Targeting ^V600EB-Raf and Akt3 Using Nanoliposomal-Small Interfering RNA Inhibits Cutaneous Melanocytic Lesion Development

Melanoma is a cancer of pigmented skin cells called melanocytes residing at the epidermal-dermal junction (1). Most genetic alterations promoting development of early melanomas are yet to be identified. One frequent change occurring in ∼90% of normal moles is a T1799A mutation of B-Raf in the mitogen-activated protein (MAP) kinase pathway (2–5). Approximately 60% of advanced-stage metastatic melanomas also contain the activating V600E mutation leading to 10.7 times higher activity than occurs in normal cells, thereby promoting cellular proliferation (2–5).

A second frequent alteration promoting melanoma development is increased Akt3 activity in ∼70% of melanomas (6). Akt3 is activated through loss of the PTEN phosphatase and increased gene copy number, leading to deregulation of apoptosis, thereby promoting chemoresistance (6, 7). Targeting ^V600EB-Raf or Akt3 signaling decreases melanoma tumor development, and simultaneous inhibition can lead to cooperatively acting tumor inhibition (6, 8, 9). Thus, agents specifically targeting ^V600EB-Raf and/or Akt3 could have significant potential to treat moles, early melanocytic lesions, or skin metastases in which these proteins are deregulated.

Inhibition of B-Raf or Akt3 using small interfering RNA (siRNA) specifically designed to inhibit ^V600EB-Raf but not wild-type B-Raf or Akt3 but not Akt1 or Akt2 protein expression can effectively decrease melanoma tumor development in mice (6, 8, 10). Because these agents target specific genes, there would be fewer potential side effects than with drugs inhibiting both mutant and wild-type B-Raf or all three Akt isoforms (6, 8, 10). Introduction of siRNA in cultured cells is easily accomplished using standard transfection methods (11–13). However, its use as a targeted chemopreventive or therapeutic agent is limited due to degradation in animal systems (14, 15).

Encapsulating siRNA into nanoliposomes can be used for protection and systemic delivery, leading to tumor inhibition (16–18). In contrast, topical administration on skin leads to uptake confined to the upper epidermal layers without complete skin penetration (19). This is due, in part, to skin architecture, with epidermis and dermis protected by a surface keratinized layer (19, 20), which prevents most topically applied agents from reaching early melanocytic lesions developing at the epidermal-dermal junction or from reaching locally invasive cutaneous melanoma metastases in the skin dermis (21). One promising strategy for delivery through the keratinized layer is the use of low-frequency ultrasound to permeabilize the skin, enabling passage of macromolecules, including antisense oligonucleotides without significant skin damage (21–26). Combining the protective qualities of nanoliposomes with the permeabilizing activities of low-frequency ultrasound is a novel unexplored strategy for delivering siRNA into melanocytic lesions present in skin to prevent or treat cutaneous melanoma.

This study details the development of a unique nanoliposomal-ultrasound–mediated approach for delivering siRNA targeting ^V600EB-Raf and Akt3 protein expression into melanocytic lesions present in skin to inhibit early lesion development or prevent cutaneous melanoma metastasis.

Cell lines and culture conditions. Human melanoma cell lines, UACC 903, UACC 903-GFP, 1205 Lu, and C8161.Cl9, and human fibroblasts, FF2441, were maintained in DMEM (Invitrogen) supplemented with 10% fetal bovine serum (FBS; Hyclone). WM35 and WM35-GFP cells were grown as described previously (27). Human keratinocytes, HFK, were grown in Epilife Medium with Human Keratinocyte Growth Supplement (Cascade Biologics). Mouse melanocytes, melan-a, were grown in RPMI 1640 (Mediatech) supplemented with 10% FBS, 2 mmol/L l-glutamine, and 200 nmol/L 12-O-tetradecanoylphorbol-13-acetate (Sigma-Aldrich).

Liposome composition, characterization, and siRNA loading. Nanoliposomes were created by combining 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP), 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE), and 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[amino(polyethylene glycol)2000] [DSPE-PEG(2000); Avanti Polar Lipids] in chloroform at a 4.75:4.75:0.5 molar ratio (28, 29). siRNA and nanoliposomes complexing occurred at specific weight ratios of 1:5, 1:10, and 1:15 for 0.5 to 24 h. A quasielastic light scattering system (Malvern Nanosizer, Malvern Instruments) was used to measure the particle diameter of nanoliposomes ± siRNA 1 d after preparation. Nanoliposomal loading of siRNA was measured by mixing fluorescent, Alexa Fluor 546–tagged siRNA (Qiagen) with nanoliposomes at specific weight ratios (1:5, 1:10, 1:15) and complexing at room temperature for 0.5, 3, or 6 h. Loading was considered complete when no free siRNA was evident with complex remaining in well of a 2% agarose gel. To measure liposome-mediated protection of siRNA, nanoliposomes complexed with siRNA overnight were treated with 50% FBS for 10, 30, 60, 180, or 360 min. Following FBS treatment, half of the complex at each time point was treated with 0.5% SDS for 10 min at 37°C to disrupt the complexes and release free siRNA. siRNA alone was also digested with 50% FBS at each time point to serve as a control. All samples were run on a 2% agarose gel at 100 V for 20 min and visualized under UV.

Nanoliposomal toxicity. To assess cytotoxicity of nanoliposomes complexed with siRNA in normal or cancer cells, 5 × 10³ fibroblasts (FF2441), keratinocytes (HFK), melanocytes (melan-a), and melanoma (1205 Lu) cells were plated into 96-well plates and, 48 h later, treated with ghost nanoliposomes for 24 h at nanoliposomal concentrations of 12.5, 25, and 50 μmol/L or exposed to 125, 250, 500, 1,000, or 2,000 nmol/L of siRNA complexed with nanoliposomes and compared with untreated control cells. Cytotoxicity was analyzed using the CellTiter 96 aqueous nonradioactive cell proliferation assay (Promega).

Uptake of nanoliposomal-siRNA into cells. Uptake and internalization of nanoliposomal-siRNA into cells was determined using Alexa Fluor 546– and/or fluorescein-tagged siRNA (Invitrogen). 1205 Lu cells (1 × 10⁵) were plated and, following overnight growth, nanoliposomal-siRNA complex (100 nmol/L) or ghost nanoliposomes were added. Three hours later, cells were trypsinized and plated onto coverslips followed by overnight incubation. Uptake of siRNA-nanoliposomal complex by fibroblasts (FF2441), keratinocytes (HFK), or melanocytes (melan-a) was measured by plating 5 × 10³ cells onto coverslips and, 2 d later, adding nanoliposomal-siRNA (50, 100, or 200 nmol/L) for 3 h. Following all conditions, cells were rinsed with PBS, fixed in 4% parafomaldehyde (Fisher Scientific) for 30 min, washed in PBS, mounted in 4′,6-diamidino-2-phenylindole–containing mountant (Vector Laboratories), and analyzed by fluorescence microscopy to measure uptake.

Nucleofection of siRNA into cells. Stealth siRNA (1.5–200 pmol; Invitrogen) was introduced via an Amaxa Nucleofector into 1 × 10⁶ UACC 903, UACC 903-GFP, or C8161.Cl9 cells using Solution R/Program K-17 or into WM35-GFP cells using NHEM-NEO Solution/Program U-20. Resulting transfection efficiency was ∼99% (8). Duration of siRNA-mediated protein knockdown at 0, 2, 4, 6, and 8 d following nucleofection was measured for B-Raf or C-Raf by Western blot analysis. Cells used in reconstructed skin were nucleofected with buffer, siMutB-Raf (100, 50, or 12.5 pmol), siC-Raf (100 pmol), or scrambled (100 pmol) siRNA; plated onto culture dishes; and, 2 d later, mixed with keratinocytes, which was added onto the dermis as detailed below. Protein lysates of remaining cells were analyzed by Western blotting.

Western blot analysis. Western blot analysis was conducted as described previously (30). Blots were probed with primary antibodies to pErk, pMek (Cell Signaling Technologies), B-Raf, extracellular signal–regulated kinase (Erk)-2, Cyclin D1, and C-Raf and secondary antibodies conjugated with horseradish peroxidase (Santa Cruz Biotechnology). Immunoblots were developed with the enhanced chemiluminescence detection system (Amersham Pharmacia Biotech). Blots were normalized to Erk2 and quantified using ImageJ software (31).

Nanoliposomal-siRNA–mediated protein knockdown. 1205 Lu cells (5 × 10⁵) were plated onto six-well plates for 2 d and nanoliposomal-siMutB-Raf or nanoliposomal-siScrambled complex (1 μmol/L) was added. Protein lysates were collected 18 and 32 h later for Western blot analysis to measure B-Raf and Erk2 expression. Protein levels were quantified from three independent blots using ImageJ software, normalized to Erk2, and average B-Raf levels were calculated (31).

Generation of skin containing melanocytic lesions. To create skin in a culture dish, human fibroblasts were trypsinized and resuspended in 10% reconstitution buffer, 10% 10× DMEM (Mediatech), 2.4 μL/mL of 10 mol/L NaOH, and 80% collagen I (Becton Dickinson) at a concentration of 2.5 × 10⁵/mL on ice (32). Mixture was then aliquoted into 6- or 12-well plates and incubated at 37°C for 3 h. E-medium was added to each well to equilibrate the dermal matrix (32). After 2 d of growth, keratinocytes and melanoma cells (WM35-GFP or UACC 903-GFP) were trypsinized and resuspended at a 1:10 ratio of melanoma cells (nucleofected or untreated) to keratinocytes in E-medium. One milliliter of cell suspension was added to each well on top of the dermal layer. Following 2 d of growth, reconstructed skin was transferred onto wire grids and fed via diffusion from E-medium below the platforms.

Ultrasound transducer design and calibration. The cymbal transducer is a novel thin flextensional transducer capable of producing very low frequencies (33, 34). It has a compact, lightweight structure with an adjustable resonance frequency. Caps on the lead zirconate-titanate ceramic contain a shallow cavity beneath the inner surface. The fundamental mode of vibration is flexing of the end caps caused by radial motion of the ceramic. Therefore, the overall displacement of the device is a combination of the axial motion of disk plus radial motion amplified by the end caps (23, 24). A radiofrequency signal was generated by a frequency pulse/function generator (Agilent 32250A) and amplified using an amplifier (model 40A12, Amplifier Research). Electrical impedance of the array was tuned to the output impedance of the amplifier by an external inductor-capacitor tuning network. Pulse period, duty cycle, and exposure time of radiofrequency signal from the frequency generator were monitored using an oscilloscope (Tektronix 2213A). For experiments, signal generator operated at 20 kHz with pulse duration of 200 ms and pulse repetition period of 1 s (i.e., 20% duty cycle); the amplifier gain was set to 50 dB. Pulsed ultrasound was used to avoid heat generation, which could damage array or skin.

Ultrasound treatment and analysis of damage, nanoliposomal-siRNA uptake, and tumor inhibition in reconstructed skin. Skin reconstructs were treated with ultrasound at a frequency of 20 kHz and duty cycle of 20% for 20 min by removing media and submersion in PBS (26). The nanoliposomal-siRNA complex (1 μmol/L) was then applied topically and allowed to soak into the skin. Skin damage was assessed following ultrasound treatment, addition of ghost liposome, and fixation in 4% paraformaldehyde at 4°C overnight. H&E-stained paraffin-embedded cross sections were examined for skin damage by microscopy. Skin reconstruct uptake studies involved treating with nanoliposomes containing 1 μmol/L Alexa Fluor 546–tagged siRNA or ghost nanoliposomes and uptake was analyzed using stereo fluorescence microscopy 1 h later. Skin reconstructs were treated with ultrasound followed by treatment with nanoliposomes containing B-Raf siRNA or ghost nanoliposomes on alternate days from days 10 to 21. Skin reconstructs from all experiments were fixed in 4% paraformaldehyde at 4°C overnight. After fixation, skin was stored in 0.5 mol/L EDTA (pH 8.0; Fisher Scientific). The total average area occupied by green fluorescent protein (GFP)–tagged tumor nodules present in four to six skin images per skin was used to quantify differences between treatment groups using the IP Lab software.

Cell doubling time. In vitro doubling time of UACC 903 cells nucleofected with siRNA was estimated by plating 3 × 10³ UACC 903 cells in 96-well plates and at 24-h intervals, using the SRB assay (Sigma) to measure growth rate and calculate doubling times (8).

In vivo cell proliferation. For mechanistic studies, 5 × 10⁶ UACC 903 cells nucleofected with siRNA were injected into mice, and tumors were harvested 4 d later to quantify cell proliferation in formalin-fixed tumor sections using the RPN 20 cell proliferation kit (Amersham Pharmacia Biotech; ref. 8).

Anchorage-independent growth. UACC 903 cells (1 × 10⁶) were nucleofected with siAkt3 (200 pmol), siMutB-Raf (3, 6, or 12 pmol), siScrambled, or buffer; allowed to recover for 2 d; and then plated in serum-free DMEM (2 × 10⁴ per well) onto polyHEMA-coated 96-well plates. Cell viability was assessed 3 d later using the 3-(4,5-dimethyl-thiazol-2yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium (MTS) assay (9).

Ultrasound and siRNA-nanoliposomal complex treatment of tumors. UACC 903-GFP cells (1 × 10⁶) were injected s.c. into the right flank of nude mice. After 24 h, mice were ultrasound treated for 15 min followed by topical application of 25-μg nanoliposomal-siRNA (siMutB-Raf, siAKt3, or siScrambled) complex on alternate days. For combination nanoliposomal ^V600EB-Raf and Akt3 siRNA studies, mice were treated with the same regimen as the single agents. Single agents were compensated with extra liposomes so that equivalent amounts were used. Before ultrasound treatment, mice were anesthetized using 2:1 ketamine/xylazine, and area was sterilized with alcohol pads. Aquaflex Gel Pads (Parker Laboratories, Inc.) served as an ultrasound standoff. Ultrasound was administered for 15 min at a frequency of 20 kHz and duty cycle of 20% (26). Nanoliposomal-siRNA was applied to ultrasound-treated area overlying the tumor for penetration into skin.

Toxicity analysis in mice. To establish toxicity associated with ultrasound plus siRNA-nanoliposomal complex, nude mice were ultrasound treated for 15 min followed by topical addition of nanoliposomal-siRNA (siMutB-Raf, siAkt3, siScrambled, or siMutB-Raf + siAkt3) complex on alternate days. Mice were weighed on alternate days to ascertain toxicity. Following euthanization on day 23, blood was collected from mice and serum separated to measure aspartate aminotransferase (AST), alanine aminotransferase (ALT), glucose, and alkaline phosphatase. Livers were fixed in formalin, sectioned, and stained with H&E to assess liver morphology changes.

Statistical analysis. One-way or two-way ANOVA followed by appropriate post hoc test (Tukey's or Bonferroni) or t test was used to ascertain whether significant differences occurred between groups. Differences were considered significant at P < 0.05. Combination index was calculated using the Chou-Talalay method with CalcuSyn software to establish mechanistic basis for cooperative inhibition mediated by siRNA targeting ^V600EB-Raf and Akt3 (35).

Development of stable cationic nanoliposomes that load siRNA. Because nanoliposomes have been reported as potential siRNA delivery vehicles, a number of possible formulations were evaluated and a novel DOTAP/DOPE/DSPE-PEG(2000) ratio of 4.75:4.75:0.5 was found to efficiently load siRNA. Loading was measured by adding Alexa Fluor 546–tagged siRNA to nanoliposomes at ratios of 1:5, 1:10, or 1:15 (siRNA/liposome, by weight) and complexed for 0.5, 3, or 6 hours. Measuring free fluorescently tagged siRNA present in the gel or as a siRNA-liposome complex retained in the gel well was used to directly assess siRNA loading. A 1:5 ratio following a 0.5-hour incubation showed the presence of both loaded and unloaded siRNA, which was in contrast to the 1:10 and 1:15 ratios where siRNA remained in gel wells, indicating complete complexing with nanoliposomes (Fig. 1A,, top left). Because maximal loading occurred following 30-min incubation at a 1:10 ratio, this siRNA-nanoliposomal complex formulation was used for subsequent experiments. The diameter of siRNA-liposome complexes 1 day after preparation was measured by dynamic light scattering, showing an average diameter of ∼50 nm with a range of 34 to 67 nm (Fig. 1A , top right). Ghost nanoliposomes had a similar diameter averaging ∼48 nm with a range of 32 to 64 nm. Thus, stable nanoliposomes were generated that loaded siRNA.

siRNA in nanoliposomes is protected from serum degradation. To evaluate the protective effects conferred by nanoliposomes on siRNA, fluorescently tagged free siRNA or siRNA-liposome complexes were exposed to 50% FBS, which is rich in RNA degrading factors, for 10, 30, 60, 180, or 360 minutes. Partial degradation of free siRNA was observed after 10 minutes and complete degradation was observed at 60 minutes (Fig. 1A,, bottom left). In contrast, siRNA-nanoliposomal complexes remained in gel wells at all time points, indicating protection from degradation (Fig. 1A,, bottom middle). Half of the original sample was treated with 0.5% SDS for 10 minutes, disrupting the complex and releasing siRNA, which was then run on an agarose gel. Similar levels of free siRNA were observed at all time points, indicating protection by the complex (Fig. 1A , bottom right). Thus, siRNA in nanoliposomes was protected from degradation by external factors.

Nanoliposomal-siRNA complex is nontoxic and taken up into the cytoplasm of normal as well as cancer cells. To determine nanoliposomes mediated cellular uptake of siRNA, fibroblasts, keratinocytes, and melanocytes were treated with ghost nanoliposomes or nanoliposomes containing Alexa Flour 546–tagged siRNA (50, 100, or 200 nmol/L). Fluorescence microscopy showed red Alexa fluorescence in ∼100% of all cells, indicating uptake of nanoliposomal-siRNA (Fig. 1B). To verify that siRNA was taken into the cytoplasm of cells, 1205 Lu cells were treated with 100 nmol/L fluorescein-tagged siRNA-nanoliposomal complex or ghost nanoliposomes. Cells were then trypsinized to remove cell surface–bound nanoliposomes and replated overnight followed by fixation (36). Fluorescence microscopy showed the presence of green fluorescein–tagged siRNA in cytoplasm of cells surrounding a nuclear shadow, indicating that nanoliposomal-siRNA was internalized (Fig. 1C). In contrast, no green fluorescence was observed in cells treated with ghost nanoliposomes. Next, toxicity mediated by ghost nanoliposomes was examined in fibroblast, keratinocyte, melanocyte, or 1205 Lu melanoma cells. Cells were treated with increasing concentrations of ghost nanoliposomes (12.5, 25, and 50 μmol/L) for 24 hours and analyzed by MTS assay (37). Compared with untreated cells, no significant decrease in cellular viability was observed, indicating that ghost nanoliposomes were not toxic at the concentrations examined (Fig. 1D). The toxicity of siRNA-nanoliposomal complex was next examined in 1205 Lu cells, comparing untreated cells to those treated with increasing concentrations of complex loaded with scrambled siRNA (Fig. 1D). No change in cell viability was observed, indicating negligible toxicity. Thus, the siRNA-liposome complex was nontoxic and taken up into the cytoplasm of normal and cancer cells.

siRNA targeting ^V600EB-Raf specifically decreases mutant, but not wild-type, protein expression in cells. siRNA can be designed overlapping the T1799A mutation site of ^V600EB-Raf, which has potential to selectively decrease the expression of mutant, but not wild-type, protein (8, 10, 30). To verify the specificity of siRNA targeting ^V600EB-Raf, 100 pmol (1 μmol/L) was nucleofected into melanoma cells containing mutant (UACC 903) or wild-type protein (C8161.Cl9). Western blot and densitometric analysis of cellular protein lysates showed that siRNA designed specifically against ^V600EB-Raf reduced protein levels by ∼50% in mutant UACC 903 (Fig. 2A) but not in wild-type B-Raf–expressing C8161.Cl9 cells (Fig. 2B). In contrast, siRNA designed to sequences present in both mutant and wild-type B-Raf decreased protein expression by ∼75% in both cell lines (Fig. 2A and B). Similarly, siRNA that specifically targets Akt3 but not Akt1 or Akt2 has been reported, which can effectively decrease melanoma tumor development (6). Thus, siRNA designed to specifically inhibit mutant ^V600EB-Raf or Akt3 reduces the expression of mutant but not wild-type B-Raf protein and Akt3 but not Akt1 or Akt2, respectively (6).

siRNA targeting ^V600EB-Raf can be loaded into nanoliposomes, delivered into cells, and decrease expression of mutant protein. Having validated the specificity of siRNA targeting ^V600EB-Raf and Akt3, 1 μmol/L siRNA was loaded into nanoliposomes, and protein knockdown measured in 1205 Lu cells containing ^V600EB-Raf in three independent experiments by densitometric analysis and Western blotting (Fig. 2C). Examples of protein knockdown at 18 and 32 hours following a single treatment showed a 50% to 60% decrease of ^V600EB-Raf protein compared with cells treated with scrambled siRNA-nanoliposomal complex (Fig. 2C). Typically, protein knockdown ranged from 25% to 60%, indicating some variability, but knockdown was reproducibly observed. Duration of knockdown following a single siRNA treatment was measured by introducing siRNA into cells and using Western blotting to measure protein knockdown 2, 4, 6, and 8 days later. Western blots were densitometrically scanned and knockdown of B-Raf protein was compared with control cells nucleofected with siRNA targeting C-Raf. A consistent ∼60% decrease in protein expression was observed through day 8 (Fig. 2D). Thus, siRNA specific to ^V600EB-Raf can be loaded into cationic nanoliposomes, delivered into melanoma cells, and decrease protein expression by 25% to 60% following a single treatment.

Ultrasound treatment of laboratory-generated skin containing melanocytic lesions enables delivery of siRNA-nanoliposomal complex into tumors. Skin forms a natural barrier that prevents uptake of agents including lipids and siRNA (19, 21). Therefore, skin was permeabilized before addition of siRNA-liposomal complex using a lightweight four-cymbal array ultrasound developed originally to deliver insulin through skin and into the blood stream (Fig. 3A,, left; refs. 23–26). A representative exposimetry analysis showed a Gaussian-shaped peak field intensity with an overall spatial peak-temporal intensity of 50.6 ± 1.5 mW/cm² (Fig. 3A,, right). Possible skin damage by ultrasound was evaluated on reconstructed skin treated with ultrasound for 20 minutes followed by addition of topically added ghost liposome for 1 hour followed by fixation. H&E-stained skin sections comparing untreated control to ultrasound- and ghost liposome–treated skin showed no detectable damage (Fig. 3B). Next, skin penetration by siRNA-nanoliposomal complex following ultrasound treatment was measured. Skin was ultrasound treated for 20 minutes followed by topical addition of liposome loaded with 1 μmol/L Alexa Fluor 546–tagged siRNA and, 1 hour later, fixed in 4% paraformaldehyde followed by analysis using fluorescence stereomicroscopy. Top-down views of the skin showed uptake of red Alexa Fluor 546–tagged siRNA in both keratinocytes and UACC 903-GFP melanoma cells (Fig. 3C). Cross sections of skin showed the presence of red Alexa Fluor 546–tagged siRNA in melanoma cells located in both epidermis and epidermal-dermal junction (Fig. 3D). Similar results were seen in skin reconstructs containing WM35-GFP (data not shown). Thus, permeabilizing skin using ultrasound before topical application of siRNA-nanoliposomal complex enabled delivery of siRNA to melanocytic lesions located within skin.

Following ultrasound treatment, nanoliposomal complex containing siRNA targeting ^V600EB-Raf inhibits melanocytic lesion development in skin reconstructs. The effectiveness of siRNA-nanoliposomal complex for inhibiting early melanocytic skin lesions or cutaneous metastasis development was examined by treating skin containing GFP tumors on alternate days with ultrasound for 20 minutes followed by addition of nanoliposomes containing siRNA to ^V600EB-Raf or ghost liposome (Fig. 4A). On day 21, stereo fluorescence microscopy was used to quantify average area occupied by GFP fluorescent skin lesions, which was compared with ultrasound + ghost liposome–treated skin. A statistically significant 50% to 65% reduction in area occupied was observed for WM35-GFP (Fig. 4B) and UACC 903-GFP (Fig. 4C) tumor nodules (P < 0.05, one-way ANOVA). The WM35 cell line represents early noninvasive melanocytic lesions, whereas UACC 903 cells represent invasive melanocytic lesions in the reconstructed skin model. Thus, siRNA-mediated inhibition of ^V600EB-Raf decreased the proliferative capacity of tumor cells, reducing skin nodule development.

Mechanistically, siRNA targeting ^V600EB-Raf decreases the proliferative potential of melanoma cells by inhibiting MAP kinase signaling in laboratory-generated skin. To establish the mechanistic basis by which siRNA targeting ^V600EB-Raf inhibited melanoma development, siRNA was introduced into GFP-tagged WM35 or UACC 903 cells, which were seeded into the epidermis of reconstructed skin. Ten days later, skin was fixed and fluorescence microscopy was used to quantify the average area occupied by GFP-expressing tumor nodules. Compared with cells nucleofected with buffer, scrambled siRNA, or C-Raf siRNA, a 3- to 7-fold reduction in nodule development was observed by cells exposed to siRNA targeting ^V600EB-Raf (Fig. 5A and B; P < 0.05, one-way ANOVA). Knockdown of ^V600EB-Raf protein expression was confirmed by Western blotting of protein lysates collected following creation of reconstructed skin (Fig. 5A and C).

The mechanism leading to decreased cutaneous tumor development was established by calculating in vitro doubling times as well as comparing differences in proliferation and measuring the effect on MAP kinase signaling. siRNA-mediated inhibition of ^V600EB-Raf decreased mutant protein expression, causing corresponding decreases in phosphorylated MAP kinase/Erk kinase (pMek) and pErk as well as decreased cyclin D1 expression, indicating reduction in the proliferative signaling pathways of these cells (Fig. 5A and C). The doubling time of cultured cells containing siRNA targeting ^V600EB-Raf was found to be ∼1.3-fold higher than that of control cells treated with buffer or scrambled siRNA (Fig. 5D,, left). Furthermore, a 4- to 5-fold decrease in proliferating cells was observed in tumor cells treated with siRNA targeting ^V600EB-Raf compared with control cells treated with C-Raf siRNA (Fig. 5D , right). This growth-inhibitory mechanism is consistent with prior reports (8, 30). Thus, siRNA in nanoliposomes can decrease the expression of ^V600EB-Raf protein, reducing MAP kinase signaling and the proliferative potential of melanoma cells.

Following ultrasound treatment, siRNA-nanoliposomal complex targeting ^V600EB-Raf and Akt3 cooperatively decreased melanocytic lesion development in animal skin. The effect of ultrasound followed by topical application of nanoliposomal-siRNA complex on melanoma development in animal skin was examined next. UACC 903-GFP cells (1 × 10⁶) were s.c. injected into the right flank of nude mice and, under anesthesia, treated on alternate days at the tumor site with ultrasound for 15 minutes followed by topical application of siRNA-nanoliposomal complex (Fig. 6A). Along with measurement of tumor size (Fig. 6B), mice were weighed (Fig. 6B,, inset) on alternate days. The effects of liposomes containing siRNA targeting ^V600EB-Raf or Akt3 alone or ^V600EB-Raf and Akt3 together on tumor development were compared. Ghost liposomes were added to single treatments to compensate for differences in the amount of liposomal vehicle. A statistically significant ∼30% reduction in tumor size was observed from day 15 in mice treated with complex containing siRNA targeting ^V600EB-Raf compared with animals exposed to complex containing scrambled siRNA (Fig. 6B; P < 0.05, two-way ANOVA). In contrast, no statistically significant difference was observed in cutaneous melanoma development following treatment with liposomes containing siRNA targeting Akt3 although treated tumors were 10% to 15% smaller than controls treated with liposomes containing scrambled siRNA (Fig. 6B; P > 0.05, two-way ANOVA). Significantly, the combination treatment showed cooperatively acting tumor inhibition from day 11 when compared with siScrambled-liposomal complex (Fig. 6B; P < 0.05, two-way ANOVA).

Weights were monitored on alternate days to ascertain weight-related toxicity (Fig. 6B , inset). No differences were seen in weights, indicating a lack of toxicity (P > 0.05, two-way ANOVA). The lack of systemic toxicity was confirmed by analyzing blood for ALT, AST, alkaline phosphatase, and glucose levels (data not shown), which showed no significant differences between animal groups (P > 0.05, t test). Additionally, no changes in liver morphology were detected in H&E-stained sections (data not shown).

Targeting ^V600EB-Raf and Akt3 leads to additively cooperative inhibition. Because liposomal-mediated targeting of Akt3 and ^V600EB-Raf suggested cooperatively acting tumor inhibition, the mechanistic basis for inhibition was determined. siRNA targeting Akt3 and/or ^V600EB-Raf was nucleofected into UACC 903 cells and the effect on in vitro anchorage-independent growth was examined. For these studies and subsequent Chou-Talalay analysis for measuring the mechanistic basis underlying inhibition, Akt3 siRNA was maintained at 200 pmol and B-Raf siRNA was titrated at doses of 3, 6, and 12 pmol. Simultaneous siRNA-mediated inhibition of Akt3 and B-Raf significantly decreased the growth of cells compared with either Akt3 or B-Raf targeting alone (Fig. 6C, lanes 3 and 4 versus lane 7, lanes 3 and 5 versus lane 8, and lanes 3 and 6 versus lane 9; P < 0.05, one-way ANOVA; ref. 38).

To establish whether growth inhibition would be additive or synergistic following simultaneous inhibition of Akt3 and ^V600EB-Raf, the Chou-Talalay method was used for determining the combination index using Calcusyn software (35). Using this approach, combination index values <0.9 are synergistic, >1.1 is antagonistic, and values 0.9 to 1.1 are additive (Fig. 6D). The estimated combination index values were 0.94, 0.97, and 1.10, for 3, 6, and 12 pmol of B-Raf siRNA plus 200 pmol of Akt3 siRNA, suggesting additive cooperation (Fig. 6D). Thus, targeting ^V600EB-Raf and Akt3 in culture or in animals should lead to cooperatively additive inhibition. However, tumor inhibition following liposome-mediated delivery of siRNA targeting ^V600EB-Raf and Akt3 in Fig. 6B seems to be more than additive, suggesting that it may tend toward being synergistic.

Novel therapeutic regimens need to be developed for treating early or cutaneous metastatic melanoma because, aside from surgical excision, few treatment options are available (39). Current systemic therapeutics cause toxicities and have off-target effects (19). Topical or localized treatments with agents could permit use of high local concentrations with minimal toxicity and be useful for treating cutaneous lesions not amenable to surgical removal (19, 40). A case in point is the use of interlesional injections of interleukin 2 (IL-2) and electroporation for improved topical delivery of bleomycin for melanoma treatment (40, 41). Effective topical agents could provide additional, currently unavailable, treatment options for patients with early or invasive cutaneous melanoma.

The era of personalized, targeted cancer treatment is near, but specific proteins playing key roles in particular cancer types need identification and the demonstration that simultaneous targeting inhibits tumor development in a synergistic manner. This study shows that ^V600EB-Raf and Akt3 are two such targets in melanoma, the inhibition of which led to cooperative tumor inhibition. Furthermore, a novel approach has been developed for delivering therapeutic siRNA into early or invasive cutaneous melanocytic lesions targeting these important proteins. Ultrasound treatment of skin followed by topical application of nanoliposomes containing siRNA targeting ^V600EB-Raf and Akt3 reduced the expression of these proteins, significantly decreasing melanoma development in a cooperatively additive manner.

Simultaneous inhibition of the MAP and phosphatidylinositol 3-kinase pathways has previously been reported to enhance the inhibitory effect on both signaling cascades leading to increased tumor cell apoptosis, which supports the feasibility of this approach (42–44). Furthermore, siRNA-mediated inhibition of ^V600EB-Raf and Akt3 in aggressive melanoma cells has been shown to cooperatively inhibit tumor development through enhanced effects on the signaling cascades, specifically leading to decreased cyclin D1 expression as well as increased p27 and cleaved caspase-3 levels (9). These results are not surprising considering discoveries linking the two pathways in early melanoma development (9). Approximately 90% of nevi express ^V600EB-Raf, but very few ever develop into melanoma due to the inhibitory effects associated with high MAP kinase pathway activity caused by the mutant protein. Akt3 has been shown to overcome this block by phosphorylating ^V600EB-Raf in these melanocytes to lower the activity of mutant protein and downstream MAP kinase pathway signaling to levels promoting, rather than inhibiting, melanoma development (9). Phosphorylation of ^V600EB-Raf enabled melanocytes to overcome the growth-inhibitory effects associated with high pathway activity and to develop melanoma cell–like characteristics (9). Thus, this discovery shows that siRNA can be used as a therapeutic specifically targeting multiple aberrantly expressed or mutant cancer-causing genes without inhibiting the activity of normal proteins to more effectively treat melanoma.

Systemic administration of free unmodified siRNA by i.v. injection is not clinically feasible because it is rapidly degraded by serum factors, preventing clinically useful concentrations from being reached (45). Development of novel vehicles for loading, protecting, and delivering siRNA to distant sites around the body is one strategy (16–18, 46, 47). This study shows that cationic nanoliposomes can load, protect, and deliver therapeutic siRNA into melanoma cells. Nanoliposomes are taken up by all cell types but only affects those containing ^V600EB-Raf or overexpressing Akt3. The specificity of siRNA for its target reduces toxicity and potential side effects. Furthermore, use of pegylated lipids and small average nanoliposomes with a diameter of ∼50 nm decreases immunologic or inflammatory responses, enabling the particles to bypass the reticulo-endothelial system (48, 49).

Ultrasound was used to permeabilize skin, enabling delivery of siRNA-liposomal complex throughout the epidermal and dermal layers because it had been reported to be effective for delivering various macromolecules into and throughout skin (19, 21–26). A lightweight transducer with shown capabilities of delivering insulin or antisense oligonucleotides into or through skin was used, showing that low-frequency ultrasound can permeabilize skin, enabling delivery of nanoliposomal-siRNA into melanoma cells residing at the dermal-epidermal junction to inhibit protein expression and melanoma development.

Mechanistically, a therapeutic nanoliposomal-siRNA complex targeting ^V600EB-Raf and Akt3 can effectively decrease the expression of both proteins and downstream signaling cascades to inhibit the development or spread of cutaneous melanoma (8, 10, 30). Thus, a nanoliposomal-siRNA complex targeting these proteins could be used as a nonsurgical approach to decrease the number of benign moles or inhibit early cutaneous melanocytic lesion development (2–5, 9). The latter would have the consequence of decreasing the spread at the earliest stages, thereby prolonging survival by preventing metastases. Ultrasound in conjunction with a therapeutic topical nanoliposomal-siRNA complex targeting ^V600EB-Raf and Akt3 could also provide a treatment option for patients whose disease is not amenable to surgical removal or limb perfusion when widespread cutaneous lesions are present on the trunk. No therapeutic agents for treating these patients are available. Furthermore, advanced-stage melanoma patients could benefit by administering these agents at the time of sentinel lymph node biopsy or surgery to target cells that may not have been removed during the procedure (50). Finally, this approach could be combined with more traditional therapeutic agents to improve clinical efficacy through synergistically acting mechanisms.

In conclusion, topical delivery of cationic nanoliposomes loaded with siRNA targeting ^V600EB-Raf and Akt3 in combination with low-frequency ultrasound has potential to decrease early melanocytic lesion development in skin or prevent the spread of cutaneous melanoma metastases.

M. Kester: ownership interest, Keystone Nano. The other authors disclosed no potential conflicts of interest.

Grant support: NIH/National Cancer Institute grant 1-RO3-CA128033-01, American Cancer Society grant RSG-04-053-01-GMC, The Foreman Foundation for Melanoma Research, State of Pennsylvania Non-formulary Tobacco Settlement Funds, and Department of Defense Technologies for Metabolic Monitoring grant W81XWH-05-1-0617.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank James Kaiser, Julie Anderson, Nishit Trivedi, Mitchel Cheung, and SubbaRao V. Madhunpantula for technical assistance, and Shantu Amin for critical reading of the manuscript.