From gene delivery to gene silencing: plasmid DNA-transfecting cationic lipid 1,3-dimyristoylamidopropane-2-[bis-(2-dimethylaminoethane)] carbamate efficiently promotes small interfering RNA-induced RNA interference

Abstract

The cationic lipid 1,3-dimyristoylamidopropane-2-[bis-(2-dimethylaminoethane)] carbamate (1,3lb2) was applied as a delivery system for small interfering RNA (siRNA) to inhibit the production of vascular endothelial growth factor (VEGF) in vitro in the human prostate carcinoma cell line PC-3. VEGF protein silencing peaked at 94% when cationic lipid-nucleic acid complexes (lipoplexes) were formulated at a nitrogen-to-phosphorothioate ratio (N/P) of 2 with a dose concentration of 53.7 nM, and the performance of these lipoplexes was not impeded by serum. Knockdown efficiency was maintained for at least 72 h and an IC₅₀ of 12 nM lasted for 48 h. Only 20% of the total siRNA became cell-associated at this N/P, at a rate of 25 ng h⁻¹. Lipoplexes of the optimum formulation were relatively monodisperse having an average diameter of 634 nm and a zeta potential of −21.3 mV. 1,3lb2-siRNA complexation reached 94% at N/P 2 and was positively cooperative; the binding constant was calculated in the range of 10⁵ M⁻¹ and a Hill coefficient of 3 was determined. 1,3lb2 was found to be a nontoxic and potent carrier of siRNA that binds to the nucleic acid effectively and whose lipoplexes promote long-lasting inhibition, have high biological activity at low N/Ps, and are functional in the presence of serum.

RNA interference (RNAi) was first described by the Nobel Prize-winning work of Fire and coworkers (1) more than a decade ago in the nematode Caenorhabditis elegans, followed three years later by its first description in mammalian cells (2). RNAi is an evolutionarily conserved defense mechanism against viruses and transposons (3). This naturally occurring process is triggered by duplex RNAs that induce enzymatic degradation of complementary mRNA, resulting in sequence-specific knockdown of gene expression.

Since its role in normal gene silencing was ascertained, RNAi quickly became a popular tool for functional genomics and has been used in numerous therapeutic applications for the down-regulation of disease-causing genes in ocular diseases, neurodegenerative disorders, viral infections, and cancers (4,5). Synthetic double-stranded small interfering RNAs (siRNAs) were developed to harness the RNAi pathway that is innately guided by endogenous molecules known as micro RNAs. One such siRNA targeting vascular endothelial growth factor (VEGF) was constructed as cancer therapeutics (6). VEGF is a secreted protein involved in angiogenesis and has been linked to tumor growth and metastasis (7–9).

When searching for potential siRNA transporters, it is only logical to first investigate those with proven efficacy in transfecting plasmid DNA (pDNA) and antisense oligonucleotides. Among these possibilities, cationic liposome/lipid-mediated siRNA cellular delivery (siFection) (10) holds tremendous promise. The cationic lipid 1,3-dimyristoylamidopropane-2-[bis-(2-dimethylaminoethane)] carbamate, henceforth referred to as 1,3lb2 (Scheme 1), is an established delivery system for pDNA; the biological activity and physicochemical properties of 1,3lb2, alone and in cationic lipid-nucleic acid complexes (lipoplexes) with pDNA, are detailed elsewhere (11,12). In this work, 1,3lb2 was studied as a carrier of siRNA for the suppression of VEGF production in vitro in the human prostate carcinoma cell line PC-3.

Scheme 1.

Molecular structure of 1,3lb2.

EXPERIMENTAL PROCEDURES

1. MATERIALS

VEGF-siRNA and scrambled (scr) siRNA sequences, the latter serving as a non-silencing control, were taken from Takei et al. (6) and synthesized by Qiagen (Valencia, CA USA), but modified with two thiolated strands to prevent hydrolysis of the ribonucleic acid (Scheme 2). Fluorescein isothiocyanate (FITC) -labeled siRNA was also from Qiagen, with the FITC tag covalently linked to the 5′-end of the sense strand. All siRNAs were purified by denatured ion-exchange high performance liquid chromatography (HPLC) and native reversed-phase HPLC, and the sequence and identity of each duplex were confirmed by matrix-assisted laser desorption/ionization time-of-flight mass spectrometry. The siRNAs were supplied as powders and stored at −20°C; prior to use they were reconstituted to 1 mg mL⁻¹ in RNase-free TE buffer pH 8. 1,3lb2 was synthesized and identified to purity (> 99%) as previously described (12). All other reagents and solvents were purchased from commercial vendors and used without further purification.

Scheme 2.

VEGF-siRNA and scr-siRNA sequences. The backbone phosphates were substituted for phosphorothioate groups, symbolized by ps.

2. METHODS

2.1. Cell culture

PC-3 cells (American Type Culture Collection, Manassas, VA USA) in F-12K Medium supplemented with 10% fetal bovine serum (FBS) were maintained at 37°C in a 5% CO₂ in air humidified atmosphere. The day before siFection, cells were seeded at the desired number in multiwell plates and left overnight to attach.

2.2. Cationic lipid dispersion and lipoplex preparation

A solution of 1,3lb2 in chloroform was dried under a stream of nitrogen gas followed by high vacuum desiccation. The dry lipid film was resuspended in 40 mM Tris pH 7.2, at elevated temperature with periodic vortexing, for a final concentration of 0.3 mM 1,3lb2. Lipoplexes were formed at various nitrogen-to-phosphorothioate ratios (N/Ps) in serum-free F-12K Medium (SFM) by pipetting an aliquot of siRNA solution into an appropriate dilution of lipid dispersion.

2.3. SiFection studies

2.3.1. Bioactivity

Lipoplexes were incubated with cells (300,000 cells well⁻¹, 12-well plate) for 3 h. Then lipoplexes were replaced with fresh serum medium and cells were incubated for the desired length of time. For serum studies, 250 μL lipoplexes were diluted within the wells in an equal volume of serum medium to give a final FBS concentration of 5%. A standard 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay was used to evaluate cytotoxicity.

VEGF protein was quantified from the cell media using an immunoassay kit (Invitrogen, Camarillo, CA USA) according to the manufacturer’s instructions. Inhibition of protein production was calculated by

$$\%\text{knockdown}=\frac{{[\text{VEGF}]}_{\mathit{untreated}}-{[\text{VEGF}]}_{\mathit{treated}}}{{[\text{VEGF}]}_{\mathit{untreated}}}\times 100$$ (1)

where [VEGF]_untreated and [VEGF]_treated are concentrations of protein from untreated cells and cells treated with lipoplexes, respectively.

VEGF mRNA was quantified from cell lysates by real-time reverse-transcription polymerase chain reaction (RT-PCR) using a 7300 Real-Time PCR System (Applied Biosystems, Foster City, CA), Qiagen’s QuantiTect SYBR Green RT-PCR Kit, and validated VEGF and β-actin (internal control) primer sets also from Qiagen. Total RNA was purified from cells using Qiagen’s RNeasy Plus Mini Kit. The real-time cycler conditions were as follows: 1 cycle of reverse transcription at 50°C for 15 min; PCR initialization at 95°C for 15 min; 40 cycles of denaturation at 95°C for 15 s; annealing at 60°C for 1 min; extension at 72°C for 30 s. Expression levels of β-actin were identical in untreated and treated cells (not shown). Inhibition of VEGF mRNA production was calculated by Eq. 1.

2.3.2. Cellular association

SiFection was carried out with lipoplexes composed of FITC-labeled siRNA (0.5 μg well⁻¹, 67.1 nM). Lipoplexes were removed at half-hour intervals and cells were lysed with a 0.1% Triton X-100 solution. Cell lysates were diluted in Tris buffer containing 0.1% SDS and analyzed for siRNA content at λ_ex = 495 nm using a Cary Eclipse Fluorescence Spectrophotometer (Varian Australia Pty Ltd, Victoria, Australia). 1,3lb2 caused a shift in the FITC-siRNA emission peak to a longer wavelength and a lower intensity. Upon addition of SDS the peak was restored to its original wavelength and intensity via decomplexation of the lipoplexes (Fig. S1). The peak fluorescence emission intensity of each sample was measured around 521 nm and converted to FITC-siRNA concentration from a standard curve (Fig. S2).

2.4. Particle size and zeta potential measurements

Particle sizes and zeta potentials of lipoplexes were measured with a Zetasizer Nano ZS System (Malvern Instruments Ltd, Worcestershire, UK) as described in Spelios and Savva (12).

2.5. Complexation studies

2.5.1. Fluorescent titration

In a quartz cuvette, ethidium bromide (EtBr) and siRNA (444 nM) were combined in SFM at a nucleotide-to-dye molar ratio of 44:1. At this ratio the EtBr fluorescence emission intensity fell within the initial steep linear region of the intensity versus [siRNA] plot (Fig. S3). Aliquots of lipid dispersion were added under constant stirring and the peak intensity I around 600 nm was monitored spectrofluorometrically at λ_ex = 510 nm. The extent of lipid-siRNA complexation was assessed from

$$\%\text{binding}=\left(1-\frac{I-I_{\mathit{blank}}}{I_0-I_{\mathit{blank}}}\right)\times 100$$ (2)

where I_blank is the intensity of an EtBr blank solution and I₀ is the intensity of EtBr in the presence of siRNA prior to the addition of lipid. The % binding (y) was plotted as a function of the N/P, and a sigmoidal equation of the form

$$y=A+\frac{B}{1+e^{C(D-x)}}$$ (3)

was fit to the plot with PSI-Plot (version 7.01, Poly Software International, Inc., Pearl River, NY USA) through minimization of the sum of the squared residuals with four adjustable parameters at a 5% tolerance level. A and B are y_max and y_max − y_min, respectively, C is a shape parameter, and D is the N/P corresponding to 50% binding.

➢ Hill and Scatchard analyses

Cationic lipid association with siRNA was studied using an infinite cooperativity model and an independent ligand model. For a siRNA molecule S with n binding sites of equal affinity exhibiting cooperative binding, the binding of lipid L to S can be described by the following equilibria where K_Ai is the respective association constant (1 ≤ i ≤ n):

$$L+S\leftrightarrow LS{K}_{A_1}$$ (4)

$$\begin{array}{l}L+LS\leftrightarrow L_{2}S{K}_{A_2}\\ ⋮\end{array}$$ (5)

$$L+L_{n-2}S\leftrightarrow L_{n-1}S{K}_{A_{n-1}}$$ (6)

$$L+L_{n-1}S\leftrightarrow L_{n}S{K}_{A_n}$$ (7)

The overall equation and the corresponding association constant are:

$$nL+S\leftrightarrow L_{n}SK=K_{A_1}\times K_{A_2}\cdots \times K_{A_{n-1}}\times K_{A_n}={K_A}^n$$ (8)

$${K_A}^n=\frac{[L_{n}S]}{{[L]}^n[S]}$$ (9)

The bound and unbound fractions F_B and F_U, respectively, of S are:

$$F_B=\frac{[L_{n}S]}{[S]+[L_{n}S]}$$ (10)

$$F_U=\frac{[S]}{[S]+[L_{n}S]}$$ (11)

Dividing Eq. 10 by Eq. 11 and substituting Eq. 9 for [L_nS] give:

$$\frac{F_B}{F_U}=\frac{[L_{n}S]}{[S]}={K_A}^n{[L]}^n$$ (12)

Taking the logarithm of Eq. 12 gives the Hill equation:

$$log\frac{F_B}{F_U}=nlog{K}_A+nlog[L]$$ (13)

Since F_B and F_U are ascertained from ΔI (I₀ − I ) and I, respectively, $log\frac{\mathrm{\Delta }I}{I}$ (blank-corrected intensities) was plotted against log [1,3lb2] to obtain the Hill plot from which n and K_A were determined by the slope and intercept, respectively. The free lipid concentration was determined as described by Boger and coworkers (13).

Noncooperative cationic lipid binding on siRNA was studied using the well-known Scatchard equation:

$$\frac{F_B}{[L]}=n·K_A-K_A·F_B$$ (14)

2.5.2. Agarose gel electrophoresis

Electrophoresis of lipoplexes was done on a 2% agarose gel with 0.2 μg well⁻¹ siRNA using a Horizon 58 Horizontal Gel Electrophoresis System (Life Technologies, Gibco BRL, Gaithersburg, MD USA). EtBr (0.5 μg mL⁻¹) was added to the gel and electrophoresis buffer (TAE, pH 7.4) for visualization of siRNA via a Gel Logic 200 Imaging System (Scientific Imaging Systems, Eastman Kodak Company, Rochester, NY USA). Band intensities were extracted with the Kodak 1D Image Analysis Software (version 3.6.3) and used to compute the % intensity decrease in each lane relative to the control (siRNA alone).

RESULTS AND DISCUSSION

1. SiFection studies

1.1. Bioactivity

Interaction between siRNA-containing lipoplexes and the cell surface is essential for transportation of the polyanionic duplex across the similarly negatively charged plasma membrane, and 1,3lb2 mediates this interaction very well. Fig. 1 shows the strong association between 1,3lb2-siRNA complexes and PC-3 cells. The amount of siRNA in the lipoplexes (N/P 2) was varied in order to examine the dose-dependence of gene silencing. Cell media were collected post-siFection at 24-h intervals up to 72 h for VEGF quantification by enzyme-linked immunosorbent assay (ELISA). Greater than 50% knockdown was observed at concentrations beyond 16.8 nM siRNA, and the level of knockdown at these concentrations was relatively stable over 3 days (Fig. 2A). Further examination revealed an optimum dose concentration of 53.7 nM (0.4 μg well⁻¹) above which the knockdown remained constant at 90% (Fig. 2B). The sequence specificity of siRNA-induced RNAi was verified by the absence of knockdown when scr-siRNA was used. No cytotoxicity was detected for any formulation.

Fig. 1.

Representative image of cell-associated 1,3lb2-siRNA complexes. SiRNA is seen by the fluorescence (green) of its FITC tag. The image was captured using an Axiovert 200M Inverted Microscope (Carl Zeiss, Göttingen, Germany).

Fig. 2.

(A, B) Knockdown of VEGF protein (as determined by ELISA) secreted into the media of PC-3 cells exposed to lipoplexes (N/P 2) at various doses. (B) ELISA and MTT cytotoxicity assay were performed 48 h post-siFection. Scr-siRNA served as a control. The results are expressed as the mean ± SD (n = 2).

The dose-response plots (Fig. 2A) were curve-fitted with PSI-Plot using

$$y=\frac{100}{1+{\left(\frac{{\text{IC}}_{50}}{x}\right)}^{\text{slope}}}$$ (15)

where y is the % knockdown of VEGF protein, x is the siRNA concentration (nM), IC₅₀ is the dose concentration producing exactly 50% knockdown, and the slope is a shape parameter. The IC₅₀ was found to be 12 nM at 24 and 48 h, almost doubling to a value of 20 nM at 72 h (Table 1). The slope also nearly doubled from 0.75 and 0.87 at 24 and 48 h, respectively, to 1.69 at 72 h. Additionally, the fits at 24 and 48 h were hyperbolic, whereas the fit at 72 h was sigmoidal (not shown). From the data analysis it appears that the inhibitory effect was reduced after 3 days. Most probably, at low doses the constituents of the RNAi pathway were incompletely activated and the cells recovered within the tested timeframe to resume normal VEGF production. Thus knockdown at these doses was not as pronounced 3 days post-siFection as it was after 1–2 days, resulting in a lag at early points in the dose-response plot at 72 h which shifted the IC₅₀ to a higher dose.

Table 1.

Curve-fit parameters of dose-response plots (Fig. 2A)

Time post- siFection (h)	IC50 (nM)	Slope	C.O.Da
24	12.2	0.75	0.919
48	12.3	0.87	0.883
72	20.3	1.7	0.989

^aCoefficient of determination

Subsequently, the N/P was optimized at the optimum dose. At N/P 0.5 there was no knockdown of VEGF protein, while lipoplexes at N/P 1 produced 86% knockdown (Fig. 3A). Knockdown reached a maximum of 94% at N/P 2 and was maintained at higher N/Ps. No significant knockdown was seen with naked siRNA or lipoplexes incorporating scr-siRNA (not shown).

Fig. 3.

Effect of the N/P on the % VEGF knockdown (A) under serum-free conditions and (B) in the presence of 5% FBS. Experiments were carried out with 0.4 μg well⁻¹ siRNA (53.7 nM). Samples were collected 48 h post-siFection. The results are expressed as the mean ± SD (n = 2).

To confirm that protein knockdown was due to targeted degradation of corresponding mRNA, real-time RT-PCR was performed on cell lysates to quantitate VEGF mRNA. As shown in Fig. 3A, the mRNA and its encoded protein were knocked down similarly.

The influence of serum on 1,3lb2-mediated siFection was also examined. Serum had no effect on VEGF protein knockdown except at N/P 1 where the knockdown dropped from 86% under serum-free conditions to 60% in the presence of 5% FBS (Fig. 3B).

1.2. Cellular association

The kinetics of cellular association of 1,3lb2-complexed siRNA during siFection were studied. At N/P 0.5 the amount of cell-associated siRNA did not vary with time after 0.5 h (Fig. 4 and Table 2). At N/P 1 there was a linear increase in the rate of cellular association (18 ng h⁻¹) from 0.5 until 1.5 h when the amount plateaued between 50 and 60 ng, twice the amount observed at N/P 0.5. It appears that at N/P 1 all excess free lipid not electrostatically complexed with siRNA was exhausted after 1.5 h. At N/P 0.5 depletion of excess lipid occurred much earlier (t ≤ 0.5 h), impeding knockdown. This key role of excess uncomplexed lipid in mediating cellular internalization of siRNA at the plasma membrane level is quite different from the well-documented role of excess cationic lipid in improving gene delivery by neutralizing extracellularly-interfering serum components (14,15). At N/P 2 cellular association increased linearly at a rate of 25 ng h⁻¹ between 0.5 and 3 h. Within the same time period the rate decreased two-fold as the N/P was doubled to 4. However, the initial rate (0 ≤ t ≤ 0.5 h) was higher at N/P 4 (91 versus 53, 70, and 60 ng h⁻¹ at N/P 0.5, 1, and 2, respectively).

Fig. 4.

Change in the amount of cell-associated siRNA (mean ± SD, n = 2) over time during siFection.

Table 2.

Kinetic parameters of cell association of siRNA in lipoplexes at various N/Ps

N/P	Rate of cellular association (ng h−1)		Cell-associated siRNA (T = 3 h)		% knockdown of VEGF proteina
	0 ≤ T ≤ 0.5 h	0.5 ≤ T ≤ 3 h	ng	% dose
0.5	52.8	0	29.4 ± 1.3	5.9 ± 0.3	0
1	70.3	18b, 0c	54.4 ± 3.8	10.9 ± 0.8	85.8 ± 4.1
2	60	24.5	92.6 ± 8.2	18.5 ± 1.6	94.4 ± 1.3
4	90.8	13.4	81.3 ± 5.0	16.3 ± 1.0	93.3 ± 1.1

^avalues are from Fig. 3A

^bbetween 0.5 and 1.5 h

^cbetween 1.5 and 3 h

The percent of total siRNA that became cell-associated (% dose) doubled as the N/P was doubled between 0.5 and 2 (Table 2). From N/P 2 to 4 the % dose leveled off as did the % knockdown of VEGF protein. Interestingly, only a fraction of the dose (11% at N/P 1) was sufficient to almost completely inhibit protein production. This is equivalent to approximately 2 × 10¹² siRNA molecules, or 7 × 10⁶ copies per cell. For every protein molecule knocked down, there were about 2500 cell-associated siRNA molecules. To our knowledge this is the first time a correlation has been made between the amounts of cell-associated siRNA molecules and down-regulated target protein molecules.

2. Particle size and zeta potential measurements

Fig. 5 shows particle sizes and zeta potentials of 1,3lb2-siRNA complexes. The lipoplexes were several hundred nanometers in diameter, ranging from 268 ± 4 nm at N/P 1 to 785 ± 62 nm at N/P 4. The low polydispersity index (PDI) values of the lipoplexes indicate fairly homogeneous particle size distributions. Only lipoplexes at N/P 4 had a positive zeta potential.

Fig. 5.

Particle size and zeta potential measurements of lipoplexes prepared at various N/Ps with 53.7 nM siRNA. The results are expressed as the mean ± SD (n = 2).

Interestingly, the particle sizes and zeta potentials at N/P 0.5 and 1 were nearly identical, yet at N/P 0.5 there was no inhibition of VEGF production as opposed to 86% knockdown at N/P 1. The PDI at N/P 0.5 (0.34) was 1.5 to 2-fold higher than that of the other lipoplexes, and the greater heterogeneity of the particle size distribution at this N/P may have contributed to the lack of bioactivity. More likely, as mentioned in the previous section, the absence of knockdown was attributed to a deficiency in excess free lipid available to facilitate the membrane interaction necessary for cellular internalization of siRNA and/or its release into the cytosol.

3. Complexation studies

Unlike pDNA, the much shorter and more rigid structure of siRNA (23 nucleotides versus several kilobases for pDNA) averts cationic lipid-induced compaction (10,16). Therefore, when EtBr-siRNA was titrated with 1,3lb2, the reduction in EtBr fluorescence emission intensity that was observed was not necessarily caused by the dye’s displacement from the siRNA duplex. SiRNA-bound lipid quenches the fluorescence of intercalated dye, thus the intensity decrease is an indication of binding and was used to calculate the complexation efficiency according to Eq. 2. Fig. 6A shows the % binding as a function of the N/P. The sigmoidal shape of the plot reflected positively cooperative binding; that is, the affinity of siRNA toward 1,3lb2 increased as a function of lipid saturation. At N/P 0.5, there was only 20% interaction between 1,3lb2 and siRNA. Half the maximum binding was achieved at about N/P 1. Binding reached 94% at N/P 2, and extrapolation of the curve fit revealed complete binding around N/P 3.

Fig. 6.

SiRNA complexation efficiency of 1,3lb2. (A) Plot of % binding (mean ± SD, n = 2) between 1,3lb2 and siRNA as a function of the N/P. The curve fit is represented by a dashed line. (B) Agarose gel electrophoresis of lipoplexes at N/Ps ranging from 0.25 to 4. Naked siRNA was loaded into the outermost wells. (C) Scatchard (top) and Hill (bottom) plots for the titration of a constant concentration of siRNA with 1,3lb2. The results are expressed as the mean ± SD (n = 2).

Fig. 6B shows the siRNA-complexation efficiency of 1,3lb2 via agarose gel electrophoresis. From N/P 0.25 to 2 the band intensity decreased by 2-fold as the N/P was doubled. At N/P 0.5 there was a 19% decrease in the intensity, and the reduction increased to 41% at N/P 1. The intensity was reduced by 82% at N/P 2 and was completely absent at N/P 3, signifying complete binding between 1,3lb2 and siRNA. These values match well with the corresponding binding percentages from the titration study.

Scatchard and Hill analyses (17) of the titration curve revealed a cooperative binding mechanism (Fig. 6C). The concave-downward curve of the Scatchard plot was indicative of positive cooperativity which was also confirmed by a Hill slope value of 3. The Hill plot yielded a K_A of 1.17 10⁵ M⁻¹, falling within the range of binding constants found in the literature for other cationic lipids (18,19).

CONCLUSIONS

The optimized 1,3lb2 lipoplex formulation is a simple two-component system in which siRNA is completely bound to cationic lipid, as proven by the complexation studies. Further characterization of excess free lipid and of potentially “empty” liposomes, i.e. by separating free lipid from siRNA-bound liposomes on a sucrose gradient (20,21), is not possible without disrupting the equilibrium of the system, a crucial balance between lipid monomers and lipoplexes.

The bioactivity of 1,3lb2-VEGF-siRNA (phosphorothioate) was compared to that of other cationic vectors delivering native VEGF-siRNA reported in the literature (Table 3). To achieve a knockdown level of VEGF protein similar to 1,3lb2, a 4-fold increase in the N/P (from 2 to 8), a 2-fold increase in the siRNA concentration (from 53.7 to 100 nM), and a 3-fold increase in the concentration of cationic lipid (from 2.4 to 6.9 μM) were required for commercially available Lipofectamine (6). This corresponds to a 4× greater dose (1.49 versus 0.4 μg) and a 7.5× greater amount of cationic lipid (6 versus 0.8 μg) in the Lipofectamine formulation. The cationic component of Lipofectamine, DOSPA, is pentavalent, 2.5× the charge density of 1,3lb2. Also, Lipofectamine contains an additional ingredient, the neutral colipid DOPE, which is not required for 1,3b2-mediated siFection. Similarly, optimized formulations of polyelectrolyte complexes (PECs) i.e. KALA- or PEI-PEGylated siRNA PECs were found at an N/P as high as 6 and 16, respectively, with as much as 200 nM siRNA (22). Kim et al. (23) reported a maximum of only 55% VEGF protein inhibition with 235 nM siRNA and 15 μM PECs containing a water soluble lipopolymer (WSLP) of PEI when transfecting just 80,000 cells (versus 300,000 cells with 1,3lb2).

Table 3.

Comparison of 1,3lb2-VEGF-siRNA (phosphorothioate) with other cationic carriers of VEGF-siRNA (native).

Delivery system	Cells well−1	Wells plate−1	Cationic agent		Dose		Lipid:siRNA ratio		Transfection time (h)	Cell medium incubation time (h) prior to ELISA	% VEGF protein knockdown	Reference
			μM	μg	nM	μg	N/Pa	w/w
1,3lb2 dispersion	300,000	12	2.4	0.8	53.7	0.4	2	2	3	48	94.4 ± 1.3
Lipofectamine	200,000	6b	6.9b	6b	100b	1.49b	8b	4	4	16	98.7	Takei et al. (6)
KALA PECs	200,000	12	7.2b	NA	200	NA	6	NA	4	18	90 ± 6.6	Lee et al. (22)
PEI PECs	200,000	12	NA	NA	200	NA	16	NA	4	18	Almost complete inhibition	Lee et al. (22)
WSLP PECs	80,000	24	15c	7.35b,c	235b	1.05	NA	7	4	16	55	Kim et al. (23)

^anitrogen-to-phosphate ratio, except for 1,3lb2 (nitrogen-to-phosphorothioate ratio)

^bcalculated from available information

^cvalue is with respect to the entire PEC, not the WSLP alone NA denotes not available

In summary, the cationic lipid 1,3lb2 was found to be a safe and efficacious cellular transporter of siRNA for potent knockdown of VEGF. It is a nontoxic delivery system that effectively binds siRNA and whose lipoplexes promote long-lasting inhibition, are highly bioactive at low N/Ps, and can function in the presence of serum. 1,3lb2 also has the potential for broader application, with the possibilities of successfully delivering other types of siRNA for silencing different targets as well as successful treatment in vivo in light of the high demand by pharmaceutical and biotech companies for proof of concept of RNAi-based therapeutics in primate subjects.

Supplementary Material

1_si_001

Fig. S1. Fluorescence emission scans of FITC-siRNA alone, in lipoplexes, and with both lipid and SDS.

Fig. S2. Standard curve of FITC-siRNA fluorescence emission intensity (mean ± SD) versus concentration. The solid line represents the curve fit.

Fig. S3. Plot of EtBr fluorescence emission intensity versus siRNA concentration. The solid line represents the curve fit.

1

3lb2, 1,3-dimyristoylamidopropane-2-[bis-(2-dimethylaminoethane)] carbamate

ELISA

enzyme-linked immunosorbent assay

EtBr

ethidium bromide

FBS

fetal bovine serum

FITC

fluorescein isothiocyanate

HPLC

high performance liquid chromatography

lipoplex

cationic lipid-nucleic acid complex

MTT

3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide

N/P

nitrogen-to-phosphorothioate ratio

PDI

polydispersity index

pDNA

plasmid DNA

PEC

polyelectrolyte complex

RNAi

RNA interference

RT-PCR

reverse-transcription polymerase chain reaction

scr

scrambled

SFM

serum-free medium

siFection

cationic liposome/lipid-mediated siRNA cellular delivery

siRNA

small interfering RNA

VEGF

vascular endothelial growth factor

WSLP

water soluble lipopolymer