저작자표시 - 비영리 - 변경금지 2.0 대한민국 이용자는아래의조건을따르는경우에한하여자유롭게 이저작물을복제, 배포, 전송, 전시, 공연및방송할수있습니다. 다음과같은조건을따라야합니다 : 저작자표시. 귀하는원저작자를표시하여야합니다. 비영리. 귀하는이저작물을영리목적으로이용할

Transcription

1 저작자표시 - 비영리 - 변경금지 2.0 대한민국 이용자는아래의조건을따르는경우에한하여자유롭게 이저작물을복제, 배포, 전송, 전시, 공연및방송할수있습니다. 다음과같은조건을따라야합니다 : 저작자표시. 귀하는원저작자를표시하여야합니다. 비영리. 귀하는이저작물을영리목적으로이용할수없습니다. 변경금지. 귀하는이저작물을개작, 변형또는가공할수없습니다. 귀하는, 이저작물의재이용이나배포의경우, 이저작물에적용된이용허락조건을명확하게나타내어야합니다. 저작권자로부터별도의허가를받으면이러한조건들은적용되지않습니다. 저작권법에따른이용자의권리는위의내용에의하여영향을받지않습니다. 이것은이용허락규약 (Legal Code) 을이해하기쉽게요약한것입니다. Disclaimer

2 의학박사학위논문 Liposomal delivery of mesoporous silica-coated iron oxide magnetic nanoparticles for theranostic application 치료진단적적용을위한다공성실리카코팅 산화철자성나노파티클의리포좀을이용한전달 2017 년 2 월 서울대학교의과대학원 의학과핵의학전공 장수진

3 Liposomal delivery of mesoporous silica-coated iron oxide magnetic nanoparticles for theranostic application 지도교수이동수 이논문을의학박사학위논문으로제출함 2017 년 12 월 서울대학교의과대학원 의학과핵의학전공 장수진 장수진의의학박사학위논문을인준함 2017 년 1 월 위원장정재민 ( 인 ) 부위원장이동수 ( 인 ) 위원최준영 ( 인 ) 위원이윤상 ( 인 ) 위원황도원 ( 인 )

4 논문초록 치료진단적적용을위한다공성실리카코팅산화철자성나노파티클의리포좀전달 암조직의영상화가가능하면서동시에암세포에치료제를효과적으로전달할수있는나노입자의개발을위해다양하고많은연구들이시도되었으나나노입자자체의독성이문제가되거나충분한양의치료제를담을수있는나노운반체를만들기어렵다는점때문에실제임상에서의암환자치료에도입되기어려웠다. 이연구에서는이러한제한점을극복하기위하여산화철을일련의화학반응을이용하여다공성실리카로표면처리한자성나노입자 2 ) 를개발하였으며그자성과물리적성상이안정적으로유지됨을확인하였다. 표면처리를하지않은산화철나노입자가산성용액에노출될경우해체되는반면 MNP@mSiO 2 는안정성을유지하였고그자성 (r 2 =270 mm -1 S -1 ) 은무정형실리카로처리한산화철나노입자 (70 mm -1 S -1 ) 에비해월등히높으면서산화철나노입자본연의자성 (286 mm -1 S -1 ) 과유사하게유지되었다. 그러나다공성실리카조성은버퍼용액내안정성에취약하여생물학적활용에부적합하였으므로리포좀복합체를이용하여안정성을확보하는동시에기존의나노입자보다많은양의치료제를담을수있도록하였다 (MagLipo). i

5 Cryo-TEM 분석결과약 6-8 개의 2 입자가하나의리포좀 입자내에봉입되었으며 2 와달리 MagLipo는인산염버퍼용액에서장시간안정성을유지하였다. 이렇게구성된 MagLipo 표면에 EGFR 항체를결합 (EGFR-MagLipo) 하여췌장암세포주의선택적검출을시도하였고췌장암세포에서이상증폭되는 Plk1 단백질합성을방해하는 sirna를항암제인 doxorubicin과함께탑재하여선택적으로전달하고암세포특이적환경에반응하여치료제를방출하도록기획하였다. 최종적 으로구성된, T 2 증강영상이가능함과동시에암세포의선택적검출및 치료를목적으로하는, 치료진단용나노파티클의검출효율및치료효과를췌장암세포주실험으로확인하였다. 암세포진입후리소좀내특이적환경을반영하는 ph 4.7의용액에서 doxorubicin을탑재한 MagLipo 는 30분이내에약제의 40% 를방출하였다. T 2 -w phantom 실험에서 MagLipo를처리한췌장암세포와대조군세포는유사한신호강도를보였으나 EGFR-MagLipo를처리한췌장암세포의경우유의한 T2 신호의감소 ( 80%) 를보였다. 항암제와치료용 sirna를모두탑재한 MagLipo 를처리한경우각각을따로처리한경우와비교하여치료의상승효과가있음을확인하였다. 이상의결과를바탕으로 MagLipo는다기능의나노물질로써향후암의표적치료와영상진단을동시에수행할수있는 nanovehicle의개발에기여할것으로기대된다. 주요어 : 자성나노입자, 리포좀, 다공성실리카, 치료진단나노입자, ph- 감응성약제전달학번 : ii

6 Table of contents Introduction...1 Materials and methods.5 Chemical.5 Preparation of 16-nm sized iron oxide magnetic nanoparticles (MNPs) and 2 particles with Dox...5 Drug loading capacity (HPLC analysis)...7 Preparation of liposome encapsulated 2 (MagLipo) particles..7 Encapsulation of therapeutic gene into the MagLipo 8 Antibody conjugated onto the MagLipo and toxicology study...9 T2-w MR phantom study.10 Delivery of therapeutic moieties..11 Cell viability assay Results..13 Characteristics and stability of 16-nm sized MNP, MNP@SiO 2 particle with Dox, and liposome encapsulated MNP@mSiO 2 (MagLipo) particle.. 13 Loading efficiency of drug and therapeutic sirna into MagLipo and controlled release of payload 14 Antibody conjugated MagLipo and cytotoxicity study 15 T2-w MR phantom study.. 15 Delivery of therapeutic moieties...15 Cell viability assay 16 iii

7 Discussion Conclusion...25 Rererences...47 Abstract in English. 57 List of figures Figure 1. Schematic illustration and TEM images of MagLipo Figure 2. Characteristics of various nanoparticle.30 Figure 3. Solution stability of nanoparticles in various conditions..33 Figure 4. Characterization of drug or sirna incorporation and controlled release...35 Figure 5. Assessment of cell cytotoxicity and expression level of EGFR receptor.38 Figure 6. Magnetic property of MagLipo. 40 Figure 7. Uptake of MagLipo particles in the cell Figure 8. Therapeutic applications of MagLipo particles in vitro iv

8 Introduction Nanoparticles for the biomedical applications have been studied in an effort to find the most efficient way to deliver the drug and/or gene for therapeutic purposes (1-4). Besides for the therapeutic applications, several nanomedicine formulations combining disease diagnosis and therapy have been developed, and these theranostic nanomaterials are considered as highly suitable systems for monitoring drug delivery, drug release, and drug efficacy (5). Especially, metal nanoparticles are suggested as attractive theranostic agents for cancer owing to its inherent physicochemical properties providing both carriers for the various therapeutic moieties and tools for the medical imaging. Various methods to fabricate nanoparticles for the establishment of multifunctional and biocompatible theranostic systems have been introduced (6). Among various metal nanoparticles, quantum dot or iron oxide nanoparticles possess unique optical or magnetic properties and have received attention for decades as appealing candidate materials for imaging diagnosis (7-9). However, particles consist of heavy metal ions such as Cd, or Mn presented potential toxicological issues (10,11). Ions from these particles tend to be released in acidic conditions, which cause serious toxicity problems. To overcome the decomposition of the metal ions, researchers have utilized the silica shell coating, which has a relatively low toxicity and high acid resistance, onto the metal oxide nanoparticles (12-14). In the group of nanoparticles with various inorganic compositions, iron oxide-based nanoparticles are in unique position since considerable store of experience and knowledge regarding it has been accumulated for decades (15) and 1

9 many studies have reported iron oxide nanoparticles are biocompatible materials with negligible toxicity (16,17). In biomedical field, iron oxide magnetic nanoparticles (MNPs) are utilized for various applications such as magnetic resonance imaging (MRI), drug delivery, controlled drug release, cell labelling, cell separation, and nanoscale biosensor (18). MNPs with a higher relaxivity are preferred in order to increase detection sensitivity in MR-based imaging or sensing (19,20). To increase transverse relaxivity (r 2 ), size and chemical composition of the particles need to be tuned (21-23). The value of r 2 is proportional to the size (d) and saturated magnetism (Ms, r 2 d 2 M 2 s ) (24,25). Since iron oxide MNPs are sensitive to the acid corrosion, to improve stability and safety of the metal oxide, silica shell coatings have been most commonly used for the preparation of functionalized iron oxide nanoparticles. However, the shell was spiked with a lowered value of r 2 due to a decrease in the number of water molecules adjacent to the MNP core (26). In this study, to solve the problems of stability, safety, and efficiency of magnetic nanoparticles, the mesoporous silica (msio 2 ) shell was coated onto a magnetic iron oxide nanoparticle (MNP@mSiO 2 ). Mesoporous silica, having large surface area, good biocompatibility, and chemical and thermal stability, provides suitable conditions for the biomedical applications of nanocarriers (27). Also, the shell possesses a sufficient amount of water molecules inside the pores near the core particles to minimize the decrease in magnetic relaxivity (28). Furthermore, the pore structures can provide roomy space for the incorporation of therapeutic load. We chose doxorubicin (Dox), an anticancer drug, to be packaged in the nanoparticles. Dox, one of the anthracyclines, exerts its anticancer effects by intercalating into the double-stranded DNA, inhibiting topoisomerase II enzyme, 2

10 and forming free radicals damaging cells. It is well known as a potent anticancer drug against many types of cancer (29). Since mitochondrial membranes are especially sensitive to the oxidative damage by Dox, high cardiotoxicity is most important clinical consideration in chemotherapeutic regimens with Dox because heart muscles are rich in mitochondria (30). As an effective chemotherapeutic drug with dangerous cardiotoxicity, like a double edged sword, Dox is considered as an attractive candidate for the liposomal delivery (31). Also, its intrinsic fluorescence allows intracellular localization of drug using microscopic technique (32). Together with the chemotherapeutic agent (Dox), we adopted RNA interference-mediated silencing of proto-oncogene (Polo-like kinase 1, Plk1) to promote chemosensitivity of cancer cell to Dox (33) and synergistic therapeutic effect. Plk1 is overexpressed in a broad range of human cancers and is associated with poor prognosis (34-37). Several preclinical therapeutic attempts targeting Plk1 showed promising results (38-40) and a nanoparticle formulation of sirna for inhibition of Plk1 is in a clinical trial study for the treatment of advanced solid tumors (41). Consequently, both hydrophilic drug (Dox) and the sirna for the silencing of Plk1, could be incorporated into the mesoporous silica shell of MNP@mSiO 2 for the material to possess therapeutic ability. To improve their bio-applicability, liposomal complex systems were combined with the MNP@mSiO 2 particles (MagLipo). Even though nanoparticles are expected to accumulate in the tumors with higher concentration than the normal tissue, as explained with the EPR (enhanced permeability and retention) effect (42), intratumoral distribution of nanoparticles could be insufficient due to various factors such as disordered tumor vasculature and extensive tumor stroma. Therefore, adoption of recognition moiety for the cellular surface target that is upregulated in pancreatic cancer cells, but is poorly or not 3

11 expressed in normal tissues is recommended (43). In this way, active targeting along with passive targeting can potentially confer higher detection sensitivity, advanced delivery efficacy, and minimal off-target toxicity on the nanoparticles (44). Cancer-specific antibodies for the epidermal growth factor receptor (EGFR- Ab) were anchored onto the MagLipo particles (EGFR-MagLipo) through bioconjugation chemistry to render specific targeting, based on that the EGFR is overexpressed in pancreas cancer and is a good subject for the molecular targeted therapy (45). Above mentioned processes and results of this study is well documented in the recently published article (46). In summary, an attempt has been made to construct a novel theranostic nanoparticle system combining magnetic property for MR imaging and drug delivery system for cancer therapy, which can specifically target cancer cells and release drugs in stimulus-sensitive manner. The physicochemical properties of current nanoparticle (MagLipo), as a carrier system and imaging modality, were characterized. Then, the cellular uptake, intracellular localization, ph-sensitive drug release, anticancer activity, and synergistic therapeutic efficacy of simultaneous delivery of drug and sirna were determined in human pancreas cancer cell line. 4

12 Materials and methods Chemicals Tris (acetylacetonate) iron (III) [Fe(acac) 3, 99.9%], iron (II) acetylacetonate [Fe(acac) 2, 99.95%], 1,2-hexadecanol (90%), oleic acid (OA, 99%), oleylamine (OY, 70%), 1-octadecene (ODE, 95%), chloroform (99%), dimethyl sulfoxide (99.9%), and N-hydroxysulfosuccinimide sodium salt (98.5%) were purchased from Sigma-Aldrich, and used without further modifications. The lipid chemicals were purchased from Avanti Polar Lipids Inc. (Alabaster, AL, USA). Isopropanol (99.5%), hexane (98.5%), ethanol (99.5%), and NaHCO 3 were purchased from Fisher Scientific, Korea Ltd., and used without further purification. Preparation of 16-nm sized iron oxide magnetic nanoparticles (MNPs) and MNP@SiO 2 particles with Dox We first synthesized 10 nm sized MNPs as seeds for subsequent particle growth. Fe(acac) 3 (4 mmol, 1.4 g), Fe(acac) 2 (2 mmol, 0.5 g), 1,2-hexadecanediol (10 mmol, 2.9 g), OA (6 mmol, 1.9 ml), OY (6 mmol, 2.8 ml) and ODE (20 ml) were mixed by stirring under N 2 flow for 1 hr. The mixture was heated and kept at 200 C for 2 hr. The temperature was then quickly raised to 280 C to initiate particle formation. After reflux, the mixture was cooled to room temperature and isopropanol (4 ml) was added. MNPs were collected by centrifugation (7000 rpm, 15 min) and the precipitates were re-dispersed in hexane. To make larger MNPs 5

13 from the seed particles, 10 nm MNPs (100 mg) were dissolved in hexane (10 ml) together with the same amounts of metal acetylacetonates, 1,2-hexadecanediol, OA, OY and ODE as described above. Under N 2 flow, the mixture was heated and kept at 100 C for 1 hr. The temperature was then elevated to 200 C and maintained for two hours. Finally, the mixture was heated again to 300 C and refluxed for 2 hr before being cooled down to room temperature. The 12 nm MNPs were collected by the same washing and isolation procedures aforementioned. In a similar manner, 16 nm MNPs were prepared using these 12 nm particles as seeds. To determine the magnetic relaxivity of the MNP, the particles were dispersed in H 2 O after being treated with CTAB chemicals. TD-NMR (Minispec mq20, 0.47 T, Bruker, Billerica, MA, USA) was used to determine the relaxivity of the particles (r 2 ). To generate water-soluble MNP, at first 600 L particle solution in 2 mg/ml chloroform (CHCl 3 ) and 0.05 mm hexadecyltrimethylammonium bromide (CTAB) 5 ml were mixed and vigorously stirred for 1 hr in a 2-neck round bottom flask. And then the temperature was increased to 60 C for removing organic solvent. After 20 min, added 25 ml distilled water and 1.8 ml NaOH (2 N) at 70 C. Stirred ethyl acetate (EA) 1.8 ml into this mixture and slowly dropped tetraethylorthosilicate (TEOS) 0.3 ml for 6 hr. After completion of reaction, the mixture was centrifuged for 20 min at 14,000 rpm and washed with ethanol and distilled water separately. To remove CTAB in the product completely, added HCl 0.05 ml solution and then vigorously stirred for 30 min. The schematic illustration of synthesis of MagLipo is presented in Figure 1A. The morphological features, structure, composition, and magnetic properties of the resulting 80 nm iron oxide mesoporous silica-core-shell 6

14 nanoparticles 2 ) were characterized using a transmission electron microscope (TEM; JEOL 2100, JEOL, Peabody, MA, USA), an X-ray powder diffractometer (XRD; RU300, Rigaku, Kyoto, Japan), and a vibrating sample magnetometer (VSM), respectively. To confirm the resistance of the meso-porous silica shell to acid corrosion, the MNP and MNP@mSiO 2 particles were dispersed in a 1 N HCl solution. To load therapeutic drug (Dox), µmol of Dox was dissolved in 2 ml MNP@mSiO 2 (2 mg) aqueous solution and then homogeneously stirred for 2 hr at 60 C. The Dox loaded particle solution was centrifuged to remove unloaded Dox (13,000 rpm, 10 min) and repeated washing step with cleaned H 2 O. Drug loading capacity (HPLC analysis) To evaluate the loading capacity of Dox in MNP@mSiO 2 particles, the supernatant solution of excess Dox after centrifugation was analyzed quantitatively by HPLC (hydrosphere C18 column: 4.6 x 50 mm, 3 µm, Agilent 1200, Santa Clara, CA, USA). The eluent solution (0.01% perfluorobutyric acid, 0.1% isopropanol in water) was utilized with optimized rate (1.0 ml/min) and Dox intensity was determined by equipped fluorescence spectroscopy (ex. 485 nm and em. 590 nm). Preparation of liposome encapsulated MNP@mSiO 2 (MagLipo) particles 7

15 1,2-Dipalmitoyl-sn-glycero-3-phosphatidylcholine (DPPC, 20 µmol, 15.4 mg), diacetyl phosphate (DCP, 1.8 µmol, 1.0 mg), 1,2-distearoyl-snphosphoethanolamine-N-[PDP(polyethylene glycol)-2000] (DSPE-PEG-PDP, 5.0 mg), and cholesterol (3.48 mg) were dissolved in 5 ml chloroform (99.9%, Sigma- Aldrich, St. Louis, MO, USA). The mixed solution was allowed to evaporate for 5 min at room temperature and was then freeze-dried for 24 h at -45 C to induce the formation of a phospholipid film. Subsequently, 3 ml MNP@mSiO 2 (4 mm Fe) in 30 mm NH 4 HCO 3 solution underwent sonication for 5 min at 60 C. The solution was subjected to a freeze-thaw cycle five times, using liquid nitrogen and a water bath. The prepared Lipo dispersion solution was extruded through a 200 nm filter at 60 C (mini-extruder, Avanti polar lipids). The excess organic chemicals were continually removed by centrifugation (5 min at 13,000 rpm) and washing with H 2 O. The size distribution of the prepared nanoparticles (MNP, MNP@mSiO 2 and MagLipo) was measured by dynamic light scattering (DLS) analysis. The solubility of MNP@mSiO 2 and MagLipo was tested in DW and in PBS respectively. Encapsulation of therapeutic gene into the MagLipo For the synergistic therapeutic effect, the sirna for Plk1 (siplk1) was loaded into the MagLipo particles with ph-sensitive NH 4 HCO 3 gradient method (47,48). To prepare Protamine (PA, 7.5 kda) and siplk1 complex, the siplk1 gene (50 µm) and different concentration of PA solutions (0.3 ~ 12 µm) were incubated for 1 hr at room temperature using a plate shaker. And the ratio of PA and siplk1 concentration was determined to be 12: 5 X 10-6 M by electrophoresis. To load 8

16 siplk into the liposome, the fresh PA and siplk1 complex solution was homogenized into MagLipo films (21 mg, 2 ml) for 5 min at room temperature, using electrophoresis. The MagLipo particles harboring siplk1-pa were washed twice with centrifugation (5 min at 13,000 rpm) to remove the unloaded gene complex and the loading capacity was then estimated with a UV-Vis spectrophotometer. Antibody conjugated onto the MagLipo and toxicology study Cancer-specific antibodies were anchored onto the MagLipo particles (EGFR- MagLipo) through bio-conjugation chemistry to render specific targeting. To conjugate cancer specific antibody (anti-egfr, AbCAM), the amine terminated particles (0.05 mm) were dispersed in 2 ml phosphate-buffered saline (PBS) and then added with Sulfo-SMCC (5 mg) in order to make the maleimidide chemical functional group. The mixture was stirred for 2 hr at room temperature. The nanoparticle solution was precipitated down (13,000 rpm, 15 min) and washed three times with H 2 O. On the other hand, we activated EGFR antibody with Traut s reagent for the thiolation of EGFR antibody according to Pierce protocol. Anti- EGFR antibody (0.2 mg) and cysteamine (2-MEA, 1 mg) were dissolved in 2 ml EDTA solution (10 mm, Sigma-Aldrich), followed by incubation in a shaking incubator for 1.5 hr at 37 C. The thiol-active antibody was purified with PD-10 desalting column (GE Healthcare Bio-Sciences, Little Chalfont, UK) and immediately combined with the maleimide terminated MagLipo (0.05 ml). The mixture was shaken for 6 hr at 4 C and subsequently purified through 9

17 centrifugation. The toxicological study of the MagLipo-based particles was conducted using fibroblast (293T, EGFR negative) and pancreatic cancer (BxPC3, EGFR positive) cell lines with increasing concentration of EGFR-targeting and bare MagLipo particles. The expression levels of EGFR in both cell lines were determined by western blotting. T 2 -w MR phantom study To demonstrate the utility of MagLipo particles for magnetic resonance imaging (MRI), we used the particles to detect pancreatic cancer cells in T 2 -w signal phantom studies that involved MRI instruments (Achieva 3.0 T, TR/TE: 3500/7 ms). 3 µm [Fe] concentration of the particles were incubated with EGFR-positive BxPC3 cells for 3 hr inside a 5% CO 2 incubator. Subsequently, the excess particles were removed by washing the cell culture with PBS, and the treated cells (1 x 10 6 /dish) were detached by trypsinization and subsequently centrifuged at 3,000 rpm and subjected to T 2 -w signal phantom studies using MRI. All T 2 -w MR images were acquired on a 3.0-T clinical MR scanner (Achieva Release version, Philips medical system, Netherland). T 2 -w images were obtained with a TR 3.5 s, a TE 7 ms. Using a time domain-nmr magnetic relaxometer, cellular relaxivity of the EGFR-MagLipo-treated cell solution was evaluated to measure detection sensitivity. The change in T 2 ( T 2 %) was determined based on the decrease in the number of cells and by its sensitivity towards the cancer cells. 10

18 Delivery of therapeutic moieties The confocal laser scanning microscope was used to monitor the location of the Dox in MagLipo during delivery by its inherent red fluorescence property. The cells were seeded onto coverslips placed in a six-well plate for overnight and then incubated with MagLipo particles harboring Dox for 3 hr in serum free media. Afterward, the cells on coverglass, PBS-washed 3 times, were stained with DAPI for 5 minute to localize nuclei, and then washed 3 times with PBS again. After fixation of cells with formaldehyde solution, the prepared slides were analysed by confocal laser scanning microscope (LSM700, Zeiss, Oberkochen, Germany). Intensity profiles of Dox and nucleus stained with DAPI directly after administration and 24 hr after administration were plotted to demonstrate the ability of Dox in MagLipo to penetrate the nucleus. The Plk1 expression level of various cells was also analysed using the quantitative polymerase chain reaction (PCR) analysis with RT-PCR kit from ThermoFisher scientific (Waltham, MA USA) according to the instruction from supplier. Cell viability assay A 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium (MTT) assay kit (Invitrogen) was used to evaluate cell viability in the presence of particles. 293T fibroblasts were cultured in Dulbecco s Modified Eagle Medium (DMEMH16, Gipco, Grand Island, New York, U.S.A.), supplemented with fetal bovine serum (FBS, 10%), penicillin and streptomycin (1%), L-glutamine (1%) and sodium bicarbonate (2%). BxPC3 pancreatic cancer cells were cultured in vendor-provided 11

19 media. Cells were maintained at 37 C in a humidified atmosphere containing 5% of CO 2. At confluence, the cells were washed, trypsinized, and resuspended in culture media. Cells were then seeded at a concentration of 5,000 cells/well in a 96- well tissue culture plate and allowed to grow overnight at 37 C under 5% CO 2. Aqueous particle solutions at different particle concentrations were added into the culture media, and the cells were allowed to grow for another 24 hr. For cell viability test, the culture media was replaced with MTT solution. After 3 hr of incubation at 37 C under 5% CO 2, MTT solubilization solution was added to dissolve the resulting formazan crystal. Absorbances of samples were measured spectrophotometrically at a wavelength of 570 nm, with background absorbance at 690 nm to determine cell viability. The treated cells were also measured for the viability with repect to time. 12

20 Results Characteristics and stability of 16-nm sized MNP, 2 particle with Dox, and liposome encapsulated 2 (MagLipo) particle 16-nm (diameter) iron oxide nanoparticles with a regular size distribution were simply synthesized by using a seed growing method (Fig. 1B) (49,50). Next, the mesoporous silica shell was successfully coated onto the surface of MNPs, which were treated with NH 4 OH and TEOS. Resulting MNP@mSiO 2 particles clearly exhibited mesoporous morphology (Fig. 1C, high magnification image in Fig. 2A) and maintained the core MNP structure during the over-coating process. After mixing the MNP@mSiO 2 particle solution with phospholipid films, between 6 and 8 MNP@mSiO 2 particles were located inside the liposome, which make MagLipo particles, and were characterized by cryo-tem analysis (Fig. 1D). The MNP was characterized as having a typical ferrite spinel structure with powder X-ray diffraction and vibrating sample magnetometry revealed that it had a high M s (95 emu/g[fe]) with super-paramagnetism (Fig. 2B and C). The relaxivity measurements with TD-NMR (Minispec mq20, 0.47 T, Bruker) showed that the MNP particles had a relatively high r 2 (286 mm -1 S -1 ). The r 2 value of mesoporous silica shell MNP (MNP@SiO 2 ) was comparable to the bare MNP (270 mm -1 S -1 ), but much higher than the amorphous silica shell MNP (MNP@SiO 2, 70 mm -1 S -1 ) (Fig. 2D and E). 13

21 The size distribution of the prepared nanoparticles (MNP, 2 and MagLipo) measured by DLS analysis showed relative monodispersity (Fig. 3A). The resistance of the mesoporous silica shell to acid corrosion was presented by dispersion of the MNP and 2 particles in a 1 N HCl solution (Fig. 3B). The bare MNPs quickly dissociated within 20 min (~50%), but the MNP@mSiO 2 particles showed remarkable stability and resistance. We thought that the resistance to the acidic etching was resulted from the ability of mesoporous silica trapping proton molecules into the shell by hydrogen bond formation. For biological applications, nanoparticles should be dispersed in buffer solution. However, the prepared MNP@mSiO 2 particles did not have high solubility in the PBS buffer solution. In contrast, the MagLipo particles were found to have longterm stability (Fig. 3C). Loading efficiency of drug and therapeutic sirna into MagLipo and controlled release of payload The loading capacity of Dox was determined by HPLC analysis to be 98% (Fig. 4A). The siplk1, prepared as complex with protamine (PA), was well loaded into the MagLipo and the loading efficiency of gene, analysed by UV-Vis spectroscopy (Fig. 4B), was approximately 80%. The MagLipo particles loaded with Dox, prepared with the NH 4 HCO 3 gradient method, released 40% of Dox within 30 min in ph 4.7 solution (Figure 4C). 14

22 Antibody conjugated MagLipo and cytotoxicity study The toxicological study of the MagLipo was conducted using fibroblast (293T, EGFR negative) and pancreatic cancer (BxPC3, EGFR positive) cell lines. Expression levels of EGFR in each cell lines were determined by western blotting. Both group of cells, separately treated with bare MagLipo and antibody conjugated MagLipo at different concentrations, showed high viability (> 85 %) after 24 h (Fig. 5A-C). T 2 -w MR phantom study The EGFR-MagLipo-treated cells showed significant decrease in the T 2 -w signal compared with the control (non-treated) and bare MagLipo (without antibody conjugation)-treated cells (Fig. 6A and B). The change in T 2 ( T 2 %), which is defined as the difference in T 2 between the cells treated with EGFR-MagLipo particles and the non-treated cells, is presented using a TD-NMR magnetic relaxometer (Fig. 6C) and the value found was approximately 10 cells if the threshold was chosen arbitrarily as 5% of T 2, which is similar to the single cell level. Delivery of therapeutic moieties The EGFR positive BxPC3 cells were treated with the therapeutic EGFR-targeting 15

23 MagLipo particles for 3 hr and displayed progressive uptake of particles. With bio- TEM analysis of treated cells, the particles localized in the cytoplasm were identified (Fig. 7A, white arrows indicate particles). The location of the Dox in MagLipo was monitored with the confocal laser scanning microscope by its inherent red fluorescence property, and at initial stage, the red color from Dox was broadly distributed in the cytosol and diffused into the nucleus of the cell after continuous culture for 24 hr (Fig. 7B). Cell viability assay In the Plk1 expression level of various cells, which were analysed using the polymerase chain reaction (PCR) technique, cells treated with only siplk1 showed a similar expression intensity as did the control (only cell). In contrast, the EGFR- MagLipo particles harboring both Dox and siplk1 were shown to substantially down-regulate the Plk1, with an approximate 60% decrease compared with the control (Fig. 8A). More importantly, the MagLipo without an antibody showed minimal amount of inhibition on Plk1. In comparison of cell viability, as expected, the simultaneous incorporation of Dox and the siplk1 into the MagLipo particle system was shown to have a remarkable induction effect on apoptotic cell death (~50%) after a 48-hr incubation (Fig. 8B) and reached ~70% of cell death after a 120-hr incubation. Compared with the single therapy with Dox-incorporated EGFR-MagLipo, simultaneous delivery with a drug and gene showed synergistic therapeutic effect. 16

24 Discussion The theranostic applications of nanoparticles in the biomedical science have been in the spotlight for decades, and great anticipation of nanotheranostic platform readily applicable in the cancer treatment cannot be mentioned too often. Metal nanoparticles of all are most promising candidate materials for the simultaneous diagnostic and therapeutic systems, not only because their innate properties are comparable as diagnostic and imaging purposes, also they could be sent to the cancer tissue or cells, armed with therapeutic drugs and equipped with targeting moiety. With these reasons, there exist several commercially available magnetic nanoparticles as contrast agents in magnetic resonance imaging, though magnetic sensing and magnetic hyperthermia will still need to be extensively validated before being introduced in the clinic, currently limited by the need of strong external magnetic field and by the concerns of sustained high temperature in the target organ along with a defined target volume, respectively (51). Nanoparticles with iron oxide core, as an attractive platform in the biomedical field, have been widely and vigorously investigated for the various applications such as diagnostics, therapeutics, biosensing, drug delivery, and controlled release of drug (52). With their inherent superparamagnetism, iron oxide nanoparticles can be used as contrast agent for the diagnostic purpose in MRI imaging, supported by its high biocompatibility rarely causing oxidative stress (53). However, there are still increasing concerns about toxicity from reactive oxygen species (ROS) generated by nanoparticles and the toxicity from metallic core itself, especially of quantum dots (11,54), though standardized and systematized toxicity 17

25 studies area lacking (10,55). Therefore, surface passivation of iron oxide nanoparticle with silica is considered to play important role in reducing oxidative stress and alteration of iron homeostasis, and consequently, the overall toxicity (56). And more importantly, such a structures consists of iron oxide core and organic coating have shown to be efficient for the applications utilizing MRI. Particularly, the mesoporous form of silica nanoparticles can be a good surface material of choice, because it could provide enough spaces for the loading of therapeutic drugs or genes, most importantly without compromising magnetic property of iron oxide core. Besides its several advantageous properties such as large surface area and pore volume to carry therapeutic moiety and favorable biocompatibility, the mesoporous silica is suggested to allow ready accessibility of water molecules to the magnetic core for the efficient water proton relaxation (57). In this study, the relaxivity of MNP@mSiO 2 was almost similar to that of bare MNPs, unlike with the low r 2 value of amorphous silica coated NPs, supporting the idea that the mesoporous silica can hold more water molecules than the amorphous silica can. Also, mesoporous silica nanoparticles are proposed as the basis of nanodevices for the controlled release of drugs and genes into living cells owing to its controllable morphology, porosity, and surface characteristics (58). However, even with these several advantageous properties, generally, both amorphous and mesoporous silica coated core-shell particles do not have long-term stability (solubility) in buffer solutions. Therefore, the surface of msio 2 particles needs to be coated with hydrophilic materials. Among materials for the surface coating of nanoparticles, liposomal formulations have been most popular and efficient tool for the delivery of various therapeutic agents such as 18

26 chemotherapeutic drugs, oligonuclide, antigens, and proteins since it offers several favorable properties; biocompatibility, biodegradability, and low toxicity (59,60). Antibody targeted liposomes loaded with anticancer drugs demonstrated high potential for the clinical applications (60). Both passive and active drug delivery by liposomal nanoparticles can significantly reduce the toxic side effects of anticancer drugs and enhance therapeutic efficacy of the drug delivered (61). A meaningful progress is that the liposomes, with FDA approval, have been introduced and used as drug delivery system in clinic (Doxil, Abelcet, Ambisome) (62,63). Liposomal doxorubicin is indicated for ovarian cancer, breast cancer, multiple myeloma, and kaposi s sarcoma (62). Liposomal doxorubicin statistically trended to lower cardiac event rates than doxorubicin (64). The early and positive preclinical and clinical experiences with liposomal delivery of anthracyclines made liposome popular in nanomedicine for cancer therapy (65). Therefore, we combined a liposomal complex with biocompatible phospholipid chemicals as stabilizers. A novel structure of msio 2 -coated MNP (MNP@mSiO 2 ) encapsulated liposomal particles (MagLipo) was prepared and showed good solubility in PBS buffer, suggesting this MagLipo particle system has excellent physico-chemical properties for biological applications. Despite their similar r 2 (270 mm - 1 S -1 ) values, the EGFR-MagLipo particles were shown to have a lower T 2 compared with MNP@mSiO 2 particles and it was assumed that the EGFR-MagLipo particles would have provided a much higher number of MNPs per one receptor of cell membrane than non-targeting particles, which would be akin to boosting cell uptake. Considering that the surface modifications need to be applied without compromising the magnetic property of iron oxide core, this unique feature of the MagLipo system in current study, even with amplified magnetism, can be regarded as very powerful and prominent 19

27 achievement from a diagnostic point of view. Based on the results from T 2 -w signal phantom study, BxPC3 pancreatic cancer cells treated with the EGFR-MagLipo particles showed significantly lower signal intensity than control cells or cells treated with bare MagLipo did, meaning both selective targeting and successful T2 relaxation have been accomplished. In the application of in vitro magnetic sensing, namely cellular assay with EGFR-MagLipo system, the T 2 change (%), defined as difference between cellular r 2 value of EGFR-MagLipo treated cells and of nontreated cells, showed that the only 10 EGFR-MagLipo-treated cells are required to be significantly differentiated from non-treated cells when 5% of T 2 was set as threshold, suggesting that the MagLipo is an effective diagnostic material system for the in vitro sensing of pancreatic cancer cells. Especially as a sirna delivery system, liposomes can prevent degradation of payload, accumulate preferentially in tumor tissue, deliver high concentration of payload, and with active targeting moiety, it also can aid specific targeting of sirnas to the tumor cells (34). The lipid nanoparticle formulation of a sirna directed against Plk1, known as TKM , successfully induced silencing of Plk1 expression in-vitro and in patient biopsy samples, and is being evaluated in a first-in-human, dose-escalation, phase I study in patients with advanced solid tumors (41). Among polo-like kinases (Plks), a family of serinethreonine kinases, Plk1 is the most widely understood key enzyme regulating essential steps of mitosis, overexpression of which in multiple human cancers and its association with poor prognosis have been well addressed (34,36,37,66). Since RNA interference with sirna of Plk1 induced G2/M phase arrest and apoptosis in various cancer cell lines (67), strategies adopting depletion of Plk1 in cancer therapy were suggested as promising therapeutic approaches (38-41). The Plk1 20

28 inhibitor named as Volasertib is currently undergoing investigation in phase III clinical trial. Since cancer cells are often characterized as its notorious apoptosis resistance, treatment regimens combining both genetic silencing promoting apoptosis and chemotherapeutic drugs, if simultaneously and specifically delivered to the target tissues or cells, can be powerful and efficient strategies to maximize responsiveness of cancer to the therapy (68). As an example, the liposome-based polyplex combining TNF-α and doxorubicin was successfully targeted at the EGFR, which is overexpressed in tumor cells and showed synergistic therapeutic effect (69). In current study, to assign the cancer-specific theranostic properties to the MagLipo, the therapeutic drugs (Dox) and sirnas (siplk1) were incorporated into the particles, and the surface of MagLipo was modified with an antibody to the EGFR. The loading capacity for the chemotherapeutic agent (Dox) was significantly high as 98%, and delivery of Dox into the nucleus successfully proceeded. Encapsulation of siplk1 in the MagLipo and silencing effect by this system were also well presented. These results highly suggest the capability of MagLipo as a drug delivery vehicle. The acidic microenvironment of tumor, which is one of the reasons for the multidrug resistance, especially for weakly basic chemotherapeutic drug such as doxorubicin (70), has been most utilized intrinsic stimulus for the design of stimulus-sensitive nanopreparations (71). The local stimulus developed by the abnormal ph gradient from tumor s extracellular microenvironment (ph 6.5) to the intracellular organelles like endosome (ph 5.5) and lysosomes (below ph 5.5), plays an important role in the controlled release of the drugs after penetration into the cell (71,72). Controlled release of Dox and siplk1 from the MagLipo system in acidic environment was nicely presented with minimal spillage of payload in 21

29 neutral ph. This may have been a result of the electrostatic interaction between the silicate (with a negative charge) and the Dox (with a positive charged). Pancreas cancer, the target disease in this study, is infamous for its aggressiveness, late presentation, and poor prognosis. Most pancreatic cancer patients are diagnosed in advanced stage and median survival of patients is less than 6 months (73). Even with multimodal therapies, pancreas cancer is one of the most common causes of cancer deaths (74). Outcomes of current study demonstrated that the synergistic apoptotic effect of nanoparticles carrying both Dox and siplk1 on human pancreas cancer cells. Above mentioned processes and results of this study is well presented in the recently published article (46). Considering in vivo imaging studies conducted with iron oxide MNPs for a decade, some of them presented identifiable decrease in the signal intensity of targeted lesion on the MR images (75,76). However, heterogenous, even vague change of signal intensity (77) or small dotlike low intensity in relatively large lesion or indistinguishable small lesion itself (78,79) or pre- and post-contrast images that are not exactly same level of slices (80), have been said that the significant T2 relaxation had occurred. Dense nature of tumor stroma or disordered vasculature of tumor could have hindered the particles from getting into the tumor thoroughly. Most importantly, possible changes in the composition of nanoparticles after being administered to the human body have not been properly investigated yet. Uptake and distribution of nanoparticles in the body are complicated processes which cannot be fully evaluated in vitro system. Therefore, nanomedicine suffers from the tons of fancy results produced in the laboratory based analyses with basic shortage of reliable and reproducible in vivo data, which explains why most of nanoparticles rarely make its debut in clinical field. Even though iron oxide nanoparticles have been 22

30 considered as biocompatible based on the cellular viability tests, current information regarding its potential undesirable and unexpected adverse effects to the whole organism after exposure is scarce, and even existing results from in vitro studies are controversial. Thus, besides diagnostic and therapeutic performance, comprehensive toxicological information of nanoparticles is indispensable part of prerequisite for the biomedical application of MagLipo. Consequently, this multifunctional nanovehicle, EGFR-targeted MagLipo particles showed considerable and synergistic therapeutic effect on pancreatic cancer cells and also displayed physicochemical properties good enough to be a contrast agent for MRI. We thus suggest that the co-delivery of Dox and siplk1 using the MagLipo particle system is highly selective and efficient, which could provide a promising cancer therapy method to be translated from bench to bedside. Various nanoparticle systems have adopted iron oxide core, mesoporous silica, phresponsive property, and simultaneous delivery of therapeutic genes and chemotherapeutic drugs, respectively or in the combination of two or more of them. Recent investigation presented a nanoparticle system combining mesoporous silica, ph-responsive drug delivery, and T 1 -w magnetic resonance imaging (81). However, none of those formulations have all of the properties the MagLipo have. The ideal properties as a drug delivery system, which may be listed as longevity, targetability, stimuli-sensitivity, enhanced intracellular delivery, contrast property (82), and even the co-delivery system of therapeutic gene and chemotherapeutic agent, were all incorporated in current formulation, MagLipo, which make the particle unique and powerful as a nanovehicle in biomedical field. Despite its wide applications in nanomedicine, biodistribution of iron oxide nanoparticles was reported in few studies (83-85) and therefore, there are some concerns regarding its use in human 23

31 being (18). For the clinical application of MagLipo, comprehensive and deep understanding of its biocompatibility, toxicity, biodistribution, and in vivo targeting efficiency is mandatory. 24

32 CONCLUSIONS In this study, we introduced the mesoporous silica shell onto the MNP to reduce oxidative stress from iron core while maintaining the magnetic property. Then, the 2 particles were combined with the liposomal complex system to improve solubility in the buffer solution. Finally, the EGFR antibody-anchored MagLipo particles were applied to BxPC3 pancreatic cancer cells for T 2 -w MR imaging. These particles were detected in the single cancer cell level by magnetic sensing method and proved to be a sensitive diagnostic material. Furthermore, the therapeutic drug and gene could both be incorporated into the MagLipo particles and were released at low ph value that is the hallmark of cancer microenvironment. Co-delivery of Dox and siplk1 using the EGFR-targeting MagLipo particle system showed highly selective and synergistic therapeutic effect for the treatment of pancreatic cancer cells. Combining attractive features of the theranostic nanoparticles all together, namely iron oxide core, mesoporous silica, phresponsive property, and simultaneous delivery of therapeutic genes and chemotherapeutic drugs, a novel nanovehicle platform was proposed. Based on these results, the MagLipo might be promising multifunctional nanovehicle for the specific cancer therapy and imaging at the same time. However, profound knowledge of potential toxicity and biodistribution of the nanoparticles at the organ and cellular levels after administration is indispensable for its application in clinic, as any techniques appeared flawless in vitro cannot guarantee stability, safety, and therapeutic efficacy in vivo system. There remains much to be done to predict what exactly will happen in human body after systemic application of nanoparticles. Currently, nanoparticles approved for clinical use are investigated in 25

33 large number of trials, and steady build-up of information regarding interaction and fate of nanoparticles extracted from those trials will provide basis for the validation of novel nanoparticle systems which are more advanced than currently approved ones. On condition that its path and fate in vivo system are clearly defined, it is expected that the MagLipo, as a smart and versatile nanoplatform that integrates imaging and therapeutic functions, will contribute to the advancement of highly specific and individualized treatment for cancer and simultaneous assessment of therapy response with image in the future. 26

34 27

35 28

36 Figure 1. Schematic illustration and TEM images of MagLipo (A) Synthetic scheme of antibody-conjugated MagLipo particles. (B) TEM images of the 16-nm iron oxide (Fe 3 O 4 ) magnetic nanoparticles (MNPs) showed regular size distribution. (C) TEM images of mesoporous silica shell coated iron oxide, MNP@mSiO 2 particles. (D) High-magnification image of liposome encapsulated MNP@mSiO 2 particles, MagLipo. 29

37 30

38 31

39 Figure 2. Characteristics of various nanoparticles (A) The morphology of 2 particles was clearly revealed porous shell from the high magnification image of TEM. (B) The structural information of 16- nm Fe 3 O 4 nanoparticles (MNPs) and mesoporous silica shell coated MNPs characterized by powder X-ray diffraction (XRD) analysis. Typical inverse spinel structure was observed and MNP@mSiO 2 particle showed broad peak (marked with *) at the low angle resulting from the mesoporous silica shell. (C) The magnetic property of prepared MagLipo particles clearly showed superparamagnetism and saturated magnetism with 95 emu/[fe]g by Vibrating Sample Magnetometer (VSM). (D) The relaxivity of bare MNP, amorphous silica coated MNP (MNP@SiO 2 ), and mesoporous silica shell MNP (MNP@mSiO 2 ) were measured as 286, 70, and 270 mm -1 S -1 respectively by using of TD-NMR (D, Minispec mq20, 0.47 T, Bruker). From the results, the amorphous silica shell (red line) caused decrease in magnetic relaxation, which might be due to the reduced number of water molecules around the MNPs. However, the mesoporous silica shell coated MNPs (blue line) showed relatively higher relaxivity then the amorphous silica shell coated particles and similar relaxivity (r 2 ) value with that of bare MNPs (black line). (E) Illustration of water molecules in the magnetic field in the vicinity of MNP core. The mesoporous silica shell is assumed to possess water molecules nearby the MNP magnetic field. 32

40 33

41 Figure 3. Solution stability of nanoparticles in various conditions (A) Size distributions of the prepared nanoparticles (MNP, 2 and MagLipo) were measured by dynamic light scattering (DLS) analysis and showed relative monodispersity. (B) The water soluble bare MNPs were dissociated in the 1 N HCl solutions within 20 min, but the mesoporous silica shell effectively protected the MNP core in the same condition. (C) Nanoparticles should be dispersed in buffer solution for biological application. The prepared MNP@mSiO 2 particles showed poor solubility in PBS buffer solution and precipitated within 15 min. But the MagLipo particles exhibited high stability and solubility in PBS solution. The absorbance was determined at 450 nm by UV-Vis spectroscopy. 34

42 B Absorbance (a.u.) (siplk1) 35

43 36

44 Figure 4. Characterization of drug or sirna incorporation and controlled release (A) The therapeutic drug (doxorubicin, Dox) was incorporated easily into the mesoporous silica coated MNP and the loading capacity was determined to be 98% by HPLC analysis. The supernatant solution after loading was analyzed for the amount of Dox concentration. (B) The siplk1 with negative charge was first complexed with positively charged protamine protein (PA) based on electrostatic interaction and then incorporated into the MagLipo using the NH 4 HCO 3 gradient method. The ratio of siplk1 and PA was optimized to 1.5 sirnas per PA using an electrophoretic protocol. The loading capacity measured with UV-Vis spectroscopy was 80%. (C) For ph-sensitive controlled release of incorporated therapeutic materials, the MagLipo particles were exposed to acidic conditions (ph 4.7) and found to quickly release Dox in NH 4 HCO 3 gradient samples (control indicates samples without NH 4 HCO 3 gradient) showing 40% release of Dox within 30 min, which is satisfactory for therapeutics. 37

45 38

46 Figure 5. Assessment of cell cytotoxicity and expression level of EGFR receptor (A and B) MagLipo (A) and EGFR antibody-conjugated MagLipo (EGFR-MagLipo) (B) showed high cell viability (< 90%) in the 293T and BxPC3 cell lines after treatment with various concentrations of particle solution for 3 hr. 293T is an embryonic kidney cell line (ATCC) and EGFR negative. BxPC3 is EGFR positive pancreatic cancer cell line. (C) The expression level of EGFR marker receptor protein was characterized by western blotting. The pancreatic cancer (BxPC3) was shown to be a positive cell line with EGFR. And 293T were determined as EGFR-negative cell line. 39

47 40

48 41

49 Figure 6. Magnetic property of MagLipo (A) After being treated with the particles, the EGFR-positive BxPC3 cells were imaged by T 2 - w MR as a phantom study. (B) Non antibody-conjugated MagLipo showed similar intensity as the control (nontreated) cells, but the EGFR-MagLipotreated cells exhibited significantly enhanced black contrast imaging with 80% decreasing T2 values (data are presented as the mean ± SE from triplicate measurements,** p < 0.01). (C) Quantitative magnetic cellular relaxivity for the sensitivity measurements after treatment with EGFR-MagLipo particles. The T 2 % change was the difference in T2 between the EGFR-MagLipo particle-treated cells and the non-treated cells. 42

50 43

51 Figure 7. Uptake of MagLipo particles in the cell (A) Bio-TEM images showed penetrated particles in the cytosol of BxPC3 cell (white arrows; N indicates nucleus). B) The location of Dox after treatment was monitored during delivery by confocal laser scanning microscopy. Images and line profiles are shown directly after administration (top) and 24 hr after administration (bottom). Red indicates the Dox, and the nucleus was stained with DAPI agent to blue color. Plot of intensity profiles along the green lines in the microscopic images are given on the right for red and blue color intensity. After 24 hr, the red and blue were perfectly matched, showing the ability of delivered Dox to penetrate the nucleus. 44

52 45

53 Figure 8. Therapeutic applications of MagLipo particles in vitro (A) The efficiency of siplk1 delivery was characterized by PCR analysis of treated cells and Plk1 was effectively downregulated in the EGFR-MagLipo-treated cells with a 60% decrease (data are the mean ± SE from triplicate measurements, ** p < 0.01). (B) The treated cells were measured for the viability to evaluate the therapeutic effects with respect to time. 46