Antiproliferative Effect of a Lectin- and Anti-Thy-1.2 Antibody-Targeted

Mouse monoclonal antibodies directed against tumor cell surface antigens have ..... Primary, metastatic, and Thy-1.2-transfected human colorectal cell...
0 downloads 0 Views 159KB Size
664

Bioconjugate Chem. 2000, 11, 664−673

Antiproliferative Effect of a Lectin- and Anti-Thy-1.2 Antibody-Targeted HPMA Copolymer-Bound Doxorubicin on Primary and Metastatic Human Colorectal Carcinoma and on Human Colorectal Carcinoma Transfected with the Mouse Thy-1.2 Gene B. R ˇ ´ıhova´,*,† M. Jelı´nkova´,† J. Strohalm,‡ M. Sˇ t’astny´,† O. Hovorka,† D. Plocova´,‡ M. Kova´rˇ,† L. Dra´berova´,§ and K. Ulbrich‡ Institute of Microbiology, Academy of Sciences of the Czech Republic, 142 20 Prague 4-Krcˇ, Czech Republic, Institute of Macromolecular Chemistry, Academy of Sciences of the Czech Republic, 162 06 Prague 6-Dejvice, Czech Republic, and Institute of Molecular Genetics, Academy of Sciences of the Czech Republic, 142 20 Prague 4-Krcˇ, Czech Republic . Received December 7, 1999; Revised Manuscript Received June 3, 2000

The aim of this study was to compare the potential of two plant lectins [peanut agglutinin (PNA) and wheat germ agglutinin (WGA)], monoclonal antibody (anti-Thy-1.2), its F(ab′)2 fragments, and galactosamine as targeting moieties bound to the polymer drug carrier to deliver a xenobiotic, doxorubicin, to selected cancer cell lines. We have used primary (SW 480, HT 29) and metastatic (SW 620) human colorectal cancer cell lines and a transfectant, genetically engineered SW 620 cell line with mouse gene Thy-1.2 (SW 620/T) to test the possibility of marking human cancer with xenogeneic mouse gene and use it for effective site-specific targeting. The targeting moieties and doxorubicin were conjugated to a water-soluble copolymer based on N-(2-hydroxypropyl)methacrylamide (HPMA) acting as a carrier responsible for controlled intracellular release of the targeted drug. FACS analysis showed a strong binding of WGA-FITC to all tested cell lines. Binding of PNA-FITC was considerably weaker. The in vitro antiproliferative effect of lectin-targeted HPMA carrier-bound doxorubicin evaluated as [3H]TdR incorporation reflected both the intensity of the binding and the different sensitivity of the tested cancer cells lines to doxorubicin. The antiproliferative effect of conjugates targeted with WGA was comparable to that with the conjugates targeted with the anti-Thy-1.2 monoclonal antibody or their F(ab′)2 fragments. The magnitude of the cytotoxic effect of HPMAdoxorubicin targeted with PNA was lower in all tested cell lines. While the conjugates with WGA were more cytotoxic, the conjugates with PNA were more specific as their binding is limited to cancer cells and to the sites of inflammation. Noncytotoxic conjugates with a very low concentration of doxorubicin and targeted with PNA, anti-Thy-1.2, or their F(ab′)2 fragments exerted in some lines (SW 480, SW 620) low mitogenic activity. The Thy-1.2 gene-transfected SW 620 metastatic colorectal cancer cell line was sensitive to the antiproliferative effect of Thy-1.2-targeted doxorubicin as was shown for the Thy-1.2+ EL4 cell line and for Thy-1.2+ concanavalin A-stimulated mouse T lymphocytes. These results represent the first indication of the suitability of transfection of human cancer cells with selected targeting genes for site-specific therapy of malignancies.

INTRODUCTION

Mouse monoclonal antibodies directed against tumor cell surface antigens have the potential to localize selectively within the tumor after systemic administration (Goldenberg et al., 1980). However, a number of clinical trials that were conducted to evaluate the therapeutic efficacy of such mAbs per se have shown relatively limited antitumor effects (Houghton et al., 1985). Recent therapeutic approaches to the treatment of cancer have * To whom correspondence should be addressed: Institute of Microbiology AS CR, Vı´denˇska´ 1083, 142 20 Prague 4-Krcˇ, Czech Republic. Telephone: +420 2 475 2267. Fax: +420 2 472 11 43. E-mail: [email protected]. † Institute of Microbiology, Academy of Sciences of the Czech Republic. ‡ Institute of Macromolecular Chemistry, Academy of Sciences of the Czech Republic. § Institute of Molecular Genetics, Academy of Sciences of the Czech Republic.

focused on developing novel delivery systems to increase the therapeutic indices of anticancer agents by targeting them to malignant cells. The lack of specificity of conventional chemotherapy for malignant cells with severe dose-limiting side effects has led to the development of a new class of compounds called immunotoxins (Neville and Youle, 1982; Raso, 1982; Vitetta et al., 1982; Hertler and Frankel, 1989) and other drug delivery 1 Abbreviations: AIBN, 2,2′-azobisisobutyronitrile; Con A, concanavalin A; DCCI, N,N′-dicyclohexylcarbodiimide; DMF, N,N-dimethylformamide; DMSO, dimethyl sulfoxide; DOX-HCL, doxorubicin hydrochloride; FITC, fluorescein isothiocyanate; FPLC, fast protein liquid chromatography; FCS, fetal calf serum; GPC, gel permeation chromatography; [3H]TdR, [3H]thymidine; HPMA, N-(2-hydroxypropyl)methacrylamide; HPLC, high-performance liquid chromatography; Ma, methacryloyl; mAbs, monoclonal antibodies; NK cells, natural killer cells; ONp, p-nitrophenoxy; OPA, o-phthaldialdehyde; PBS, phosphatebuffered saline; RI, refractive index; PNA, peanut agglutinin; TEA, triethylamine; WGA, wheat germ agglutinin.

10.1021/bc9901696 CCC: $19.00 © 2000 American Chemical Society Published on Web 09/18/2000

Lectin- and Antibody-Targeted Polymeric Prodrugs Based on HPMA

systems for antitumor therapy (Duncan, 1992). These drug delivery systems are directed against epitopes present on tumor cells and carry drugs interfering with tumor cell growth. Such tumor-selective cytotoxic conjugates are also produced by a conjugation of the anthracycline antibiotic doxorubicin to the water-soluble polymeric carrier based on N-(2-hydroxypropyl)methacrylamide (Kopecˇek et al., 1991) and targeted with carbohydrates (Duncan et al., 1987, 1988), antibody (R ˇ ´ıhova´ and Kopecˇek, 1985; R ˇ ´ıhova´ et al., 1988, 1992, 1996; Sˇ t’astny´ et al., 1997; Jelı´nkova´ et al., 1998; Omelyanenko et al., 1998a,b), or growth hormones (O’Hare et al., 1993) to avoid nonspecific damage of nonmalignant cells and toxicity to healthy tissues. Lectins are plant proteins or glycoproteins that bind to specific sugars in the surface glycoproteins and glycolipids of eukaryotic cells. Weaver and Bailey (1987) demonstrated a rapid receptor-mediated endocytosis of concanavalin A, and Hussain et al. (1997) suggested that lectins can be utilized for intestinal drug targeting in vivo. Altered lectin binding is common in malignant and premalignant epithelia (Ryder et al., 1994). Cancer cells often express different carbohydrates on their surfaces, which are involved in NK cell cosignaling and consequently in cancer cell eradication. It was reported that the normal adult colonic epithelium expresses receptors for wheat germ agglutinin (WGA), while neoplastic or inflammatory colonic epithelium binds peanut agglutinin (PNA) (Boland and Roberts, 1988). Wirth et al. (1998) described the usefulness of WGA as a targeting carrier protein of doxorubicin for the treatment of colon carcinoma, and Wroblewski et al. (1999) reported that conjugation of WGA and PNA to the polymeric carrier poly(HPMA) does not destroy their binding affinity to target carbohydrates. However, spontaneous human cancers often do not express surface antigens suitable for targeting. Thus, we also address the question of whether transfection of human colorectal carcinoma with the mouse Thy-1.2 gene would result in a xenogeneic-gene-marked cancer target suitable for site-specific drug delivery. The Thy-1.2 (CDw90) alloantigen is a GPI-anchored surface protein with an Mw of about 25-35 kDa, which is expressed on thymocytes and mouse peripheral T cells and on neurons. It is involved in T activation process. Thy-1.2 glycoprotein was used for efficient targeting in mouse T lymphocytes and mouse EL4 T cell lymphoma (R ˇ ´ıhova´ et al., 1996, 2000; Jelı´nkova´ et al., 1998). Thus, the purpose of this study was (a) to investigate the targeting capacities of two lectins, WGA and PNA, in several tumor cell lines and in normal concanavalin A-stimulated T lymphocytes, (b) to compare the targeting capacity of lectins with galactosamine, the monoclonal anti-Thy-1.2 antibody and its F(ab′)2 fragment, and (c) to determine if the metastatic human colorectal cancer cell line SW 620 transfected with the mouse Thy-1.2 gene is sensitive to the antiproliferative effect of anti-Thy-1.2 monoclonal antibodies-targeted HPMA carrier-bound doxorubicin. The sensitivity of transfected cell line SW 620/T was compared with the sensitivity of Thy-1.2+ EL4 mouse T cell lymphoma and with normal Thy-1.2+ splenocytes activated with concanavalin A. MATERIALS AND METHODS

Chemicals. 1-Aminopropan-2-ol, methacryloyl chloride, glycyl-L-phenylalanine, L-leucylglycine, 4-nitrophenol, triethylamine, N,N-dimethylformamide, N,N′-dicyclohexylcarbodiimide, 2,2′-azobisisobutyronitrile, phthalalde-

Bioconjugate Chem., Vol. 11, No. 5, 2000 665 Table 1. Characteristics of Polymer Precursors and Antibody (Lectin)-Targeted Poly(HPMA) Conjugates of Doxorubicin sample

structure

Mwa ONpa DOX TMb (wt %) ( SD (kDa) (mol %) (wt %)

1

16.4

9.0

4.5

25 ( 2

2

16.4

9.0

4.5

25 ( 2

3

25.1

8.8

6.4

4

18.8

8.7

4.2

25 ( 2

5

18.8

8.7

4.2

25 ( 2

17.8 17.8

6.2 6.2

0 0

25 ( 2 25 ( 2

6 7

P-GFLG-WGA P-GFLG-PNA

6.2 ( 0.9

a Polymer precursor. b Targeting moieties: monoclonal antibody anti-Thy-1.2, its F(ab′)2 fragment, PNA, WGA, or galactosamine.

hyde, 5-thio-D-glucose, galactosamine, and dimethyl sulfoxide were from Fluka AG (Buchs, Switzerland). Doxorubicin hydrochloride was obtained as a kind gift from Pharmacia-Farmitalia Carlo Erba, and peanut agglutinin and wheat germ agglutinin were from Sigma-Aldrich. All other chemicals and solvents were analytical grade. The organic solvents were dried and purified by conventional procedures and distilled before use. Stock solutions of samples and/or drug were made in PBS and diluted further in the medium. Synthesis of Monomers. N-(2-Hydroxypropyl)methacrylamide was prepared by the reaction of methacryloyl chloride with 1-aminopropan-2-ol in freshly distilled methylene chloride containing suspended anhydrous sodium hydrogencarbonate as described by Ulbrich et al. (2000). Mp: 70 °C. Anal. Calcd: C, 58.72; H, 9.15; N, 9.78. Found: C, 58.79; H, 9.02; N, 9.75. Methacryloylglycyl-DL-phenylalanyl-L-leucylglycine 4-nitrophenyl ester (Ma-GlyPheLeuGly-ONp) was prepared as previously described (Rejmanova´ et al., 1983). Mp: 134-136 °C. Amino acid analysis: 2.06:0.53:0.48:1.00: 0.02 Gly:L-Phe:D-Phe:L-Leu:D-Leu. Synthesis of Polymer Conjugates. Polymer conjugates with antibody, its F(ab′)2 fragment, galactosamine, or lectins (samples 1-7, Table 1) were prepared by twoor three-step synthesis. In the first step, polymer precursors differing in the content of reactive ester groups were prepared. Polymer precursors were prepared by radical precipitation copolymerization of HPMA with Ma-GlyPheLeuGly-ONp (initiator AIBN, 0.6 wt %; concentration of monomers, 12.5 wt %; content of Ma-GlyPheLeuGlyONp in the reaction mixture, 9 or 6 mol %; temperature, 60 °C; polymerization time, 24 h) as described previously (Rejmanova´ et al., 1983). The content of oligopeptide side chains terminated in 4-nitrophenoxy groups was determined in the copolymer by UV spectrophotometry (DMSO,  ) 9500 L mol-1 cm-1, λ ) 274 nm), and weight- and number-average molecular masses, Mw and Ma, respectively, were determined by liquid chromatography after aminolysis of reactive groups with 1-aminopropan-2-ol

666 Bioconjugate Chem., Vol. 11, No. 5, 2000

[FPLC Pharmacia, Superose 6, phosphate buffer containing 0.5 M NaCl, calibration using poly(HPMA) fractions]. In the second step, HPMA conjugates with WGA or PNA (samples 6 and 7) were obtained by aminolytic reaction of the polymer precursor in aqueous solution. The polymer precursor was dissolved in distilled water (10-15 wt %) under stirring after cooling to 10 °C and adding a solution of the respective lectin in phosphate buffer (15-25 mg/mL). The pH of the reaction mixture was adjusted to 8.2 under stirring at 10 °C by adding a saturated solution of Na2B4O7 (end-point titration; pHstat radiometer). The course of the reaction was controlled by the consumption of the borate solution. After the reaction had been completed, the polymer-lectin conjugate was purified by gel filtration (Sephadex G-25) and transferred into phosphate buffer. The conjugate was characterized by GPC and electrophoresis, and the content of the protein in the conjugate was estimated using amino acid analysis. In the second step of the synthesis of the drugcontaining conjugates (samples 1-5), DOX was bound to the polymer via aminolysis of a part of reactive ester groups. The respective polymer precursor was dissolved in DMSO (15 wt %), and 50 mol % DOX-HCl (relative to the ONp content in the precursor) was added. Under stirring, 1 equiv of TEA (related to DOX-HCl) was added in four small portions. The reaction mixture was stirred in the dark at room temperature for 2 h; the solution was filtered, and the polymer was obtained by precipitation into an acetone/diethyl ether mixture. Free, unreacted DOX was removed by gel filtration of the polymer solution in methanol using a Sephadex LH-20 column. In the third step, the polymer containing DOX and remaining ONp groups, dried in a vacuum, was dissolved in distilled water (10-15 wt %) and conjugated with the respective protein using the procedure described above. The content of DOX in the conjugate was estimated spectrophotometrically; the conjugate was characterized by GPC and electrophoresis, and the content of the protein in the conjugate was estimated using amino acid analysis. Conjugate 3 containing galactosamine as a targeting moiety was prepared by an aminolytic reaction of the rest of ONp groups of the DOX-bearing polymer (as described above) with galactosamine in DMSO as described previously (Duncan et al., 1987). Methods of Purification and Characterization of Conjugates. Polymer-drug intermediates were freed of low-molecular mass impurities (such as free DOX or 4-nitrophenol) by gel filtration using a Sephadex LH-20 column (100 cm × 3.4 cm) with methanol elution. The conjugates with antibody or lectin were purified using a Sephadex G-25 column eluted with phosphate buffer (pH 7.4). Samples were kept in the frozen state (-30 °C) until biological evaluation could be carried out. Conjugates containing both the targeting moiety and the drug were characterized and tested for the content of free polymer, drug, or protein by GPC using a Sepharose 4B/6B 100 cm × 1.6 cm column (UV and RI detection) or FPLC with a system equipped with a Superose 6 column, and by electrophoresis (PharmaciaLKB Phast System, SDS-PAGE, gradient gels 4-15) (R ˇ ´ıhova´ and Kopecˇek, 1985). The protein content in the conjugates was estimated by amino acid analysis (LDC Analytical, precolumn OPA derivatization, calibration with the respective protein) and the DOX content by UV spectrophotometry in water ( ) 11 500 mol L-1 cm-1, λ ) 485 nm). Neither GPC nor electrophoresis showed

Rˇ ´ıhova´ et al.

significant amounts of free antibody, free lectin, or free drug in the conjugates. GPC was carried out on a Sepharose 4B/6B column (1.6 cm × 100 cm) using phosphate buffer (pH 7.2) at a flow rate of 2 mL/min, via RI and UV detection. The composition of amino acids in Ma-GlyPheLeuGlyONp and the content of proteins in the conjugates were determined by amino acid analysis with precolumn OPA derivatization on an LDC Analytical analyzer using the reverse-phase column (Tessek SGX C18, 250 mm × 4 mm), via elution with a solvent gradient of sodium acetate buffer and methanol, and fluorescence detection with a model 4100 fluoromonitor (LDC Analytical). Before analysis, the samples were hydrolyzed with 6 M HCl at 115 °C for 16 h. The hydrolysate was dried over NaOH and, prior to the analysis, dissolved in distilled water. The concentration and the ratio of amino acids were determined using OPA-3-mercaptopropionic acid derivatization, with detection at an emission wavelength of 229 nm and an excitation wavelength of 450 nm. The optical purity of amino acids was determined using OPA-5-thio-D-glucose derivatization, with detection at an emission wavelength of 450 nm and an excitation wavelength of 229 nm. Targeting Moieties. Two lectins were used in this study: PNA from Arachis hypogea, which binds to galactose β-1,3-N-acetylgalactosamine, and WGA from Triticum vulgare, which binds to N-acetyl-D-glucosamine. Both lectins as well as their fluorescein isothiocyanate (FITC) analogues were obtained from Sigma-Aldrich. Monoclonal antibody directed against Thy-1.2 mouse alloantigen, clone 1aG4/C5, is of the IgG3 isotype (Dra´ber et al., 1980). Antibodies were produced in BALB/c mice by injection of hybridoma clone 1aG4/C5 and purified from mouse ascites fluid by a combination of 2-fold 45% SAS precipitation and protein A affinity chromatography (R ˇ ´ıhova´ et al., 1996). The concentration was determined by spectrophotometry (λ ) 280 nm), and their purity was assessed by SDS-PAGE. mAb reactivities were checked before use by indirect immunofluorescence using FITClabeled anti-mouse IgG3 and FACSorter (Becton Dickinson, Mountain View, CA) against appropriate cell lines. The F(ab′)2 fragment of anti-Thy-1.2 antibodies was produced by digestion with pepsin as described previously (R ˇ ´ıhova´ et al., 2000). Cell Lines. The following cell lines were used: human colorectal cancer cell line SW 480 (primary), human colorectal cancer cell line SW 620 (metastatic), human colorectal cancer cell line HT 29 (primary), and mouse EL4 T cell lymphoma (H-2b origin), all of them purchased from the American Type Culture Collection (ATCC, Rockville, MD). (Thy-1.2+) SW 620 Transfectant Cells. Cells. SW 620 cells were originally cultured in L-15 medium supplemented with 10% fetal calf serum (FCS). The cells were gradually adapted to grow in RPMI-1640 medium supplemented with 10% FCS and antibiotics and cloned in this medium. Clones with improved cloning efficiency were recloned, and cell lines with a relatively high cloning efficiency (about 30%) were used in these experiments. Plasmids. Two plasmid vectors were used. Plasmid pThy-1.2, a pUC13-derived vector containing the 8.1 kb EcoRI fragment derived from the genomic bacteriophage library of MOPC41 plasmacytoma (Evans et al., 1984), was kindly obtained from G. A. Evans of The Salk Institute of Biological Studies (La Jolla, CA). For selection of transfected cells, we used plasmid pSTneoB in which the bacterial aminoglycosyl 3′-phosphotransferase type II (AGPT) gene is under the control of tandem promoters

Lectin- and Antibody-Targeted Polymeric Prodrugs Based on HPMA

Bioconjugate Chem., Vol. 11, No. 5, 2000 667

Figure 1. (a-g) Flow cytometry analysis of WGA and PNA expression in EL4, SW 480, SW 620, SW 620/T, and HT 29 human colorectal cancer cell lines and in control and concanavalin A-stimulated A/J splenocytes. Cells were treated with WGA-FITC and PNA-FITC: left, WGA-FITC; right, PNA-FITC; gray, control cells; dotted line (left), 400 ng/mL WGA-FITC; dotted line (right), 10 µg/mL PNA-FITC; solid line (left), 80 ng/mL WGA-FITC; solid line (right), 2 µg/mL PNA-FITC.

of SV40 and HSVtk (Katoh et al., 1987). This vector, which confers resistance to the neomycin analogue G418 on mammalian cells, was kindly provided by K. Katoh of Kyoto University (Kyoto, Japan). Both plasmids were isolated by the alkaline lysis method and by two equilibrium centrifugations in CsCl (Sambrook et al., 1989). DNA Transfection and G418 Selection. DNA transfection was performed by a modification of the method of Chen and Okayama (1987). SW 620 cells (6 × 105) were aliquoted into 60 mm tissue culture dishes and incubated in culture medium. After 24 h, the medium was replaced with fresh culture medium and 0.5 mL of plasmid mix was added. Plasmid mix was prepared by mixing 9 µg of pThy-1.2 and 1 µg of pSTneoB in 0.5 mL of BES buffer containing 25 mM BES (pH 7.0), 140 mM NaCl, 0.75 mM Na2HPO4, and 125 mM CaCl2. After incubation for 6 h at 37 °C and in 5% CO2 in air, the medium was removed and 2 mL of 25% DMSO in BES buffer was added. After 4 min, the DMSO was removed, the cells were washed twice with PBS, and culture medium was added. The next day, the cells were trypsinized and transferred into three 60 mm tissue culture dishes with culture medium containing 1 mg/mL G418 (Geneticin, Sigma Chemical Corp., St. Louis, MO). Detached cells were removed by washing on the third, sixth, ninth, and fifteenth day after DNA transfection, and G418-resistant colonies were isolated 21 days after transfection.

The colonies were expanded and tested for the presence of surface Thy-1.2 by indirect immunofluorescence using 1aG4/C5 mAb (Dra´ber et al., 1980) and FACSorter analysis. Initial screening showed that only a small fraction of transfected cells exhibited surface fluorescence, and therefore, Thy-1.2+ cells were enriched by a two-step immunomagnetic separation procedure using 1aG4/C5 mAb in the first step and Dynabeads M-450 coated with goat anti-mouse IgG (Dynal A. S., Oslo, Norway) in the second step. After purification, the cells were cloned and recloned in the absence of G418, and only subclones in which more than 95% of the cells expressed surface Thy-1.2, as determined by FACSorter analysis (see above), were used in further experiments (Figure 1). Tissue Culture Condition. Routine culture and all experiments with cancer cell lines were conducted at 37 °C in a humidified atmosphere of 5% CO2/95% air in RPMI 1640 medium (Gibco Laboratories) supplemented with 10% v/v heat-inactivated fetal calf serum (FCS) selected for low mitogenicity, 2 mM L-glutamine (Gibco Laboratories), 50 mM 2-mercaptoethanol (Fluka), 100 units/mL penicillin, and 100 µg/mL streptomycin. Mice. Mice of inbred strain A/J were obtained from the breeding colony of the Institute of Physiology (Academy of Sciences of the Czech Republic) and kept in standard housing. The mice were used when they were

668 Bioconjugate Chem., Vol. 11, No. 5, 2000

12 weeks old. All experiments were approved by the Animal Welfare Committee of the Institute of Microbiology of the Academy of Sciences of the Czech Republic. Fluorescence-Activated Cell Sorter Analysis (FACS). Flow cytometric measurements were carried out on a FACSorter as described previously (R ˇ ´ıhova´ et al., 1996, 2000). Labeled cells were analyzed using a forward versus side scatter gate for the inclusion of single-cell populations and exclusion of debris and cell aggregates. At least 5000 cells were tested for each measurement. Cytotoxicity Assay. Inhibition of [3H]thymidine incorporation was used to evaluate the cytostatic action of targeted HPMA conjugates of doxorubicin in vitro as described previously (R ˇ ´ıhova´ et al., 1992). Tumor cells were harvested from the exponential phase of growing cultures, centrifuged for 10 min at 800g, resuspended in fresh medium (RPMI 1640), and counted by hemocytometer. NUNCLON 96-well FB tissue culture plates (NUNC) were seeded with 0.1 mL of cell suspension (EL4 at a density of 5 × 105 cells/mL for 24 h; SW 620, SW 480, SW 620/T, and HT 29 at a density of 1 × 105 cells/mL for 72 h; splenocytes at a density of 5 × 106 cells/mL for 72 h), and various concentrations of free doxorubicin, doxorubicin bound to HPMA targeted with lectins, mAbs, their F(ab′)2 fragment or galactosamine, and nontargeted HPMA-bound lectins were added to the wells to achieve the desired concentrations and a final well volume of 0.25 mL. Triplicate wells were used for each test condition. At the end of the incubation period, 1 µCi/50 µL of [3H]thymidine (Amersham Corp.) was added per well, followed by incubation for 6 h. Cells were then collected using a cell harvester (Tomtec) onto glass fiber filters (filtermat, Wallac). When the mixture was dried, a sheet of solid scintillator-Meltilex (Wallac) was placed together with a filtermat containing 96 samples in a sample bag and run together through a heat sealer (Microsealer, Wallac) in MicroBeta TriLux (Wallac). Statistical Analysis. The Wilcoxon’s statistics (multiple-comparison method) and a Student’s t test were used to evaluate the differences between experimental groups. P values of