Members of the signaling pathways that confer hyperproliferative and anti-apoptotic potential to tumor cells are currently exploited as possible anti-neoplastic target molecules. A promising strategy to specifically suppress the expression of target proteins in tumor cells utilizes the ability of short double-stranded RNAs, termed small interfering RNAs (siRNAs), to induce degradation of the respective RNAs by virtue of their sequence complementary to the target RNA by a process termed RNA interference.1, 2 In this process, the antisense strand of the siRNAs directs the RNA-induced silencing complex to cleave its target RNA bearing a complementary sequence.3 Recently, siRNA-mediated gene silencing has been demonstrated in nonhuman primates,4 further reinforcing the interest in exploring this potent and specific gene-silencing mechanism for tumor therapy.
Evidence has accumulated that Polo-like kinase 1 (Plk1), a serine/threonine kinase, which regulates mitotic progression in mammalian cells, might be a suitable target for an anti-tumor therapy. Plk1 is required for centrosome maturation, bipolar spindle formation and chromosomal segregation, and it is associated with microtubules and centrosomes at various stages of mitosis.5, 6 Plk1 is overexpressed in a broad range of human tumors, and its level of expression inversely correlates with patient survival.7, 8, 9 Inhibition of Plk1 blocks cell cycle progression and increases apoptosis of tumor cells.6, 10, 11, 12, 13, 14, 15, 16, 17, 18 In contrast, normal cells are less affected by inhibition of Plk1 than cancer cells,12, 13, 14, 15, 16 supporting the suggestion that Plk1 might be a feasible cancer therapeutic target.
siRNAs targeting Plk1 inhibit proliferation of cultured tumor cells, and they reduce tumor progression in mouse models of prostate and bladder cancer.17, 18 Nevertheless, major problems of a systemic application of siRNAs are insufficient delivery into the target cells, a short half life of the siRNAs in vivo because of their degradation and renal elimination, and off-target effects.19 Whereas siRNA appears to be quite stable within cells,20 our recent data indicate that siRNAs are degraded by RNAse A-like enzymes in human serum, resulting in a loss of silencing activity.21 Our study showed that serum caused a loss of the anti-neoplastic potential of a siRNA directed against Plk1, which was prevented by inhibition of RNAse A-like enzymes.
Material and methods
siRNAs and annealing
The RNAs were obtained as single strands from Biomers (Ulm, Germany). The single stranded RNAs were resuspended in annealing buffer (50 mM Tris/HCl, pH 7.5; 100 mM NaCl). For annealing, 2 complementary strands were mixed, heated to 93°C for 3 min and allowed to cool down to room temperature within 1 hr. The sequences (sense strand) of the siRNAs used in the present study were: 5′-agaccuaccuccggaucaa-3′ (Plk1-siRNA1) and 5′-aacuggguaagcgggcgca-3′ (control siRNA). The siRNAs contained additional two-base overhangs (dTdT) at the 3′-ends.
Preincubation of siRNA in human serum
Plk1-siRNA (1.5 μg), which was diluted in 4 μl of TE-buffer, was incubated in 15 μl of human serum (prepared from healthy donors) for various duration at 37°C. Where indicated, 40–70 units of RNAseOUT (Invitrogen, Carlsbad, CA) were added to the serum prior to the addition of the siRNA. The extraction of the duplex RNA was performed as described recently.21
RNA gel electrophoresis
siRNA samples were electrophoretically separated on 20% native polyacrylamide gels, followed by their visualization by silver staining.
Cell culture and transfection
LoVo, Hela and HCT116 cells were obtained from American Type Culture Collection (Manassas, VA). All cell lines were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum (FCS) and antibiotics (Invitrogen). One hour prior to transfection, the cells were switched to serum-reduced medium (Opti-MEM I, Invitrogen). Untreated or serum-preincubated siRNAs were transfected into the cells using Lipofectamine 2000 (Invitrogen) according to the recommendation of the manufacturer. After 6 hr of incubation, the medium was replaced by complete medium.
Protein electrophoresis and immunoblotting
Cells were lyzed in RIPA (50 mM Tris-HCl, pH 7.4, 1% NP-40, 0.25% Na-deoxycholate, 150 mM NaCl, 1 mM EDTA, 1 mM phenylmethylsulfonyl fluoride; 1 μg/ml aprotinin, 1 mM Na3VO4). Immunoblotting of the lysates with anti-Plk1 (Zymed, San Francisco, CA) or anti-β-actin (Sigma, St. Louis, MO) was performed as recently described.22 Antigen-antibody complexes were visualized using horseradish peroxidase-conjugated secondary antibodies (Santa Cruz, Santa Cruz, CA) and enhanced chemiluminescence (Pierce, Rockford, IL).
The number of cells was evaluated by direct cell counting after trypan blue staining.
Determination of the viable cell mass
The cells (5 × 104 cells/well) were seeded in 24-well, flat-bottomed plates. We added 50 μL of 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) solution (5 mg per mL sterile PBS) (Sigma, St. Louis, MO) to the wells, and the cells were incubated for 30 min at 37°C in a CO2 incubator. MTT crystals were dissolved in a solution containing 10% SDS and 0.5% acetic acid in DMSO. The optical density of the samples was determined photometrically at 570 nm wavelength and 630 nm reference wavelength.
Apoptotic cells were assessed by Hoechst 33342 staining (Sigma, St. Louis, MO) as described recently.22
The data presented are means ± standard deviation. Statistical significance (p < 0.05) was calculated using Students unpaired t-test.
Plk1 is a suitable target for anti-neoplastic therapy
Screening of a number of different siRNAs targeting the mRNA of Plk1 for their ability to specifically down-modulate the expression of Plk1 in Hela cells led to the identification of a particularly effective siRNA (Plk1-siRNA1),21 which was therefore used in the subsequent experiments.
Transfection of 3 different cancer cell lines (Hela, LoVo and HCT116) with Plk1-siRNA1 resulted in a strong reduction of Plk1 expression in each cell line (Fig. 1a). Down-modulation of Plk1 expression by Plk1-siRNA1 was accompanied by a reduction of cell proliferation as revealed by the MTT assay (Fig. 1b) or direct cell counting (Fig. 1c). The effects of the Plk1-siRNA1 on the viable cell mass were most pronounced in Hela cells in which p53 expression is downregulated. Determination of apoptosis by Hoechst staining revealed that the Plk1-siRNA1 drastically increased the rate of apoptosis (Figs. 1d and 1e). The control siRNA only slightly increased the rate of apoptosis. These data support the suggestion that Plk1 is an excellent anti-neoplastic target.
Preincubation of the Plk1-siRNA in human serum causesgradual loss of its anti-tumor activity
To mimic the influence of RNAses from the blood on the anti-tumor activity of the Plk1-siRNA1 during a possible systemic human therapeutic application of siRNAs, Plk1-siRNA1 was preincubated in human serum followed by its extraction from the serum and determination of its anti-tumor activity upon transfection into Hela cells. Preincubation of the siRNA in human serum led to a gradual loss of its ability to down-modulate Plk1 expression (Fig. 2a). When Plk1-siRNA1 was preincubated in human serum for 2 hr, followed by extraction and transfection into the tumor cell lines, no significant reduction of the viable cell mass (Fig. 2b) or increase in apoptosis of the tumor cells were observed (Fig. 2c). Similar data were obtained in human sera prepared from the blood of different healthy donors (data not shown). These results revealed that serum inhibited the anti-tumor activity of Plk1-siRNA1.
An obvious possibility is that the anti-neoplastic activity of the siRNA targeting Plk1 declined upon preincubation in serum is due to the degradation of the siRNA by serum-derived RNAses, leading to a loss of Plk1 silencing activity. Preincubation of the siRNA in serum led to the degradation of the Plk1-siRNA1 into a fragment shorter than 19 bp (Fig. 2d), which possess little or no silencing activity.2 These data indicate that the loss of the biological activity of Plk1-siRNA1 in serum was indeed due to its shortening by serum RNAses.
Inhibition of RNAse A-like RNAses in human serum prevents the loss of anti-tumor activity of Plk1-siRNA
To investigate whether the loss of anti-tumor activity of the Plk1-siRNA1 might be due to the action of RNAse A-like enzymes and could thus be prevented by inhibition of RNAse A-like enzymes, Plk1-siRNA1 was incubated in human serum with or without RNAseOUT, followed by extraction from the serum and transfection into Hela, LoVo and HCT116 cells. As illustrated in Figure 3a, RNAseOUT prevented the serum-induced loss of the ability of Plk1-siRNA1 to reduce the viable cell mass in the 3 different tumor cell lines. These data suggested that RNAse A-like enzymes reduced the anti-tumor activity of the Plk1-siRNA1 in serum and that the anti-neoplastic potential of Plk1-siRNA1 could be protected in serum by inhibition of RNAse A-like enzymes.
RNAseOUT completely prevented the shortening of the siRNA in serum (Fig. 2d) as well as the loss of its ability to down-modulate Plk1 expression in HeLa, LoVo and HCT116 cells (Fig. 3b). In addition, preincubation of the Plk1-siRNA1 in human serum for 2 hr reduced the proapoptotic effect of the siRNA in all cell lines tested (Fig. 3c). The loss of the proapoptotic effect of Plk1-siRNA1 in serum was prevented when the serum-preincubation was performed in the presence of RNAseOUT (Fig. 3c).
Other commercially available RNAse A inhibitors, such as RiboLOCK or RNAsin mimicked the effects of RNAseOUT, and the depletion of RNAse A-like RNAses by a specific antibody had the same effect (data not shown). These data indicate that the loss of the anti-tumor activity of the Plk1-siRNA1 in serum was due to RNAse A-like enzyme-mediated degradation of Plk1-siRNA1.
During a potential anti-tumor application of siRNAs in humans, the siRNAs will have to resist degradation by RNAses present in the serum compartment. Within cells, siRNA appears to be relatively stable.20 Experiments in mice may underestimate the kinetics of inactivation of siRNAs in human serum because siRNAs are more rapidly degraded in human than in mouse serum.21 Thus, the influence of serum during a potential future human therapeutic application of siRNAs can not be estimated solely from experiments in mice. The present study investigated the influence of human serum on the anti-tumor activity of siRNA targeting the mitosis-associated serine/threonine kinase Plk1. The data showed that serum caused inactivation of the anti-neoplastic potential of anti-Plk1-siRNA. However, this could be prevented by inhibition of RNAse A-like enzymes. These findings may facilitate the development of strategies to avoid inactivation of siRNAs in the patient during a potential future anti-tumor therapy with siRNAs.
The stability of siRNAs in serum can be enhanced by chemical modifications.23 However, these modifications often reduce the silencing efficiency. The present study indicates that stabilization of siRNAs can also be achieved by inhibition of siRNA degrading enzymes, thereby avoiding the necessity of stabilizing chemical modifications in the siRNAs.
Plk1 is a promising target for an anti-tumor therapy because it is overexpressed in a broad range of human tumors, its level of expression inversely correlates with patient survival, and its inhibition decreases cell cycle progression and increases apoptosis.6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18 Importantly, there is strong indication that nontransformed cells are less sensitive to Plk1 depletion or inactivation than cancer cells.6, 10, 11, 12, 13, 14, 15, 16, 17 In the present study, we identified a siRNA that potently down-modulated Plk1 expression in various tumor cells, which was accompanied by loss of cell number and viability and increase in the rate of apoptosis. Using anti-Plk1-siRNA, we show that RNAse A-like enzymes present in human serum rapidly diminish the anti-tumor effect of this siRNA. Importantly, the loss of anti-tumor effect of Plk1-siRNA was inhibited when the siRNA had been preincubated in serum in the presence of recombinant RNAse A inhibitor (RNAseOUT), an acidic protein with a molecular weight of ∼52 kDa that inhibits pancreatic type RNAses, such as human pancreatic RNAse, RNase A, RNase B and RNase C.24 The loss of the anti-tumor activity of the Plk1-siRNA1 in serum correlated with its shortening to less than 19 bp, which does not allow effective gene silencing.2 RNAse A inhibition prevented the degradation of Plk1-siRNA1 in human serum. Together, these data indicate that prevention of RNAse A-like enzyme-mediated inactivation of siRNAs may strongly increase the anti-tumor effect of siRNAs.
An important mechanism of unintended off-target effects by siRNAs is activation of the interferon response. Recent data indicate that activation of dsRNA signaling is strongly promoted by 21–27 mer siRNAs when they lacked the 2-nucleotide 3′-overhangs.25 Because inhibition of RNAse A-like enzymes in serum appears to conserve the molecular structure of the siRNA, including the 2-nucleotide 3′-overhangs (present study and21), inhibition of RNAse A-like enzymes might reduce undesired off-target effects of siRNAs resulting from the production of blunt end dsRNA by removal of 2-nucleotide 3′-overhangs from siRNAs.
Despite encouraging progress, there are significant obstacles to resolve on the way to an anti-neoplastic application of siRNAs in humans, which include difficulties with delivery, bio-stability, pharmacokinetics and off-target effects.19 The present study supports the concept that siRNAs targeting Plk1 possess anti-tumor therapeutic potential and that inactivation of the anti-neoplastic potential of this siRNA by RNAse A-like enzymes in serum can be prevented by RNAse A inhibition. This may have important implications for the development of a human anti-tumor therapeutic application of siRNAs.
The authors thank Dr. R.M. Biondi for advice and criticism, and Swantje Martens for advice concerning the statistical analysis of the data. They are indebted to Dr. R. Elez († 26.12.2003).