Effect of PEG anchor in PEGylation of folate-modified cationic liposomes with PEG-derivatives on systemic siRNA delivery into the Tumor

Abstract In this study, we prepared small interfering RNA (siRNA)/cationic liposome complexes (lipoplexes) modified with folate (FA)-polyethylene glycol (PEG, MW 2000, 3400 or 5000)-1,2-distearoyl-sn-glycero-3-phosphoethanolamine (DSPE) to facilitate their uptake into tumor cells via folate receptor (FR), and with PEG1600-cholesterol (PEG1600-Chol) or PEG2000-chondroitin sulfate conjugate (PEG2000-CS), to enhance their systemic stability. Among the FA-PEG-modified siRNA lipoplexes, 0.5 mol% FA-PEG5000-DSPE-modified lipoplexes with 2.5 mol% PEG2000-CS or PEG1600-Chol (LP-0.5F5/2.5P2-CS and LP-0.5F5/2.5P1.6-CL, respectively) exhibited selective growth inhibition of human nasopharyngeal carcinoma KB cells through transduction with polo-like kinase 1 (PLK1) siRNA. Furthermore, the LP-0.5F5/2.5P2-CS and LP-0.5F5/2.5P1.6-CL lipoplexes exhibited decreased agglutination with erythrocytes through PEGylation, and markedly decreased the accumulation of siRNA in murine lungs after systemic injection. Finally, systemic injection of LP-0.5F5/2.5P2-CS and LP-0.5F5/2.5P1.6-CL lipoplexes resulted in accumulation of siRNA in KB tumor xenografts. These findings suggest that PEGylation of FA-PEG5000-DSPE-modified siRNA lipoplexes with PEG2000-CS or PEG1600-Chol might improve their systemic stability without the loss of selective transfection activity in tumor cells.


Introduction
Synthetic small interfering RNA (siRNA) therapy has great potential for use in the treatment of tumors [1], because it can knock down target genes by cleaving the target mRNA using a sequence complementary siRNA [2]. The strategy of siRNA therapy is to treat the tumor by silencing specific tumor-promoting genes or anti-apoptotic genes using rationally designed siRNAs. For example, polo-like kinase 1 (PLK1) is essential for regulating the cell division, and its overexpression is found in many types of human tumors [3,4]. The knockdown of PLK1 expression by transfection with PLK1 siRNA can inhibit the proliferation of tumor cells [5,6]; therefore, PLK1 is one of potential targets for siRNA therapeutics to tumor. Cationic liposomes have been well studied as carriers for effective delivery of siRNAs into tumor cells [7,8]. However, for clinical applications, siRNAs must be delivered more selectively into tumor cells by cationic liposomes [9].
Overexpression of folate receptor (FR) has been observed in epithelial tumors such as ovary, cervical, colorectal, brain, prostate, and nasopharyngeal tumors [10][11][12]; and folic acid (FA) has been utilized as an attractive ligand for tumor targeting of siRNAs by cationic liposomes [13][14][15][16][17], because FA or its conjugates are taken up by tumor cells via receptor-mediated endocytosis after binding to FR [17]. Therefore, modification of cationic liposomes with FA-polyethylene glycol (PEG)-lipid has been used to facilitate the tumor uptake of siRNA lipoplexes.
For systemic siRNA delivery with cationic liposomes into tumors, siRNA/cationic liposome complexes (siRNA lipoplexes) must be stabilized in the systemic circulation by preventing their interaction with blood components such as erythrocytes, because the agglutinates formed by interaction between siRNA lipoplexes and blood components get entrapped in the highly extended lung capillaries [18,19], resulting in low accumulation of siRNA in the tumor. Therefore, cationic liposomes are often PEG-modified (PEGylated) for systemic siRNA delivery, for which a PEG-lipid derivative is incorporated into the membrane of cationic liposome via a lipid-anchor. For the PEGylation of cationic liposomes, a commonly used lipid-derivative of PEG is PEG-1,2-distearoylsn-glycero-3-phosphoethanolamine (PEG-DSPE) [20]. The presence of PEG on the surface of siRNA lipoplexes can prolong retention in the blood circulation by preventing siRNA lipoplexes from interacting with blood components and avoiding macrophage capture [21]. However, PEGylation of cationic liposomes with PEG-DSPE decreases gene knockdown by siRNA lipoplexes, as the amount of PEG-DSPE increases in the formulation of cationic liposomes [20], a phenomenon generally known as the PEG dilemma [21,22]. Conjugation of folate to the distal terminal of the PEG chain of PEG-DSPE can restore cellular association with FR-positive tumors; however, increasing the amount of FA-PEG-DSPE in the formulation of cationic liposomes also strongly reduces cellular association with tumor cells by steric hindrance of PEG on the surface of FA-PEG-modified siRNA lipoplexes [17]. Therefore, for the development of FR-targeting siRNA lipoplexes without the loss of gene knockdown activity caused by FA-PEG-modification, an optimal amount of FA-PEG-lipid must be included in the cationic liposomal formulation.
Generally, inclusion of PEG-lipids with longer and saturated diacyl chains into liposomal formulation can increase the plasma concentration of siRNA lipoplexes after systemic injection [23]. However, stable PEGylation of siRNA lipoplexes with PEG-DSPE strongly inhibits cellular uptake and/or endosomal escape, resulting in a significant loss of gene-knockdown activity [20]. Previously, we reported that PEGylation of siRNA lipoplexes with PEG-cholesterol (PEG-Chol) or PEG-chondroitin sulfate conjugate (PEG-CS) could improve systemic stability without loss of transfection activity by PEGylation [20]. We speculated that PEG-Chol and PEG-CS might be gradually released from the PEGylated siRNA lipoplexes, temporally stabilizing siRNA lipoplexes in the blood circulation. These findings suggested that the use of PEG-Chol and PEG-CS for PEGylation of FA-PEG-modified siRNA lipoplexes might improve systemic stability without reducing gene knockdown activity.
In this study, we prepared siRNA lipoplexes modified with FA-PEG-DSPE to facilitate uptake by tumor cells via FR, and with PEG-Chol or PEG-CS to enhance the systemic stability. Thus, FA-PEG-and PEG-modified siRNA lipoplexes were designed to gradually release PEG-derivatives after systemic injection, so that they could be efficiently taken up by tumor cells via FR without steric hindrance of PEG. We examined whether PEGylation of FA-PEG-DSPE-modified siRNA lipoplexes with PEG-Chol and PEG-CS improved systemic stability without loss of transfection activity in tumor tissues.
DDAB, DOPE, FA-PEG-DSPE and PEG-lipid derivative were dissolved in chloroform, and chloroform was evaporated under vacuum on a rotary evaporator at 60 °C to obtain a thin film. The thin film was hydrated with sterilized water at 60 °C by vortexing, followed by sonication in a bath-type sonicator to produce liposomes with a final size of approximately 100 nm.
To prepare siRNA lipoplexes, each FA-PEG-and PEG-modified cationic liposome suspension was mixed with siRNA at a charge ratio (+:−) of 4:1, by vortexing for 10 s, and left at room temperature for 15 min. The theoretical charge ratio (+:−) was defined as the molar ratio of nitrogen of cationic lipids in cationic liposomes to siRNA phosphate.
For preparation of FA-PEG-DSPE-and PEG 2000 -CS-modified siRNA lipoplexes, the 0.5 mol% FA-PEG-DSPE-modified cationic liposome suspension was mixed with siRNA at a charge ratio (+:−) of 4:1 by vortexing for 10 s, and left for 15 min at room temperature. To prepare ternar y complexes with PEG 2000 -CS, the FA-PEG-DSPE-modified siRNA lipoplexes were mixed with PEG 2000 -CS to be 1.5 or 2.5 mol% PEGylated.

Sizes and ζ-potentials of FA-PEG-and PEG-modified cationic liposomes and siRNA lipoplexes
The particle size distributions and ζ-potentials of FA-PEG-and PEG-modified cationic liposomes and the respective siRNA lipoplexes were measured by light-scattering photometry (ELS-Z2, Otsuka Electronics Co., Ltd., Osaka, Japan) as previously reported [20].

Cell culture
Human nasopharyngeal tumor KB cells (a subline of human cervix adenocarcinoma HeLa) were obtained from the Cell Resource Center for Biomedical Research, Tohoku university (Miyagi, Japan). using STR DNA profile analysis, it was confirmed that the KB cells used in this study were identical to the KB and HeLa cells registered in the database of Japanese Collection of Research Bioresources Cell Bank (JCRB, National Institute of Biomedical Innovation, Japan). KB cells stably expressing firefly pGL4 luciferase (KB-Luc cells) were established as previously reported [17]. KB and KB-Luc cells were cultured in RPMI-1640 medium with 10% heat-inactivated fetal bovine serum (FBS) and 100 μg/mL kanamycin, in humidified atmosphere containing 5% CO 2 at 37 °C.

Gene knockdown by FA-PEG-and PEG-modified siRNA lipoplexes
KB-Luc cells were seeded in 6-well culture plate at a density of 3 × 10 5 cells per well, 24 h prior to transfection. FA-PEG-and PEG-modified siRNA lipoplexes with 50 pmol Luc siRNA were diluted in 1 ml of folate-deficient RPMI-1640 medium supplemented with 10% FBS, and the mixture was then added to the cells (50 nM siRNA final concentration). Forty-eight hours after the transfection, luciferase activity was measured as counts per second (cps)/μg protein using PicaGene MelioraStar-LT Luminescence Reagent (Toyo Ink Co. Ltd., Tokyo, Japan) and BCA reagent (Pierce™ BCA Protein Assay Kit, Thermo Fisher Scientific, Rockford, IL, uSA) as previously reported [20]. Luciferase activity (%) was calculated as relative to the luciferase activity (cps/μg protein) of untransfected cells.

Agglutination assay
Erythrocyte suspension was prepared from the whole blood of female BALB/c mice (8 weeks of age; Sankyo Labo Service Corp. Inc., Tokyo, Japan) as previously reported [20]. FA-PEG-modified siRNA lipoplexes or FA-PEG-and PEG-modified siRNA lipoplexes with 2 μg siRNA were added to 100 μL of 2% (v/v) erythrocyte suspension. After incubation for 15 min at 37 °C, the sample was placed on a glass plate and agglutination was observed by microscopy.

Biodistribution of siRNA after systemic injection of FA-PEGand PEG-modified siRNA lipoplexes into mice
Female BALB/c mice (8 weeks of age) and female BALB/c nu/nu mice (8 weeks of age, CLEA Japan, Inc., Tokyo, Japan) were housed in a temperature (24 °C) and humidity (55%) controlled room with a 12 h light/dark cycle (lights on at 8:00 a.m.) with ad libitum access to food and water. The mice were maintained on a folate-deficient rodent diet (Oriental Yeast Co., Ltd., Tokyo, Japan) for the duration of the study. The non-PEGylated or FA-PEG-and PEG-modified siRNA lipoplexes with 20 μg Cy5-siRNA were administered intravenously via the lateral tail vein into BALB/c mice. One hour after injection, the mice were sacrificed, and the tissues were frozen on dry ice and sliced into 16-μm sections. The localization of Cy5-siRNA was examined using an Eclipse TS100-F microscope (Nikon, Tokyo, Japan).
For investigation of siRNA accumulation in tumor, we generated KB tumor xenografts by subcutaneous inoculation of 1 × 10 7 KB cells suspended in 50 μL of phosphate-buffered saline (PBS, pH 7.4) in the flank region of BALB/c nu/nu mice. The tumor volume was calculated using the following formula: tumor volume = 0.5 × a × b 2 , where a and b are the larger and smaller diameters, respectively. When the average volume of the tumors reached 200-300 mm 3 , FA-PEG-and PEG-modified siRNA lipoplexes with 20 μg Cy5-siRNA were systemically injected into the mice bearing KB xenografts. Twenty-four hours after injection, Cy5-fluorescence imaging of the tumors was performed using a NightOWL LB981 NC100 system (Berthold Technologies, Bad Wildbad, Germany) as previously reported [20]. The tumors were frozen on dry ice and sliced into 16-μm sections. The localization of Cy5-siRNA was examined using an Eclipse TS100-F microscope.

Statistical analysis
The statistical significance of differences between means was determined by Student t-tests, or analysis of variance followed by one-way analysis of variance on ranks with post-hoc Tukey tests, using GraphPad Prism 4.0 (GraphPad Software Inc., San Diego, CA, uSA). Statistical significance was set at p < .05.

Characterization of FA-PEG-modified cationic liposomes and siRNA lipoplexes
We used DDAB as a cationic lipid and DOPE as a neutral helper lipid ( Figure 1) for the preparation of cationic liposomes, because this formulation has exhibited high gene knockdown activity in vitro and in vivo in previous studies [25]. For FA-PEG-modification of cationic liposomes, FA-PEG-DSPE was incorporated into the liposomal membrane via the DSPE anchor. First, to investigate the effect of the amount of FA-PEG-DSPE and PEG-length between FA and DSPE in FA-PEG-modified siRNA lipoplexes on FR-targeting, we prepared FA-PEG-modified cationic liposomes with 0.5, 1, 2 or 3 mol% FA-PEG 2000 -DSPE, FA-PEG 3400 -DSPE and FA-PEG 5000 -DSPE ( Figure 2 and Table 1). As non-targeted controls for FA-PEGmodified cationic liposomes, 0.5, 1, 2, or 3 mol% PEG 2000 -DSPE, PEG 3400 -DSPE and PEG 5000 -DSPE, were included into the cationic liposome formulation. The sizes of PEG-modified or FA-PEGmodified cationic liposomes were approximately 100-150 nm, and the ζ-potentials were approximately +36-56 mV (Table 3). When the liposomes were mixed with siRNA, the lipoplex sizes were approximately 120-230 nm and their ζ-potentials were approximately +31-48 mV. Next, we examined the association of siRNA with FA-PEG-or PEG-modified cationic liposomes using an exclusion assay with SYBR® Green I. SYBR® Green I fluorescence markedly decreased upon the addition of PEG-modified or FA-PEG-modified cationic liposomes into the siRNA solution ( Figure S1). The amount of unbound siRNA in the siRNA lipoplex suspension decreased with increasing charge ratio (+:−), and saturated at 10-20% unbound   siRNA above charge ratios (+:−) of 2:1. This result suggested that siRNAs were completely bound to FA-PEG-or PEG-modified cationic liposomes regardless of the amount of FA-PEG-DSPE added and the PEG-length between FA and DSPE.

Effects of PEG length and modification ratio of FA-PEG-DSPE in FA-PEG-modified PLK1 siRNA lipoplexes on in vitro tumor growth inhibition
KB cells are often used for the evaluation of FR-targeting carriers, because they overexpress FR [26]. PLK1 is a potential therapeutic target for treating tumors, because it is expressed in multiple types of human tumors [3], and the inhibition of its expression reduces tumor cell proliferation [5]. We previously reported that transfection with FA-PEG-modified PLK1 siRNA lipoplexes could specifically suppress the expression of PLK1 mRNA in KB cells, resulting in a decrease in proliferation [17]. In the present study,  Figure S2). These results indicated that the modification by FA-PEG-DSPE with longer PEG-length decreased the optimal amount of FA-PEG-DSPE needed for selective tumor growth inhibition by FA-PEG-modified lipoplexes with PLK1 siRNA. Therefore, in the formulation of FA-PEG-modified liposomes, the optimal amount of FA-PEG-DSPE for selective transfection via FR may be largely affected by the PEG length between FA and DSPE.

Interaction between erythrocytes and FA-PEG-modified siRNA lipoplexes
For selective delivery into tumor by systemic injection, aggregation of siRNA lipoplexes with blood components such as erythrocytes must be prevented. Therefore, we examined how the amount of FA-PEG on the surface of siRNA lipoplexes affected aggregation with erythrocytes. It was observed that regardless of the PEG length in PEG-DSPE and FA-PEG-DSPE, 0.5 mol% PEG-modified or FA-PEG-modified siRNA lipoplexes formed large agglutinates after mixing with erythrocytes ( Figure 4). However, modification of siRNA lipoplexes with above 2 mol% PEG 2000 -DSPE, FA-PEG 2000 -DSPE, PEG 3400 -DSPE and FA-PEG 3400 -DSPE, and above 1 mol% PEG 5000 -DSPE and FA-PEG 5000 -DSPE, could prevent agglutination with erythrocytes. These results indicated that inclusion of above 1-2 mol% FA-PEG-DSPE or PEG-DSPE into siRNA lipoplexes might be needed for preventing agglutination with erythrocytes.

Characterization of FA-PEG-and PEG-modified cationic liposomes and siRNA lipoplexes
Stable PEGylation of siRNA lipoplexes with FA-PEG-DSPE can retain FA via PEG-chain on the surface of the lipoplexes, and the inclusion of 0.5 mol% FA-PEG 5000 -DSPE into cationic liposomes was enough for the selective inhibition of tumor growth via FR by FA-PEG-modified siRNA lipoplexes ( Figure 3); however, additional PEG-modification was needed for preventing agglutination with erythrocytes ( Figure 4). We previously reported that 3 mol% PEGylation of siRNA lipoplexes with PEG-Chol or PEG-CS prevented agglutination with erythrocytes, and decreased the accumulation of siRNA in the lungs after systemic injection, without significantly inhibiting the gene knockdown activities of siRNA lipoplexes [20]. Therefore, for stabilizing FR-targeting siRNA lipoplexes in blood circulation without loss of FR-mediated transfection activity by PEGylation, we included 0.

Interaction between erythrocytes and FA-PEG-and PEGmodified siRNA lipoplexes
To examine the effect of PEGylation of 0.5 mol% FA-PEG-DSPE-modified siRNA lipoplexes with 2.5 mol% PEG-derivatives on the aggregation of siRNA lipoplexes with erythrocytes, FA-PEG-and PEG-modified siRNA lipoplexes were added to erythrocyte suspensions. All the FA-PEG-and PEG-modified siRNA lipoplexes resisted agglutination ( Figure 9) and exhibited low levels in hemolysis ( Figure S4). These results indicated that PEGylation of 0.5 mol% FA-PEG-DSPE-modified siRNA lipoplexes with 2.5 mol% PEG-derivatives prevented aggregation with erythrocytes regardless of the anchor type of the PEG-derivatives.

Biodistribution of siRNA after systemic injection of PEGylated siRNA lipoplexes
To investigate the effects of the PEGylation of 0.5 mol% FA-PEG-DSPE-modified siRNA lipoplexes with 2.5 mol% Table 5. Particle size and ζ-potential of fa-Peg-and Peg-CS-modified sirna lipoplexes.    PEG-derivatives on the biodistribution of siRNA, we systemically administrated FA-PEG-and PEG-modified lipoplexes with Cy5-labeled siRNA into mice. In case of 0.5 mol% FA-PEG-DSPE-modified siRNA lipoplexes, accumulation of the LP-0.5F 2 , LP-0.5F 3.4 , and LP-0.5F 5 lipoplexes in the lungs was comparable to LP lipoplexes ( Figure 10), indicating that the presence of 0.5 mol% FA-PEG on the surface of siRNA lipoplexes could not prevent the interaction of siRNA lipoplexes with blood components. In contrast, PEGylation of LP-0.5F 2 , LP-0.5F 3.4 , and LP-0.5F 5 lipoplexes with 2.5 mol% PEG 2000 -DSPE, PEG 2000 -DSG, PEG 1600 -Chol or PEG 2000 -CS, significantly decreased siRNA accumulation in the lungs, suggesting that PEGylation can stabilize FA-PEG-modified siRNA lipoplexes in the blood circulation.

Discussion
Several researchers have reported the application of FA-PEGmodified cationic liposomes for siRNA delivery into FR-expressing cells [13][14][15][16]. For siRNA delivery with FA-PEG-modified cationic liposomes, 1 mol% FA-PEG 2000 -DSPE, 1 mol% FA-PEG 5000 -DSPE, and 0.5 mol% FA-PEG 3350 -Chol were included in the liposomal formulations composed of dioctadecyldimethylammonium chloride (DODAC) [14], 3β-[N-(N' ,N'-dimethylaminoethane)-carbamoyl]cholesterol (DC-Chol) [13], and 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP) [15], respectively, as cationic lipids. Previously, we also reported that the inclusion of 2 mol% FA-PEG 2000 -DSPE into cationic liposomes composed of cholesteryl (3-((2-hydroxyethyl) amino)propyl)carbamate (HAPC-Chol) and DOPE could enhance in vitro siRNA delivery via FR into tumor cells; however, it did not enhance the inhibition of tumor growth in KB tumor xenografts after intratumoral injection of PLK1 siRNA lipoplexes, compared with that of 2 mol% PEG 2000 -DSPE [17], indicating that the formulations optimized by in vitro transfection were not necessarily correlated with the ones optimized by in vivo transfection. Therefore, the in vivo optimization of liposomal formulation for For systemic stabilization, increasing the amount of FA-PEG-DSPE in the liposomal formulation can increase siRNA accumulation in tumor after systemic injection of FA-PEG-modified siRNA lipoplexes; however, it significantly reduces cellular association with tumor cells by steric hindrance of PEG, resulting in a decrease of gene knockdown activity in tumor cells. Regarding the optimal concentration of FA-PEG-DSPE in the formulation of FA-PEG-modified cationic liposomes, LP-0.5F 5 lipoplexes with PLK1 siRNA selectively inhibited cell growth, indicating that the presence of 0.5 mol% FA on the surface of siRNA lipoplexes was enough for tumor-targeting via FR; however, LP-0.5F 2 and LP-0.5F 3.4 lipoplexes with PLK1 siRNA slightly enhanced the inhibition of cell growth compared to LP-0.5P 2 and LP-0.5P 3.4 lipoplexes, respectively (Figure 3). In contrast, PLK1 siRNA lipoplexes modified with 2-3 mol% FA-PEG 2000 -DSPE or 1-2 mol% FA-PEG 3400 -DSPE selectively inhibited cell growth. These results are in agreement with previous reports that a low level of FA-PEG-modification with a sufficiently long PEG chain in FA-PEGmodified liposomes exhibited the highest FA-mediated association with the cells [27,28]. We speculated that in LP-0.5F 2 and LP-0.5F 3.4 lipoplexes, FA-PEG on the surface of siRNA lipoplexes could not reduce charge-medicated siRNA transfer owing to the shorter length of PEG; therefore, large amount of FA-PEG 2000 -DSPE and FA-PEG 3400 -DSPE might be needed for selective association between FA-PEG-modified siRNA lipoplexes and FR.
contrast, in case of PEGylation with 2.5 mol% PEG 2000 -DSPE or PEG 2000 -DSG, all the lipoplexes showed abolition of gene knockdown activity by Luc siRNA (Figure 8) and cell growth inhibition by PLK1 siRNA (Figure 7). These results indicated that stable PEGylation of FA-PEG-DSPE-modified siRNA lipoplexes with PEG 2000 -DSPE or PEG 2000 -DSG might inhibit cellular association by steric hindrance of PEG; however, PEG 1600 -Chol and PEG 2000 -CS were gradually released from FA-PEG-modified siRNA lipoplexes during the transfection, preventing the inhibition of cellular uptake and transfection activity. Regarding the PEG length of FA-PEG-DSPE, the inclusion of PEG 1600 -Chol or PEG 2000 -CS into FA-PEG 5000modified siRNA lipoplexes resulted in high selectivity for FR-mediated inhibition of tumor growth, compared with that into FA-PEG 2000 -or FA-PEG 3400 -modified siRNA lipoplexes (Figure 7). For strong association between FA and FR, FA must be situated on the distal terminal of PEG to stay outside the hydration layer of PEG in PEG-modified siRNA lipoplexes ( Figure 5).
In our preliminary study, we evaluated the anti-tumor effect by systemic injection of LP-0.5F 5 /2.5P 2 -CS or LP-0.5F 5 /2.5P 1.6 -CL lipoplexes with PLK1 siRNA on KB tumor xenografts ( Figure S5). Injection of LP-0.5F 5 /2.5P 2 -CS or LP-0.5F 5 /2.5P 1.6 -CL lipoplexes with Cont siRNA or PLK1 siRNA was performed three times, with a 2 day interval between each injection. Systemic injections of LP-0.5F 5 /2.5P 2 -CS lipoplexes with PLK1 siRNA inhibited tumor growth at days 4-8 compared with the Cont siRNA, although the difference was not significant. In contrast, systemic injections of LP-0.5F 5 /2.5P 1.6 -CL lipoplexes with PLK1 siRNA did not inhibit tumor growth at the day 8 compared with those with the Cont siRNA. LP-0.5F 5 /2.5P 2 -CS lipoplexes had negative charge in ζ-potentials, but LP-0.5F 5 /2.5P 1.6 -CL lipoplexes had positive charge, indicating that the lipoplexes with negative charge might be able to associate selectively with FR on the surface of tumor cells after accumulation in tumor. However, in future study, we will need to measure PLK1 mRNA level in tumor after systemic injection of LP-0.5F 5 /2.5P 2 -CS lipoplexes with PLK1 siRNA. Among FR-targeted cationic liposomes, LP-0.5F 5 /2.5P 2 -CS and LP-0.5F 5 /2.5P 1.6 -CL may have potential as a vector for siRNA delivery to the tumor by systemic injection.

Conclusions
In this study, we assessed the effect of PEG length and modification ratio in FA-PEG-modified cationic liposomes on the inhibition of tumor growth by PLK1 siRNA lipoplexes. Among the FA-PEG-and PEG-modified siRNA lipoplexes, 0.5 mol% FA-PEG 5000 -DSPE-modified siRNA lipoplexes with 2.5 mol% PEG 2000 -CS or PEG 1600 -CL exhibited selective inhibition of in vitro tumor growth by PLK1 siRNA, and they increased siRNA accumulation in the tumor after systemic injection. These findings indicate that PEGylation of FA-PEG 5000 -DSPE-modified siRNA lipoplexes with PEG 2000 -CS or PEG 1600 -Chol may improve their systemic stability without the loss of selective transfection activity in tumor cells. The in vivo optimization of PEGylation of FA-PEGmodified siRNA lipoplexes might be necessary for successful in vivo siRNA delivery via FR.