-FAM are 492 and 517 nm,tively.

r cell capture using anti-EGFR aptamer/ody substratesll experiments, the cells were centrifuged and thenatants were removed. Sterilized 1× PBS solution5 mmol/L MgCl2) was added to dilute the centrifugedAbout 50 μL of cell suspension were placed on eachrate. The substrates were incubated for 30, 60, ornutes at 37°C (15) and then washed with PBS solu-n a shaker (Boekel Scientific) at 90 rpm for 6 tonutes in orbital and reciprocal movements. Thef incubation was also studied for saturation effects.was no difference seen in the results for the threeent groups of 30-, 60-, and 90-minute incubation.uffer evaporation was seen for longer incubation.ubsequent incubations of cells were thus done fornutes. For tumor-specific isolation studies, the hGBMwere mixed with fibroblasts in a 1:1 ratio. Mutanter-functionalized substrates were used as controls.xperiments of EGFR capture with antibodies followedy the same procedure.

oring of the dynamic interactions betweenmor cells and the anti-EGFR aptamer-ionalized substratesvisualize tumor cell capture via the anti-EGFR aptamerates, these were placed on a custom-designed neuro-l microfluidic platform, and the interaction betweencells and surface-grafted aptamers was monitored asbed previously (35). Briefly, the substrates were placed
platform, which maintained 5% CO2 at 37°C and highity for live cell imaging. The interactions between cells
Cancer Res; 70(22) November 15, 2010 9373



and ausingevery30 minsubstrclosely

QuanFor

areas)were aber ofsurfac(numbdiamedata w20 μmtages

Discu

Theis of gthe affis to emuchbe unhigh aof consyntheidentiffirmedcells a

AptamThe

were haptamgliomaspecifnoncofluoretumoraptamincubaptamblasts.SimilaspecifThesepressEGFRmousemer isresultligandfalls oin EGF

CaptuderiveThe

bifunc

Figuretumor cDNA printeractCO2. AA, scheB, overthe celwavelefluoresc

Wan et al.

Cance9374


nti-EGFR aptamer surfaces were closely monitoredan inverted microscope (63×: DIC). Images were taken30 seconds, and the interaction was recorded forutes. The tumor cells were also seeded on the controlates (with mutant aptamer), and the interaction wasmonitored in a similar manner.

tification and statistical analysisanalysis, 5 representative images (of 25 total substratewere randomly taken from each substrate. The imagesnalyzed with Image-Pro Plus software. The total num-captured cells and their relevant diameters on thee were counted automatically, and the cell densitieser of cells per mm2) were calculated. To show theter of tumor cells on aptamer-grafted substrates, theere sorted into six groups based on cell sizes (fromto the maximum; 5-μm interval), and relevant percen-were obtained.

ssion

use of microfluidic devices to isolate rare tumor cellsreat importance. The capture of tumor cells requiresinity recognition of specific biomarkers. The challengefficiently isolate a small number of tumor cells from alarger pool of normal cells (5). Aptamers may prove toiquely useful for lab-on-chip devices because of theirnd specific affinities for analytes, and the versatilityjugation and labeling inherent in their chemicalsis. Before we used the RNA aptamer substrates toy and isolate EGFR-overexpressed cancer cells, we con-the specific binding between mouse-derived glioma

nd the anti-EGFR aptamers.

er binding to mouse-derived tumor cellsprobe oligonucleotides modified with 6-FAM dyeybridized to the anti-EGFR and mutant (as control)ers, and specific binding to cultured mouse-derivedcells was observed (Fig. 1). An additional control for

icity was to incubate the anti-EGFR aptamer with agnate cell, primary fibroblasts. After washing, greenscence was observed only with the mouse-derivedcell surface incubated with the labeled anti-EGFRers (Fig. 1B). The controls included (a) mutant aptamerated with mouse-derived glioma cells, (b) anti-EGFRer with fibroblast, and (c) mutant aptamer with fibro-The fluorescence intensity data are shown in Fig. 1C.r results were obtained for hGBM cells. This showed theic binding of anti-EGFR aptamer with tumor cells.mouse-derived tumor cells and primary hGBM cells ex-the endogenous wild-type EGFR as well as mutantvIII, but the expression level of wild-type EGFR on-derived tumor cells is much lower. Because the apta-also known to compete with EGF for binding, theses are consistent with the aptamer binding to the-binding domain III of the extracellular region, which
utside of the deletion encompassed by exons 2 to 7RvIII.
any aoligon

r Res; 70(22) November 15, 2010


re and morphologic characteristics of mouse-d tumor cells overexpressing EGFRfunctionalization of the substrates yielded a homo-tional layer of PDITC that can be used to immobilize

ngth of 6-FAM are 492 and 517 nm, respectively. C, averageence intensity of each group.

1. Anti-EGFR aptamer binding to the cultured mouse-derivedell. The RNA aptamer was annealed to 6-FAM–modifiedobe. The RNA aptamer–DNA probe complex was allowed toand bind to mouse-derived tumor cells at 37°C for 30 min in 5%fter binding, the cells were washed with 1× PBS three times.matic depicting mouse-derived cell bound with aptamer complex.laid fluorescent and DIC images. The green fluorescent showsl-bound aptamer molecules; the excitation and emission

mine-modified molecules. An amine-bearing captureucleotide was conjugated to the surface, which in turn

Cancer Research



alloweof unradsorpnucleomanyany futhe ussubstrstericsurfacthe apalso reactingGen

on the(Fig. 2derivethe anan isoloriginanumbmutanThe

the uscleic anonspstaticadsorpThe dcancegroupcance

and thaptamspecifcouldturedcomphas thtion, aprobea robuand seThe

tinct bstratecaptu(SD, 1esting30 μmbetwespreadthereand cconclunormaanti-Ean imphenostudie

Captu

Figuresteps oto scaleprobeson thehybridizanti-EGfor 2 h,with tumAfter inwere wfor 8 m


www.a


d the capture of the extended aptamer. The cappingeacted PDITC end groups ensured that nonspecifiction of aptamer did not occur. The use of probe oligo-tides for the functionalization of the substrates hasadvantages: it shows a generalized approach to capturenctional nucleic acid, a distinct advantage relative toe of proteins; it increases the distance between theate surface and the aptamer, alleviating the effects ofand/or electrostatic hindrance that may come frome tethering; and it increases the radius of gyration oftamer, thereby potentially increasing reactivity. Thissulted into very distinct behavior of cells when inter-with aptamers (discussed later).etically engineered mouse glioma cells were incubatedaptamer substrates and washed with warmed 1× PBS; ref. 36). A significantly higher number of the mouse-d EGFR-overexpressed tumor cells were seen bound toti-EGFR aptamer-functionalized surfaces (Fig. 3), withation efficiency (the ratio of captured cell number to thel number of cells on the surface) of 62.32%. A very smaller of cells (average 7 cells/mm2) were captured on thet aptamer-functionalized control substrates.se results may reveal an additional important feature ofe of nucleic acids on possible lab-on-chip devices. Nu-cids may provide a passivation layer that minimizesecific adsorption. The hydrophilic surface and electro-repulsionmay have prevented any nonselective physicaltion of the cells on mutant aptamer substrates (37–39).ensity and amount of sialylation on the surface ofr cells is higher than on normal ones (40–42). Carboxyl
s from sialic acid cause a net negative surface charge onr cells. The repulsion between neg
EGF

in.

acrjournals.org


e negative charges from the surface-functionalizeder can be another putative reason for the lack of non-ic adsorption. On the other hand, the cancer cells thatselectively interact and bind to the aptamer got cap-on the surface, even in the presence of above-statedeting forces. The use of capture oligonucleotides thuse advantage of reducing nonspecific binding or adsorp-dding to the selectivity of the substrates. The use ofDNA to covalently immobilize aptamers thus providesst passivation of the surface that provides functionalitylectivity with screening effects of surface charges.density and diameters of captured cells showed dis-ehavior on the anti-EGFR and mutant aptamer sub-s (Fig. 3B). On average, there were ∼392 cancer cellsred per mm2 on 12 anti-EGFR aptamer substrates43.3), with the size ranges depicted in the figure. Inter-ly, ∼70% of the captured cells had diameters above, whereas the size of these cells in suspension rangeden 25 and 30 μm. This indicates that cancer cells wereing on the anti-EGFR aptamer substrates. Althoughare reports that show size differences of normal cellsancer cells as discriminating factor, there is still notsive evidence that cancer cells are bigger than thel cells. In any case, the new class of binding betweenGFR aptamer substrates and EGFR-expressing cells isportant analytic tool, serving as a novel and importantmenon that can be a discriminating factor in cytologics for the confirmation of the captured tumor cells.

re of hGBM cells
R expression level on hGBM cells was ∼50% com-
atively charged cells pared with the genetically engineered mouse glioma cells

2. Schematics showingf experiments (not drawn). The amine-modified DNAwere first immobilizedglass substrates. Afteration with 1 μmol/LFR RNA aptamer at 37°Csubstrates were incubatedor cells at 37°C for 30 min.

cubation, the substratesashed with 1× PBS




(levelrelativturedEGFRbind t

Figurewere inA, aver(averagmm2; S(averag2.8). *,anti-EGFR aptamer substrates. C and D, representative pictures of tumorcells on (C) mutant aptamer and (D) anti-EGFR aptamer-grafted surfaces.

Figurewith hGthe musubstrasubstra56 cellsaptameper mm

Wan et al.

Cancer Res; 70(22) November 15, 20109376



verified by Western blot; data not shown). Despite theely lower EGFR expression level, hGBM cells were cap-with comparable sensitivity and specificity by the anti-
aptamer–functionalized substrates. These cells did noto mutant aptamer control substrates (Fig. 4).
3. The density and size ranges of captured cells. Substratescubated with mouse-derived tumor cells and washed with 1× PBS.age tumor cell density on 12 anti-EGFR aptamer substratese, 392 cells per mm2; max, 831 cells per mm2; min, 284 cells perD, 143.3) and on 12 control substrates with mutant aptamere, 7 cells per mm2; max, 11 cells per mm2; min, 0 cells per mm2; SD,P < 0.01. B, distribution of the diameters of tumor cells on 12

4. The hGBM cells on the substrates. Substrates were incubatedBM and washed with PBS. The hGBM cells captured on (A)tant aptamer control substrate and (B) the anti-EGFR aptamerte. C, average hGBM cell density on 12 anti-EGFR aptamertes (average, 117 cells per mm2; max, 228 cells per mm2; min,per mm2; SD, 44.4) and on 12 control substrates with mutant

2 2
r (average, 4 cells per mm ; max, 13 cells per mm ; min, 0 cells2; SD, 4.1). *, P < 0.01.
Cancer Research



Figushapesubstrcells,adsorbent (don av

anti-Eand 5of 38.densitdensitrisonof capcells (termsly engdensit

IsolatIn t

and paptamly mormer.substrior, ablast cdensitsubstrthe recells wcent ifaces,and fluified tentiatdid nostrateshowestrateSD, 6.imagefibrobfirst-ocaptursubstr

Tab

CapUnc

NOTcaucomresp4.3)

FigureSubstrawashedfrom ththat we


www.a


re 4 shows the clear difference in the number and cells of hGBM cells on anti-EGFR and mutant aptamerates. Along with difference in the numbers of capturedthe shapes of the cells bound with aptamers anded on mutant aptamer surfaces were also quite differ-

with PBS. A and B, DIC and fluorescent images, respectively,e same position. The circles in A indicate a few fibroblastsre captured and cannot be seen in B.

iscussed later). Analysis of 12 substrates showed thaterage 117 hGBM cells were captured per mm2 on

suspenthe su

le 1. Comparison betwee

ntibody

hGBM Fibrob

, respectively.

acrjournals.org


GFR aptamer substrate (SD, 44.4; max and min of 2286 cells per mm2, respectively), with isolation efficiency74%. In the control mutant substrate group, averagey of 4 cells per mm2 was seen (SD, 3.1; max and minies of 13 and 0 cells per mm2, respectively). In compa-with the mouse-derived tumor cells, a smaller numbertured hGBM cells than that for mouse-derived tumordiscussed in previous section) can be explained inof an overly high number of EGFR that were genetical-ineered in mouse-derived tumor cells. The decreasedy of captured hGBM cells was thus as expected.

ion of cancer cells from cell mixturehe above two experiments (mouse-derived tumor cellsrimary hGBM cells), we confirmed that the anti-EGFRer-functionalized substrates could capture significant-e tumor cells compared with that of the mutant apta-Toward the application of aptamer-functionalizedates in isolating tumor cells and to study their behav-cell mixture was used. A mixture of hGBM and fibro-ells was prepared in a ratio of 1:1, and the average celly on the surface was 303 per mm2 (SD, 11.95). Theates were incubated in the mixture and washed, andsults were imaged. In parallel experiments, only hGBMere incubated with substrates. Both DIC and fluores-mages were taken (Fig. 5A and B). In hGBM-only sur-the fluorescent intensities were not uniform when DICorescent images were overlaid (hGBM cells were mod-o express m-cherry fluorescent protein for clear differ-ion). There were ∼16.7% cells from 12 substrates thatt show any fluorescence (189 of 1,133 cells from 12 sub-s; average, 16; SD, 5.1). The data from the mixture groupd no fluorescence from ∼27.5% cells from 12 sub-s (378 of 1,376 cells from 12 substrates; average, 31.5;8). The cells that did not show up in fluorescences included captured hGBM and nonspecifically boundlast cells. The difference of the two percentages, as arder approximation, shows that on average ∼10.8%ed cells were fibroblasts. Thus, the anti-EGFR aptamerates can selectively isolate and enrich a 1:1 mixture
sion ratio of fibroblasts to cancer cells, to 1:8.24 onrface. In the EGFR antibody substrate control group,
aptamer and
n capture efficiency of anti-EGFR
hGBM

Ca

. © 2010 America

EGFR-specific antibody

aptamer

Fibrob

ncer Res; 70(22) November 15

n Association for Cancer

EGFR a

cells
last
Anti-EGFR

cells
last
tured 188 68 84 10aptured 30 150 134 208

E: The average number of cells on 10 substrate surfaces is 436 (5 with EGFR antibody and 5 with anti-EGFR aptamer). Be-se the hGBMs were mixed with fibroblast in the ratio of 1:1, 218 hGBMs and 218 fibroblast cells were considered for thisparison. The average captured hGBMs and fibroblasts on EGFR antibody substrates are 188 (SD, 17.2) and 68 (SD, 11.0),ectively, and the average captured hGBMs and fibroblasts on anti-EGFR aptamer substrates are 84 (SD, 22.8) and 10 (SD,

5. The hGBM and fibroblast cells on the substrate surfaces.tes were incubated with mixture of hGBM and fibroblast and

, 2010 9377


the raThe sption wresultscal scecan enconcecationsubstrimporanti-Ether sting ra

ShapefunctiIn a

distincirregulevels

far lesma cetheirEGFRboundtumorical ingraftesurfac(43). Tas larshoweFig. 6derivetant atakenare evalso b

ureou

nd Cceltantpectr onamend Bthefacechatant aptamer substrates.

Wan et al.

Cance9378


tio of captured fibroblasts to cancer cells was 1:2.77.ecificity of aptamer and antibody on cancer cell isola-as thus 94.82% and 68.81%, respectively (Table 1). Theshow that aptamer has higher specificity. In a practi-nario, as a lower limit, the aptamer-grafted substratesrich the amount of cancer cells by an order from thentration in the solution. In a cyclic iteration appli-, a sample can be run for multiple times over theates to increase the capture efficiency. It may betant to note here that “mean capture yield” usingpCAM antibodies has been shown to be ∼65% (5). Fur-udies are under way to verify this efficiency with vary-tios of tumor cells and cancer xenografts.

and size of cancer cells ononalized substratesll the experiments, the cell shapes and sizes showed at behavior: fibroblasts altered their fusiform, stellar, or
lar shape to spherical. Mouse fibroblasts showed lowof wild-type EGFR expression on their cell membrane,
the in(Supp

r Res; 70(22) November 15, 2010


s than primary hGBM cells or the mouse-derived glio-lls. To bind with anti-EGFR aptamer, the cells alteredshape to decrease their surface area to increase thedensity that would come in contact with the surface-aptamer. The temporal images of mouse-derivedcells also showed changes in cell shapes from spher-suspension to semi-elliptic and flat on the aptamer-

d surfaces (Fig. 6). The different cell behavior on thee may be a result of different elasticity in cancer cellshe EGFR-overexpressed cells were reshaping to coverge of an area as possible. Temporal imaging alsod tumor cell migration on surfaces. The images forA and C were taken at the beginning when mouse-d tumor cells were seeded on the anti-EGFR and mu-ptamer surfaces, and the images for Fig. 6B and D wereafter 30 minutes. Changes in cells shapes and flatnessident in going from Fig. 6A to B only. The size of cellsecame bigger, and many pseudopodia formed during

Figof mA athemureslateaptA aonsurnomu

cubation period on anti-EGFlementary Video S1). The vide

. © 2010 American Associati

6. The changes in shapesse-derived tumor cells., taken 3 min after seeding

ls on the anti-EGFR andaptamer substrates,ively. B and D, taken 30 minthe anti-EGFR and mutantr substrates, respectively., changes of cell shapesanti-EGFR aptamer-graftedin 30 min. C and D,

nges in cell shapes on

R aptamer substrateso also shows that the

Cancer Research

on for Cancer


tumorwere aFig. 6Cthe mrepulschargely, the30 mishowglobulof candalitylogic fphysicfiltratigicalinadeqthe lepossibmaticafunctimargitumorthe exyond tmors,enrichfor virboth ticalmstep in

Conc

It hsubstr

canceaptamexpressensitcallyThe ibetwethe ceactivitThe sisolatidramatically changing intervention and prognosis ofmetas

Disclosure of Potential Conflicts of Interest

No p

Ackn

We thelp with experimental setup; Waseem Asghar and Melissa Johnson for helpwith manuscript preparation; Kailash Karthikeyan for help with videocapturiequipm

Grant

TheCAREERat AustCancer

Theof pageaccorda

Refe1. Mo

a rIEE

2. Hutio98

3. Butissin49

4. MatouHo

5. NacirNa

6. Fencelmic

7. Tibsel20


www.a


cells on the substrate surface had strong activity andrbitrarily changing their shape. In contrast, going fromto D, first, only much fewer cells did get capture on

utant aptamer surfaces, and second, those too facedion from the hydrophilic glass surface and the negatives from the immobilized oligonucleotides. Consequent-cells had almost no change in their sizes during thenutes. The data on hGBM cell shapes in Fig. 4A alsothat cells on mutant aptamer substrates maintainedar shape as discussed above. The spreading and flatnesscer cells on aptamer surfaces can be an important mo-for detection as an additional method to support histo-indings and further identify tumor cells based on theiral behaviors. In addition, hGBM cells are diffusively in-ve and currentmethods to define tumormargins for sur-resection, using magnetic resonance imaging, areuate. Histologic evidence suggests that tumor cells atading edge may express high levels of EGFR. It isle, therefore, that freshly resected tumor could be enzy-lly dissociated and captured on the anti-EGFR aptamer-onalized substrates, in real time, to better define tumorns. Such information, thus, can help guide the extent ofresection as well. There is considerable evidence thattent of resection is directly related to overall survival. Be-he specific application for management of hGBM tu-our findings are especially important given thatment of rare circulating tumor cells may be difficulttually any lab-on-chip device. The use of aptamers leadso high passivation and the presentation of unique phys-orphologies, and thusmay be a novel first-level detectionpoint-of-care examination of circulating tumor cells.

lusions

as been shown that anti-EGFR RNA aptamernize, capture, and isolate

ReceOnlineF

be AG, de Grooth BG, Greve J, et al. Magnetic field design forecting and aligning immunomagnetic labeled cells. Cytometry02;47:163–72.

8. Lemo

9. HecirPro

10. To77

11. Baresp526

12. Dacotes

13. BuMi

14. SuNa

15. Faapca

16. Ch

acrjournals.org


r cells that are known to overexpress EGFR. Theers can capture both human and murine GBM cellssing wild-type EGFR and mutant EGFRvIII with highivity and specificity. Aptamer substrates also specifi-isolated hGBM cells from a mixture of fibroblasts.solation efficiency depended on strong bindingen aptamer and the amount of EGFR expression onll membrane. The change in cell shape and cellulary can serve as a novel way of identifying tumor cells.ubstrates can also be used for identification andon of circulating tumor cells from peripheral blood,

tasis.

otential conflicts of interest were disclosed.

owledgments

hank Charles Huang, Priyanka P. Ramachandran, and Swati Goyal for

ng; and staff of Nanotechnology Research and Teaching Facility forent training.

Support

work at the University of Texas at Arlington was supported with NSFgrant ECCS-0845669 (S.M. Iqbal). The work at the University of Texas

in was supported by Welch Foundation grant F-1654 and NationalInstitute Award Number 5R01CA119388-05.costs of publication of this article were defrayed in part by the paymentcharges. This article must therefore be hereby marked advertisement innce with 18 U.S.C. Section 1734 solely to indicate this fact.

ived 02/17/2010; revised 08/24/2010; accepted 09/14/2010; publishedirst 11/09/2010.
ates can specifically recog
renceshamed H, McCurdy LD, Szarowski DH, et al. Development ofare cell fractionation device: application for cancer detection.E Trans Nanobioscience 2004;3:251–6.ang LR, Cox EC, Austin RH, et al. Continuous particle separa-n through deterministic lateral displacement. Science 2004;304:7–90.stin SA, Gyselman VG, Siddiqi S, et al. Cytokeratin 20 is not aue-specific marker for the detection of malignant epithelial cellsthe blood of colorectal cancer patients. Int J Surg Invest 2000;2:–57.cDonald MP, Neale S, Paterson L, et al. Cell cytometry with a lightch: sorting microscopic matter with an optical lattice. J Biol Regulmeostatic Agents 2004;18:200–5.grath S, Sequist LV, Maheswaran S, et al. Isolation of rareculating tumour cells in cancer patients by microchip technology.ture 2007;450:1235–9.g J, Soper SA, McCarley RL, et al. Separation of breast cancerls from peripherally circulating blood using antibodies fixed inrochannels. Proc SPIE 2004;5312:278–93.

e H, Sun E, Ham D, et al. Chip-NMR biosensor for detection andlecular analysis of cells. Nat Med 2008;14:869–74.W, Wang H, Hartmann LC, et al. In vivo quantitation of rare

culating tumor cells by multiphoton intravital flow cytometry.c Natl Acad Sci U S A 2007;104:11760–5.ner M, Irimia D. Blood-on-a-chip. Annu Rev Biomed Eng 2005;7:–103.as IO, van den Berg FM, Mulder JW, et al. Potential false-positiveults with antigen enhancement for immunohistochemistry of the3 gene product in colorectal neoplasms. J Pathol 1996;178:4–7.lle F, Lopez J, Caillot D, et al. False-positive results caused bytton swabs in commercial Aspergillus antigen latex agglutinationt. Eur J Clin Microbiol Infect Dis 2002;21:130–2.nka DH, Stockley PG. Aptamers come of age—at last. Nat Revcrobiol 2006;4:588–96.llenger BA, Gilboa E. Emerging clinical applications of RNA.ture 2002;418:252–8.rokhzad OC, Jon S, Khademhosseini A, et al. Nanoparticle-
tamer bioconjugates: a new approach for targeting prostatencer cells. Cancer Res 2004;64:7668–72.en HW, Medley CD, Sefah K, et al. Molecular recognition of



sm99

17. Napedoind

18. Usop

19. Ching199

20. BatheEx

21. Phimm

22. MeAm

23. Fraoffor200

24. MaEG

25. Chcepthevat20

26. Ca197

27. Cakin

28. Wiizagro

29. OsanBio

30. Möstuloid

31. Iqbsel

32. Memi20

33. RaanBio

34. Barecdifcy

35. Kimfor9:2

36. Chcach

37. MoPs

38. Haenalb

39. Tato

40. Kotyp

41. Khac15

42. Silin h41

Wan et al.

Cance9380


all-cell lung cancer cells using aptamers. ChemMedChem 2008;3:1–1001.gel-Wolfrum K, Buerger C, Wittig I, et al. The interaction of specificptide aptamers with the DNA binding domain and the dimerizationmain of the transcription factor Stat3 inhibits transactivation anduces apoptosis in tumor cells. Mol Cancer Res 2004;2:170–82.man N, Blatt LM. Nuclease-resistant synthetic ribozymes: devel-ing a new class of therapeutics. J Clin Invest 2000;106:1197–202.arlton J, Sennello J, Smith D. In vivo imaging of inflammation us-an aptamer inhibitor of human neutrophil elastase. Chem Biol7;4:809–16.tes PJ, Laber DA, Miller DM, et al. Discovery and development ofG-rich oligonucleotide AS1411 as a novel treatment for cancer.

p Mol Pathol 2009;86:151–64.illips JA, XuY, Xia Z, et al. Enrichment of cancer cells using aptamersobilized on a microfluidic channel. Anal Chem 2008;81:1033–9.ndelsohn J. EGF receptors as a target for cancer therapy. TransClin Climatol Assoc 2004;115:249–53.novic A, Gunaratnam L, Smith K, et al. Translational up-regulationthe EGFR by tumor hypoxia provides a nonmutational explanationits overexpression in human cancer. Proc Natl Acad Sci U S A7;104:13092–7.heswaran S, Sequist LV, Nagrath S, et al. Detection of mutations inFR in circulating lung-cancer cells. N Eng J Med 2008;359:366–77.akravarti A, Loeffler JS, Dyson NJ. Insulin-like growth factor re-tor I mediates resistance to anti-epidermal growth factor receptorrapy in primary human glioblastoma cells through continued acti-ion of phosphoinositide 3-kinase signaling. Cancer Res 2002;62:0–7.rpenter G, Cohen S. Epidermal growth factor. Annu Rev Biochem9;48:193–216.rpenter G. The biochemistry and physiology of the receptor-ase for epidermal growth factor. Mol Cell Endocrinol 1983;31:1–19.kstrand CJ, McLendon RE, Friedman AH, et al. Cell surface local-tion and density of the tumor-associated variant of the epidermalwth factor receptor, EGFRvIII. Cancer Res 1997;57:4130–40.
borne SE, Matsumura I, Ellington AD. Aptamers as therapeuticd diagnostic reagents: problems and prospects. Curr Opin Cheml 1997;1:5–9.
43. LeneAc

r Res; 70(22) November 15, 2010


ller R, Csáki A, Köhler JM, et al. DNA probes on chip surfacesdied by scanning force microscopy using specific binding of col-al gold. Nucleic Acids Res 2000;28:e91.al SM, Akin D, Bashir R. Solid-state nanopore channels with DNAectivity. Nat Nanotechnol 2007;2:243–8.zzasoma L, Bacarese-Hamilton T, Di Cristina M, et al. Antigencroarrays for serodiagnosis of infectious diseases. Clin Chem02;48:121–30.j J, Herzog G, Manning M, et al. Surface immobilisation oftibody on cyclic olefin copolymer for sandwich immunoassay.sens Bioelectron 2009;24:2654–8.choo RM, Maher EA, Ligon KL, et al. Epidermal growth factoreptor and Ink4a/Arf Convergent mechanisms governing terminalferentiation and transformation along the neural stem cell to astro-te axis. Cancer Cell 2002;1:269–77.Y, Karthikeyan K, Chirvi S, et al. Neuro-optical microfluidic plat-

m to study injury and regeneration of single axons. Lab Chip 2009;576–81.eung LSL, Zheng X, Stopa A, et al. Detachment of capturedncer cells under flow acceleration in a bio-functionalized micro-annel. Lab Chip 2009;9:1721–31.risaki H. A novel system for the analysis of cell attachment ineudomonas syringae. J Microbiol Methods 1995;22:69–78.zen KC, Brawner DL, Riesselman MH, et al. Differential adher-ce between hydrophobic and hydrophilic yeast cells of Candidaicans. Infect Immun 1991;59:907–12.mada Y, Ikada Y. Effect of preadsorbed proteins on cell adhesionpolymer surfaces. J Colloid Interface Sci 1993;155:334–9.koglulu E, Sonmez H, Uslu E, et al. Sialic acid levels in variouses of cancer. Cancer Biochem Biophys 1992;13:57–64.adapkar SV, Sheth NA, Bhide SV. Independence of sialicid levels in normal and malignant growth. Cancer Res 1975;35:20–3.lanaukee P, Pönniö M, Jääskeläinen IP. Occurrence of sialic acidsealthy humans and different disorders. Eur J Clin Invest 1999;29:

3–25.kka M, Laidler P, Ignacak J, et al. The effect of chitosan on stiff-
ss and glycolytic activity of human bladder cells. Biochim Biophysta 2001;1540:127–36.
Cancer Research



Correction

Correction: Surface-Immobilized Aptamers forCancer Cell Isolation and Microscopic Cytology

In this article (Cancer Res 2010:70;9371–80), which was published in the November 15,2010 issue of Cancer Research (1), several of the author affiliations were incorrect. Thecorrect affiliations are provided below.

Yuan Wan1,2, Young-tae Kim1,2, Na Li3, Steve K. Cho4,5, Robert Bachoo4–6, Andrew D.Ellington3, and Samir M. Iqbal2,7,8

1Department of Bioengineering, 2Nanotechnology Research and Teaching Facility,University of Texas at Arlington, Arlington; 3Institute for Cell and Molecular Biology,University of Texas at Austin, Austin; 4Internal Medicine, 5Annette G. Strauss Centerfor Neuro-Oncology; 6Department of Neurology, University of Texas SouthwesternMedical Center, Dallas; 7Department of Electrical Engineering, 8Joint GraduateCommittee of Bioengineering Program, University of Texas at Arlington, and Uni-versity of Texas Southwestern Medical Center at Dallas, University of Texas atArlington, Arlington, Texas

Reference1. Wan Y, Kim Y, Li N, Cho SK, Bachoo R, Ellington AD, et al. Surface-immobilized aptamers for cancer

cell isolation and microscopic cytology. Cancer Res 2010:70;9371–80.

Published OnlineFirst January 11, 2011�2011 American Association for Cancer Research.doi: 10.1158/0008-5472.CAN-10-4187

CancerResearch

Cancer Res; 71(2) January 15, 2011626

2010;70:9371-9380. Published OnlineFirst November 9, 2010.Cancer Res Yuan Wan, Young-tae Kim, Na Li, et al. Microscopic CytologySurface-Immobilized Aptamers for Cancer Cell Isolation and

Updated version

10.1158/0008-5472.CAN-10-0568doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerres.aacrjournals.org/content/suppl/2010/11/05/0008-5472.CAN-10-0568.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerres.aacrjournals.org/content/70/22/9371.full#ref-list-1

This article cites 43 articles, 10 of which you can access for free at:

Citing articles

http://cancerres.aacrjournals.org/content/70/22/9371.full#related-urls

This article has been cited by 3 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected] at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/70/22/9371To request permission to re-use all or part of this article, use this link

Research. on May 17, 2018. © 2010 American Association for Cancercancerres.aacrjournals.org Downloaded from


http://cancerres.aacrjournals.org/lookup/doi/10.1158/0008-5472.CAN-10-0568

http://cancerres.aacrjournals.org/content/suppl/2010/11/05/0008-5472.CAN-10-0568.DC1

http://cancerres.aacrjournals.org/content/70/22/9371.full#ref-list-1

http://cancerres.aacrjournals.org/content/70/22/9371.full#related-urls

http://cancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://cancerres.aacrjournals.org/content/70/22/9371


cancer research surface-immobilized aptamers for...

Documents