Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

The enzyme DNA MeTase has been shown to play a key role in normal cellular differentiation as well as in neoplasia. In many cancers, DNA MeTase has been shown to be overexpressed and is implicated as a causal factor in the maintenance of a transformed phenotype (Szyf, 1996). Thus, one therapeutic strategy would be to decrease the extent of DNA methylation by inhibition of DNA MeTase gene expression through the use of antisense oligonucleotides. Cell culture studies have suggested that a phosphorothioate antisense referred to as MT-AS can inhibit DNA MeTase expression in a dose-dependent manner (0.1-20 µM), inhibit growth in soft agar and suppress tumor formation by NCI-H446 small-cell lung cancer cells in immunodeficient mice (Szyf et al., 1996) and by Y1 adrenocortical carcinoma cells in syngeneic LAF1 mice (Ramchandani et al., 1997).

Pharmacokinetic studies of phosphorothioate ODNs have been conducted in mice (Agrawal et al., 1989; Bigelow et al., 1990; Chen et al., 1990; Crooke et al., 1996; Goodarzi et al., 1992; Sands et al., 1994), rats (Cossum et al., 1993; Zhang, 1995a) and nonhuman primates (Agrawal et al., 1995 and Srinivasan and Iversen 1995), as well as in humans (Crooke, 1995; Srinivasan and Iversen, 1995; Zhang et al., 1995b). These investigations have revealed some general pharmacokinetic properties of ODNs, such as relative stability in plasma and the fact that elimination occurs primarily via the liver and kidneys. However, because most pharmacokinetic analyses have been based on radioactivity measurements, they have yielded little understanding of the kinetic characteristics of individual ODNs. In addition, only limited data are available on the tissue distribution of ODNs, particularly their tumor uptake, and on how this is influenced by dose.

In the present study, the pharmacokinetics and tissue distribution of MT-AS were characterized in nude mice bearing a subcutaneous NCI-H446 cell line by using a novel HPLC method and, in some samples, a CGE method.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals and reagents. Phosphorothioate ODNs were synthesized using phosphoramadite chemistry on an Oligo Pilot II synthesizer (Pharmacia Biotech, Piscataway, NJ) and purified by HPLC. MT-AS, a 20-mer phosphorothioate ODN, was used at a purity of approximately 97% with 2% to 3% of a 19-mer impurity. The nucleotide sequence of MT-AS was 5-CAT CTG CCA TTC CCA CTC TA-3. Sequences for the 15-mer to 19-mer species were exactly the same as for the parent 20-mer MT-AS with serial 1 nucleotide deletions from the 3 end. The 15-mer to 19-mer ODNs were synthesized using phosphoramadite chemistry on a DNA synthesizer model 394 (Applied Biosystems, Foster City, CA). Oxidation of the oligonucleotides to form the phosphorothioate linkages was performed using TETD/acetonitrile sulfurizing reagent according to the manufacturer's instructions. Lithium bromide was obtained from Fluke Chemical Corp (Ronkonkoma, NY). Phenol was obtained from Amresco (Solon, OH). TEMED was purchased from Fisher Scientific (Fair Lawn, NJ). NP-40, formamide, chloroform, proteinase K and all other chemicals were obtained from Sigma (St. Louis, MO) at the highest grade available. Solid-phase cartridges with anion-exchange membranes (Ultrafree-MC-DEAE) were purchased from Millipore (Bedford, MA). Centrifree micropartition systems were obtained from Amicon (Danver, MA). HPLC water was deionized distilled water filtered through a Millipore Milli-Q Ultra-pure water system.

Cell culture. The NCI-H446 cell line was obtained from the American Type Culture Collection (Rockville, MD) and maintained at 37°C, 5% CO₂ in RPMI 1640 culture medium containing 10% fetal bovine serum. The cells were cultured to about 75% confluence. Then they were harvested with 0.04% trypsin/EDTA and washed twice with Dulbecco's PBS to remove serum proteins. The cells were resuspended at a concentration of 3 × 10⁶ cells per 100 µl of Dulbecco's PBS and kept on ice until further use.

Animal studies. Female athymic nude mice on a Swiss background (NIH-nu/nusf), aged 4 to 5 weeks and weighing between 20 g and 30 g, were used in all studies. The animal protocol was approved by the institutional Animal Care and Use Committee at Fox Chase Cancer Center, in accordance with the Guide for the Care and Use of Laboratory Animals provided by the NIH.

Protein binding and stability of MT-AS in plasma. Binding of MT-AS to plasma proteins was determined by filtration of human plasma, SCID mouse plasma, or human serum albumin through a Centrifree micropartition system using a YMT membrane with 30,000 molecular weight cutoff. Before filtration, 10 µl of MT-AS was added to 500 µl of each protein sample at initial concentrations of 2, 5, 10, 20, 50, 100 and 200 µM and then incubated at 37°C for 25 min. The protein samples were centrifuged at 1164 × g at 25°C for 25 min. For human plasma, the latter centrifugation step was also completed at 37°C. Fifty microliters of the resultant ultrafiltrate was analyzed by HPLC for MT-AS. PBS solutions (pH 7.4) containing the same MT-AS concentrations as the protein samples were treated in an identical fashion to determine the nonspecific binding of MT-AS to the Centrifree system. Triplicate samples were analyzed for each protein sample and for MT-AS concentration.

Measurement of MT-AS and catabolites in biological samples. Quantitation of parent MT-AS (20-mer) and 15-mer to 19-mer catabolites was achieved with HPLC and CGE assays (Chen et al., 1997). The HPLC method provided a measure of total MT-AS, parent 20-mer MT-AS plus 15-mer to 19-mer catabolites, because these species elute as a single chromatographic peak. The CGE method, which does resolve individual oligonucleotides, was used to determine the relative percentages of the parent MT-AS and of catabolites of between 15-mer and 19-mer. A brief summary of the methods (Chen et al., 1997) follows.

HPLC sample analysis. The liquid chromatographic system (Hewlett-Packard Series 1050, Palo Alto, CA) consisted of a gradient pump, an autoinjector and a variable-wavelength UV detector. Anion-exchange liquid chromatography was performed with a 20 × 1 mm I.D. guard column (Upchurch Scientific, Oak Harbor, WA) hand-packed with spherical 13-mm Dionex Nucleopak PA-100 support (Dionex Chromatography, Sunnyvale, CA). A ternary solvent gradient elution method was employed that consisted of mobile phase A (25 mM Tris, 1 mM EDTA, pH 7.0), mobile phase B (25 mM Tris, 1 mM EDTA, 2 M LiBr, pH 7.0) and mobile phase C (formamide). The initial mobile-phase composition was set at 60% A, 10% B, 30% C used at a flow rate of 1 ml/min and was then brought to 30% A, 40% B, 30% C over 1.2 min and held for 0.8 min. MT-AS was detected spectrophotometrically at 267 nm.

CGE assays. The relative percentages of MT-AS and of its 15-mer to 19-mer catabolites in biological samples were determined by a CGE assay (Chen et al., 1997). Briefly, tissue homogenates or urine samples were centrifuged to remove debris. An aliquot of the resultant supernatant was diluted with 340 µl of loading buffer (70% 25 mM Tris, 1 mM EDTA, pH 7.0; 30% formamide). Plasma samples were directly diluted with the loading buffer. All samples were extracted through a solid-phase cartridge with an anion-exchange membrane (Ultrafree-MC DEAE, Millipore Co. Bedford, MA). The collected effluent was dialyzed against deionized water with a 0.025-µm membrane (Millipore Co. Bedford, MA) for 20 min. The resultant dialysates were injected electrokinetically into gel-filled capillaries composed of fused-silica tubing (I.D. = 75 µm, O.D. = 375 µm, effective length = 30 cm, total length = 50 cm; Polymicro Technologies, Phoenix, AZ) filled with a degassed solution of 18% polymerizing linear acrylamide in 30% (v/v) formamide media (0.1 M Tris-borate, 2.5 mM EDTA-2Na buffer (pH 8.3) containing 7 M urea). Polymerization was initiated by amonium persulfate/TEMED chemistry (Drossman et al., 1990). An electric field of 500 V/cm was applied, which resulted in a current of 5 to 10 µA.

Pharmacokinetic analysis. Noncompartmental analysis was used to calculate the pharmacokinetic parameters from mean plasma or tissue total ODNs (15-mer to 20-mer) concentration-time data. AUC and AUMC were obtained using Lagrange polynomial integration from time zero to the last measured sample time, with extrapolation to time infinity using the least-squares terminal slope by means of the NCOMP computer program (Laub and Gallo, 1996). From these areas and terminal slopes, CL_t, f_e, V_ss, and t_1/2 were calculated from standard formulas (Gibaldi and Perrier, 1982). Also, t_max and C_max were recorded for each tissue.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

After i.v. bolus administration, total MT-AS (i.e., 15-mer to 20-mer species) was rapidly cleared from blood (see fig. 1), with an elimination half-life ranging from 45 to 240 min at doses from 10 mg/kg to 300 mg/kg. At all but the lowest dose level of 10 mg/kg, the concentration-time profiles were multiphasic, particularly at the highest dose of 300 mg/kg. At the three lower dose levels, plasma concentrations were below the limit of quantitation after 6 h and probably contributed to the shorter half-lives compared with the 240 min value obtained at the 300-mg/kg level. Because CL_t values were comparable at all dose levels, the contribution of the fractional post-6 h AUC values, had plasma concentrations been measured after 6 h at the lower doses, to the total AUC values is likely to have been small. Table 1 gives the estimated pharmacokinetic parameters for MT-AS. MT-AS was relatively stable in nude mouse plasma at a concentration of 2 µM at ambient temperature, with a degradation half-life of 102 h, which indicates that chemical and enzymatic degradation during sample preparation and analysis was unlikely.

View larger version (20K): [in this window] [in a new window]   Fig. 1.   Total (HPLC analysis of 15-mer to 20-mer) MT-AS oligonucleotides plasma concentration-time profiles after i.v. bolus administration at four dose levels. Symbols represent mean concentration and 1 positive standard deviation.

Total clearance of MT-AS, based on total 15-mer to 20-mer concentrations, varied from 7.9 ml/min/kg to 15.2 ml/min/kg over the dose range of 10 mg/kg to 300 mg/kg; however, there was no systematic trend. In fact, CL_t appears relatively constant except for the minimum value observed at 30 mg/kg. Because studies of this nature rely on mean concentrations, statistical evaluations of dose-dependent pharmacokinetics cannot be made.

The V_ss value of MT-AS (based on 15-mer to 20-mer species) also exhibited variability (table 1), but unlike CL_t, it showed a very substantial increase at a dose of 300 mg/kg. This 3- to 4-fold increase in V_ss may have a physiological basis in saturable plasma protein binding. Table 2 gives the percentages MT-AS bound to plasma proteins from various sources. Regardless of the protein matrix (i.e., human and SCID mouse plasma, or human serum albumin), there is saturation of binding sites above an MT-AS concentration of about 50 µM, concentrations that were achieved in vivo particularly at the highest dose level (300 mg/kg). The saturation of protein binding was confirmed in human plasma when the ultrafiltration step was completed at 37°C. Under these conditions, the % protein binding varied from 99.1% to 72% over the MT-AS concentration range from 2 µM to 200 µM. The similarity in the percentage bound between mouse and human plasma, and between human albumin and human plasma, indicates that the mouse is a suitable model for plasma protein binding of ODNs and that albumin is the predominant protein-binding species.

HPLC analyses indicated that excretion of total MT-AS (15-mer to 20-mer) in urine was minimal, being 0.4% at 10 mg/kg, 4.1% at 30 mg/kg, 5.1% at 100 mg/kg and 5.3% at 300 mg/kg in the first 6 h after drug administration. At the level of 300 mg/kg, urine was collected for 48 h, and it was found that 12.4% of the administered dose was excreted as 15-mer to 20-mer ODNs.

Extensive tissue distribution was observed (see fig. 2) after i.v. bolus administration of different doses of MT-AS to tumor-bearing nude mice. It was found that the rank order of tissue distribution based on the total 15-mer to 20-mer MT-AS species was dose-dependent. At the dose level of 10 mg/kg, based on a partial AUC to 8 h, the rank order of tissue distribution was kidney > spleen > liver > muscle > lung > tumor. At 300 mg/kg, it became kidney > liver > tumor > muscle > spleen > lung > heart. At all dose levels, the minimum amount of MT-AS was in brain. Table 3 provides the observed T_max, C_max and partial AUC_tissue values in each tissue at all four dosage levels. Examination of the AUC_tissue values for tumor and heart indicates more than proportional increases in going from 100 mg/kg to 300 mg/kg. On the basis of the observed T_max values, liver and tumor showed the slowest rates of MT-AS accumulation, with values ranging from 4 h to 12 h, depending on the dose.

View larger version (45K): [in this window] [in a new window]   Fig. 2.   Total MT-AS (15-mer to 20-mer) concentration in tissue as functions of dose and time. Data are presented as mean (n  3) plus 1 standard deviation.

CGE analysis (fig. 3; table 4) of certain plasma, urine and tissues samples was used to determine the relative percentage of parent and catabolites based on the total MT-AS (i.e., 15-mer to 20-mer) concentrations obtained from HPLC analysis. In plasma, from 2 h to 8 h after the 300-mg/kg dose, the percentage of 20-mer MT-AS was relatively constant at approximately 52%. The remainder of the total ODN concentration consisted of 19-mer to 15-mer ODNs in decreasing percentages. Urine analyzed over a 24-h period for the 300 mg/kg group revealed that about 50% to 60% of the total ODN was the 20 mer parent drug, the remainder being 15-mer to 19-mer catabolites. The pattern of catabolite composition was relatively constant over 24 h in urine and was similar to that observed in plasma. Regardless of the tissue (see table 4), the mean percentage of parent MT-AS (20-mer) was between about 43% and 63%. At 2 h after dosing, the greatest extent of metabolism was in the liver, the parent MT-AS accounting for 43.8%. MT-AS was degraded in kidney by about 51.5% at 2 h, the rest of the catabolites being 19-mer (30%), 18-mer (11.6%) and 17-mer (17%).

View larger version (11K): [in this window] [in a new window]   Fig. 3.   CGE electropherograms of individual oligonucleotides of MT-AS at different time-points in plasma (panel A) and kidney (panel B) of mouse that received 300 mg/kg of MT-AS i.v. The last peak at about 60 min is 20-mer MT-AS, preceded in order by 19-mer to 15-mer MT-AS.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The pharmacokinetics and metabolism of ODNs have been based on a variety of bioanalytical assays, such as radiochemical (Agrawal et al., 1991, 1995, 1996; Chen et al., 1990; Crooke, 1995; Goodarzi et al., 1992; Sands et al., 1994; Srinivasan and Iversen, 1995; Zhang et al., 1995a,b), hybridization (Temsamani et al., 1993) and chromatographic (Bigelow et al., 1990; Bourque and Cohen, 1993; Sands et al., 1994) methods. Direct fluorescence and immunoenzymatic histological methods were also applied to a tissue distribution study with ODNs (Plenat et al., 1995). However, these methods suffer from one or more problems, such as inability to distinguish catabolites, lack of sensitivity, inconvenience, and lack of validation data. In order to characterize accurately the in vivo disposition of antisense ODNs, an HPLC assay combined with CGE (Bourque and Cohen, 1994) and most recently a CGE assay alone (Leeds et al., 1996) were developed to discriminate full-length oligonucleotide from its catabolites in plasma. However, further reports on the application of these methods to the pharmacokinetics and tissue distribution of ODNs are unavailable. In the present study, the HPLC assay provided a rapid and reproducible determination of total 15-mer to 20-mer MT-AS in biological samples. The time-consuming yet reproducible CGE assay permitted measurement of parent MT-AS and individual catabolites and thus provided analyte-specific data on ODN disposition. Monia et al. (1992, 1993) found that in order to have sufficient mRNA affinity for biological activity, phosphorothioate ODNs should be at least 15-mer to 17-mer in length. Although the HPLC assay cannot qualify as a parent drug-specific assay, it does provide a measure of the most relevant biologically active species. This aspect, combined with its convenience, speed and reproducibility, make it an attractive tool for pharmacokinetic studies.

Although little is known about its in vivo metabolic fate, catabolism of MT-AS is presumed to occur by endonucleases and exonucleases. Saturation of these enzymes, although possible, was not supported by a reduction in CL_t at the high dose. Unlike kinetic profiles of other phosphorothioates ODNs (Agrawal et al., 1995, 1996; Crooke and Bennett 1996; Srinivasan and Iversen, 1995; Zhang, et al., 1995a, b), which reported half-lives of greater than 24 h in experimental animals and humans, the elimination of MT-AS was relatively fast, with elimination half-lives from 45 min to 240 min at doses levels from 10 to 300 mg/kg. It should be emphasized that the former investigations utilized nonspecific quantitation methods, such as radiolabeled assays, that could lead to an overestimation of the terminal elimination phase, particularly for compounds eliminated primarily by metabolism as the ODNs (Crooke, 1995). Although differences in molecular weight, ODN sequence and animal species may account for differences in kinetic properties, differences in analytical methods cannot be discounted. On the basis of the CGE analysis of parent 20-mer MT-AS, the terminal elimination phase of MT-AS was between about 1 and 4 h, shorter than one would have anticipated from earlier studies using radiolabeled compounds.

In early studies with phosphorothioate ODNs in rodents, primates and humans, urinary recovery of radioactivity was reported to be between 30% and 60% of the total dose (Agrawal et al., 1991; 1995), with 30% of the radioactivity accounted for by intact drug (Crooke 1995). In the present study, approximately 10% of the dose was accounted for by total (15-mer to 20-mer) MT-AS in urine. Because only radioactivity was followed in previous studies, it can be assumed that metabolites as well as parent compound were included in these measurements. However, in a result consistent with previous reports, about 50% of the recovered amount was parent ODNs. The fact that some mice had "black urine" within 4 to 8 h after i.v. bolus of 300 mg/kg suggests a possible dose-related hemolysis. Trace amounts of blood were also found in monkey urine after ODN administration (Srinivasan and Iversen, 1995; Crooke, 1995). The clinical implications of these findings need further clarification.

The volume distribution at steady state demonstrated dose-dependence, which is consistent with saturable plasma protein binding. It has been suggested that serum proteins such as serum albumin are a repository for phosphorothioate ODN (Bigelow et al., 1990; Crooke et al., 1996; Sands et al., 1994; Srinivasan and Iversen, 1995). The in vitro protein binding studies revealed that MT-AS is appreciably bound to plasma proteins, that albumin is the major binding species and that saturable binding occurred above total MT-AS concentrations of 50 µM. Saturation of plasma protein binding sites by MT-AS would greatly increase the unbound fraction in plasma, making more MT-AS available for tissue uptake, and would lead to an increase in the volume of distribution.

MT-AS accumulated in kidney and to a smaller extent in liver, a result consistent with previous investigations of other ODNs in preclinical models (Agrawal, 1996; Crooke et al., 1996 and Srinivasan and Iversen, 1995). The rate of distribution and clearance of MT-AS from tissues seemed to be both dose- and tissue-dependent. Because tissues were thoroughly perfused before collection, the measured total ODN concentrations probably represent tissue parenchyma concentrations rather than those contaminated by residual blood, as suggested by Plenat et al. (1995). Moreover, MT-AS concentrations did not parallel plasma concentrations in the terminal phases consistent with saturable distribution. Of special interest were the appreciable tumor MT-AS concentrations obtained at the higher dose levels. A similar observation with other ODNs was also obtained in tumor-bearing nude mice by using a histological method (Plenat et al., 1995). On the basis of cell-culture studies (Mukhooadhyay et al., 1991; Szyf et al., 1996), it would be anticipated that the in vivo MT-AS tumor concentrations would be therapeutically active and would further facilitate investigations of meaningful pharmacodynamic relationships in this model.

Overall, MT-AS is characterized by moderate clearance, primarily metabolic, and by saturable volume of distribution. The combined HPLC/CGE assays provide pharmacokinetic data on the relevant therapeutic moieties, as well as individual ODN, and thus they rationally support drug-delivery, pharmacodynamic and clinical trial investigations.