*3.5. Association of FA-PEG-Modified Cationic Liposomes with siRNA*

Next, we evaluated the association of FA-PEG-modified cationic liposomes with siRNA. The association of siRNA with each cationic liposome was monitored by gel retardation electrophoresis (Figure 5). The migration pattern of siRNA in siRNA lipoplexes was changed when the siRNA was mixed with cationic liposomes at charge ratios (+:−) from 1:1 to 4:1, and the migration of siRNAs ceased gradually as the charge ratio (+:−) increased. In HAPC-Chol-based liposomes, no migration was observed beyond charge ratios (+:−) of 2:1 in LP-HAPC, LP-HAPC-1mol%PEG2000, and LP-HAPC-1mol%FA-PEG2000 lipoplexes, of 3:1 in LP-HAPC-2mol%PEG2000, LP-HAPC-2mol%FA-PEG2000, and LP-HAPC-3mol%PEG2000, and of 4:1 in LP-HAPC-3mol%FA-PEG2000, (Figure 5A).

This result suggested that the association of siRNA with the cationic liposomes was inhibited with increasing amounts of FA-PEG2000-DSPE or PEG2000-DSPE. In siRNA transfection, we used a charge ratio (+:−) of 7:1 for the preparation of siRNA lipoplexes; therefore, siRNAs were completely bound to LP-HAPC regardless the PEG- or FA-PEG-modification. In addition, in LP-HAPC-1mol%PEG5000 and LP-HAPC-1mol%FA-PEG5000 lipoplexes, no migration was observed beyond charge ratios (+:−) of 2:1 (Figure 5A), indicating that the decrease in gene silencing activity in LP-HAPC-1mol%FA-PEG5000 lipoplexes (Figure 2B) was not caused by a decrease in the association of cationic liposomes with siRNA by the long PEG chain. Furthermore, LP-OH- and LP-OH-C-based liposomes, beyond charge ratios (+:−) of 4:1 in LP-OH, LP-OH-1mol%PEG2000, LP-OH-1mol%FA-PEG2000, LP-OH-C, LP-OH-C-2mol%PEG2000, LP-OH-C-2mol%FA-PEG2000 lipoplexes, no migration or decreased

migration was observed (Figure 5B,C). These results indicated that OH-Chol- and OH-C-Chol-based liposomes might make a weaker association with siRNA than HAPC-Chol-based liposomes.

**Figure 5.** Effect of FA-PEG-modification of cationicliposomes on association of siRNAwith FA-PEG-modified cationic liposomes. siRNA association by cationic liposomes was analyzed by gel retardation assay. (**A**) LP-HAPC-1–3mol%PEG2000, LP-HAPC-1mol%PEG5000, LP-HAPC-1–3mol%FA-PEG2000, and LP-HAPC-1mol%FA-PEG5000, (**B**) LP-OH-1–3mol%PEG2000 and LP-OH-1–3mol%FA-PEG2000, (**C**) LP-OH-C-1–3mol%PEG2000 and LP-OH-C-1–3mol%FA-PEG2000 were used. Each liposome was formed with siRNA at various charge ratios (+:−) from 1:1 to 4:1, and were analyzed using 18% acrylamide gel electrophoresis.

Furthermore, we examined the association of siRNA with each cationic liposome using an exclusion assay with SYBR® Green I. SYBR® Green I is a DNA/RNA-intercalating agent whose fluorescence was dramatically enhanced upon binding to unbound siRNA in cationic lipoplexes. As a result, in all the cationic liposomes, the fluorescence of SYBR® Green I was markedly decreased by the addition of cationic liposomes into the siRNA solution beyond charge ratios (+:−) of 2:1 or 3:1, compared with that in siRNA solution (Figure 6A–C). This result suggested that siRNAs were completely bound to each cationic liposome regardless of PEG or FA-PEG modification of the cationic liposomes. Although a discrepancy between the results from the accessibility of SYBR® Green I (Figure 6) and gel retardation electrophoresis (Figure 5) was observed, siRNAs might be released from siRNA lipoplexes by electrophoresis due to the weak association between siRNA and the cationic liposomes. From the result shown in Figure 6, for all the cationic liposomes used in this study, siRNAs might be completely bound to the cationic liposomes when mixed beyond a charge ratio (+:−) of 3:1.

**Figure 6.** Effect of FA-PEG-modification of cationic liposomes on association of siRNA with FA-PEG-modified cationic liposome. siRNA association by cationic liposomes was analyzed by exclusion assay using an SYBR® Green I Nucleic Acid Gel Stain. (**A**) LP-HAPC-1–3mol%PEG2000, LP-HAPC-1mol%PEG5000, LP-HAPC-1–3mol%FA-PEG2000, and LP-HAPC-1mol%FA-PEG5000, (**B**) LP-OH-1–3mol%PEG2000 and LP-OH-1–3mol%FA-PEG2000, (**C**) LP-OH-C-1–3mol%PEG2000 and LP-OH-C-1–3mol%FA-PEG2000 were used. The siRNA lipoplexes were formed at various charge ratios (+:−) from 1:1 to 4:1. As a control, the value of fluorescence obtained upon addition of free siRNA solution was set as 100%. The amount of siRNA available to interact with the SYBR® Green I was expressed as a percentage of the control. Each column represents the mean + SD (*n* = 3).

#### *3.6. Suppression of EGFP Expression by FA-PEG-Modified siRNA Lipoplexes*

To examine the effect of FA-PEG-modification in cationic liposomes on gene knockdown using FA-PEG-modified siRNA lipoplexes, KB-EGFP cells were incubated with siRNA lipoplexes, and then the gene silencing effect was assessed by assaying the fluorescence intensity in the cells. Here, we decided to use LP-HAPC-2mol%FA-PEG2000, LP-OH-1mol%FA-PEG2000, and LP-OH-C-2mol%FA-PEG2000 as optimal FA-PEG-modified liposomes. In addition, we used Lipofectamine RNAiMax as a commercially available in vitro transfection reagent for siRNAs.

LP-HAPC lipoplexes with EGFP siRNA strongly suppressed EGFP expression (~50% knockdown, compared with Cont siRNA); however, LP-OH and LP-OH-C lipoplexes with EGFP siRNA were suppressed moderately (20–30% knockdown, compared with Cont siRNA) (Figure 7). In contrast, LP-HAPC-2mol%PEG2000, LP-OH-1mol%PEG2000, and LP-OH-C-2mol%PEG2000 lipoplexes did not exhibit suppression of EGFP expression in the cells. However, LP-HAPC-2mol%FA-PEG2000, LP-OH-1mol%FA-PEG2000, and LP-OH-C-2mol%FA-PEG2000 lipoplexes restored the gene silencing effect by FA-PEG-modification. Among the FA-PEG modified siRNA lipoplexes, LP-HAPC-2mol%FA-PEG2000 lipoplexes strongly suppressed the expression of EGFP in the cells (~40% knockdown, compared with Cont siRNA), similar to Lipofectamine RNAiMax. Although LP-HAPC-2mol%FA-PEG2000, LP-OH-1mol%FA-PEG2000, and LP-HAPC-2mol%FA-PEG2000 lipoplexes exhibited strong gene silencing effects in KB-Luc cells (Figure 2), they showed moderate gene silencing efficacy in KB-EGFP cells (Figure 7). It has been reported that the t1/<sup>2</sup> of firefly luciferase protein is about 2–3 h [33,34], but those of GFP and EGFP are more than 24 h [35], indicating that EGFP expression could not be suppressed completely by siRNA due to the long t1/<sup>2</sup> of EGFP. Among the FA-PEG-modified cationic liposomes, LP-HAPC-2mol%FA-PEG2000 appeared to be the most effective in FR-mediated gene silencing.

#### *3.7. Antiproliferative Activity*

PLK1 is a potential target in tumor therapy, because PLK1 is overexpressed in various types of human tumors [3], and its inhibition is potently antiproliferative for tumor cells. To examine whether transfection of PLK1 siRNA into KB cells with FA-PEG modified siRNA lipoplexes could selectivity inhibit tumor growth, we measured cell viability 48 h after transfection of PLK1 siRNA into KB cells. Transfection of LP-HAPC, LP-OH, and LP-OH-C lipoplexes with PLK1 siRNA largely inhibited cell growth; however, LP-HAPC-2mol%PEG2000, LP-OH-1mol%PEG2000, and LP-OH-C-2mol%PEG2000 lipoplexes with PLK1 siRNA showed a decreased cytotoxic effect with the PEG-modification (Figure 8). In contrast, LP-HAPC-2mol%FA-PEG2000, LP-OH-1mol%FA-PEG2000, and LP-OH-C-2mol%FA-PEG2000 lipoplexes with PLK1 siRNA strongly decreased cell proliferation, similar to Lipofectamine RNAiMax.

**Figure 7.** Effect of FA-PEG-modification of cationic liposomes on suppression of EGFP expression in KB-EGFP cells after transfection with FA-PEG-modified siRNA lipoplexes. As cationic liposomes, LP-HAPC, LP-HAPC-2mol%PEG2000, LP-HAPC-2mol%FA-PEG2000, LP-OH, LP-OH-1mol%PEG2000, LP-OH-1mol%FA-PEG2000, LP-OH-C, LP-OH-C-2mol%PEG2000, and LP-OH-C-2mol%FA-PEG2000 were used. The siRNA lipoplexes with Cont siRNA or EGFP siRNA were added to KB-Luc cells at 50 nM siRNA, and the EGFP expression levels were determined on the basis of EGFP-fluorescence by flow cytometry. The EGFP expression level (%) was calculated as relative to the fluorescence intensity of untransfected KB-EGFP cells. Each column represents the mean + SD (*n* = 3). Lipofectamine RNAiMax was used as a control. \*\* *p* < 0.01, compared with Cont siRNA.

**Figure 8.** Effect of FA-PEG-modification of cationic liposomes on antiproliferative activities 48 h after transfection with FA-PEG-modified PLK1 siRNA lipoplexes into KB cells. As cationic liposomes, LP-HAPC, LP-HAPC-2mol%PEG2000, LP-HAPC-2mol%FA-PEG2000, LP-OH, LP-OH-1mol%PEG2000, LP-OH-1mol%FA-PEG2000, LP-OH-C, LP-OH-C-2mol%PEG2000, and LP-OH-C-2mol%FA-PEG2000 were used. The siRNA lipoplexes with Cont siRNA or PLK1 siRNA were added to KB-Luc cells at 50 nM siRNA. At 48 h after transfection, cell viability was measured. Each column shows the mean + SD (*n* = 4). Lipofectamine RNAiMax was used as a control. \*\* *p* < 0.01, compared with Cont siRNA.

Next, to investigate whether these cytotoxic effects by transfection of PLK1 siRNA were induced by decreased expression level of PLK1 mRNAs, we measured PLK1 mRNA levels at 24 h after transfection

of PLK1 siRNA with each cationic liposome. LP-HAPC, LP-OH, and LP-OH-C lipoplexes with PLK1 siRNA significantly inhibited the expression of PLK1 mRNA, compared with Cont siRNA, which was similar to Lipofectamine RNAiMax with PLK1 siRNA (Figure 9). However, LP-HAPC-2mol%PEG2000, LP-OH-1mol%PEG2000, and LP-OH-C-2mol%PEG2000 lipoplexes with PLK1 siRNA did not greatly suppress the expression of PLK1 mRNA. In contrast, transfection of LP-HAPC-2mol%FA-PEG2000, LP-OH-1mol%FA-PEG2000, or LP-OH-C-2mol%FA-PEG2000 siRNA lipoplexes with PLK1 siRNA markedly decreased the expression of PLK1 mRNA in the cells. These results indicated that PLK1 siRNA transfected by FA-PEG-modified lipoplexes could specifically suppress the expression of PLK1 mRNA in the cells, and that suppression of PLK1 mRNA did affect the decrease in proliferation.

**Figure 9.** Effect of FA-PEG-modification of cationic liposomes on suppression of PLK1 mRNA expression by transfection with FA-PEG-modified PLK1 siRNA lipoplexes into KB cells. As cationic liposomes, LP-HAPC, LP-HAPC-2mol%PEG2000, LP-HAPC-2mol%FA-PEG2000, LP-OH, LP-OH-1mol%PEG2000, LP-OH-1mol%FA-PEG2000, LP-OH-C, LP-OH-C-2mol%PEG2000, and LP-OH-C-2mol%FA-PEG2000 were used. The siRNA lipoplexes with Cont siRNA or PLK1 siRNA were added to KB-Luc cells at 50 nM siRNA. At 24 h after transfection, the expression levels of PLK1 mRNA in the cells were analyzed by quantitative RT-PCR. Each result represents the mean + SD (*n* = 3). \* *p* < 0.05, \*\* *p* < 0.01, compared with Cont siRNA.
