Cancer Letters 330 (2013) 84–95

Contents lists available at SciVerse ScienceDirect

Cancer Letters

journal homepage: www.elsevier .com/ locate/canlet

RNA aptamers targeting cancer stem cell marker CD133

0304-3835/$ - see front matter Crown Copyright � 2012 Published by Elsevier Ireland Ltd. All rights reserved.http://dx.doi.org/10.1016/j.canlet.2012.11.032

⇑ Corresponding authors. Addresses: Department of Pathology and ForensicScience, Dalian Medical University, 9 Western Section, Lyshun Sooth Street, Dalian116044, China. Tel.: +86 411 8611 0050; fax: +86 411 8611 0050 (L. Li), School ofMedicine, Deakin University, Pigdons Road, Waurn Ponds, Victoria 3217, Australia.Tel.: +61 3 5227 1149; fax: +61 3 5227 2945 (W. Duan).

E-mail addresses: lilianhong917@126.com (L. Li), wduan@deakin.edu.au (W.Duan).

Sarah Shigdar a, Liang Qiao b, Shu-Feng Zhou c, Dongxi Xiang a, Tao Wang a,d, Yong Li e, Lee Yong Lim f,Lingxue Kong g, Lianhong Li h,⇑, Wei Duan a,h,⇑a School of Medicine, Deakin University, Pigdons Road, Waurn Ponds, Victoria 3217, Australiab Storr Liver Unit at The Westmead Millennium Institute, Department of Medicine and Western Clinical School of The University of Sydney, Westmead, NSW 2145, Australiac Department of Pharmaceutical Sciences, College of Pharmacy, University of South Florida, Tampa, FL 33612, USAd The Nursing College of Zhengzhou University, Zhengzhou 450052, Chinae Cancer Care Centre, St. George Hospital, and St. George Clinical School, Faculty of Medicine, University of New South Wales, Kensington, NSW 2052, Australiaf Laboratory for Drug Delivery, Pharmacy, University of Western Australia, Crawley, WA 6009, Australiag Institute for Frontier Materials, Deakin University, Pigdons Road, Waurn Ponds, Victoria 3217, Australiah Department of Pathology and Forensic Science, Dalian Medical University, Dalian 116044, China

a r t i c l e i n f o a b s t r a c t

Article history:Received 14 September 2012Received in revised form 19 November 2012Accepted 19 November 2012

Keywords:AC133Cancer stem cellCD133RNA aptamerSELEXTherapeutic

The monoclonal antibody against the AC133 epitope of CD133 has been widely used as a cell surface mar-ker of cancer stem cells in several different cancer types. Here, we describe the isolation and character-isation of two RNA aptamers, including the smallest described 15 nucleotide RNA aptamer, whichspecifically recognise the AC133 epitope and the CD133 protein with high sensitivity. As well, both theseaptamers show superior tumour penetration and retention when compared to the AC133 antibody in a 3-D tumour sphere model. These novel CD133 aptamers will aid future development of cancer stem celltargeted therapeutics and molecular imaging.

Crown Copyright � 2012 Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction

CD133, also known as Prominin-1, is a pentaspan, highly glycos-ylated, membrane glycoprotein that is associated with cholesterolin the plasma membrane [1,2]. Though this protein is known to de-fine a broad population of cells, including somatic stem and pro-genitor cells, and is expressed in various developing epithelialand differentiated cells, its exact function is still being elucidated.It has, however, been linked to the Notch-signalling pathwaywhich is critical for binary cell fate, differentiation of intestinal epi-thelium, and lymphopoiesis [3]. CD133 has gained its prominencein the cancer research field due to its reported role as a marker ofcancer stem cells (CSCs) in glioblastomas [4]. Indeed, growing evi-dence has shown that CD133 is an important cell surface markerfor CSCs in a variety of solid cancers, including those of the brain,prostate, pancreas, melanoma, colon, liver, lung and ovarian can-

cers [5] by virtue of the enhanced tumourigenic potential ofCD133+ cells versus their negative counterparts in immunodefi-cient mice [6].

Most of the work in isolating CD133+ putative cancer stem cellsubpopulation from the bulk cancer cells utilises one monoclonalantibody, AC133 [7]. However, there has been some controversyregarding the notion that CD133 can be used as a marker for CSCs,as investigators from independent laboratories showed that theCD133� cells are also tumourigenic in immunocompromised mice[8–10]. This contention is further complicated by the near ubiqui-tous expression of CD133 on non-CSCs as well as CSCs, especially intumours of the colon [11]. A recent study has shed some light onthis, with a conformational change postulated to hide an epitopeon the second extracellular membrane loop of CD133 during thedifferentiation process [5]. Kemper and co-workers suggested thatthe CD133 protein becomes differentially folded as a result of gly-cosylation, thus masking the AC133 epitope. These results werefurther supported by later studies [5,12] which suggested thatthe AC133 epitope, rather than the complete CD133 protein, isthe marker for CSCs. As well, a recent report has shown that theAC133 epitope is lost upon cell differentiation, suggesting that thisepitope is a marker of primitive cells [5]. While the AC133 epitopehas been shown to be a marker for CSCs, not all cells positive for

S. Shigdar et al. / Cancer Letters 330 (2013) 84–95 85

the AC133 epitope are CSCs. In fact, the first description of this epi-tope was by Yin et al. in 1997, who described this as a marker ofhaematopoietic stem and progenitor cells [13]. However, theexpression of AC133 on these cells is approximately 1000-fold low-er than that observed in CSCs [14]. Despite the on-going debate ofthe utility of using AC133 to identify cancer stem cells, a retrospec-tive study on colorectal patients showed that a high level of AC133expression was associated with a poorer prognosis [15], though thesole use of the AC133 antibody is not recommended as it is thoughtto underestimate the level of CD133 expression [1,16].

AC133-positive cells have been shown to have an increasedresistance to radiation therapy due to activation of the DNA dam-age checkpoint proteins, and an increased chemoresistance due toan increased Akt/PKB and Bcl-2 cell survival response [17]. Thesedata suggest that a more targeted response is required to eradicatethis population of cells, especially given the increasing evidenceregarding the roles that CSCs play in the relapse of cancer after ini-tial treatments. Immunotherapy has had a great impact on thetreatment of cancer in recent years [18,19]. However, the use ofantibodies, even humanised antibodies, can lead to adverse side ef-fects with fatal consequences [20]. This has led to the search for‘bigger and better’ options. There have been several attempts touse nucleic acids as therapeutics though these have met with dis-appointing results, not least because of the failure of these nucleicacids to enter the cell [21]. The reports in 1990 by two separategroups describing the generation of nucleic acids that can bind tar-get molecules in the same manner as antibodies seemed to be theanswer [22,23]. These chemical antibodies, termed aptamers, havebeen increasingly utilised for clinical applications recently. Indeed,one RNA aptamer has been approved by the FDA and several moreare in clinical trials [21,24]. The increased interest in these apta-mers is due to the fact that they exhibit no immunogenicity, littlebatch-to-batch variation due to being chemically synthesised, andare more stable than conventional antibodies. Due to their smallsize, aptamers also show superior tumour penetration. One impor-tant feature of these chemical antibodies is their versatility as theycan be attached to nanoparticles, drugs, imaging agents or othernucleic acid therapeutics without loss-of-function [25,26]. Thisfunctionalisation is leading to new and more targeted therapies,with fewer side effects than current treatment modalities [25].When compared to conventional treatment which is largely a pas-sive process, targeted delivery systems are much more effective.For an aptamer to be an effective drug delivery agent, the aptamermust be efficiently internalised upon binding to its target on thecell surface [27].

In this study, we performed iterative rounds of an in vitro selec-tion process, known as the systematic evolution of ligands byexponential enrichment (SELEX), to identify RNA aptamers thatspecifically bind to CD133. Further studies identified one aptamer,CD133-A15, specifically bound to the same epitope as the AC133antibody, while the other aptamer, CD133-B19, bound to the extra-cellular domain of the CD133 protein. These aptamers were effi-ciently internalised into CD133-positive cancer cells and showedsuperior penetration of three-dimensional tumour sphere.

2. Materials and methods

2.1. Cell lines and cell culture

The cell lines of human origin used in this study were purchased from Americantype Culture Collection. They are human colorectal cancer HT-29; human hepato-cellular carcinoma Hep3B; human glioblastoma multiform carcinoma T98G; humanembryonic kidney cells HEK293T; human ductal breast carcinoma, T47D; humanlung adenocarcinoma, A549; human ovarian teratocarcinoma, PA-1; human hepa-toma, PLC/PRC/5; and human prostate carcinoma, DU145. Cells were grown andmaintained in culture with Dulbecco’s Modified Eagle Medium (DMEM) (Invitrogen,

Victoria, Australia) supplemented with 10% foetal calf serum (FCS) (HEK293T andHT-29), or Minimum Essential Medium (MEM) (Invitrogen) supplemented with10% FCS (Hep3B and T98G). Cells were maintained at 37 �C in a 5% CO2 atmosphere.

2.2. Protein expression and cell SELEX

Human CD133 cDNA was purchased from Invitrogen and cloned into a mamma-lian expression vector, pcDNA 3.1/V5-His-TOPO (Invitrogen). The recombinant 6�Histine-tagged CD133 was transiently expressed in HEK293T cells. Briefly, HEK293Tcells were seeded in 100 mm or 60 mm dishes to reach 70% confluency after 24 hincubation, and transfected with a total of 24 or 8 lg, respectively, of CD133 usingLipofectamine 2000 (Invitrogen Life Technologies) in antibiotic-free mediumaccording to the manufacturer’s instructions. Following a 72 h incubation, thetransfected cells were used as the target for cell SELEX. Successful transfectionand expression of the recombinant CD133 were confirmed using flow cytometryand the AC133-APC antibody (Miltenyl Biotech) prior to each round of SELEX.

2.3. SELEX selection

A DNA library containing a central 40-nt randomised sequence (50-TAA TAC GACTCA CTA TAG GGA GAC AAG AAT AAA CGC TCA A-N40-TTC GAC AGG AGG CTC ACAACA GGC, with the T7 RNA polymerase promoter sequence underlined) was syn-thesised (GeneWorks, Australia). The double stranded DNA pool was generatedfrom the original synthetic library via a large scale PCR using primers flankingthe randomised sequence, 50-TAA TAC GAC TCA CTA TAG GGA GAC AAG AAT AAACGC TCA A-30 and 50-GCC TGT TGT GAG CCT CCT GTC GAA-30 . A portion of thelarge-scale PCR products (�1014 sequences) was used as a template for in vitro tran-scription to produce the initial 20-fluoropyrimidine modified RNA pool using aDurascribe� T7 Transcription kit (EPICENTRE� Biotechnologies, USA). For SELEX,RNA, at a concentration of 5 lM for initial selection or 1 lM for each iterativerounds, was diluted in 100 lL of wash buffer (Dulbecco’s phosphate buffered salinecontaining 2.5 mM MgCl2) and denatured at 85 �C for 5 min, allowed to cool toroom temperature for 10 min, and annealed at 37 �C for 15 min, before incubationwith the target protein expressed in HEK293T cells for 1 h at 4 �C. Following incu-bation and extensive washes, the bound RNA was reverse transcribed using Super-Script III Reverse Transcriptase (Invitrogen), followed by PCR amplification andin vitro transcription and used for the next round of SELEX. Counter-selection stepswere included from round 4, using a His-tagged irrelevant protein expressed inHEK293T cells, to decrease the enrichment of species specifically recognising theHis-tag, the HEK293T cells or the tissue culture plate. The number of PCR amplifi-cation cycles was also optimised to prevent over-amplification of non-specific ‘‘par-asite’’ PCR products. In addition, the stringency of the selection process wasenhanced to promote the selection of high-affinity aptamers through adjustmentsto aptamer concentration, incubation times, and the number of washes. To acquireaptamers of high specificity, the number of cells used was progressively decreasedwhile the washing stringency increased during the progression of SELEX, with neg-ative selections included from round four. Enrichment was monitored using restric-tion fragment length polymorphism (RFLP) and flow cytometry using live cells.

2.4. RFLP analysis

The enrichment of aptamer candidates during selection was determined byRFLP. Briefly, RFLP was performed as previously described [28,29], with minor mod-ifications. Approximately 5 ng of cDNA from iterative cycles was amplified by PCRfor eight cycles. The amplified DNA was digested with four restriction enzymes,Afa I, Alu I, Hha I and Xsp I that recognise four nucleotides (frequent cutters) in Buf-fer T supplied by the manufacturer (Takara) with 0.1% (w/v) bovine serum albuminat 37 �C overnight. Following the overnight digestion, the DNA was heated to 65 �C,cooled on ice, and separated via electrophoresis on a native 20% polyacrylamide gelin TBE buffer. The gel was then stained in GelStar and visualised using a standardgel imaging system.

2.5. Flow cytometry assays

Cells were harvested at 80% confluence with trypsin digestion and resuspendedin washing buffer (DPBS with 2.5 mM MgCl2) and enumerated. Following centrifu-gation (1000g for 5 min), the cell pellet was resuspended in DMEM with 10% FCSand diluted to 1 � 106/mL. The cells were allowed to re-establish their cell surfacemarkers for a period of 2 h prior to binding analysis. To confirm aptamer binding tothe target protein, RNA from iterative rounds were labelled at the 30-ends with fluo-rescein isothiocyanate (FITC) according to a previously described method [30]. Am-ber tubes were used throughout to minimise photo-bleaching. Briefly, sampleswere oxidised with sodium periodate. The oxidation was terminated with the addi-tion of 10 mM ethylene glycol, followed by ethanol precipitation. FITC was added ata 30-M excess, and the reaction was completed overnight at 4 �C. 1 lM of FITC-la-belled RNA was incubated with trypsinised 5 � 105 HEK293T transfected withCD133 protein or non-transfected HEK293T cells in 100 lL of binding buffer (DPBSwith 2.5 mM MgCl2, 0.1 mg/mL tRNA and 0.1 mg/mL salmon sperm DNA) for 1 h onice, followed by washing three times and resuspension in 300 lL of binding buffer.

86 S. Shigdar et al. / Cancer Letters 330 (2013) 84–95

Fluorescent intensity was determined with a FACS Canto II flow cytometer (BectonDickinson) by counting 50,000 events for each sample. The FITC-labelled RNA fromthe unselected library was used to determine non-specific binding. The binding foreach round was calculated after subtracting the mean fluorescence intensity of thebinding of round zero RNA to target cells as well as that for binding to negative con-trol cells according to a method described by Ellington and colleagues [31].

2.6. Cloning, sequencing and structural analysis of selected aptamers

Following RFLP and flow cytometric analyses of iterative rounds, round sixdemonstrated a sufficient enrichment of RNA sequences that selectively recognisedthe target protein. This enriched pool was amplified by PCR for ten cycles and thePCR products were cloned into the plasmid pCR�4-TOPO� (Invitrogen). PlasmidDNA from individual clones was prepared and their sequence determined usingan automated DNA sequencing procedure. The aptamer sequences were analysedusing ClustalX2 [32]. Secondary structures were predicted using the program RNA-fold [33].

2.7. Determination of aptamer affinity

The dissociation constant (KD) of successful 20-fluoropyrimidine RNA aptamerspecies to native CD133 expressed on the cell surface was determined using flowcytometry. HEK293T cells transfected with CD133 protein, or non-transfectedHEK293T cells (5 � 105) were first incubated with blocking buffer for 1 h on ice(binding buffer containing 0.2% (w/v) sodium azide) followed by two washes withbinding buffer prior to incubation with serial concentrations (approximately 10-fold above and below the apparent KD) of FITC-labelled aptamer in a 100 lL volumeof binding buffer for 1 h on ice. The cells were washed three times with binding buf-fer, resuspended in 150 lL binding buffer and subjected to flow cytometric analy-ses. The FITC-labelled unselected library was used as a negative control. Themean fluorescence intensity (MFI) of the unselected library was subtracted fromthat of the aptamer-target cell to generate the MFI of specific binding. The KD foreach aptamer was determined by Scatchard analysis according to the followingequation:

½Bound aptamer�=½aptamer� ¼ �ð1=KDÞ � ½bound aptamer� þ ð½T�tot=KDÞ

where [T]tot represents the total target concentration.

2.8. Aptamer truncation and determination of specificity

To generate the truncated aptamers, the sense and antisense DNA oligonucleo-tides of desired sequences were synthesised. CD133-A58 (1st Truncation) was de-rived from a sense oligonucleotide, 50-TAA TAC GAC TCA CTA TAG AGA CAA GAATAA ACG CTC AAC CCA CCC TCC TAC ATA GGG AGG AAC GAG TTA CTA TAG-30 ,and antisense oligonucleotide, 50-CTA TAG TAA CTC GTT CCT CCC TAT GTA GGAGGG TGG GTT GAG CGT TTA TTC TTG TCT C-30; CD133-A35 (2nd Truncation) wasderived from a sense oligonucleotide: 50-TAA TAC GAC TCA CTA TAG CTC AACCCA CCC TCC TAC ATA GGG AGG AAC GAG T-30and an antisense oligonucleotide,50-ACT CGT TCC TCC CTA TGT AGG AGG GTG GGT TGA GC-30; CD133-A21 (3rd Trun-cation) was derived from a sense oligonucleotide, 50-TAA TAC GAC TCA CTA TACCAC CCT CCT ACA TAG GGT GG-30 and an antisense oligonucleotide, 50-CCA CCCTAT GTA GGA GGG TGG-30; and CD133-A15 (4th Truncation) was derived from asense oligonucleotide, TAA TAC GAC TCA CTA TAC CCT CCT ACA TAG GG-30 andan antisense oligonucleotide, 50-CCC TAT GTA GGA GGG-30 . CD133-B19 (1st Trunca-tion) was derived from a sense oligonucleotide 50-TAA TAC GAC TCA CTA TAC AGAACG TAT ACT ATT CTG-30 and an antisense oligonucleotide, 50-CAG AAT AGT ATACGT TCT G-30 (T7 RNA promoter sequence is underlined). The relevant pairs of oli-gonucleotides were mixed in equal molar ratios in 1 � PEI buffer (0.1 M Tris–HCl pH8, 0.1 M MgCl2 0.5 M NaCl and 0.1 M dithiothreitol), heated for 5 min at 90 �C andcooled slowly to room temperature prior to ethanol precipitation. In vitro transcrip-tion and FITC-labelling was performed as described above. The final truncations ofthese clones (CD133-A15 and CD133-B19), 50-DY647-CCC UCC UAC AUA GGG-dT-30

and 50-DY647-CAG AAC GUA UAC UAU UCU G-dT-30 , were also chemically synthes-ised with a 50-DY647 fluorescent tag and a 30-inverted deoxythymidine (Dharma-con). The binding affinities of these two aptamers and a negative control aptamerwas determined by flow cytometric analysis using CD133 positive (HT-29 andHep3B) and CD133 negative cell lines (T98G and HEK293T). The blocking stepwas performed at 4 �C using blocking buffer containing 5% (v/v) foetal calf serum,whilst the binding of the aptamers was performed at 37 �C for 30 min. Furthermore,the binding specificity of these aptamers was confirmed with additional CD133-po-sitive (PA-1 and PLC/PRF/5) and -negative cell lines (A549, DU145 and T47D). Apta-mers and antibody were incubated with the cells at 100 nM or 1:11 dilution aspreviously described for 30 min at 37 �C, washed three times and quantified usingthe flow cytometer.

2.9. Confocal microscopy

Twenty-four hours prior to labelling, cells were seeded at a density of 75,000cells per cm2 in an 8-chamber slide (Lab-Tek II, Nunc). DY647-CD133-A15,DY647-CD133-B19 and the control aptamer were prepared in the same manneras for flow cytometry. Following removal of media, cells were incubated in blockingbuffer containing 5% (v/v) foetal calf serum at 37 �C for 15 min, washed twice inbinding buffer prior to incubation with 200 nM aptamer for 30 min at 37 �C. Bis-benzimide Hoechst 33342 (3 lg/mL) (Sigma) was added to the cells during the final15 min of incubation. The aptamer solution was removed and the cells washedthree times for 5 min each in binding buffer prior to visualisation using a FluoViewFV10i laser scanning confocal microscope (Olympus).

2.10. Inhibition of endocytosis

This was performed essentially as described for confocal microscopy with min-or modifications. Briefly, cells were pre-treated with a potassium-depleted buffer(50 mM HEPES, 140 mM NaCl, 2.5 mM MgCl2, and 1 mM CaCl2) for 1 h at 37 �C priorto incubation with the aptamers. These buffers were also used in the incubationstep with aptamers and all rinsing steps. The effectiveness of these treatments ininhibiting endocytosis was evaluated by qualitatively characterising the internali-sation of human transferrin conjugated to Alexa Fluor 488 (Invitrogen Life Technol-ogies). Transferrin (5 lg/mL) was added to the cells following pre-treatmentfollowed by a 30 min incubation at 37 �C. The cells were washed three times in theirrespective buffers and visualised using the FluoView FV10i confocal microscope.

2.11. Tumour sphere preparation and incubation with aptamers and antibody

Two thousand HT29 and HEK293T cells were plated out in ultralow attachmentwells and allowed to form spheres for 7 days in DMEM/F12 media (Invitrogen LifeTechnologies) containing 10 ng/mL epidermal growth factor, 10 ng/mL basic fibro-blast growth factor, 50 lg/mL insulin and 100 units/mL B27. At 7 days, the sphereswere washed three times in PBS containing 2.5 mM MgCl2 and blocked for 20 minusing binding buffer. The spheres were then incubated with 100 nM of aptamer orAC133 antibody (30 lM, Miltenyl Biotec) for 30 min, 60 min, 120 min, or 240 min.Following each time point, the spheres were washed three times with PBS priorto visualisation using the FluoView FV10i confocal microscope. To determine theretention of aptamers within the tumour sphere, HT-29 tumour spheres were incu-bated with CD133 aptamers or CD133 antibody for a total of 4 h, washed threetimes in PBS, followed by incubation in sphere medium for a further 24 h beforebeing imaged.

2.12. Differentiation assay with sodium butyrate

Twenty-four hours prior to differentiation, HT-29 cells were seeded at a densityof 500,000 cells in a 12 well plate. Upon reaching 70% confluence, cells were treatedwith 5 mM sodium butyrate for a total of 24 or 48 h and were then incubated witheither the AC133 antibody, CD133-A15 or CD133-B19 aptamer for 30 min at 37 �C.The AC133 antibody was used at a concentration of 30 lM as recommended by themanufacturer. After washing with PBS, the binding of AC133 antibody or aptamerswas assessed using flow cytometric analysis.

2.13. Western blotting

HT-29 cells were treated with sodium butyrate as described in the previous sec-tion. The expression of CD133 protein in HT-29 cells was analysed by Western anal-ysis as previously described [34]. Fifteen microlitres of each sample was loadedonto a 12% NuPAGE Bis–Tris mini gel (Invitrogen) along with a Precision Plus dualcolour protein standard (BioRad). Following electrophoresis for 45 min at 200 V, theprotein was transferred to a nitrocellulose membrane (Invitrogen) and blocked with5% skimmed milk for 3 h at 25 �C, before being incubated with either anti-b-actin(Sigma) diluted 1:2000, or the anti-CD133 antibody, CD133/1 (clone W6B3C1, Mil-tenyl Biotec) diluted 1:200 in 1% skimmed milk, overnight at 4 �C. Chemilumines-cence was detected using an ImageQuant™ LAS 4000 Biomolecular Imager (GEHealthcare).

3. Results

3.1. Cell SELEX facilitates the selection of aptamers against cell surfacetargets

CD133 is a complex pentaspan protein containing two extracel-lular loops. To effectively select aptamers against only the extracel-lular portion of the protein, it was necessary to devise a procedurethat allowed us to express CD133 so as to preserve its native con-formation. To this end, we sought to transiently express the protein

S. Shigdar et al. / Cancer Letters 330 (2013) 84–95 87

on the surface of HEK293T cells. Using Lipofectamine 2000, the C-terminally His-tagged CD133 was transfected into HEK293T cellsand allowed to express for 72 h prior to SELEX experiments, withexpression confirmed by AC133 antibody staining. Similar to ourprevious SELEX experiments [26], a random RNA library of approx-imately 1 � 1014 species containing 20-fluoro-modified ribose onall pyrimidines was incubated with HEK293T cells expressing hu-man CD133. Unbound RNA was removed via several washing stepsprior to RT-PCR, and the process was repeated for a total of 12rounds. Non-specific binding was eradicated through negativeselection using an irrelevant His-tagged protein transfected intoHEK293T cells. Non-radioactive RFLP was performed to confirmevolution of the species during iterative rounds and confirmationof enrichment was determined using flow cytometry using trans-fected and non-transfected HEK293T cells (Fig. 1A). As shown inFig. 1B, round six showed a greater than 2.5-fold increase in bind-ing to CD133-transfected HEK293T cells as compared to non-transfected HEK293T cells and that of the unselected library.Indeed, following round six, an almost complete loss of bindingspecies was observed. This could be due to a proportion of thePCR amplicons being products of non-specific amplification whichbound to the target via non-specific binding. These amplicons arepreferentially amplified and possibly led to a decrease in the pro-portion of binding species in subsequent rounds [35]. Round sixand round eleven were cloned and sequenced and a high propor-tion of the clones from round eleven displayed truncated sequences.

3.2. Post-SELEX engineering generated the smallest RNA aptamer

DNA from round six was cloned and sequenced and the cloneswere fluorescently tagged with FITC using an in-house method.The binding specificity of each clone was determined usingCD133-negative HEK293T cells and HEK293T cells transfected withthe expression construct of His-tagged CD133. The most encourag-ing results were shown with two aptamers, designated CD133-Aand CD133-B (Fig. 2 A(i) and B(i)). These two clones were sequen-tially truncated to determine the shortest number of bases re-

Fig. 1. Isolation of CD133 aptamers using systematic evolution of ligands by exponentialfrom iterative rounds of SELEX to CD133-transfected HEK293T cells. Fluorescein-labelledby flow cytometric analysis. (B) Comparison of the binding capacity of relative roundfluorescent intensity of the binding of unselected library RNA to target cells, as well asrandom library.

quired to maintain the structure of the binding region of theaptamer (Fig. 2 and Table 1). Clone CD133-A was truncated a totalof four times to confirm the binding region of the aptamer (Fig. 2Aand Table 1). This clone was successfully truncated to 15 nucleo-tides, making it the smallest published RNA aptamer against a can-cer stem cell marker protein, and equivalent in size to the smallestpublished DNA aptamer directed against thrombin [36]. A secondclone, CD133-B, was also investigated for its potential to bind withhigh affinity and specificity to CD133. This aptamer was truncatedto 19 nucleotides (Fig. 2B(ii)), similar in size to our published apt-amer targeting EpCAM [27]. In order for these CD133 aptamers tobecome effective cancer targeting agents, they must have negligi-ble interactions with cells that do not express CD133. Therefore,both CD133-positive (HT-29 and Hep3B) and CD133-negative(T98G and HEK293T) cell lines of human origin were used to studythe sensitivity and specificity of these two aptamers. As shown inFig. 3A and B, determination of the equilibrium dissociation con-stant for both truncated aptamers showed a moderate bindingaffinity with CD133-positive cancer cells (33.85–145 nM) (Table 2),with the smallest aptamer, CD133-A15, having the better bindingaffinity. These results are consistent with results observed fromprevious experiments [27,37]. In addition, both aptamers did notbind to CD133-negative cells (Fig. 3C and D. The specific interac-tion between the CD133 aptamers was further verified usingCD133-positive and -negative cell lines. The cell line HT-29 wasused as a positive control to determine the specificity of our twoaptamers. As shown in Fig. 4, the interaction of our aptamers withthe different cell lines showed a similar staining pattern to that ob-served with the AC133 antibody, thus confirming the specificity ofthe aptamers.

3.3. CD133-specific aptamers are internalised via receptor-mediatedendocytosis

For an aptamer to be developed into an effective cancer therag-nostic, it must be efficiently internalised following binding to itstarget [27,38,39]. We studied whether the newly isolated CD133

enrichment (SELEX). (A) Flow cytometric binding analysis of FITC-labelled aptamersRNA from each round was incubated with target cells at 37 �C for 30 min, followed

s of SELEX. The binding of each round was calculated after subtracting the meanthat for binding to negative control cells. R, round in SELEX cycle; R0, unselected

Fig. 2. Post-selection engineering of CD133 aptamers. (A) CD133-A was serially truncated a total of four times (i: CD133-A; ii: CD133-A58, iii: CD133-A35, iv: CD133-A21, v:CD133-A15); (B) CD133-B (i: was truncated once, ii: CD133-B19).

Table 1Sequences of CD133 aptamers and their truncations.

Aptamer Sequence Basepairs

CD133-A GAG ACA AGA AUA AAC GCU CAA CCC ACC CUCCUA CAU AGG GAG GAA CGA GUU ACU AUA GAGCUU CGA CAG GAG GCU CAC AAC

81

CD133-A58 GAG ACA AGA AUA AAC GCU CAA CCC ACC CUCCUA CAU AGG GAG GAA CGA GUU ACU AUA G

58

CD133-A35 GCU CAA CCC ACC CUC CUA CAU AGG GAG GAACGA GU

35

CD133-A21 CC ACC CUC CUA CAU AGG GUG G 21CD133-A15 CC CUC CUA CAU AGG G 15CD133-B GAG ACA AGA AUA AAC GCU CAA GGA AAG CGC

UUA UUG UUU GCU AUG UUA GAA CGU AUA CUAUUU CGA CAG GAG GCU CAC AAC AGG C

85

CD133-B19 CAG AAC GUA UAC UAU UCU G 19

88 S. Shigdar et al. / Cancer Letters 330 (2013) 84–95

aptamers could internalise upon binding. To this end, we incubatedboth CD133-positive and CD133-negative cells with CD133 apta-mers at 37 �C for 30 min followed by confocal microscopy. Asshown in Fig. 5A CD133 aptamers were efficiently internalisedupon binding. Such internalisation was specific as there was nofluorescent signal in CD133-negative cell lines. Furthermore, weconfirmed that the mode of internalisation of CD133 aptamerswas via receptor-mediated endocytosis as the aptamer fluores-cence was found outside the cells and displayed a ring patternalong the plasma membrane upon the pre-treatment of endocyticblockers, such as potassium-depletion and hypertonic treatments(Fig. 5B). The effectiveness of these treatments in blocking recep-

tor-mediated endocytosis has been previously confirmed usingtransferrin as a positive control [27].

3.4. CD133-specific aptamers are superior in penetrating tumourspheres than CD133 antibodies

To study the effectiveness of our aptamers as cancer theranos-tics, we investigated the potential of our aptamers to penetrate atumour mass using an in vitro 3-dimensional culture, tumoursphere, as a model. We generated tumour sphere models of HT-29 and HEK293T cell lines under ultralow attachment condition.Upon reaching 200 lm in size, these spheres were incubated with100 lM CD133-A15, CD133-B19 or 30 lM AC133 antibody. As pre-sented in Fig. 6A, no tumour penetration was discernable with theAC133 antibody even after 4 h incubation and limited penetrationwas seen with our control RNA aptamer, which is of the same (19nucleotide long) length as CD133-B19 but does not bind to CD133,after 4 h incubation. In sharp contrast, a superior penetration intothe centre of the tumour sphere was seen with both our aptamersfrom 30 min up to 4 h during our assay period. A moderately en-hanced signal was seen from the CD133-B19 aptamer as comparedwith CD133-A15. As well, only a weak signal was discernable withthe HEK293T tumour spheroids which served as a negative controlin this study (Fig. 6B). Importantly, 24 h after washing and furtherincubation in sphere medium, the fluorescence signals for CD133aptamers in the tumour sphere were clearly detectable (Fig. 6C),demonstrating that the CD133 aptamers can not only penetrateinto the core of the tumour sphere but can also be retained bythe tumour cells in the centre of the sphere for at least 24 h.

Fig. 3. Determination of equilibrium dissociation constants (KD) for the interaction of truncated clones of CD133 aptamers with CD133 positive or negative cell lines.Representative binding curves were determined at varying concentrations of CD133 aptamers (1–200 nM) using a cell density of 5 � 105 cells/mL. (A) HT-29 cells; (B) Hep3Bcells; (C) T98G cells; (D) HEK293T cells.

Table 2CD133 dissociation binding constants against CD133-positive (HT-29 and Hep3B) andCD133-negative (T98G and HEK293T) cells.

Cell line CD133-A15 KD (nM ± SEM) CD133-B19 KD (nM ± SEM)

HT-29 83.2 ± 82.9 145.1 ± 75.4Hep3B 33.9 ± 12.5 52.3 ± 15.8HEK293T >1000 >1000T98G >1000 >1000

S. Shigdar et al. / Cancer Letters 330 (2013) 84–95 89

3.5. CD133-specific aptamers recognise both the AC133 epitope andthe CD133 protein

As discussed, the AC133 epitope, instead of the entire CD133protein, has been used as a marker of CSCs. Treatment with sodiumbutyrate has been shown to differentiate colorectal cancer cells,thus decreasing the expression of the AC133 epitope [40,41]. Usingthe colorectal cancer cell line HT-29 as a model system, we studiedwhether our CD133 aptamers could recognise both the AC133 epi-tope and the non-AC133 extracellular segment of CD133 displayedon the cancer cell surface. Following differentiation treatment withsodium butyrate for 24 h, HT-29 cells were analysed by flowcytometry to determine the binding epitope of our aptamers usedat 100 nM concentration. The binding of the AC133 antibody at30 lM to the treated cells was used as a benchmark and showeda 20% decrease in AC133-positive cells following treatment withsodium butyrate for 24 h (Fig. 7A). These results are consistentwith previous reports showing a decrease in the AC133 epitopeduring differentiation [5,7,40–43]. In parallel, the binding of theCD133-A15 aptamer also showed a similar decrease in binding tothe differentiated HT-29 cells (Fig. 7A). Given that the level of bind-ing of both the CD133-A15 aptamer and AC133 antibody decreaseswith differentiation, it is plausible that this aptamer marks thesame population of cells as the AC133 antibody, the CSCs. In con-trast, the binding of the other aptamer, CD133-B19, to the differen-tiated HT-29 cells was increased, indicating that CD133-B19 bindsto the extracellular domain of the CD133 protein, but not theAC133 epitope. Indeed, using another CD133 antibody which rec-

ognises a spatially distinct epitope to AC133 [44], we confirmedthat there was no substantial change of the total CD133 protein le-vel in the HT-29 cells before and after butyrate-induced differenti-ation in a separate Western analysis (Fig. 7B and C).

4. Discussion

Cancer stem cells (CSCs) are considered to be the root of cancerresponsible for cancer recurrence. This model has gained accep-tance because it explains radiation- and chemotherapy-resistance[45], and has led to numerous attempts to specifically target thispopulation of cells within the tumour. While there is not one spe-cific marker which defines all CSCs, a number of markers, includingCD133, CD44, ALDH, EpCAM and ABCG2 [45,46], have proven use-ful for defining the CSC population in solid tumours. CD133 hasbeen implicated as a marker of the CSC population in brain, pros-tate, pancreas, melanoma, colon, liver, lung and ovarian cancers[5], and it has been suggested to be the most important markerof CSCs so far [47,48]. While the function of CD133 is yet to be elu-cidated, this marker is upregulated in hypoxic conditions and hasbeen associated with vasculogenic mimicry in triple negativebreast cancer and prostate cancer [49,50], indicating the impor-tance of CD133 in tumour growth and metastasis. As well, giventhe conformational change that occurs during the differentiationprocess, the AC133 epitope represents a unique target for the treat-ment of various tumour types. Given how critical these CD133+

cells could be to the continuing spread of the tumour, we have iso-lated RNA aptamers against CD133 that are 25–32 times smaller insize than a monoclonal antibody.

We have previously described the success of our SELEX proce-dures to isolate aptamers targeting another CSC marker [27]. Theisolation of aptamers targeting CD133 required a modification toour protocol due to the pentaspan nature of this protein and thenecessity to use proteins in their physiological conformation forselection of aptamers. The success of our selection protocol wasshown using flow cytometric binding assays, as well as RFLP anal-ysis. Successful evolution was shown following the sixth SELEX cy-cle, and several aptamers were cloned. Two aptamers were chosen

Fig. 4. Determination of specificity of CD133 aptamers. CD133-A15 or CD133-B19 were incubated at a concentration of 100 nM with CD133-positive or -negative cell lines for30 min at 37 �C. Cells were also incubated with the AC133 antibody to confirm the presence of AC133 epitope. (A) HT-29 cells incubated with AC133-APC; (B) HT-29 cellsincubated with CD133-A15 and CD133-B19; (C) PA-1 cells incubated with AC133-APC; (D) PA-1 cells incubated with CD133-A15 and CD133-B19; (E) PLC/PCF/5 cellsincubated with AC133-APC; (F) PLC/PRF/5 cells incubated with CD133-A15 and CD133-B19; (G) A549 cells incubated with AC133-APC; (H) A549 cells incubated with CD133-A15 and CD133-B19; (I) DU145 cells incubated with AC133-APC; (J) DU145 cells incubated with CD133-A15 and CD133-B19; (K) T47D cells incubated with AC133-APC; (L)T47D cells incubated with CD133-A15 and CD133-B19. Black: Autofluorescence; Purple: AC133 antibody; Blue: CD133-A15 aptamer; Orange: CD133-B19 aptamer. (Forinterpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

90 S. Shigdar et al. / Cancer Letters 330 (2013) 84–95

for further characterisation using both CD133-positive and -nega-tive cell lines. These aptamers were also truncated to determinethe minimal size required to maintain binding affinity. One ofour aptamers, CD133-A15 was truncated to a size of 15 bases. Thistruncation makes this the smallest RNA aptamer described so far.Both of these aptamers were shown to be sensitive and specific,

and of more importance, these two aptamers were rapidly interna-lised by receptor-mediated endocytosis following binding to theirtarget. This latter feature of the aptamers is a necessary require-ment for these aptamers to be modified as theragnostic reagents.

Aptamers possess many benefits that make them ideal escortmodalities for both treatment and imaging of tumour masses. Their

Fig. 5. CD133 aptamers are endocytosed via receptor mediated endocytosis following binding to CD133-positive cells but not to CD133-negative cells. (A) DY647-labelledCD133 aptamers were incubated with indicated cancer cells for 30 min at 37 �C, followed by imaging using laser scanning confocal microscopy. (B) CD133 positive cells wereincubated with potassium depleted buffer to inhibit endocytosis. Transferrin or DY647-labelled CD133 aptamers were then incubated with HT-29 or Hep3B cells for 30 min at37 �C, followed by imaging using laser scanning confocal microscopy. Transferrin was used as a positive control to confirm inhibition of endocytosis. (C) HT-29 or Hep3B cellswere incubated with transferrin for 30 min at 37 �C without endocytic inhibitors. Data are representative of three independent experiments. For each pair of panels,fluorescent images are on the top, and optical (phase) images are on the bottom. Red: CD133 aptamer; blue: nuclear stain; green: transferrin. Scale bar = 20 lm. (Forinterpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

S. Shigdar et al. / Cancer Letters 330 (2013) 84–95 91

small size means that they are capable of penetrating the tumourmuch more efficiently than conventional immunotherapy options,and these nucleic acids also lack immunogenicity, leading to far

fewer side effects. One of the problems associated with the useof monoclonal antibodies for tumour therapy is the lack of penetra-tion into the tumour mass. We have shown that our aptamers are

Fig. 6. CD133 aptamers show superior tumour penetration. AC133 antibody, CD133 aptamers or control aptamer were incubated with HT-29 or HEK293T tumour spheroidsfor up to 240 min at 37 �C. The tumour spheres were then washed three times in PBS and then either imaged using laser scanning confocal microscopy or incubated for afurther 24 h in PBS. (A) Antibody and aptamer staining of HT-29 cells; (B) Antibody and aptamer staining of HEK293T spheres for up to 240 min. Images are representative of240 min incubation. (C) Following a 4 h incubation with CD133-A15 or CD133-B19, the HT-29 tumour spheroids were washed with PBS and incubated in PBS for 24 h andimaged using laser scanning confocal microscopy. Scale bar = 200 lm.

92 S. Shigdar et al. / Cancer Letters 330 (2013) 84–95

capable of not only penetrating a tumour sphere, but are also re-tained for a minimum of 24 h, suggesting that these aptamerswould act as ideal drug delivery modalities. In contrast, theAC133 antibody was not capable of penetrating the tumour spherecore, even at a 300-fold higher concentration than that for theaptamers. This important property of the aptamers in this studyhighlights the invaluable attribute of aptamers for molecular imag-ing and targeted therapy. Not only are these aptamers internalisedefficiently within a short amount of time by the tumour cells, theyare also capable of penetrating the tumour mass, indicating theirpotential to target all of the cells in the tumour, including thosethat are usually poorly accessible due to their location in the tu-mour core when used as a molecular imaging probe. Through addi-tional functionalisation, either by direct conjugation to achemotherapeutic agent, e.g. doxorubicin or siRNA, or by surfaceconjugation onto nanoparticles containing a drug cargo, aptamerscan function as very effective drug escorts [21]. While some apta-mers can be effective solely by binding to their target, the majorityof aptamers are much more successful as targeting ligands to im-prove the therapeutic index of cytotoxic agents. Therefore, by con-

jugating chemotherapeutic drugs directly to the CD133 aptamers,the drugs would not only be delivered to the tumour cells, butwould also be retained for a sufficient amount of time within thetumour mass for the chemotherapeutic to kill the tumour cells.We are currently investigating the effectiveness of functionalisingour aptamers for future theragnostic applications. Further studieswill be required to confirm that these aptamer-nanoparticle andaptamer-drug constructs possess the same ability as aptamers topenetrate the tumour mass.

It is important to distinguish between the CD133 protein andthe AC133 epitope when discussing the expression of either inrelation to CSCs. AC133 is generally considered to be the markerof CSCs [51–53] though it should be pointed out that there is thesame heterogeneity in CSCs as seen in normal cancer cells [17].While AC133 may be a marker of CSCs, not all CSCs expressAC133. There is also discordance over whether AC133 is truly aCSC marker as studies have shown that both CD133+ and CD133�

cells are capable of propagating tumours in mouse models [7–10]. There are a number of reasons for this: there is a heterogeneityin cell surface markers on CSCs [54,55]; the studies have used an

Fig. 7. CD133 aptamers bind to different epitopes. (A) HT-29 cells were incubated with 5 mM sodium butyrate for 24 h. The cells were then incubated with either the AC133antibody, CD133-A15 or CD133-B19 aptamer for 30 min at 37 �C. The cells were then washed three times prior to flow cytometric analysis. Graphical representation of atleast three independent experiments. Data are shown as means ± SEM. �Statistically significant decrease. p = 0.05; (B) CD133 expression was confirmed by Western blottingusing the W6B3C1 antibody; (C) relative expression of CD133 was compared to b-actin.

S. Shigdar et al. / Cancer Letters 330 (2013) 84–95 93

antibody directed against an epitope other than AC133 to deter-mine CD133 expression [56]; and the AC133 antibody is unableto determine low expression or variants of the AC133 epitope soa ‘negative’ AC133 cell population does not necessarily mean thispopulation of cells are not cancer stem cells [42,44,57]. It shouldbe noted, however, that differentiation studies examining theexpression of CD133 has consistently shown a down-regulationin the expression of the AC133 epitope as compared to CD133[40,41,43,58]. This data indicates that the AC133 is an importantmarker of more primitive cells, such as CSCs. We examined if eitherof our CD133 aptamers were able to distinguish this epitope versusthe CD133 protein using sodium butyrate differentiation treatmentof HT-29 cells. The AC133 antibody was used as a positive controland showed a more than 20% decrease in expression levels follow-ing 24 h treatment with 5 mM sodium butyrate. Interestingly, apt-amer CD133-A15 also showed a similar decrease in cell binding,indicating that this aptamer marks the same population of cellsas identified by the AC133 epitope. An additional antibody usedto confirm the expression of CD133 did not change following dif-ferentiation treatment. This antibody recognises the same epitopeas AC133, and no significant decrease was observed followingtreatment as determined by Western blot analysis. This has previ-ously been reported to show that the AC133 epitope is maskedupon differentiation but is detectable by Western blot indicatingthe cells still retain their CD133 protein. As well, no decreasewas seen in the percentage of positive cells recognised by aptamerCD133-B19, indicating that this aptamer not only recognises a dif-ferent epitope to CD133-A15, but is also capable of identifying dif-ferentiated cancer cells. The use of both of these aptamers togetherhave the potential to target the bulk cancer cells and CSCs in a tu-mour, thus reducing the bulk of the disease, as well as reducing themetastatic potential at the same time. Indeed, it is becoming wellrecognised that this is the only way we will successfully eradicatecancer in the future [59].

An important aspect of aptamers in cancer research is their lackof immunogenicity, a problem often encountered when antibodiesare used for therapeutics or for flow cytometric cell sorting prior toxenotransplantation. A recent study demonstrated that the Fc-fragment of antibodies can cause a profound inhibitory effecton engraftment of tumour cells in xenograft models due to

antibody-mediated clearance [60]. This study on antibody-coatedcells flow cytometrically sorted based on cell surface markers sug-gests that the proportion of CSCs could be highly under-estimatedin xenotransplant studies using AC133 antibody-sorted cells. Inaddition, a further study has suggested that by stimulating the im-mune system, and thus targeting and attacking the bulk cancercells, a reduction in the size of the tumour results in an increasein the proportion of CSCs [61]. These results taken together, alongwith prior knowledge of the effect that the majority of therapeuticantibodies have on the immune system, would indicate thatmodalities other than antibodies are required both for researchinto CSCs and also for the complete eradication of cancer in clinicalstudies. Aptamers, which are identical to monoclonal antibodies intheir ability to bind to their targets with a high degree of selectiv-ity, are superior to antibodies due to the lack of the Fc fragmentand thus devoid of an immune response when injected intomammals. This attribute alone makes aptamers the ideal choicefor cancer therapeutics, in addition to be a valuable tool in cellsorting for the isolation of CSCs. The ability to functionalisethese nucleic acids without loss-of-function is an added benefit.As well, these aptamers do not suffer from the disadvantagesassociated with other nucleic acid species, such as RNAi or anti-sense oligonucleotides [21] due to their inability of entering thetarget cells, as we select aptamers that are internalised uponbinding.

The targeting of CD133+ cells in certain tumour types wouldseem to be a highly effective strategy to assist with the eradicationof cancer [62]. As CD133+ cells have been implicated in vasculogen-ic mimicry, this could explain the lack of efficacy of anti-angiogenictreatments, given that removing the oxygen supply from the tu-mour increases the number of CD133+ cells. It is possible that bycombining the targeting of CD133+ cells continually alongsideanti-angiogenic therapy may increase the efficacy of these treat-ments, though care must be taken that this does not increase theinvasiveness of the tumour. Further studies are required to deter-mine the effectiveness of targeting CD133 cells alone, or in combi-nation with other therapeutic options using the novel CD133 RNAaptamers.

In summary, we have isolated the smallest RNA aptamer againstCD133. Our AC133-targeting aptamers represent the ideal

94 S. Shigdar et al. / Cancer Letters 330 (2013) 84–95

molecules for the development of CSC targeted theranostics with-out the sides effects associated with immunogenecity.

Acknowledgements

This work was supported by the Australia–India Strategic Re-search Fund Grant ST010013 and Victorian Cancer Agency GrantPTCP-02 to WD. Dr. L. Qiao is supported by the Career Develop-ment and Support Fellowship Future Research Leader Grant ofthe NSW Cancer Institute, Australia (Grant ID: 08/FRL/1-04).

References

[1] D. Mizrak, M. Brittan, M.R. Alison, CD133: molecule of the moment, J. Pathol.214 (2008) 3–9.

[2] J. Jaszai, C.A. Fargeas, M. Florek, W.B. Huttner, D. Corbeil, Focus on molecules:prominin-1 (CD133), Exp. Eye Res. 85 (2007) 585–586.

[3] I.V. Ulasov, S. Nandi, M. Dey, A.M. Sonabend, M.S. Lesniak, Inhibition of Sonichedgehog and Notch pathways enhances sensitivity of CD133(+) glioma stemcells to temozolomide therapy, Mol. Med. 17 (2011) 103–112.

[4] S.K. Singh, C. Hawkins, I.D. Clarke, J.A. Squire, J. Bayani, T. Hide, R.M.Henkelman, M.D. Cusimano, P.B. Dirks, Identification of human brain tumourinitiating cells, Nature 432 (2004) 396–401.

[5] K. Kemper, M.R. Sprick, M. de Bree, A. Scopelliti, L. Vermeulen, M. Hoek, J.Zeilstra, S.T. Pals, H. Mehmet, G. Stassi, J.P. Medema, The AC133 epitope, butnot the CD133 protein, is lost upon cancer stem cell differentiation, Cancer Res.70 (2010) 719–729.

[6] C. Dittfeld, A. Dietrich, S. Peickert, S. Hering, M. Baumann, M. Grade, T. Ried, L.A.Kunz-Schughart, CD133 expression is not selective for tumor-initiating orradioresistant cell populations in the CRC cell lines HCT-116, Radiother. Oncol.92 (2009) 353–361.

[7] P. Grosse-Gehling, C.A. Fargeas, C. Dittfeld, Y. Garbe, M.R. Alison, D. Corbeil, L.A.Kunz-Schughart, CD133 as a biomarker for putative cancer stem cells in solidtumours: limitations, problems and challenges, J. Pathol. (2012).

[8] D. Beier, P. Hau, M. Proescholdt, A. Lohmeier, J. Wischhusen, P.J. Oefner, L.Aigner, A. Brawanski, U. Bogdahn, C.P. Beier, CD133(+) and CD133(-)glioblastoma-derived cancer stem cells show differential growthcharacteristics and molecular profiles, Cancer Res. 67 (2007) 4010–4015.

[9] J. Wang, P.O. Sakariassen, O. Tsinkalovsky, H. Immervoll, S.O. Boe, A. Svendsen,L. Prestegarden, G. Rosland, F. Thorsen, L. Stuhr, A. Molven, R. Bjerkvig, P.O.Enger, CD133 negative glioma cells form tumors in nude rats and give rise toCD133 positive cells, Int. J. Cancer 122 (2008) 761–768.

[10] X. Zheng, G. Shen, X. Yang, W. Liu, Most C6 cells are cancer stem cells: evidencefrom clonal and population analyses, Cancer Res. 67 (2007) 3691–3697.

[11] S.V. Shmelkov, J.M. Butler, A.T. Hooper, A. Hormigo, J. Kushner, T. Milde, R. StClair, M. Baljevic, I. White, D.K. Jin, A. Chadburn, A.J. Murphy, D.M. Valenzuela,N.W. Gale, G. Thurston, G.D. Yancopoulos, M. D’Angelica, N. Kemeny, D. Lyden,S. Rafii, CD133 expression is not restricted to stem cells, and both CD133+ andCD133� metastatic colon cancer cells initiate tumors, J. Clin. Invest. 118(2008) 2111–2120.

[12] A.B. Mak, K.M. Blakely, R.A. Williams, P.A. Penttila, A.I. Shukalyuk, K.T. Osman,D. Kasimer, T. Ketela, J. Moffat, CD133 protein N-glycosylation processingcontributes to cell surface recognition of the primitive cell marker AC133epitope, J. Biol. Chem. 286 (2011) 41046–41056.

[13] A.H. Yin, S. Miraglia, E.D. Zanjani, G. Almeida-Porada, M. Ogawa, A.G. Leary, J.Olweus, J. Kearney, D.W. Buck. AC133, a Novel Marker for HumanHematopoietic Stem and Progenitor Cells. Blood 90 (1997) 5002–5012.

[14] N.N. Waldron, D.S. Kaufman, S. Oh, Z. Inde, M.K. Hexum, J.R. Ohlfest, D.A.Vallera, Targeting tumor-initiating cancer cells with dCD133KDEL showsimpressive tumor reductions in a xenotransplant model of human head andneck cancer, Mol. Cancer Ther. 10 (2011) 1829–1838.

[15] D. Horst, L. Kriegl, J. Engel, T. Kirchner, A. Jung, CD133 expression is anindependent prognostic marker for low survival in colorectal cancer, Br. J.Cancer 99 (2008) 1285–1289.

[16] G. Ferrandina, M. Petrillo, G. Bonanno, G. Scambia, Targeting CD133 antigen incancer, Expert Opin. Ther. Targets 13 (2009) 823–837.

[17] S. Bidlingmaier, X. Zhu, B. Liu, The utility and limitations of glycosylatedhuman CD133 epitopes in defining cancer stem cells, J. Mol. Med. (Berl) 86(2008) 1025–1032.

[18] L.M. Weiner, J.C. Murray, C.W. Shuptrine, Antibody-based immunotherapy ofcancer, Cell 148 (2012) 1081–1084.

[19] I. Mellman, G. Coukos, G. Dranoff, Cancer immunotherapy comes of age,Nature 480 (2011) 480–489.

[20] T.T. Hansel, H. Kropshofer, T. Singer, J.A. Mitchell, A.J. George, The safety andside effects of monoclonal antibodies, Nat. Rev. Drug Discov. 9 (2010) 325–338.

[21] S. Shigdar, A.C. Ward, A. De, C.J. Yang, M. Wei, W. Duan, Clinical applications ofaptamers and nucleic acid therapeutics in haematological malignancies, Br. J.Haematol. 155 (2011) 3–13.

[22] A.D. Ellington, J.W. Szostak, In vitro selection of RNA molecules that bindspecific ligands, Nature 346 (1990) 818–822.

[23] C. Tuerk, L. Gold, Systematic evolution of ligands by exponentialenrichment: RNA ligands to bacteriophage T4 DNA polymerase, Science 249(1990) 505–510.

[24] L. Gryziewicz, Regulatory aspects of drug approval for macular degeneration,Adv. Drug Deliv. Rev. 57 (2005) 2092–2098.

[25] L. Meng, L. Yang, X. Zhao, L. Zhang, H. Zhu, C. Liu, W. Tan, Targeted delivery ofchemotherapy agents using a liver cancer-specific aptamer, PLoS ONE 7 (2012)e33434.

[26] H. Xing, N.Y. Wong, Y. Xiang, Y. Lu, DNA aptamer functionalized nanomaterialsfor intracellular analysis, cancer cell imaging and drug delivery, Curr. Opin.Chem. Biol. (2012).

[27] S. Shigdar, J. Lin, Y. Yu, M. Pastuovic, M. Wei, W. Duan, RNA aptamer against acancer stem cell marker epithelial cell adhesion molecule, Cancer Sci. 102(2011) 991–998.

[28] M. Das, C. Mohanty, S.K. Sahoo, Ligand-based targeted therapy for cancertissue, Expert Opin. Drug Deliv. 6 (2009) 285–304.

[29] S. Missailidis, A. Hardy, Aptamers as inhibitors of target proteins, Expert Opin.Ther. Pat. 19 (2009) 1073–1082.

[30] D. Willkomm, R. Hartmann, Handbook of RNA Biochemistry, in: R.K.Hartmann, A. Schön, E. Westhof (Eds.), Wiley-VCH GmbH & Co. KGaA,Weinheim, 2005, pp. 86–94.

[31] N. Li, J.N. Ebright, G.M. Stovall, X. Chen, H.H. Nguyen, A. Singh, A. Syrett, A.D.Ellington, Technical and biological issues relevant to cell typing with aptamers,J. Proteome Res. 8 (2009) 2438–2448.

[32] M.A. Larkin, G. Blackshields, N.P. Brown, R. Chenna, P.A. McGettigan, H.McWilliam, F. Valentin, I.M. Wallace, A. Wilm, R. Lopez, J.D. Thompson, T.J.Gibson, D.G. Higgins, Clustal W and Clustal X version 2.0, Bioinformatics 23(2007) 2947–2948.

[33] I.L. Hofacker, Vienna RNA secondary structure server, Nucleic Acids Res. 31(2003) 3429–3431.

[34] S.S. Yeong, Y. Zhu, D. Smith, C. Verma, W.G. Lim, B.J. Tan, Q.T. Li, N.S. Cheung,M. Cai, Y.Z. Zhu, S.F. Zhou, S.L. Tan, W. Duan, The last 10 amino acid residuesbeyond the hydrophobic motif are critical for the catalytic competence andfunction of protein kinase calpha, J. Biol. Chem. 281 (2006) 30768–30781.

[35] Z. Tang, P. Parekh, P. Turner, R.W. Moyer, W. Tan, Generating aptamers forrecognition of virus-infected cells, Clin. Chem. 55 (2009) 813–822.

[36] L.R. Paborsky, S.N. McCurdy, L.C. Griffin, J.J. Toole, L.L. Leung, The single-stranded DNA aptamer-binding site of human thrombin, J. Biol. Chem. 268(1993) 20808–20811.

[37] A. Wochner, M. Menger, M. Rimmele, Characterisation of aptamers fortherapeutic studies, Expert Opin. Drug Discov. 2 (2007) 1205–1224.

[38] U. Zangemeister-Wittke, Antibodies for targeted cancer therapy – technicalaspects and clinical perspectives, Pathobiology 72 (2005) 279–286.

[39] S. Hussain, A. Pluckthun, T.M. Allen, U. Zangemeister-Wittke, Antitumoractivity of an epithelial cell adhesion molecule targeted nanovesicular drugdelivery system, Mol. Cancer Ther. 6 (2007) 3019–3027.

[40] N. Haraguchi, M. Ohkuma, H. Sakashita, S. Matsuzaki, F. Tanaka, K. Mimori, Y.Kamohara, H. Inoue, M. Mori, CD133+CD44+ population efficiently enrichescolon cancer initiating cells, Ann. Surg. Oncol. 15 (2008) 2927–2933.

[41] A.Sgambato,M.A.Puglisi,F.Errico,F.Rafanelli,A.Boninsegna,A.Rettino,G.Genovese,C. Coco, A. Gasbarrini, A. Cittadini, Post-translational modulation of CD133expression during sodium butyrate-induced differentiation of HT29 human coloncancer cells: implications for its detection, J. Cell. Physiol. 224 (2010) 234–241.

[42] T.L. Osmond, K.W. Broadley, M.J. McConnell, Glioblastoma cells negative forthe anti-CD133 antibody AC133 express a truncated variant of the CD133protein, Int. J. Mol. Med. 25 (2010) 883–888.

[43] H.L. Feng, Y.Q. Liu, L.J. Yang, X.C. Bian, Z.L. Yang, B. Gu, H. Zhang, C.J. Wang, X.L.Su, X.M. Zhao, Expression of CD133 correlates with differentiation of humancolon cancer cells, Cancer Biol. Ther. 9 (2010) 216–223.

[44] F. Gambelli, F. Sasdelli, I. Manini, C. Gambarana, G. Oliveri, C. Miracco, V.Sorrentino, Identification of cancer stem cells from human glioblastomas:growth and differentiation capabilities and CD133/prominin-1 expression, CellBiol. Int. 36 (2012) 29–38.

[45] J.E. Visvader, G.J. Lindeman, Cancer stem cells: current status and evolvingcomplexities, Cell Stem Cell 10 (2012) 717–728.

[46] Y. Hu, L. Fu, Targeting cancer stem cells: a new therapy to cure cancer patients,Am. J. Cancer Res. 2 (2012) 340–356.

[47] S.K. Singh, I.D. Clarke, M. Terasaki, V.E. Bonn, C. Hawkins, J. Squire, P.B. Dirks,Identification of a cancer stem cell in human brain tumors, Cancer Res. 63(2003) 5821–5828.

[48] G. Rappa, O. Fodstad, A. Lorico, The stem cell-associated antigen CD133(prominin-1) is a molecular therapeutic target for metastatic melanoma, StemCells 26 (2008) 3008–3017.

[49] T.J. Liu, B.C. Sun, X.L. Zhao, X.M. Zhao, T. Sun, Q. Gu, Z. Yao, X.Y. Dong, N. Zhao,N. Liu, CD133(+) cells with cancer stem cell characteristics associates withvasculogenic mimicry in triple-negative breast cancer, Oncogene (2012).

[50] R. Liu, K. Yang, C. Meng, Z. Zhang, Y. Xu, Vasculogenic mimicry is a marker ofpoor prognosis in prostate cancer, Cancer Biol. Ther. 13 (2012).

[51] F. Crea, L. Fornaro, G. Masi, A. Falcone, R. Danesi, W. Farrar, Faithful markers ofcirculating cancer stem cells: is CD133 sufficient for validation in clinics?, JClin. Oncol. 29 (2011) 3487–3488 (author reply 3488–90).

[52] B. Campos, C.C. Herold-Mende, Insight into the complex regulation of CD133 inglioma, Int. J. Cancer 128 (2011) 501–510.

[53] G. Bertolini, L. Roz, P. Perego, M. Tortoreto, E. Fontanella, L. Gatti, G. Pratesi, A.Fabbri, F. Andriani, S. Tinelli, E. Roz, R. Caserini, S. Lo Vullo, T. Camerini, L.Mariani, D. Delia, E. Calabro, U. Pastorino, G. Sozzi, Highly tumorigenic lung

S. Shigdar et al. / Cancer Letters 330 (2013) 84–95 95

cancer CD133+ cells display stem-like features and are spared by cisplatintreatment, Proc. Natl. Acad. Sci. USA 106 (2009) 16281–16286.

[54] R. Castriconi, A. Dondero, F. Negri, F. Bellora, P. Nozza, B. Carnemolla, A. Raso, L.Moretta, A. Moretta, C. Bottino, Both CD133+ and CD133� medulloblastomacell lines express ligands for triggering NK receptors and are susceptible to NK-mediated cytotoxicity, Eur. J. Immunol. 37 (2007) 3190–3196.

[55] Z. Zhu, X. Hao, M. Yan, M. Yao, C. Ge, J. Gu, J. Li, Cancer stem/progenitor cellsare highly enriched in CD133+CD44+ population in hepatocellular carcinoma,Int. J. Cancer 126 (2010) 2067–2078.

[56] M.J. Pfeiffer, J.A. Schalken, Stem cell characteristics in prostate cancer cell lines,Eur. Urol. 57 (2010) 246–254.

[57] Y. Wu, P.Y. Wu, CD133 as a marker for cancer stem cells: progresses andconcerns, Stem Cells Dev. 18 (2009) 1127–1134.

[58] M.A. Puglisi, M. Barba, M. Corbi, M.F. Errico, E. Giorda, N. Saulnier, A.Boninsegna, A.C. Piscaglia, R. Carsetti, A. Cittadini, A. Gasbarrini, A.

Sgambato, Identification of endothelin-1 and NR4A2 as CD133-regulatedgenes in colon cancer cells, J. Pathol. 225 (2011) 305–314.

[59] C. Scheel, R.A. Weinberg, Cancer stem cells and epithelial-mesenchymaltransition: Concepts and molecular links, Semin. Cancer Biol. (2012).

[60] D.C. Taussig, F. Miraki-Moud, F. Anjos-Afonso, D.J. Pearce, K. Allen, C. Ridler, D.Lillington, H. Oakervee, J. Cavenagh, S.G. Agrawal, T.A. Lister, J.G. Gribben, D.Bonnet, Anti-CD38 antibody-mediated clearance of human repopulating cellsmasks the heterogeneity of leukemia-initiating cells, Blood 112 (2008) 568–575.

[61] H. Enderling, L. Hlatky, P. Hahnfeldt, Immunoediting: evidence of themultifaceted role of the immune system in self-metastatic tumor growth,Theor. Biol. Med. Model. 9 (2012) 31.

[62] G. Liu, X. Yuan, Z. Zeng, P. Tunici, H. Ng, I.R. Abdulkadir, L. Lu, D. Irvin, K.L.Black, J.S. Yu, Analysis of gene expression and chemoresistance of CD133+cancer stem cells in glioblastoma, Mol. Cancer 5 (2006) 67.