Pharmacokinetics, Toxicity, and Efficacy of Ends-modified raf Antisense Oligodeoxyribonucleotide Encapsulated in a Novel Cationic Liposome¹

Raf-1, a cytoplasmic serine and threonine kinase, plays an important role in mitogen and stress-induced signaling response, cell survival, and proliferation (1, 2, 3). Depending on the cell type and stimulus, a number of growth factors, growth factor receptors, cytokines, and antiapoptotic molecules have been identified as components of the Raf-1 signaling pathway(s) (4, 5). Although the classical Ras/Raf/MEK/ERK pathway is generally associated with cell proliferation and survival via engagement of downstream transcription factors (4, 5, 6, 7, 8, 9), growing evidence suggests that the survival effects of Raf-1 may also be a result of its mitochondrial translocation and interaction with the members of the Bcl family (10, 11) or the phosphatidylinositol 3′-kinase/AKT survival module (12). Reports using genetic knockouts of mouse c-raf-1 gene imply an MEK-independent role of Raf-1 in differentiation and survival (13, 14). Raf-1 has been linked with cell cycle regulation via its interaction with the retinoblastoma protein (15) and with cytoskeletal reorganization via its association with vimentin, vimentin kinases, and myosin phosphatase (16, 17). Raf-1 also regulates the transcription of a number of diverse molecules (9, 18, 19). The importance of Raf-1 activity has been demonstrated both in solid tumors and hematological malignancies (2, 6, 20). These and other data showing an association of the expression of exogenous Raf-1 with increased survival against radiation or chemotherapeutic agents (2, 7, 21, 22) imply that a relatively complex balance between the anti- and proapoptotic molecules may be a determinant of the Raf-1-mediated survival response. Antisense therapeutics is an attractive target-specific strategy for cancer drug development (23, 24, 25, 26, 27, 28). We and others have demonstrated that human tumor cells transfected with antisense c-raf-1 cDNA show inhibition of Raf-1 expression, delayed tumor growth, and enhanced sensitivity to anticancer agents (2, 22). Down-regulation of Raf-1 expression via antisense raf oligonucleotides is associated with enhanced radiosensitivity, tumor growth arrest, activation of a proapoptotic molecule caspase-3, poly(ADP-ribose) polymerase cleavage, and apoptosis (20, 29, 30, 31). These observations underscore the importance of Raf-1 as a critical target in cancer therapy.

Liposomal delivery systems for anticancer drugs have demonstrated significant clinical benefits including extended circulation time, reduced drug-related toxicity, and improved efficacy (32, 33, 34, 35, 36). More recently, a variety of lipid formulations have been used for intracellular delivery of nuclease-resistant phosphorothioate, P-ethoxy, or methylphosphonate oligonucleotides (37, 38, 39, 40). Cationic liposomes have been shown to enhance cellular uptake and alter intracellular trafficking of phosphorothioate oligonucleotides. While free phosphorothioate oligonucleotides are localized within endosomes or lysosome-like structures, liposome-complexed phosphorothioates are present in cytosolic and nuclear compartments, thereby reducing the efflux of oligonucleotides (41).

Limited information is available on the in vivo liposomal delivery of nuclease sensitive phosphodiester oligonucleotides. Although anionic liposomes have been shown to protect unmodified oligonucleotides against serum nuclease degradation, understandably these carriers exhibit relatively low oligonucleotide entrapment efficiency (42). To date, preclinical evaluation of nuclease-sensitive, liposome-entrapped antisense oligonucleotides has not been conducted. We have designed novel cationic liposomes for in vivo delivery of a 15-mer antisense oligonucleotide complementary to the translation initiation region of c-raf-1 mRNA, rafAON⁴ (43). The phosphorothioate modification in rafAON is limited to only the terminal base at the 5′ and 3′ ends. Using fluorescence-labeled rafAON, we have demonstrated that rafAON is entrapped in the lipid bilayer of the liposome, facilitating endocytosis via association of the cationic lipids with negatively charged plasma membrane. Plasma pharmacokinetic and tissue distribution studies indicate that although free rafAON undergoes degradation within minutes after i.v. administration in athymic mice, LErafAON is stable in plasma and tissues for up to at least 24–48 h (Ref. 43 and unpublished data). The cationic liposome formulation was also shown to improve the antitumor effects of a fully phosphorothioate antisense raf oligonucleotide complementary to a 3′-untranslated region of c-raf-1 mRNA (44). The goal here was to perform the preclinical studies of ends-modified and nuclease-sensitive rafAON entrapped in liposomes using rodent and primate model systems. The pharmacokinetics, biodistribution, safety, and efficacy data presented here demonstrate that LErafAON is a promising cancer therapeutic agent.

Human prostate carcinoma cells (PC-3) were grown as a monolayer in Improved Minimum Essential Medium (Biofluids Inc., Rockville, MD) supplemented with 10% heat-inactivated fetal bovine serum, 100 μg/ml streptomycin, and 100 units/ml penicillin.

Antisense oligodeoxyribonucleotide sequence (5′-GTGCTCCATTGATGC-3′) directed toward the translation initiation site of human c-raf-1 mRNA (rafAON), the complementary sense strand sequence (5′-GCATCAATGGAGCAC-3′; rafSON), and a mismatch sequence (5′-GTGTTCGACCTATGC-3′; MMAON) were custom synthesized at Hybridon Specialty Products (Milford, MA). The terminal base linkage at the 5′ and 3′ ends of these sequences was modified to a phosphorothioate group. The sequence of the GMP grade product (97% pure) was confirmed by matrix-assisted laser desorption ionization-time of flight mass spectrometry by the manufacturer. Dimethyldioctadecyl ammonium bromide, egg phosphatidylcholine, and cholesterol were purchased from Avanti Polar Lipids (Alabaster, AL). Alternatively, GMP preparations of lyophilized rafAON and lipids were provided by NeoPharm, Inc. (Lake Forest, IL). All other chemicals were reagent grade.

LErafAON formulation was prepared by one of the following procedures. In method 1, lipids (dimethyldioctadecyl ammonium bromide, egg phosphatidylcholine, and cholesterol) were dissolved in chloroform and evaporated to dryness in a round-bottomed flask using a rotary vacuum evaporator in a 37°C water bath for ∼20 min. LErafAON was prepared by hydrating the dried lipid film overnight at 4°C with rafAON solution in normal saline using rafAON to lipid ratio of 1:15 (w/w). The film was dispersed by vigorous vortexing, and the liposome suspension was sonicated for 10 min in a bath-type sonicator (Model XL 2020; Misonix, Inc., Farmingdale, NY). In method 2, the lipids were dissolved in t-butanol and lyophilized. Lyophilized lipids were reconstituted at room temperature with rafAON solution as above. The mixture was vortexed vigorously for 2 min, followed by hydration at room temperature for 2 h, and sonication for 10 min. Alternatively, lyophilized lipids supplied by NeoPharm, Inc., were reconstituted with rafAON solution as above (method 3). LErafAON was stored at 4°C and used within 3 days (mice and rabbit studies) or 8 h after preparation (monkey studies). Control LE-MMAON was prepared as above, and the BLs were prepared in the absence of rafAON.

The amount of rafAON entrapped in liposomes was determined by adding traces of ³²P-end-labeled rafAON to the excess of unlabeled rafAON, and the entrapment efficiency was calculated by scintillation counting as reported earlier (43). Alternatively, the entrapment efficiency and stability of intact rafAON were determined by the denaturing gel electrophoresis method. An aliquot of the LErafAON preparation was designated as the prewash sample. The unentrapped rafAON was removed from the remainder of the preparation by ultracentrifugation at 100,000 × g for 20 min, followed by washing the liposomes twice with normal saline (postwash sample). RafAON was extracted from pre- and postwash samples with phenol-chloroform, and the extracts were electrophoresed on a 20% polyacrylamide/8 m urea gel, followed by electroblotting on to a nylon membrane in 0.5× Tris-borate EDTA buffer at 20 V for 1 h. The blot was probed with ³²P-end-labeled complementary rafSON probe in ExpressHyb solution (Clontech, Palo Alto, CA). The radiolabeled rafSON was generated by 5′-end labeling with γ³²P-ATP using T4 polynucleotide kinase and purification over Chroma Spin-10 columns (Clontech). The autoradiographs were scanned and the rafAON band (15-mer) quantified using ImageQuant software (Personal Densitometer; Molecular Dynamics, Sunnyville, CA). The percentage of entrapment efficiency of rafAON in liposomes was calculated by the formula: (postwash signal/prewash signal) × 100. To determine the stability of LErafAON, freshly prepared sample was stored at room temperature, and the entrapment efficiency was determined at various time points after preparation by the gel electrophoresis method as described above.

Male CD2F1 mice, 20–22 g, and male BALB/c nu/nu mice, 5–6 weeks old, were purchased from the National Cancer Institute (Frederick, MD). Male and female New Zealand white rabbits, 2.3–3.0 kg, were purchased from Covance Research Products, Inc. (Denver, PA). Mice and rabbits were maintained at the AAALAC-accredited Research Resource Facility of the Division of Comparative Medicine, Georgetown University Medical Center. Cynomolgus monkeys (Macaca fascicularis), 1.7–2.2 kg, were purchased from Covance Research Products, Inc., and maintained according to the Food and Drug Administration’s GLP regulations at Covance Laboratory, Inc. (Vienna, VA).

Male CD2F1 mice received multiple i.v. injections of 5.0, 15.0, 25.0, or 35.0 mg/kg LErafAON (n = 5). LErafAON was administered via tail vein on days 1, 2, 4, 5, 7, 8, 10, 11, 13, 14, 16, and 17. Control groups received BLs containing LErafAON equivalent lipid doses or were left untreated. Mice in each group were observed daily and group-weighed at least twice a week. Two mice/group were sacrificed on day 19, 48 h after the last dose and representing the short-term response. The remaining three mice/group were sacrificed on day 32, 15 days after the last dose and representing the long-term response. Blood was collected via intracardiac stick under anesthesia, and tissues from euthanized animals were collected in 10% NBF. Specimens were processed for clinical pathology, including CBC count and serum chemistry and histopathology evaluations. Gross and histopathological findings were graded as minimal, slight, moderate, or severe based on the degree of involvement relative to the UT control.

The two i.v. doses of 3.75 and 6.5 mg/kg in rabbits were selected based on allometric scaling of 15.0 and 25.0 mg/kg doses in mice, respectively. However, the total number of injections in rabbits had to be reduced because of technical limitations. Two male and two female rabbits received eight injections of 3.75 mg/kg LErafAON via marginal ear vein over 12 days (days 1, 2, 4, 5, 8, 9, 11, and 12; cohort I), and two male and two female rabbits received six i.v. injections of 6.5 mg/kg LErafAON over 8 days (days 1, 2, 4, 5, 7, and 8; cohort II). Control groups, one male and one female rabbit, received BLs containing LErafAON equivalent lipid doses, and one male and female rabbit received normal saline on treatment days. The study was terminated 24–72 h after the last dose. Blood samples for CBC and chemistry panel were collected on day 1 from the auricular artery (predose) and before euthanasia under sedation through cardiac stick. Blood samples for coagulation profile (prothrombin time, partial thromboplastin time, and fibrinogen) were collected within 30 min after injection on day 12 in cohort I and within 30 min after injection on days 5 and 8 and before euthanasia in cohort II. Blood specimens for complement activity (CH50) were obtained within 30 min after injection on day 12 in cohort I and before euthanasia in cohort II. Various tissues were collected in 10% NBF for histopathology. Gross and histopathological findings were graded as minimal, slight, moderate, or severe based on the degree of involvement relative to the normal saline group.

Monkeys received nine slow bolus injections of 3.75 mg/kg (three males) or 6.25 mg/kg LErafAON (three males and three females) administered via saphenous vein over 15 days (days 1, 2, 4, 5, 8, 9, 11, 12, and 15). Control group received BLs containing a lipid dose equivalent to 6.25 mg/kg LErafAON on treatment days (three males). Blood samples for CBC and chemistry panel were obtained from femoral vein before initiation of dosing and before study termination. Blood specimens were also collected on day −7 and at 10–15 min after dose on days 5, 12, and 15 for coagulation tests (prothrombin time and activated partial thromboplastin time) and complement marker evaluations (CH50, C3a, and Bb). Animals were anesthetized and exsanguinated 24 h after the last dose, and various tissues were collected in 10% NBF for histopathology. Pathological findings were graded relative to the vehicle control.

Plasma pharmacokinetic analysis was performed in male CD2F1 mice and in male BALB/c nu/nu mice bearing a s.c. PC-3 tumor. Mice received injections of 30.0 mg/kg single i.v. bolus dose of LErafAON via tail vein. Blood samples were obtained via cardiac puncture under anesthesia in microtainers containing sodium heparin as an anticoagulant at predose and at 5, 15, and 30 min and 1, 2, 4, 8, 24, and 48 h after LErafAON administration. Athymic mice were euthanized, and liver, spleen, kidney, lung, heart, and PC-3 tumor tissue were rapidly excised, rinsed in ice-cold normal saline, and snap frozen on dry ice. Blood samples were centrifuged at 3000 rpm for 10 min at 4°C to separate the plasma. The plasma and tissue samples were stored at −80°C until additional analysis.

RafAON concentrations in plasma, and normal and tumor tissue samples were determined as described earlier (43). Briefly, rafAON was isolated from plasma samples using the phenol-chloroform extraction method and from tissue samples using a DNA extraction kit (Stratagene, La Jolla, CA). The rafAON concentration standards were prepared by adding known amounts of rafAON in blank plasma or blank tissue samples, followed by extraction. The extracts were analyzed by gel electrophoresis, blotting, and hybridization as detailed above for entrapment efficiency. A 10–50-fold excess of the probe was used to ensure saturation of all bands. The autoradiographs were scanned and quantified using ImageQuant software (Molecular Dynamics), and the amounts of rafAON in various samples were calculated by comparison to standards.

Plasma pharmacokinetic parameters were assessed by standard methods (43, 44). The elimination rate constant (β) was calculated from the linear regression analysis of plasma concentration-time curve. The area under the curve (AUC_0→∞) was calculated using the linear trapezoidal method with extrapolation of the terminal phase to infinity (C_last/β), where C_last is the last measured concentration. Other parameters calculated were: total body clearance (Cl) = dose/AUC; volume of distribution (Vd) = Cl/β; and elimination half-life (t_1/2β) = 0.693/β.

Monkeys received a slow bolus injection of LErafAON via saphenous vein (3 males, 3.75 mg/kg; 3 males and 3 females, 6.25 mg/kg). Heparinized blood samples were obtained from each monkey at following time points: predose and at 5, 10, and 30 min and 1, 2, 4, and 24 h after LErafAON administration. Blood tubes were centrifuged at 3000 rpm, and plasma was separated and stored at −80°C until additional analysis as described above. Predose specimens spiked with known amounts of rafAON were used as concentration standards.

Logarithmically growing PC-3 cells were injected s.c. (5 × 10⁶ cells) in the left flank region of male BALB/c nu/nu mice. Tumors were allowed to grow to a mean tumor volume of ∼60 mm³ before initiation of treatment. Tumor volumes were determined from caliper measurements of the three major axes (a,b,c) and calculated using abc/2, an approximation for the volume of an ellipse (π abc/6). Tumor-bearing mice were randomly divided into various treatment groups (n = 6–8). LErafAON and LErafAON + IR treatment groups received 10 i.v. injections of 25.0 mg/kg/dose LErafAON over 14 days (day 0, 1, 3, 4, 6, 7, 9, 10, 12, and 13). Radiation was delivered to the tumors of mice in the IR alone and LErafAON + IR treatment groups once daily from day 7 through day 11 using a [¹³⁷Cs] irradiator (J. L. Shepard Mark I; 2.37 Gy/min, 3.8 Gy/day) as described earlier (44). In the combination group, the two treatments were given 6–8 h apart. Control groups received i.v. doses of BLs or liposome-entrapped mismatch oligo (LE-MMAON) on the same dosing schedule as LErafAON or were left untreated.

Tumor sizes were monitored once or twice weekly. Individual tumor volume was calculated as the percentage of pretreatment tumor volume (day 0, the first day of dosing; 100%) and the mean tumor volume (percentage of initial) ± SE for each treatment group was plotted.

Normal and tumor tissues of PC-3 tumor-bearing BALB/c nu/nu mice were excised within 6–12 h after the last LErafAON treatment given as above. Raf-1 protein expression was examined in whole cell extracts from tissues by immunoprecipitation with polyclonal anti-Raf-1 antibody (Santa Cruz Biotechnology, Santa Cruz, CA), followed by immunoblotting with monoclonal anti-Raf-1 antibody (Transduction Labs, Lexington, KY) using standard procedures as detailed earlier (3, 17, 43, 44, 45). Raf-1 protein levels were quantified using ImageQuant software (Molecular Dynamics).

Student’s t test was performed to determine the significance of changes in clinical pathology values obtained in CD2F1 mice treated with four different of doses of LErafAON or BLs versus untreated control mice. Analysis included both short-term and long-term treatment groups of these mice. ANOVA (one-way ANOVA) was used to compare the clinical pathology values observed in male monkeys treated with two different doses of LErafAON versus BLs. One-way ANOVA was also performed to determine the statistical significance of a treatment-related change in mean tumor volume in BALB/c nu/nu mice.

The liposome entrapment efficiency of LErafAON at 0 h after preparation was found to be >85.0% (four independent experiments, cumulative n = 14). A reduction in the hydration time from 2 h to 30 min had no effect on the entrapment of rafAON (percentage of entrapment efficiency, 89.0 ± 6.9 n = 3). Stability of entrapment was determined by comparing the entrapment efficiency of the LErafAON formulation stored at room temperature for 8 h with its freshly prepared counterpart. This formulation was stable at room temperature for up to at least 8 h after preparation (percentage of stability: hydration time 2h, 93.0 ± 7.2, n = 5; hydration time 30 min, 94.1 ± 4.2, n = 2). In other studies, particle size analysis of LErafAON was carried out using Nicomp particle sizer (Particle Sizing Systems, Inc., Santa Barbara, CA). The vesicle size was found to be 467.2 ± 72.0 nm (n = 2).

The toxicity of LErafAON was evaluated after i.v. administration in CD2F1 mice, NZW rabbits, and cynomolgous monkeys as described in “Materials and Methods.” No morbidity or mortality was observed in any of the groups during the course of the study.

CD2F1 mice received a total of 12 i.v. injections of LErafAON at a dose of 5.0, 15.0, 25.0, or 35.0 mg/kg administered over 17 days. The cumulative dose of LErafAON administered in these groups was 60.0, 180.0, 300.0, and 420.0 mg/kg, respectively (n = 5). Control groups received BLs at LErafAON equivalent lipid doses (BL) or were left untreated (n = 5). Mice in untreated control group gained weight throughout the duration of study with a maximum group weight gain of 18.5% at the time of sacrifice (day 32, 15 days after last dose). At 5.0 mg/kg, BL and LErafAON groups exhibited group weight loss of 0.5 and 0.2%, respectively, by day 7, with return to weight gain by day 10. At 15.0 mg/kg, BL group had a moderate group weight gain, whereas LErafAON group showed a mild weight loss of 3.3% on day 12, with weight gain thereafter. Mice receiving 25.0 and 35.0 mg/kg BL had a maximum weight loss of 6.0 and 10.5%, respectively, by day 18, followed by weight gain by day 25. Mice treated with 25.0 and 35.0 mg/kg LErafAON had a maximum weight loss of 11.0 and 13.4%, respectively, by day 18 and weight gain by day 21. Weight loss did not exceed 14% in any of the groups, and all groups demonstrated positive growth curves within 5 days after treatment cessation.

No hematological abnormalities were noted in any of the groups of mice except for a dose-dependent increase in total neutrophil count with the total WBC count remaining in the normal range. This effect was seen in BL and LErafAON groups as compared with UT controls and was attributed to liposome administration. Within 48 h after the last dose, serum chemistries revealed dose-related elevations in liver enzymes (Table 1, ALT, short-term response). However, these elevations were not statistically significant as compared with control untreated mice. Additionally, the total bilirubin or alkaline phosphatase remained unchanged as compared with control untreated mice (data not shown). Fifteen days after the last dose, AST was significantly elevated in BL and LErafAON groups, and ALT was significantly elevated in only the highest dose LErafAON group (Table 1, AST/ALT, long-term response). The total bilirubin or alkaline phosphatase level remained unchanged as compared with control untreated mice (data not shown).

Changes in liver enzymes were manifested histologically. Mild to moderate diffuse hepatitis with minimal hepatocyte necrosis was observed in LErafAON and BL groups of mice as compared with untreated mice. These effects appeared to be dose related and relatively more pronounced in the LErafAON groups (Table 1). No clear difference was observed between the short-term and long-term responses of BL or LErafAON animals. In addition, mild to moderate splenic enlargements were seen in mice receiving mid to high doses of BL and LErafAON as compared with UT controls. However, no clinical chemistry or histological correlation was found. No other gross or microscopic pathology was observed in any of the groups.

The cumulative i.v. doses of LErafAON administered in rabbits were 30.0 and 39.0 mg/kg corresponding to allometric mouse equivalent doses of 120.0 and 156.0 mg/kg, respectively. Control groups received BL at LErafAON equivalent lipid doses or normal saline. Evaluations of hematology, serum chemistry, coagulation profile, serum complement, and gross and microscopic pathology revealed no drug-related toxicity as compared with BL or normal saline-treated control groups (data not shown).

The cumulative i.v. doses of LErafAON administered in monkeys were 33.75 and 56.25 mg/kg corresponding to allometric mouse equivalent doses of 135.0 and 225.0 mg/kg, respectively. The control group received BLs at 6.25 mg/kg LErafAON equivalent lipid dose. Hematological and serum chemistry evaluations were unremarkable except for mild increase in monocytes in all groups from day 7 to day 15 and attributed to liposome administration. Coagulation profiles of LErafAON and BL groups were comparable with their respective baseline values. Interestingly, complement activation was documented in all treatment groups (Fig. 1). A 62–74% decrease in the CH50 was observed on days 5 and 12 in all treatment groups compared with their respective baseline values (day −7). Additionally, increases in complement products C3a (∼3-fold) and Bb (∼5–12-fold) were noted in all treatment groups on day 5 and/or day 12 compared with their respective baseline values (day −7). By day 15, complement levels appeared to return to the baselines. Statistical analysis showed that transient changes in complement profile in monkeys treated with LErafAON versus BLs were identical and appear to be related to the liposome treatment per se. Histomorphological changes characterized by the presence of vacuolated macrophages were observed in spleen, liver, and kidneys of monkeys in LErafAON and BL groups and are likely to be associated with liposomes. In addition, mild chronic inflammation was recorded in liver and kidney tissues in all treatment groups.

After i.v. administration of 30 mg/kg LErafAON, the peak plasma concentration of 10.17 ± 2.62 μg/ml was achieved in BALB/c nu/nu mice bearing PC-3 tumor, and intact rafAON (15-mer) could be detected in plasma for up to 48 h (Fig. 2). The decrease in plasma concentration of LErafAON followed a biexponential pattern with initial distribution half-life (t_1/2α) of 1.09 ± 0.53 h and a terminal half-life (t_1/2β) of 26.95 ± 0.99 h, and area under the plasma concentration-time curve (AUC) was 10.12 ± 5.37 μg.h/ml (Table 2). Comparable pharmacokinetic parameters were observed in CD2F1 mice treated with a similar dose of LErafAON (Table 2).

The tissue distribution profile of LErafAON in PC-3 tumor-bearing BALB/c athymic mice is presented in Fig. 3. Liver, spleen, and kidneys were the preferential sites of rafAON accumulation with intact rafAON detectable up to at least 48 h (Fig. 3, bottom panel). Heart and lungs also showed substantial uptake of rafAON up to 48 h. Remarkably, intact rafAON was also detected in PC-3 tumor tissue for up to at least 48 h (Fig. 3, top and bottom panels). A peak rafAON concentration of 5.56 ± 3.47 μg/g tumor tissue was obtained at 30 min to 1 h after dose (n = 2). The tissue AUC_{(0 → 48 h)} (μg·h/g) values were: liver, 2248.26; spleen, 2031.00; kidneys, 796.00; lungs, 423.30; heart, 111.63; and PC-3 tumor, 25.26 ± 6.38.

The plasma pharmacokinetic parameters of LErafAON after a single bolus i.v. dose of 3.75 or 6.25 mg/kg in cynomolgus monkeys are shown in Table 3. The peak plasma concentrations achieved were 0.85 ± 0.63 μg/ml (n = 3) and 1.25 ± 0.53 μg/ml (n = 6) at 3.75 mg/kg dose and 6.25 mg/kg, respectively. Intact rafAON could be detected up to 24 h (data not shown). The decrease in plasma concentration of LErafAON followed a biexponential pattern with an initial half-life (t_1/2α) of 9.0 min seen at both dose levels. The terminal half-lives (t_1/2β) were 36.39 and 30.36 h at low and high dose, respectively. The AUC increased from 2.35 ± 0.76 μg·h/ml at 3.75 mg/kg dose to 5.73 ± 6.30 μg·h/ml at 6.25 mg/kg dose.

The effect of multiple i.v. injections of LErafAON on Raf-1 protein expression was examined in limited sets of normal tissues from athymic mice and monkeys. Within 24 h after the last dose of LErafAON (25 mg/kg, 10×), Raf-1 protein levels in liver, kidney, and splenic tissues of mice treated with LErafAON were 7.2 ± 6.1, 17.0 ± 3.7, and 4.8 ± 2.3%, respectively, compared with BL-treated group (100%; n = 2; Fig. 4, top panel). BL treatment did not alter Raf-1 expression in these tissues compared with UT controls (data not shown). In monkeys, within 24 h after the last dose of LErafAON (6.25 mg/kg, 9×), Raf-1 protein expression in liver specimens from male and female monkeys was found to be 60.9 ± 14.1 and 56.6 ± 26.6%, respectively, as compared with BL-treated group (n = 3; Fig. 4, bottom panel). A lower dose of LErafAON (3.75 mg/kg, 9×) was found to be ineffective.

Effect of LErafAON alone or LErafAON in combination with IR on tumor growth was examined using hormone-independent human prostate carcinoma (PC-3) models. Fig. 5 shows the antitumor activities of LErafAON, IR, and combination treatments in PC-3 tumor-bearing athymic mice. LErafAON alone caused tumor growth arrest compared with control groups of mice (untreated and BL) for up to 12 days after the last dosing. This was followed by a steady increase in mean percentage tumor volume. Similar observations were made in the IR group. The combination of LErafAON and IR treatments caused a highly remarkable and steady decline in the mean percentage tumor volume (Fig. 5,A). By day 46, the mean ± SE percentage tumor volumes compared with initial (pretreatment) tumor volumes (100%, day 0) were: UT, 870.0 ± 192.0%; BL, 963.0 ± 397.0%; LE-MMAON, 803.0 ± 325%; LErafAON, 230.0 ± 55.0%; IR, 220.0 ± 44.0%; and LErafAON + IR, 21.0 ± 7.0%. Single treatment groups (LErafAON and IR) displayed significant tumor growth arrest versus control UT group (P < 0.001). Most importantly, the LErafAON + IR combination group displayed significantly greater antitumor activity versus either treatments alone or control groups (Fig. 5 B, day 46; P < 0.001), maintaining this response for at least up to day 59 (data not shown). No significant differences were observed between the LErafAON and IR groups. The mismatch antisense group LE-MMAON continued to show significant tumor growth compared with LErafAON group (mean ± SE percentage of initial tumor volume, day 52: LE-MMAON, 1209 ± 577%; LErafAON, 390 ± 106%). The data demonstrates that a combination of LErafAON and IR treatments is highly efficacious in a hormone-refractory prostate cancer.

Within 6–12 h after the last LErafAON treatment, Raf-1 expression in tumor tissue was 23.4 ± 15.4% of the UT control group. In PC-3 tumors from mice treated with a combination of LErafAON and IR, Raf-1 expression was 13.2 ± 2.7% of the UT control group. Raf-1 expression in PC-3 tumors of mice treated with IR, LE-MMAON (MM), or BL was comparable with UT control PC-3 tumors (Fig. 5 C). These data indicate that rafAON selectively inhibits Raf-1 protein expression in vivo.

Raf-1 protein kinase is an established molecular target for cancer drug discovery (2, 20, 46). Although recent drug discovery efforts are directed at the structure-based design of suitable peptides or chemicals that may interfere with Raf-1 signaling (47, 48), the antisense oligonucleotide approach is at the forefront of clinical endeavors to block Raf-1 expression and activity (26, 49). With the advent of high-throughput technologies, the number of clinically relevant biological molecules is fast growing; however, delivery systems for safe and effective targeting of such agents to advanced lesions are relatively limited (50, 51, 52, 53). This study offers a novel systemic liposomal oligonucleotide delivery system. Our data demonstrate favorable pharmacokinetic and biodistribution profiles of liposomal formulation of rafAON as compared with rafAON reported earlier (43), tumor tissue uptake of intact rafAON, sequence-specific inhibition of Raf-1 protein in normal and tumor tissues, and rafAON-associated enhanced radiation response in a human hormone-refractory prostate tumor model.

The antisense oligonucleotide chemistry, validity of the antisense sequence-specific inhibition of gene expression, and pharmaceutical application of the antisense technology have generated considerable interest in the past decade (23, 24, 25, 26, 27, 28). Earlier, we have established the sequence specificity of the 15-mer rafAON oligonucleotide used in this study (29, 43). In vitro and in vivo data using control sense oligonucleotide (29, 43) or a mismatch antisense oligonucleotide (present study) have established the rafAON sequence as an antisense oligonucleotide capable of inhibiting Raf-1 protein expression in cells (29, 43) in normal and tumor tissues of athymic mice (Ref. 43 and present study), and in liver specimens of monkeys. The specimens obtained from BL or untreated animals served as additional controls. The LErafAON formulation developed here shows high entrapment efficiency of rafAON (>85%), and LErafAON is stable at room temperature for at least up to 8 h. Toxicology studies indicated that cumulative doses of 420.0, 156.0, and 56.0 mg/kg LErafAON are well tolerated in mice, rabbits, and monkeys, respectively. A transient increase in complement activity observed in monkeys is restored to the baseline value during the course of treatment with BLs or LErafAON (Fig. 1). Unlike phosphorothioate oligonucleotides where complement activity has been related to plasma concentrations of the oligonucleotide (54), rafAON per se does not appear to contribute to changes in complement activity observed in monkeys. These changes are likely to be attributed to liposomes rather than rafAON. Additional investigations should address the mechanism of liposome-mediated changes in complement activity. All other toxicology parameters observed in monkeys are clinically insignificant. Pharmacokinetic studies in mice and monkeys indicated that intact rafAON is detectable in plasma for up to at least 24 h. The intact rafAON could be detected in all normal and tumor tissues examined in mice for up to at least 48 h. These data demonstrate that liposomes entrapment protects circulating rafAON and clearly enhances the biodistribution profile of these oligonucleotides. On the basis of our observations of normal clinical chemistry and histopathology in monkeys and a concomitant inhibition of Raf-1 expression observed in liver specimens from the same monkeys, it appears that inhibition of Raf-1 per se does not cause any toxicity in normal tissues. Furthermore, antitumor efficacy studies in athymic mice indicated that 50% inhibition of Raf-1 expression in prostate tumor tissues is sufficient to cause tumor regression in conjunction with radiation treatment (Fig. 5).

The knowledge of the mechanisms of radiation-induced signal transduction and biological response has increased enormously during the past decade (3, 7, 45, 55, 56, 57, 58, 59, 60); however, a limited number of radiation/chemosensitizers are currently approved for clinical use by the United States Food and Drug Administration (61, 62). Systemic administration of LErafAON treatment also results in enhanced radiation sensitivity in human head and neck squamous cell carcinoma (SQ-20B) and pancreatic carcinoma models (AsPC-1; unpublished data). In addition, LErafAON i.v. treatment causes tumor growth inhibition and chemosensitization of human lung (A549), breast (MDA-MB 231), prostate (PC-3), and pancreatic tumor xenografts grown in athymic mice (AsPC-1 and COLO-357; Refs. 63, 64, 65). Recently, the clinical Phase I studies of LErafAON as a monotherapy and LErafAON in combination with radiation therapy have been initiated (66, 67, 68, 69). The present report has a broad application in the area of evolving antisense technology. Our data demonstrate that cationic liposome entrapment is a safe and effective mode of in vivo delivery of intact antisense oligonucleotides. This delivery system eliminates the need for a significant modification of the oligonucleotide chemistry and preserves the two salient features of the antisense drug: potency and target selectivity. The cationic liposomes developed here may be used to deliver more than one type of antisense oligonucleotide or decoy oligonucleotides (70) to validate components of a signaling pathway and associated biological response.

We thank Dr. Ralph Weichselbaum for the SQ-20B cells and Dr. Lisa Portnoy, Laura Rutter-Call, Carrie Silver, and Amy Durham for technical assistance. In-life monkey GLP procedures were conducted by Dr. Michael D. DuVall at Covance Laboratories, Inc. PC-3 cells were obtained from the Tissue Culture Shared Resource facility of the Lombardi Cancer Center (Georgetown University Medical Center). Mice and rabbit clinical pathology and histopathology tests were performed by Antech Diagnostics, and data were evaluated by a board certified veterinary pathologist (Pathology Associates International); rabbit complement activity tests were performed by Laboratory Corporation of America; and monkey complement markers were analyzed by Anilytics, Inc. The liposomal formulation and its use reported in this study are patented (United States Patent Nos. 6,126,965 and 6,333,314).