Lipid-Based Liquid Crystalline Nanoparticles Facilitate Cytosolic

21 Mar 2018 - Lares, M. R.; Rossi, J. J.; Ouellet, D. L. Trends Biotechnol. 2010, 28 (11), 570– 9, DOI: 10.1016/j.tibtech.2010.07.009. [Crossref], [...
1 downloads 0 Views 8MB Size
Download PDF
Letter Cite This: Nano Lett. XXXX, XXX, XXX−XXX

pubs.acs.org/NanoLett

Lipid-Based Liquid Crystalline Nanoparticles Facilitate Cytosolic Delivery of siRNA via Structural Transformation Shufang He,†,‡ Weiwei Fan,†,‡ Na Wu,† Jingjing Zhu,† Yunqiu Miao,† Xiaran Miao,§ Feifei Li,† Xinxin Zhang,*,† and Yong Gan*,† †

Shanghai Institute of Materia Medica, Chinese Academy of Sciences, Shanghai 201203, China University of Chinese Academy of Sciences, Beijing 100049, China § Shanghai Synchrotron Radiation Facility, Shanghai Institute of Applied Physics, Chinese Academy of Sciences, Shanghai 201204, China ‡

S Supporting Information *

ABSTRACT: RNA interference (RNAi) technology has shown great promise for the treatment of cancer and other genetic disorders. Despite the efforts to increase the target tissue distribution, the safe and effective delivery of siRNA to the diseased cells with sufficient cytosolic transport is another critical factor for successful RNAi clinical application. Here, the constructed lipid-based liquid crystalline nanoparticles, called nano-Transformers, can transform thestructure in the intracellular acidic environment and perform high-efficient siRNA delivery for cancer treatment. The developed nano-Transformers have satisfactory siRNA loading efficiency and low cytotoxicity. Different from the traditional cationic nanocarriers, the endosomal membrane fusion induced by the conformational transition of lipids contributes to the easy dissociation of siRNA from nanocarriers and direct release of free siRNA into cytoplasm. We show that transfection with cyclin-dependent kinase 1 (CDK1)-siRNA-loaded nano-Transformers causes up to 95% reduction of relevant mRNA in vitro and greatly inhibits the tumor growth without causing any immunogenic response in vivo. This work highlights that the lipid-based nano-Transformers may become the next generation of siRNA delivery system with higher efficacy and improved safety profiles. KEYWORDS: Lipids, liquid crystalline nanoparticles, siRNA delivery, endosomal escape, synthetic lethality

T

It should further be noted that most nonviral systems used in academic studies and clinical trials are excessively positivecharged, which is required for effective siRNA condensation but will induce significant systemic toxicity and lead to low cell selectivity.20 The cationic complexes are capable of destabilizing the endosomal membrane or buffering the endosomal vesicle, leading to endosomal swelling and lysis, thus releasing siRNA into the cytoplasma.21,22 Although many cationic siRNA delivery systems have been proved to promote decent transfection in cell culture, researches have reported that the cationic-related toxicities, such as radical-mediated pulmonary toxicity, ATPase decrease, cell autophagy, and immunogenic response, limited the broad application of cationic siRNA delivery systems in clinic.23−26 In recent years, a couple of systems such as pH-sensitive charge-reversal copolymers and cationic lipid-assisted PEG−PLA nanoparticles have been reported to exhibit negative charge in the physiological environment and shift to positive charge in the acidic

he RNA interference (RNAi) therapy has gained the world’s attention and become a promising new class of medicines that can silence disease-causing genes in cancer, autoimmune disease, and genopathy.1−4 To activate the RNAi pathway, small interfering RNA (siRNA) molecules must be delivered across the plasma membrane of the target cells and be incorporated into the RNAi machinery.5 As the plasma membrane is semipermeable, naked siRNA molecules are difficult to diffuse across the cell membranes alone. 6 Consequently, various systems have been explored for siRNA delivery with varying levels of success in specific gene silencing.7−9 Virus have shown high siRNA delivery efficiency both in vitro and in vivo.10,11 The low selectivity and potential safety issues, nevertheless, restrict the clinical application of viral systems.12,13 In contrast, nonviral vectors such as polymers, liposomes, and dendrimers have the clear merit of better safety.14−16 In particular, lipid-based nanocarriers have attracted interests for siRNA delivery due to their high biocompatibility. They are basically composed of natural or bioinspired lipid materials and have been used to silence therapeutic genes in several clinical trials.17−19 © XXXX American Chemical Society

Received: December 27, 2017 Revised: March 4, 2018 Published: March 21, 2018 A

DOI: 10.1021/acs.nanolett.7b05430 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 1. Transformation behavior of the DCLC nano-Transformers. (A) Schematic illustration of the synthesis of DCLC nano-Transformers and the delivery mechanism. (B) SEM and cryo-FESEM images of DCLC nano-Transformers before (pH 7.4) and after (pH 5.0) transformation showed hollow needles and spheres, respectively. SEM scale bar = 500 nm, cryo-FESEM scale bar = 100 nm. (C) Size distribution and (D) zeta potential of DCLC nano-Transformers before (pH 7.4) and after (pH 5.0) transformation (n = 3; mean ± SD). (E) Bright field, fluorescent and merged confocal images of DCLC nano-Transformers before (0 min, pH 7.4) and after (30 and 90 min, pH 5.0) incubating in acidic environment, showing the relation between DC degradation and morphology change. Scale bar = 1 μm. (F) 1H NMR spectrum and (G) small-angle X-ray diffraction patterns of DCLC nano-Transformers before (pH 7.4) and after incubation in the endosomal acidic environment (pH 5.0).

Transformers) exhibited endosomal environment-triggered structural transformation property, which could facilitate endosomal escape of siRNA through high endosomal membrane fusion efficiency and directly release free siRNA molecules into cytosol. Negative to neutral surface charge allowed the nano-Transformers to avoid cellular toxicity and immunological recognition. As a proof of concept, transformability, safety, and other physicochemical properties were characterized first. Cytosolic delivery efficiency along with the underlying mechanism was then investigated in human hepatocellular carcinoma (HCC) cell line, and in vivo studies were finally taken to assess the therapeutic effect and the immunological response. The present study focuses on the design of liquid crystalline nanoparticles with high efficiency and low toxicity for siRNA delivery. In order to accomplish this, we designed a lipid material to constitute the nano-Transformers. This lipid material was obtained by linking the primary amine of dioleostearin-3-amino-1, 2-propanediol (DOA) with citraconic

endosomal environment, decreasing the immunogenicity during systemic circulation.27,28 However, even these methods cannot completely avoid the in vivo toxicity induced by cationic materials after their endosomal escaping process.29 In addition, difficulty in dissociation between siRNA and positive-charged materials hinders siRNA molecules to incorporate into the RNAi machinery, decreasing the overall siRNA transfection efficiency.30 It is significantly important to develop a new generation of siRNA delivery system to improve the gene transfection efficiency and decrease the cationic-related toxicity.31 Some lipid molecule complexes with rational arranged structures (gyroid cubic lipid matrix for example) are reported to possess effective gene silencing. 32,33 Here, we synthesized an amphiphilic negative charged lipid and constructed an intelligent lipid-based liquid crystalline nanoparticle platform to improve the gene transfection efficiency and decrease the cationic-related toxicity. Different from the proton sponge effect, the designed nanoparticles (also called the nanoB

DOI: 10.1021/acs.nanolett.7b05430 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 2. siRNA encapsulation and characterization of the DCLC nano-Transformers. (A) Ternary phase diagram of DC/PBS/oleic oil at 37 °C. (B) Agarose gel electrophoresis assay of the siRNA packaging efficiency in nano-Transformers (Lane 1, Naked-siRNA; Lanes 2−7, volume ratios of DC/PBS/oleic oil = 8/11/1, 10/9/1, 12/7/1, 14/5/1, 16/3/1, and 18/1/1; each well contains 0.25 μg of siRNA and 0.5 μg of DCLC nanoTransformers). (C) Size and zeta potential of siRNA-loaded DCLC nano-Transformers barely changed even when concentration of nanoTransofrmers was diluted to 0.006 mg/mL with serum containing PBS (n = 3; mean ± SD). (D) Electrophoresis strips of naked siRNA and siRNA extracted from nano-Transformers after incubation with fetal bovine serum for several time periods. (E) Cell viability and (F) cellular ATP level of HepG2 cells incubated with different formulations containing various concentrations of scrambled siRNA (n = 3; mean ± SD). Significance levels were shown as *p < 0.05 and ***p < 0.001.

lyzation of DC to DOA. During the following 60 min, these nanofragments assembled into the spherical nanoparticles with more fluorescamine signals associated with the nanoparticles, suggesting the acidic environment triggered morphological transformation. The lipid constituent in DCLC nano-Transformers before and after the acidic environment incubation was further validated by 1H NMR. It has been reported that the citraconic amide is stable in both neutral and basic pH but became unstable in acidic pH.34−36 Thus, DC could degrade back into DOA under this condition. As shown in Figure 1F, two characteristic 1H NMR peaks of DC, including methyl group and olefinic bond of citraconic acid at δ = 2.1 and 6.1 ppm disappeared after incubation. In addition, the frame structure of 3-amino-1,2-propanediol at 2.7−4.1 ppm were observed, which was coordinated to the spectrum of DOA (Figure S2B). It was demonstrated that DC molecules degraded back to DOA after incubation in the acidic environment. According to the Israelachvili theory, DC and DOA molecules were expected to show cylindrical and reversed-conical conformations, respectively.37 The geometrical change of the building blocks could explain the microstructural transformation of DCLC nano-Transformers.38 The crystal lattice changing was then determined using smallangle X-ray scattering (SAXS). With the Bragg peak of DCLC nano-Transformers (Qa = 1.337 nm−1, consistent with the Lβ gel phase) gradually decreasing in the acidic environment, two intense peaks (Qb = 1.461 nm−1 and Qc = 1.698 nm−1) and one broad peak at approximately 2.087 nm−1 then emerged (Figure 1G). Since the amplitude of each peak was proportional to the quantity of relevant repeated structures in that phase, it was reasonable to conclude that the microstructural transformation occurred under this condition. In addition, similar SAXS results were obtained under the pH 6.0 PBS condition (Figure S4),

anhydride through ring-opening reaction and was called DOACA (DC). Detailed synthesis procedures were shown in the Supporting Information and Figure S1. Structures of both DOA and DC were verified by 1H NMR (Figure S2). As schematically illustrated in Figure 1A, DC molecules could self-assemble into DC liquid crystalline (DCLC) nanoparticles due to the amphiphilic property. When assessed using scanning electron microscopy (SEM) and cryo-field emission scanning electron microscopy (cryo-FESEM), DCLC nano-Transformers exhibited uniform needle shape with hollow interior (upper panel in Figure 1B). When incubated in the acidic environment (pH 5.0 PBS) at 37 °C, it was also found that the morphology of DCLC nano-Transformers would change into nanospheres with nodules inside (lower panel in Figure 1B), and the diameter would change from 170.3 ± 10.85 nm to 286.1 ± 14.66 nm (Figure 1C). Moreover, polarized pattern of DCLC nano-Transformers changed from rod-like to spot-like (Figure S3) after incubation. Zeta potential of DCLC nano-Transformers increased from −30.7 ± 2.87 to 6.4 ± 1.80 mV (Figure 1D), which was close to neutral comparing with the commercially available cationic liposome Lipofectamine 2000 (Lipo 2000, 32.7 ± 3.41 mV). These results indicated a transformation process of DCLC nano-Transformers that occurred during incubation. To systematically investigate how the nano-Transformers evolve during transformation, fluorescamine was used to label the amino group of DOA, and the transformation process was monitored using a confocal microscope (Figure 1E). The needle structured crystalline nanoparticles gradually disappeared after incubation in the pH 5.0 PBS for the first 30 min. The colocalization of the fluorescamine (blue) signal with the nanofragments indicated that the shape transformation of DCLC nano-Transformers could be a result of the hydroC

DOI: 10.1021/acs.nanolett.7b05430 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

remained the same (Figure S7). All these results indicated that the anionic DCLC nano-Transformers provided a safer way for siRNA delivery, which might ensure much more reliable experimental data by decreasing the side effects and false positive rate. Cellular uptake is an important step for siRNA delivery.6 We used flow cytometry, laser scanning confocal microscopy, and super-resolution microscopy (SRM) to investigate the endocytosis and intracellular distribution of FAM-siRNA (green) in HepG2 cells. The fluorescent intensities of DCLC nano-Transformer group and Lipo 2000 group were significantly higher than free siRNA group after incubation (Figure 3A,B), indicating the enhanced cellular uptake of siRNA delivered by both nanocarriers. Confocal and SRM images (top panels in Figure 3C) showed the colocalization of siRNA and endosomes after incubation for 30 min, and the siRNA molecules locating inside the endosomes (red fluorescence ring). These results demonstrated that siRNA delivered by both nanocarriers were taken up into cells through the endocytosis pathway. The release of free siRNA molecules into cytosol is another key procedure related to overall siRNA silencing efficiency. Successful cytosolic delivery of siRNA by DCLC nanoTransformers was shown by the separation of the green siRNA signals from red endosomal signals (Figure 3C). The overlap coefficiency of DCLC nano-Transformer group at 120 min was only 0.186 (Figure 3D), lower than that of 0.264 in Lipo 2000 group, indicating that most siRNAs have escaped from the endosomal entrapment. In addition, siRNA must dissociate from its nanocarrier and release into the cytoplasma before they bind to RNA-induced silencing complex.30 It has been found that the condensed siRNA in DCLC nanoTransformers or Lipo 2000 showed bright green spots, while the free siRNA showed diffused fluorescence (Figure S8). The efficient delivery of free form siRNA by DCLC nanoTransformers was further validated by diminishing the green spots and the diffusion of siRNA signal in cytosol at 120 min. In contrast, siRNA molecules condensed by Lipo 2000 were still accumulated in cytosol even though escaped from endosomes (arrows in Figure 3C), and the average number of siRNA clusters was significantly higher than that in the DCLC nanoTransformer group (Figure 3E). This could be explained by the strong electrostatic interaction between positively charged Lipo 2000 and negatively charged siRNA. Instead of using electrostatic interactions, encapsulation of siRNA into DCLC nano-Transformers took advantage of the solubility, leading to an easier dissociation of siRNA from the delivery system during endosomal escape. It was also worth mentioning that almost all siRNA failed to escape from endosomes by 120 min with the presence of the proton pump inhibitor bafilomycin A1 (Figure S9), which hindered the endosomal acidification and DCLC nanoparticles’ transformation. This indicated that the endosomal escape of siRNA delivered by DCLC nano-Transformers was transformation-dependent. Since only fully dissociated siRNA that escaped from endosomes could form the RNAinduced silencing complex and then bind with the downstream mRNA, siRNA delivered by DCLC nano-Transformers was thought to enable more efficient gene silencing. Additional efforts were devoted to characterize the endosomal escape process in further detail by live-cell fluorescence imaging with HepG2 cells. Here, Alexa 647-labeled dextran was added into the culture medium to label endosomes.40,41 Dextran was water-soluble and membrane-impermeable.

indicating that the transformation of DCLC nano-Transformers would occur only if there were enough H+s for DC degradation. Considering that pH 5.0 was often used to represent the endosomal acidic environment, pH 5.0 PBS was then chosen for the following experiments in this study. All these results demonstrated that nano-Transformers with endosomal acidic environment-induced transforming property were successfully synthesized. Particle size of DCLC nano-Transformers was sonicationdependent (Figure S5). To optimize the formulation of DCLC nano-Transformers, various ratios of the material components were then investigated and drug loading efficiency was used as a criterion. The ternary phase diagram of the DC/PBS/oleic oil system was shown in Figure 2A, and the blue area represented the condition of DCLC nano-Transformer growth. The broad area of DCLC nano-Transformer growth demonstrated an easy preparation, which was important for commercialization. As shown in Figure 2B, siRNA encapsulation efficiency was promoted along with the increasing volume ratio of DC. When the ratio of DC raised to 80%, siRNA molecules were all encapsulated (loading yield = 50%). Therefore, we chose the volume ratio of DC/PBS/oleic oil = 16/3/1 to prepare the siRNA-loaded DCLC nano-Transformers. The desirable siRNA encapsulation efficiency of DCLC nano-Transformers was attributable to the internal hydrophilic hollow space formed by amphiphilic lipid DC. It is essential for DCLC nanoTransformers to remain stable when diluted into serum, as siRNA is vulnerable to serum nuclease after being systemically administrated.39 Stability of DCLC nano-Transformers during dilution was investigated by measuring the size distribution and zeta potential. As shown in Figure 2C, size and zeta potential of siRNA-loaded DCLC nano-Transformers barely changed. In addition, all siRNA-loaded DCLC nano-Transformers showed similar polarized pattern during dilution (Figure S6). These results ensured the stability of siRNA-loaded nano-Transformers facing dilution, which might be benefited from the highly negative charged hydrophilic surface. Serum stability assay was taken by incubating siRNA-loaded DCLC nanoTransformers with fetal bovine serum for a series of time intervals. As shown in Figure 2D, encapsulation of DCLC nano-Transformers prolonged the time of siRNA against nuclease. Electrophoresis strip of disrupted DCLC nanoTransformers was still detectable at 12 h, but the strip of the naked siRNA molecules disappeared at 10 h. It could be explained by the hollow space inside DCLC nano-Transformers (Figure 1B), which provided shelter for siRNA molecules from degradation. It has been reported that the cationic lipids will induce positive surface charge-dependent cellular toxicity and ATPase activity inhibition.25,26 To investigate the safety of siRNAloaded DCLC nano-Transformers, we measured the cell viability and cellular ATP level using HepG2 cells treated with DCLC nano-Transformers containing scrambled siRNA. Lipo 2000-siRNA nanoparticles were used as positive control. As shown in Figure 2E and 2F, DCLC nano-Transformer group had less influence on both cell viability and cellular ATP level compared with Lipo 2000 group. Statistical difference in cell viability between the DCLC nano-Transformer group and the Lipo 2000 group appeared at a siRNA working concentration of 200 nM, and statistical difference in the cellular ATP level appeared at 50 nM. In addition, siRNA concentrations higher than 200 nM of Lipo 2000 group induced remarkable cellular morphology change, while the DCLC nano-Transformer group D

DOI: 10.1021/acs.nanolett.7b05430 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

labeled, red). and nanoparticles (DiO-labeled, green) were visualized in both merged and individual channels. Early, after internalization, endosomes containing siRNA-loaded DCLC nano-Transformers were visible as bright, globular fluorescent structures. Approximately 10−12 min after monitoring, the fluorescent signal of siRNA quickly disappeared (red arrows) and the integrated fluorescent density showed a significant decrease (Figure 4B,C), while the fluorescent signals of both dextran and DCLC nano-Transformers were still colocalized and visible. These results indicated a quick escape and complete dispersion of siRNA into cytosol without endosomal membrane rupture, which was reasonable for the membrane fusion mechanism. However, for Lipo 2000 group (Figure 4A), the fluorescent signal of dextran quickly disappeared about 15−17 min after monitoring (yellow arrows), and signals of both siRNA and Lipo 2000 were colocalized and visible after escaping from endosomes. Comparing with DCLC nanoTransformer group, the disappearance of dextran signal in Lipo 2000 group (Figure 4D) revealed a leakage of dextran into the cytosol, further indicating the damage of the endosomal membrane. Consistent with the results in Figure 3, more than 50% siRNA delivered by Lipo 2000 remained condensed in the cytosol after escaping from endosomes (Figure 4E), which might decrease the transfection efficiency of siRNA. Moreover, similar to the confocal microscopy results (Figure S9), pretreatment of the cells with bafilomycin A1 also inhibited the release of siRNA when delivered by DCLC nano-Transformers (Figure S10), verifying that the fusion between DCLC nano-Transformers and the endosomal membrane was acidification- and transformation-dependent. Overall, different from the disruptive mode of escape mediated by Lipo 2000, DCLC nano-Transformers achieved endosomal escape of siRNA through highly efficient membrane fusion by taking advantage of the transforming process. To verify the enhanced endosomal escape was a result of promoted membrane fusion by transformation, the mimetic endosomes were prepared to visualize the fusion process. DiOlabeled DCLC nano-Transformers (green) containing scrambled siRNA were loaded into DiI-labeled model endosomes (red, composed of DOPC bilayer and PBS) to imitate the membrane fusion process occurred inside the cell (Figure 4F). Similar with the previous results, transformation of DCLC nano-Transformers (pH 5.0) markedly facilitated the fusion with lipid bilayer, as was evident by the distribution of the green fluorescent signals along with the red signals. For the Lipo 2000 group, though nanoparticles quickly adhered to the membrane through electrostatic interactions, the diffusion of green signals was even slower than DCLC nano-Transformers without transformation (pH 7.4). To directly test the membrane fusion effect, the lysosomes isolated from HepG2 cells and siRNAloaded nanoparticles were labeled with two fluorescence dyes that constituted a Förster resonance energy transfer (FRET) pair. As shown in Figure 4G, it was observed that DCLC nanoTransformers showed a rapid increase of fluorescence intensity at the emission wavelength around 565 nm within 60 min, indicating the excellent fusion of lysosomes with DCLC nanoTransformers. However, the fusion effect was partial and slow in Lipo 2000 group, indicated by the noticeable lower fluorescence intensity at 565 nm. These results indicated that the DCLC nano-Transformers were advantageous in promoting the membrane fusion with endo/lysosomes. Molecular dynamics simulation was used to investigate the underlying mechanisms. As shown in Figure 4H, insertion of

Figure 3. Cellular uptake and intracellular distribution of siRNA molecules. (A) Representative flow cytometry histograms of HepG2 cells treated with PBS, free FAM-siRNA, FAM-siRNA-loaded DCLC nano-Transformers, and FAM-siRNA-loaded Lipo 2000. (B) Mean fluorescence intensities from panel A (n = 3; mean ± SD). (C) Confocal and SRM images of HepG2 cells transfected with FAMsiRNA-loaded DCLC nano-Transformers or Lipo 2000, showing the relative localization of the Alexa 555 labeled endosomes (red) and the FAM-siRNA (green) at incubation times of 30 and 120 min, respectively. Cell nuclei were stained with DAPI (blue). Scale bar = 30 μm. Magnified confocal microscopy images of the regions were shown at the top right corner of each group. Scale bar = 5 μm. SRM images of magnified endosomes were shown at the bottom right corner of each group. Scale bar = 200 nm. (D) Pearson’s coefficient for each formulation was measured in five frames (15−30 cells per frame) that were acquired through random sampling (n = 3 independent experiments; mean ± SD). (E) Average number of green fluorescent dots (siRNA molecules gathered within nanoparticles or endosomes) was quantitated in 30 cells for each formulation at 120 min (n = 3; mean ± SD). Significance levels were shown as n.s., p > 0.05; *p < 0.05; **p < 0.01; and ***p < 0.001.

Thus, fluorescent dextran could be used in measuring the integrity of the endosomal membrane during intracellular tracing, as dextran would leak into the cytosol only if the endosomal membrane was damaged. As shown in Figure 4A and Movies 1 and 2 (Supporting Information), the timedependent appearances of the endosomes (blue), siRNA (Cy3E

DOI: 10.1021/acs.nanolett.7b05430 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 4. Intracellular tracing of the DCLC nano-Transformers and cytosolic delivery mechanism. (A) Intracellular tracing of siRNA molecules, endosomes, and nanocarriers. HepG2 cells were incubated with DCLC nano-Transformers or Lipo 2000 containing siRNA, and the interaction was monitored by live-cell imaging. Representative frames from Movies 1 and 2 were shown. The endosomes were labeled by Alexa 647-dextran (blue). siRNA molecules were labeled by Cy3 (red). Nanocarriers were labeled by DiO (green). Scale bar = 10 μm. (B) Relative integrated densities of siRNA and dextran over time after incubation with DCLC nano-Transformers (Movie 1). (C) Relative integrated densities of siRNA and dextran before and after DCLC nano-Transformers escaping from endosomes were quantitated in five traced particles that were acquired through random sampling (n = 3 independent experiments; mean ± SD). (D) Relative integrated densities of siRNA and dextran over time after incubation with Lipo 2000 (Movie 2). (E) Relative integrated densities of siRNA and dextran before and after Lipo 2000 escaping from endosomes were quantitated in five traced particles that were acquired through random sampling (n = 3 independent experiments; mean ± SD). (F) Fluorescent microscopy images showing the interaction between DiO-labeled nanoparticles (green) and the DiI-labeled model endosomal membrane (red) from 0 to 60 min in acidic endosomal environment (pH 5.0) or proton-inhibited neutral environment (pH 7.4). Scale bar = 5 μm. (G) FRET activity between DiOlabeled nanocarriers and DiI-labeled lysosomes. Fluorescence emission spectra for each formulation at incubation time of 0, 15, 30, and 60 min were shown. Spectra of purified lysosome without nanocarriers were taken as control. Data were representative of three independent experiments. (H) Models established for molecular dynamic simulation. The phosphorus and nitrogen of DOPC were shown as yellow and red spheres, respectively. Hydrophobic tails of DOPC were shown as gray spheres. DC molecules before and after degradation were colored by purple and cyan, respectively. Significance levels were shown as n.s., p > 0.05, and ***p < 0.001.

polar parts of the molecules. Changing of the molecular geometry from cylindrical to reversed-conical disturbed the arrangement of DOPC molecules and influenced the van der Waals interactions, resulting in the conspicuous enhancement of the relative free energy. In this way, conformational transition of DC could lead to a considerable decrease of free energy cost for membrane destabilization, which would further

the DC molecule into the lipid bilayer without conformational transition (pH 7.4) increased the free energy by ΔE1 = 131 kJ mol−1 of the membrane system. Notably, this free energy difference markedly increased to ΔE2 = 412 kJ mol−1 after DC degradation and conformational transition (pH 5.0). Since the total number of atoms did not change during this simulation, the relative energy calculated for each system was only dependent on the van der Waals interactions between the F

DOI: 10.1021/acs.nanolett.7b05430 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters accelerate the membrane fusion process and finally facilitated the release of siRNA molecules into cytosol. After escaping from the endosomal entrapment, free siRNA molecules could combine to downstream mRNA to achieve the RNAi effect. Here, RT-qPCR and Western blot were used to investigate whether siRNA delivered by DCLC nano-Transformers could play an effective role. As shown in Figure 5A, treatment of HepG2 cells with DCLC nano-Transformers containing CDK1-siRNA (siRNA working concentration: 100 nM) showed a 95% reduction on CDK1 mRNA and a corresponding decrease in CDK1 protein expression, which was dramatically lower than Lipo 2000 (approximately 50% reduction on CDK1 mRNA). This remarkable gene silencing ability of DCLC nano-Transformers permitted the application in broader fields compared to Lipo 2000. It has been reported that elevated c-Myc expression rendered HCC cells more proliferative and synthetically lethal with CDK1 inhibition.42,43 Thus, the downregulation of CDK1 protein caused by the nano-Transformers was able to kill HCC cell line HepG2 more effectively and spare normal cells. Cell cycle and apoptosis after treating with nanocarriers containing CDK1-siRNA were then investigated using flow cytometry. Compared with Lipo 2000 group, siRNA delivered by DCLC nano-Transformers obviously caused the arrest of cell cycle at the G2/M-phase (Figure 5B) and enhanced cellular apoptosis ratio (68.6% for nano-Transformers group and 38.1% for Lipo 2000 group, Figure 5C). Moreover, as shown in Figure 5D, inhibition of CDK1 expression reduced viability of the HepG2 cells to 25.8%. As for normal liver cell line L02, silencing of the CDK1 expression did not affect the cell viability (Figure 5E). These results indicated that the CDK1-siRNA-loaded DCLC nano-Transformers were safe and effective for HCC treatment in vitro. To obtain further insight into the antitumor activity, tumor tissue RT-qPCR and tumor volume measurement were taken to monitor the in vivo gene silencing and therapeutic efficacy. Balb/c nude mice bearing HepG2 tumors were peritumorally injected with CDK1-siRNA-loaded DCLC nano-Transformers, scrambled siRNA-loaded DCLC nano-Transformers, CDK1siRNA-loaded Lipo 2000, or PBS every other day. As shown in Figure 6A, CDK1 mRNA reduction for the DCLC nanoTransformer group reached up to 79.6% (48 h after administration), which was significantly higher than that for the Lipo 2000 group (10.5% reduction). Moreover, DCLC nano-Transformer group significantly inhibited tumor growth compared to Lipo 2000 and scramble groups (Figure 6B). The percentage of tumor growth inhibition in the DCLC nanoTransformer group was 84.84% compared to PBS control after 21 days of treatment (Figure 6C,D). Hematoxylin and eosin (H&E) staining results also showed that DCLC nanoTransformer treatment led to more severe apoptosis in the tumor when compared with other groups (Figure 6E). To evaluate the cytosolic siRNA delivery efficiency of DCLC nano-Transformers in vivo, frozen tissue sections and immunofluorescent staining were used to analyze the siRNA distribution in tumor tissue. As shown in Figure 6F, siRNA delivered by DCLC nano-Transformers diffused through the cytoplasm while the spot-like structure (condensed siRNA molecules) still existed in the Lipo 2000 group, when comparative siRNA fluorescent signals were observed in tumor tissues of both groups (Figure 6G). It indicated that DCLC nano-Transformers facilitated siRNA dissociation from nanocarriers during endosomal escape process and could

Figure 5. Gene silencing and in vitro therapeutic efficacy of CDK1siRNA-loaded DCLC nano-Transformers. (A) Relative CDK1 mRNA expressions determined by RT-qPCR after treated with PBS, CDK1siRNA-loaded DCLC nano-Transformers, CDK1-siRNA-loaded Lipo 2000, and scrambled siRNA-loaded DCLC nano-Transformers (n = 3; mean ± SD). CDK1 protein expressions were determined by Western blot, and actin was used as the control group. (B) Cell cycle results were acquired after transfection with CDK1-siRNA and PI staining. (C) Percentage of apoptotic cells in each sample was assessed by flow cytometry after transfection with different formulations (n = 3; mean ± SD). (D) Cell viabilities of HepG2 cells were assessed after incubation with different formulations for 36 h (n = 3; mean ± SD). (E) Relative CDK1 mRNA expressions and cell viability of L02 cells after incubation with different formulations (n = 3; mean ± SD). Significance levels were shown as n.s., p > 0.05, **p < 0.01, and *** p < 0.001.

achieve remarkable cytosolic delivery efficiency even in the complicated in vivo environment. These results were consistent with in vitro observations (Figure 3C). It is reported that cationic siRNA carriers may induce high levels of inflammatory cytokines associated with innate immunogenicity.44,45 When applied to the immune-responsive ICR mice, scrambled siRNAG

DOI: 10.1021/acs.nanolett.7b05430 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters

Figure 6. In vivo antitumor activity of DCLC nano-Transformers. (A) CDK1 mRNA expression in tumors after administration of different formulations (CDK1-siRNA-loaded DCLC nano-Transformers, CDK1-siRNA-loaded Lipo 2000, and scrambled siRNA-loaded DCLC nanoTransformers at 20 μg siRNA per injection containing 40 μg of nanocarrier) and PBS for 48 h (n = 5; mean ± SD). (B) Tumor volume curves of mice during treatment with formulations described above (n = 5; mean ± SD). (C) Tumor volumes at day 21 after treatment with formulations described above (n = 5; mean ± SD). (D) Representative photographs showing the appearance of tumor-bearing nude mice and the tumors at the end of treatment. (E) Representative H&E stained histological sections of tumor tissues after treatment. Scale bar = 100 μm. (F) siRNA distribution in tumor tissues. Tumor sections from mice injected with Lipo 2000 or DCLC nano-Transformers containing FAM-siRNA (green) were immunelabeled for endosomes (red). Nuclei were stained by DAPI and shown in gray. Scale bar = 50 μm. (G) Integrated density of green fluorescent signal (siRNA molecules) in tumor tissues for each formulation was measured in five frames (about 200 cells per frame) that were acquired through random sampling (n = 3 independent experiments; mean ± SD). (H) Immune response for nanoparticles. TNF-α and IL-12 (p70) level in mice blood were tested using ELISA kit 24 h after systematic administration with formulations described above (n = 5; mean ± SD). Significance levels were shown as n.s., p > 0.05, *p < 0.05, **p < 0.01, ***p < 0.001, and ****p < 0.0001.

cationic-related toxicity both in vitro and in vivo. The desirable safety along with high gene transfection efficiency provided a broad application potential of DCLC nano-Transformers, even in the clinical treatment of cancer. The work presented here could help to provide principles for the development of nextgeneration, conformational transformable lipid material for efficient siRNA delivery.

or CDK1-siRNA-loaded DCLC nano-Transformers achieved lower immune response than Lipo 2000-siRNA nanoparticles with an equal amount of DCLC nano-Transformers per unit of siRNA to Lipo 2000 (Figure 6H). This result indicated the superior biocompatibility of the DCLC nano-Transformers. Collectively, we can draw a conclusion that both the transformable capacity and nonimmunogenicity attributed to the satisfactory therapeutic results of DCLC nano-Transformers. Achieving the nonimmunogenic and highly efficient siRNA delivery ability, DCLC nano-Transformers were safe and efficient for HCC treatment in vivo. In this work, DCLC nano-Transformers composed of synthesized lipid material DC were established to achieve high siRNA delivery efficiency with low toxicity. The constructed DCLC nano-Transformers could transform the three-dimensional structure at acidic pH, greatly facilitating the endosomal escape of siRNA through highly efficient membrane fusion mechanism. This transformation ability was derived from the specific structural features of DC molecules. Linking the small head of DOA to another bigger headgroup with an endosomal acidic environment-degradable amido bond allowed the molecular conformational transition during degradation. Moreover, DCLC nano-Transformers were anionic in physiological pH and nearly neutral in endosomal pH, avoiding the

■

ASSOCIATED CONTENT

S Supporting Information *

The Supporting Information is available free of charge on the ACS Publications website at DOI: 10.1021/acs.nanolett.7b05430.

■

Experimental procedures and additional figures (PDF) Supporting movies (AVI) (AVI)

AUTHOR INFORMATION

Corresponding Authors

*E-mail: [email protected]. *E-mail: [email protected]. ORCID

Yong Gan: 0000-0002-4579-994X H

DOI: 10.1021/acs.nanolett.7b05430 Nano Lett. XXXX, XXX, XXX−XXX

Letter

Nano Letters Notes

(26) Wei, X. W.; Shao, B.; He, Z. Y.; Ye, T. H.; Luo, M.; Sang, Y. X.; Liang, X.; Wang, W.; Luo, S. T.; Yang, S. Y.; Zhang, S.; Gong, C. Y.; Gou, M. L.; Deng, H. X.; Zhao, Y. L.; Yang, H. S.; Deng, S. Y.; Zhao, C. J.; Yang, L.; Qian, Z. Y.; Li, J.; Sun, X.; Han, J. H.; Jiang, C. Y.; Wu, M.; Zhang, Z. R. Cell Res. 2015, 25 (2), 237−253. (27) Yang, X.; Grailer, J. J.; Rowland, I. J.; Javadi, A.; Hurley, S. A.; Matson, V. Z.; Steeber, D. A.; Gong, S. ACS Nano 2010, 4 (11), 6805−17. (28) Chen, W.; Meng, F.; Cheng, R.; Zhong, Z. J. Controlled Release 2010, 142 (1), 40−6. (29) Lonez, C.; Vandenbranden, M.; Ruysschaert, J. M. Prog. Lipid Res. 2008, 47 (5), 340−7. (30) Kwon, Y. J. Acc. Chem. Res. 2012, 45 (7), 1077−1088. (31) Rietwyk, S.; Peer, D. ACS Nano 2017, 11 (8), 7572−7586. (32) Leal, C.; Bouxsein, N. F.; Ewert, K. K.; Safinya, C. R. J. Am. Chem. Soc. 2010, 132 (47), 16841−7. (33) Kim, H.; Leal, C. ACS Nano 2015, 9 (10), 10214−26. (34) Li, Z.; Dong, K.; Huang, S.; Ju, E.; Liu, Z.; Yin, M.; Ren, J.; Qu, X. Adv. Funct. Mater. 2014, 24 (23), 3612−3620. (35) Lee, Y.; Fukushima, S.; Bae, Y.; Hiki, S.; Ishii, T.; Kataoka, K. J. Am. Chem. Soc. 2007, 129 (17), 5362−3. (36) Guo, S.; Huang, Y.; Jiang, Q.; Sun, Y.; Deng, L.; Liang, Z.; Du, Q.; Xing, J.; Zhao, Y.; Wang, P. C.; Dong, A.; Liang, X. J. ACS Nano 2010, 4 (9), 5505−11. (37) Israelachvili, J. N.; Mitchell, D. J.; Ninham, B. W. Biochim. Biophys. Acta, Biomembr. 1977, 470 (2), 185−201. (38) Goodby, J. W.; Saez, I. M.; Cowling, S. J.; Gortz, V.; Draper, M.; Hall, A. W.; Sia, S.; Cosquer, G.; Lee, S. E.; Raynes, E. P. Angew. Chem., Int. Ed. 2008, 47 (15), 2754−87. (39) Bumcrot, D.; Manoharan, M.; Koteliansky, V.; Sah, D. W. Nat. Chem. Biol. 2006, 2 (12), 711−9. (40) Zanoni, I.; Ostuni, R.; Marek, L. R.; Barresi, S.; Barbalat, R.; Barton, G. M.; Granucci, F.; Kagan, J. C. Cell 2011, 147 (4), 868−80. (41) Mellman, I.; Fuchs, R.; Helenius, A. Annu. Rev. Biochem. 1986, 55, 663−700. (42) Goga, A.; Yang, D.; Tward, A. D.; Morgan, D. O.; Bishop, J. M. Nat. Med. 2007, 13 (7), 820−827. (43) Horiuchi, D.; Kusdra, L.; Huskey, N. E.; Chandriani, S.; Lenburg, M. E.; Gonzalez-Angulo, A. M.; Creasman, K. J.; Bazarov, A. V.; Smyth, J. W.; Davis, S. E.; Yaswen, P.; Mills, G. B.; Esserman, L. J.; Goga, A. J. Exp. Med. 2012, 209 (4), 679−96. (44) Xue, H. Y.; Liu, S.; Wong, H. L. Nanomedicine (London, U. K.) 2014, 9 (2), 295−312. (45) Landesman-Milo, D.; Peer, D. Bioconjugate Chem. 2016, 27 (4), 855−62.

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS This work was supported by the National Natural Science Foundation of China (grant number: 81571796 and 61433017), Youth Innovation Promotion Association of the Chinese Academy of Sciences (grant number: 2015229), and the SA-SIBS Scholarship Program. We gratefully acknowledge the Shanghai Synchrotron Radiation Facility (SSRF) and the National Center for Protein Science Shanghai (NCPSS) for providing the facility of the beamline BL16B1 and spin-disk confocal microscope, respectively. We also appreciate Yuxiang Wang, Le Wang, Shuang Guo, Ran An, and Yanli Yao for helpful discussions with the paper.

■

REFERENCES

(1) Swamy, M. N.; Wu, H.; Shankar, P. Adv. Drug Delivery Rev. 2016, 103, 174−86. (2) Stokman, G.; Qin, Y.; Racz, Z.; Hamar, P.; Price, L. S. Adv. Drug Delivery Rev. 2010, 62 (14), 1378−89. (3) Wu, S. Y.; Lopez-Berestein, G.; Calin, G. A.; Sood, A. K. Sci. Transl. Med. 2014, 6 (240), 240ps7. (4) Lares, M. R.; Rossi, J. J.; Ouellet, D. L. Trends Biotechnol. 2010, 28 (11), 570−9. (5) Filipowicz, W. Cell 2005, 122 (1), 17−20. (6) Whitehead, K. A.; Langer, R.; Anderson, D. G. Nat. Rev. Drug Discovery 2009, 8 (2), 129−38. (7) Kanasty, R.; Dorkin, J. R.; Vegas, A.; Anderson, D. Nat. Mater. 2013, 12 (11), 967−77. (8) Kesharwani, P.; Gajbhiye, V.; Jain, N. K. Biomaterials 2012, 33 (29), 7138−50. (9) Wong, J. K. L.; Mohseni, R.; Hamidieh, A. A.; MacLaren, R. E.; Habib, N.; Seifalian, A. M. Trends Biotechnol. 2017, 35 (5), 434−451. (10) Yoo, J. W.; Irvine, D. J.; Discher, D. E.; Mitragotri, S. Nat. Rev. Drug Discovery 2011, 10 (7), 521−35. (11) Kotterman, M. A.; Chalberg, T. W.; Schaffer, D. V. Annu. Rev. Biomed. Eng. 2015, 17, 63−89. (12) Thomas, C. E.; Ehrhardt, A.; Kay, M. A. Nat. Rev. Genet. 2003, 4 (5), 346−358. (13) Waehler, R.; Russell, S. J.; Curiel, D. T. Nat. Rev. Genet. 2007, 8 (8), 573−87. (14) Pack, D. W.; Hoffman, A. S.; Pun, S.; Stayton, P. S. Nat. Rev. Drug Discovery 2005, 4 (7), 581−93. (15) Mintzer, M. A.; Simanek, E. E. Chem. Rev. 2009, 109 (2), 259− 302. (16) Yin, H.; Kanasty, R. L.; Eltoukhy, A. A.; Vegas, A. J.; Dorkin, J. R.; Anderson, D. G. Nat. Rev. Genet. 2014, 15 (8), 541−55. (17) Ozcan, G.; Ozpolat, B.; Coleman, R. L.; Sood, A. K.; LopezBerestein, G. Adv. Drug Delivery Rev. 2015, 87, 108−19. (18) Allen, T. M.; Cullis, P. R. Adv. Drug Delivery Rev. 2013, 65 (1), 36−48. (19) Ozpolat, B.; Sood, A. K.; Lopez-Berestein, G. Adv. Drug Delivery Rev. 2014, 66, 110−6. (20) Verdurmen, W. P.; Brock, R. Trends Pharmacol. Sci. 2011, 32 (2), 116−24. (21) Cheng, X.; Lee, R. J. Adv. Drug Delivery Rev. 2016, 99 (Pt A), 129−37. (22) Khalil, I. A.; Kogure, K.; Akita, H.; Harashima, H. Pharmacol Rev. 2006, 58 (1), 32−45. (23) Tseng, Y. C.; Mozumdar, S.; Huang, L. Adv. Drug Delivery Rev. 2009, 61 (9), 721−31. (24) Wan, C.; Allen, T. M.; Cullis, P. R. Drug Delivery Transl. Res. 2014, 4 (1), 74−83. (25) Lv, H.; Zhang, S.; Wang, B.; Cui, S.; Yan, J. J. Controlled Release 2006, 114 (1), 100−9. I

DOI: 10.1021/acs.nanolett.7b05430 Nano Lett. XXXX, XXX, XXX−XXX