International Journal of Pharmaceutics 469 (2014) 206–213
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
Highly efficient delivery of siRNA to a heart transplant model by a novel cell penetrating peptide-dsRNA binding domain Hua Li a,b , Xiangtao Zheng a , Viktoria Koren a , Yogesh Kumar Vashist a , Tung Yu Tsui a, * a b
The Department of General, Visceral and Thoracic Surgery, University Medical Center Hamburg-Eppendorf, Hamburg, Germany Institute for Regenerative Medicine and Cancer, Huzhou University School of Medicine, Huzhou, China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 20 March 2014 Received in revised form 18 April 2014 Accepted 19 April 2014 Available online 23 April 2014
Small interfering RNAs (siRNAs) delivery remains a bottleneck for RNA interference (RNAi) – based therapies in the clinic. In the present study, a fusion protein with two cell-penetrating peptides (CPP), Hph1–Hph1, and a double-stranded RNA binding domain (dsRBD), was constructed for the siRNA delivery: dsRBD was designed to bind siRNA, and CPP would subsequently transport the dsRBD/siRNA complex into cells. We assessed the efficiency of the fusion protein, Hph1–Hph1–dsRBD, as a siRNA carrier. Calcium-condensed effects were assessed on GAPDH and green fluorescent protein (GFP) genes by western blot, real time polymerase chain reaction (RT-PCR), and flow cytometry analysis in vitro. Evaluations were also made in an in vivo heart transplantation model. The results demonstrated that the fusion protein, Hph1–Hph1–dsRBD, is highly efficient at delivering siRNA in vitro, and exhibits efficiency on GAPDH and GFP genes similar to or greater than lipofectamine. Interestingly, the calcium-condensed effects dramatically enhanced cellular uptake of the protein–siRNA complex. In vivo, Hph1–Hph1–dsRBD transferred and distributed ^ targeted siRNA throughout the whole mouse heart graft. Together, these results indicate that Hph1–Hph1–dsRBD has potential as an siRNA carrier for applications in the clinic or in biomedical research. ã 2014 Elsevier B.V. All rights reserved.
Keywords: Cell penetrating peptide RNA interference Gene delivery Double-stranded RNA binding domain
1. Introduction Nowadays, the development of RNA interference (RNAi) as an innovative biomedical tool provides new perspectives for pharmaceutical and medical research. However, until recently, small interfering RNA (siRNA) delivery constituted the greatest challenge for siRNA application in the clinic due to its high negative charge, large molecular weight, and hydrophility. Cell penetrating peptide (CPP) composed of 4–30 amino acid residues with arginine or lysine rich peptide (Lindgren et al., 2000), have been shown to transfer various substances, including siRNA, into different cell types (Eguchi et al., 2009). It was recently reported that CPP transfers siRNA through 2 strategies: covalent and noncovalent associations. During delivery of siRNA into target cells, the negative charge of siRNA may neutralize the positive charge of CPP, resulting in inhibition of siRNA delivery efficiciency. To overcome these issues, we have previously showed the CPP type and the number of repetitive CPP at the N terminal of fusion protein
of CPP-dsRBD (double-stranded RNA (dsRNA) binding domain (dsRBD)) play the key role of siRNA delivery and found the CPP of Hph1 was more efficacy than the CPP of TAT in siRNA delivery (Li and Tsui, 2014). In order to get a high efficient siRNA carrier, here we further provided a fusion protein consisting of 2 CPPs (Hph1 (YARVRRRGPRR)), which could provide efficient transfer into a primary cell (Choi et al., 2008, 2006), and one dsRBD for efficient siRNA delivery. The dsRBD was designed to bind siRNA, and the CPP would subsequently transport the dsRBD/siRNA into cells. The efficiency of Hph1–Hph1–dsRBD as a siRNA carrier for siRNA delivery was assessed in vitro, and the effectiveness of RNAi was determined. Moreover, the affect of the calcium-condensed step in siRNA delivery of Hph1–Hph1–dsRBD was also studied. Finally, the efficacy of CPP-dsRBD in siRNA delivery in vivo was investigated in a mouse heart transplantation model. 2. Material and methods 2.1. Materials
* Corresponding author at: Department of General, Visceral and Thoracic Surgery, University Medical Center Hamburg-Eppendorf, Martinistr. 52, 20246 Hamburg, Germany. Tel.: +49 40 741050822; fax: +49 40 7410 46756. E-mail address: [email protected] (T.Y. Tsui). http://dx.doi.org/10.1016/j.ijpharm.2014.04.050 0378-5173/ ã 2014 Elsevier B.V. All rights reserved.
The genes of TAT-dsRBD and Hph1–Hph1–TAT–TAT–dsRBD (the dsRBD was from double stranded RNA activated protein kinase (Homo sapiens), amino acid numbers 11–78, GenBank: AAF13156.1), were synthesized by Eurofin (Ebersberg, Germany).
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
Rosetta(DE3)pLysS and TOP10 competetent Escherichia coli cells were purchased from Merck (Darmstadt, Germany) and Invitrogen (Darmstadt, Germany), respectively. Ni-NTA column, complete protease inhibitor, and PD-10 columns were purchased from Qiagen (Hilden, Germany), Roche (Mannheim, Germany), and Invitrogen, respectively. Cell culture medium reagents, including DMEM, RIPM 160, FBS, DMSO, penicillin/streptomycin, pyruvate, mercaptoefhenol, and glutamine were obtained from Invitrogen. 18S rRNA and GAPDH primers for real time polymerase chain reaction (RT-PCR), and GAPDH siRNA were ordered from Qiagen. Negative siRNA, FAM-siRNA, and GFP siRNA were purchased from Ambion (Darmstadt, Germany). HeLa cell line was purchased from American Type Culture Collection (ATCC, Manassas, USA), and the GFP stable expressing HeLa cell line (GFP HeLa) was generated by using the pcDNA3.1+ GFP construct (Invitrogen, Steinheim, Germany). All other reagents were obtained from Sigma–Aldrich (Steinheim, Germany) unless otherwise noted.
207
performed equally to the HeLa cell line with the addition of 1.5 mg/ml gentamicin to the medium (Invitrogen). 2.6. Cell uptake assay HeLa cells (100 ml) were preplated in 96-well plates at a density of 1 104 cells/well and incubated for 24 h. Cells were then rinsed twice with serum-free DMEM medium. Supplements of serumfree DMEM medium and the different ratios of Hph1–Hph1– dsRBD/FAM-labeled siRNA complex were added to the cells. The final siRNA concentration was 100 nM. After an incubation period of 1–5 h, cells were visualized under Leica DMI4000 B Microscope (Leica, Wetzlar, Germany). Cells were visualized at different incubation times, or cells were detached and approximately 1 104 healthy cells were analyzed by BD FACS Calibur flow cytometer (BD, Heidelberg, Germany). 2.7. Cell viability assay
2.2. Hph1–Hph1–dsRBD protein construction purification TAT–dsRBD and Hph1–Hph1–TAT–TAT–dsRBD genes were constructed into PET 28b vector by restrict enzyme FastDigest NheI and FastDigest EcoRI (Fermentas, Leon-Rot, Germany). The plasmids of PET28b–TAT–dsRBD and PET28b–Hph1–Hph1–TAT– TAT–dsRBD were further cleaved by the restrict enzyme FastDigest PstI and FastDigest EcoRI (Fermentas, Leon-Rot, Germany), harvested, and dsRBD and PET28b–Hph1–Hph1 ligated dsRBD to obtain the novel plasmid Hph1–Hph1–dsRBD (see Supplementay data). The plasmid was transferred into E. coli (Rosetta(DE3) pLysS), and cells were induced by 500 mM isopropyl-b-D-thiogalactoside (IPTG) for 6 h at room temperature. The protein of Hph1–Hph1– dsRBD was isolated by Ni-NTA column and desalted by the PD-10 to 100 mM NaCl including 10% glycerol nuclease-free water. The CPP–EGFP was prepared using our previously reported method (Ma et al., 2009). 2.3. Preparation of Hph1–Hph1–dsRBD/siRNA Hph1–Hph1–dsRBD/siRNA and CaCl2-condensed Hph1–Hph1– dsRBD/siRNA were prepared as previously described (Baoum et al., 2012; Eguchi et al., 2009; Hua Li, 2014). Hph1–Hph1–dsRBD/siRNA prepared as follows, briefly, siRNA was dissolved in nuclease-free water. In the final 10 Ul solution, 2 Ul of (10 UM) siRNA was mixed with different mol rates of Hph1–Hph1–dsRBD, and then incubated on ice for 30 min to form the complex of Hph1– Hph1–dsRBD/siRNA. The preparation of CaCl2-condensed Hph1– Hph1–dsRBD/siRNA was similar process, siRNA was mixed with different mol ratio of Hph1–Hph1–dsRBD in the final of 6 ml solution, and then it was mixed with 6 Ul CaCl2 (e.g., 0, 20, 46.2, 138.4, 300 mM) solution with final CaCl2 concentrations of 0, 10, 23.1, 46.2, 69.2, or 150 mM and incubated on ice for 30 min prior to use. All the complex was freshly prepared before experiment.
Cell viability of Hph1–Hph1–dsRBD was assessed according to Cell Counting Kit-8 from Dojindo instructions (Kumamoto, Japan). Approximately 1 104 HeLa cells/well were preplated into 96-well plates. The following day, Hph1–Hph1–dsRBD in 20 ml DMEM medium was added and incubated for 48 h. Finally, 10 ml Cell Counting Kit-8 solution was added to cells and incubated for 1 h at 37 C. Cell viability was detected at an absorbance of 405 nm by Infinite1 M200 (Tecan, Männedorf, Schweiz). 2.8. RNA interference The gene silencing effect was determined on the exogenous gene GFP and endogenous gene GAPDH by siGFP or siGAPDH transferred by Hph1–Hph1–dsRBD. Cells were seeded into 96-well plates at a density of 1 104 cells/well. The following day cells were rinsed with DMEM medium. Hph1–Hph1–dsRBD/siRNA complexes or CaCl2-condensed Hph1–Hph1–dsRBD/siRNA complexes in DMEM medium were added to the cells and incubated at 37 C in 5% CO2 for 4 h. The medium was exchanged for complete culture medium. Cells were cultured for 24 h for RT-PCR or 48 h for fluorescent activated cell sorter (FACS) or western blot analysis. Negative siRNA (nontarget to the known gene) was applied as the negative control. The target siRNA transferred by Lipofectamin RNAiMax (Invitrogen) was used as a positive control. 2.9. Western blot HeLa cells were lyzed on ice with RIPA buffer. Cell lysates were separated by 12.5% SDS-PAGE and transferred to PVDF membrane. Membranes were then treated with primary antibody (anti-HSC70, anti GAPDH) according to the manufacturer’s instructions (2 component Kit – ECLTM western blot detection reagents (Invitrogen, Germany)). Signals were detected on X-ray film captured in a dark room.
2.4. Agarose gel retardation assay 2.10. RT-PCR The siRNA binding ability was determined by agarose gel retardation assay. Protein (Hph1–Hph1–dsRBD)/siRNA (0:1; 1:1; 2:1; 4:1; 8:1; 16:1, 100 pmol siRNA was used in each sample) complexes were prepared at various mol ratios and analyzed with gel retardation assay. 2.5. Cell culture HeLa cell line was cultured in DMEM medium supplemented with 10% heat inactivated FBS and 1% penicillin/streptomycin at 37 C and 5% CO2. The HeLa cell line steadily expressing GFP
RT-PCR was used to measure the efficiency of GAPDH gene silencing in vitro at 24 h after treatment. Cells were washed with PBS, detached, and harvested. Total RNA of the cell or graft heart was extracted by the RNeasy Mini Kit (Qiagen, Hilden, German) according to the manufacturer’s instructions. The concentration of extracted RNA was measured by Infinite1 M200 (Tecan, Männedorf, Schweiz). Approximately 1000 ng RNA was reverse transcribed into sscDNA using oligo-dT as the primer based on the protocol of the QuantiTect reverse transcription Kit (Qiagen, Hilden, German). Then, RT-PCR was performed with QuantiTect SYSR Green
208
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
DDC T ¼ DC TðtreatmentcontrolÞ ðC TGAPDHmRNA C T1BSrRNA Þ
(Fig. 1A). The recombinant proteins (length 151 amino acids, MW 17407.8 Da, and pI 11.44) were expressed in E. coli with one 6 His tag at the N terminal for purification, and isolated with high purity as shown in Fig. 1B. To determine siRNA binding ability, different molar rates of protein to siRNA were prepared. Fig. 1C shows the migration of protein/siRNA complex by agarose gel retardation assay. The degree of siRNA binding was dependent on the concentration of Hph1–Hph1–dsRBD. When the molar ratio reached 16:1 or above, siRNA remained in the wells indicating that all siRNA molecules were entrapped in Hph1–Hph1–dsRBD in the form of a stable protein–siRNA complex.
2.11. siRNA distribution in the heart and histological analysis
3.2. Cellular uptake of Hph1–Hph1–dsRBD siRNA
Male C57BL/6J (H-2b) mice weighing 25–30 g and aged between 8 and 20 weeks were obtained from the Animal Department of Universitätsklinikum Hamburg-Eppendorf for use in heart transplant experiments. All of the mice were treated in accordance with the animal guidelines approved by the local Animal Care and Research Committee. The heart transplant model was constructed following the method available in our laboratory. Briefly: mice were anesthetized with an i.p. injection of a mixed solution containing xylazine (2 mg/ml, Bayer, Germany) and ketamine (15 mg/ml, Bayer, Germany) at a dose of 100 ml/10 g body weight. After ensuring complete anesthesia, mice were intravenously perfused with 10 ml ice cold PBS to exchange the serum in the heart. Then, the aortic artery and the pulmonary trunk were dissected and ligated with a 5-0 suture. Surrounding vessels were also ligated. Then, all vessels connecting the heart were cut, the condensed protein/siRNA solution (50 ml) was injected into the ligated aortic artery, and the heart grafts were preserved in ice cold PBS for 1 h. After siRNA distribution, the hearts were immediately frozen in optimal cutting temperature solution (Sakura Finetechnical Co.,), and stored at 80 C. Heart was sliced into 5-mm sections with a cryostat, and immediately analyzed under fluorescence microscope.
To determine the cellular uptake of Hph1–Hph1–dsRBD/siRNA, studies were performed on HeLa cells. Compared with siRNA without Hph1–Hph1–dsRBD, strong green fluorescent signals were detected in HeLa cells treated with the Hph1–Hph1– dsRBD/FAM-siRNA complex for 1 h or 4 h (Fig. 2A). The results of flow cytometer analysis revealed that almost 100% of HeLa cells were FAM-siRNA positive; however, no positive cells were detected with siRNA treatment alone (Fig. 2B).
PCR Kit (Qiagen, Hilden, German) and the ABI 7500 Fast Real-time RT-PCR System (Applied Biosystems, Darmstadt, Germany) according to the manufacturer’s instructions. 18S rRNA was used as internal control gene and all samples were reproduced in triplicate. Data were analyzed according to Schmittgen and Livak (Schmittgen and Livak, 2008); the change in gene expression of GAPDH due to treatment was compared with the internal control, 18s rRNA. Foldchangeduetotreatment ¼ 2DDCT ;
2.12. Statistical analysis The experiment was assayed in triplicate on at least three independent occasions to get accurate results. Experimental data were analyzed using a Student’s t-test. 3. Results 3.1. Hph1–Hph1–dsRBD fusion protein preparation and characterization In our study, a novel fusion protein with two CPPs (Hph1–Hph1) and a single dsRBD was designed for the effective siRNA delivery
3.3. Gene silencing in vitro The Hph1–Hph1–dsRBD/anti-GFP siRNA (siGFP) complex induced GFP gene silencing in GFP stable-expression HeLa cells with molar ratio dependence (Fig. 3). Following 48-h treatment, the Hph1–Hph1–dsRBD/siGFP complex induced a 26.4%, 82.6%, 92.7%, and 98.5% knock down of GFP genes in cells at the Hph1–Hph1– dsRBD/siGFP ratio of 20:1, 30:1, 45:1, and 60:1, respectively. Protein combined with the negative control siRNA demonstrated a 5.45% silencing effect. Moreover, transferring siGFP with LipofectamineTM RNAiMAX (Lipofectamine) generated a knock down effect equal to that observed with Hph1–Hph1–dsRBD/siGFP at ratio of 30:1. The ability of the Hph1–Hph1–dsRBD/anti-GAPDH siRNA (siRNAGAPDH) complex to reduce endogenous gene GAPDH expression on protein (Fig. 4A) and mRNA (Fig. 4B) levels were further examined in HeLa cells 24 h or 48 h after transfection, respectively. The GAPDH mRNA level was significantly suppressed by Hph1–Hph1–dsRBD-mediated siRNAGAPDH delivery to only 29.2% and 37% of the initial GAPDH expression at the protein– siRNA ratio of 30:1 and 20:1, respectively. These results demonstrated that Hph1–Hph1–dsRBD was more efficient than LipofectamineTM RNAiMAX (43.6%) as a siRNA carrier. Similar results were also obtained from protein levels from western blot analysis as shown in Fig. 4A. These results were indicating that Hph1–Hph1–dsRBD is a highly efficient siRNA delivery vector in vitro.
Fig. 1. (A) Structure of the Hph1–Hph1–dsRBD, the green fragment is the cell penetrating peptide (Hph1); the red part stands for the dsRNA binding domain. (B) SDS Page analysis the purity of the protein stained with Coomassie blue. (C) Gel retardation assays of Hph1–Hph1–dsRBD binding siRNA at different molar rates (0:1, 1:1, 2:1, 4:1, 8:1, 16:1 mol/mol). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
209
Fig. 2. (A) Fluorescent microscope analysis concerning about Hph1–Hph1–dsRBD delivery of siRNA into HeLa cell. Fluorescent images regarded to the Hph1–Hph1–dsRBD combined with FAM-siRNA (green) at ratio of 30:1 for 1 h (2), or 4 h (3), or with FAM-siRNA at ratio of 1:1 condensed with 23.1 mM CaCl2 for 1 h (1). HeLa cells were exposed to PBS–siRNA-FAM complex for 1 h as the control (4). From left to right: figures were taken under bright field and green filter, and then an overlay of above was presented. The scale bar was defined as 20 mm. (B) HeLa cells were exposed to naked siRNA or Hph1–Hph1–dsRBD/siRNA-FAM complex for 1 h, and the cell uptake was measured by the FACS. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
210
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
Fig. 3. Gene Silencing Effect of Hph1–Hph1–dsRBD on GFP expression in HeLa cells. Flow cytometry analysis of RNAi efficiency on GFP gene in HeLa GFP cells after treatment with siGFP, different ratios of Hph1–Hph1–dsRBD:siRNA (20:1, 30:1, 45:1, 60:1) or Lipofectamine RNAiMAX, or with PBS or negative siRNA (negative control) as indicated.
3.4. Calcium enhances the efficiency of Hph1–Hph1–dsRBD in siRNA delivery Fig. 2A(1) showed that the Hph1–Hph1–dsRBD/siRNAGFP complex at a ratio of 1:1 was dramatically improved to form microparticles when condensed in 23.1 mM CaCl2. However, in the absence of calcium, fluorescence particles were not detected. Fig. 5A showed that the expression of GFP was unaffected when the protein–siRNA complex was condensed in 0 and 10 mM CaCl2 buffer. However, it was clearly inhibited at concentrations equal to or greater than 23.1 mM. The Hph1–Hph1–dsRBD:siRNA (P/S) complex, at ratios of 1:1 and 5:1, yielded very high RNAi efficiency at CaCl2 concentrations of 34.6 mM and 69.2 mM. This exceeded the performance of Lipofectamine. The maximum knock down level was achieved by the complex (P/S, 1:1) condensed at 69.2 mM CaCl2, which resulted in 94% GFP silencing. Irrespective of mixing
ratios, the P/S complex had a decreased tendency of siRNA silencing effects at higher CaCl2 concentrations (150 mM CaCl2). The knocked-down of GFP gene was evaluated in the presence of complexes with different protein:siRNA ratios (1:1, 5:1, and 10:1). These complexes were condensed at 10 mM (Fig. 5B) and 23.1 mM CaCl2 (Fig. 5C). The results indicated that the efficiency of RNAi was dependent on the amount of protein, and was proportional to the protein/siRNAGFP ratio when the complex was condensed at 10 mM CaCl2. The condensed complex could efficiently suppress GAPDH gene expression and showed higher GAPDH knock down at P/S ratios of 1:1 (83.4% knock down) and 5:1 (57% knocked down). The silencing effect of LipofectamineTM RNAiMAX was approximately 56.4% (Fig. 5D). Data from the western blot (Fig. 4A) also suggested that GAPDH was knocked down by Hph1–Hph1–dsRBD/siGPDH. Together, these data provide direct evidence that the generated
Fig. 4. (A) Knock down of GAPDH gene in HeLa cells on protein levels. Western blot analysis of GAPDH expression was measured 48 h after treatment of the complex (p/siRNA 30:1) with GAPDH siRNA or negative siRNA, or 23.1 mM CaCl2 complex of p/siRNA at ratio of 1:1. (B) Knock down of GAPDH gene in HeLa cells on mRNA levels. Quantitative RTPCR analysis of endogenous GAPDH mRNA expression in HeLa cells were carried out at 24 h after treatment of the complex (p/siRNA 20:1 or 30:1) including GAPDH siRNA or negative siRNA, GAPDH siRNA plus Lipofectamine RNAi MAX. Error bars represent standard deviation (n = 3). Asterisk (**) indicates p < 0.01 based on analysis of one sample ttest and one way ANOVA post hoc multiple comparisons (LSD as the variances assumed).
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
211
Fig. 5. (A) GFP knocked down condensed at various CaCl2 concentration. Flow cytometry analysis of RNAi efficiency on GFP gene in GFP HeLa cells at 48 h after treatment with the Hph1–Hph1–dsRBD/siRNAGFP complex at ratios of 1:1 or 5:1, respectively, which were condensed in the different concentration of CaCl2 (0, 10, 23.1, 34.6, 69.2 and 150 mM), or with Lipofectamine plus siGFP (positive control). Negative control was the negative siRNA with Hph1–Hph1–dsRBD (P/siRNA 2/1) as indicated. (B) Flow cytometry analysis of the percentage of GFP knocked down at various concentration of siRNAGFP (100, 50, 25 and 12.5 nM) prepared with different ratios of protein/siGFP (1:1, 1:5 and 1:10), which were condensed in the 10 mM (B) or 23.1 mM (C) CaCl2, respectivly. (D) Calcium condensed Hph1–Hph1–dsRBD/siRNA complexes affect the efficient of GAPDH gene expression. Quantitative RT-PCR analysis of endogenous GAPDH mRNA expression in HeLa cells at 24 h after treatment of the Hph1–Hph1–dsRBD with GAPDH siRNA or with negative siRNA (p/siRNA 1/1 or 5/1), which were condensed in 23.1 mM CaCl2. The GAPDH siRNA attached with Lipofectamine RNAiMAX was designed as the positive control. Error bars represent standard deviation (n = 3). Asterisk (**) indicates p < 0.01 based on analysis of one sample t-test and one way ANOVA post hoc multiple comparisons (LSD as the variances assumed).
Hph1–Hph1–dsRBD/siRNA complexes achieve better gene silencing effects than the commercially available siRNA delivery vectors, LipofectamineTM RNAiMAX. 3.5. Cytotoxicity of Hph1–Hph1–dsRBD Cell proliferation was not significantly suppressed in the medium containing Hph1–Hph1–dsRBD compared to PBS control (Fig. 6). Furthermore, cell viability studies demonstrated that Hph1–Hph1–dsRBD exhibited significantly low cytotoxicity. Similar results were obtained with Lipofectamine RNAiMAX. These data demonstrate that the Hph1–Hph1–dsRBD carrier was associated with low cytotoxicity. 3.6. Application of Hph1–Hph1–dsRBD in vivo
Fig. 6. Cell viability of Hph1–Hph1–dsRBD. Cell viability was analyzed using CCK8 Kit by measuring the absorbance in each well at OD 450. Error bars represent standard deviation (n = 3). “0.1, 1.5 and 4.5 mM Ca2+ condensed” was 0.1, 1.5 and 4.5 mM of Hph1–Hph1–dsRBD was condensed in 23.1 mM Ca2+ solution; 0.1, 1.5, 4.5, 6.0, 9.0 mM was respectively treated with 0.1, 1.5, 4.5, 6.0, 9.0 mM of Hph1–Hph1– dsRBD. The PBS and the lipofectamine was the control group.
The heart transplantation model is shown in Fig. 7A. Images revealed that FAM-siRNA bound with Hph1–Hph1–dsRBD entered the heart and was distributed throughout (Fig. 7B(3)). Similar observations were made with CPP–EGFP (enhanced green fluorescent protein) (Fig. 7B(2)). However, no significant fluorescent signals were detected with naked FAM-siRNA, which had not been
212
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
Fig. 7. (A) Administration Hph1–Hph1–dsRBD/siRNA-FAM. The heart from mouse was perfused using 10 ml PBS, and then the Hph1–Hph1–dsRBD/siRNA was injected into the heart and incubated PBS on ice for 30 mins. (B) The siRNA distribution in the mouse heart, 1, naked siRNA labeled FAM (green); 2, mouse heart transferred with CPP–EGFP; 3, mouse heart transferred with Hph1–Hph1–dsRBD/FAM-siRNA (green). Each sample was triplicated and only one of them was shown here. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
conjugated with Hph1–Hph1–dsRBD (Fig. 7B(1)). These results confirm the ability of Hph1–Hph1–dsRBD to transfer and distribute siRNA throughout the heart. 4. Discussion CPPs are capable of transferring various substrates into target cells. However, direct transfer of siRNA with CPP results in charge neutralization of CPP, aggregation/precipitation of the CPP/siRNA complex, which limits the application of CPP in siRNA delivery (Eguchi et al., 2009; Palm-apergi et al., 2011). In this study, we designed a novel fusion protein (Hph1–Hph1–dsRBD) to act as a siRNA carrier. The results demonstrated that this novel fusion protein was capable of delivering siRNA in vitro and in vivo with high efficiency. Although Hph1–Hph1–dsRBD completely binds to siRNA at ratio of 16:1, a slightly higher preparation ratio of at least 20:1 is necessary for effective siRNA delivery into cells and to achieve a desirable silencing effect. One possible reason for this phenomenon may be that two positive CPPs (Hph1) together with the dsRBD in the Hph1–Hph1–dsRBD complex could themselves bind to siRNA. This undesired binding of Hph1 with siRNA may lead to charge neutralization and invalidity of Hph1 (Eguchi et al., 2009). Notably, Hph1–Hph1–dsRBD was more efficient than naked CPP as a siRNA carrier. Compared with CPP noncovalently bonded with siRNA, Hph1–Hph1–dsRBD demands much less consumption of protein to achieve similar silencing effects, e.g., CPP of MPG (GALFLGFLGAAGSTMG) at ratio of 84:1 (Simeoni et al., 2003), and oligoarginine at ratio of 56:1 (Veldhoen et al., 2006). Additionally, Hph1–Hph1–dsRBD showed higher efficiency than CPP as a siRNA carrier. For example, the Hph1–Hph1–dsRBD/siRNA complex could inhibit up to a maximum 80% of the target gene expression at
100 nM concentrations of endogenous (GAPDH) or exogenous gene (GFP) siRNA, however, siRNA delivery by CPP of MPG, penetratin (RQIKIWFQNRRMKWKK), TP10 (AGYKKGKINLKALAALAKKIL), or EB1 (LIRLWSHLIHIWFQNRRLKWKKK) achieved only 60% (Simeoni et al., 2003), 0%, 18%, and 55% (Lundberg et al., 2007) at a final concentration of 100 nM siGFP, respectively, and oligoarginine reached 43% inhibition at double the concentration of siRNA (Veldhoen et al., 2006). In contrast to CPP covalently linked with siRNA, such as siRNA delivery by Tat (YGRKKRRQRRR) on HeLa cells, the carrier Hph1–Hph1–dsRBD demonstrated superior advantages. Not only is its preparation simple, but it also has a high transfection rate (80% knocked down at 100 nM Hph1–Hph1– dsRBD–siRNA compared with 70% at 200 nM siRNA) (Chiu et al., 2004). Although there are some reports that siRNA delivered by the CPP of penetratin generates RNAi efficiency of 85% in neuron cells (Davidson et al., 2004), others report much lower target gene silencing effects in the region of 20% (Moschos et al., 2007) and 53% (Muratovska and Eccles, 2004). In summary, Hph1–Hph1–dsRBD demonstrated considerably better transfection performance compared with other covalently linked or noncovalently linked siRNA carriers. We also investigated the effects of calcium condensation on the siRNA delivery efficiency of the Hph1–Hph1–dsRBD/siRNA complex according to established methods (Baoum et al., 2012). There was evidence that the efficiency of Hph1–Hph1–dsRBD delivery of siRNA positively correlated with calcium concentrations at siRNA conditions of 12.5 nM. The highest RNAi effect was reached when the Hph1–Hph1–dsRBD/siRNA complex was condensed with 69.2 mM CaCl2. However, similar results were observed when condensed at 23.1 mM CaCl2 and 25 nM siGFP, in which almost complete suppression of GFP genes was observed. Since complexes condensed in 10 mM CaCl2 did not demonstrate significant gene
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
knock down, we can conclude that a certain amount of calcium might compensate or reinforce the function of protein–siRNA complexes. The mechanism of action of calcium can be elucidated from previously published data. The first hypothesis is that the complex of protein–siRNA could form nanoparticles when condensed in concentrations of CaCl2 above 23.1 mM (Baoum et al., 2012; Khondee et al., 2011). Furthermore, calcium could also disrupt the endosomes (Hoyer and Neundorf, 2012; Liu et al., 2012), in addition, calcium part plays an important role in CPP based cell penetrating (Lorents et al., 2012; Nascimento et al., 2012; PalmApergi et al., 2009). In vivo application, the FAM labeled siRNA delivery propelled by Hph1–Hph1–dsRBD could enter and distribute in the heart tissue, and no significant signal was observed in the control group without Hph1–Hph1–dsRBD. This result demonstrated that Hph1– Hph1–dsRBD could successfully transfer siRNA to the heart graft in vivo. To sum up, Hph1–Hph1–dsRBD as siRNA carrier presented very promising efficiency both in vitro and in vivo. Conflict of interest There is no conflict of interest. Acknowledgment This work was was partly supported by funding from Deutsche Forschungsgemeinschaft (TS175/1-1). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. ijpharm.2014.04.050. Reference Baoum, A., Ovcharenko, D., Berkland, C., 2012. Calcium condensed cell penetrating peptide complexes offer highly efficient, low toxicity gene silencing. International Journal of Pharmaceutics 427, 134–142. Chiu, Y.-L., Ali, A., Chu, C.-Y., Cao, H., Rana, T.M., 2004. Visualizing a correlation between siRNA localization, cellular uptake, and RNAi in living cells. Chemistry and Biology 11, 1165–1175. Choi, J.-M., Ahn, M.-H., Chae, W.-J., Jung, Y.-G., Park, J.-C., Song, H.-M., Kim, Y.-E., Shin, J.-A., Park, C.-S., Park, J.-W., Park, T.-K., Lee, J.-H., Seo, B.-F., Kim, K.-D., Kim, E.-S., Lee, D.-H., Lee, S.-K., Lee, S.-K., 2006. Intranasal delivery of the cytoplasmic domain of CTLA-4 using a novel protein transduction domain prevents allergic inflammation. Nature Medicine 12, 574–579.
213
Choi, J.-M., Kim, S.-H., Shin, J.-H., Gibson, T., Yoon, B.-S., Lee, D.-H., Lee, S.-K., Bothwel l, A.L.M., Lim, J.-S., Lee, S.-K., 2008. Transduction of the cytoplasmic domain of CTLA-4 inhibits TcR-specific activation signals and prevents collagen-induced arthritis. Proceedings of the National Academy of Sciences of the United States of America 105, 19875–19880. Davidson, T.J., Harel, S., Arboleda, V.A., Prunell, G.F., Shelanski, M.L., Greene, L.A., Troy, C.M., 2004. Highly efficient small interfering RNA delivery to primary mammalian neurons induces microRNA-like effects before mRNA degradation. Journal of Neuroscience 24, 10040–10046. Eguchi, A., Meade, B.R., Chang, Y.-C., Fredrickson, C.T., Willert, K., Puri, N., Dowdy, S.F., 2009. Efficient siRNA delivery into primary cells by a peptide transduction domain-dsRNA binding domain fusion protein. Nature Biotechnology 27, 567– 571. Hoyer, J., Neundorf, I., 2012. Peptide vectors for the nonviral delivery of nucleic acids. Accounts of Chemical Research 45, 1048–1056. Li, H., Tsui, T.Y., 2014. Six-cell penetrating peptide based fusion protein for siRNA delivery. Drug Delivery Epub ahead of print. Khondee, S., Baoum, A., Siahaan, T.J., Berkland, C., 2011. Calcium condensed LABL– TAT complexes effectively target gene delivery to ICAM-1 expressing cells. Molecular Pharmaceutics 8, 788–798. Lindgren, M., Llbrink, H.M.P.A., Langel, U., 2000. Cell-penetrating peptides. Trends in Pharmacological Sciences 21, 99–103. Liu, B.R., Lin, M.-D., Chiang, H.-J., Lee, H.-J., 2012. Arginine-rich cell-penetrating peptides deliver gene into living human cells. Gene 505, 37–45. Lorents, A., Kodavali, P.K., Oskolkov, N., Langel, Ü., Hällbrink, M., Pooga, M., 2012. Cell-penetrating peptides split into two groups based on modulation of intracellular calcium concentration. Journal of Biological Chemistry 287, 16880– 16889. Lundberg, P., El-Andaloussi, S., Sütlü, T., Johansson, H., Langel, U., 2007. Delivery of short interfering RNA using endosomolytic cell-penetrating peptides. FASEB Journal 21, 2664–2671. Ma, J., Lau, C.K., Obed, A., Dada, A., Doenecke, A., Fan, S.T., Schlitt, H.J., Tsui, T.Y., 2009. A cell penetrating heme oxygenase protein protects heart graft against ischemia/reperfusion injury. Gene Therapy 16, 320–328. Moschos, S.A., Jones, S.W., Perry, M.M., Williams, A.E., Erjefalt, J.S., Turner, J.J., Barnes, P.J., Sproat, B.S., Gait, M.J., Lindsay, M.A., 2007. Lung delivery studies using siRNA conjugated to TAT (48–60) and penetratin reveal peptide induced reduction in gene expression and induction of innate immunity. Bioconjugate Chemistry 18, 1450–1459. Muratovska, A., Eccles, M.R., 2004. Conjugate for efficient delivery of short interfering RNA (siRNA) into mammalian cells. FEBS Letters 558, 63–68. Nascimento, F.D., Sancey, L., Pereira, A., Rome, C., Oliveira, V., Oliveira, E.B., Nader, H. B., Yamane, T., Kerkis, I., Tersariol, I.L.S., Coll, J.-L., Hayashi, M.A.F., 2012. The natural cell-penetrating peptide crotamine targets tumor tissue in vivo and triggers a lethal calcium-dependent pathway in cultured cells. Molecular Pharmaceutics 9, 211–221. Palm-Apergi, C., Lorents, A., Padari, K., Pooga, M., Hällbrink, M., 2009. The membrane repair response masks membrane disturbances caused by cellpenetrating peptide uptake. FASEB Journal 23, 214–223. Palm-apergi, C., Eguchi, A., Dowdy, S.F., 2011. PTD–DRBD siRNA Delivery. Methods in Molecular Biology 683, 339–347. Schmittgen, T.D., Livak, K.J., 2008. Analyzing real-time PCR data by the comparative CT method. Nature Protocols 3, 1101–1108. Simeoni, F., Morris, M.C., Heitz, F., Divita, G., 2003. Insight into the mechanism of the peptide-based gene delivery system MPG: implications for delivery of siRNA into mammalian cells. Nucleic Acids Research 31, 2717–2724. Veldhoen, S., Laufer, S.D., Trampe, A., Restle, T., 2006. Cellular delivery of small interfering RNA by a non-covalently attached cell-penetrating peptide: quantitative analysis of uptake and biological effect. Nucleic Acids Research 34, 6561–6573.
Contents lists available at ScienceDirect
International Journal of Pharmaceutics journal homepage: www.elsevier.com/locate/ijpharm
Highly efficient delivery of siRNA to a heart transplant model by a novel cell penetrating peptide-dsRNA binding domain Hua Li a,b , Xiangtao Zheng a , Viktoria Koren a , Yogesh Kumar Vashist a , Tung Yu Tsui a, * a b
The Department of General, Visceral and Thoracic Surgery, University Medical Center Hamburg-Eppendorf, Hamburg, Germany Institute for Regenerative Medicine and Cancer, Huzhou University School of Medicine, Huzhou, China
A R T I C L E I N F O
A B S T R A C T
Article history: Received 20 March 2014 Received in revised form 18 April 2014 Accepted 19 April 2014 Available online 23 April 2014
Small interfering RNAs (siRNAs) delivery remains a bottleneck for RNA interference (RNAi) – based therapies in the clinic. In the present study, a fusion protein with two cell-penetrating peptides (CPP), Hph1–Hph1, and a double-stranded RNA binding domain (dsRBD), was constructed for the siRNA delivery: dsRBD was designed to bind siRNA, and CPP would subsequently transport the dsRBD/siRNA complex into cells. We assessed the efficiency of the fusion protein, Hph1–Hph1–dsRBD, as a siRNA carrier. Calcium-condensed effects were assessed on GAPDH and green fluorescent protein (GFP) genes by western blot, real time polymerase chain reaction (RT-PCR), and flow cytometry analysis in vitro. Evaluations were also made in an in vivo heart transplantation model. The results demonstrated that the fusion protein, Hph1–Hph1–dsRBD, is highly efficient at delivering siRNA in vitro, and exhibits efficiency on GAPDH and GFP genes similar to or greater than lipofectamine. Interestingly, the calcium-condensed effects dramatically enhanced cellular uptake of the protein–siRNA complex. In vivo, Hph1–Hph1–dsRBD transferred and distributed ^ targeted siRNA throughout the whole mouse heart graft. Together, these results indicate that Hph1–Hph1–dsRBD has potential as an siRNA carrier for applications in the clinic or in biomedical research. ã 2014 Elsevier B.V. All rights reserved.
Keywords: Cell penetrating peptide RNA interference Gene delivery Double-stranded RNA binding domain
1. Introduction Nowadays, the development of RNA interference (RNAi) as an innovative biomedical tool provides new perspectives for pharmaceutical and medical research. However, until recently, small interfering RNA (siRNA) delivery constituted the greatest challenge for siRNA application in the clinic due to its high negative charge, large molecular weight, and hydrophility. Cell penetrating peptide (CPP) composed of 4–30 amino acid residues with arginine or lysine rich peptide (Lindgren et al., 2000), have been shown to transfer various substances, including siRNA, into different cell types (Eguchi et al., 2009). It was recently reported that CPP transfers siRNA through 2 strategies: covalent and noncovalent associations. During delivery of siRNA into target cells, the negative charge of siRNA may neutralize the positive charge of CPP, resulting in inhibition of siRNA delivery efficiciency. To overcome these issues, we have previously showed the CPP type and the number of repetitive CPP at the N terminal of fusion protein
of CPP-dsRBD (double-stranded RNA (dsRNA) binding domain (dsRBD)) play the key role of siRNA delivery and found the CPP of Hph1 was more efficacy than the CPP of TAT in siRNA delivery (Li and Tsui, 2014). In order to get a high efficient siRNA carrier, here we further provided a fusion protein consisting of 2 CPPs (Hph1 (YARVRRRGPRR)), which could provide efficient transfer into a primary cell (Choi et al., 2008, 2006), and one dsRBD for efficient siRNA delivery. The dsRBD was designed to bind siRNA, and the CPP would subsequently transport the dsRBD/siRNA into cells. The efficiency of Hph1–Hph1–dsRBD as a siRNA carrier for siRNA delivery was assessed in vitro, and the effectiveness of RNAi was determined. Moreover, the affect of the calcium-condensed step in siRNA delivery of Hph1–Hph1–dsRBD was also studied. Finally, the efficacy of CPP-dsRBD in siRNA delivery in vivo was investigated in a mouse heart transplantation model. 2. Material and methods 2.1. Materials
* Corresponding author at: Department of General, Visceral and Thoracic Surgery, University Medical Center Hamburg-Eppendorf, Martinistr. 52, 20246 Hamburg, Germany. Tel.: +49 40 741050822; fax: +49 40 7410 46756. E-mail address: [email protected] (T.Y. Tsui). http://dx.doi.org/10.1016/j.ijpharm.2014.04.050 0378-5173/ ã 2014 Elsevier B.V. All rights reserved.
The genes of TAT-dsRBD and Hph1–Hph1–TAT–TAT–dsRBD (the dsRBD was from double stranded RNA activated protein kinase (Homo sapiens), amino acid numbers 11–78, GenBank: AAF13156.1), were synthesized by Eurofin (Ebersberg, Germany).
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
Rosetta(DE3)pLysS and TOP10 competetent Escherichia coli cells were purchased from Merck (Darmstadt, Germany) and Invitrogen (Darmstadt, Germany), respectively. Ni-NTA column, complete protease inhibitor, and PD-10 columns were purchased from Qiagen (Hilden, Germany), Roche (Mannheim, Germany), and Invitrogen, respectively. Cell culture medium reagents, including DMEM, RIPM 160, FBS, DMSO, penicillin/streptomycin, pyruvate, mercaptoefhenol, and glutamine were obtained from Invitrogen. 18S rRNA and GAPDH primers for real time polymerase chain reaction (RT-PCR), and GAPDH siRNA were ordered from Qiagen. Negative siRNA, FAM-siRNA, and GFP siRNA were purchased from Ambion (Darmstadt, Germany). HeLa cell line was purchased from American Type Culture Collection (ATCC, Manassas, USA), and the GFP stable expressing HeLa cell line (GFP HeLa) was generated by using the pcDNA3.1+ GFP construct (Invitrogen, Steinheim, Germany). All other reagents were obtained from Sigma–Aldrich (Steinheim, Germany) unless otherwise noted.
207
performed equally to the HeLa cell line with the addition of 1.5 mg/ml gentamicin to the medium (Invitrogen). 2.6. Cell uptake assay HeLa cells (100 ml) were preplated in 96-well plates at a density of 1 104 cells/well and incubated for 24 h. Cells were then rinsed twice with serum-free DMEM medium. Supplements of serumfree DMEM medium and the different ratios of Hph1–Hph1– dsRBD/FAM-labeled siRNA complex were added to the cells. The final siRNA concentration was 100 nM. After an incubation period of 1–5 h, cells were visualized under Leica DMI4000 B Microscope (Leica, Wetzlar, Germany). Cells were visualized at different incubation times, or cells were detached and approximately 1 104 healthy cells were analyzed by BD FACS Calibur flow cytometer (BD, Heidelberg, Germany). 2.7. Cell viability assay
2.2. Hph1–Hph1–dsRBD protein construction purification TAT–dsRBD and Hph1–Hph1–TAT–TAT–dsRBD genes were constructed into PET 28b vector by restrict enzyme FastDigest NheI and FastDigest EcoRI (Fermentas, Leon-Rot, Germany). The plasmids of PET28b–TAT–dsRBD and PET28b–Hph1–Hph1–TAT– TAT–dsRBD were further cleaved by the restrict enzyme FastDigest PstI and FastDigest EcoRI (Fermentas, Leon-Rot, Germany), harvested, and dsRBD and PET28b–Hph1–Hph1 ligated dsRBD to obtain the novel plasmid Hph1–Hph1–dsRBD (see Supplementay data). The plasmid was transferred into E. coli (Rosetta(DE3) pLysS), and cells were induced by 500 mM isopropyl-b-D-thiogalactoside (IPTG) for 6 h at room temperature. The protein of Hph1–Hph1– dsRBD was isolated by Ni-NTA column and desalted by the PD-10 to 100 mM NaCl including 10% glycerol nuclease-free water. The CPP–EGFP was prepared using our previously reported method (Ma et al., 2009). 2.3. Preparation of Hph1–Hph1–dsRBD/siRNA Hph1–Hph1–dsRBD/siRNA and CaCl2-condensed Hph1–Hph1– dsRBD/siRNA were prepared as previously described (Baoum et al., 2012; Eguchi et al., 2009; Hua Li, 2014). Hph1–Hph1–dsRBD/siRNA prepared as follows, briefly, siRNA was dissolved in nuclease-free water. In the final 10 Ul solution, 2 Ul of (10 UM) siRNA was mixed with different mol rates of Hph1–Hph1–dsRBD, and then incubated on ice for 30 min to form the complex of Hph1– Hph1–dsRBD/siRNA. The preparation of CaCl2-condensed Hph1– Hph1–dsRBD/siRNA was similar process, siRNA was mixed with different mol ratio of Hph1–Hph1–dsRBD in the final of 6 ml solution, and then it was mixed with 6 Ul CaCl2 (e.g., 0, 20, 46.2, 138.4, 300 mM) solution with final CaCl2 concentrations of 0, 10, 23.1, 46.2, 69.2, or 150 mM and incubated on ice for 30 min prior to use. All the complex was freshly prepared before experiment.
Cell viability of Hph1–Hph1–dsRBD was assessed according to Cell Counting Kit-8 from Dojindo instructions (Kumamoto, Japan). Approximately 1 104 HeLa cells/well were preplated into 96-well plates. The following day, Hph1–Hph1–dsRBD in 20 ml DMEM medium was added and incubated for 48 h. Finally, 10 ml Cell Counting Kit-8 solution was added to cells and incubated for 1 h at 37 C. Cell viability was detected at an absorbance of 405 nm by Infinite1 M200 (Tecan, Männedorf, Schweiz). 2.8. RNA interference The gene silencing effect was determined on the exogenous gene GFP and endogenous gene GAPDH by siGFP or siGAPDH transferred by Hph1–Hph1–dsRBD. Cells were seeded into 96-well plates at a density of 1 104 cells/well. The following day cells were rinsed with DMEM medium. Hph1–Hph1–dsRBD/siRNA complexes or CaCl2-condensed Hph1–Hph1–dsRBD/siRNA complexes in DMEM medium were added to the cells and incubated at 37 C in 5% CO2 for 4 h. The medium was exchanged for complete culture medium. Cells were cultured for 24 h for RT-PCR or 48 h for fluorescent activated cell sorter (FACS) or western blot analysis. Negative siRNA (nontarget to the known gene) was applied as the negative control. The target siRNA transferred by Lipofectamin RNAiMax (Invitrogen) was used as a positive control. 2.9. Western blot HeLa cells were lyzed on ice with RIPA buffer. Cell lysates were separated by 12.5% SDS-PAGE and transferred to PVDF membrane. Membranes were then treated with primary antibody (anti-HSC70, anti GAPDH) according to the manufacturer’s instructions (2 component Kit – ECLTM western blot detection reagents (Invitrogen, Germany)). Signals were detected on X-ray film captured in a dark room.
2.4. Agarose gel retardation assay 2.10. RT-PCR The siRNA binding ability was determined by agarose gel retardation assay. Protein (Hph1–Hph1–dsRBD)/siRNA (0:1; 1:1; 2:1; 4:1; 8:1; 16:1, 100 pmol siRNA was used in each sample) complexes were prepared at various mol ratios and analyzed with gel retardation assay. 2.5. Cell culture HeLa cell line was cultured in DMEM medium supplemented with 10% heat inactivated FBS and 1% penicillin/streptomycin at 37 C and 5% CO2. The HeLa cell line steadily expressing GFP
RT-PCR was used to measure the efficiency of GAPDH gene silencing in vitro at 24 h after treatment. Cells were washed with PBS, detached, and harvested. Total RNA of the cell or graft heart was extracted by the RNeasy Mini Kit (Qiagen, Hilden, German) according to the manufacturer’s instructions. The concentration of extracted RNA was measured by Infinite1 M200 (Tecan, Männedorf, Schweiz). Approximately 1000 ng RNA was reverse transcribed into sscDNA using oligo-dT as the primer based on the protocol of the QuantiTect reverse transcription Kit (Qiagen, Hilden, German). Then, RT-PCR was performed with QuantiTect SYSR Green
208
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
DDC T ¼ DC TðtreatmentcontrolÞ ðC TGAPDHmRNA C T1BSrRNA Þ
(Fig. 1A). The recombinant proteins (length 151 amino acids, MW 17407.8 Da, and pI 11.44) were expressed in E. coli with one 6 His tag at the N terminal for purification, and isolated with high purity as shown in Fig. 1B. To determine siRNA binding ability, different molar rates of protein to siRNA were prepared. Fig. 1C shows the migration of protein/siRNA complex by agarose gel retardation assay. The degree of siRNA binding was dependent on the concentration of Hph1–Hph1–dsRBD. When the molar ratio reached 16:1 or above, siRNA remained in the wells indicating that all siRNA molecules were entrapped in Hph1–Hph1–dsRBD in the form of a stable protein–siRNA complex.
2.11. siRNA distribution in the heart and histological analysis
3.2. Cellular uptake of Hph1–Hph1–dsRBD siRNA
Male C57BL/6J (H-2b) mice weighing 25–30 g and aged between 8 and 20 weeks were obtained from the Animal Department of Universitätsklinikum Hamburg-Eppendorf for use in heart transplant experiments. All of the mice were treated in accordance with the animal guidelines approved by the local Animal Care and Research Committee. The heart transplant model was constructed following the method available in our laboratory. Briefly: mice were anesthetized with an i.p. injection of a mixed solution containing xylazine (2 mg/ml, Bayer, Germany) and ketamine (15 mg/ml, Bayer, Germany) at a dose of 100 ml/10 g body weight. After ensuring complete anesthesia, mice were intravenously perfused with 10 ml ice cold PBS to exchange the serum in the heart. Then, the aortic artery and the pulmonary trunk were dissected and ligated with a 5-0 suture. Surrounding vessels were also ligated. Then, all vessels connecting the heart were cut, the condensed protein/siRNA solution (50 ml) was injected into the ligated aortic artery, and the heart grafts were preserved in ice cold PBS for 1 h. After siRNA distribution, the hearts were immediately frozen in optimal cutting temperature solution (Sakura Finetechnical Co.,), and stored at 80 C. Heart was sliced into 5-mm sections with a cryostat, and immediately analyzed under fluorescence microscope.
To determine the cellular uptake of Hph1–Hph1–dsRBD/siRNA, studies were performed on HeLa cells. Compared with siRNA without Hph1–Hph1–dsRBD, strong green fluorescent signals were detected in HeLa cells treated with the Hph1–Hph1– dsRBD/FAM-siRNA complex for 1 h or 4 h (Fig. 2A). The results of flow cytometer analysis revealed that almost 100% of HeLa cells were FAM-siRNA positive; however, no positive cells were detected with siRNA treatment alone (Fig. 2B).
PCR Kit (Qiagen, Hilden, German) and the ABI 7500 Fast Real-time RT-PCR System (Applied Biosystems, Darmstadt, Germany) according to the manufacturer’s instructions. 18S rRNA was used as internal control gene and all samples were reproduced in triplicate. Data were analyzed according to Schmittgen and Livak (Schmittgen and Livak, 2008); the change in gene expression of GAPDH due to treatment was compared with the internal control, 18s rRNA. Foldchangeduetotreatment ¼ 2DDCT ;
2.12. Statistical analysis The experiment was assayed in triplicate on at least three independent occasions to get accurate results. Experimental data were analyzed using a Student’s t-test. 3. Results 3.1. Hph1–Hph1–dsRBD fusion protein preparation and characterization In our study, a novel fusion protein with two CPPs (Hph1–Hph1) and a single dsRBD was designed for the effective siRNA delivery
3.3. Gene silencing in vitro The Hph1–Hph1–dsRBD/anti-GFP siRNA (siGFP) complex induced GFP gene silencing in GFP stable-expression HeLa cells with molar ratio dependence (Fig. 3). Following 48-h treatment, the Hph1–Hph1–dsRBD/siGFP complex induced a 26.4%, 82.6%, 92.7%, and 98.5% knock down of GFP genes in cells at the Hph1–Hph1– dsRBD/siGFP ratio of 20:1, 30:1, 45:1, and 60:1, respectively. Protein combined with the negative control siRNA demonstrated a 5.45% silencing effect. Moreover, transferring siGFP with LipofectamineTM RNAiMAX (Lipofectamine) generated a knock down effect equal to that observed with Hph1–Hph1–dsRBD/siGFP at ratio of 30:1. The ability of the Hph1–Hph1–dsRBD/anti-GAPDH siRNA (siRNAGAPDH) complex to reduce endogenous gene GAPDH expression on protein (Fig. 4A) and mRNA (Fig. 4B) levels were further examined in HeLa cells 24 h or 48 h after transfection, respectively. The GAPDH mRNA level was significantly suppressed by Hph1–Hph1–dsRBD-mediated siRNAGAPDH delivery to only 29.2% and 37% of the initial GAPDH expression at the protein– siRNA ratio of 30:1 and 20:1, respectively. These results demonstrated that Hph1–Hph1–dsRBD was more efficient than LipofectamineTM RNAiMAX (43.6%) as a siRNA carrier. Similar results were also obtained from protein levels from western blot analysis as shown in Fig. 4A. These results were indicating that Hph1–Hph1–dsRBD is a highly efficient siRNA delivery vector in vitro.
Fig. 1. (A) Structure of the Hph1–Hph1–dsRBD, the green fragment is the cell penetrating peptide (Hph1); the red part stands for the dsRNA binding domain. (B) SDS Page analysis the purity of the protein stained with Coomassie blue. (C) Gel retardation assays of Hph1–Hph1–dsRBD binding siRNA at different molar rates (0:1, 1:1, 2:1, 4:1, 8:1, 16:1 mol/mol). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
209
Fig. 2. (A) Fluorescent microscope analysis concerning about Hph1–Hph1–dsRBD delivery of siRNA into HeLa cell. Fluorescent images regarded to the Hph1–Hph1–dsRBD combined with FAM-siRNA (green) at ratio of 30:1 for 1 h (2), or 4 h (3), or with FAM-siRNA at ratio of 1:1 condensed with 23.1 mM CaCl2 for 1 h (1). HeLa cells were exposed to PBS–siRNA-FAM complex for 1 h as the control (4). From left to right: figures were taken under bright field and green filter, and then an overlay of above was presented. The scale bar was defined as 20 mm. (B) HeLa cells were exposed to naked siRNA or Hph1–Hph1–dsRBD/siRNA-FAM complex for 1 h, and the cell uptake was measured by the FACS. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
210
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
Fig. 3. Gene Silencing Effect of Hph1–Hph1–dsRBD on GFP expression in HeLa cells. Flow cytometry analysis of RNAi efficiency on GFP gene in HeLa GFP cells after treatment with siGFP, different ratios of Hph1–Hph1–dsRBD:siRNA (20:1, 30:1, 45:1, 60:1) or Lipofectamine RNAiMAX, or with PBS or negative siRNA (negative control) as indicated.
3.4. Calcium enhances the efficiency of Hph1–Hph1–dsRBD in siRNA delivery Fig. 2A(1) showed that the Hph1–Hph1–dsRBD/siRNAGFP complex at a ratio of 1:1 was dramatically improved to form microparticles when condensed in 23.1 mM CaCl2. However, in the absence of calcium, fluorescence particles were not detected. Fig. 5A showed that the expression of GFP was unaffected when the protein–siRNA complex was condensed in 0 and 10 mM CaCl2 buffer. However, it was clearly inhibited at concentrations equal to or greater than 23.1 mM. The Hph1–Hph1–dsRBD:siRNA (P/S) complex, at ratios of 1:1 and 5:1, yielded very high RNAi efficiency at CaCl2 concentrations of 34.6 mM and 69.2 mM. This exceeded the performance of Lipofectamine. The maximum knock down level was achieved by the complex (P/S, 1:1) condensed at 69.2 mM CaCl2, which resulted in 94% GFP silencing. Irrespective of mixing
ratios, the P/S complex had a decreased tendency of siRNA silencing effects at higher CaCl2 concentrations (150 mM CaCl2). The knocked-down of GFP gene was evaluated in the presence of complexes with different protein:siRNA ratios (1:1, 5:1, and 10:1). These complexes were condensed at 10 mM (Fig. 5B) and 23.1 mM CaCl2 (Fig. 5C). The results indicated that the efficiency of RNAi was dependent on the amount of protein, and was proportional to the protein/siRNAGFP ratio when the complex was condensed at 10 mM CaCl2. The condensed complex could efficiently suppress GAPDH gene expression and showed higher GAPDH knock down at P/S ratios of 1:1 (83.4% knock down) and 5:1 (57% knocked down). The silencing effect of LipofectamineTM RNAiMAX was approximately 56.4% (Fig. 5D). Data from the western blot (Fig. 4A) also suggested that GAPDH was knocked down by Hph1–Hph1–dsRBD/siGPDH. Together, these data provide direct evidence that the generated
Fig. 4. (A) Knock down of GAPDH gene in HeLa cells on protein levels. Western blot analysis of GAPDH expression was measured 48 h after treatment of the complex (p/siRNA 30:1) with GAPDH siRNA or negative siRNA, or 23.1 mM CaCl2 complex of p/siRNA at ratio of 1:1. (B) Knock down of GAPDH gene in HeLa cells on mRNA levels. Quantitative RTPCR analysis of endogenous GAPDH mRNA expression in HeLa cells were carried out at 24 h after treatment of the complex (p/siRNA 20:1 or 30:1) including GAPDH siRNA or negative siRNA, GAPDH siRNA plus Lipofectamine RNAi MAX. Error bars represent standard deviation (n = 3). Asterisk (**) indicates p < 0.01 based on analysis of one sample ttest and one way ANOVA post hoc multiple comparisons (LSD as the variances assumed).
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
211
Fig. 5. (A) GFP knocked down condensed at various CaCl2 concentration. Flow cytometry analysis of RNAi efficiency on GFP gene in GFP HeLa cells at 48 h after treatment with the Hph1–Hph1–dsRBD/siRNAGFP complex at ratios of 1:1 or 5:1, respectively, which were condensed in the different concentration of CaCl2 (0, 10, 23.1, 34.6, 69.2 and 150 mM), or with Lipofectamine plus siGFP (positive control). Negative control was the negative siRNA with Hph1–Hph1–dsRBD (P/siRNA 2/1) as indicated. (B) Flow cytometry analysis of the percentage of GFP knocked down at various concentration of siRNAGFP (100, 50, 25 and 12.5 nM) prepared with different ratios of protein/siGFP (1:1, 1:5 and 1:10), which were condensed in the 10 mM (B) or 23.1 mM (C) CaCl2, respectivly. (D) Calcium condensed Hph1–Hph1–dsRBD/siRNA complexes affect the efficient of GAPDH gene expression. Quantitative RT-PCR analysis of endogenous GAPDH mRNA expression in HeLa cells at 24 h after treatment of the Hph1–Hph1–dsRBD with GAPDH siRNA or with negative siRNA (p/siRNA 1/1 or 5/1), which were condensed in 23.1 mM CaCl2. The GAPDH siRNA attached with Lipofectamine RNAiMAX was designed as the positive control. Error bars represent standard deviation (n = 3). Asterisk (**) indicates p < 0.01 based on analysis of one sample t-test and one way ANOVA post hoc multiple comparisons (LSD as the variances assumed).
Hph1–Hph1–dsRBD/siRNA complexes achieve better gene silencing effects than the commercially available siRNA delivery vectors, LipofectamineTM RNAiMAX. 3.5. Cytotoxicity of Hph1–Hph1–dsRBD Cell proliferation was not significantly suppressed in the medium containing Hph1–Hph1–dsRBD compared to PBS control (Fig. 6). Furthermore, cell viability studies demonstrated that Hph1–Hph1–dsRBD exhibited significantly low cytotoxicity. Similar results were obtained with Lipofectamine RNAiMAX. These data demonstrate that the Hph1–Hph1–dsRBD carrier was associated with low cytotoxicity. 3.6. Application of Hph1–Hph1–dsRBD in vivo
Fig. 6. Cell viability of Hph1–Hph1–dsRBD. Cell viability was analyzed using CCK8 Kit by measuring the absorbance in each well at OD 450. Error bars represent standard deviation (n = 3). “0.1, 1.5 and 4.5 mM Ca2+ condensed” was 0.1, 1.5 and 4.5 mM of Hph1–Hph1–dsRBD was condensed in 23.1 mM Ca2+ solution; 0.1, 1.5, 4.5, 6.0, 9.0 mM was respectively treated with 0.1, 1.5, 4.5, 6.0, 9.0 mM of Hph1–Hph1– dsRBD. The PBS and the lipofectamine was the control group.
The heart transplantation model is shown in Fig. 7A. Images revealed that FAM-siRNA bound with Hph1–Hph1–dsRBD entered the heart and was distributed throughout (Fig. 7B(3)). Similar observations were made with CPP–EGFP (enhanced green fluorescent protein) (Fig. 7B(2)). However, no significant fluorescent signals were detected with naked FAM-siRNA, which had not been
212
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
Fig. 7. (A) Administration Hph1–Hph1–dsRBD/siRNA-FAM. The heart from mouse was perfused using 10 ml PBS, and then the Hph1–Hph1–dsRBD/siRNA was injected into the heart and incubated PBS on ice for 30 mins. (B) The siRNA distribution in the mouse heart, 1, naked siRNA labeled FAM (green); 2, mouse heart transferred with CPP–EGFP; 3, mouse heart transferred with Hph1–Hph1–dsRBD/FAM-siRNA (green). Each sample was triplicated and only one of them was shown here. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
conjugated with Hph1–Hph1–dsRBD (Fig. 7B(1)). These results confirm the ability of Hph1–Hph1–dsRBD to transfer and distribute siRNA throughout the heart. 4. Discussion CPPs are capable of transferring various substrates into target cells. However, direct transfer of siRNA with CPP results in charge neutralization of CPP, aggregation/precipitation of the CPP/siRNA complex, which limits the application of CPP in siRNA delivery (Eguchi et al., 2009; Palm-apergi et al., 2011). In this study, we designed a novel fusion protein (Hph1–Hph1–dsRBD) to act as a siRNA carrier. The results demonstrated that this novel fusion protein was capable of delivering siRNA in vitro and in vivo with high efficiency. Although Hph1–Hph1–dsRBD completely binds to siRNA at ratio of 16:1, a slightly higher preparation ratio of at least 20:1 is necessary for effective siRNA delivery into cells and to achieve a desirable silencing effect. One possible reason for this phenomenon may be that two positive CPPs (Hph1) together with the dsRBD in the Hph1–Hph1–dsRBD complex could themselves bind to siRNA. This undesired binding of Hph1 with siRNA may lead to charge neutralization and invalidity of Hph1 (Eguchi et al., 2009). Notably, Hph1–Hph1–dsRBD was more efficient than naked CPP as a siRNA carrier. Compared with CPP noncovalently bonded with siRNA, Hph1–Hph1–dsRBD demands much less consumption of protein to achieve similar silencing effects, e.g., CPP of MPG (GALFLGFLGAAGSTMG) at ratio of 84:1 (Simeoni et al., 2003), and oligoarginine at ratio of 56:1 (Veldhoen et al., 2006). Additionally, Hph1–Hph1–dsRBD showed higher efficiency than CPP as a siRNA carrier. For example, the Hph1–Hph1–dsRBD/siRNA complex could inhibit up to a maximum 80% of the target gene expression at
100 nM concentrations of endogenous (GAPDH) or exogenous gene (GFP) siRNA, however, siRNA delivery by CPP of MPG, penetratin (RQIKIWFQNRRMKWKK), TP10 (AGYKKGKINLKALAALAKKIL), or EB1 (LIRLWSHLIHIWFQNRRLKWKKK) achieved only 60% (Simeoni et al., 2003), 0%, 18%, and 55% (Lundberg et al., 2007) at a final concentration of 100 nM siGFP, respectively, and oligoarginine reached 43% inhibition at double the concentration of siRNA (Veldhoen et al., 2006). In contrast to CPP covalently linked with siRNA, such as siRNA delivery by Tat (YGRKKRRQRRR) on HeLa cells, the carrier Hph1–Hph1–dsRBD demonstrated superior advantages. Not only is its preparation simple, but it also has a high transfection rate (80% knocked down at 100 nM Hph1–Hph1– dsRBD–siRNA compared with 70% at 200 nM siRNA) (Chiu et al., 2004). Although there are some reports that siRNA delivered by the CPP of penetratin generates RNAi efficiency of 85% in neuron cells (Davidson et al., 2004), others report much lower target gene silencing effects in the region of 20% (Moschos et al., 2007) and 53% (Muratovska and Eccles, 2004). In summary, Hph1–Hph1–dsRBD demonstrated considerably better transfection performance compared with other covalently linked or noncovalently linked siRNA carriers. We also investigated the effects of calcium condensation on the siRNA delivery efficiency of the Hph1–Hph1–dsRBD/siRNA complex according to established methods (Baoum et al., 2012). There was evidence that the efficiency of Hph1–Hph1–dsRBD delivery of siRNA positively correlated with calcium concentrations at siRNA conditions of 12.5 nM. The highest RNAi effect was reached when the Hph1–Hph1–dsRBD/siRNA complex was condensed with 69.2 mM CaCl2. However, similar results were observed when condensed at 23.1 mM CaCl2 and 25 nM siGFP, in which almost complete suppression of GFP genes was observed. Since complexes condensed in 10 mM CaCl2 did not demonstrate significant gene
H. Li et al. / International Journal of Pharmaceutics 469 (2014) 206–213
knock down, we can conclude that a certain amount of calcium might compensate or reinforce the function of protein–siRNA complexes. The mechanism of action of calcium can be elucidated from previously published data. The first hypothesis is that the complex of protein–siRNA could form nanoparticles when condensed in concentrations of CaCl2 above 23.1 mM (Baoum et al., 2012; Khondee et al., 2011). Furthermore, calcium could also disrupt the endosomes (Hoyer and Neundorf, 2012; Liu et al., 2012), in addition, calcium part plays an important role in CPP based cell penetrating (Lorents et al., 2012; Nascimento et al., 2012; PalmApergi et al., 2009). In vivo application, the FAM labeled siRNA delivery propelled by Hph1–Hph1–dsRBD could enter and distribute in the heart tissue, and no significant signal was observed in the control group without Hph1–Hph1–dsRBD. This result demonstrated that Hph1– Hph1–dsRBD could successfully transfer siRNA to the heart graft in vivo. To sum up, Hph1–Hph1–dsRBD as siRNA carrier presented very promising efficiency both in vitro and in vivo. Conflict of interest There is no conflict of interest. Acknowledgment This work was was partly supported by funding from Deutsche Forschungsgemeinschaft (TS175/1-1). Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j. ijpharm.2014.04.050. Reference Baoum, A., Ovcharenko, D., Berkland, C., 2012. Calcium condensed cell penetrating peptide complexes offer highly efficient, low toxicity gene silencing. International Journal of Pharmaceutics 427, 134–142. Chiu, Y.-L., Ali, A., Chu, C.-Y., Cao, H., Rana, T.M., 2004. Visualizing a correlation between siRNA localization, cellular uptake, and RNAi in living cells. Chemistry and Biology 11, 1165–1175. Choi, J.-M., Ahn, M.-H., Chae, W.-J., Jung, Y.-G., Park, J.-C., Song, H.-M., Kim, Y.-E., Shin, J.-A., Park, C.-S., Park, J.-W., Park, T.-K., Lee, J.-H., Seo, B.-F., Kim, K.-D., Kim, E.-S., Lee, D.-H., Lee, S.-K., Lee, S.-K., 2006. Intranasal delivery of the cytoplasmic domain of CTLA-4 using a novel protein transduction domain prevents allergic inflammation. Nature Medicine 12, 574–579.
213
Choi, J.-M., Kim, S.-H., Shin, J.-H., Gibson, T., Yoon, B.-S., Lee, D.-H., Lee, S.-K., Bothwel l, A.L.M., Lim, J.-S., Lee, S.-K., 2008. Transduction of the cytoplasmic domain of CTLA-4 inhibits TcR-specific activation signals and prevents collagen-induced arthritis. Proceedings of the National Academy of Sciences of the United States of America 105, 19875–19880. Davidson, T.J., Harel, S., Arboleda, V.A., Prunell, G.F., Shelanski, M.L., Greene, L.A., Troy, C.M., 2004. Highly efficient small interfering RNA delivery to primary mammalian neurons induces microRNA-like effects before mRNA degradation. Journal of Neuroscience 24, 10040–10046. Eguchi, A., Meade, B.R., Chang, Y.-C., Fredrickson, C.T., Willert, K., Puri, N., Dowdy, S.F., 2009. Efficient siRNA delivery into primary cells by a peptide transduction domain-dsRNA binding domain fusion protein. Nature Biotechnology 27, 567– 571. Hoyer, J., Neundorf, I., 2012. Peptide vectors for the nonviral delivery of nucleic acids. Accounts of Chemical Research 45, 1048–1056. Li, H., Tsui, T.Y., 2014. Six-cell penetrating peptide based fusion protein for siRNA delivery. Drug Delivery Epub ahead of print. Khondee, S., Baoum, A., Siahaan, T.J., Berkland, C., 2011. Calcium condensed LABL– TAT complexes effectively target gene delivery to ICAM-1 expressing cells. Molecular Pharmaceutics 8, 788–798. Lindgren, M., Llbrink, H.M.P.A., Langel, U., 2000. Cell-penetrating peptides. Trends in Pharmacological Sciences 21, 99–103. Liu, B.R., Lin, M.-D., Chiang, H.-J., Lee, H.-J., 2012. Arginine-rich cell-penetrating peptides deliver gene into living human cells. Gene 505, 37–45. Lorents, A., Kodavali, P.K., Oskolkov, N., Langel, Ü., Hällbrink, M., Pooga, M., 2012. Cell-penetrating peptides split into two groups based on modulation of intracellular calcium concentration. Journal of Biological Chemistry 287, 16880– 16889. Lundberg, P., El-Andaloussi, S., Sütlü, T., Johansson, H., Langel, U., 2007. Delivery of short interfering RNA using endosomolytic cell-penetrating peptides. FASEB Journal 21, 2664–2671. Ma, J., Lau, C.K., Obed, A., Dada, A., Doenecke, A., Fan, S.T., Schlitt, H.J., Tsui, T.Y., 2009. A cell penetrating heme oxygenase protein protects heart graft against ischemia/reperfusion injury. Gene Therapy 16, 320–328. Moschos, S.A., Jones, S.W., Perry, M.M., Williams, A.E., Erjefalt, J.S., Turner, J.J., Barnes, P.J., Sproat, B.S., Gait, M.J., Lindsay, M.A., 2007. Lung delivery studies using siRNA conjugated to TAT (48–60) and penetratin reveal peptide induced reduction in gene expression and induction of innate immunity. Bioconjugate Chemistry 18, 1450–1459. Muratovska, A., Eccles, M.R., 2004. Conjugate for efficient delivery of short interfering RNA (siRNA) into mammalian cells. FEBS Letters 558, 63–68. Nascimento, F.D., Sancey, L., Pereira, A., Rome, C., Oliveira, V., Oliveira, E.B., Nader, H. B., Yamane, T., Kerkis, I., Tersariol, I.L.S., Coll, J.-L., Hayashi, M.A.F., 2012. The natural cell-penetrating peptide crotamine targets tumor tissue in vivo and triggers a lethal calcium-dependent pathway in cultured cells. Molecular Pharmaceutics 9, 211–221. Palm-Apergi, C., Lorents, A., Padari, K., Pooga, M., Hällbrink, M., 2009. The membrane repair response masks membrane disturbances caused by cellpenetrating peptide uptake. FASEB Journal 23, 214–223. Palm-apergi, C., Eguchi, A., Dowdy, S.F., 2011. PTD–DRBD siRNA Delivery. Methods in Molecular Biology 683, 339–347. Schmittgen, T.D., Livak, K.J., 2008. Analyzing real-time PCR data by the comparative CT method. Nature Protocols 3, 1101–1108. Simeoni, F., Morris, M.C., Heitz, F., Divita, G., 2003. Insight into the mechanism of the peptide-based gene delivery system MPG: implications for delivery of siRNA into mammalian cells. Nucleic Acids Research 31, 2717–2724. Veldhoen, S., Laufer, S.D., Trampe, A., Restle, T., 2006. Cellular delivery of small interfering RNA by a non-covalently attached cell-penetrating peptide: quantitative analysis of uptake and biological effect. Nucleic Acids Research 34, 6561–6573.
