Novel cationic lipid with reduction-responsive cleavable hydrophobic tail for siRNA delivery        
Yi Yan1, Shihe Cui1, Jing Sun1, Piaopiao Li1, Haitao Zhang1,2, Jiancheng Wang1*     
1. Beijing Key Laboratory of Molecular Pharmaceutics and New Drug Delivery Systems, School of Pharmaceutical Sciences, Peking University Health Science Center, Beijing 100191, China 
2.School of Pharmaceutical Sciences, Central South University, Changsha 410013, Hunan, China 
Abstract: To achieve a higher transfection efficiency and lower toxicity, a novel herringbone-like cationic lipid (2ssHLL) composed of hydrophilic aspartic acid linked with two reduction-responsive cleavable hydrophobic oleic acid tails was synthesized and assessed in this study. In our results, the cationic nanoplexes with a uniform spherical shape and a particle size of ~150 nm were successfully prepared by the electrostatic interaction between siRNAs and 2ssHLL-based liposomes. From the results evaluated in HepG2 cells, it was shown that the nanoplexes exhibited high cellular uptake of siRNA with a low cytotoxicity. Moreover, the significant down-regulation effects of 2ssHLL/siEGFR nanoplexes on target mRNA were displayed by RT-PCR analysis, which were similar to those of Lipofectamine2000. It suggested that the enhanced siRNA gene silencing efficiency was probably attributed to the detachment of hydrophobic tail chains induced by reduction-responsive cleavage. This mechanism was also confirmed by the changes of size distribution and siRNA release of nanoplexes in the reductive environment and DTT-absence condition. Overall, we believed that the redox-active herringbone-like 2ssHLL would be a potential nanocarrier towards siRNA delivery.                      
Keywords: siRNA delivery; Disulfide bond; Reduction-responsive; Nanoplexes; Cleavage 
CLC number: R943                Document code: A                 Article ID: 10031057(2018)638314
1. Introduction
Small interference RNA (siRNA), a double-stranded RNA with 21 to 25 nucleotides, is a representative nucleic acid therapeutic for RNA interference and has been extensively explored for efficacious cancer therapy[1,2].However, the deficiencies of high hydrophilicity, negativecharge and nuclease degradation limit the clinical application of siRNAs[35]. In the past decades, many non-viral vectors have been shown as excellent potential delivery systems for siRNA applications[69], especiallycationic liposomes composed of different lipids, such as DOTMA, 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP), 1,2-dilinoleyloxy-3-dimethylaminopropane (DLinDMA) and so on[1012]. These cationic liposomes,with a stable bilayer membrane self-assembled by hydrophobic tail chains[6,13], can effectively complex with negatively charged nucleic acids into lipoplexes by electrostatic interaction[7,10]. However, the clinical applications of siRNAs are now still restricted with the systemic toxicities related to the residue positive-charge of lipoplexes[14]. Up to now, some other alternative safe vectors are still continuing to be explored[8,12].
In recent years, some endogenous natural amino acids, including lysine (Lys) and arginine (Arg), have been used as head moieties to improve the toxicity of cationic lipids. For instance, Yosuke Obata et al.[15]have developed different cationic lipids bearing lysine, histidine or arginine as cationic headgroup and demonstrated that their cytotoxicities were only one-fifth that of Lipofectamine2000[16]. Yi Zheng et al. have developed a gemini-like cationic lipid with two lysine headgroups, showing efficient gene silencing with low toxicity. Kim et al. have prepared a series of cationic lipids containing lysine headgroup and aspartate backbone for reducing cytotoxicity[18]. These results imply that the use of amino acid as headgroup of lipids can achieve good biosafety to a certain degree.
Moreover, various pH- or redox-responsive cleavable functional linkers are introduced in the structures of cationic lipids for promoting disassembly of nanostructures and improving transfection efficiency of siRNAs. One strategy that a cleavable linker bridged between two amphiphilic fragments can benefit to the disassembly of gemini-like lipids. In our previous study[19], an H-shapedcationic lipid composed of two hydrophilic lysine heads and two hydrophobic oleyl alcohol tails linked with disulfide (-S-S-) bridge was used for siRNA delivery, exhibiting higher gene silencing based on the increased release of siRNA from the lysosome after cleavage of redox-active disulfide-bridge. Another strategy that cleavable linkers incorporated between hydrophilic heads and hydrophobic tails of cationic lipids can benefit to the disassembly of herringbone-like lipids. For example, Sean C Semple et al. have proved that the introduction of a pH-responsive ketal ring linker between dimethylamino headgroups and alkyl chains in LNPs exhibited ~2.5-foldmore potent effects on reducing protein levels relative to the DLinDMA benchmark[10]. These results demonstratethat the disassembly of cationic lipid-based nanoparticles can play a crucial role in the enhancement of siRNA transfection.
To achieve a higher transfection efficiency of siRNA and lower toxicity of cationic lipids, a novel herringbone-like cationic lipid (2ssHLL) composed of one endogenous aspartic acid and two reduction-responsive cleavable hydrophobic tails was synthesized, and the resulting cationicliposomes could efficiently complex with siRNA into nanoplexes. After cell internalization, the nanoplexes would promote the cleavage of hydrophobictails and release of siRNA in cytosol based on the significantdifferent glutathione concentrations (102–103 fold) between intra- and extra-cellular milieu, and thus improve siRNA transfection (Scheme 1).

Scheme 1.
Schematic illustration of the reduction-responsive release of siRNA from 2ssHLL/siRNA nanoplexes. 
2. Experimental sections
2.1.Materials and cell lines
4-Dimethyl aminopyridine (DMAP) and N,N-dimethyl-carbonimide (DCC) were obtained from J&K ScientificLtd. (Beijing, China). L-Asp, oleic acid (OA), cystamine dihydrochloride (CA?2HCl), triethylamine(Et3N) and Boc2O were purchased from GL Biochem Co., Ltd. (Shanghai, China). Lipofectamine2000 and OPTI-MEM were supplied by Invitrogen (NY, USA). Agarose was obtained from GENE Company (Hong Kong, China),and(3-(4,5-dimethylthiazolyl-2)-2,5-diphenyltetrazolium bromide) (MTT) was purchased from Sigma Aldrich Co. (St. Louis, MO, USA).Anti-EGFR siRNA (sense strand: 5′-AGGAAUUAAGAGAAGCAACAUdTdT-3′;antisense strand: 5′-AUGUUGCUUCUCUUAAUUCC-UdTdT-3′, named as siEGFR), negative control siRNA (sense strand: 5′-UUCUCCGAACGUGUCACGUTT-3′;antisense strand: 5′-ACGUGACACGUUCGGAGAATT-3′. named as siNC) and fluorescein-labeled siRNA (5′ end of the sense strand, FAM-siRNA) were synthesized and purified with HPLC by Gene Pharma Co., Ltd. (Shanghai, China).
The HepG2 (human hepatocellular carcinoma cell) cell line was obtained from the Institute of Basic Medical Science, Chinese Academy of Medical Sciences (Beijing, China). The cells were cultured in DMEM medium (Macgene, Beijing, China) supplemented with 10% foetalbovine serum (FBS), 100 units/mL penicillin and 100 mg/mL streptomycin at 37 ºC in a humidified atmosphere containing 5% CO2. The cells for all experiments were in the logarithmic phase of growth.  
2.2. Synthesis of 2ssHLL materials
2.2.1. Synthesis of compound 1
Under stirring condition, the OA (740 mg, 2.21 mmol)and DCC reagent (540 mg, 2.62 mmol) were sequentiallyadded into 15 mL anhydrous dichloromethane to activate the carboxyl groups for 30 min. Then, Cystaminedihydrochloride (CA?2HCl) (1 g, 4.44 mmol), triethylamine(2.5 mL, 18.07 mmol) and DMAP (300 mg, 2.46 mmol) were respectively added into the reaction mixture. After stirring overnight at room temperature, the reaction mixture was filtered, and the organic solvent was removed in vacuum. Residue was purified on a silica column using dichloromethanemethanol (20:1, v/v) to obtain the white powder.
2.2.2. Synthesis of compound 2
L-Asp (1 g, 7.51 mmol) was dissolved in 15 mL saturated NaOH aqueous solution, and then the solutionof Boc2O (1.6 g, 7.51 mmol) dioxane (5 mL) was dropwised into the reaction mixture. After stirring overnight, HCl was added into the reaction under ice bath conditions to adjust the pH ~2. Then the residue was extracted with ethyl acetate, and organic phase was dried over anhydrous Na2SO4 for 4 h. The organic solvent was removed in vacuum to obtain the product.
2.2.3. Synthesis of compound 3
Both DCC reagent (1.4 g, 6.76 mmol) and compound 2 (750 mg, 3.21 mmol) were added into the anhydrous dichlormethane to activate the carboxyl groups for 30 min. And then the intermediate 1 (6.75 mmol) and the DMAP (6.75 mmol) were respectively added to the reactionmixture. After stirring at room temperature for 12 h, the reaction mixture was filtered, and the organic solvent was removed in vacuum. The residue was purified in column chromatography with (dichloromethanemethanol, 30:1, v/v) to obtain the compound 3. 
2.2.4.Synthesis of 2ssHLL material
Compound 3 (200 mg, 0.194 mmol) was dissolved in 20 mL anhydrous dichloromethane under stirring at 0 ºC, and 4 mL TFA was slowly droped into the reactionmixture above. After stirring at room temperature for 3 h, the reaction solvent was removed under vacuum evaporation for 1 h to obtain the final product.
2.3.Preparation and characterization of 2ssHLL/siRNA nanoplexes
2.3.1. Preparation of 2ssHLL/siRNA nanoplexes
The 2ssHLL nanoparticles were prepared by thin film hydration method. Briefly, 2ssHLL materials were dissolved into methanol in a flask with appropriate concentration, and then organic solvent was removed by rotary evaporation method at 37 ºC for 15 min. The formed film was hydrated by 5% glucose solution pretreated with diethy pyrocarbonate (DEPC). Subsequently, it was sonicated at 60 ºC for 30 min. As a result, 2ssHLL nanoparticles were obtained and stored at 4 ºC for use. To form the 2ssHLL/siRNA nanoplexes,the 2ssHLL nanoparticles and siRNA were mixed together with N/P=10/1 or other different N/P ratios, and the nanoplexes were incubated at 37 ºC for 15 min and then diluted to certain concentration for use.
2.3.2.Particle size and zeta potential of nanoplexes
 The particle sizes and zeta potentials of blank 2ssHLL nanoparticles and 2ssHLL/siRNA nanoplexes were determined using dynamic light scattering (Malvern Zetasizer Nano ZS, Malvern, UK) at 25 ºC and at a scattering angle 90º. The final concentration of siRNA was 100 nM.
2.3.3.Morphological observation of nanoplexes
The morphology of nanoplexes was observed by transmission electron microscopy (TEM, JEOL, Japan). Briefly, the 2ssHLL/siRNA samples (the final siRNA concentration was 1 μM) were dropped on a copper grid or a silicon pellet (25 mm2). Then the sample was stained with 10 μL 2% uranyl acetate solution (Sigma-Aldrich) for 2 min and dried under vacuum overnight. After that a JOEL 100CX transmission electron microscope (100 kV) was used to obtain images of the samples.
2.4.Gel retardation assay
Various 2ssHLL/siRNA nanoplexes were prepared at different N/P ratios as described above, the final siRNA concentration was fixed at 1 μM. Then these prepared nanoplexes were tackled with 6× loading buffer and subjected to electrophoresis on 1% agarose gel containing 0.5 mg/mL Gelred(a special luminant dye for siRNA staining). Electrophoresis was performed at 80 mV for 3 min, and subsequently 100 mV for 15 min. These resulting gels were photographed under UV-illumination. Free naked siRNA was used as the control (the final concentration of siRNA was 1 μM).
2.5. In vitro disassembly of nanoplexes triggered by DTT
The disassembly of redox-sensitive nanoplexes in vitro was simulated by adding the reducing agent DTT (Dithiothreitol).
2.5.1. Effect of DTT on particle size variation
The nanoplexes (N/P = 10/1) were prepared by the method described above.After that the nanoplexes was diluted with DTT (the final concentration of DTT and siRNA was 10 mM and 100 nM, respectively) and then incubated in a shaking bed at 37 ºC. The particle size was monitored by dynamic light scattering (Malvern Zetasizer Nano ZS, Malvern, UK) to observe the size change at 0 h, 1 h, 2 h and 4 h. For comparison, the same nanoplexes incubated without DTT were designed as control. 
2.5.2.Effect of DTT on the loading capacity
Various 2ssHLL/siRNA nanoplexes were prepared at different N/P ratios as described above, and the final siRNA concentration was fixed at 1 μM. Then these prepared nanoplexes were incubated with 6× loading buffer containing 10 mM DTT for 2 h and subjected to electrophoresis on 1% agarose gel containing 0.5 mg/mL Gelred (a special luminant dye for siRNA staining) at 80 mV for 3 min, subsequently 100 mV for 15 min. The resulting gels were photographed under UV-illumination. Free naked siRNA was used as the control (the final concentration of siRNA was 1 μM).
2.5.3. Effect of DTT on the siRNA release in vitro
To investigate the release profile of siRNA from nanoplexes in response to the reducing environment, the 2ssHLL/Cy5-siRNA nanoplexes (N/P = 10/1, the final concentration of Cy5-siRNA was 100 nM) were dialyzed (MWCO = 25 KDa) in 40 mL of PBS in the absence or presence of DTT (the final concentration of DTT was 10 mM). The buffer solution was placed in an incubation shaker at 37 ºC at 100 r/min. At desired time intervals, the concentration of the free Cy5-siRNA in the supernatant was measured by Multifunctional Microplate Reader (Tecan Infinite 200 PRO, Switzerland)with a fluorescence detector fixed at 650 nm for excitation and 670 nm for emission. The final formulation was resuspended in PBS as a positive control.
2.6. Cytotoxicity Assay
In vitro cytotoxicity of 2ssHLL and 2ssHLL/siRNC nanoplexes to HepG2 cells was determined by MTT assay. In the case of biomaterial MTT assay, HepG2 cells were seeded in 96-well plates at a density of 6000 cells per well. After 24 h proliferation, cells were treated with 200 μL OPTI-MEM containing serial concentrationsof different nanoplexes. The medium was replaced by DMEM containing 10% FBS after 4 h incubation, followed by another 24 h incubation. After addition of 20 μL MTT (5 mg/mL in PBS) to each well for 4 h, 150 μL DMSO was added to dissolve the formazan crystals on a gyratory shaker for 10 min. Then, absorbancevalues were read on a microplate reader at the wavelength of 490 nm. The cell viability (%) was calculated according to the following formula: cell viability (%) = [OD490 (sample)/OD490 (control)] × 100%, where OD490 (sample) is the absorbance from the cells treated with various nanoparticles, and OD490 (control) is that from the cells treated with OPTI-MEM. All the results were normalized to the untreated group.
2.7. Cellular uptake assay of 2ssHLL/siRNA nanoplexes
HepG2 cells (2×105 per well) were seeded in 6-well plates. After 24 h of proliferation, cells were exposed to various formulations containing FAM-labeled siRNAwith different N/P ratios at the fixed final concentration of 100 nM and incubated for an additional 4 h at 37 ºC. After incubation, the cells were harvested and washed three times with pre-cooled PBS, and intracellular fluorescence intensities were detected by a FACS Calibur flow cytometer (Becton Dickinson, San Jose, CA, USA) immediately with the excitation wavelength of 488 nm and the emission wavelength of 518 nm.
Confocal microscopy was also used to observe the internalization efficacy of 2ssHLL/siRNA nanoplexes. HepG2 cells were seeded into glass-bottom dishes at a density of 2×105 cells per dish. The cells were proliferatedfor 24 h. Then the culture medium was replaced by 1.5 mL fresh OPTI-MEM containing different FAM-siRNA nanoparticles. The final concentration of FAM-siRNA was 100 nM. After transfected for 4 h, the cells were washed three times with PBS, followed by fixation with 4% paraformaldehyde at room temperature for 15 min. Then, F-actin and nucleus were individually stained by rhodamine labeled phallacidin and Hoechst 33342 according to the manufacturer’s instructions. The cells were visualized under a Leica TCS SP8 confocalfluorescence microscope (Leica Microsystems, Heidelberg, Germany).
2.8. In vitro gene silencing effects
The expression of EGFR at the mRNA level in the cells treated with different nanoparticles (N/P = 10/1, final siRNA concentration was 100 nM) was analyzed by RT-PCR. To detect mRNA expression, HepG2 cells (2×105 cells per well) were seeded into 6-well plates. After 24 h proliferation, cells were exposed to various nanoplexes and incubated for an additional 4 h at 37 ºC. After refreshing the cell culture medium, the cells were incubated for another 24 h for proliferation. Total RNA was extracted using the TRIOL® reagent, and then reversely transcribed into cDNA with a GoScript™ Reverse Transcription System (A5001, Promega, USA). PCR was performed using GoTaq®qPCR Master Mix (A6002, Promega, USA) in triplicates, and data collectionwere performed on a real-time PCR amplifier (MX3005P, Stratagene, USA). Glyceraldehyde-phosphate dehydrogenase (GAPDH) was selected as the housekeeping gene. The relative expression level for each target gene was calculated by 2(CtCc) (Ct and Cc were the mean threshold cycle differences after normalizing to GAPDH).
2.9. Statistical analysis
All the results were expressed as means±standard deviation (SD). For comparisons between two groups, unpaired Student’s t-test (two-tailed) was used. For statistical analysis among multiple groups, one-way ANOVA with Tukey test was applied. All statistical analyses were performed by GraphPad Prism Software (Version 6.0, San Diego, CA). P<0.05 was considered as statistically significant.
3. Results and discussion
3.1. Synthesis and characterization of 2ssHLL material and key intermediates
The synthetic procedure of 2ssHLL material was shown in Scheme 2. Briefly, using both cysteine and L-Asp as the starting reagents, the intermediates of 1 and 2 were respectively amidated by OA and protected by Boc2O reagent, and then the amidation between intermediates 1 and 2 was carried out to get thecompound 3. Finally, the 2ssHLL material was obtainedby de-protection under trifluoroacetic acid (TFA) condition. The 2ssHLL and intermediates were confirmed by 1H NMR as follows (Fig. 1).
Scheme 2. Synthetic route of 2ssHLL material. Reagents and conditions: i. DCC, DMAP/DCM, r.t.; ii. Boc2O/Dioxane, NaOH aq, r.t. iii. DCC, DMAP/DCM, r.t. iv. TFA/DCM, r.t.
Figure 1.1H NMR spectra of compound 13 and 2ssHLL material from A to D. 
Compound 1:1H NMR (400 MHz, d6-DMSO) δ: 8.18 (t, 1H,-NH-), 5.45–5.28 (m, 2H, -CH=CH-), 3.52–3.24 (m, 8H,-CH2-CH2-S-S-CH2-CH2-), 2.82 (t, 2H, NH-CH2-), 2.05 (dt, 4H,C*H2-C*H=), 1.72–1.35 (m, 2H, NH2-), 1.23 (t, 22H, CH2 on long chain), 0.89 (t, J 6.7 Hz, 3H,-CH3).Compound 2:1H NMR (400 MHz, DMSO-d6) δ: 12.52 (s, 2H, double chain-COO*H), 7.09 (t, 1H, -NH-CO-), 4.25 (tt, 1H, -CH-COOH), 2.67–2.51 (DD, 2H, -CH2-), 1.38 (s, 9H, Boc-). Compound3:1H NMR (400 MHz, Chloroform-d1) δ: 6.43 (s, 3H), 6.00 (s, 6H), 5.50 (s, 3H), 5.34 (s, 11H), 4.94 (s, 2H), 3.54 (s, 15H), 3.04 (s, 3H), 2.82 (s, 14H), 2.69 (s, 9H), 2.13 (s, 7H), 2.02 (s, 15H), 1.53 (s, 8H), 1.42 (s, 27H), 1.33 (s, 21H), 1.31–1.27 (m, 47H), 1.26 (s, 48H), 0.89 (s, 12H). 2ssHLL material:1H NMR (400 MHz, DMSO-d6)δ: 8.52 (dt, 2H, -NH2), 8.25–7.90 (m, 4H, tetra-N*H-CO-), 5.44–5.26 (m, 4H, double chain-C*H=C*H-), 3.95 (d, 1H, NH2-CH-CO(CH2)), 3.57–3.23 (m, 8H, -NH-C*H2), 2.85–2.60 (m, 10H, -CH-CH2-CO and double chain-C*H2-*S-*S-C*H2-), 2.04 (dt, 12H, double chain-CO-C*H2-CH2- and double chain -CH2-CH=CH-CH2), 1.55–1.44 (m, 4H, double chain-CO-CH2-C*H2), 1.29 (d, J 19.6 Hz, 40H, double chain-on long chains), 0.88 (t, 6H, double chain-CH3).
3.2.Characterization of the 2ssHLL/siRNA nanoplexes
As seen from Table 1 and Figure 2, 2ssHLL material could form the stable nanoparticles with an average size around 140 nm and zeta potential about 40 mV. When they were complexed with siRNA, the particle size was gradually decreased, and the nanoplexes tended to be stable with the increase of 2ssHLL concentration (N/P>5) (Fig. 2A). As shown inFigure 2B, 2ssHLL could effectively complex with siRNA and form the uniform nanoplexes at N/P = 10/1 with a mean size of (150.2±6.1) nm and zeta potential of (20.0±0.3) mv, which was suitable for intravenous administration.The morphology was observed by TEM, showing that the nanoplexes were uniformly spherical in shape without obvious aggregation (Fig. 2C). The size determined by TEM was much smaller than that by DLS measurementbecause dehydration effects occurred during TEM sample preparation.

Table 1. Characteristics of nanoparticles (n = 3).

Figure 2.
Characterization of the prepared nanoparticles. (A) Size and zeta potential of nanoplexes with different N/P ratios were detected by DLS, and the data were shown as means±SD (n = 3).Blank 2ssHLL nanoparticles and 2ssHLL/siRNA (N/P = 10) nanoplexesdetected by DLS (B), and TEM images of 2ssHLL/siRNA nanoplexes (C). 

   The results of loading efficiency for siRNA were confirmed in
Figure 3A–B. It showed thatsiRNA could be completely compacted at N/P ratio as low as 10/1 by gel retardation assay. After nanoplexes were incubated with DTT for 2 h, a clear strip of intact siRNAcould be seen in electrophoresis images, implying that the encapsulated siRNA was structurally stable in nanoplexes. Cytotoxicity assay (Fig. 3C) demonstrated that from N/P=5 to N/P=15, cell viability could still reach at least 80% when the concentration of siRNA was 100 nM. Therefore, N/P=10/1 could be selected for the preparation of nanoplexes in the further experiments.

Figure 3.
Gel retardation assay for the capacity of nanoplex loading siRNA in the absence (A)or presence of DTT (B) at different N/P ratios. (C) Cytotoxicity of various nanoplexes in HepG2 cells. The siRNA concentration was 100 nM. All data were shown as means±SD (n = 3).
3.3. Cellular uptake of2ssHLL/siRNA nanoplexes
High cellular uptake of siRNA is an important step for gene transfection[20]and silencing effects[13]. In this study, FAM-labelled siRNA was used as a fluorescenceindicator to detect the intracellular fluorescence intensityand distribution by flow cytometry (FCM) and confocal laser scan microscope (CLSM). As seen in Figure 4A, after 4 h incubation, 2ssHLL/FAM-siRNA nanoplexes showed much higher fluorescence intensity in HepG2 cells than that in MCF-7 cells and SW1990 cells at almost all N/P ratios. Moreover, when the N/P ratio was up to 10/1 or more, the fluorescence intensity of siRNA was no longer elevated. Therefore, the N/P of 10 was further confirmed to be the suitable preparation ratio of nanoplexes in the following experiments.
Consistently with results of FCM, CLSM images in Figure 4B showed that much more fluorescence was appeared in the cytoplasm of HepG2 cells after 4 h incubation with nanoplexes compared with the naked FAM-siRNA, indicating that these nanoplexes could be efficiently internalized into HepG2 cells. More nanoplexes were attached to the cell membrane than Lipofectamin2000(lipo2000), implying that lipo2000 presented a stronger intracellular transfection efficiency.

Figure 4.
Cellular uptake of 2ssHLL/siRNA nanoplexes with different N/P ratios in HepG2 cells, MCF-7 cells and SW1990 cells. (A)The intracellular intensities were detected by FCM after 4 h of incubation (the final concentration of FAM-siRNA was 100 nM). The data were shown as means±SD (n = 3).**P<0.01, ***P<0.001.(B) CLSM images of 2ssHLL/FAM-siRNA nanoplexes (N/P=10/1) in HepG2 cells. The Rhodamine-labelled phalloidin (green) was used to show the cytoskeleton, and Hoechst 33342 (blue) was used to stain nucleus. 

3.4. Cytotoxicity in HepG2 cells
The cytotoxicities of 2ssHLL nanoparticles and 2ssHLL/siRNA nanoplexes were evaluated in HepG2 cells via MTT assay. As shown in Figure 5A, average 80% viability could be observed in the cells treated with blank 2ssHLL nanoparticles at a series of concentrations. After complexed with scrambled siRNA (siNC), the resulting 2ssHLL/siNC nanoplexes showed nearly 60% cell viability within the siRNA concentration range from 10 nM to 400 nM (Fig. 5B). These results further suggested that the 2ssHLL cationic lipid possessed a lower cytotoxicity and could have a potential use forin vivo treatment. Based on the results, it could be attributed to the use of natural and endogenous aspartic acids[21], which might be involved in the metabolism of ornithine cycle pathway[22,23]. Therefore, the cationic 2ssHLL displayed low toxicity during siRNA delivery in vitro.

Figure 5.
(A) Cytotoxicity of the blank 2ssHLL nanoparticles with a series of concentrations in HepG2 cells. (B)Cytotoxicity of the 2ssHLL/siNC nanoplexes (N/P=10/1) with different concentrations of siRNA in HepG2 cells. The data were shown as means±SD (n = 3).
3.5. In vitro disassembly of nanoplexes triggered by DTT
As well known, the concentration of glutathione (GSH) in intracellular cytosol (310 mM) is maintained approximately three orders of magnitude higher than that in extracellular matrix (~2.8 μM)[24,25]. Moreover, the cytoplasmic environment of tumor cells has a much higher reducing potential (approximately 20 mM GSH) compared with the normal cells[26,27]. It has been suggestedthat the redox gradient can provide a potent stimulative condition for release of siRNA from disulfide-bridge based 2ssHLL/siRNA nanoplexes[19]. In order to study the reduction-triggered cleavage of hydrophobic tails, 10 mM DTT was used to simulate the reducible intracellular environment. As shown in Figure 6, with the expansion of incubation time from 0 h to 4 h, the particle size was significantly increased from ~200 nm to ~2000 nm, and also a sharp increase of siRNA release was observed accordingly. In the contrary, the control group without DTT did not present any change of particle size, and the cumulative release of siRNA was also less than 20%. The similar results were also confirmed by gel retardation assay (Fig. 3B). These outcomes would further validate our assumptions that due to the reductive characteristics of the disulfide bonds, the hydrophilic heads of 2ssHLL were separated from the hydrophobic chains, and then this material would no longer be assembled into tight nanoparticles. Therefore, the positive charge density of the 2ssHLL surface would be reduced accordingly, and not enough to compass or absorb siRNA, resulting in the release of siRNA to the cytoplasm.     

Figure 6.
In vitro disassembly of nanoplexes triggered by DTT. (A) Effect of DTT on particle size variation. (B) Effect of DTT on the Cy5-siRNA release in vitro (N/P=10/1, the final concentration of siRNA was 100 nM). The data were shown as means±SD (n = 3).
3.6. In vitro gene silencing effects
Gene silencing efficiencies of nanoplexes were determined in HepG2 cells with a model siRNA targeting EGFR protein, which has been over-expressedin various cancers and involved in the tumor occurrence, invasion and metastasis mechanisms[28]. The expression of EGFR at the mRNA level was detected by RT-PCR analysis (Fig. 7). It was shown that neither naked siEGFR nor 2ssHLL/siNC had any efficacy. In contrast, 2ssHLL/siEGFR nanoplexes could effectively down-regulate EGFR mRNA by 40% after incubation with cells for 24 h. Meanwhile, there was no significant difference between 2ssHLL/siEGFR nanoplexes and lipo2000/siEGFR (the final concentration of siRNA was 100 nM), even though lipo2000/siEGFR showed a much higher cellular uptake than that of 2ssHLL/siEGFR nanoplexes in HepG2 cells. The mechanism was hypothesized as follows: due to the cleavage of reduction-responsive hydrophobic tails, 2ssHLL/siEGFRnanoplexes exhibited the rapid and complete disassembly of nanoplexes, resulting in the rapid rise of osmotic pressure, which could induce more siEGFR escaping from lysosomes into cytoplasm[29,30]. As a result, the gene silencing effects on target mRNA of siEGFR would be enhanced[31]. In spite of the absent comparable data of uncleavable materials, the present transfection efficiency similar to lipofectamine2000 suggested that the 2ssHLL-based cationic liposomes could be a promising nanocarrier for siRNA delivery. 

Figure 7.
Relative expression of EGFR at the mRNA level in HepG2 cells treated with the 2ssHLL/siRNA nanoplex-loading siRNA against EGFR or NC by RT-PCR. The data were shown as means±SD (n = 3) and **P<0.01. 
4. Conclusions
In summary, we successfully synthesized a herringbone-like cationic lipid named 2ssHLL with reduction-responsive cleavable hydrophobic tails. From the changesof particle size and siRNA release with or without DTT-reductive condition, it was shown that the detachment of hydrophobic tail chains was induced by disulfide bonds cleavage, leading to the rapid disassembly of nanoplexes and enhanced release of siRNA, thus exhibiting strong gene silencing effects. Meanwhile, the utilization of biocompatible aspartic acid heads made the 2ssHLL/siRNA nanoplexes also possess a good biosafety. Therefore, 2ssHLL would be a potential cationic lipid for siRNA delivery.
The authors acknowledge financial support from the projects of National Natural Science Foundationof China (Grant No. 81473158, 81690264 and 81773650), the New Drug R&D program of China (Grant No. 2018ZX09721003-004) and the Opening Project of Key Laboratory of Drug Targeting and Drug Delivery System, Ministry of Education (Sichuan University).
[1] Castanotto, D.; Rossi, J.J. The promises and pitfalls of RNA-interference-based therapeutics. Nature.2009, 457, 426433.
[2] Meister, G.; Tuschl, T. Mechanisms of gene silencing by double-stranded RNA. Nature. 2004, 431, 343349.
[3] Dowdy, S.F. Overcoming cellular barriers for RNA therapeutics. Nat. Biotechnol. 2017, 3, 222229.
[4] Ozcan, G.; Ozpolat, B.; Coleman, R.L.; Anil, S.; Berestein,G. Preclinical and clinical development of siRNA-based therapeutics. Adv. Drug Delive. Rev. 2015, 87, 108119.
[5] Zuckerman, J.E.; Davis, M.E. Clinical experiences with systemically administered siRNA-based therapeutics in cancer. Nat. Rev. Drug Discov. 2015, 14, 843856.
[6] Lin, Q.; Chen, J.; Zhang, Z.; Zheng, G. Lipid-based nanoparticles in the systemic delivery of siRNA. Nanomedicine-UK. 2014, 9, 105120.
[7] Ozpolat, B.; Sood, A.K.; Lopez-Berestein, G. Liposomal siRNA nanocarriers for cancer therapy. Adv. Dru. Delive. Rev. 2014, 66, 110116.
[8] TsengY, C.; Mozumdar, S.; Huang, L. Lipid-based systemic delivery of siRNA. Adv. Drug Delive. Rev. 2009, 9, 721731.
[9] Majzoub, R.N.; Ewert, K.K.; Safinya, C.R. Cationic liposome-nucleic acid nanoparticle assemblies with applications in gene delivery and gene silencing. Philos. Transact. Ser. A Math. Phys. Eng. Sci.2016, 2072, 114.
[10] Semple, S.C.; Akinc, A.; Chen, J.; Sandhu, A.P.; Mui, B.L.; Cho, C.K.; Sah, D.W.; Stebbing, D.; Crosley, E.J.; Yaworski, E.D.; Hafez, I.M.; Dorkin, J.R.; Qin, J.; Lam, K.; Rajeev, K.G.; Wong, K.F.; Jeffs, L.B.; Nechev, L.; Eisenhardt, M.L.; Jayaraman, M.; Kazem, M.; Maier, M.A.; Srinivasulu, M.; Weinstein, M.J.; Chen, Q.; Alvarez, R.; Barros, S.A.; De, S.; Klimuk, S.K.; Borland, T.; Kosovrasti,V.; Cantley, W.L.; Tam, Y.K.; Manoharan, M.;Ciufolini, M.A.; Tracy, M.A.; Fougerolles, A.; MacLachlan, I.; Cullis, P.R.; Madden, T.D.; Hope, M.J. Rational design of cationic lipids for siRNA delivery. Nat. Biotechnol. 2010, 28, 172178.
[11] Jayaraman, M.; Ansell, S.M.; Mui, B.L.; Tam, Y.K.; Chen, J.X.; Du, X.J.; Butler, D.B.; Eltepu, L.; Matsuda, S.; Narayanannair, J.K.; Rajeev, K.G.; Hafez, I.M.; Akinc, A.; Maier, M.A.;Tracy, M.A.; Cullis, P.R.; Madden, T.D.; Manoharan, M.; Hope, M.J. Maximizing the Potency of siRNA Lipid Nanoparticles for Hepatic Gene Silencing in vivo. Angew. Chem. Int. Ed. Engl. 2012, 1, 85298533.
[12] Zhang, S.; Zhao, B.; Jiang, H.; Wang, B.; Ma, B.C. Cationic lipids and polymers mediated vectors for delivery of siRNA. J. Controlled Release. 2007, 123, 110. 
[13] Rehman, Z.U.; Zuhorn, I.S.; Hoekstra, D. How cationic lipids transfer nucleic acids into cells and across cellular membranes: Recent advances. J. Controlled Release. 2013, 166, 4656.
[14] Lv, H.; Zhang, S.; Wang, B.; Cui, S.H.; Yan, J. Toxicity of cationic lipids and cationic polymers in gene delivery.J. Controlled Release. 2006, 114, 100109.
[15] Obata, Y.; Tajima, S.; Takeoka, S. Evaluation of pH-responsiveliposomes containing amino acid-based zwitterionic lipids for improving intracellular drug deliveryin vitro and in vivo. J. Controlled Release. 2010, 142, 267276.
[16] Obata, Y.; Suzuki, D.; Takeoka, S. Evaluation of Cationic Assemblies Constructed with Amino Acid Based Lipids for Plasmid DNA Delivery. Bioconjug. Chem. 2008, 19, 10551063.
[17] Zheng, Y.; Guo, Y.G.; Li, Y.T.; Wu, Y.; Zhang, L.H.; Yang, Z.J. A novel gemini-like cationic lipid for the efficient delivery of siRNA. New J. Chem. 2014, 38, 49524962.
[18] Wang, B.; Yi, W.J.; Zhang, J.; Zhang, Q.F.; Xun, M.M.; Yu, X.Q. TACN-based cationic lipids with amino acid backbone and double tails: materials for non-viral gene delivery. Bioorg Med. Chem. Lett. 2014, 24, 17711775.
[19] Ma, X.F.; Sun, J.; Qiu, C.; Wu, Y.F.; Yu, M.Z.; Pei, X.W.; Wei, L.; Niu, Y.J.; Pang, W.H.; Yang, Z.J.; Wang, J.C.; Zhang, Q. The role of disulfide-bridge on the activities of H-shape gemini-like cationic lipid based siRNA delivery.J. Controlled Release. 2016, 235, 99111.
[20] Breunig, M.; Hozsa, C.; Lungwitz, U.; Kazuo, W.; Umeda, I.; Kato, H.; Goepferich, A. Mechanistic investigation of poly(ethylene imine)-based siRNA delivery: Disulfide bonds boost intracellular release of the cargo. J. Controlled Release. 2008, 130, 5763.
[21] Darpolor, M.M.; Basu, S.S.; Worth, A.; Nelson, D.; Clarke-Katzengerg, R.H.; Glickson, J.D.; Kaplan, D.E.; Blair, L.A. The aspartate metabolism pathway is differentiable in human hepatocellular carcinoma: transcriptomics and 13C-isotope based metabolomics. NMR. Biomed. 2014, 27, 381389.
[22] Sivashanmugam, M.; Umashankar, J.V.; Sulochana, K.N. Ornithine and its role in metabolic diseases: An appraisal.Biomed. Pharmacother. 2017, 86, 185194.
[23] Hiraishi, T. Poly(aspartic acid) (PAA) hydrolases and PAA biodegradation: current knowledge and impact on applications. Appl. Microbiol. Biotechnol. 2016, 100, 16231630.
[24] Son, S.; Ran, N.; Kim, J.; Singha, K.; Kim, W.J. Bioreducible Polymers for Gene Silencing and Delivery. Accounts. Chem. Res. 2012, 45, 11001122.
[25] Saito, G.; Swanson, J.A.; Lee, K.D. Drug delivery strategy utilizing conjugation via reversible disulfide linkages: role and site of cellular reducing activities. Adv. Drug Delive. Rev. 2003, 55, 199215.
[26] Yin, T.; Wang, L.; Yin, L.; Jianpin, Z.; Meirong, H. Co-delivery of hydrophobic paclitaxel and hydrophilic AURKA specific siRNA by redox-sensitive micelles for effective treatment of breast cancer. Biomaterials. 2015, 61, 1025.
[27] Schafer, F.Q.; Buettner, G.R. Redox environment of the cell as viewed through the redox state of the glutathione disulfide/glutathione couple. Free Radical Biology & Medicine. 2001, 30, 11911210.
[28] Nicholson, R.I.; Gee, J.M.; Harper, M.E. EGFR and cancer prognosis. Eur. J. Cancer. 2001, 37, S9S15.
[29] Varkouhi, A.K.; Scholte, M.; Storm, G.; Haisma, H. Endosomal escape pathways for delivery of biologicals. J. Controlled Release. 2011, 151, 220228. 
[30] Wattiaux, R.; Laurent, N.; Wattiaux-De, C.S.; Jadot, M.Endosomes, lysosomes: their implication in gene transfer. Adv. Drug Delive. Rev. 2000, 41, 201208.
[31] Juliano, R.L.; Carver, K. Cellular uptake and intracellular trafficking of oligonucleotides. Adv. Drug Delive. Rev. 2015, 87, 3545.
闫仪1, 崔仕贺1, 孙晶1, 李飘飘1, 张海涛1,2, 王坚成1*
1. 北京大学 医学部 药学院 分子药剂学与新释药系统北京市重点实验室, 北京 100191
2. 中南大学药学院, 湖南 长沙 410013  
摘要: 为了获得更高的转染效率和更低的毒性, 本研究合成了一种新型的人字形阳离子脂质(2ssHLL), 它由亲水性天冬氨酸和两条还原敏感可断裂的疏水油酸尾链组成。通过siRNA和基于2ssHLL的脂质体间的静电相互作用, 成功制备了粒径约150 nm的均匀的球形阳离子纳米复合物。从评估结果可以看出, 该纳米复合物在HepG2细胞试验中表现出较好的细胞摄取和较低的细胞毒性。RT-PCR分析结果表明, 2ssHLL/siEGFR纳米复合物可表现出与Lipofectamine2000相类似的对靶mRNA显著的下调效应。这些增强的siRNA基因沉默效率可能归因于还原响应性断裂所诱导的疏水尾链的分离。在还原环境中观察到的纳米复合物的粒径和siRNA释放的变化也证实了上述的机制。由此, 我们认为氧化还原响应型2ssHLL将有可能成为siRNA递送的潜在纳米载体。 
关键词: 小干扰RNA的递送; 二硫键; 还原敏感; 纳米复合物; 断裂
Received: 2018-05-06, Revised: 2018-05-28, Accepted: 2018-06-03.
Foundation items: National Natural Science Foundation of China (Grant No. 81473158, 81690264 and 81773650), the New Drug R&D program of China (Grant No. 2018ZX09721003-004) and the Opening Project of Key Laboratory of Drug Targeting and Drug Delivery System, Ministry of Education (Sichuan University).
*Corresponding author. Tel.: +86-010-82805932, E-mail:        
本作品采用知识共享署名-非商业性使用 4.0 国际许可协议进行许可。