Contents lists available at ScienceDirect

Journal of Controlled Release

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

RNA-based micelles: A novel platform for paclitaxel loading and delivery

Yi Shub,1, Hongran Yina, Mehdi Rajabib,2, Hui Lia,b, Mario Viewegera, Sijin Guoa, Dan Shua,⁎,Peixuan Guoa,⁎

a Center for RNA Nanobiotechnology and Nanomedicine, Division of Pharmaceutics and Pharmaceutical Chemistry/College of Pharmacy, College of Medicine, Dorothy M.Davis Heart and Lung Research Institute and James Comprehensive Cancer Center, The Ohio State Universtiy, Columbus, OH 43210, United StatesbNanobiotechnology Center, Markey Cancer Center and Department of Pharmaceutical Sciences/College of Pharmacy, University of Kentucky, Lexington, KY 40536,United States

A R T I C L E I N F O

Keywords:pRNA three-way junctionRNA nanotechnologyRNA micellesSelf-assemblyDrug loadingSystemic delivery

A B S T R A C T

RNA can serve as powerful building blocks for bottom-up fabrication of nanostructures for biotechnological andbiomedical applications. In addition to current self-assembly strategies utilizing base pairing, motif piling andtertiary interactions, we reported for the first time the formation of RNA based micellar nanoconstruct with acholesterol molecule conjugated onto one helical end of a branched pRNA three-way junction (3WJ) motif. Theresulting amphiphilic RNA micelles consist of a hydrophilic RNA head and a covalently linked hydrophobic lipidtail that can spontaneously assemble in aqueous solution via hydrophobic interaction. Taking advantage of pRNA3WJ branched structure, the assembled RNA micelles are capable of escorting multiple functional modules. As aproof of concept for delivery for therapeutics, Paclitaxel was loaded into the RNA micelles with significantlyimproved water solubility. The successful construction of the drug loaded RNA micelles was confirmed andcharacterized by agarose gel electrophoresis, atomic force microscopy (AFM), dynamic light scattering (DLS),and fluorescence Nile Red encapsulation assay. The estimate critical micelle formation concentration rangesfrom 39 nM to 78 nM. The Paclitaxel loaded RNA micelles can internalize into cancer cells and inhibit theirproliferation. Further studies showed that the Paclitaxel loaded RNA micelles induced cancer cell apoptosis in aCaspase-3 dependent manner but RNA micelles alone exhibited low cytotoxicity. Finally, the Paclitaxel loadedRNA micelles targeted to tumor in vivo without accumulation in healthy tissues and organs. There is also no orvery low induction of pro-inflammatory response. Therefore, multivalence, cancer cell permeability, combinedwith controllable assembly, low or non toxic nature, and tumor targeting are all promising features that makeour pRNA micelles a suitable platform for potential drug delivery.

1. Introduction

Functional nanoparticles fabricated via molecular self-assemblyhold great promise for applications in biotechnology and biomedicine[1–4]. Among these, RNAs have been utilized as unique biomaterials toconstruct a wide variety of nanoparticles via self-assembly [3,5–16].These RNA nanoparticles have been demonstrated to be further func-tionalized for biomedical applications [2,17–22]. The strategies forRNA nanoparticles self-assembly were previously reported. ExploitingRNA base pairing and tertiary interactions provides versatile RNA as-semblies with accurate control over their composition, structure, andfunction at the nanometer scale [10,23]. In this present study, we ex-plore another novel strategy to fabricate RNA self-assemblies with mi-cellar properties through intermolecular hydrophobic interactions.

We have discovered and manufactured a stable phi29 pRNA three-way junction (3WJ) motif that can be used as a scaffold to constructmultivalent RNA nanoparticles with high chemical and thermodynamicstability [5,8]. The resulting branched RNA nanoparticles are uniformin size and shape. They can harbor different functionalities while re-taining their tertiary fold and independent functionalities both in vitroand in vivo [8,12,19,21,24–28]. We have also shown that fluorescentdye molecules [29,30] and specific cell targeting ligands [19–21,27,31]can be covalently attached onto RNA strands either through RNA solidphase synthesis or by post-transcriptional chemical conjugationmethods [7,10,32]. The feasibility of modifying RNA molecules withdifferent chemical and functional moieties proved the potential of RNAnanoparticles as a multi-functional drug delivery platform.

Further inspired by DNA micelle construction [33–42], we are able

https://doi.org/10.1016/j.jconrel.2018.02.014Received 7 October 2017; Received in revised form 8 February 2018; Accepted 9 February 2018

⁎ Corresponding authors at: Center for RNA Nanobiotechnology and Nanomedicine, 912 Biomedical Research Tower(BRT), 460 W 12th Ave., Columbus, OH 43210, United States.

1 Lester and Sue Smith Breast Cancer Center, Baylor College of Medicine, Houston, TX 77030, United States.2 The Pharmaceutical Research Institute(PRI), Albany College of Pharmacy and Health Science, NY 12144, United States.

E-mail address: guo.1091@osu.edu (P. Guo).

Journal of Controlled Release 276 (2018) 17–29

Available online 14 February 20180168-3659/ © 2018 Elsevier B.V. All rights reserved.

T

to convert a hydrophilic RNA molecule into an amphiphilic construct byfusing a lipophilic moiety onto one of pRNA-3WJ branched helicalends. This amphiphilic construct behaves similarly as a phospholipidwhile it spontaneously self-assembles into monodispersed, three-di-mensional micellar nanostructures with a lipid inner core and a bran-ched exterior RNA corona as the results of intermolecular hydrophobicinteractions in aqueous solution. Unlike current DNA micellar systems,which are limited in application by their mono-functionality [34,37],our pRNA micelles are capable of covalently attaching diverse types offunctional moieties to a single particle, including chemo-drugs forcancer therapeutics, imaging moieties for nanoparticles tracking, andco-delivered RNAi components for combination treatment.

Paclitaxel (PTX), isolated from the bark of Pacific Yew (Taxus bre-vifolia) [43], is one of the most effective chemotherapeutic drugs for anumber of cancer types [44–46]. The mechanism of paclitaxel forcancer treatment is to promote and stabilized microtubules and furtherinhibit G2 or M phases of the cell cycle followed by cell death [47].However, Paclitaxel has been classified as an IV chemical drugs ac-cording to the Biopharmaceutical classification system (BCS), because ithas both low water solubility (~0.4 μg/mL) and low permeability. Thefirst formulation of paclitaxel to be used was in a 1:1 (v:v) blend ofCremophor EL(polyethoxylated castor oil) and dehydrated ethanol di-luted 5–20 fold with 0.9% sodium chloride or 5% dextrose solution fori.v. administration [48]. However, formulation with Cremophor oil hasbeen observed to cause severe side effects [49] as well as unpredictablenon-linear plasma pharmacokinetics [50]. Therefore, alternative Pa-clitaxel formulations have been explored extensively, particularly withnanoparticle-based delivery systems [51]. Taking advantage of nano-scaled size, tumor targeted delivery, and biocompatibility, the en-capsulation of Paclitaxel in a nano-delivery system can increase drugcirculation half-life, lower its systemic toxicity, reduce side effects,improve pharmacokinetic and pharmacodynamic profiles, and demon-strate better patient compliance. Paclitaxel albumin-bound nano-particles (Abraxane®) have been approved by the FDA in 2005. Pacli-taxel liposomes (Lipusu®), polymeric micelles (Genexol PM®) andpolymeric conjugate with polyglutamate (Xyotax®) are currently com-mercial available paclitaxel formulations. In addition, there are varioustypes of Paclitaxel nanoparticle formulations either under developmentor in clinical trials, such as polymer-based nanoparticles [52–54], lipid-based nanoparticles [55,56], polymer-drug conjugates [57,58], in-organic nanoparticles [59], carbon nanotubes [60], nanocrystals [61],etc. However, formulating Paclitaxel in a RNA based nano-deliveryplatform has never been reported.

The present study demonstrates, to the best of our knowledge, thefirst design and construction of a well-defined RNA-micelles system toload Paclitaxel through conjugation to pRNA-3WJ-lipid. It is also thefirst time to report RNA-Paclitaxel micelles with significantly enhancedPaclitaxel water solubility and tumor permeability. The resulting RNAmicelles showed low critical micelle formation concentration (CMC),excellent cell binding and permeability, and efficient inhibition ofcancer cell proliferation by induction of Caspase-3 dependent cellapoptosis in vitro. Finally, we demonstrated that these drug-loaded RNAmicelles can be systemically delivered into tumor with advantageoustumor targeting, while minimizing retention of drug in healthy tissuesand crucial organs. There is also no or very low induction of pro-in-flammatory response. All of these findings indicate the strong potentialof the RNA-micelle system as a suitable drug delivery platform forcancer treatment. We are aimed at achieving tumor specific targeting,reducing the drug effective dose and further diminishing the adverseeffects of chemotherapy. Our ultimate goal is to explore the applicationof the RNA micelles as a robust and safe nano-delivery system to carryanti-cancer drugs for specific tumor targeting and treatment withminimum adverse effects to combat cancer and improve the life qualityof patients.

2. Materials and methods

2.1. Synthesis of paclitaxel conjugated RNA strand via “click chemistry”

To synthesize PTX-Azide, 1:2:2:1 eq. of Paclitaxel (Alfa Aesar), 6-Azido-hexanoic acid (Azido-HA) (Chem-IMPEX INT'L Inc), N, N′-Dicyclohexyl-carbodiimide (DCC) (Acros Organics), and 4-(Dimethylamino) pyridine (DMAP) (Sigma Aldrich) were weighed intoa two-neck round bottom flask. The reaction mixture was dissolved in~20mL dried Dichloromethane (DCM) and reacted at room tempera-ture while stirring under nitrogen atmosphere for 24 h. The reactionsolution was filtered and concentrated on a rotary evaporator. Theconcentrated reaction was further purified by slica gel chromatographyunder serial solvent wash with n-Hexane:ethyl acetate. Fractions con-taining pure product were combined and dried. The purified finalproduct PTX-Azide was characterized by Nuclear Magnetic Resonance(NMR) Spectroscopy.

5′-Alkyne-RNA was synthesized via standard RNA solid phase che-mical synthesis using 5′-Hexynyl phosphoramidite (Glen ResearchCorp.). The theoretical yield of labeled RNA strand is ~68%(0.9819= 0.68) with an average coupling efficiency of 98%.

To 5 μL of a 2mM RNA-Alkyne solution in water, 2 μL of PTX-Azide(50mM, 5 eq. in 3:1 (v/v) DMSO (Dimethyl sulfoxide, extra dried,Acros Organics)/tBuOH (tert-Butanol, anhydrous, Sigma Aldrich)), 3 μLof freshly prepared “Click Solution” containing 0.1 M CuBr (Copper(I)bromide, Sigma Aldrich) and 0.1M TBTA (Tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl]amine, Sigma Aldrich) in a 1:2 molar ratio in 3:1DMSO/tBuOH were added. The reaction mixture was thoroughly mixedand reacted under room temperature for 3 h. The success of the reactionwas determined by 20% 8M Urea PAGE in TBE (89mM Tris base-bo-rate, 2 mM EDTA) buffer. The reaction was subsequently diluted with100 μL 0.3M NaOAc (Sodium Acetate) and 1mL 100% Ethanol for RNAprecipitation. The precipitates were dissolved in water for purificationusing Ion-Pair Reverse Phase HPLC on an Agilent PLRP-S 4.6×250mm300A column. PTX-labeled RNA was separated from unreacted RNA-Alkyne in an Acetonitrile ramp. The column was equilibrated in 95%solvent A (0.1 M TEAA, HPLC grade H2O) and 5% solvent B (0.1MTEAA, 75% Acetonitrile, 25% HPLC grade H2O) at a flow rate of1.5 mL/min. The samples were filtered through a 0.2 μm spinfilter,loaded onto the column and then eluted by ramping from 5% to 100% Bover the course of 30min. Fractions were collected, combined andbuffer exchanged. The final RNA-PTX conjugate was characterized byMass Spectrometry.

2.2. Design and construction of 2′-F modified pRNA-3WJ-PTX micelles

The RNA micelles were constructed using a bottom-up approach[6]. The pRNA-3WJ-PTX micelles consisting of three fragments, a3WJ,b3WJ and c3WJ, were functionalized with Paclitaxel (Alfa Aesar) on a3WJ

5′-end (a3WJ-5′PTX), as a therapeutic module; Cholesterol on b3WJ 3′-end (b3WJ-3′chol), as lipophilic module; and Alexa647 (Alexa Fluor®647, Invitrogen) on c3WJ 3′-end (c3WJ-3′Alexa647), as a near-infra-red(NIR) imaging module. The control RNA nanoparticles are particleswithout therapeutic module named as pRNA-3WJ micelles, without li-pophilic module named as pRNA-3WJ-PTX, and without both lipophilicmodule and therapeutic module named as pRNA-3WJ.

The following pRNA-3WJ scaffold [6] sequences were used: a3WJ 5′-UUgCCaUgUgUaUgUggg-3′; b3WJ 5′-CCCaCaUaCUUUgUUgaUCC-3′;c3WJ 5′-ggaUCaaUCaUggCaa-3′ (upper case indicates 2′ Fluoro (2′-F)modified nucleotide). The RNA fragments were synthesized by standardsolid phase chemical synthesis [60] with commercially availablephosphoramidite monomers of 2′-TBDMS Adenosine (n-bz) CED, 2′-TBDMS Guanosine (n-ibu) CED, 2′-Fluoro Cytidine (n-ac) CED, and 2′-Fluoro Uridine CED, followed by deprotection according to manu-facturer (Azco Biotech) provided protocol. Paclitaxel (PTX) was con-jugated onto a3WJ 5′-end by chemoselective Cu(I)-catalyzed Huisgen

Y. Shu et al. Journal of Controlled Release 276 (2018) 17–29

18

1,3-dipolar cycloaddition reaction (“Click chemistry”) and purifiedthrough reverse phase HPLC as mentioned above. Cholesterol was at-tached to the 3′-end of b3WJ strand by using 3′-Cholesteryl-TEG CPGsupport (Glen Research Corp.), following manufacturer instructions.The theoretical yield of b3WJ-3′chol is ~68% (0.9819= 0.68) with anaverage coupling efficiency of 98%. Synthesized RNA is purified fromterminated strands using an ion-pair reverse phase column which ty-pically yields complete labeling with cholesterol during solid phasesynthesis. Alexa647 labeled RNA strand was purchased from TriLinkBio Technologies, LLC.

pRNA-3WJ-PTX micelles were assembled by mixing 3WJ strands(a3WJ, b3WJ and c3WJ) at equal molar concentrations in TMS buffer(50mM Tris pH=8.0, 100mM NaCl, 10mM MgCl2) or PBS buffer(137mM NaCl, 2.7 mM KCl, 10mM Na2HPO4 and 2mM KH2PO4,pH 7.4) followed by heating to 80 °C for 5min and slowly cooling to37 °C over the course of 40min with a subsequent 1 h incubation at37 °C.

2.3. Characterization of the assembled pRNA-3WJ micelles

The assembly of the functionalized 3WJ nanoparticles was char-acterized by 1% (w/v) agarose gel shift assays in TAE (40mM Tris-acetate, 1 mM EDTA) buffer. After electrophoresis, the gel was stainedby ethidium bromide and visualized by Typhoon FLA 7000 (GEhealthcare).

The apparent hydrodynamic diameters of preassembled pRNA-3WJ-PTX (20 μM in PBS buffer) and pRNA-3WJ-PTX micelles (10 μM in PBSbuffer) were measured by Zetasizer nano-ZS (Malvern Instrument, LTD)at 25 °C, respectively. The data were obtained from three independentmeasurements. The Zeta potential of pRNA-3WJ-PTX micelles (1 μM inPBS buffer) was also mearsured by Zetasizer nano-ZS (MalvernInstrument, LTD) at 25 °C.

The size and shape of the RNA micelles were also determined byAtomic Force Microscopy (AFM). For AFM, 10 μM solution of the pRNA-3WJ-PTX micelles in TMS buffer were deposited onto freshly cut micasurface and dried overnight. After two consecutive rinsing steps withHPLC grade water, the mica was mounted onto Bruker Multimode IVAFM and imaged in taping mode.

Successful formation of the RNA micelles was also determined by aNile Red encapsulation assay as reported previously [62,63]. A 5mMstock solution of Nile Red in acetone was used for all experiments.Briefly, 0 μM, 1 μM, 5 μM, and 10 μM of assembled pRNA-3WJ micelleswere incubated with 100 μM of Nile Red in TMS buffer, respectively.The mixture was heated to 80 °C for 5min and slowly cooled to 37 °Cover 40min followed by 1 h incubation at 37 °C. The fluorescence in-tensity of Nile Red versus RNA micelles concentration was measured byFluorolog spectrofluorometer (Horiba Jobin Yvon) with an excitationwavelength of 535 nm and emission spectra taken from 560 nm to760 nm. The pRNA-3WJ was used as a control.

2.4. Determination of critical micelle formation concentration (CMC)

CMC of pRNA-3WJ micelles was determined by both fluorescentNile Red encapsulation assay as previously reported [63] and agarosegel electrophoresis. Briefly, 2-fold serial diluted RNA micelle samples(in the range of 5 μM to 0.005 μM) were incubated with 100 μM of NileRed in a final volume of 50 μL. The samples were heated to 80 °C for5min and slowly cooled to 37 °C over 40min followed by 1 h incuba-tion at 37 °C. The fluorescence intensity of Nile Red versus RNA micellesconcentration was measured by Fluorolog spectrofluorometer (HoribaJobin Yvon) with an excitation wavelength of 535 nm and emissionspectra taken from 560 nm to 760 nm. 2-fold serial diluted RNA micellesamples (in the range of 2.5 μM to 0.0390625 μM) were directly loadedonto a 1% (w/v) agarose gel for electrophoresis under 120 V in TAEbuffer. The gel was stained by ethidium bromide and visualized byTyphoon FLA 7000.

2.5. Formulation stability assay

Final 2 μM of pRNA-3WJ micelles were incubated in acidic(pH=4), neutral (pH=7.4), and basic (pH=12) buffers at 37 °C for1 h. Final 2 μM of pRNA-3WJ micelles were also incubated at differenttemperatures (4 °C, 37 °C, and 65 °C) for 1 h. All samples after incuba-tion were loaded onto a 1% (w/v) agarose gel for electrophoresis under120 V in TAE buffer. The gel was stained by ethidium bromide andvisualized by Typhoon FLA 7000.

2.6. Cell culture

Human KB cells (American Type Culture Collection, ATCC) weregrown and cultured in RPMI-1640 (Thermo Scientific) containing 10%fetal bovine serum(FBS) in a 37 °C incubator under 5% CO2 and a hu-midified atmosphere. Mouse macrophage-like RAW 264.7 cells weregrown in Dulbecco's Modified Eagle's Medium (DMEM) supplementedwith 10% FBS, 100 units/mL penicillin and 100mg/mL streptomycin at37 °C in humidified air containing 5% CO2.

2.7. In vitro binding assay using flow cytometry

250 nM, 500 nM, and 1 μM Alexa647 labeled pRNA-3WJ-PTX mi-celles and the control pRNA-3WJ-PTX nanoparticles without lipophilicmodule were each incubated with 2× 105 KB cells at 37 °C for 1 h.After washing with PBS twice, the cells were resuspended in PBS. FlowCytometry was performed by the University of Kentucky FlowCytometry & Cell Sorting core facility to observe the cell binding effi-cacy of the Alexa647 labeled pRNA-3WJ-PTX micelles. The data wasanalyzed by FlowJo 7.6.1 software.

2.8. In vitro binding and internalization assay using confocal microscopy

KB cells were grown on glass slides overnight. 1 μM Alexa647 la-beled pRNA-3WJ-PTX micelles and the control pRNA-3WJ-PTX nano-particles without lipophilic module were each incubated with the cellsat 37 °C for 1 h. After washing with PBS, the cells were fixed by 4%paraformaldehyde (PFA) and washed 3 times by PBS. The cytoskeletonof the fixed cells was treated with 0.1% Triton-×100 in PBS for 5min toimprove cell membrane permeability and then stained by Alexa Fluor488 Phalloidin (Life Technologies) for 30min at room temperature andthen rinsed with PBS for 3×10min. The cells were mounted withProlong® Gold antifade reagent with DAPI (Life Technologies) and DAPIwas used for staining the nucleus. The cells were then assayed forbinding and cell entry by FluoView FV1000-Filter Confocal MicroscopeSystem (Olympus Corp.).

2.9. In vitro drug releasing assay for pRNA-3WJ-PTX

10 μM of The pRNA-3WJ-PTX conjugate was incubated with 50%FBS at 37 °C and samples were collected at different time points (1 h,4 h, 8 h, 12 h, 15 h, 22 h, 24 h, 28 h, and 36 h) and froze immediately at−80 °C. After completing sample collection through entire time course,all samples were run on 15% native PAGE in TBM buffer (89mM Trisbase-borate, 5 mM MgCl2). The gel bands were quantified by Image Jand the percentage of intact particle was calculated by intact particle%= [(intensity of upper band) / (intensity of upper band+ lowerband)]× 100%. Upper band indicates intact RNA-drug conjugate whilelower band indicates RNA oligo after drug releasing.

2.10. MTT assay

In order to assay the cytotoxicity effects after pRNA-3WJ-PTX mi-celles treatment, CellTiter 96 Non-Radioactive Cell Proliferation Assay(Promega) was used to assay cell viability changes following manu-facturer instructions. Briefly, 1×104 KB cells were seeded into 96-well

Y. Shu et al. Journal of Controlled Release 276 (2018) 17–29

19

plates a day prior to the assay. On the second day, 1 μM, 800 nM,600 nM, 400 nM, and 200 nM of pRNA-3WJ-PTX micelles were addedto the wells in triplicate. Paclitaxel alone, pRNA-3WJ micelles, pRNA-3WJ-PTX, and pRNA-3WJ were used as controls in the same testingconcentrations. The plate was incubated at 37 °C for 48 h in a humidi-fied, 5% CO2 atmosphere. After incubation, 15 μl of the Dye Solutionwas added to each well and the plate was incubated at 37 °C for up to4 h in a humidified, 5% CO2 atmosphere. Next, 100 μL of theSolubilization Solution/Stop Mix was added to each well and incubatedfor 1 h. Finally, the contents of the wells were mixed to get a uniformlycolored solution and their absorbance at 570 nm was recorded usingSynergy 4 microplate reader (Bio-Tek).

2.11. Apoptosis study in in vitro cell model

FITC Annexin V Apoptosis Detection Kit (BD Pharmingen) andCaspase-3 Assay Kit (BD Pharmingen) were used as previously reported[64] to study cell apoptosis induced by pRNA-3WJ-PTX micellestreatment. For FITC Annexin V staining assay, KB cells were seeded on12-well plates overnight. Cells (~30% confluence) were then treatedwith 1 μM of pRNA-3WJ-PTX micelles. The controls include Paclitaxel,

pRNA-3WJ micelles, and pRNA-3WJ. According to manufacturer's in-struction, 48 h after incubation with the RNA micelles, KB cells weretrypsinized to single cell suspension. After two times PBS wash, the cellswere re-suspended in 100 μL 1× Annexin V-FITC binding buffer. Then5 μL AnnexinV-FITC and 5 μL propidium iodide (PI) were added intoeach sample and incubated at room temperature for 25min. The sam-ples were finally added to a flow tube which contained 200 μL 1×binding buffer for FACS analysis within 1 h.

For Caspase-3 assay, KB cells were seeded on 24-well plates over-night. Cells (~30% confluence) were then treated with 1 μM of pRNA-3WJ-PTX micelles. The controls include Paclitaxel, pRNA-3WJ micelles,and pRNA-3WJ. The cellular Caspase-3 activity was measured andcompared by Caspase-3 Assay Kit (BD Pharmingen) according to man-ufacturer's instruction. Briefly, Cell lysates (1–10× 106 cells/mL) atdifferent time points (4 h, 8 h, 12 h, and 24 h) after induction of apop-tosis were prepared using cold Cell Lysis Buffer provided by the kit, andincubated for 30min on ice. For each sample, 25 μL of cell lysate wasadded with 2 μL reconstituted Ac-DEVD-AMC in 80 μL of 1× HEPESbuffer and incubated at 37 °C for 1 h. The amount of AMC liberatedfrom Ac-DEVD-AMC was measured at an excitation wavelength of380 nm over an emission wavelength range of 400–500 nm on a

Fig. 1. Structure base and assembly principle of pRNA-3WJ-PTX micelles. A. pRNA-3WJ motif from bacteriophage phi29 packaging RNA. B. The angled branch structure of pRNA-3WJ. C.Conjugating pRNA-3WJ with a lipophilic module (cholesterol, blue), a therapeutic module (PTX, green), and a reporter module (Alexa dye, red). D. Illustration of the formation of pRNA-3WJ micelles via hydrophobic interaction of conjugated lipophilic module in aqueous solutions. E. Assay the assembly of pRNA-3WJ micelles by 1% TAE agarose gel electrophoresis.Upper gel: EtBr channel; lower gel: Alexa647 channel (M: 1 kb plus DNA ladder). (For interpretation of the references to color in this figure legend, the reader is referred to the webversion of this article.)

Y. Shu et al. Journal of Controlled Release 276 (2018) 17–29

20

Fluorolog spectroflurometer (Horiba Jobin Yvon).

2.12. Animal models

All protocols involving animals are performed under the supervisionof the University of Kentucky Institutional Animal Care and UseCommittee (IACUC). To generate xenograft model, female athymic nu/nu mice, 4–8weeks old, were purchased from Taconic. Subcutaneoustumor xenografts were established by injecting 2×106 cells/site KBcells resuspended in sterile PBS into the left shoulder of nude mice.When the tumor nodules had reached a volume of 50mm3, approxi-mately 5 days post-injection, the mice were used for tumor targetingstudies.

2.13. NIR fluorescence imaging to detect the targeting of RNA micelles tocancer xenografts in vivo

To investigate the delivery of pRNA-3WJ-PTX micelles in vivo, a

fluorescence imaging study was performed after tail vein injection of100 μL 20 μM Alexa 647 labeled pRNA-3WJ-PTX micelles and pRNA-3WJ-PTX (estimated final concentration in blood is about 1 μM) intomice bearing KB tumor, respectively. PBS injected mice were used asfluorescence negative controls. The whole body images were taken at30min, 1 h, 2 h, 4 h, 8 h, 24 h. Mice were sacrificed at 24 h post injec-tion by inhalation of CO2 followed by cervical dislocation, and majorinternal organs including heart, lungs, liver, spleen, kidneys togetherwith tumor from the sacrificed mice were collected and subjected tofluorescence imaging for assessment of biodistribution profiles usingIVIS Spectrum station (Caliper Life Sciences) with excitation at 640 nmand emission at 680 nm.

2.14. Evaluation of pro-inflammatory induction of pRNA-3WJ micelles

For in vitro evaluation of pro-inflammatory cytokines induction,2.5× 105 RAW 264.7 cells per well were seeded in 24-well plate andcultured overnight. pRNA-3WJ micelles (1 μM or 200 nM) as well as

Fig. 2. Characterization of pRNA-3WJ micelles. A. AFM image. Scale bar: 200 nm. B. The apparent hydrodynamic diameters measurement by DLS. C. The Zeta potential measurement byDLS. D. Validate the assembly of pRNA-3WJ micelles via Nile Red binding assay. (For interpretation of the references to color in this figure legend, the reader is referred to the web versionof this article.)

Y. Shu et al. Journal of Controlled Release 276 (2018) 17–29

21

Lipopolysaccharide (LPS, 3.6 μg/mL, equal amount as 200 nM pRNA-3WJ micelles) were diluted in DMEM (Life Technologies) and added tocells for incubation in triplicate. After 16 h incubation, cell culture su-pernatants were collected and stored at −80 °C immediately for furtherassay. TNF-α and IL6 in collected supernatants were examined by usingMouse ELISA MAX Deluxe sets (BioLegend) while IFN-α was examinedusing Mouse IFN Alpha ELISA Kit (PBL Assay Science), following themanufacturers' instruction respectively.

For in vivo evaluation of pro-inflammatory cytokines and chemo-kines induction, pRNA-3WJ micelles (1 μM), LPS (10 μg per mouse),and DPBS control were injected into 4–6weeks male C57BL/6 mice viatail vein. 3 h post-injection, blood samples were harvested from mice bycardiac puncture and centrifuged at 12,800×g for 10min to collectserum. Concentrations of TNF-α, IL6 and IFN-α in serum were ex-amined by ELISA as described above following manufactures' instruc-tion. Chemokine induction was conducted by using Mouse ChemokineArray Kit (R&D Systems) following manufacturer's instruction. LiCorwas used for blotting spot quantification.

2.15. Statistical analysis

Each experiment was repeated 3 times in duplicates for each sample

tested. The results are presented as mean ± SD unless otherwise in-dicated. Statistical differences were evaluated using Student's t-test, andp < 0.05 was considered significant.

3. Results and discussion

3.1. Construction and characterization of pRNA-3WJ micelles

The pRNA-3WJ-PTX micelles utilized a modular design composingof three short RNA fragments from pRNA 3WJ motif [6] (Fig. 1A). Thelipophilic module, cholesterol, was conjugated to the 3′-end of b3WJ byconsidering the global folding structure of the pRNA 3WJ motif. Thecrystal structure of the pRNA 3WJ motif revealed that the angles amongthree helices (H1, H2, and H3) of the pRNA 3WJ are ~60° (H1−H2),~120° (H2−H3), and ~180° (H1−H3) [65] (Fig. 1B). In order to avoidinterference of micelle formation by steric hindrance due to the bran-ched 3WJ structure, the cholesterol molecule was placed onto H3 whereit is furthest from the other two helices, H1 and H2. Our current designfunctioned H1 with therapeutic module (Paclitaxel) and imagingmodule (Alexa 647 dye) without interfering in micelle formation(Fig. 1C, Fig. 1D). The functions of conjugated chemotherapeutic drugand detection dye were also well retained as shown in the following

Fig. 3. RNA-PTX conjugates. A. The rational design of RNA-PTX conjugates. PTX-N3 can react with end Alkyne labeled RNA via Click chemistry and PTX can be later released out fromRNA strand by hydrolysis. B. Assay of successful RNA-PTX conjugation by 20% 8M Urea PAGE in TBE buffer. C. The experimental mass prediction of a3WJ-PTX conjugates by MassSpectrometry. D. In vitro PTX releasing profile along time.

Y. Shu et al. Journal of Controlled Release 276 (2018) 17–29

22

study. The unoccupied helix H2 can be further activated with RNAimodules, such as siRNA or microRNA, for combined treatment to gen-erate enhanced or synergetic therapeutic effects.

Upon mixing the individual strands in equal molar ratio in PBS orTMS buffer, the complex assembled with high efficiency, as shown inthe 1% Agarose gel shift assay (Fig. 1E). Alexa 647 labeled RNA mi-celles also showed fluorescent bands corresponding to the band in-dicating successful micelle formation (Fig. 1E). The assembly of theRNA micelles was further demonstrated by AFM which showinghomogeneous sphere-shaped architectures (Fig. 2A). DLS assaysshowed that the average hydrodynamic diameter of pRNA-3WJ-PTXmicelles was 118.7 ± 14.50 nm compared to 7.466 ± 1.215 nm forpRNA-3WJ core scaffold (Fig. 2B). The constructed pRNA-3WJ-PTXmicelles were also negatively charged. The Zeta potential was−26.1 ± 10.4mV as shown in Fig. 2C. Finally, the RNA micelles for-mation was also assayed by a Nile Red encapsulation assay. Nile Red isa hydrophobic dye and nearly non-emissive in bulk aqueous solution,but its inclusion in a nonpolar microenvironment such as the lipid coreof the micellar structure results in an intense fluorescence enhancement[66]. Therefore, the increase in fluorescence intensity associated withthe incubation of RNA micelles at varying concentrations with fixamounts of Nile Red dye indicated the formation of the RNA micelles inthe buffer solution (Fig. 2D). In contrast, no significant increase offluorescent intensity was observed in the control pRNA-3WJ without alipid core (Fig. 2D).

To make the pRNA-3WJ micelles chemically stable in vivo, we used2′-F modified U and C nucleotides [67,68] during RNA strand synthesis.The 2′-F modified RNA nanoparticles were proved to be chemicallystable and exhibit longer half-life in circulation compared to its

unmodified RNA counterparts [6,67]. The presence of 2′-F nucleotidesnot only makes the RNA nanoparticles resistant to RNase degradation,but also enhances the melting temperature of pRNA-3WJ [69], withoutcompromising the authentic folding and functionalities of the core andincorporated modules [6,70].

The pRNA-3WJ formulation stability was further assayed versus pH(acidic pH 4, neutral pH 7.4, and basic pH 12) and temperature (4 °C,37 °C, and 65 °C). The results indicated that pRNA-3WJ micelles werestable across a wide range of temperature and in acidic and neutralcondition. Although pRNA-micelles showed partial dissociation pro-blem at basic condition as reported in Supplementary Fig. 1, pRNA-3WJmicelles formulation was chemically and thermodynamically stable inphysiological condition (pH 7.4, 37 °C).

3.2. Synthesis of paclitaxel conjugated RNA strand via “click chemistry”

In order to load pRNA micelles with therapeutic module, PTX wasfirstly functionalized with -Azide (-N3) in order to further react withalkyne modified RNA. An azide group was introduced to 2'-OH positionof PTX using 6-azidohexanoic acid linker via esterification, in the pre-sence of N,N′-dicyclohexylcarbodiimide (DCC) and 4-(dimethylamino)pyridine (DMAP) in dry dichloromethane to afford azide-functionalizedPTX (PTX-N3) as the predominant product (Supplementary Fig. 2A).Although both hydroxyl groups at the 7′ and 2′ positions of PTX arechemically reactive, the 2′-OH group typically showed higher reactivitythan the 7′-OH group, because acetylation of 7′-OH group is very un-stable in aqueous media, losing the C-7 substituent rapidly [71]. PTX-N3 with functionality at the 2′ position was obtained in 73% yield afterpurification. The 1H NMR spectroscopy was employed to confirm the

Fig. 4. Assay tumor cell binding and internalization of pRNA-3WJ-PTX micelles in vitro by A. Flow Cytometry and B. Confocal microscopy. Blue: nuclei staining; Green: cytoskeletonstaining; Red: pRNA-3WJ-PTX micelles binding. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Y. Shu et al. Journal of Controlled Release 276 (2018) 17–29

23

chemical structure of the PTX-N3 in CDCl3. In the 1H NMR spectrum ofPTX-N3, the 7′-CH resonates at 4.4 ppm and no significant chemicalshift change was observed for the 7-CH-OH signal (at 4.40 ppm) beforeand after esterification (Supplementary Fig. 2B) while the resonance of2′-CH proton was shifted from 4.78 ppm (before esterification) to5.46 ppm (after esterification) (Supplementary Fig. 2C).

PTX-N3 reacted with - Alkyne modified RNA (a3WJ) at above 90%efficiency via Copper (I) mediated Click reaction as shown in Fig. 3A.The successful conjugation was confirmed by 20% 8M Urea PAGE(Fig. 3B) and the Mass Spectrometry. The experimental mass of a3WJ-PTX determined by Mass Spectrometry is 6939.2 (m/z) which is close tothe theoretic masses (6936.82 (m/z)) calculated based on the chemicalformula (Fig. 3C). The ester bond formed between Paclitaxel and RNAcan then be hydrolyzed in presence of either esterase or in aqueoussolution, which allows slow releasing of the loaded drug in a con-trollable manner as indicated by in vitro drug releasing assay (Fig. 3A,D). This is the first report to “click” a RNA molecule with a che-motherapeutic drug, Paclitaxel. PTX conjugation did not interfere withthe stepwise assembly of pRNA-3WJ (Supplementary Fig. 3A). It is also

the first time demonstrating that the water insolubility of Paclitaxel canbe altered by fusing with a RNA oligo. 1 mM of RNA-Paclitaxel con-jugates dissolved in DEPC H2O showed clear solution compared tocloudy solution prepared using equal concentration of Paclitaxel inDEPC H2O (Supplementary Fig. 3B). The significantly improved watersolubility of Paclitaxel after conjugation with RNA oligo and assemblyinto micellar nanostructures allows the use of normal saline solution forin vivo administration instead of Cremophor EL formulation. In additionto Click chemistry, functionalizing RNA oligos and candidate drugswith other chemical coupling reactive groups, such as –NH2/-NHS–NH2/-COOH, or –SH/-maleimide, are also feasible RNA-drug con-jugation approaches which can be applied for drug candidates that arenot suitable for Click chemistry.

3.3. Determine the critical micelle formation concentration

In order to determine the Critical Micelle Formation Concentration(CMC) of pRNA-3WJ-PTX micelles, the Nile Red assay was used aspreviously reported [63]. Nile red is a hydrophobic dye which has low

Fig. 5. Cytotoxicity and apoptotic effects of PTX loaded pRNA-3WJ micelles in vitro. A. Assay cytotoxicity effects of pRNA-3WJ-PTX micelles by MTT assay. Error bars represent StandardDeviation. B. Assay apoptotic effects of pRNA-3WJ-PTX micelles by PI/Annexin V-FITC dual staining and FACS analysis. C. Caspase-3 assay.

Y. Shu et al. Journal of Controlled Release 276 (2018) 17–29

24

fluorescence in water and other polar solvents but emits strong fluor-escence in nonpolar environments such as the lipid core of the micellarstructure. Therefore, Nile Red fluorescence emission intensity is utilizedas an indicator for micellar structure formation. To determine CMC, thefluorescence intensity of Nile Red was plot as a function of the sampleconcentration. As shown in Supplementary Fig. 4A, Nile Red exhibitedlow fluorescence intensity at the concentration below 0.078 μM in-dicating that the Nile Red was in water and few micelles were present.With increasing concentration, the fluorescence intensity increaseddramatically demonstrating that Nile Red was encapsulated in the lipidcore of RNA micelles. The CMC can be estimated as the range from39 nM to 78 nM since the Nile Red fluorescence intensity exhibits ob-vious increase starting at 39 nM to 78 nM (Supplementary Fig. 4B). 1%TAE agarose gel further confirmed the CMC is ~150 nM due to sensi-tivity limit (Supplementary Fig. 4C). All the pRNA micelles con-centration in vitro and in vivo experiments we performed is above CMCto ensure all the micelles are formed.

3.4. Binding and internalization of pRNA-3WJ-PTX micelles into KB cells

In order to assay the tumor cell targeting in vitro, the pRNA-3WJ-PTX micelles and control pRNA-3WJ-PTX without lipophilic modulewere incubated with KB cells, trypsinized, washed and then analyzed byFluorescence-Activated Cell Sorting (FACS) assay. Strong binding (al-most 100%) was observed for pRNA-3WJ-PTX micelles compared topRNA-3WJ-PTX scaffold control (0% binding) (Fig. 4A). Confocal mi-croscopy images further confirmed the efficient binding and inter-nalization of pRNA-3WJ-PTX micelles into cancer cells, as demon-strated by excellent overlap of fluorescent RNA nanoparticles (red colorin Fig. 4B) and cytoplasm (green color in Fig. 4B). Very low signal wasobserved for control pRNA-3WJ-PTX without lipophilic modules. Theseresults indicated that the RNA micelles had high affinity for cancer cellbinding. Although the current RNA micelles design did not include atumor specific targeting module, the highly charged RNA micelles (asRNA is negatively charged, also shown in Zeta potential study) wereable to disintegrate themselves and insert into the cell membrane wheninteracting with the cells similar as DNA micelles as reported previously[35]. We suspected the internalization of the RNA micelles was medi-ated possibly by subsequent endocytosis after inserting into the cell

membrane.

3.5. Effects of pRNA-3WJ-PTX micelles on growth and apoptosis of cancercells in vitro

In order to determine the cellular effects of RNA micelles treatment,MTT assay was performed to assay the cell viability after treatment. TheRNA micelles harboring PTX can successfully inhibit tumor cell growthat or above 250 nM, compared to RNA micelles only and 3WJ controlswithout Paclitaxel conjugation (Fig. 5A). We did not test the con-centrations below 125 nM in order to keep experiment condition higherthan CMC. Cell growth inhibition caused by pRNA-3WJ-PTX micellesindicated release of the Paclitaxel from RNA strands due to the linkerester hydrolysis in aqueous solution. There was very little possibility offree Paclitaxel contamination to the testing constructs because of all theRNA-PTX conjugates went through thorough HPLC purification. Inaddition, pRNA-3WJ micelles without Paclitaxel loading showed noeffects on cell viability and proliferation which indicating the low cy-totoxicity of RNA micelles scaffolds and their potential as a safe drugdelivery platform.

FITC labeled Annexin V staining confirmed that the majority of thecancer cell death was due to cell apoptosis as shown in Fig. 5B. Loss ofplasma membrane is one of the earliest features in apoptotic cells. Theexternalization of membrane phospholipid phosphatidylserine (PS)from the inner to the outer leaflet of the plasma membrane allows FITClabeled Annexin V to bind PS for detecting cells that are undergoingapoptosis. Staining with FITC Annexin V is used in conjunction with PIto identify early apoptotic cells (Q3: PI negative, FITC Annexin V po-sitive) and cells that are in late apoptosis or already dead (Q2: FITCAnnexin V and PI positive). In our study, over 30% of the cells wereundergoing apoptosis after 48 h treatment with pRNA-3WJ-PTX mi-celles compared to the control micelles without PTX (4.52%) or pRNA-3WJ without PTX (3.7%), indicating that the cancer viability change isdue to the induction of cell apoptosis.

Caspase-3 is an early cellular apoptotic marker. The increase of theCaspase-3 activity is closely related to cell apoptosis. The elevation ofCaspase-3 activity, as reflected by increased fluorescence intensity,appeared 12 h after treatment with pRNA-3WJ-PTX micelles(Supplementary Fig. 5). We found that pRNA-3WJ-PTX micelles treated

Fig. 6. In vivo tumor targeting of pRNA-3WJ-PTX micelles in mouse xenografts. A. whole body image obtained 4 h post injection. B. Organ image obtained 24 h post injection.

Y. Shu et al. Journal of Controlled Release 276 (2018) 17–29

25

cell lysates showed the highest fluorescence emission similar toPaclitaxel alone compared to the control nanoparticles (pRNA-3WJmicelles and pRNA-3WJ), which indicate the induction of cell apoptosisin a Caspase-3 dependent manner (Fig. 5C).

3.6. Specific targeting of tumors in xenografted animal models via imagingusing NIR fluorescent pRNA-3WJ-PTX micelles

Tumor targeting efficiency by pRNA-3WJ-PTX micelles was in-vestigated by collecting in situ fluorescence images of tumor xenografts

in nude mice at different post injection time points. Images of the tumorarea became readily defined 4 h post injection (Fig. 6A, SupplementaryFig. 6). Ex vivo images of normal tissues, organs, and tumors taken fromthe RNA micelles-injected mice showed that the tumors harvested 24 hpost injection exhibited the strongest signal (Fig. 6B). In terms of tumoraccumulation kinetics, RNA nanoparticles reached their highest accu-mulation 4 h post injection and remained longer in the tumor comparedto healthy organs and tissues, which indicates a high tumor targetingefficiency and tumor retention capability of the constructed RNA mi-celles. Such distinct tumor retention behavior of RNA micelles

Fig. 7. Assay inductions of pro-inflammatory cytokines and chemokines by pRNA-3WJ micelle formulation. A. In vitro evaluation of the TNF-α, IL6, and IFN-α production after incubatingpRNA-3WJ micelles with mouse macrophage-like RAW 264.7 cells by ELISA assay. Error bars represent Standard Deviation. B. In vivo evaluation of the TNF-α, IL6, and IFN-α productionafter injecting pRNA-3WJ micelles into C57BL/6 mice by ELISA assay. Error bars represent Standard Deviation. C. In vivo chemokines induction profiling for pRNA-3WJ micelleformulation.

Y. Shu et al. Journal of Controlled Release 276 (2018) 17–29

26

suggested this delivery system makes the advantage of the EPR (en-hanced permeability and retention) effect due to its nano-scale size andparticle shape. In addition, the RNA micelles constructs showed pro-longed tumor retention possible due to its larger particle size. Thespecificity of in vivo tumor targeting of RNA micelle nanoparticle can befurther secured by including a tumor targeting modules, such as folate[6,27,65,72,73] and RNA aptamers [19,21,74], to the empty helicalbranch on the RNA micelles.

3.7. No or low induction of pro-inflammatory response by pRNA-3WJmicelles

Pro-inflammatory response could potentially be induced by bothRNA component [75] and cholesterol component [76] in pRNA-3WJmicelles formulation. In order to address this concern, we evaluated theproduction of pro-inflammatory cytokines and chemokines upon pRNA-3WJ micelles treatment both in vitro and in vivo. Tumor necrosis factor-α (TNF-α) is cytokine involved in systemic inflammation and one of thecytokines that make up the acute phase reaction [77]. Interleukin 6(IL6) is an interleukin that acts as both a pro-inflammatory cytokineand an anti-inflammatory myokine. IL6 is secreted to stimulate immuneresponse during infection [78,79]. IFN-α, which belongs to type I in-terferon, is also cytokines involved in pro-inflammatory reaction re-leased in response to the presence of viral pathogens. The results inFig. 7A showed that pRNA-3WJ micelles at higher dose (1 μM) andlower dose (200 nM) did not induce either IL6 or IFN-α productioncompared to LPS positive control while incubating with mouse mac-rophage-like RAW 264.7 cells in vitro. TNF-α induction was not de-tectable with lower dose of pRNA-3WJ-micelles treatment but wasslightly raised up upon higher dose of RNA micelles treatment. All threecytokines were not induced after in vivo injection of pRNA-3WJ-mi-celles into immune competent C57BL/6 mice compared to LPS controlas shown in Fig. 7B.

Chemokines are the main proinflammatory mediators [80,81]. Weprofiled 25 chemokine production upon pRNA-micelles treatment invivo. From the result as shown in Fig. 7C, we can conclude that pRNA-3WJ micelles did not induce new chemokines compared to PBS group.Three chemokines, macrophage inflammatory protein-1 gamma (MIP-1γ), Chemokine10 (C10), and Monocyte chemoattractant protein 2(MCP2) showed elevated induction compared to PBS control as high-lighted by red square in the Supplementary Fig. 7. In summary, pRNA-3WJ micellar formulation induced no or very low pro-inflammatoryresponse.

In conclusion, we have designed and constructed well-defined pRNAbased micelles composed of a hydrophobic lipid core and a hydrophilicpRNA-3WJ corona. Chemotherapeutic drug Paclitaxel has been loadedinto RNA micelles with significantly improved water solubility. We alsodemonstrated that the Paclitaxel loaded pRNA micelles exhibit ex-cellent tumor cell binding and internalization as well as effective in-duction of cytotoxicity effects on tumor cells as PTX in vitro. There isalso no or very low induction of pro-inflammatory responses uponpRNA-micelles injection. Tumor targeting was achieved upon systemicinjection of pRNA micelles into xenografted mouse models withoutaccumulation into normal organs and tissues. The branched pRNA-3WJcorona gives incomparable versatility that can be designed and con-structed in any desired combination. For example, the multifunctionalpRNA-3WJ micelles can be achieved by combing targeting, imaging andtherapeutic modules all in one nanoparticle. pRNA-3WJ micelles canalso simultaneously deliver siRNAs or microRNAs against multiplegenes or different location of one gene to generate synergetic effects. Inaddition, different types of anti-cancer drugs can be loaded onto onepRNA-3WJ micelles to enhance the therapeutic effect or overcome thedrug resistance by combination therapy. Therefore, this innovative RNAbased micellar nano-delivery platform holds great potential for clinicalapplications.

Acknowledgement

The research in P.G.'s lab was supported by the National Institutes ofHealth [U01CA207946, R01EB019036, U01CA151648] to Peixuan Guoas well as [P50CA168505, R21CA209045] and DOD Award [W81XWH-15-1-0052] to Dan Shu. The authors would like to thank Daniel Jasinskifor the helpful discussion and assistance in RNA solid phase synthesis.We also would like to thank Dr. Haining Zhu at University of Kentuckyfor the help of performing Mass Spectrometry. P.G.'s Sylvan G. FrankEndowed Chair position in Pharmaceutics and Drug Delivery is fundedby the CM Chen Foundation. PG is the consultant of Oxford NanoporeTechnologies and Nanobio Delivery Pharmaceutical Co. Ltd., as well asthe cofounder of Exonano RNA LLC, Shenzhen P&Z Bio-medical Co. Ltd.and its subsidiary US P&Z Biological Technology LLC.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.jconrel.2018.02.014.

References

[1] R. Paulmurugan, R. Bhethanabotla, K. Mishra, R. Devulapally, K. Foygel, T.V. Sekar,J.S. Ananta, T.F. Massoud, A. Joy, Folate receptor-targeted polymeric micellar na-nocarriers for delivery of orlistat as a repurposed drug against triple-negative breastcancer, Mol. Cancer Ther. 15 (2016) 221–231.

[2] J.B. Lee, J. Hong, D.K. Bonner, Z. Poon, P.T. Hammond, Self-assembled RNA in-terference microsponges for efficient siRNA delivery, Nat. Dent. Mater. 11 (2012)316–322.

[3] H. Li, K. Zhang, F. Pi, S. Guo, L. Shlyakhtenko, W. Chiu, D. Shu, P. Guo, Controllableself-assembly of RNA tetrahedrons with precise shape and size for cancer targeting,Adv. Dent. Mater. 28 (2016) 7501–7507.

[4] S.M. Douglas, H. Dietz, T. Liedl, B. Hogberg, F. Graf, W.M. Shih, Self-assembly ofDNA into nanoscale three-dimensional shapes, Nature 459 (2009) 414–418.

[5] P. Guo, The emerging field of RNA nanotechnology, Nat. Nanotechnol. 5 (2010)833–842.

[6] D. Shu, Y. Shu, F. Haque, S. Abdelmawla, P. Guo, Thermodynamically stable RNAthree-way junctions for constructing multifuntional nanoparticles for delivery oftherapeutics, Nat. Nanotechnol. 6 (2011) 658–667.

[7] Y. Shu, M. Cinier, D. Shu, P. Guo, Assembly of multifunctional phi29 pRNA nano-particles for specific delivery of siRNA and other therapeutics to targeted cells,Methods 54 (2011) 204–214.

[8] F. Haque, D. Shu, Y. Shu, L. Shlyakhtenko, P. Rychahou, M. Evers, P. Guo,Ultrastable synergistic tetravalent RNA nanoparticles for targeting to cancers, NanoToday 7 (2012) 245–257.

[9] K.A. Afonin, M. Kireeva, W.W. Grabow, M. Kashlev, L. Jaeger, B.A. Shapiro, Co-transcriptional assembly of chemically modified RNA nanoparticles functionalizedwith siRNAs, Nano. Lett. 12 (2012) 5192–5195.

[10] Y. Shu, D. Shu, F. Haque, P. Guo, Fabrication of pRNA nanoparticles to delivertherapeutic RNAs and bioactive compounds into tumor cells, Nat. Protoc. 8 (2013)1635–1659.

[11] E. Khisamutdinov, H. Li, D. Jasinski, J. Chen, J. Fu, P. Guo, Enhancing im-munomodulation on innate immunity by shape transition among RNA triangle,square, and pentagon nanovehicles, Nucleic Acids Res. 42 (2014) 9996–10004.

[12] D. Jasinski, E.F. Khisamutdinov, Y.L. Lyubchenko, P. Guo, Physicochemically tun-able poly-functionalized RNA square architecture with fluorogenic and ribozymaticproperties, ACS Nano 8 (2014) 7620–7629.

[13] E.F. Khisamutdinov, D.L. Jasinski, H. Li, K. Zhang, W. Chiu, P. Guo, Fabrication ofRNA 3D nanoprism for loading and protection of small RNAs and model drugs, Adv.Mater. 28 (2016) 100079–100087.

[14] K.A. Afonin, M. Viard, A.Y. Koyfman, A.N. Martins, W.K. Kasprzak, M. Panigaj,R. Desai, A. Santhanam, W.W. Grabow, L. Jaeger, E. Heldman, J. Reiser, W. Chiu,E.O. Freed, B.A. Shapiro, Multifunctional RNA nanoparticles, Nano Lett. 14 (2014)5662–5671.

[15] S.M. Dibrov, J. McLean, J. Parsons, T. Hermann, Self-assembling RNA square, Proc.Natl. Acad. Sci. U.S.A. 108 (2011) 6405–6408.

[16] K.A. Afonin, E. Bindewald, A.J. Yaghoubian, N. Voss, E. Jacovetty, B.A. Shapiro,L. Jaeger, In vitro assembly of cubic RNA-based scaffolds designed in silico, Nat.Nanotechnol. 5 (2010) 676–682.

[17] T.J. Lee, F. Haque, D. Shu, J.Y. Yoo, H. Li, R.A. Yokel, C. Horbinski, T.H. Kim, S.-H. Kim, I. Nakano, B. Kaur, C.M. Croce, P. Guo, RNA nanoparticles as a vector fortargeted siRNA delivery into glioblastoma mouse model, Oncotarget 6 (2015)14766–14776.

[18] D. Cui, C. Zhang, B. Liu, Y. Shu, T. Du, D. Shu, K. Wang, F. Dai, Y. Liu, C. Li, F. Pan,Y. Yang, J. Ni, H. Li, B. Brand-Saberi, P. Guo, Regression of gastric cancer by sys-temic injection of RNA nanoparticles carrying both ligand and siRNA, Sci. Rep. 5(2015) 10726.

[19] D. Shu, H. Li, Y. Shu, G. Xiong, W.E. Carson, F. Haque, R. Xu, P. Guo, Systemicdelivery of anti-miRNA for suppression of triple negative breast cancer utilizing

Y. Shu et al. Journal of Controlled Release 276 (2018) 17–29

27

RNA nanotechnology, ACS Nano 9 (2015) 9731–9740.[20] P. Rychahou, F. Haque, Y. Shu, Y. Zaytseva, H.L. Weiss, E.Y. Lee, W. Mustain,

J. Valentino, P. Guo, B.M. Evers, Delivery of RNA nanoparticles into colorectalcancer metastases following systemic administration, ACS Nano 9 (2015)1108–1116.

[21] D. Binzel, Y. Shu, H. Li, M. Sun, Q. Zhang, D. Shu, B. Guo, P. Guo, Specific deliveryof MiRNA for high efficient inhibition of prostate cancer by RNA nanotechnology,Mol. Ther. 24 (2016) 1267–1277.

[22] K.A. Afonin, W.W. Grabow, F.M. Walker, E. Bindewald, M.A. Dobrovolskaia,B.A. Shapiro, L. Jaeger, Design and self-assembly of siRNA-functionalized RNAnanoparticles for use in automated nanomedicine, Nat. Protoc. 6 (2011)2022–2034.

[23] E.F. Khisamutdinov, M.N. Bui, D. Jasinski, Z. Zhao, Z. Cui, P. Guo, Simple methodfor constructing RNA triangle, square, pentagon by tuning interior RNA 3WJ anglefrom 60 degrees to 90 degrees or 108 degrees, Methods Mol. Biol. 1316 (2015)181–193.

[24] Y. Shu, F. Haque, D. Shu, W. Li, Z. Zhu, M. Kotb, Y. Lyubchenko, P. Guo, Fabricationof 14 different RNA nanoparticles for specific tumor targeting without accumula-tion in normal organs, RNA 19 (2013) 766–777.

[25] E.F. Khisamutdinov, D.L. Jasinski, P. Guo, RNA as a boiling-resistant anionicpolymer material to build robust structures with defined shape and stoichiometry,ACS Nano 8 (2014) 4771–4781.

[26] H. Li, P.G. Rychahou, Z. Cui, F. Pi, B.M. Evers, D. Shu, P. Guo, W. Luo, RNA na-noparticles derived from three-way junction of Phi29 motor pRNA are resistant to I-125 and Cs-131 radiation, Nucleic Acid Ther. 25 (2015) 188–197.

[27] T.J. Lee, J.Y. Yoo, D. Shu, H. Li, J. Zhang, J.G. Yu, A.C. Jaime-Ramirez, M. Acunzo,G. Romano, R. Cui, H.L. Sun, Z. Luo, M. Old, B. Kaur, P. Guo, C.M. Croce, RNAnanoparticle-based targeted therapy for glioblastoma through inhibition of onco-genic miR-21, Mol. Ther. 25 (2017) 1544–1555.

[28] D. Shu, E. Khisamutdinov, L. Zhang, P. Guo, Programmable folding of fusion RNAcomplex driven by the 3WJ motif of phi29 motor pRNA, Nucleic Acids Res. 42(2013).

[29] D. Shu, H. Zhang, J. Jin, P. Guo, Counting of six pRNAs of phi29 DNA-packagingmotor with customized single molecule dual-view system, EMBO J. 26 (2007)527–537.

[30] D.L. Jasinski, H. Yin, Z. Li, P. Guo, The hydrophobic effect from conjugated che-micals or drugs on in vivo biodistribution of RNA nanoparticles, Hum. Gene Ther.29 (2017) 77–86.

[31] F. Pi, D.W. Binzel, T.J. Lee, Z. Li, M. Sun, P. Rychahou, H. Li, F. Haque, S. Wang,C.M. Croce, B. Guo, B.M. Evers, P. Guo, Nanoparticle oreientation to control RNAsurface display on extracellular vesicles for efficient cell targeting, siRNA deliveryand cancer regression, Nat. Nanotechnol. 13 (2017) 82–89.

[32] M. Raouane, D. Desmaele, G. Urbinati, L. Massaad-Massade, P. Couvreur, Lipidconjugated oligonucleotides: a useful strategy for delivery, Bioconjug.Chem 23(2012) 1091–1104.

[33] C. Gosse, A. Boutorine, I. Aujard, M. Chami, A. Kononov, E. Cogne-Laage,J.F. Allemand, J. Li, L. Jullien, Micelles of lipid-oligonucleotide conjugates: im-plications for membrane anchoring and base pairing, J. Phys. Chem. B 108 (2004)6485–6497.

[34] Y. Wu, K. Sefah, H. Liu, R. Wang, W. Tan, DNA aptamer-micelle as an efficientdetection/delivery vehicle toward cancer cells, Proc. Natl. Acad. Sci. U.S.A. 107(2010) 5–10.

[35] H. Liu, Z. Zhu, H. Kang, Y. Wu, K. Sefan, W. Tan, DNA-based micelles: synthesis,micellar properties and size-dependent cell permeability, Chemistry 16 (2010)3791–3797.

[36] T. Chen, C.S. Wu, E. Jimenez, Z. Zhu, J.G. Dajac, M. You, D. Han, X. Zhang, W. Tan,DNA micelle flares for intracellular mRNA imaging and gene therapy, Angew.Chem. Int. Ed. Engl. 52 (2013) 2012–2016.

[37] J.H. Jeong, T.G. Park, Novel polymer-DNA hybrid polymeric micelles composed ofhydrophobic poly(D,L-lactic-co-glycolic acid) and hydrophilic oligonucleotides,Bioconjug. Chem. 12 (2001) 917–923.

[38] P.M. Dentinger, B.A. Simmons, E. Cruz, M. Sprague, DNA-mediated delivery of li-pophilic molecules via hybridization to DNA-based vesicular aggregates, Langmuir22 (2006) 2935–2937.

[39] F.E. Alemdaroglu, K. Ding, R. Berger, A. Herrmann, DNA-templated synthesis inthree dimensions: introducing a micellar scaffold for organic reactions, Angew.Chem. Int. Ed. Engl. 45 (2006) 4206–4210.

[40] T. Trinh, P. Chidchob, H.S. Bazzi, H.F. Sleiman, DNA micelles as nanoreactors: ef-ficient DNA functionalization with hydrophobic organic molecules, Chem.Commun. (Camb.) 52 (2016) 10914–10917.

[41] Z. Li, Y. Zhang, P. Fullhart, C.A. Mirkin, Reversible and chemically programmablemicelle assembly with DNA block-copolymer amphiphiles, Nano Lett. 4 (2004)1055–1058.

[42] F.E. Alemdaroglu, N.C. Alemdaroglu, P. Langguth, A. Herrmann, DNA block co-polymer micelles- a combinatorial tool for cancer nanotechnology, Adv. Dent.Mater. 20 (2008) 899–902.

[43] M.C. Wani, H.L. Taylor, M.E. Wall, P. Coggon, A.T. McPhail, Plant antitumor agents.VI. The isolation and structure of taxol, a novel antileukemic and antitumor agentfrom Taxus brevifolia, J. Am. Chem. Soc. 93 (1971) 2325–2327.

[44] C.M. Spencer, D. Faulds, Paclitaxel. A review of its pharmacodynamic and phar-macokinetic properties and therapeutic potential in the treatment of cancer, Drugs48 (1994) 794–847.

[45] E.K. Rowinsky, R.C. Donehower, Paclitaxel (taxol), N. Engl. J. Med. 332 (1995)1004–1014.

[46] M.A. Jordan, L. Wilson, Microtubules as a target for anticancer drugs, Nat. Rev.Cancer 4 (2004) 253–265.

[47] S.B. Horwitz, Mechanism of action of taxol, Trends Pharmacol. Sci. 13 (1992)134–136.

[48] A.K. Singla, A. Garg, D. Aggarwal, Paclitaxel and its formulations, Int. J. Pharm. 235(2002) 179–192.

[49] H. Gelderblom, J. Verweij, K. Nooter, A. Sparreboom, Cremophor EL: the draw-backs and advantages of vehicle selection for drug formulation, Eur. J. Cancer 37(2001) 1590–1598.

[50] A. Sparreboom, T.O. Van, W.J. Nooijen, J.H. Beijnen, Nonlinear pharmacokineticsof paclitaxel in mice results from the pharmaceutical vehicle Cremophor EL, CancerRes. 56 (1996) 2112–2115.

[51] S.C. Kim, D.W. Kim, Y.H. Shim, J.S. Bang, H.S. Oh, K.S. Wan, M.H. Seo, In vivoevaluation of polymeric micellar paclitaxel formulation: toxicity and efficacy, J.Control. Release 72 (2001) 191–202.

[52] T. Hamaguchi, K. Kato, H. Yasui, C. Morizane, M. Ikeda, H. Ueno, K. Muro,Y. Yamada, T. Okusaka, K. Shirao, Y. Shimada, H. Nakahama, Y. Matsumura, Aphase I and pharmacokinetic study of NK105, a paclitaxel-incorporating micellarnanoparticle formulation, Br. J. Cancer 97 (2007) 170–176.

[53] Y. Dong, S.S. Feng, Methoxy poly(ethylene glycol)-poly(lactide) (MPEG-PLA) na-noparticles for controlled delivery of anticancer drugs, Biomaterials 25 (2004)2843–2849.

[54] K. Kim, J.H. Kim, H. Park, Y.S. Kim, K. Park, H. Nam, S. Lee, J.H. Park, R.W. Park,I.S. Kim, K. Choi, S.Y. Kim, K. Park, I.C. Kwon, Tumor-homing multifunctionalnanoparticles for cancer theragnosis: simultaneous diagnosis, drug delivery, andtherapeutic monitoring, J. Control. Release 146 (2010) 219–227.

[55] Y. Yoshizawa, Y. Kono, K. Ogawara, T. Kimura, K. Higaki, PEG liposomalization ofpaclitaxel improved its in vivo disposition and anti-tumor efficacy, Int. J. Pharm.412 (2011) 132–141.

[56] H. Yuan, J. Miao, Y.Z. Du, J. You, F.Q. Hu, S. Zeng, Cellular uptake of solid lipidnanoparticles and cytotoxicity of encapsulated paclitaxel in A549 cancer cells, Int.J. Pharm. 348 (2008) 137–145.

[57] J.J. Khandare, S. Jayant, A. Singh, P. Chandna, Y. Wang, N. Vorsa, T. Minko,Dendrimer versus linear conjugate: influence of polymeric architecture on the de-livery and anticancer effect of paclitaxel, Bioconjug. Chem. 17 (2006) 1464–1472.

[58] A.Y. Bedikian, C. Plager, N. Papadopoulos, O. Eton, J. Ellerhorst, T. Smith, Phase IIevaluation of paclitaxel by short intravenous infusion in metastatic melanoma,Melanoma Res. 14 (2004) 63–66.

[59] J.R. Hwu, Y.S. Lin, T. Josephrajan, M.H. Hsu, F.Y. Cheng, C.S. Yeh, W.C. Su,D.B. Shieh, Targeted Paclitaxel by conjugation to iron oxide and gold nanoparticles,J. Am. Chem. Soc. 131 (2009) 66–68.

[60] C.L. Lay, H.Q. Liu, H.R. Tan, Y. Liu, Delivery of paclitaxel by physically loading ontopoly(ethylene glycol) (PEG)-graft-carbon nanotubes for potent cancer therapeutics,Nanotechnology 21 (2010) 065101.

[61] J. Deng, L. Huang, F. Liu, Understanding the structure and stability of paclitaxelnanocrystals, Int. J. Pharm. 390 (2010) 242–249.

[62] T.G.W. Edwardson, K.M.M. Carneiro, C.K. McLaughlin, C.J. Serpell, H.F. Sleiman,Site-specific positioning of dendritic alkyl chains on DNA cages enables their geo-metry-dependent self-assembly, Nat. Chem. 5 (2013) 868–875.

[63] A. Zhang, Z. Zhang, F. Shi, J. Ding, C. Xiao, X. Zhuang, C. He, L. Chen, X. Chen,Disulfide crosslinked PEGylated starch micelles as efficient intracellular drug de-livery platforms, Soft Matter 9 (2013) 2224–2233.

[64] S. Guo, N. Tschammer, S. Mohammed, P. Guo, Specific delivery of therapeutic RNAsto cancer cells via the dimerization mechanism of phi29 motor pRNA, Hum. Gene.Ther. 16 (2005) 1097–1109.

[65] H. Zhang, J.A. Endrizzi, Y. Shu, F. Haque, C. Sauter, L.S. Shlyakhtenko,Y. Lyubchenko, P. Guo, Y.I. Chi, Crystal structure of 3WJ Core revealing divalention-promoted thermostability and assembly of the Phi29 hexameric motor pRNA,RNA 19 (2013) 1226–1237.

[66] P. Greenspan, E.P. Mayer, S.D. Fowler, Nile red: a selective fluorescent stain forintracellular lipid droplets, J. Cell Biol. 100 (1985) 965–973.

[67] M.A. Behlke, Chemical modification of siRNAs for in vivo use, Oligonucleotides 18(2008) 305–319.

[68] H. Vestweber, J. Pommerehne, R. Sander, R.F. Mahrt, A. Greiner, W. Heitz,H. Bassler, Majority carrier injection from Ito anodes into organic light-emitting-diodes based upon polymer blends, Synth. Met. 68 (1995) 263–268.

[69] D.W. Binzel, E.F. Khisamutdinov, P. Guo, Entropy-driven one-step formation ofPhi29 pRNA 3WJ from three RNA fragments, Biochemistry 53 (2014) 2221–2231.

[70] J. Liu, S. Guo, M. Cinier, L.S. Shlyakhtenko, Y. Shu, C. Chen, G. Shen, P. Guo,Fabrication of stable and RNase-resistant RNA nanoparticles active in gearing thenanomotors for viral DNA packaging, ACS Nano 5 (2011) 237–246.

[71] M. Skwarczynski, Y. Hayashi, Y. Kiso, Paclitaxel prodrugs: toward smarter deliveryof anticancer agents, J. Med. Chem. 49 (2006) 7253–7269.

[72] P. Rychahou, Y. Shu, F. Haque, J. Hu, P. Guo, B.M. Evers, Methods and assays forspecific targeting and delivery of RNA nanoparticles to cancer metastases, MethodsMol. Biol. 1297 (2015) 121–135.

[73] S. Guo, F. Huang, P. Guo, Construction of folate-conjugated pRNA of bacteriophagephi29 DNA packaging motor for delivery of chimeric siRNA to nasopharyngealcarcinoma cells, Gene Ther. 13 (2006) 814–820.

[74] F. Pi, H. Zhang, H. Li, V. Thiviyanathan, D.G. Gorenstein, A.K. Sood, P. Guo, RNAnanoparticles harboring annexin A2 aptamer can target ovarian cancer for tumor-specific doxorubicin delivery, Nanomedicine 13 (2016) 1183–1193.

[75] S. Guo, H. Li, M. Ma, J. Fu, Y. Dong, P. Guo, Size, shape and sequence-dependentimmunogenicity of RNA nanoparticles, Mol. Ther. Nucleic Acids 9 (2017) 399–408.

[76] A.R. Tall, L. Yvan-Charvet, Cholesterol, inflammation and innate immunity, Nat.Rev. Immunol. 15 (2015) 104–116.

[77] U. Jaffer, R.G. Wade, T. Gourlay, Cytokines in the systemic inflammatory responsesyndrome: a review, HSR Proc. Intensive Care Cardiovasc. Anesth. 2 (2010)

Y. Shu et al. Journal of Controlled Release 276 (2018) 17–29

28

161–175.[78] T. Tanaka, M. Narazaki, T. Kishimoto, IL-6 in inflammation, immunity, and disease,

Cold Spring Harb. Perspect. Biol 6 (2014) a016295.[79] J. Scheller, A. Chalaris, D. Schmidt-Arras, S. Rose-John, The pro- and anti-in-

flammatory properties of the cytokine interleukin-6, Biochim. Biophys. Acta 1813(2011) 878–888.

[80] Z.M. Wang, C. Liu, R. Dziarski, Chemokines are the main proinflammatory med-iators in human monocytes activated by Staphylococcus aureus, peptidoglycan, andendotoxin, J. Biol. Chem. 275 (2000) 20260–20267.

[81] M.D. Turner, B. Nedjai, T. Hurst, D.J. Pennington, Cytokines and chemokines: at thecrossroads of cell signalling and inflammatory disease, Biochim. Biophys. Acta 1843(2014) 2563–2582.

Y. Shu et al. Journal of Controlled Release 276 (2018) 17–29

29