RNA-Based TWIST1 Inhibition via Dendrimer Complex to Reduce Breast Cancer Cell Metastasis

Research ArticleRNA-Based TWIST1 Inhibition via Dendrimer Complex toReduce Breast Cancer Cell Metastasis

James Finlay123 Cai M Roberts12 Gina Lowe1 Joana Loeza4

John J Rossi2 and Carlotta A Glackin1

1Department of Neurosciences City of Hope Beckman Research Institute 1500 East Duarte Road Duarte CA 91010 USA2Irell amp Manella Graduate School of Biological Sciences City of Hope Beckman Research Institute 1500 East Duarte RoadDuarte CA 91010 USA3Division of Comparative Medicine City of Hope Beckman Research Institute 1500 East Duarte Road Duarte CA 91010 USA4Department of Biological Sciences California State University Los Angeles 5151 State University Drive Los AngelesCA 90032-8201 USA

Correspondence should be addressed to Carlotta A Glackin cglackincohorg

Received 20 June 2014 Revised 31 December 2014 Accepted 31 December 2014

Academic Editor Eric W-F Lam

Copyright copy 2015 James Finlay et al This is an open access article distributed under the Creative Commons Attribution Licensewhich permits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Breast cancer is the leading cause of cancer-related deaths among women in the United States and survival rates are lower forpatients with metastases andor triple-negative breast cancer (TNBC ER PR and Her2 negative) Understanding the mechanismsof cancer metastasis is therefore crucial to identify new therapeutic targets and develop novel treatments to improve patientoutcomes A potential target is the TWIST1 transcription factor which is often overexpressed in aggressive breast cancers and is amaster regulator of cellular migration through epithelial-mesenchymal transition (EMT) Here we demonstrate an siRNA-basedTWIST1 silencing approach with delivery using a modified poly(amidoamine) (PAMAM) dendrimer Our results demonstratethat SUM1315 TNBC cells efficiently take up PAMAM-siRNA complexes leading to significant knockdown of TWIST1 and EMT-related target genes Knockdown lasts up to one week after transfection and leads to a reduction in migration and invasion asdetermined by wound healing and transwell assays Furthermore we demonstrate that PAMAM dendrimers can deliver siRNAto xenograft orthotopic tumors and siRNA remains in the tumor for at least four hours after treatment These results suggest thatfurther development of dendrimer-based delivery of siRNA for TWIST1 silencing may lead to a valuable adjunctive therapy forpatients with TNBC

1 Introduction

Breast cancer is the leading cause of cancer-related deathsamong women in the United States with over 235000 newdiagnoses and roughly 40000 deaths expected in 2014 [1]Nonmetastatic breast cancer is relatively well-managed withchemotherapy radiation and surgery However metastaticbreast cancer (MBC) that has spread to the liver bonebrain and lungs is frequently incurable [2 3] Triple-negativebreast cancer (TNBC) (ER-negative progesterone receptor(PR) negative and HER2-negative) is of particular interestbecause it is aggressive and metastatic and does not respondto current therapies Understanding the metastatic mecha-nisms of the aberrant cancer cells that allow them to spread to

distant sites in the body and develop intometastatic tumors istherefore crucial in order to identify new therapeutic targetsand develop novel treatments that can be used in conjunctionwith current therapies to prevent metastases and improvepatient outcomes

A key mechanism for the spread of cancer cells isepithelial-mesenchymal transition (EMT)During EMT can-cer cells undergo changes enabling them to detach fromthe primary tumor and invade into surrounding tissuesthe lymphatic system and blood vessels [4ndash7] EMT alsoallows mobile cancer cells to migrate out of blood vesselsand into distant organs The TWIST1 transcription factoractivates EMT in cancer cells [8 9] and activates severaltarget genes that promote cellular dedifferentiation and cell

Hindawi Publishing CorporationBioMed Research InternationalVolume 2015 Article ID 382745 12 pageshttpdxdoiorg1011552015382745

2 BioMed Research International

+

Nuc

NucNucNuc

TWIST1knockdowndownstream

gene expressionchanges

reversal of EMT

1 2

3 4 5 6

Figure 1 Mechanism of dendrimer-mediated siRNA delivery and TWIST1 knockdown (1) Negatively charged siRNA is electrostaticallyattracted to positive charges on the YTZ3-15 dendrimer leading to the formation of 6ndash8 nm diameter micelles coated with siRNA (2) Thesedendriplexes are administered to tumor cells (3) Dendriplexes are taken up via macropinocytosis (4) Dendriplexes are trafficked to lateendosomes (5) Due to the proton sponge effect electrostatic interactions between the dendrimer and siRNA are disrupted and siRNA escapesfrom the disrupted endosome into the cytosol (6) Once in the cytosol siRNA recruits the endogenous RNAi machinery to degrade TWIST1mRNA Following TWIST1 knockdown TWIST1 target gene expression is altered to reduce invasive capacity

mobility In addition to promoting EMT in cancer cellsTWIST1 is thought to promote the cancer stem cell (CSC)phenotype [10] inhibit apoptosis [11 12] and contributeto chemotherapy resistance [13 14] TWIST1 has also beenshown to be overexpressed in numerous solid tumors [1115ndash19] including aggressive and metastatic forms of breastcancer [20ndash22] however it is not expressed in normal adulttissues This expression profile coupled with the establishedrole of TWIST1 in numerousmetastasis-promoting pathwayssuggests it is a promising novel target for MBC therapy [23]

The therapeutic use of small interfering RNA (siRNA)for cancer has gained considerable interest since its genesilencing properties were first described [24ndash26] Once inthe cytoplasm the siRNA unwinds and associates with Argo-naute2 forming an RNA-induced silencing complex (RISC)which leads to sequence-specific mRNA degradation andgene silencing [27] However while promising the develop-ment of siRNA therapy has encountered challenges includingsusceptibility to enzymatic degradation delivery to targettissues endosomal escape immune activation and off-targeteffects [28ndash31] Effective siRNA delivery (both to the tissueof interest and across the cell membrane) remains one ofthe main barriers to developing clinically relevant therapies[32 33]

The success of an siRNA-based gene silencing therapeuticapproach requires that the siRNA enters the cytoplasmwithout being degraded [34 35] Recent studies have demon-strated that poly(amidoamine) (PAMAM) dendrimers arecapable of functional siRNA delivery both in vitro and

in vivo by protecting siRNA (via electrostatic interactionsand aggregation) from enzymatic degradation prior tomacropinocytosis and eventual release for the endosome(Figure 1) [36 37] Recently a modified third generationamphiphilic PAMAM dendrimer (YTZ3-15) was shown toeffectively deliver siRNA and cause gene knockdown invivo via intratumoral (IT) administration [38] ComplexingYTZ3-15 with TWIST1 siRNAmay therefore have the poten-tial to allow delivery of potent siRNAs to breast tumor cellsto reduce TWIST1-mediated expression of EMT target genesand inhibit metastatic potential

In the current study we investigated whether anti-TWIST1 siRNA could be functionally delivered to metastaticbreast cancer cells (SUM 1315 cell line) using YTZ3-15 Wetested the ability of the YTZ3-15-delivered siRNA to knockdown TWIST1 reduce expression of EMT-related targetgenes and alter the phenotypic characteristics associatedwith cancer cell migrationinvasion We also evaluated thetumor-specific delivery capability of YTZ3-15 in vivo using amouse breast cancer model

2 Materials and Methods

21 Cell Culture Transfection and Stable Cell Line ProductionSUM 1315 breast cancer cells were obtained from ATCC(Manassas VA) Cells were maintained at 37∘C 5 CO

2 and

90 humidity in a tissue culture incubator Media for SUM1315 cells consisted of a 50-50 mixture of DMEM and F12media supplemented with 5120583gmL insulin 10 ngmL EGF

BioMed Research International 3

5 fetal bovine serum (FBS) and 1 penicillinstreptomycinCells were passaged using 025 trypsin (Genesee Sci-entific San Diego CA) every 2-3 days as they reachedconfluency Transient transfection of SUM 1315 cells wasperformed using Lipofectamine 2000 (Thermo Fisher Sci-entific Waltham MA) or YTZ3-15 (obtained from Dr LingPeng Centre Interdisciplinaire de Nanoscience de MarseilleFrance) siRNA sequenceswere as follows siTwistA-sense 51015840-GGACAAGCUGAGCAAGAUU-31015840 siTwistA-antisense 51015840-AAUCUUGCUCAGCUUGUCCUU-31015840 siTwistB-sense 51015840-GCGACGAGCUGGACUCCAA-31015840 siTwistB-antisense 51015840-UUGGAGUCCAGCUCGUCGCUU-31015840 All custom siRNAswere synthesized by IDT (Integrated DNA TechnologiesInc Skokie Illinois) and arrived lyophilized and were resus-pended in water prior to being reannealed Two nega-tive control siRNAs were used siQ (labeled with eitherAlexaFluor 488 or 647) was AllStars Negative ControlsiRNA from Qiagen (Valencia CA) and siCtrl was anonlabeled control siRNA with the following sequencesiCtrl-sense 51015840-ACUCCAAGAUGGCAAGCUG-31015840 siCtrl-antisense 51015840-CAGCUUGCCAUCUUGGAGU-31015840 Lipofec-tamine 2000 was diluted fiftyfold in OptiMEM (ThermoFisher Scientific Waltham MA) and incubated with 10120583Lof 10 120583M siRNA for 20min at room temperature YTZ3-15 was diluted to 175 120583M in Opti-MEM and mixed with10 120583L of 10 120583M siRNA at an NP ratio of 5 for a finalsiRNA concentration of 50 nM when applied to cells in 2mLof media per well in a 6-well plate Dendrimer complexes(dendriplexes) were incubated 20min at room temperaturewhich resulted in dendriplex aggregates of roughly 100ndash200 nm in size (Figure 1) [39] Incubation of the YTZ3-15dendriplexes with SUM 1315 cells was done at the tissueculture conditions described previously for up to 7 days withfresh media being added to existing media as needed

Stable transfections of SUM 1315 cells were performedusing lentivirus To examine the effects of TWIST1 knock-down in SUM 1315 cells without the possible confound-ing variables of the delivery mechanism itself we devel-oped cell lines that stably expressed short hairpin RNAs(shRNAs) against TWIST1 (shTwistA shTwistB) or a scram-bled control shRNA (shScram) as a negative controlThe mRNA target sequences for shTwistA and shTwistBare identical to the sense sequences of siTwistA andsiTwistB (mentioned in the previous paragraph) respec-tively The mRNA ldquotargetrdquo sequence for shScram was 51015840-UUCUCCGAACGUGUCACGU-31015840 Cells were either stablytransfected with shTwistA shTwistB or shScram as describedpreviously [40] Cells expressing eGFP-firefly luciferasefusion protein (eGFP + luc) were created by transfectingthe CMV cassette as described previously [41] Immortalizedhuman mesenchymal stem cells (hMSCs) were used for invivo experiments as described previously [42]

The YTZ3-15 dendrimer has been described in detailpreviously [38 39] but will be briefly summarized hereThe chemical formula for each YTZ3-15 dendrimer isC125

H247

N37O20

with a molecular weight of 25869448 dal-tons They are formed by click chemistry and consist of twolipid tails at one end and a dendron with eight terminalamines on the opposite end The dendrimer was purified

by column chromatography on silica gel with PetroleumetherEtOAc [38]These dendrimers spontaneously aggregateto form micelles ranging in size from 100 nm to 200 nmWhen in the presence of siRNA these dendrimers rearrangeinto smaller (6ndash8 nm) substructures as part of the largermicelles in order to allow for more electrostatic interactionsbetween the negatively charged siRNA and the positivelycharged amines on the dendrimer (Figure 1) [39]

22 Wound Healing Assay To examine directional cellmigration in vitro wound healing assays were performedas described previously [43] In summary SUM 1315 cells(parental eGFP + luc shScram shTwistA and shTwistB)were grown in the conditions described above in Section 21in 6-well tissue culture plates Once cells reached 80confluency a sterile 200 120583L pipette tip was used to scratch aline in themonolayer of cells Images were taken immediatelyafter the scratch and at several time points thereafter using aNikon TE-2000S microscope and SPOT Advanced software(Diagnostic Instruments Sterling Heights MI) Care wastaken to always capture images in the same location for eachtime point Additionally SUM 1315 cells were transfectedwith siQ- siTwistA- or siTwistB-YTZ3-15 dendriplexes at a50 nM final siRNA concentration (175 120583M YTZ3-15) Cellswere incubated with dendriplexes for 24 hours at 37∘C 5CO2 and 90 humidity in a tissue culture incubator the

plates were then scratched with the 200120583L pipette tip andimaged as described above

23 Invasion Assay 25 times 105 SUM 1315 cells were transfectedusing YTZ3-15 complexed with siQ siTwistA or siTwistBas described above in Section 22 After 24 hours incuba-tion 25 times 105 SUM 1315 cells were lifted from the platewith 025 trypsin washed with PBS and seeded ontotranswell inserts (8120583m pore diameter Millipore DarmstadtGermany) Inserts (which nest inside wells in a 24-welltissue culture plate) were precoated with Matrigel (3mgmL60 120583L diluted with serum-free medium) (BD BiosciencesSan Jose CA) which was allowed to solidify in a tissueculture incubator for 30 minutes prior to the addition of thetransfected cells In order to stimulate cell invasion the topchamber containing the transfected SUM 1315 cells contained1 FBS (400 120583L) while the lower chamber contained 600 120583Lof complete media with 20 FBS Cells were permitted toinvade for 24 hours in a tissue culture incubator After thisperiod the Matrigel and any remaining cells in the upperchamber were removed with a cotton-tipped swab Transwellmembranes were then washed twice with PBS and stainedwith Crystal Violet Five images of each membrane weretaken and cells were counted manually

24 Quantitative PCR Total cellular RNA was isolated usingthe RNeasy Plus kit (Qiagen Valencia CA) RNA quantityand quality were measured and analyzed (with 260280 nmand 260230 nm spectra measurements) using a NanoDropND-1000 and its associated software (Thermo Fisher Scien-tific Waltham MA) An equal amount of RNA for all condi-tions was used as a template for cDNA synthesis using the


iScript cDNA Synthesis kit with provided random primers(Bio-Rad Hercules CA) Quantitative RT-PCR (qPCR) wasperformed in triplicate using 500 ngwell cDNA and Max-ima SYBR Green Master Mix (Thermo Fisher ScientificWaltham MA) in 25120583L reactions Cycling was conductedin a Bio-Rad iQ5 thermal cycler for 40 cycles (95∘C 15 s57∘C 60 s 79∘C 30 s) followed by melt curve analysis Datawere analyzed using Bio-Rad iQ5 software (2minusΔΔCt methodnormalized to 120573-Actin) Primers used were TWIST for-ward 1 51015840-CTATGTGGCTCACGAGCGGCTC-31015840 TWISTreverse 1 51015840-CCAGCTCCAGAGTCTCTAGACTGTCC-31015840TWIST forward 2 51015840-TCTTACGAGGAGCTGCAGACG-CA-31015840 TWIST reverse 2 51015840-ATCTTGGAGTCCAGCTCG-TCGCT-31015840 N-Cadherin forward 51015840-GGGACAGTTCCT-GAGGGATCAA-31015840 N-Cadherin reverse 51015840-TGGAGC-CTGAGACACGATTCTG-31015840 Vimentin forward 51015840-TCG-TCACCTTCGTGAATACCAAGA-31015840 Vimentin reverse 51015840-CCTCAGGTTCAGGGAGGAAAAGTT-31015840 120573-Actin for-ward 51015840-CCGCAAAGACCTGTACGCCAAC-31015840 120573-Actinreverse 51015840-CCAGGGCAGTGATCTCCTTCTG-31015840

25 Western Blot Cells were seeded at 250000 cells perwell in 6-well tissue culture plates and treated as describedin Section 21 Cells were pelleted and lysed in RIPA bufferand protein concentration was determined using a BCAAssay (Thermo Fisher Scientific Waltham MA) 30120583g totalprotein per lane was run on 4 stacking and 10ndash12resolving polyacrylamide gels and transferred to Immobilon-P PVDF membrane (Millipore Billerica MA) using a Trans-Blot SD Semi-Dry Transfer Cell (Bio-Rad Hercules CA)Membranes were blocked with 5 dry milk dissolved in1X PBS with 01 Tween-20 Antibodies were diluted inblocking buffer Antibodies used were anti-TWIST TWIST2c1a (Santa Cruz Biotech Dallas TX) anti-120573-Actin A1978(SigmaAldrich St LouisMO) andHorse Radish Peroxidase(HRP) conjugated anti-mouse secondary antibodies ECLPlus chemiluminescent substrate (Pierce Thermo FisherScientific Waltham MA) and Blue Devil Film (GeneseeScientific San Diego CA) were used for capturing images

26 Confocal Microscopy SUM 1315 cells were transfectedwith dendriplexes comprised of siQ (labeled with AlexaFluor647) and YTZ3-15 and incubated for 24 hours in a tissueculture incubator Cells were then treated with LysoTrackerRed (Thermo Fisher Scientific Waltham MA) according tothe manufacturerrsquos protocol Confocal images were obtainedusing the Zeiss LSM 700 Confocal Microscope and ZEN 2012microscopy software (Zeiss AG Oberkochen Germany)

27 Tumor Engraftment A total of six female NODCg-Prkdcscid Il2rgtm1WjlSzJ (NSG) mice (The Jackson LaboratoryBar Harbor ME) were used in a pilot study to engraft theSUM 1315 eGFP + luc breast cancer cells (8 weeks old attime of inoculation) The six mice were to be divided intotwo groups intratumoral (IT) and intravenous (IV) Whileunder anesthesia (Isoflurane 25ndash4)mice received bilateralinoculations of cells into the 4thmammary fat pad Inoculumfor each mammary fat pad consisted of 1 times 106 SUM 1315

eGFP + luc cells together with 2 times 105 hMSCs suspended in50 120583L 3mgmL Matrigel Injections were delivered into themammary fat pad adjacent to the nipple Mice were thenallowed to recover in a clean cage Two NSG mice receivingno cells were used as controls

28 In Vivo Imaging After tumor cell inoculations the micewere imaged every two weeks using the Xenogen IVIS 100biophotonic imaging system (STTARR Toronto OntarioCanada) to monitor tumor growth To obtain in vivo imagesmice were given a 200120583L intraperitoneal (IP) injection of25mgmL D-Luciferin (PerkinElmer Waltham MA) Aftera 10-minute waiting period animals were anesthetized usingIsoflurane (2ndash5) and placed in a black box in the biopho-tonic imager Bioluminescent images were captured over aperiod of one minute Once tumors had reached 05ndash075 cmby caliper measurement three mice were given a singleintravenous (IV) injection of the YTZ3-15 + siQ dendriplex(15 120583L of 240120583M YTZ3-15 and 10 120583L of 10 120583M siQ) dilutedin 200120583L PBS A separate group of three animals was givenintratumoral injections of the YTZ3-15 + siQ dendriplex(15 120583L of 240120583MYTZ3-15 and 10 120583L of 10 120583M siQ) diluted in100 120583L PBS After the injections animals underwent in vivofluorescent imaging using the IVIS 100 (Cy55 filter) Imageswere captured at 5 10 15 and 240 minutes after injectionof the dendriplexes After the final time point (4 hours) allanimals were euthanized and tissues were collected Tumorsspleen kidney and liver from each animal were imaged exvivo using the IVIS 100 to detect the AlexaFluor 647-labeledsiQ without the hindrance of the skin and fur TwoNSGmicereceiving no cells were used as controls for in vivo imaging

29 Statistics and Replications Wound healing assays wererepeated three times as were the Western blot analysesInvasion assay was repeated twice with identical conditionsFive images were captured for each invasion assay conditionand the numbers of cells were countedmanually and standarddeviations were calculated using Excel (Microsoft RedmanWA) qPCR experiments were done in triplicate and analyzedusing the 2minusΔΔCt method in the Bio-Rad iQ5 software Threeanimals per group (along with two control animals) wereused for biodistribution purposes only

3 Results and Discussion

31 Stable TWIST1 Knockdown in SUM 1315 Cells The rela-tionship between TWIST1 expression and EMT has beenestablished for breast cancer [44] To examine the effects ofTWIST1 knockdown in SUM 1315 cells without the possibleconfounding variables that a delivery mechanism may causewe developed cell lines that stably expressed shRNAs againstTWIST1 (shTwistA and shTwistB) and a scrambled shRNA(shScram) as a negative control TWIST1 expression in theSUM 1315 shTwistA and shTwistB cell lines demonstratedexcellent knockdown of TWIST1 compared to the parentalline and the shScram line (Figures 2(a) and 2(b))

To test the effect of TWIST1 knockdown on cell migra-tion we performed a wound healing assay Our results


Pare

ntal

shSc

ram

shTw

istA

shTw

istB

TWIST1

Actin

(a)

0

02

04

06

08

1

12

14

SUM 1315parental

SUM 1315shScram

SUM 1315shTwistA

SUM 1315shTwistB

Relat

ive e

xpre

ssio

n

(b)

shScram shTwistA shTwistB

0ho

urs

96

hour

s

(c)

NTF

Lipo

onl

y

siTw

istA

siTw

istB

TWIST1

Actin

siCtr

l

(d)

00

02

04

06

08

10

12

14

NTF Lipo only siCtrl siTwistA siTwistB

Relat

ive e

xpre

ssio

n

(e)

Figure 2 Stable and transient RNA-mediatedTWIST1 knockdown in SUM1315 (a)Western blotting demonstrated robust TWIST1 knockdownin both shTwistA and shTwistB lines (b) qPCR confirmed TWIST1 knockdown at the mRNA level for both stable knockdown lines Errorbars represent standard deviation (c) SUM 1315 cells expressing shTwistA or shTwistB exhibited decreased directional migration comparedto those expressing shScram control in wound healing assays Dashed lines indicate migratory front and were addedmanually Images shownare representative data from experiments performed in triplicate (d) Western blot demonstrated substantial TWIST1 knockdown in SUM1315 cells transfected with siTwistA and siTwistB using Lipofectamine 2000 when compared to nontransfected (NTF) Lipofectamine 2000alone (Lipo only) or control siRNA (siCtrl) (e) qPCR results mirrored those seen in the Western blot


demonstrated that the SUM 1315 shTwistA and shTwistB celllines had reduced directional migratory abilities compared tothe SUM 1315 shScram cell line (Figure 2(c)) Taken togetherthese data suggest not only that shTwistA and shTwistBsignificantly knock down expression of TWIST1 in SUM 1315cells but also that the downregulation of TWIST1 results ina phenotypic change consistent with diminished migratoryability

32 siRNA-Mediated TWIST1 Knockdown in SUM 1315 CellsThe SUM 1315 shRNA results described above demonstratenot only a significant reduction in the amount of TWIST1expression but also a phenotypic change in cell migra-tion suggesting that these shRNA sequences were effectivein knocking down TWIST1 expression We thus designedsiRNA sequences (siTwistA and siTwistB) based on theseshRNA sequences To test the efficacy of siTwistA andsiTwistB SUM 1315 cells were transfected using Lipofec-tamine 2000 Transfection with either siTwistA or siTwistBresulted in knockdown of TWIST1 with siTwistB givingslightly more knockdown than siTwistA at both the proteinand mRNA levels (Figures 2(d) and 2(e)) Next we testedthe delivery of siRNA into SUM 1315 cells using the YTZ3-15 dendrimer Cellular uptake of AlexaFluor 647-labeled siQ(acting as a surrogate for unlabeled siTwistA and siTwistB)was greater than 90 after 24 hours as measured by flowcytometry and fluorescent microscopy (Figures 3(a) and3(b)) The presence of siQ in cells transfected using YTZ3-15dendriplexes was confirmed as far out as 7 days from the timeof transfection (Figure 3(b))These findings confirmpreviouswork [38] performed with this PAMAM dendrimer anddemonstrate its ability to safely deliver siRNA across the cellmembrane because we did not appreciate any increase in celldeath Cellular uptake using YTZ3-15 + siQ was comparablewhen tested in other cell lines including other breast ovarianuterine and prostate cancer cell lines (data not shown)

While uptake of the dendriplexes can be appreciated withfluorescent microscopy and flow cytometry these methodsdo not indicate the location of the siRNA within the cellTo examine this we used LysoTracker Red (dye taken upinto acidic organelles) to show where siQ is localized Ourresults show that much of the siQ signal colocalizes withthe mid to late endosome in the SUM 1315 eGFP + luc cellline (Figure 3(c)) siRNA localization to these organelles isdesirable to take advantage of the ldquoproton sponge effectrdquowhich is thought to be essential for siRNA release [45 46]

After confirming the function of siTwistA and siTwistBwith Lipofectamine 2000 and the cellular uptake of siQusing YTZ3-15 we tested siTwistA and siTwistB with YTZ3-15-based delivery TWIST1 levels were measured usingqPCR and found to be significantly reduced at 24 hoursand one week after transfection (Figure 4(a)) Two EMT-related TWIST1 target genes (Vimentin and N-Cadherin)also showed reduced mRNA expression Vimentin and N-Cadherin were both substantially reduced at the 24-hourtime point however Vimentin showed a slight return atthe one week time point whereas N-Cadherin continued todecrease (Figure 4(b)) While reduced expression of these

genes was noted renewed expression of the epithelial markerE-Cadherin was not observed (data not shown) This isa noted difference from previous studies [4] The possiblecauses for this discrepancy are the different cell lines usedbetween previous studies and ours and that E-Cadherinis not entirely controlled by TWIST1 [19 47] Reducedexpression of these EMT-related genes is a positive indicationthat migration and invasion would be hindered

Next we performed awoundhealing assay to validate thatYTZ3-15-delivered siRNA against TWIST1 not only reducesthe expression of TWIST1 and its target genes but alsoinhibits the migratory action of SUM 1315 cells This assaydemonstrated decreased directional migration of SUM 1315cells transfected with siTwistA (Figure 4(c))

The EMT process consists of migration and invasion andTWIST1 is a major factor in allowing cancer cells to infiltratesurrounding tissues blood vessels and the lymphatic system[40 48] To investigate whether the invasive phenotype isreduced following siRNA-mediated TWIST1 knockdown weperformed a transwell invasion assay Results indicated thatthe YTZ3-15 + siRNA-treated cells had diminished abilitiesto invade the Matrigel matrix and pass through the porousmembrane thus indicating a reduction in the invasive pheno-type (Figure 4(d)) TWIST1 overexpression is associated withcancers that are more metastatic and therefore invasive [22]and these data show that TWIST1 silencing following den-driplex delivery of siRNA decreases metastatic potentialThisin turn suggests that as a therapeutic approach for patientswith MBC this delivery method and target could have asignificant impact on improving survival and outcomes forMBC patients if preclinical and clinical trials show similarresults

33 In Vivo Distribution of PAMAM Dendrimers In vivostudies were completed to determine the optimum deliveryroute (IV versus IT) of siQ using YTZ3-15 Five minutes afterthe IV or IT injection of the YTZ3-15 + siQ dendriplex abright signal was noted at the site of the tumor (Figure 5(a))The signal at the tumor site continued to be evident inmice that received IT injections at 10 15 and 240 minuteswhereas no signal was seen at the tumor site after 5 minutesin mice that received IV injections (Figure 5(a)) Howeverex vivo imaging of tumors spleen liver and kidneys after240 minutes revealed a robust AlexaFluor 647 signal in thetumors but little to no signal in other examined organs(Figure 5(b)) This ex vivo tumor-centric signal was evidentfor all mice regardless of the route of administration (IVversus IT)

While the YTZ3-15 dendrimer does not have any inher-ent tumor-targeting capabilities results from our in vivostudies demonstrate that these dendriplexes do accumulatepreferentially in the orthotopic breast cancer tumors It ispossible that localization to the tumor is due to the enhancedpermeability and retention (EPR) effect which has been seenwith other PAMAM dendrimers and nanoparticle deliveryvehicles [49ndash51] The inherent leakiness of tumor vasculaturecoupled with minimal lymphatic drainage results in particles


Nontransfected

Forw

ard

scat

ter

AlexaFluor-647signal intensity


60K

40K

20K

0

100

101

102

103

104

100

101

102

103

104

011Alexa-647 subset

992Alexa-647 subset

YTZ3-15 + siQ

(a)

Phas

eA

lexa

Fluo

r 488

Mer

ge

1 day 7 days

(b)

Merge

Phase

LysoTracker

GFP

siQ

(c)

Figure 3 YTZ3-15 effectively delivers siRNA to SUM 1315 cells (a) Left nontransfected SUM 1315 cells had low background fluorescenceRight more than 99 of YTZ3-15 transfected cells were positive for AlexaFluor 647-labeled siQ (b) Fluorescent microscopy revealed thatAlexaFluor 488-labeled siQ was taken up into cells within one day and AlexaFluor signal was still detectable in cells at seven days aftertransfection (c) Confocal images of SUM 1315 cells stably expressing eGFP + luc and transiently transfected with AlexaFluor 647-labeled siQusing YTZ3-15 LysoTracker dye revealed that siQ primarily colocalized with mid to late endosomes after 24-hour incubation with YTZ3-15siRNA dendriplexes

becoming trapped and consequently concentrated in thetumor environment This effect is magnified as the tumorenlarges and promotes angiogenesis which may explain whysiQ concentration was noted only after orthotopic tumorsreached 05 times 05 cm in size (data not shown)

In addition to our promising results there are otherimportant reasons to focus on the knockdown of TWIST1 incancer cells TWIST1 plays an essential role in early embry-onic development as evidenced in mice and humans withheterozygous gene mutations where both have craniofacialabnormalities (Saethre-Chotzen syndrome in humans) [52

53] Additionally TWIST1 knockout mice are embryoniclethal [54 55] Given that TWIST1 is crucial in early devel-opment it is not surprising that it maintains the CSC phe-notype [5ndash7 10 56] The CSC phenotype is associated withan undifferentiated cellular morphology increased mobilityself-renewal resistance to apoptosis and chemoresistance[7] thus silencing of TWIST1 may aid in weakening thosecells that are most resilient to current therapeutic modalities

A TWIST1 siRNA therapeutic approach to assist inthe treatment of MBC is also attractive because it couldcomplement and augment current treatment regimens For


00

02

04

06

08

10

12

14

Rela

tive e

xpre

ssio

n

Twist

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(a)

N-Cadherin Vimentin

00

02

04

06

08

10

12

14

Relat

ive e

xpre

ssio

n

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(b)

siQTw

A

0 hours 24 hours 48 hours

(c)

SiQ TwA TwB CellsField200

150

100

50

0

SiQ TwA TwB

(d)

Figure 4 TWIST1 knockdown following YTZ3-15 delivery of siTwist decreases cell motility and downstream EMT marker expression (a)Compared to siQ control (at seven days) siTwistA (TwA) and siTwistB (TwB) delivered via YTZ3-15 produced gt90 TWIST1 knockdown atthe mRNA level Knockdown lasted seven days after transfection (b) Compared to siQ control (at seven days) TwA and TwB delivered viaYTZ3-15 produced knockdown of the TWIST1 targets N-Cadherin and Vimentin N-Cadherin mRNA levels decreased by gt40 after oneday and by approximately 90 after seven days Vimentin mRNA was nearly undetectable after one day and remained at lt10 after sevendays (c) YTZ3-15 transfection of siTwistA decreased directional migration compared to siQ transfected cells (control) in wound healingassays Dashed lines indicate migratory front and were placed manually Images shown are representative data from experiments performedin triplicate (d) Left YTZ3-15 transfection of TwA or TwB resulted in gt50 decrease in invasion of SUM 1315 cells through Matrigel Cellswere allowed to migrate for one day following one day incubation with YTZ3-15-siRNA dendriplexes Five fields per condition were imaged(representative images shown) Right quantification of image data Bars represent mean and standard deviation of five fields per condition


fflucIV

cont

rol

Intr

aven

ous

Intr

atum

oral

5min 10min 15min 240min

70

60

50

40

30

20

10

times106

times106

300

250

200

150

(a)

Kidney

Spleen

Liver

Tumor

Spleen

Kidney

Liver Tumors

Kidney

Spleen

Liver

Intr

aven

ous

Intr

atum

oral

IV co

ntro

l

50 100 150 200 times106

(b)

Figure 5 YTZ3-15 concentrates in orthotopic breast cancer tumors in vivo (a) Representative animals from the mice that received YTZ3-15 +siQ via intratumoral (IT) and intravenous (IV) injections Control animals received IV injections of the dendriplexes but had no tumorsMicereceiving IT injections showed accumulation of siQ lasting at least 15 minutes post injection whereas mice receiving IV injections showedlittle accumulation after 5 minutes Control animals do not show accumulation of siQ due to the absence of tumors (b) Ex vivo imaging ofspleen kidney liver and tumors (where applicable) from the three animals shown in Figure 5(a) demonstrating concentration of YTZ3-15+ siQ dendriplexes in the tumors but not in other organs Images were obtained 240mins after the administration (IT or IV) of YTZ3-15 +siQ The units for the scale bars in this figure are photonsseccm2steradian

example it is known that TWIST1 overexpression in breastcancer is associated with a poorer prognosis partly dueto downregulation of estrogen receptor 120572 (ER-120572) [13 1421] a reduction in ER-120572 leads to a diminished sensitivityto hormone therapies Furthermore TWIST1 expression isassociated with resistance to commonly used chemother-apy agents in many human carcinomas [15 57] and ithas been demonstrated that a reduction of TWIST1 canresensitize tumor cells to chemotherapy [58] TWIST1 isalso an intriguing therapeutic target because for almostall adult tissues TWIST1 expression is nonexistent [8 59]Therefore if a TWIST1-specific therapy could be deliveredthe side effects onnontumor tissuewould beminimal becausethere is no TWIST1 to knock down Taken together thesedata and observations suggest that siRNA-based knockdown

of TWIST1 could be used in conjunction with hormonaltherapy or chemotherapy to achieve a synergistic effectSuch a combined approach (chemotherapy plus anti-TWIST1siRNA) is currently being explored by our laboratory andothers using various types of nanoparticles that allow forsimultaneous delivery into breast cancer cells [60ndash62]

4 Conclusions

Our studies demonstrate successful delivery and utilizationof two siRNAs against TWIST1 Delivery was realized using amodified third generation PAMAM dendrimer and resultedin significant knockdown of TWIST1 and other EMT-relatedtarget genes in vitro TWIST1 knockdown resulted in a reduc-tion in cellular migration and invasion as has been observed


previously [9 11 40 48 63] Finally delivery of an siRNA byYTZ3-15 was shown to have a specific concentrating ability inorthotopic tumors in a TNBC mouse model

These data add to the growing evidence that TWIST1 isan important and potentially clinically significant therapeutictarget for the treatment of MBC as well as other solid tumorcancers [23 64 65]While TWIST1 knockdown via PAMAMdendrimer-delivered siRNAcould not reasonably be used as asole means of treatment for MBC it could serve as a valuabletool and adjuvant therapy to reduce migrationinvasionchemoresistance and antiapoptotic tendencies associatedwith aggressive tumors Novel results from this study serveto validate a multimodal approach to cancer treatment byfocusing on a transcription factor associated with breastcancer tumor types that have minimal treatment options(eg TNBC) Furthermore these data support further inves-tigations (both in vitro and in vivo) into the use of siRNAcoupled with nanoparticles to treat malignant breast cancerby knocking down TWIST1 and its associated EMT targets

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Acknowledgments

The authors would like to thank Dr Nicola Solomon PhDfor editorial assistance and critical review of the paper Theyalso wish to thank Drs John Rossi PhD and Jiehua ZhouPhD for their assistance in obtaining and testing the YTZ3-15 dendrimer Also they thank Dr Ling Peng for providingYTZ3-15 reagents They are grateful to Dr Shan Li PhD forthe initial creation of the SUM 1315 shRNA stable cell lines

References

[1] R Siegel J Ma Z Zou and A Jemal ldquoCancer statistics 2014rdquoCA Cancer Journal for Clinicians vol 64 no 1 pp 9ndash29 2014

[2] O Hussein and S V Komarova ldquoBreast cancer at bonemetastatic sites recent discoveries and treatment targetsrdquo Jour-nal of Cell Communication and Signaling vol 5 no 2 pp 85ndash992011

[3] N U Lin E Claus J Sohl A R Razzak A Arnaout andE P Winer ldquoSites of distant recurrence and clinical outcomesin patients with metastatic triple-negative breast cancer highincidence of central nervous system metastasesrdquo Cancer vol113 no 10 pp 2638ndash2645 2008

[4] F Vesuna P van Diest J H Chen and V Raman ldquoTwist isa transcriptional repressor of E-cadherin gene expression inbreast cancerrdquo Biochemical and Biophysical Research Commu-nications vol 367 no 2 pp 235ndash241 2008

[5] B G Hollier K Evans and S A Mani ldquoThe epithelial-to-mesenchymal transition and cancer stem cells a coalitionagainst cancer therapiesrdquo Journal of Mammary Gland Biologyand Neoplasia vol 14 no 1 pp 29ndash43 2009

[6] D Kong Y Li ZWang and F H Sarkar ldquoCancer stem cells andepithelial-to-mesenchymal transition (EMT)-phenotypic cellsare they cousins or twinsrdquo Cancers vol 3 no 1 pp 716ndash7292011

[7] N Takebe R Q Warren and S P Ivy ldquoBreast cancer growthand metastasis interplay between cancer stem cells embryonicsignaling pathways and epithelial-to-mesenchymal transitionrdquoBreast Cancer Research vol 13 no 3 article 211 2011

[8] OWatanabeH Imamura T Shimizu et al ldquoExpression of twistand wnt in human breast cancerrdquo Anticancer Research vol 24no 6 pp 3851ndash3856 2004

[9] A E Vernon and C LaBonne ldquoTumor metastasis a new twiston epithelial-mesenchymal transitionsrdquoCurrent Biology vol 14no 17 pp R719ndashR721 2004

[10] F Vesuna A Lisok B Kimble and V Raman ldquoTwist modulatesbreast cancer stem cells by transcriptional regulation of CD24expressionrdquo Neoplasia vol 11 no 12 pp 1318ndash1328 2009

[11] M-Y Feng K Wang H-T Song et al ldquoMetastasis-inductionand apoptosis-protection by TWIST in gastric cancer cellsrdquoClinical and Experimental Metastasis vol 26 no 8 pp 1013ndash1023 2009

[12] R Maestro A P Dei Tos Y Hamamori et al ldquoTwist is a poten-tial oncogene that inhibits apoptosisrdquo Genes and Developmentvol 13 no 17 pp 2207ndash2217 1999

[13] J Fu L Zhang THe et al ldquoTWIST represses estrogen receptor-alpha expression by recruiting the NuRD protein complex inbreast cancer cellsrdquo International Journal of Biological Sciencesvol 8 no 4 pp 522ndash532 2012

[14] F Vesuna A Lisok B Kimble et al ldquoTwist contributes to hor-mone resistance in breast cancer by downregulating estrogenreceptor-alphardquo Oncogene vol 31 no 27 pp 3223ndash3234 2012

[15] Y Chen L Li J Zeng et al ldquoTwist confers chemoresistanceto anthracyclines in bladder cancer through upregulating P-glycoproteinrdquo Chemotherapy vol 58 no 4 pp 264ndash272 2012

[16] M C Elias K R Tozer J R Silber et al ldquoTWIST is expressedin human gliomas and promotes invasionrdquoNeoplasia vol 7 no9 pp 824ndash837 2005

[17] N Matsuo H Shiraha T Fujikawa et al ldquoTwist expressionpromotes migration and invasion in hepatocellular carcinomardquoBMC Cancer vol 9 article 240 2009

[18] H-F Yuen Y-P Chan M L-Y Wong et al ldquoUpregulation ofTwist in oesophageal squamous cell carcinoma is associatedwith neoplastic transformation and distant metastasisrdquo Journalof Clinical Pathology vol 60 no 5 pp 510ndash514 2007

[19] H-F Yuen C-W Chua Y-P Chan Y-C Wong X Wangand K-W Chan ldquoSignificance of TWIST and E-cadherinexpression in the metastatic progression of prostatic cancerrdquoHistopathology vol 50 no 5 pp 648ndash658 2007

[20] T AMartin A Goyal GWatkins andWG Jiang ldquoExpressionof the transcription factors snail slug and twist and theirclinical significance in human breast cancerrdquo Annals of SurgicalOncology vol 12 no 6 pp 488ndash496 2005

[21] J G H van Nes E M de Kruijf H Putter et al ldquoCo-expression of SNAIL and TWIST determines prognosis inestrogen receptor-positive early breast cancer patientsrdquo BreastCancer Research and Treatment vol 133 no 1 pp 49ndash59 2012

[22] J Yang S AMani J LDonaher et al ldquoTwist amaster regulatorof morphogenesis plays an essential role in tumor metastasisrdquoCell vol 117 no 7 pp 927ndash939 2004

[23] M A Khan H-C Chen D Zhang and J Fu ldquoTwist amolecular target in cancer therapeuticsrdquo Tumor Biology vol 34no 5 pp 2497ndash2506 2013

[24] A Fire S Xu M K Montgomery S A Kostas S E Driver andC CMello ldquoPotent and specific genetic interference by double-stranded RNA in Caenorhabditis elegansrdquo Nature vol 391 no6669 pp 806ndash811 1998


[25] K A Whitehead R Langer and D G Anderson ldquoKnockingdown barriers advances in siRNA deliveryrdquo Nature ReviewsDrug Discovery vol 8 no 2 pp 129ndash138 2009

[26] A J Hamilton and D C Baulcombe ldquoA species of smallantisense RNA in posttranscriptional gene silencing in plantsrdquoScience vol 286 no 5441 pp 950ndash952 1999

[27] P Guo O Coban N M Snead et al ldquoEngineering RNA fortargeted siRNA delivery and medical applicationrdquo AdvancedDrug Delivery Reviews vol 62 no 6 pp 650ndash666 2010

[28] L Aagaard and J J Rossi ldquoRNAi therapeutics principlesprospects and challengesrdquoAdvancedDrugDelivery Reviews vol59 no 2-3 pp 75ndash86 2007

[29] M A Behlke ldquoProgress towards in vivo use of siRNAsrdquoMolecular Therapy vol 13 no 4 pp 644ndash670 2006

[30] K Gavrilov and W M Saltzman ldquoTherapeutic siRNA princi-ples challenges and strategiesrdquoThe Yale Journal of Biology andMedicine vol 85 no 2 pp 187ndash200 2012

[31] N M Snead and J J Rossi ldquoBiogenesis and function ofendogenous and exogenous siRNAsrdquo Wiley InterdisciplinaryReviews RNA vol 1 no 1 pp 117ndash131 2010

[32] D Castanotto and J J Rossi ldquoThe promises and pitfalls of RNA-interference-based therapeuticsrdquo Nature vol 457 no 7228 pp426ndash433 2009

[33] J Wang Z Lu M G Wientjes and J L-S Au ldquoDelivery ofsiRNA therapeutics barriers and carriersrdquo The AAPS Journalvol 12 no 4 pp 492ndash503 2010

[34] S David B Pitard J-P Benoıt and C Passirani ldquoNon-viralnanosystems for systemic siRNA deliveryrdquo PharmacologicalResearch vol 62 no 2 pp 100ndash114 2010

[35] C Foged ldquosiRNA delivery with lipid-based systems promisesand pitfallsrdquo Current Topics in Medicinal Chemistry vol 12 no2 pp 97ndash107 2012

[36] Y Tang Y-B Li B Wang et al ldquoEfficient in vitro siRNAdelivery and intramuscular gene silencing using PEG-modifiedPAMAM dendrimersrdquo Molecular Pharmaceutics vol 9 no 6pp 1812ndash1821 2012

[37] J Zhou C P Neff X Liu et al ldquoSystemic administration ofcombinatorial dsiRNAs via nanoparticles efficiently suppressesHIV-1 infection in humanizedmicerdquoMolecularTherapy vol 19no 12 pp 2228ndash2238 2011

[38] T Yu X Liu A-L Bolcato-Bellemin et al ldquoAn amphiphilicdendrimer for effective delivery of small interfering RNA andgene silencing in Vitro and in vivordquo Angewandte ChemieInternational Edition vol 51 no 34 pp 8478ndash8484 2012

[39] X Liu J Zhou T Yu et al ldquoAdaptive amphiphilic dendrimer-based nanoassemblies as robust and versatile siRNA deliverysystemsrdquo Angewandte ChemiemdashInternational Edition vol 53no 44 pp 11822ndash11827 2014

[40] S Li S E Kendall R Raices et al ldquoTWIST1 associates withNF-120581B subunit RELA via carboxyl-terminal WR domain topromote cell autonomous invasion through IL8 productionrdquoBMC Biology vol 10 article 73 2012

[41] C E Brown R Starr CMartinez et al ldquoRecognition and killingof brain tumor stem-like initiating cells by CD8+ cytolytic Tcellsrdquo Cancer Research vol 69 no 23 pp 8886ndash8893 2009

[42] S Samineni C Glackin and J E Shively ldquoRole of CEACAM1ECM and mesenchymal stem cells in an orthotopic model ofhuman breast cancerrdquo International Journal of Breast Cancervol 2011 Article ID 381080 10 pages 2011

[43] C-C Liang A Y Park and J-L Guan ldquoIn vitro scratchassay a convenient and inexpensive method for analysis of cell

migration in vitrordquo Nature Protocols vol 2 no 2 pp 329ndash3332007

[44] Y Teng and X Li ldquoThe roles of HLH transcription factorsin epithelial mesenchymal transition and multiple molecularmechanismsrdquo Clinical and Experimental Metastasis vol 31 no3 pp 367ndash377 2014

[45] D Ouyang H Zhang H S Parekh and S C Smith ldquoTheeffect of pH on PAMAM dendrimer-siRNA complexationmdashendosomal considerations as determined by molecular dynam-ics simulationrdquo Biophysical Chemistry vol 158 no 2-3 pp 126ndash133 2011

[46] R V Benjaminsen M A Mattebjerg J R Henriksen S MMoghimi and T L Andresen ldquoThe possible lsquoproton spongersquoeffect of polyethylenimine (PEI) does not include change inlysosomal pHrdquo Molecular Therapy vol 21 no 1 pp 149ndash1572013

[47] C-H Zhang G-L Xu W-D Jia et al ldquoActivation of STAT3signal pathway correlates with twist and E-cadherin expressionin hepatocellular carcinoma and their clinical significancerdquoTheJournal of Surgical Research vol 174 no 1 pp 120ndash129 2012

[48] C Li Y Ren X Jia et al ldquoTwist overexpression promotedepithelial-to-mesenchymal transition of human peritonealmesothelial cells under high glucoserdquo Nephrology DialysisTransplantation vol 27 no 11 pp 4119ndash4124 2012

[49] H Kobayashi R Watanabe and P L Choyke ldquoImproving con-ventional enhanced permeability and retention (EPR) effectswhat is the appropriate targetrdquo Theranostics vol 4 no 1 pp81ndash89 2014

[50] K Greish ldquoEnhanced permeability and retention (EPR) effectfor anticancer nanomedicine drug targetingrdquoMethods inMolec-ular Biology vol 624 pp 25ndash37 2010

[51] S Biswas and V P Torchilin ldquoDendrimers for siRNA deliveryrdquoPharmaceuticals vol 6 no 2 pp 161ndash183 2013

[52] WKress C SchroppG Lieb et al ldquoSaethre-Chotzen syndromecaused by TWIST 1 gene mutations functional differentiationfrom Muenke coronal synostosis syndromerdquo European Journalof Human Genetics vol 14 no 1 pp 39ndash48 2006

[53] V El Ghouzzi E Lajeunie M Le Merrer et al ldquoMutationswithin or upstream of the basic helix-loop-helix domain ofthe TWIST gene are specific to Saethre-Chotzen syndromerdquoEuropean Journal of Human Genetics vol 7 no 1 pp 27ndash331999

[54] P Bialek B Kern X Yang et al ldquoA twist code determines theonset of osteoblast differentiationrdquo Developmental Cell vol 6no 3 pp 423ndash435 2004

[55] K Soo M P OrsquoRourke P-L Khoo et al ldquoTwist function isrequired for the morphogenesis of the cephalic neural tube andthe differentiation of the cranial neural crest cells in the mouseembryordquo Developmental Biology vol 247 no 2 pp 251ndash2702002

[56] S Isenmann A Arthur A C W Zannettino et al ldquoTWISTfamily of basic helix-loop-helix transcription factors mediatehumanmesenchymal stem cell growth and commitmentrdquo StemCells vol 27 no 10 pp 2457ndash2468 2009

[57] X Wang M T Ling X-Y Guan et al ldquoIdentification of anovel function of Twist a bHLH protein in the developmentof acquired taxol resistance in human cancer cellsrdquo Oncogenevol 23 no 2 pp 474ndash482 2004

[58] Q-Q Li J-D Xu W-J Wang et al ldquoTwist1-mediatedadriamycin-induced epithelial-mesenchymal transition relatesto multidrug resistance and invasive potential in breast cancer


cellsrdquo Clinical Cancer Research vol 15 no 8 pp 2657ndash26652009

[59] Y Mironchik P T Winnard Jr F Vesuna et al ldquoTwistoverexpression induces in vivo angiogenesis and correlates withchromosomal instability in breast cancerrdquo Cancer Research vol65 no 23 pp 10801ndash10809 2005

[60] H Meng M Xue T Xia et al ldquoUse of size and a copolymerdesign feature to improve the biodistribution and the enhancedpermeability and retention effect of doxorubicin-loaded meso-porous silica nanoparticles in amurine xenograft tumormodelrdquoACS Nano vol 5 no 5 pp 4131ndash4144 2011

[61] J LuM Liong Z Li J I Zink and F Tamanoi ldquoBiocompatibil-ity biodistribution and drug-delivery efficiency of mesoporoussilica nanoparticles for cancer therapy in animalsrdquo Small vol 6no 16 pp 1794ndash1805 2010

[62] H Meng M Xue T Xia et al ldquoAutonomous in vitro anticancerdrug release from mesoporous silica nanoparticles by pH-sensitive nanovalvesrdquo Journal of the American Chemical Societyvol 132 no 36 pp 12690ndash12697 2010

[63] J Zhang P Wang F Wu et al ldquoAberrant expression of thetranscriptional factor Twist1 promotes invasiveness in ALK-positive anaplastic large cell lymphomardquoCellular Signalling vol24 no 4 pp 852ndash858 2012

[64] W K Kwok M-T Ling T-W Lee et al ldquoUp-regulation ofTWIST in prostate cancer and its implication as a therapeutictargetrdquo Cancer Research vol 65 no 12 pp 5153ndash5162 2005

[65] H Wallerand G Robert G Pasticier et al ldquoThe epithelial-mesenchymal transition-inducing factor TWIST is an attractivetarget in advanced andor metastatic bladder and prostatecancersrdquo Urologic Oncology vol 28 no 5 pp 473ndash479 2010


+

Nuc

NucNucNuc

TWIST1knockdowndownstream

gene expressionchanges

reversal of EMT

1 2

3 4 5 6

Figure 1 Mechanism of dendrimer-mediated siRNA delivery and TWIST1 knockdown (1) Negatively charged siRNA is electrostaticallyattracted to positive charges on the YTZ3-15 dendrimer leading to the formation of 6ndash8 nm diameter micelles coated with siRNA (2) Thesedendriplexes are administered to tumor cells (3) Dendriplexes are taken up via macropinocytosis (4) Dendriplexes are trafficked to lateendosomes (5) Due to the proton sponge effect electrostatic interactions between the dendrimer and siRNA are disrupted and siRNA escapesfrom the disrupted endosome into the cytosol (6) Once in the cytosol siRNA recruits the endogenous RNAi machinery to degrade TWIST1mRNA Following TWIST1 knockdown TWIST1 target gene expression is altered to reduce invasive capacity

mobility In addition to promoting EMT in cancer cellsTWIST1 is thought to promote the cancer stem cell (CSC)phenotype [10] inhibit apoptosis [11 12] and contributeto chemotherapy resistance [13 14] TWIST1 has also beenshown to be overexpressed in numerous solid tumors [1115ndash19] including aggressive and metastatic forms of breastcancer [20ndash22] however it is not expressed in normal adulttissues This expression profile coupled with the establishedrole of TWIST1 in numerousmetastasis-promoting pathwayssuggests it is a promising novel target for MBC therapy [23]

The therapeutic use of small interfering RNA (siRNA)for cancer has gained considerable interest since its genesilencing properties were first described [24ndash26] Once inthe cytoplasm the siRNA unwinds and associates with Argo-naute2 forming an RNA-induced silencing complex (RISC)which leads to sequence-specific mRNA degradation andgene silencing [27] However while promising the develop-ment of siRNA therapy has encountered challenges includingsusceptibility to enzymatic degradation delivery to targettissues endosomal escape immune activation and off-targeteffects [28ndash31] Effective siRNA delivery (both to the tissueof interest and across the cell membrane) remains one ofthe main barriers to developing clinically relevant therapies[32 33]

The success of an siRNA-based gene silencing therapeuticapproach requires that the siRNA enters the cytoplasmwithout being degraded [34 35] Recent studies have demon-strated that poly(amidoamine) (PAMAM) dendrimers arecapable of functional siRNA delivery both in vitro and

in vivo by protecting siRNA (via electrostatic interactionsand aggregation) from enzymatic degradation prior tomacropinocytosis and eventual release for the endosome(Figure 1) [36 37] Recently a modified third generationamphiphilic PAMAM dendrimer (YTZ3-15) was shown toeffectively deliver siRNA and cause gene knockdown invivo via intratumoral (IT) administration [38] ComplexingYTZ3-15 with TWIST1 siRNAmay therefore have the poten-tial to allow delivery of potent siRNAs to breast tumor cellsto reduce TWIST1-mediated expression of EMT target genesand inhibit metastatic potential

In the current study we investigated whether anti-TWIST1 siRNA could be functionally delivered to metastaticbreast cancer cells (SUM 1315 cell line) using YTZ3-15 Wetested the ability of the YTZ3-15-delivered siRNA to knockdown TWIST1 reduce expression of EMT-related targetgenes and alter the phenotypic characteristics associatedwith cancer cell migrationinvasion We also evaluated thetumor-specific delivery capability of YTZ3-15 in vivo using amouse breast cancer model

2 Materials and Methods

21 Cell Culture Transfection and Stable Cell Line ProductionSUM 1315 breast cancer cells were obtained from ATCC(Manassas VA) Cells were maintained at 37∘C 5 CO

2 and

90 humidity in a tissue culture incubator Media for SUM1315 cells consisted of a 50-50 mixture of DMEM and F12media supplemented with 5120583gmL insulin 10 ngmL EGF





H247

N37O20



















Pare

ntal

shSc

ram

shTw

istA

shTw

istB

TWIST1

Actin

(a)

0

02

04

06

08

1

12

14

SUM 1315parental

SUM 1315shScram

SUM 1315shTwistA

SUM 1315shTwistB

Relat

ive e

xpre

ssio

n

(b)


0ho

urs

96

hour

s

(c)

NTF

Lipo

onl

y

siTw

istA

siTw

istB

TWIST1

Actin

siCtr

l

(d)

00

02

04

06

08

10

12

14


Relat

ive e

xpre

ssio

n

(e)













Nontransfected

Forw

ard

scat

ter



60K

40K

20K

0

100

101

102

103

104

100

101

102

103

104

011Alexa-647 subset

992Alexa-647 subset

YTZ3-15 + siQ

(a)

Phas

eA

lexa

Fluo

r 488

Mer

ge

1 day 7 days

(b)

Merge

Phase

LysoTracker

GFP

siQ

(c)







00

02

04

06

08

10

12

14

Rela

tive e

xpre

ssio

n

Twist

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(a)

N-Cadherin Vimentin

00

02

04

06

08

10

12

14

Relat

ive e

xpre

ssio

n

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(b)

siQTw

A


(c)


150

100

50

0

SiQ TwA TwB

(d)



fflucIV

cont

rol

Intr

aven

ous

Intr

atum

oral


70

60

50

40

30

20

10

times106

times106

300

250

200

150

(a)

Kidney

Spleen

Liver

Tumor

Spleen

Kidney

Liver Tumors

Kidney

Spleen

Liver

Intr

aven

ous

Intr

atum

oral

IV co

ntro

l

50 100 150 200 times106

(b)




4 Conclusions







Acknowledgments


References










































































H247

N37O20



















Pare

ntal

shSc

ram

shTw

istA

shTw

istB

TWIST1

Actin

(a)

0

02

04

06

08

1

12

14

SUM 1315parental

SUM 1315shScram

SUM 1315shTwistA

SUM 1315shTwistB

Relat

ive e

xpre

ssio

n

(b)


0ho

urs

96

hour

s

(c)

NTF

Lipo

onl

y

siTw

istA

siTw

istB

TWIST1

Actin

siCtr

l

(d)

00

02

04

06

08

10

12

14


Relat

ive e

xpre

ssio

n

(e)













Nontransfected

Forw

ard

scat

ter



60K

40K

20K

0

100

101

102

103

104

100

101

102

103

104

011Alexa-647 subset

992Alexa-647 subset

YTZ3-15 + siQ

(a)

Phas

eA

lexa

Fluo

r 488

Mer

ge

1 day 7 days

(b)

Merge

Phase

LysoTracker

GFP

siQ

(c)







00

02

04

06

08

10

12

14

Rela

tive e

xpre

ssio

n

Twist

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(a)

N-Cadherin Vimentin

00

02

04

06

08

10

12

14

Relat

ive e

xpre

ssio

n

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(b)

siQTw

A


(c)


150

100

50

0

SiQ TwA TwB

(d)



fflucIV

cont

rol

Intr

aven

ous

Intr

atum

oral


70

60

50

40

30

20

10

times106

times106

300

250

200

150

(a)

Kidney

Spleen

Liver

Tumor

Spleen

Kidney

Liver Tumors

Kidney

Spleen

Liver

Intr

aven

ous

Intr

atum

oral

IV co

ntro

l

50 100 150 200 times106

(b)




4 Conclusions







Acknowledgments


References


















































































Pare

ntal

shSc

ram

shTw

istA

shTw

istB

TWIST1

Actin

(a)

0

02

04

06

08

1

12

14

SUM 1315parental

SUM 1315shScram

SUM 1315shTwistA

SUM 1315shTwistB

Relat

ive e

xpre

ssio

n

(b)


0ho

urs

96

hour

s

(c)

NTF

Lipo

onl

y

siTw

istA

siTw

istB

TWIST1

Actin

siCtr

l

(d)

00

02

04

06

08

10

12

14


Relat

ive e

xpre

ssio

n

(e)













Nontransfected

Forw

ard

scat

ter



60K

40K

20K

0

100

101

102

103

104

100

101

102

103

104

011Alexa-647 subset

992Alexa-647 subset

YTZ3-15 + siQ

(a)

Phas

eA

lexa

Fluo

r 488

Mer

ge

1 day 7 days

(b)

Merge

Phase

LysoTracker

GFP

siQ

(c)







00

02

04

06

08

10

12

14

Rela

tive e

xpre

ssio

n

Twist

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(a)

N-Cadherin Vimentin

00

02

04

06

08

10

12

14

Relat

ive e

xpre

ssio

n

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(b)

siQTw

A


(c)


150

100

50

0

SiQ TwA TwB

(d)



fflucIV

cont

rol

Intr

aven

ous

Intr

atum

oral


70

60

50

40

30

20

10

times106

times106

300

250

200

150

(a)

Kidney

Spleen

Liver

Tumor

Spleen

Kidney

Liver Tumors

Kidney

Spleen

Liver

Intr

aven

ous

Intr

atum

oral

IV co

ntro

l

50 100 150 200 times106

(b)




4 Conclusions







Acknowledgments


References







































































Pare

ntal

shSc

ram

shTw

istA

shTw

istB

TWIST1

Actin

(a)

0

02

04

06

08

1

12

14

SUM 1315parental

SUM 1315shScram

SUM 1315shTwistA

SUM 1315shTwistB

Relat

ive e

xpre

ssio

n

(b)


0ho

urs

96

hour

s

(c)

NTF

Lipo

onl

y

siTw

istA

siTw

istB

TWIST1

Actin

siCtr

l

(d)

00

02

04

06

08

10

12

14


Relat

ive e

xpre

ssio

n

(e)













Nontransfected

Forw

ard

scat

ter



60K

40K

20K

0

100

101

102

103

104

100

101

102

103

104

011Alexa-647 subset

992Alexa-647 subset

YTZ3-15 + siQ

(a)

Phas

eA

lexa

Fluo

r 488

Mer

ge

1 day 7 days

(b)

Merge

Phase

LysoTracker

GFP

siQ

(c)







00

02

04

06

08

10

12

14

Rela

tive e

xpre

ssio

n

Twist

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(a)

N-Cadherin Vimentin

00

02

04

06

08

10

12

14

Relat

ive e

xpre

ssio

n

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(b)

siQTw

A


(c)


150

100

50

0

SiQ TwA TwB

(d)



fflucIV

cont

rol

Intr

aven

ous

Intr

atum

oral


70

60

50

40

30

20

10

times106

times106

300

250

200

150

(a)

Kidney

Spleen

Liver

Tumor

Spleen

Kidney

Liver Tumors

Kidney

Spleen

Liver

Intr

aven

ous

Intr

atum

oral

IV co

ntro

l

50 100 150 200 times106

(b)




4 Conclusions







Acknowledgments


References

















































































Nontransfected

Forw

ard

scat

ter



60K

40K

20K

0

100

101

102

103

104

100

101

102

103

104

011Alexa-647 subset

992Alexa-647 subset

YTZ3-15 + siQ

(a)

Phas

eA

lexa

Fluo

r 488

Mer

ge

1 day 7 days

(b)

Merge

Phase

LysoTracker

GFP

siQ

(c)







00

02

04

06

08

10

12

14

Rela

tive e

xpre

ssio

n

Twist

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(a)

N-Cadherin Vimentin

00

02

04

06

08

10

12

14

Relat

ive e

xpre

ssio

n

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(b)

siQTw

A


(c)


150

100

50

0

SiQ TwA TwB

(d)



fflucIV

cont

rol

Intr

aven

ous

Intr

atum

oral


70

60

50

40

30

20

10

times106

times106

300

250

200

150

(a)

Kidney

Spleen

Liver

Tumor

Spleen

Kidney

Liver Tumors

Kidney

Spleen

Liver

Intr

aven

ous

Intr

atum

oral

IV co

ntro

l

50 100 150 200 times106

(b)




4 Conclusions







Acknowledgments


References







































































Nontransfected

Forw

ard

scat

ter



60K

40K

20K

0

100

101

102

103

104

100

101

102

103

104

011Alexa-647 subset

992Alexa-647 subset

YTZ3-15 + siQ

(a)

Phas

eA

lexa

Fluo

r 488

Mer

ge

1 day 7 days

(b)

Merge

Phase

LysoTracker

GFP

siQ

(c)







00

02

04

06

08

10

12

14

Rela

tive e

xpre

ssio

n

Twist

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(a)

N-Cadherin Vimentin

00

02

04

06

08

10

12

14

Relat

ive e

xpre

ssio

n

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(b)

siQTw

A


(c)


150

100

50

0

SiQ TwA TwB

(d)



fflucIV

cont

rol

Intr

aven

ous

Intr

atum

oral


70

60

50

40

30

20

10

times106

times106

300

250

200

150

(a)

Kidney

Spleen

Liver

Tumor

Spleen

Kidney

Liver Tumors

Kidney

Spleen

Liver

Intr

aven

ous

Intr

atum

oral

IV co

ntro

l

50 100 150 200 times106

(b)




4 Conclusions







Acknowledgments


References







































































00

02

04

06

08

10

12

14

Rela

tive e

xpre

ssio

n

Twist

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(a)

N-Cadherin Vimentin

00

02

04

06

08

10

12

14

Relat

ive e

xpre

ssio

n

siQ1 day 7 days

siRNATime

TwA TwBTwA TwB

(b)

siQTw

A


(c)


150

100

50

0

SiQ TwA TwB

(d)



fflucIV

cont

rol

Intr

aven

ous

Intr

atum

oral


70

60

50

40

30

20

10

times106

times106

300

250

200

150

(a)

Kidney

Spleen

Liver

Tumor

Spleen

Kidney

Liver Tumors

Kidney

Spleen

Liver

Intr

aven

ous

Intr

atum

oral

IV co

ntro

l

50 100 150 200 times106

(b)




4 Conclusions







Acknowledgments


References







































































fflucIV

cont

rol

Intr

aven

ous

Intr

atum

oral


70

60

50

40

30

20

10

times106

times106

300

250

200

150

(a)

Kidney

Spleen

Liver

Tumor

Spleen

Kidney

Liver Tumors

Kidney

Spleen

Liver

Intr

aven

ous

Intr

atum

oral

IV co

ntro

l

50 100 150 200 times106

(b)




4 Conclusions







Acknowledgments


References











































































Acknowledgments


References




























































































































Date post:	21-Nov-2023
Category:	Documents
Upload:	coh
View:	0 times
Download:	0 times

RNA-Based TWIST1 Inhibition via Dendrimer Complex to Reduce Breast Cancer Cell Metastasis

Documents