Comparison of strategies targeting Raf-1 mRNA in ovarian cancer



In this study, we characterize the uptake and specificity of a first-generation Raf-1 antisense oligonucleotide (ASO) (ISIS 5132) and compare it with a second-generation ASO (ISIS 13650) and an RNA interference approach. All three approaches resulted in inhibition of both Raf-1 expression and cellular growth. Specificity of the Raf-1 ASOs was confirmed by comparison with ASOs targeted against another Raf isoform (B-Raf) as well as mismatch sequences. Cellular uptake studies with FAM-labelled ISIS 5132 revealed that whilst the majority of cells treated at a low-intermediate plating density were labelled within 3 hr, cells treated at high density demonstrated neither Raf-1 protein knockout nor significant growth inhibition, following similar treatment. This lack of response at high cell densities was associated with reduced pERK and Raf-1 inhibition. Cell cycle analysis revealed that whilst SKOV-3 cells both accumulated in the S-phase of the cell cycle and showed enhanced annexin V levels, following Raf-1 ASO treatment; these effects were also demonstrated with first-generation but not second-generation mismatch oligonucleotides. Bromodeoxyuridine incorporation analysis suggested that these effects may indeed be partly attributable to sequence nonspecific effects. Finally, the combination of ISIS 5132 with either carboplatin or taxol showed enhanced growth inhibition, supporting the view that such ASOs may have a more effective clinical role when used in combination with cytotoxic agents. © 2005 Wiley-Liss, Inc.

One of the key pathways regulating mammalian cell growth is the Ras, Raf, MEK and ERK signal transduction pathway, which transmits cell surface receptor-mediated signals to the nucleus via a cascade of specific phosphorylation events.1, 2 The family of Raf protein serine/threonine kinases (consisting of Raf-1, A-Raf and B-Raf) is central to this pathway and is involved in cellular processes that regulate proliferation, differentiation and apoptosis.3 All three isoforms are widely expressed4 and share highly conserved amino-terminal regulatory regions and a carboxy-terminal catalytic kinase domain.5 It has been suggested that they have overlapping but unique regulatory functions.4 Raf-1 is frequently activated in many tumors, either through mutation or through overexpression, and is essential for the regulation of cellular proliferation. Since overexpression of Raf-1 is associated with reduced survival in ovarian cancer,6, 7 we have explored strategies to reduce its expression.

Antisense Oligonucleotides (ASOs) are short, synthetic, single-stranded pieces of DNA designed to hybridize to complementary sequences of a target mRNA, leading to a subsequent reduction in levels of the target protein. The most extensively used ASOs have been phosphorothioate-modified oligodeoxynucleotides, which differ from natural DNA in that one of the nonbridging oxygen atoms in the phosphodiester linkage is substituted with sulfur. ISIS 5132, an ASO targeted to the 3′-end of Raf-1 mRNA, is an example of such a first-generation phosphorothioate ASO and has shown antiproliferative activity in a variety of cell types in vitro and in vivo.8, 9, 10 In clinical studies, ISIS 5132 has reduced Raf-1 expression in blood cells11, 12 and demonstrated evidence of disease stabilization11, 12, 13, 14, 15 and changes in surrogate markers in both colon cancer (CEA)11, 12 and ovarian cancer (CA125).13 Whilst this suggests that when used as a single agent, ISIS 5132 may have cystostatic effects in some cancer types, this is of limited value in advanced disease where future clinical evaluation is quite likely to involve combination with cytotoxic agents. Further modifications in the sugar moiety have been engineered into these ASOs to generate a variety of second-generation structures that include 2′-O-methoxyethyl-substituted (2′-MOE) ribonucleosides.16 These oligonucleotides have increased RNA binding activity, superior nuclease resistance, greater potency and selective inhibition of mRNA/protein expression but less toxicity than the phosphorothioate-only structures.16 ISIS 13650 is an example of a 2′-MOE ASO targeted against Raf-1 and has the same base sequence as that of ISIS 5132. This ASO has been shown to have higher RNA-binding affinity than ISIS 5132 and inhibits Raf-1 at a lower IC50.16

An alternative strategy of mRNA knockdown, termed RNA interference (RNAi), has been developed recently, following the observation that double-stranded RNAs (dsRNAs) elicit potent degradation of complementary RNA sequences.17 The active component of this pathway (small interfering RNAs or siRNAs) can be chemically synthesized and designed to target specific mRNA sequences.

We have previously demonstrated a growth-regulatory role for Raf-1 in ovarian cancer models.6, 7 In the present study, we extend these investigations by comparing the effects of first-generation ASOs with second-generation ASOs and also with siRNAs targeted against Raf-1 mRNA. We demonstrate the importance of a high level of transfection for effective target mRNA reduction and blockade of downstream signaling resulting in growth inhibition. Finally, we have evaluated the value of combining ISIS 5132 with cytotoxic agents.

Material and methods

Cell lines

SKOV-3 cells were obtained from the American Type Culture Collection (Manassas, VA) and routinely grown as monolayer cultures in RPMI 1640 media supplemented with 10% heat-inactivated fetal calf serum (FCS) and 100 IU/ml penicillin/streptomycin in a humidified atmosphere of 5% CO2 at 37°C. The PE01 cell line was derived within the Edinburgh Medical Oncology Unit.18 Growth inhibition experiments were set up using log-phase cells seeded into 24 well tissue culture plates (1 × 104 in 1 ml) and incubated to reach 40–60% confluence after 2–3 days, prior to exposure. Cells for flow cytometric analysis or Western blotting were similarly set up in 60 mm diameter petri dishes and 75 cm2 flasks, respectively.

Antisense oligonucleotides

First-generation (ISIS 5132) and second-generation (ISIS 13650) human Raf-1 ASOs were obtained from Isis Pharmaceuticals (Carlsbad, CA), along with first- and second-generation mismatch oligonucleotides (ISIS 10353 and ISIS 16971), respectively. Specificity throughout the Raf family was assessed with the aid of second-generation ASOs targeted against human B-Raf (ISIS 15344). All oligonucleotide sequences are shown in Table I. Media was removed prior to oligonucleotide exposure and cells were washed (×2) with PBS before adding 250 μl of “Optimem” media containing lipofectin (Gibco-BRL; 6 μl/ml). Oligonucleotides (25–200 nM) were then added from 50 μM stock solutions. Cells were incubated at 37°C for 3 hr in the absence of serum, washed with PBS (×2), replenished with RPMI/10% FCS and replaced in the incubator for the remainder of the time course. Cells were either trypsinized from 24 well trays and counted using a Coulter Counter “Z2” or lysed from 75 cm2 flasks for western blot analysis.

Table I. Oligonucleotide Sequences
Target proteinIdentifierOligonucleotide sequence
Mismatch (1st generation)ISIS 10353TCC CGC GCA CTT GAT GCA TT
Mismatch (2nd generation)ISIS 16971TCA CAT TGG CGC TTA GCC GT

siRNA oligoribonucleotides

Four target DNA sequences were identified and paired sequences designed before obtaining from Qiagen-Xeragon (Germantown, MD). All oligoribonucleotide-paired sequences are shown in Table II. To each oligoribonucleotide tube was added 1 ml of the provided buffer (100 mM potassium acetate, 30 mM HEPES-KOH, 2 mM magnesium acetate, pH 7.4) to yield a 20 μM stock solution. Tubes were then heated to 90°C for 1 min before being incubated at 37°C for 60 min. Samples were aliquoted and stored at −20°C. Media was removed and cells were washed (×2) with PBS before adding 250 μl of optimem media containing lipofectamine (Gibco-BRL; 6 μl/ml), which had been preincubated with siRNAs for 15 min at 37°C. Cells were incubated at 37°C for 4 hr and then RPMI/15% FCS was added (125 μl to give a final concentration of 5%) and replaced in the incubator for the remainder of the time course. Cells were again either trypsinized and counted or lysed for Western blot analysis.

Table II. siRNA Sequences
PairSense sequenceAntisense sequenceTargeted protein sequence

Western blotting

Following exposure to oligonucleotiodes in 75 cm2 flasks, SKOV-3 cells were grown to 70% confluence (48 hr), washed with PBS (×2) and lysed in 600 μl ice-cold hypotonic lysis buffer (50 mM Tris-HCl (pH 7.5), 5 mM EGTA (pH 8.5), 150 mM NaCl, 1% Triton X-100, 2 mM sodium orthovanadate, 50 mM sodium fluoride, 1 mM phenylmethanesulfonylfluoride, 10 μg/ml leupeptin, 10 μg/ml aprotinin and 10 mM sodium molybdate). Lysates were centrifuged at 4°C (6 min at 16,000g) in a microcentrifuge and protein concentrations of supernatants determined using the Bio-rad Protein Assay Kit (Bio-rad, Richmond, CA). Cell lysates (30 μg) were resolved on 10 or 12% SDS-PAGE and then transferred electrophoretically overnight onto Immobilon-P membranes (Millipore, Bedford, MA). After transfer, membranes were blocked with 1% blocking agent in TBS (20 mM Tris-HCl, 137 mM NaCl, pH 7.5) before probing overnight at 4°C with the appropriate primary antibody: anti-Raf1 (R19120, Transduction Laboratories, Lexington, KY) or anti-pERK (9101, Cell Signaling). Immunoreactive bands were detected using enhanced chemiluminescent reagents (1520709, Roche) and Hyperfilm ECL film (Amersham, Buckinghamshire, UK).

Cell cycle analysis

Flow cytometric DNA analysis of treated cells was carried out using methodology described elsewhere.19 After trypsinization, cells were resuspended in 100 μl of citrate buffer and stored at −20°C prior to flow cytometric DNA analysis using a FACSCalibre flow cytometer (Becton Dickinson). Analysis was carried out using Modfit software.

Annexin V assay

SKOV-3 cells were treated with both first- and second-generation antisense and mismatch oligonucleotides, and apoptosis was measured using the TACS Annexin V-FITC kit (R & D Systems), following the prescribed protocol.

Bromodeoxyuridine (BrdU) incorporation

Pretreated cells were incubated in RPMI containing 5′-fluoro-2′-deoxyuridine (20 μM) and 5′-bromo-2′-deoxyuridine (20 μM) for 20 min at room temperature, washed in PBS and transferred back to RPMI/FCS for further 30 min before trypsinizing. BrdU incorporation and staining was performed by washing 2 × 106 cells free of DMSO and resuspending in ethanol (−20°C) on ice for 30 min. Tubes were aspirated and resuspended in Pepsin (0.4 mg/ml PBS; 0.1 M HCl) for 45 min at 37°C, resuspended in 2 ml detergent (0.5% Triton X-100 in 4 N HCl) for 30 min in the dark and washed twice in 0.1 M sodium tetraborate, pH 8.5 (2 ml). Two further washes were then carried out in PBS/0.5% Tween 20 before resuspending in PBS/1% FCS/0.5% Tween 20 (50 μl). Staining was carried out by adding anti-BRDU/FITC (20 μl) for 30 min in the dark before washing in PBS/1% FCS/0.5% Tween 20 and finally resuspending in propidium iodide (0.1 mg/ml in PBS; 0.9 ml). Samples were then analyzed by flow cytometry.

Antisense fluorescence uptake studies

The ability of SKOV-3 cells to incorporate antisense molecules was assessed using FAM-labeled ISIS 5132. Cells were plated out in petri dishes and exposed to FAM-5132, as described above, before being harvested and resuspended in PBS (0.5 ml). Fluorescence (FL1) was then measured in 20,000 cells, using a Becton Dickinson FACSCalibur flow cytometer.

Carboplatin and taxol combination studies

SKOV-3 and PE01 cells were seeded out into 24 well plates and subsequently exposed to carboplatin (1–20 μM) or taxol (0.1–0.6 nM) alone, together or in combination with ISIS 5132 (50 nM). Cells were trypsinized and counted on day 5 by cell counter.

Results and discussion

To demonstrate sequence specificity between ASOs targeted against Raf-1 and mismatch oligonucleotides, SKOV-3 cells were exposed to either first- (ISIS 5132) or second-generation (ISIS 13650) Raf-1 ASOs, and Raf-1 expression was assessed by Western blotting (Fig. 1a). Both ASOs resulted in partial or complete protein knockout when used at concentrations between 25 and 200 nM. Targets such as total ERK or actin were unaffected at these concentrations (data not shown). In contrast, no inhibition of Raf-1 protein expression was seen with either of the respective first (ISIS 10353) and second-generation (ISIS 16971) mismatch sequences. A second-generation ASO targeted against B-Raf similarly had no effect on Raf-1 expression (Fig. 1b)—it did however reduce expression of its own target isoform B-Raf (data not shown). This finding confirms sequence specificity. The reduction of Raf-1 expression using Raf-1 ASO treatment was accompanied by a dose-dependent inhibition of cellular growth, which was not seen with either a B-Raf or mismatch targeted ASOs (Fig. 1c). Reductions in Raf-1 protein levels were also obtained using a siRNA approach. Four siRNAs targeted against Raf-1 were designed (Table II) and their effects assessed against SKOV-3 cells. Treatment with all 4 siRNAs resulted in reduction of Raf-1 protein expression, with 3 of the siRNAs (sequences 2–4) producing more complete inhibition of expression (Fig. 1d). These 4 siRNAs also produced growth inhibition (Fig. 1e). These results substantiate the association of Raf-1 inhibition with reduced growth and demonstrate that both ASO and siRNA strategies can selectively and effectively reduce Raf-1 protein expression.

Figure 1.

Inhibition of Raf-1 by ASOs or siRNA in SKOV-3 cells. (a) Cells were treated with Raf-1 ASOs or control mismatch sequences. Both first- and second-generation ASOs (ISIS 5132 and ISIS 13650) produced a dose-dependent reduction in Raf-1 protein levels over the range 25–200 nM. In contrast, first- (ISIS 10353) and second-generation (ISIS 16971) mismatch control ODNs showed no evidence of reduced expression. (b) Sequence specificity was confirmed using second-generation antisense oligonucleotides targeted against Raf-1 (ISIS 13650), B-Raf (ISIS 15344) or mismatch oligonucleotide (ISIS 16971). Only the Raf-1 ASO resulted in a marked reduction of Raf-1 protein expression. (c) Cells were exposed to Raf-1 (ISIS 13650), B-Raf (ISIS 15344) or mismatch (ISIS 16971) oligonucleotides for 72 hr and cells were then counted by Coulter Counter. Only Raf-1 ASO resulted in growth inhibition. (d) Raf-1 protein levels were also successfully reduced using 4 siRNAs designed to target Raf-1 mRNA when compared with a control sequence (Ambion). Three of the siRNAs (sequences 2–4) produced more complete inhibition of expression. ISIS 5132 (antisense) and ISIS 10353 (mismatch) oligonucleotides are included as controls. (e) Growth inhibition was also demonstrated with all 4 siRNAs.

The effects of ASOs on cell kinetics were next investigated. Analysis of SKOV-3 cells treated with first- and second-generation ASOs (200 nM) was carried out to determine their effects on annexin V positivity (a marker of early-stage apoptosis), cell-cycle distribution and bromodeoxyuridine (BrdU) incorporation. With the exception of the second-generation mismatch oligonucleotide, all oligonucleotides caused a dose-dependent increase in annexin V positivity, following treatment (25–200 nM). These effects were greater with the second-generation ASOs, particularly at the highest concentration. A first-generation mismatch oligonucleotide caused similar changes but of lower magnitude. Treatment with the second-generation mismatch oligonucleotide showed no appreciable differences relative to untreated cells. Cell cycle changes reflected annexin V expression, where both first- and second-generation ASOs again caused major dose-dependent alterations to the distribution of cells throughout the cell cycle (Fig. 2b). Both first- and second-generation ASOs caused a large increase in the number of cells accumulating in S-phase at the expense of those in both G0/G1 and G2/M phases. A first-generation mismatch sequence, however, also caused similar profound changes. Treatment with the second-generation mismatch oligonucleotide showed no appreciable differences in cell-cycle distribution when compared to untreated control cells. Finally, studies investigating BrdU incorporation revealed that increased cellular uptake was seen with both first-generation antisense and mismatch oligonucleotides (Fig. 2c). In contrast, the second-generation ASO showed only slightly enhanced incorporation, following treatment (200 nM), and the corresponding second-generation mismatch oligonucleotide showed no changes at all, being identical to untreated cells. These results indicate that while Raf-1 inhibition produces marked S phase accumulation and enhanced apoptosis, there are also sequence-independent effects demonstrated in the mismatch sequence, probably linked to the first-generation backbone chemistry that are not evident with the second-generation control. These data support the view that second-generation structures are more specific in their action.

Figure 2.

Characterization of cell cycle effects in SKOV-3 cells after treatment with first- and second-generation Raf-1 ASOs along with respective control oligonucleotide sequences. Analysis was carried out 48 hr after treatment in all cases. (a) Both ASOs showed a dose-dependent (25–200 nM) increase in Annexin V levels—these effects were greatest with second-generation ASOs. A smaller dose-dependent increase was seen with the first-generation mismatch oligonucleotide. Only the second-generation mismatch sequence had no effect whatsoever. (b) Similar treatment also caused an accumulation of S-phase cells at the expense of those in G0/G1 and G2/M phases of the cycle. These effects were most profound using the first-generation ASO compounds (i, iii), which also showed greater growth inhibition. Whilst the first-generation mismatch sequence was as potent as either of the antisense compounds in terms its effect upon cell cycle distribution (ii), the second-generation mismatch sequence (iv) was similar to untreated control cells. (c) Increased BrdU incorporation was seen with both first-generation ASO (i) and mismatch (ii) oligonucleotide sequences whilst the second-generation ASO (iii) showed only slight enhanced incorporation. The second-generation mismatch sequence (iv) was identical to untreated cells.

The impact of cell density on ASO uptake and efficacy was next assessed. Cells were cultured in the absence or presence of lipofectin and uptake monitored using FAM-labelled ISIS 5132 detected by flow cytometry. Fluorescence uptake increased over a 3 hr period in the presence of lipofectin. In contrast, no uptake was seen by 3 hr in the absence of lipofectin (Fig. 3a). Cells incubated with FAM-labelled ISIS 5132 showed a reduction in both the proportion of cells staining and the intensity of cellular staining when grown at a higher density (Fig. 3b). This was associated with profoundly different effects on cell growth, with cells being markedly growth-inhibited at low density while minimal effects were demonstrated at high density (Fig. 3c). Investigation of the amount of reduced Raf-1 expression at these different densities indicated marked inhibition of expression at low cell density, intermediate inhibition at medium densities and little or no inhibition at high cell density. In turn, these effects were associated with reduced inhibition of phospho-ERK as cell density increased (Fig. 3d). These data demonstrate the marked influence of cell density on antisense effect, and suggest that for effective cell signaling blockade and growth inhibition, most cells have to incorporate the oligonucleotide. This may well account for the observed difference reported for antisense efficacy on small and large tumors10 and may be part of the explanation as to why this agent has not demonstrated greater potency in clinical studies. This has led to the view that ASOs are quite likely to be more effective when used in combination with cytotoxic therapies,20 and to address this, we explored the activity of ISIS 5132 in combination with the drugs most extensively used to treat ovarian cancer, carboplatin and taxol. When combined with either agent-alone (Figs. 4a and b) or in combination with both carboplatin and taxol, ISIS 5132 produced an additive growth inhibitory effect (Fig. 4c).

Figure 3.

Impact of cell density on cellular signalling and growth inhibition after treatment with Raf-1 ASOs. (a) Incorporation of FAM-labelled ISIS 5132 into SKOV-3 cells was performed in the absence or presence of lipofectin. Only in the presence of lipofectin was uptake seen to increase over the 3 hr incubation period; negligible uptake was seen in the absence of lipofectin. (b) Both the number of positively staining cells as well as the staining intensity of FAM-labelled cells was greater in cells treated at low density when compared to cells treated at higher densities, demonstrating that oligonucleotide uptake was more efficient at lower cell densities. (c) Growth inhibition following Raf-1 ASO treatment became more profound as cell density at the time of ASO exposure was reduced. (d) Density-related effects were also seen at the protein level where ASOs were unable to abolish Raf-1 and pERK expression at high cell density.

Figure 4.

Growth inhibition of SKOV-3 and PE01 cells by Raf-1 ASO (ISIS 5132) in combination with carboplatin and taxol. (a) Cells were treated with Raf-1 ASO alone or in combination with carboplatin and then counted after 72 hr. Growth inhibition in both cell lines was shown to be greater when ASO treatment was combined with carboplatin exposure. (b) Similar treatment showed growth inhibition to be similarly greater when ASO treatment was combined with taxol. (c) The dual combination of taxol and carboplatin could be further enhanced when exposure was combined with Raf-1 ASO.

In conclusion, these results indicate that both first- and second-generation ASOs can effectively inhibit Raf-1 expression, as can RNAi, but first-generation structures may be associated with more nonspecific growth effects. For effective reduction of target expression (and consequent reduction of downstream signaling), a high percentage of cells require transfection and this might be achieved in vivo by targeting low-volume disease. The presence of a carrier such as lipofectin is critical to the activity of these agents in vitro, and even if cells are transfected in vivo in the absence of a carrier, it is quite likely that approaches such as liposomal encapsulation will enhance drug delivery.21 Finally, because the cytostatic effects produced by ISIS 5132 as a single agent are enhanced by combination with cytotoxic agents, combination strategies should be further evaluated.