Cellular delivery of cationic lipid nanoparticle-based SMAD3 antisense oligonucleotides for the inhibition of collagen production in keloid fibroblasts

European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 19–26

Contents lists available at SciVerse ScienceDirect

European Journal of Pharmaceutics and Biopharmaceutics journal homepage: www.elsevier.com/locate/ejpb

Research paper

Cellular delivery of cationic lipid nanoparticle-based SMAD3 antisense oligonucleotides for the inhibition of collagen production in keloid fibroblasts Su-Eon Jin a, Chong-Kook Kim a,b,⇑, Yang-Bae Kim a,⇑ a b

College of Pharmacy, Seoul National University, Seoul, Republic of Korea College of Pharmacy, Inje University, Gimhae, Gyeongnam, Republic of Korea

a r t i c l e

i n f o

Article history: Received 12 January 2012 Accepted in revised form 23 May 2012 Available online 15 June 2012 Keywords: Cationic lipid nanoparticles SMAD3 Antisense oligonucleotides Cellular delivery Keloid fibroblasts Fibrosis

a b s t r a c t SMAD3 is a key player in the TGFb signaling pathway as a primary inducer of fibrosis. The inhibition of SMAD3 production is one strategy to alleviate fibrosis in keloid fibroblasts. In the present study, antisense oligonucleotides (ASOs) against SMAD3 were designed to specifically block the expression of SMAD3. The cationic lipid nanoparticles (cLNs) were formulated to enhance an intracellular activity of SMAD3 ASOs in keloid fibroblasts. This formulation was prepared using melt-homogenization method, composed of 3-[N-(N0 ,N0 -dimethylaminoethane)-carbamol] cholesterol (DC-Chol), dioleoylphosphatidylethanolamine (DOPE), Tween20, and trimyristin as a lipid core (1:1:1:1.3, w/w). The size and zeta potential of cLNs and cLN/ASO complexes were measured using light scattering. AFM was used to confirm the morphology and the size distribution of cLNs and cLN/ASO complexes. The prepared cLNs had a nano-scale sized spherical shape with highly positive charge, which were physically stable without aggregation during the storage. The cLN/SMAD3 ASO complexes were successfully generated and internalized onto keloid fibroblasts without toxicity. After the treatment with cLN/ASO complexes, SMAD3 was inhibited and collagen type I was also significantly suppressed in keloid fibroblasts. These results suggest that SMAD3 ASOs complexed with cLNs have a therapeutic potential to suppress collagen deposition in fibrotic diseases. Therefore, this strategy might be developed to lead to anti-fibrotic therapies. Ó 2012 Elsevier B.V. All rights reserved.

1. Introduction Keloids are a dermal fibroproliferative tumor characterized by an excessive accumulation of extracellular matrix, especially collagen. Collagen deposition in keloids is mediated by an abnormal pathological response to cutaneous wound healing, creating disfiguring scars beyond the original site of skin injury [1]. Current treatment of keloids, simple excision or corticosteroid injection frequently undertaken, results in a very high recurrence rate over 50% [2]. There are no satisfactory treatments in keloid scarring although patients are painful and extensively pruritic [3]. The pathogenesis of keloids is based on the role of several cells such as monocytes and fibroblasts, while its etiology is still largely unknown [4,5]. From the pathological process in the cells, SMAD3 is a key signaling molecule in the TGFb signaling pathway, which is linked with fibrosis [5,6]. Previous studies have shown that SMAD3 plays a crucial role in fibrosis and wound healing and also has predominantly pro-inflammatory and pro-fibrotic activity [5,7]. It ⇑ Corresponding authors. College of Pharmacy, Seoul National University, 1 Gwanak-ro Gwanak-gu, Seoul 151-742, Republic of Korea. Tel.: +82 2 880 7867. E-mail addresses: [email protected], [email protected] (C.-K. Kim), ybkim@ snu.ac.kr (Y.-B. Kim). 0939-6411/$ - see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ejpb.2012.05.015

is also very specific to keloidal disease state compared with normal tissues [8]. Thus, pathway-specific inhibitors with increased specificity can be developed based on the critical role of SMAD3 in mediating the pathogenic effects of TGFb. For the inhibition of SMAD3 protein, antisense oligonucleotides (ASOs) can be introduced to inhibit the production of target protein designed to specifically hybridize to heterogeneous nuclear RNA or mature mRNA sequences. In this study, we used the ASOs against SMAD3 which would inhibit the SMAD3 and down-regulate the collagen production in keloid fibroblasts. Although ASOs were developed to treat several diseases as socalled magic bullets [9,10], only one drug, formivirsen (VitraveneÒ), has been approved by the FDA for the treatment of ocular cytomegalovirus retinitis in AIDS patients [11]. The ASObased therapeutic agents still have limitations of delivery to their eventual sites of action due to their characteristics (e.g., large molecular weight, highly negative charge, rapid degradation by nuclease, etc.). The poor uptake of ASOs into cells can be overcome by the use of cationic lipids that bind to ASOs [12,13]. Current research in this area has demonstrated the potential of cationic lipids (e.g., cholesterol-based lipids, amine-linked synthetic lipids, etc.) and substantial progress has been made in designing and developing novel cationic lipid-based carriers [14].

20

S.-E. Jin et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 19–26

Recently, lipid nanoparticles (LNs) have been developed as an alternative delivery system due to their improved physical stability compared with conventional delivery systems such as liposomes and emulsions [15]. In addition, cationic LNs (cLNs) can be produced carriers of nucleic acids using cationic lipids for the cellular delivery of ASOs [16–18]. We previously studied novel cLNs for enhanced p53 gene transfer into lung cancer cells [19]. The results of this earlier study led us to speculate on the impact of cLNs on cellular delivery of ASOs. To be pharmacologically active, ASOs must be protected and transferred specifically to target cells. In the present study, we proposed that cLN-based SMAD3 ASOs would inhibit the SMAD3 production and successfully suppress the collagen production linked with fibrosis in keloid fibroblasts. The physicochemical properties of cLNs and cLN/ASO complexes were characterized, and their effects of complexes on fibrosis were determined in vitro. 2. Materials and methods 2.1. Reagents

(PBS, pH 7.4) to prepare the cLN/ON complexes. The 1 lg of ONs were separately diluted with PBS for complexation. The complexes were prepared by mixing cLNs and ONs in 200 ll of PBS at the mass ratios of cationic lipid of cLNs to ONs ranged from 1 to 30 (0.00015–0.0044% of total lipids, w/v). The mixtures were vortexmixed and incubated for 20 min at room temperature prior to use. 2.4. Measurement of particle size and zeta potential Particle size of the cLNs and cLN/ASO complexes was measured using light scattering spectrophotometer (Nicomp 370, Particle Sizing System, CA, USA). For the measurements, cLN/ASO complexes were diluted with PBS prior to size measurement. The mean diameters of cLNs and cLN/ASO complexes were monitored in Gaussian mode. Zeta potential of cLNs and cLN/ASO complexes was measured using electrophoretic light scattering spectrophotometer (ELS8000, Otsuka Electronics Co. Ltd., Japan) at the room temperature to assess the surface charge of nanoparticles.

3-[N-(N0 ,N0 -dimethylaminoethane)-carbamol] cholesterol (DCChol) and dioleoylphosphatidylethanolamine (DOPE) were purchased from Avanti Polar Lipids (Albaster, AL, USA). Tween 20 and trimyristin (TM) were obtained from Sigma (Sigma, St. Louis, Missouri, USA). Dulbecco’s modified Eagle’s medium (DMEM), trypsin–ethylenediaminetetraacetic acid (EDTA), LipofectinÒ, and fetal bovine serum (FBS) were purchased from Invitrogen (GibcoBRL, Invitrogen, San Diego, CA, USA). The sequences of SMAD3 ASOs were 50 -GGCCATCGCCACAGGCGGCA-30 , which are designed to inhibit the translation site of smad3 transcripts and has a phosphorothioate modification in each end of three sequences (underlined 50 -GGC and 30 -ACG). The control ONs (ONs) were mismatched form as the same length of SMAD3 ASOs. All of these ONs were purchased from Bioneer (Seoul, Korea). The ONs were dissolved in sterile water and stored at 20 °C. Deionized water was used after filtration through a 0.2-lm filter. All other chemicals were of reagent grade and were used without further purification.

2.5. Physical stability of cLNs and cLN/ASO complexes

2.2. Cell culture

2.6. Gel retardation assay of cLN/ASO complexes

Keloid fibroblasts (KEL FIB) were obtained from the ATCC (American Type Culture Collection, USA). These cells were cultured in DMEM supplemented with 10% FBS, 100 U/ml penicillin and 100 lg/ml streptomycin at 37 °C in humidified atmosphere containing 10% CO2.

The complexation between ASOs and cLNs was determined by agarose gel retardation assay. To detect the ASO bands in this gel, the amounts of ASOs were fixed at 2 lg. The cLNs (2.2% of total lipids, w/v) were diluted with filtered PBS to prepare the cLN/ASO complexes at the ratios of 1–30 (cationic lipid to ASOs, w/w). Ten ll of ASOs (2 lg) and 15 ll of diluted cLNs (0.0003–0.0088% of total lipids, w/v) were vortex-mixed to 25 ll of total suspension and incubated for 20 min. The aliquots of samples (20 ll) were resolved onto a 1.0% agarose gel with ethidium bromide and visualized with the UV transilluminator (Chemi-Imager 4400, Alpha Innotech Co., CA, USA).

2.3. Preparation of cLNs and cLN/ASO complexes The cLNs were prepared by melt homogenization method with slight modification [20]. Briefly, DC-Chol, DOPE, Tween 20 and TM (1:1:1:1.3, w/w) were mixed and dissolved in approximately 1 ml of tertiary butylalcohol. After rapid freezing of suspension in liquid nitrogen tank, mixtures were dried in Ultra 35 EL freeze-dryer (Virtis, USA). Finely dispersed cakes were obtained after overnight drying, and then, cakes were put in water bath at 60–65 °C. Preheated water at 60–65 °C for injection was slowly added to the melts and sonicated in bath type sonicator for 90 min at 60– 65 °C until crude and milky emulsions were obtained. These crude emulsions were homogenized for 7 cycles at 60–65 °C and 100 MPa using a high pressure homogenizer (Emulsiflex EF-B3, Avestin Inc., Canada) wired with heating tape (Thermolyne). The cLNs were produced by subsequent cooling of homogenized emulsions in liquid nitrogen, then thawed and stored at 4 °C. The cLNs were formulated at the concentration of 2.2% of total lipids (w/v). The cLNs were diluted in phosphate buffered saline

The physical stability of the cLNs was investigated with monitoring both particle size and zeta potential during the storage. The cLNs from three batches were stored at room temperature and 4 °C for 50 days. For the preparation of samples, the stored cLNs were vortex-mixed before dilution. The cLNs were diluted with filtered PBS and vigorously vortex-mixed to measure particle size and zeta potential. The method of measuring particle size and zeta potential of cLNs was mentioned above in Section 2.4 of materials and methods. For the stability of cLN/ASO complexes, particle size of cLN/ASO complexes was monitored at various time points (0, 2, 12, and 24 h) after the preparation. The cLN/ASO complexes were prepared and stored at room temperature for 24 h. The cLNs were diluted with filtered PBS and vortex-mixed for 3 s. The particle size was measured as mentioned above in Section 2.4.

2.7. Nuclease protection assay of cLN/ASO complexes The cLNs were mixed with 0.5 lg of ASOs at a cLN/ASO ratio of 27 (w/w) and then, the complexes were incubated at room temperature for 20 min. For digestion by DNase I, cLNs were prepared in water instead of PBS to avoid possible interference of salt on DNase I activity. ASOs, and cLN/ASO complexes containing the same amount of ASOs were exposed to DNase I (0.25 U and 0.50 U) for 30 min at 37 °C. LipofectinÒ was used as a positive control. ASOs were retrieved by phenol/chloroform extraction followed by ethanol precipitation. The ASOs were then visualized on a 1% agarose gel containing ethidium bromide under UV transillumination (Chemi-Imager 4400).

S.-E. Jin et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 19–26

2.8. Atomic force microscopy The morphological features of cLNs and cLN/ASO complexes were studied by atomic force microscopy (AFM). For the dilution of samples, filtered water was used to prevent salt crystal formation for AFM measurement. Five ll of the cLN/ASO complexes at the mass ratio of 27 was deposited onto freshly cleaved mica for 5 min. Then, sample solution was removed by gently swirling, and 20 ll of water was spotted onto the mica for 10 s. Water was removed by gentle swirling and mica was air-dried for 30 min. Images were acquired using Scanning Probe Microscope (AutoProbe CPTM, PSIA, USA) in a tapping mode. 2.9. MTT assay Cell viability of cLNs and cLN/ASO complexes was determined using 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. The cells were seeded at 1  103 cells/well onto 96-well culture plates and incubated for 24 h. The cLNs and the cLN/ASO complexes in serum-free medium were added into the cells. After 4 h treatment, medium was removed and replaced with the serum-containing growth medium. After 24 h incubation, 20 ll of MTT solution (5 mg/ml in PBS) was added to each well and incubated for 4 h. Then, solution was removed from each well and formazan crystals produced by living cells were dissolved in 100 ll of dimethylsulfoxide. The absorbance was monitored at 540 nm in a microplate reader (MCC340, Multiskan, Belgium). The cell viability (%) was defined relative to untreated control cells.

Cell viability ð%Þ ¼ ½OD540 ðsampleÞ=OD540 ðcontrolÞ  100 2.10. Confocal microscopy Confocal microscopy was used to confirm the uptake of cLN/ ASO complexes into the keloid fibroblasts. The cells were grown on 24  24 mm coverslips at 1  104 cells/cm2 density. In this case, fluorescein isothiocyanate (FITC)-labeled DOPE was added to cLNs (0.1% of DOPE) to prepare the cLNs for confocal microscopy (FITClabeled cLNs). The untreated and the ASO-unloaded cLNs were used as controls. The FITC-labeled cLN/ASO complexes were prepared in serum-free medium. After 1, 2, and 4 h incubation of complexes, cells were rinsed with PBS and observed on confocal microscopy (Leica TCS NT, Leica Microsystems, Wetzlar, Germany) supplemented with an argon–krypton laser and equipped with a 200 magnification. The excitation wavelength was 488 nm. The acquisitions were recorded as normal representations. 2.11. Treatment of cLN/ASO complexes Keloid fibroblasts were seeded onto 6-well culture plates at a cell density of 1  105 cells/well. These cells were allowed to adhere as a monolayer overnight and to achieve 70–80% confluence. The cLN/ASO complexes in 500 ll of serum-free medium were added to each well. The amount of ASOs was used at 1 lg, which was mixed with cLNs at the mass ratio of 15 and 27. After 4 h incubation, cells were rinsed with PBS and incubated for 24–48 h in 2 ml of serum-containing growth medium. 2.12. Western blot for SMAD3 The SMAD3 expression in keloid fibroblasts was determined using Western blot analysis. The treated cells were harvested and lyzed in lysis buffer (150 mM NaCl; 20 mM Tris, pH 5.1; 1 mM PMSF; 1 mM Na3VO4; 25 mM NaF; 1% Aprotinin; 10 mg/ml Leuprotinin; 1% Triton X-100; 1% NP-40) in ice for 30 min. Then samples were centrifuged at 12,000 rpm for 10 min, and the protein

21

concentrations of supernatants were determined using Bio-Rad detergent-compatible microprotein assay using bovine serum albumin (BSA) as a standard protein. Ten-microgram aliquots of protein were separated by sodium dodecylsulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and transferred to nitrocellulose filters. The nitrocellulose membrane was incubated for 1 h in a blocking buffer (5% nonfat milk in PBS) followed by incubation with the rabbit antihuman SMAD3 antibody (1:200, Zymed, San Francisco, USA). 2.13. Determination of collagen type I using ELISA The amounts of collagen type I were measured using ELISA as manufacturer’s instruction with slight modification (Chondron, USA). For the calculation of collagen titer, the samples were serially diluted to express the titer as a relative log-scale value (2x) and checked the optical densities of active response between samples from collagen type I and collagen type I specific antibody. Briefly, ELISA plates were coated with 150 ll of the appropriate dilutions of samples in 0.1 M carbonate/bicarbonate buffer (pH 9.6) at 4 °C overnight. After coating, 200 ll of 2% BSA (w/v) were added into each well and incubated for 1.5 h at 37 °C to block non-specific protein binding. The 100 ll of collagen type I specific antibody (1:500, Southern Biotech. Associates, Birmingham, AL, USA) was added into each well After incubation at 37 °C for 2 h, 100 ll of goat anti-rabbit IgG-alkaline phosphatase conjugated complex (1:1000, Sigma) was added following washings. After incubation at 37 °C for 2 h and subsequent washing, p-nitro-phenylphosphate (PNPP, Sigma) as a substrate for alkaline phosphatase was added. The optical density was determined at 405 nm after 30 min incubation. 2.14. Statistical analysis All results are expressed as means ± SD. Statistical analysis of the data was performed using Student’s t-test and ANOVA. A p value of less than 0.05 was considered significant. 3. Results 3.1. Size and stability of cLNs for storage To characterize the physicochemical properties of the cLNs, particle size and zeta potential were measured using light scattering. We monitored the size and zeta potential of cLNs during the storage at 4 °C and room temperature. The cLNs themselves were physically stable with a nano-scale size (80–100 nm) (Fig. 1a) and a strongly positive charge (30–40 mV) (Fig. 1b). As shown in Fig. 1a, the cLNs were maintained in terms of particle size after storage at 4 °C for 50 days. For the zeta potential, cLNs stored at room temperature had relatively low values (25–30 mV) at the early time points of storage (Fig. 1b). Therefore, we stored the cLNs at 4 °C to maintain nano-particular properties with highly positive surface charge in this study. 3.2. Complexation of cLNs and ASOs For the delivery of ASOs, we prepared cLN/ASO complexes via charge–charge interaction between cLNs and ASOs. As the mass ratio of cLNs increased from 3 to 30, particle size decreased from 1.8 lm to 194 nm and the zeta potential increased from 25 up to +38 mV (Fig. 2a). The zeta potential of the complexes reached a saturated positive charge at a mass ratio of 15. The particle size of complexes at the ratios of 15–30 was 180–220 nm, which was proposed to use for cellular delivery of ASOs into keloid fibroblasts in terms of the stable complexation between cLNs and ASOs. We

22

S.-E. Jin et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 19–26

Fig. 2. Physicochemical properties of cLN/SMAD3 ASO complexes at the various mass ratios of cationic lipid to SMAD3 ASOs. (a) Particle size and zeta potential of cLN/ASO complexes. The results are expressed as mean ± SD (n = 3). (b) Gel retardation of cLN/ASO complexes on agarose gel. The mass ratios of cationic lipid to SMAD3 ASOs were from 1 to 30. ASO Bands from cLN/SMAD3 ASO complexes were retarded and disappeared in a mass ratio-dependent manner of cLN/ASO complexes.

Fig. 1. Temperature-based physical stability of cLNs during the storage. (a) Particle size and (b) zeta potential were monitored at the storage of 4 °C and room temperature. The results are expressed as mean ± SD (n = 3).

additionally confirmed these results using gel retardation assay with cLN/ASO complexes. The band of cLN/ASO complexes was retarded on the agarose gel as the mass ratio of cLNs increased from 3 to 30 compared with the band of ASOs alone (Fig. 2b). In Fig. 2b, the band of complexes was completely disappeared at the ratio of 27. From the results, two ratios of 15 and 27 were selected to use for further study determining ASO activity. 3.3. Stability of cLN/ASO complexes 3.3.1. Physical stability of cLN/ASO complexes The stability of cLN/ASO complexes was determined using the measurement of particle size (Fig. 3a). The cLN/ASO complexes were used at the mass ratio of 27 for this stability study. The size of cLN/ASO complexes was not significantly changed for 24 h after the preparation of complexes, which were 180–200 nm (Fig. 3a). In addition, no aggregation was detectable after complex formation. 3.3.2. Nuclease protection of cLNs from cLN/ASO complexes To confirm the resistance of the complexes to degradation by DNase I, the cLN/ASO complexes at the ratio of 27 were treated with DNase I, followed by gel electrophoresis to confirm the ASO band. In Fig. 3b, ASOs in cLN/ASO complexes were protected from DNase I at both concentrations of DNase I (0.25 and 0.50 U), while

Fig. 3. Stability of cLN/ASO complexes. (a) Particle size of cLN/ASO complexes at room temperature for 24 h. The cLN/ASO complexes were used at the mass ratio of 27. The results are expressed as mean ± SD (n = 3). (b) Nuclease protection of cLNs from cLN/ASO complexes. DNase I was used for nuclease digestion at the concentration of 0.25 and 0.50 U. The cLN/ASO complexes were used at the mass ratio of 27. LipofectinÒ was used as a positive control.

S.-E. Jin et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 19–26

23

determined whether the cell integration of cLNs and cLN/ON complexes for 4 h incubation could be affected to cell viability after further 24 h incubation. These were examined using MTT assay. The cLNs from 0.001% to 0.1% (the concentration of total lipids, w/v) (Fig. 5a) and the cLN/ON complexes at mass ratios varying from 3 to 27 (Fig. 5b) were transfected into keloid fibroblasts. The concentrations of cLNs from cLN/ON complexes matched for 0.00044– 0.00396% of total lipids (w/v), which are the mass ratios of 3–27. Cell viability of the cLNs was over 80% at the ranges of 0.001– 0.01% (Fig. 5a). The cLN/ON complexes had a minimum toxicity showing over 80% of cell viability (Fig. 5b). Therefore, the cLNs could be used for cellular delivery of ASOs without cytotoxicity.

the ASOs were cleaved by enzymatic digestion even at the lower concentration (0.25 U). The ASOs were treated as described in the experimental procedures, except for DNase I treatment, as controls. The nuclease-protection effect of the cLNs on ASOs was comparable with that of the commercial formulation, LipofectinÒ, which was utilized as a positive control. 3.4. Morphological analysis using AFM AFM reveals the particle size distribution and the surface morphology of cLNs and cLN/ASO complexes. The cLN/ASO complexes at the ratio of 27 were used for the examination of complex formation. Fig. 4 shows the homogeneous dispersion of cLNs (Fig. 4a) with spherical shape and the formation of complexes between cLNs and ASOs (Fig. 4b). The each size of cLNs and cLN/ ASO complexes was 80–100 nm and 180–220 nm, respectively. After the complex formation, free ASOs from cLN/ASO complexes were not detectable in AFM images. These results of particle size and size distribution were consistently obtained using both light scattering and AFM.

3.6. Cellular uptake of cLN/ON complexes into keloid fibroblasts To confirm the cellular delivery of ASOs, FITC-labeled cLN-based ON complexes (27:1, w/w) were transfected into keloid fibroblasts and monitored using confocal microscopy. Fig. 6a presents the schematic diagram of the general mechanism of cellular uptake of cLN/ON complexes and the expected ASO action in the cells. Fig. 6cde summarizes confocal images captured from 1 h to 4 h incubation after the treatment of cLN/ON complexes. Green fluorescence in Fig. 6cde represents FITC-labeled cLN/ON complexes. For the controls, Fig. 6b and f show green fluorescence from an untreated control (blank) and ON-unloaded cLNs (cLNs),

3.5. Cytotoxicity The cLNs were characterized to have a toxic effect on keloid fibroblasts as a reagent for cellular delivery of ASOs. We

(a')

(a) 0.8

0.6

0.4

0.2

0

0.2 0.00

0.4

0.6

0

0.8

[nm]

[µm] 21.05

(b)

(b') 0.8

0.6

0.4

0.2

0

0.2 0.00

0.4 [nm]

0.6

0

0.8

[µm] 36.79

Fig. 4. Representative AFM images with 3D views of (aa0 ) cLNs and (bb0 ) cLN/SMAD3 ASO complexes (27:1, w/w). AFM images were acquired using Scanning Probe Microscope in a tapping mode. Scan areas for cLNs and cLN/ASO complexes were 1.0 lm  1.0 lm (2.0 Hz). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

24

S.-E. Jin et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 19–26

Fig. 5. Cell viability of (a) cLNs and (b) cLN/ON complexes in keloid fibroblasts. The concentrations of cLNs were used at the range of 0.001–0.1% (concentration of total lipids, w/v). The complexes were used at the ratios of cationic lipid to control ONs (w/w) from 3 to 27. Cell viability was obtained using MTT assay. The results are expressed as mean ± SD (n = 3).

respectively. Increase in green fluorescence from cLN/ON complexes was examined in the cells as incubation time increased from 1 h to 4 h (Fig. 6cde). The extended green fluorescence spots were observed like aggregation (pointed by red arrows) at 2 h and 4 h incubation of complexes (Fig. 6de). Green fluorescence was homogeneously distributed in the cell after 4 h incubation (Fig. 6e) compared with 1 h (Fig. 6c) and 2 h (Fig. 6d) incubation. After 4 h treatment of cLN/ON complexes, the cellular uptake was comparable with that of cLN alone (Fig. 6f). Fig. 6bcde shows an incubation time-dependent accumulation of cLN/ON complexes into the cell. 3.7. Anti-fibrotic activity of SMAD3 ASOs with cLNs in keloid fibroblasts 3.7.1. Inhibition of SMAD3 We monitored SMAD3 expression in keloid fibroblasts after the treatment of SMAD3 ASOs with cLNs. The relative expression of SMAD3 in the cells was analyzed by Western blotting. Fig. 7a shows the representative image of SMAD3 expression in the cells. The relative band density (RBD) of SMAD3 was analyzed using Image J (NIH). When SMAD3 expression of PBS-treated keloid fibroblasts (PBS, negative control) was normalized as 1.00, the relative expression of SMAD3 was calculated. In Fig. 7a, SMAD3 ASOs from cLN/ASO complexes (RBD – 15:1; 0.28 and 27:1; 0.21) remarkably inhibited SMAD3 expression compared with PBS (RBD – 1.00) and SMAD3 ASOs alone (RBD – 0.51), indicating cLNs enhanced the uptake of SMAD3 ASOs into the cells. Compared with LipofectinÒ/ASO (RBD – 0.23) as a positive control, cLN/ASO complex

Fig. 6. Internalization of cLN/ASO complexes into keloid fibroblasts. (a) Schematic diagram of the cellular uptake of cLN/ASO complexes and the expected ASO action in the cells. Uptake of cLN/ON complexes into keloid fibroblasts at (c) 1, (d) 2, and (e) 4 h incubation. (b) The untreated cells (blank) and (f) ON-unloaded cLNs-treated cells at 4 h incubation (cLNs) were used as controls. These images were obtained at 200 magnification using confocal microscopy to confirm the cellular uptake of complexes into keloid fibroblasts. Green spots (cde) represent FITC-labeled cLNs from cLN/ON complexes at the ratio of 27. In particular, extended green spots like aggregation were pointed by red arrows. Scale bar; 10 lm. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

had a comparable repression of SMAD3 expression (RBD – 15:1; 0.28 and 27:1; 0.21), suggesting SMAD3 expression was successfully down-regulated by SMAD3 ASOs from cLN/SMAD3 ASO complexes. 3.7.2. Suppression of collagen type I To examine the possible effect of SMAD3 ASOs from cLN/ASO complexes on the extracellular matrix, the level of collagen type I was analyzed using ELISA (Fig. 7b). Collagen type I was significantly decreased by 4- to 9-fold after the treatment of cLN/SMAD3 ASO compared with the treatment of ASOs. In addition, cLN/SMAD3 ASO had a comparable suppressive potential of collagen type I to LipofectinÒ/SMAD3 ASO. These results also indicate that the cLNbased SMAD3 ASOs enhanced the intracellular activity of ASOs through complexing with cLNs. 4. Discussion Herein, we report enhanced cellular delivery of SMAD3 ASOs with cLNs in keloid fibroblasts. TGFb/SMAD3 signaling pathway is majorly involved in the multiple processes of collagen formation that are linked with fibrosis and wound healing in macrophages and fibroblasts. SMAD3 ASOs in this study were designed to inhibit the translation of smad3 gene, which could ultimately suppress the overexpression of collagen through blocking the TGFb/SMAD3 signaling pathway in the cells. At this point, the cLNs were prepared for the cellular delivery of SMAD3 ASOs to overcome the several disadvantages of ASO-based therapeutics (e.g., highly

S.-E. Jin et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 19–26

Fig. 7. The effect of cLN/SMAD3 ASO complexes on SMAD3 and collagen type I levels in keloid fibroblasts. (a) SMAD3 inhibition by cLN/ASO complexes. Inhibition of SMAD3 production was monitored using Western blotting at 24 h incubation after 4 h treatment of cLN/SMAD3 ASO complexes. The values of relative band density (RBD) were analyzed using Image J (NIH). (b) Suppression of collagen type I. The each sample was collected at 24 h and 48 h incubation after 4 h treatment of cLN/SMAD3 ASO complexes and diluted appropriately to measure collagen titer using ELISA. The results are expressed as mean ± SD (n = 3). Denotes p < 0.05.

negative charge, large molecular weight, instability in the cells, etc.) and to enhance intracellular activity of ASOs in the cells. In this study, we formulated the physically stable cLNs, which were nano-sized and positively charged (Fig. 1). The cLNs were composed of DC-Chol, DOPE, Tween 20 and TM. DC-Chol is a commonly used cholesterol-based cationic lipid to give positive charge to the cLNs for the formation of complexes with ASOs. DOPE is a fusogenic lipid for improved association of the cLNs onto cell membrane and for enhanced cellular uptake to keloid fibroblasts. Similar observations to physicochemical characteristics of DC-Chol/ DOPE-based formulations were previously reported by Zhang et al. [21] and Kearn et al. [22]. Tween20, a non-ionic surfactant, can cover the surface of this cLN formulation to prepare the stable and spherical nanoparticles in this study. TM (melting point, 54 °C), a triglyceride as a lipid core of the cLNs, inhibits particle aggregation and coalescence. Because the cLNs include a solid lipid core, a lipid dispersion is physically stable without aggregation during the storage [23]. Moreover, we have previously studied TM-cored LNs carrying paclitaxel with comparable characteristics for this formulation [20]. The cLN formulation was essential to have a sufficient stability and binding potential to ASOs as a cellular delivery system. The cLNs prepared in this study successfully formed complexes coupling with ASOs via charge–charge interaction (positive from cLNs and negative from ASOs). Fig. 2a shows the size and the zeta potential of cLN/ASO complexes vs mass ratios of cationic lipid to ASOs (w/w). The stable complexes were generated over the mass ratio of 15 maintaining 180–220 nm and 35–40 mV. Fig. 2b presents the gel retardation of cLN/ASO complexes at the given mass ratio preventing the mobility of ASOs on the gel in an electric field. From

25

the results, we confirmed that the cLNs had the high binding and condensing potentials of ASOs over the ratio of 15 of cLN/ASO complexes. It suggests that the immobilization of ASOs on the cLNs was majorly governed by the surface characteristics of the cLNs based on the size and the zeta potential of the cLNs with a stable surface layer and the stable interaction between cationic lipid and solid lipid [24]. For the stability of cLN/ASO complexes, we monitored the size of complexes for 24 h and the protection effect of complexes on the nuclease digestion of ASOs. The size of cLN/ASO complexes was maintained for 24 h after complexation (Fig. 3a). In addition, the cLN/ASO complexes had a protective property to inhibit the digestion of ASOs from nuclease comparing to LipofectinÒ/ASO complexes (Fig. 3b). From these results, the cLN/ASO complexes were physically stable and enzyme-protective. It strongly suggests that the cLNs can be applied to carry ASOs with their therapeutic potential. AFM images of the cLNs provide valuable information of morphology and physical characteristics of formulations - size, particular distribution, shape, and structure [25]. Fig. 4 shows the representative AFM images of cLNs and cLN/ASO complexes (27:1, w/w) with 3D structures in a tapping mode. The spherical cLNs were narrowly distributed without aggregation. In addition, the cLN/ASO complexes were visualized to confirm the generation of complexes between cLNs and ASOs. The size results from AFM were comparable to those from light scattering. All results of physicochemical characterization consistently indicate that the cLNs were successfully formulated and cLN/ASO complexes were readily prepared to apply for the cellular delivery of ASOs. The cytotoxicity of cLNs and cLN/ON complexes was monitored using MTT assay (Fig. 5). The cytotoxicity study was considered not only to estimate the working concentration of cLNs but also to show no toxicity of all ratios of cLN/ON complexes at the working concentration of cLNs. The cLNs at 0.001–0.01% (w/v) as a concentration of total lipids had a minimum toxicity over 80% of cell viability (Fig. 5a). Furthermore, none of complexes at any ratios was toxic to cells (Fig. 5b). These concentrations of cLNs and cLN/ON complexes were acceptable to use for further transfection study of complexes in the cells. Next, we examined whether the cLN/ASO complexes could be internalized and localized into the cells to have a therapeutic effect of ASOs on the keloid fibroblasts. Fig. 6 presents the schematic diagram of proposed cellular uptake mechanism of complexes (Fig. 6a) and the cellular uptake of cLN/ASO complexes into the keloid fibroblasts (Fig. 6bcde). For the general uptake mechanism of cLN/ASO complexes, the cLN-based ASOs can be internalized to cells via charge-based membrane interaction [26]. The endocytosis is a major mechanism of intracellular uptake of ASOs with nanocarriers. Then, complexes lead to entrap the vesicles such as endosome and lysosome. After the release of ASOs from intracellular vesicles, ASOs in cytoplasm have an expected activity to bind target mRNA inhibiting production of target protein. Additionally, the released ASOs in cytoplasm also enter the nucleus. Nuclear entry of ASOs, specifically phosphorothioate backbone-based ASOs, occurs to continuously shuttle between nucleus and cytoplasm, which is an active process using nuclear pore structure [27]. This step may not be a necessary to have an activity of ASOs, but ASOs are able to encounter their therapeutic target in the nucleus. It can synergistically enhance the pharmacological activity of ASOs to inhibit the target protein in the cells. We determined the expected mechanism-based cellular uptake of cLN/ASO complexes using confocal microscopy. Intensity of green fluorescence from FITC-labeled cLN/ASO complexes increased with longer incubation time of complexes (1 h < 2 h 6 4 h). The uptake of cLN/ASO complexes was comparable to that of ASO-unloaded cLNs at 4 h of incubation indicating the enhanced

26

S.-E. Jin et al. / European Journal of Pharmaceutics and Biopharmaceutics 82 (2012) 19–26

uptake of cLN/ASO complexes based on the improvement of stability of complexes. The stability of complexes attributed to physicochemical characteristics of complexes, which were highly positive (high zeta potential) and physically stable (homogenous nano-suspension maintaining particle size after complex formation). In addition, nuclease-protection effect of cLN/ASO complexes illustrated the protection of ASOs from the interaction between cLN/ASO complexes and macromolecules (e.g., enzymes). The extended spot size of green fluorescence like aggregation (pointed by red arrows) was detected as incubation time increased in Fig. 6d (2 h) and e (4 h). These extended spots were also detectable in ASO-unloaded cLNs at 4 h incubation (Fig. 6f). These results might be explained by an accumulation of the cLNs or the cLN/ASO complexes based on the cellular uptake process, or a rupture of vesicular compartments after cellular uptake of complexes although intracellular trafficking of complexes was needed to be further determined with multiple fluorescence labels. The enhanced uptake potential of cLNs is comparable to that of previous study performed by Choi et al., which shows the intracellular uptake of plasmid DNA-loaded cLNs [19]. The cLN-based SMAD3 ASOs not only significantly inhibited SMAD3 protein expression (Fig. 7a), but also suppressed collagen type 1 levels indirectly (Fig. 7b) in the keloid fibroblasts. Keloid fibroblasts were used to screen the intracellular therapeutic effect of SMAD3 ASOs on target protein, in which SMADs-mediated TGFb signaling pathway was detected from previous reports [28,29]. Our data consistently shows reduced production of SMAD3 in keloid fibroblasts following treatment with cLN/ASO complexes. Then, inhibition of SMAD3 was related to collagen reduction based on the pathological process of keloid fibroblasts. Similar observation of these finding was reported by Wang et al. [29]. The inhibition of SMAD3 could additionally suppress the TGFb at the cells based on the critical function of SMAD3 and the feedback mechanism in the TGFb pathway from the previous study [6,30]. It suggests that the inhibition of SMAD3 as well as the down-regulation of TGFb could help the strong effect of SMAD3 ASO on the reduction of collagen. Based on the delivery of ASOs with the cLNs, SMAD3 ASOs can be used to block the translation of SMAD3 that are critical to fibrotic disease onset and progression. It suggests that the profound cellular effects of SMAD3 ASOs can provide potential tools and therapeutics reagents for developing anti-fibrotic therapeutics. 5. Conclusions The prepared cLNs have an enhanced cellular uptake potential of SMAD3 ASOs into keloid fibroblasts, which were physically stable and easily prepared complexes with ASOs. This strategy of cLNs/SMAD3 ASO complexes can be developed for the in vivo clinical use of nanoparticle-based fibrosis therapies inhibiting SMAD3 production and suppressing collagen type I. Acknowledgements This study was financially supported by the Ministry of Education and Human Resources Development (MOE), the Ministry of Commerce, Industry and Energy (MOCIE), and the Ministry of Labor (MOLAB) through the fostering project of the Lab of Excellency. References [1] O. Seifert, U. Mrowietz, Keloid scarring: bench and bedside, Arch. Dermatol. Res. 301 (2009) 259–272. [2] A. Al-Attar, S. Mess, J.M. Thomassen, C.L. Kauffman, S.P. Davison, Keloid pathogenesis and treatment, Plast. Reconstr. Surg. 117 (2006) 286–300.

[3] O. Bock, G. Schmid-Ott, P. Malewski, U. Mrowietz, Quality of life of patients with keloid and hypertrophic scarring, Arch. Dermatol. Res. 297 (2006) 433– 438. [4] D.G. Greenhalgh, The role of apoptosis in wound healing, Int. J. Biochem. Cell Biol. 30 (1998) 1019–1030. [5] K.C. Flanders, Smad3 as a mediator of the fibrotic response, Int. J. Exp. Pathol. 85 (2004) 47–64. [6] P. Bonniaud, P.J. Margetts, K. Ask, K. Flanders, J. Gauldie, M. Kolb, TGF-beta and Smad3 signaling link inflammation to chronic fibrogenesis, J. Immunol. 175 (2005) 5390–5395. [7] J. Zhao, W. Shi, Y.L. Wang, H. Chen, P. Bringas Jr., M.B. Datto, J.P. Frederick, X.F. Wang, D. Warburton, Smad3 deficiency attenuates bleomycin-induced pulmonary fibrosis in mice, Am. J. Physiol. Lung Cell Mol. Physiol. 282 (2002) L585–593. [8] G.S. Chin, W. Liu, Z. Peled, T.Y. Lee, D.S. Steinbrech, M. Hsu, M.T. Longaker, Differential expression of transforming growth factor-b receptors I and II and activation of Smad 3 in keloid fibroblasts, Plast. Reconstr. Surg. 108 (2001) 423–429. [9] C.A. Stein, Y.C. Cheng, Antisense oligonucleotides as therapeutic agents – is the bullet really magical?, Science 261 (1993) 1004–1012 [10] S.T. Crooke, Progress in antisense technology, Annu. Rev. Med. 55 (2004) 61– 95. [11] J. Temsamani, G.S. Pari, P. Guinot, Antisense approach for the treatment of cytomegalovirus infection, Expert. Opin. Investig. Drugs 6 (1997) 1157– 1167. [12] I.S. Zuhorn, J.B. Engberts, D. Hoekstra, Gene delivery by cationic lipid vectors: overcoming cellular barriers, Eur. Biophys. J. 36 (2007) 349–362. [13] M. Junghans, S.M. Loitsch, S.C. Steiniger, J. Kreuter, A. Zimmer, Cationic lipid– protamine–DNA (LPD) complexes for delivery of antisense c-myc oligonucleotides, Eur. J. Pharm. Biopharm. 60 (2005) 287–294. [14] C. Tros de Ilarduya, Y. Sun, N. Düzgünesß, Gene delivery by lipoplexes and polyplexes, Eur. J. Pharm. Biopharm. 40 (2010) 159–170. [15] R.H. Muller, K. Mader, S. Gohla, Solid lipid nanoparticles (SLN) for controlled drug delivery – a review of the state of the art, Eur. J. Pharm. Biopharm. 50 (2000) 161–177. [16] C. Olbrich, U. Bakowsky, C.-M. Lehr, R.H. Müller, C. Kneuer, Cationic solid-lipid nanoparticles can efficiently bind and transfect plasmid DNA, J. Control. Release 77 (2001) 345–355. [17] N. Pedersen, S. Hansen, A.V. Heydenreich, H.G. Kristensen, H.S. Poulsen, Solid lipid nanoparticles can effectively bind DNA, streptavidin and biotinylated ligands, Eur. J. Pharm. Biopharm. 62 (2006) 155–162. [18] A. del Pozo-Rodríguez, D. Delgado, M.A. Solinís, A.R. Gascón, J.L. Pedraz, Solid lipid nanoparticles: formulation factors affecting cell transfection capacity, Int. J. Pharm. 339 (2007) 261–268. [19] S.H. Choi, S.-E. Jin, M.-K. Lee, S.-J. Lim, J.-S. Park, B.-G. Kim, W.S. Ahn, C.-K. Kim, Novel cationic solid lipid nanoparticles enhanced p53 gene transfer to lung cancer cells, Eur. J. Pharm. Biopharm. 68 (2008) 545–554. [20] M.-K. Lee, S.-J. Lim, C.-K. Kim, Preparation, characterization and in vitro cytotoxicity of paclitaxel-loaded sterically stabilized solid lipid nanoparticles, Biomaterials 28 (2007) 2137–2146. [21] Y. Zhang, H. Li, J. Sun, J. Gao, W. Liu, B. Li, Y. Guo, J. Chen, DC-Chol/DOPE cationic liposomes: a comparative study of the influence factors on plasmid pDNA and siRNA gene delivery, Int. J. Pharm. 390 (2010) 198–207. [22] M.D. Kearns, Y.N. Patel, M. Savva, Physicochemical characteristics associated with transfection of cationic cholesterol-based gene delivery vectors in the presence of DOPE, Chem. Phys. Lipids 163 (2010) 755–764. [23] B. Siekmann, K. Westesen, Submicron-sized parenteral carrier systems based on solid lipids, Pharm. Pharmacol. Lett. 1 (1992) 123–126. [24] S. Doktorovova, R. Shegokar, E. Rakovsky, E. Gonzalez-Mira, C.M. Lopes, A.M. Silva, P. Martins-Lopes, R.H. Muller, E.B. Souto, Cationic solid lipid nanoparticles (cSLN): structure, stability and DNA binding capacity correlation studies, Int. J. Pharm. 420 (2011) 341–349. [25] J. Sitterberg, A. Özcetin, C. Ehrhardt, U. Bakowsky, Utilising atomic force microscopy for the characterisation of nanoscale drug delivery systems, Eur. J. Pharm. Biopharm. 74 (2010) 2–13. [26] A. del Pozo-Rodríguez, D. Delgado, M.A. Solinís, A.R. Gascón, J.L. Pedraz, Solid lipid nanoparticles for retinal gene therapy: transfection and intracellular trafficking in RPE cells, Int. J. Pharm. 360 (2008) 177–183. [27] R.L. Juliano, X. Ming, O. Nakagawa, Cellular uptake and intracellular trafficking of antisense and siRNA oligonucleotides, Bioconjug. Chem. 23 (2) (2011) 147– 157. [28] T.T. Phan, I.J. Lim, O. Aalami, F. Lorget, A. Khoo, E.K. Tan, A. Mukhopadhyay, M.T. Longaker, Smad3 signalling plays an important role in keloid pathogenesis via epithelial–mesenchymal interactions, J. Pathol. 207 (2005) 232–242. [29] Z. Wang, Z. Gao, Y. Shi, Y. Sun, Z. Lin, H. Jiang, T. Hou, Q. Wang, X. Yuan, X. Zhu, H. Wu, Y. Jin, Inhibition of Smad3 expression decreases collagen synthesis in keloid disease fibroblasts, J. Plast. Reconstr. Aesthet. Surg. 60 (2007) 1193– 1199. [30] G. Latella, A. Vetuschi, R. Sferra, V. Catitti, A. D’Angelo, G. Zanninelli, et al., Targeted disruption of Smad3 confers resistance to the development of dimethylnitrosamine-induced hepatic fibrosis in mice, Liver Int. 29 (7) (2009) 997–1009.