Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

The microorganism Actinomadura melliaura produces the two antibiotics AT2433-A1 and AT2433-B1 that only differ by the presence of a chlorine atom on the indolocarbazole chromophore (Fig. 1). In both cases, the large heterocyclic ring system is substituted by a unique disaccharide consisting of a methoxyglucose and an aminosugar subunit, 2,4-dideoxy-4-methylamino-L-xylose. These two antibiotics were first isolated in 1989 (Golik et al., 1989; Matson et al., 1989) and their total synthesis was accomplished 10 years later (Chisholm et al., 1999; Chisholm and Van Vranken, 2000) but their mechanism of action remains largely unknown.

View larger version (34K): [in this window] [in a new window]   Fig. 1.   Structures of the mono- and bis-glycosylated indolocarbazole derivatives used in this study.

AT2433-A1 and -B1 are structurally analogous to the monosaccharide indolocarbazole antibiotic rebeccamycin (Fig. 1) produced by Saccharothrix aerocolonigenes (Nettleton et al., 1985; Bush et al., 1987). This natural product containing an uncharged methoxyglucose residue has profoundly inspired the development of antitumor agents targeting topoisomerase I, a ubiquitous enzyme essential to the control of DNA topology in cells. A large number of synthetic monosaccharide rebeccamycin analogs have been developed and a few of them, such as the drug NB-506 (also known as J-107185 or L-753,000; Fig. 1), have revealed promising anticancer activities in vivo (Arakawa et al., 1995; Kanzawa et al., 1995; Yoshinari et al., 1995). Topoisomerase I is a primary target for NB-506 and related monosaccharide compounds (Urasaki et al., 2001; Woo et al., 2002).

The structural analogy between AT2433-A1 and rebeccamycin prompted us to postulate that the disaccharide antibiotics can also bind to DNA and interfere with the DNA cleavage activity of topoisomerase I. To test this hypothesis and investigate the structure-activity relationships in the indolocarbazole disaccharide series, we selected four compounds, AT2433-A1, AT2433-B1, iso-AT2433-B1, and the compound JDC-277, with an uncharged D-melibiose sugar unit (Fig. 1). Iso-AT2433-B1 is a diastereoisomer of the natural aminodisaccharide and corresponds to the incorrect structure originally proposed for AT2433-B1 (Chisholm et al., 1999). The four diglycosides contain the same N-methyl indolo[2,3-a]carbazole chromophore and for this reason, we choose the D-glucose derivative JDC-108 as a control monoglycoside (Fig. 1). A range of biophysical and biochemical techniques was used to compare the effects of these compounds on DNA and topoisomerase I at the molecular and cellular levels.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Drugs. The syntheses of the antibiotics AT2433-A1/B1 and the analogs used in this study have been described previously (Chisholm et al., 1999; Chisholm and Van Vranken, 2000). Camptothecin was purchased from Sigma-Aldrich (St. Louis, MO) and a sample of NB-506 (J-107185) was kindly provided by Dr. Tomoko Yoshinari (Banyu Pharmaceutical Co., Ltd., Tsukuba, Japan). Except for the surface plasmon resonance (SPR) experiments, the drugs were dissolved in DMSO at 5 mM. The stock DMSO solutions of drugs were kept at 20°C and freshly diluted with water to the desired concentration immediately before use.

Absorption Spectra and Melting Temperature Studies. Melting curves were measured using an Uvikon 943 spectrophotometer (Kontron, Zurich, Switzerland) coupled to a Neslab RTE111 cryostat. For each series of measurements, 12 samples were placed in a thermostatically controlled cell-holder, and the quartz cuvettes (10-mm pathlength) were heated by circulating water. Measurements were performed in BPE buffer, pH 7.1 (6 mM Na₂HPO₄, 2 mM NaH₂PO₄, and 1 mM EDTA). The temperature inside the cuvette was measured with a platinum probe; it was increased over the range 20-100°C with a heating rate of 1°C/min. The "melting" temperature, Tm, was taken as the mid-point of the hyperchromic transition. The Uvikon 943 spectrophotometer was also used to record the absorption spectra.

Surface Plasmon Resonance. The 5' biotin-labeled DNA hairpins d(CATATATATCCCCATATATATG) and d(CGCGCGCGTTTTCGCGCGCG) (hairpin loop underlined, polyacrylamide gel electrophoresis-purified; Eurogentec, Seraing, Belgium) were used for the SPR studies. Samples of hairpin DNA oligomers in HBS-EP buffer (10 mM HEPES, pH 7.4, 150 mM NaCl, 3 mM EDTA, and 0.0005% Surfactant P20) at 25 nM concentration were applied to flow cells in streptavidin-derivatized sensor chips (BiaCore SA-chips; BiaCore, Uppsala, Sweden) by direct flow at 2 µl/min in a four-channel 3000 optical biosensor system (BiaCore). The sensor chips were conditioned with three consecutive 1-min injections of 1 M NaCl in 50 mM NaOH followed by extensive washing with buffer. Nearly the same amount of all oligomers was immobilized on the surface by noncovalent capture, leaving one of the flow cells blank as a control. Manual injection was used with a flow rate of 2 µl/min to achieve long contact times with the surface and to control the amount of the DNA bound to the surface. All procedures for binding studies were automated as methods using repetitive cycles of sample injection and regeneration. Steady-state binding analysis was performed with multiple injections of different compound concentrations over the immobilized DNA surfaces for a 10-min period at a flow rate of 20 µl/min and 25°C. Solutions of drug with known concentrations were prepared in filtered and degassed buffer by serial dilutions from stock solution and were injected from 7-mm plastic vials with pierceable plastic crimp caps (BiaCore).

Circular Dichroism. Circular dichroic (CD) spectra were recorded on a J-810 dichrograph (Jasco, Tokyo, Japan). Solutions of drugs, nucleic acids, and their complexes (1 ml in 1 mM sodium cacodylate buffer, pH 7.0) were scanned in 1-cm quartz cuvettes. Measurements were made by progressive dilution of drug-DNA complex at a high phosphate/drug (P/D) ratio with a pure ligand solution to yield the desired drug/DNA ratio. Three scans were accumulated and automatically averaged.

Preparation of Chromatin Fibers and Electric Linear Dichroism. The chromatin was extracted from chicken erythrocytes and purified according to procedures described in Hagmar et al. (1989). The alternating double-stranded polymers poly(dAT)₂ and poly(dGC)₂ (Pharmacia AB, Uppsala, Sweden) were used without further purification. Calf thymus DNA was deproteinized with sodium dodecyl sulfate (protein content <0.2%) and all nucleic acids were dialyzed against 1 mM sodium cacodylate buffered solution, pH 7.0. Details of the procedures used for electric linear dichroism (ELD) experiments were reported previously (Bailly et al., 1990; Colson et al., 1996). In the ionic strength conditions used for ELD measurements, chromatin (10-80 nucleosomes long) is mostly present as the 10-nm fiber. The optical setup incorporating a high sensitivity T-jump instrument equipped with a Glan polarizer was used under the following conditions: bandwidth, 3 nm, sensitivity limit, 0.001 in A/A; and response time, 3 µs. All experiments were conducted at 20°C with a 10-mm pathlength Kerr cell having 1.5-mm electrode separation (Houssier, 1981).

Purification and Radiolabeling of DNA Restriction Fragments. The 117- and 265-base pair DNA fragment were prepared by 3'-³²P-end labeling of the EcoRI-PvuII double digest of the plasmid pBS (Stratagene, La Jolla, CA) using -[³²P]dATP (3000 Ci/mmol) and avian myeloblastosis virus reverse transcriptase. The same procedure was applied for the 174-mer EcoRI-PvuII fragment from plasmid pKS (Stratagene). In each case, the digestion products were separated on a 6% polyacrylamide gel under native conditions in Tris-borate/EDTA-buffered solution (89 mM Tris-borate, pH 8.3, and 1 mM EDTA). After autoradiography, the band of DNA was excised, crushed, and soaked in water overnight at 37°C. This suspension was filtered through a 0.22-µm filter (Millipore Corporation, Bedford, MA) and the DNA was precipitated with ethanol. After washing with 70% ethanol and vacuum drying of the precipitate, the labeled DNA was resuspended in 10 mM Tris adjusted to pH 7.0 containing 10 mM NaCl.

DNase I Footprinting, Electrophoresis, and Quantitation by Storage Phosphorimaging. Experiments were performed essentially as described previously (Bailly and Waring, 1995). Briefly, reactions were conducted in a total volume of 10 µl. Samples (3 µl) of the labeled DNA fragments were incubated with 5 µl of the buffered solution containing the ligand at appropriate concentration. After 30-min incubation at 37°C to ensure equilibration of the binding reaction, the digestion was initiated by the addition of 2 µl of a DNase I solution whose concentration was adjusted to yield a final enzyme concentration of about 0.01 unit/ml in the reaction mixture. After 3 min, the reaction was stopped by freeze drying. Samples were lyophilized and resuspended in 5 µl of an 80% formamide solution containing tracking dyes. The DNA samples were then heated at 90°C for 4 min and chilled in ice for 4 min before electrophoresis.

Sequencing of Topoisomerase I-Mediated DNA Cleavage Sites. Each reaction mixture contained 2 µl of 3' end ³²P-labeled DNA (~1 µM), 5 µl of water, 2 µl of 10× topoisomerase I buffer, and 10 µl of drug solution at the desired concentration (1-100 µM). After 10-min incubation to ensure equilibration, the reaction was initiated by addition of 2 µl (20 units) of calf thymus topoisomerase I (Invitrogen, Carlsbad, CA). Samples were incubated for 45 min at 37°C before adding SDS to 0.25% and proteinase K to 250 µg/ml to dissociate the drug-DNA-topoisomerase I-cleavable complexes. The DNA was precipitated with ethanol and then resuspended in 5 µl of formamide-Tris borate-EDTA loading buffer, denatured at 90°C for 4 min, and then chilled in ice for 4 min before loading on to the sequencing gel. DNA cleavage products were resolved by polyacrylamide gel electrophoresis under denaturing conditions, as described above for the footprinting experiments.

DNA Relaxation Experiments. Supercoiled pKMp27 DNA (0.5 µg) was incubated with 4 units of human topoisomerase I or II (TopoGEN, Columbus, OH) at 37°C for 1 h in relaxation buffer (50 mM Tris pH 7.8, 50 mM KCl, 10 mM MgCl₂, 1 mM dithiothreitol, and 1 mM EDTA) in the presence of varying concentrations of the drug under study. Reactions were terminated by adding SDS to 0.25% and proteinase K to 250 µg/ml. DNA samples were then added to the electrophoresis dye mixture (3 µl) and electrophoresed in a 1% agarose gel at room temperature for 2 h at 120 V. Gels were stained with ethidium bromide (1 µg/ml), washed, and photographed under UV light. Similar experiments were performed using ethidium-containing agarose gels.

Cell Cultures and Survival Assay. Human CEM and CEM/C2 leukemia cells were obtained from the American Type Culture Collection (Manassas, VA). Cells were grown at 37°C in a humidified atmosphere containing 5% CO₂ in RPMI 1640 medium, supplemented with 10% fetal bovine serum, L-glutamine (2 mM), 1.5 g/l sodium bicarbonate, 4.5 g/l glucose, 10 mM HEPES, 1 mM sodium pyruvate, 100 IU/ml penicillin, and 100 µg/ml streptomycin. The cytotoxicity of the test compounds was assessed using a cell proliferation assay developed by Promega (Madison, WI) (CellTiter 96 AQ_ueous one solution cell proliferation assay). Briefly, 3 × 10⁴ exponentially growing cells were seeded in 96-well microculture plates with various drug concentrations in a volume of 100 µl. After 72-h incubation at 37°C, 20 µl of the tetrazolium dye was added to each well and the samples were incubated for a further 2 h at 37°C. Plates were analyzed on a Multiskan MS (type 352) reader (Labsystem, Helsinki, Finland) at 492 nm.

Cell Cycle Analysis. For flow cytometric analysis of DNA content, 10⁶ cells in exponential growth were treated with graded concentrations of the test drug for 24 h and then washed three times with citrate buffer. The cell pellet was incubated with 250 µl of trypsin-containing citrate buffer for 10 min at room temperature and then with 200 µl of citrate buffer containing a trypsin inhibitor and RNase (10 min) before adding 200 µl of propidium iodide (PI) at 125 µg/ml. Samples were analyzed on a FACScan flow cytometer (BD Biosciences, San Jose, CA) using the LYSYS II software, which is also used to determine the percentage of cells in the different phases of the cell cycle. PI was excited at 488 nm, and fluorescence was analyzed at 620 nm on channel Fl-3.

Bromodeoxyuridine Incorporation. Cells were cultured in complete RPMI 1640 medium with the test drug at different concentrations for 24 h before harvesting and then pulse-labeled with 10 µM BrdU for 60 min in complete medium. After two washes in phosphate-buffered saline, pH 7.3, with 0.1% sodium azide, cells were fixed in ethanol 70% and incubated 1 h at 4°C. After another wash in phosphate-buffered saline, cells were denatured in 2 N HCl for 15 min at 37°C (or 30 min at room temperature) under gentle stirring. pH was adjusted by a short incubation (5 min) in 3 ml of 0.1 M Na₂B₄O₇, pH 8.5, before centrifugation (5 min, 500g, 4°C). The cell pellet was then washed with 5 ml of buffer containing phosphate-buffered saline, 0.05% Tween, and 0.1% bovine serum albumin (fraction V), resuspended in 50 µl of this buffer, and then incubated with the fluorescein isothiocyanate-conjugated anti-BrdU monoclonal antibody (BD Biosciences) for 30 min at room temperature in the dark. For the negative controls, the pellet was incubated without antibody. All cell pellets were washed with 1 ml of buffer containing phosphate-buffered saline, 0.05% Tween, and 0.1% bovine serum albumin. Cells were collected by centrifugation and counterstained with 10 µg/ml of propidium iodide and treated with RNase (1 µg/ml). Samples were analyzed on a FACScan flow cytometer (BD Biosciences) using the LYSYS II software.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

DNA Binding Affinities. Different spectroscopic methods were used to compare the affinity of the drugs for DNA. Absorption measurements provided the first indication that the test compounds exhibit distinct binding affinities. Figure 2A shows the absorption spectra of AT2433-B1 and JDC-277 in the absence and presence of calf thymus DNA. The extent of hypochromism at the composite band with two maxima at 287 and 317 nm is much higher with AT2433-B1 compared with JDC-277 and the bathochromism is slightly more pronounced (8 versus 4 nm). Melting temperature measurements also suggested that the DNA affinity of AT2433-B1 is significantly higher than that of JDC-277. The histograms in Fig. 3 compare the Tm values (Tm = Tm^complex  Tm^DNA) values measured with each drug interacting with calf thymus DNA or the polynucleotide poly(dAT)₂ at a drug/DNA-phosphate ratio of 0.5. AT2433-B1 stabilizes duplex DNA against heat denaturation more strongly than its chlorinated analog AT2433-A1 and the neutral mono- and bis-glycosyl compounds are much less efficient than the analogs containing an amino-diglycoside. This is the first indication that the chemical structure of the sugar unit is essential for DNA binding.

View larger version (38K): [in this window] [in a new window]   Fig. 2.   Absorption (A) and circular dichroism spectra (B) of AT2433-B1 and JDC-277 in the presence of increasing concentrations of calf thymus DNA. DNA titrations of the drugs were performed in 1 mM sodium cacodylate buffer at pH 7.0. To 3 ml of drug solution at 20 µM was added aliquots of a concentrated calf thymus DNA solution. The phosphate-DNA/drug ratio increased from 0 to 20 in the direction indicated by the arrows.

View larger version (27K): [in this window] [in a new window]   Fig. 3.   Variation of the Tm (Tmdrug-DNA complex  TmDNA alone, in °C) of the complexes between the test compounds and () calf thymus DNA or () poly(dAT)2. Melting temperature measurements were performed in BPE buffer at pH 7.0 with a drug/DNA ratio of 0.5.

View larger version (40K): [in this window] [in a new window]   Fig. 4.   SPR sensorgrams for binding of AT2433-A1 and JDC-277 to the [AT]4 and [CG]4 DNA hairpin oligomers in HBS-EP buffer. In each case, the concentration of the unbound ligand in the flow solution varies from 0 to 5 µM (top curve).

DNA Binding Mode. Two spectroscopic methods using polarized light, CD and ELD, were deployed to investigate the orientation of the indolocarbazole chromophore with respect to the DNA helix. The CD spectra for AT2433-B1 and JDC-277 interacting with DNA (Fig. 2B) confirm the weaker binding of the neutral disaccharide compared with the cationic parent compound. For both compounds, the decrease of the CD band at 300 to 330 nm is concomitant to an increase of the CD amplitude at 340 to 360 nm, with a relatively well resolved isodichroic crossover. This type of CD behavior may be assigned to excitonic coupling between adjacent intercalated molecules. The ELD measurements also strongly suggest intercalation. The dependence of the reduced dichroism (A/A) on the wavelength and the electric field is shown in Fig. 5, A and B, for AT2433-B1 and JDC-277. For each drug-DNA complex (at a DNA/drug ratio of 25), A/A was measured at 320 nm and compared with that obtained with calf thymus DNA alone at 260 nm (Fig. 5C). In all cases, negative A/A values were measured, as expected for intercalating agents. With compounds AT2433-A1, JDC-277, and iso-AT2433-B1, the reduced dichroism was close to that of DNA alone, whereas more negative A/A values were obtained with AT2433-B1 and JDC-108. The difference most likely reflects the stronger binding capacity of these two compounds, as indicated by the Tm, SPR, and CD data. Although the amplitude of the ELD signals differs from one compound to another, the reduced dichroism is always negative in sign in the 300- to 350-nm region where the indolocarbazole chromophore absorbs the light and this is characteristic of an intercalative binding mode to DNA. It is possible that the indolocarbazole unit of the antibiotic is more tilted with respect to the base pair plane than that of JDC-277, or the two drugs exert different stiffening and unwinding effects on the DNA helix, but both compounds can be considered as DNA intercalators.

View larger version (21K): [in this window] [in a new window]   Fig. 5.   ELD data for the binding to DNA. Dependence of the reduced dichroism A/A on the wavelength (A) and electric field strength (B) for AT2433-B1 () and JDC-277 (). Conditions: 13.6 kV/cm, P/D = 20 (200 µM DNA, 10 µM drug) (A); and 320 nm, P/D = 20 for the DNA-drug complexes and 260 nm for the DNA alone (B). C, variation of the reduced dichroism (A/A) of the complexes between calf thymus DNA and the test compounds. A/A was measured in the absorption band of the ligand-DNA complex (320 nm) or at 260 nm for DNA alone, at 13.5 kV/cm and at a DNA/drug ratio of 25. All measurements were performed in 1 mM sodium cacodylate buffer, pH 7.0.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   ELD data for the binding to chromatin. Dependence of the reduced dichroism A/A on the wavelength (A), P/D ratio (B), and electric field strength (C) for AT2433-B1 (), JDC-277 (), or DNA alone (). D, A/A values measured for each drug-chromatin complex at 320 nm or for chromatin alone at 260 nm. Conditions: 13.6 kV/cm, P/D = 25 (250 µM chromatin, 10 µM drug) (A); 320 nm, 13.6 kV/cm (B); 320 nm, P/D = 25 for the DNA-drug complexes and 260 nm for the DNA alone (C) in 1 mM sodium cacodylate buffer, pH 7.0.

Sequence Selectivity. A DNase I footprinting study was carried out using three DNA restriction fragments of 117, 174, and 265 base pairs labeled at the 3' end with ³²P. With each fragment, the products of digestion by DNase I in the absence and presence of the test drugs were resolved by polyacrylamide gel electrophoresis. Typical gels obtained with AT2433-A1, AT2433-B1, and iso-AT2433-B1 are shown in Fig. 7. The two other uncharged compounds, JDC-108 and JDC-277, showed no effect on the cleavage of DNA by DNase I (data not shown) in sharp contrast to the antibiotic AT2433-B1, which profoundly perturbed the enzyme activity. Both the modification of the orientation of the sugar, as in iso-AT2433-B1, and the addition of a chlorine atom on the indolocarbazole chromophore, as in AT2433-A1, reduce significantly the capacity of the antibiotic to recognize defined sequences in DNA. Band intensities in the different gels were quantified by PhosphorImaging to obtain the differential cleavage plots presented in Fig. 8. The comparison between AT2433-B1 and its diastereoisomer fully confirms our previous observations that the configuration of the xylose subunit of the antibiotic is essential to the DNA interaction (Carrasco et al., 2002). The 3S,4S stereochemistry of the 2,4-dideoxy-4-methylamino-L-xylose subunit of AT2433-B1 seems to be well adapted to the DNA groove surface allowing the antibiotic to bind preferentially to sequences with a high GC content, such as 5'-CGGCCAG, 5'-GTCACG, and 5'-ACGGCC. On the contrary, the 3R,4R stereochemistry of the sugar moiety of iso-AT2433-B1 is apparently unfavorable for DNA sequence recognition. It is not rare to observe that a modification of the sugar moiety of a DNA binding antibiotic has a profound impact on its capacity to read the genetic information.

View larger version (71K): [in this window] [in a new window]   Fig. 7.   Sequence selective binding. The three gels show DNase I footprinting with the 117-mer (A), 174-mer (B), and 265-mer PvuII-EcoRI restriction fragments (C) in the presence of graded concentrations of AT2433-A1, AT2433-B1, and iso-AT2433-B1. In each case, the DNA was 3' end-labeled at the EcoRI site with -[32P]dATP in the presence of avian myeloblastosis virus reverse transcriptase. The products of nuclease digestion were resolved on an 8% polyacrylamide gel containing 7 M urea. Control tracks (Cont) contained no drug. Guanine-specific sequence markers obtained by treatment of the DNA with dimethyl sulfate followed by piperidine were run in the lanes marked G. Numbers on the left side of the gel refer to the standard numbering scheme for the nucleotide sequence of the DNA fragment, as indicated in Fig. 8.

View larger version (44K): [in this window] [in a new window]   Fig. 8.   Differential cleavage plots comparing the susceptibility of the 117-mer (A), 174-mer (B), and 265-mer PvuII-EcoRI restriction fragments (C) to DNase I cutting in the presence of AT2433-A1, AT2433-B1, and iso-AT2433-B1 (10 µM each). Deviation of points toward the lettered sequence (negative values) corresponds to a ligand-protected site and deviation away (positive values) represents enhanced cleavage. Vertical scales are in units of ln(fa)  ln(fc), where fa is the fractional cleavage at any bond in the presence of the drug and fc is the fractional cleavage of the same bond in the control, given a closely similar extent of digestion in each case.

Topoisomerase I Inhibition. The effect of the drugs on the catalytic activity of human topoisomerase I was investigated using a conventional plasmid DNA relaxation assay (Bailly, 2001). Two types of effects can be detected depending on the experimental conditions. In the absence of ethidium bromide in the agarose gel during the electrophoresis, the relaxation of supercoiled DNA by topoisomerase I gives a population of topoisomers and the presence of an intercalating agent affects the distribution of the topoisomers population due to an unwinding effect. The typical gel presented in Fig. 9A shows that JDC-277 produces dose-dependent alterations in plasmid linking number. With 5 µM JDC-277, the plasmid becomes fully relaxed and at higher concentrations the accumulation of positive supercoils reflects the unwinding of the DNA typical of an intercalating agent. The unwinding effect was more pronounced with AT2433-B1 than with JDC-277 (data not shown). To better differentiate the specific (poisoning) and nonspecific effects the same experiments were performed in parallel using ethidium bromide-containing agarose gels. In this case, the relaxed DNA migrates faster than the supercoiled plasmid due to ethidium-induced DNA unwinding effects. As shown in Fig. 9B, an increase in the intensity of the band corresponding to nicked DNA molecules can be detected with JDC-277, although the effect is weak compared with what can be achieved with the reference topoisomerase I poison camptothecin (lane CPT in Fig. 9B). However, there is no doubt that JDC-277 stabilizes topoisomerase I-DNA complexes. Similar experiments performed with the five compounds and the results are presented in Fig. 9C. The antibiotic AT2433-B1, which has the highest affinity for DNA in the series, strongly reduces the electrophoretic mobility of the supercoiled DNA band but does not stabilize topoisomerase I-DNA complexes. In contrast, the two compounds with the lowest affinity for DNA, JDC-277 and JDC-108, both induce an increase of the intensity of the band corresponding to the nicked form of DNA. Tight binding to DNA is not required, and even detrimental, for trapping the covalent DNA-topoisomerase I complexes.

View larger version (23K): [in this window] [in a new window]   Fig. 9.   Topoisomerase I inhibition. Effect of increasing concentrations of JDC-277 on the relaxation of plasmid DNA by human topoisomerase I in the absence (A) or in the presence (B) of ethidium bromide during the electrophoresis. C, effect of the different drugs (20 µM each). Native supercoiled pKMp27 DNA (0.5 µg) (lane DNA) was incubated with 4 units topoisomerase I in the absence (lane TopoI) or presence of drug at the indicated concentration (micromolar). Reactions were stopped with sodium dodecylsulfate and treatment with proteinase K. DNA samples were separated by electrophoresis on 1% agarose gels. The gel was stained with ethidium bromide after the electrophoresis (A) or contained ethidium (1 µg/ml) before the electrophoresis (B and C). Nck, nicked; Rel, relaxed; Sc, supercoiled.

1

1

1

View larger version (38K): [in this window] [in a new window]   Fig. 10.   Cleavage of the 174-mer (A) and the 117-mer DNA fragments (B) by human topoisomerase I in the presence of the indolocarbazoles. In both cases, the 3' end-labeled fragment (DNA) was incubated in the absence (lane TopoI) or presence of the test drug at the indicated µM concentration. CPT was used at 50 µM. Topoisomerase I cleavage reactions were analyzed on an 8% denaturing polyacrylamide gel. Numbers at the left side of the gels show the nucleotide positions, determined with reference to the guanine tracks labeled G. The nucleotide positions and sequences to the cleavage sites are indicated.

Cytotoxicity. A tetrazolium-based assay was applied to determine the drug concentration required to inhibit cell growth by 50% after incubation in the culture medium for 72 h. The calculated IC₅₀ values with the CEM human leukemia cell line are collated in Table 2. The two diglycoside antibiotics AT2433-A1 and AT2433-B1 proved to be the most cytotoxic compounds of the series with IC₅₀ values 5 to 10 times lower than that of the uncharged mono and diglycosides JDC-108 and JDC-277. Iso-AT2433-B1 was the least cytotoxic compound with an IC₅₀ value 20 times higher than that of the parent diastereoisomer AT2433-B1. The altered configuration of iso-AT2433-B1 is highly detrimental to the cytotoxic action and the substitution of the indolocarbazole chromophore with a chlorine atom also slightly reduces the cytotoxicity.

Cell Cycle Effects. Treatment of human leukemia CEM cells with increasing concentrations of AT2433-B1, but not with its diastereoisomer, for 24 h led to marked changes of the cell cycle profiles (Fig. 11). The flow cytometric analysis of propidium iodide-labeled cells indicates that the treatment with 0.5 µM AT2433-A1 or -B1 induces a significant accumulation of cells in the S phase. The S-cell population increases from 23% in the control to 41% in the presence of 1 µM AT2433-B1. To get a more precise view of the drug effect at the S-phase level, cells were labeled with BrdU, a thymidine analog that incorporates into newly synthesized strands of DNA. As shown in Fig. 12, fewer BrdU⁺ cells, labeled in S phase at the time of treatment, were detected after 24 h of treatment with AT2433-B1 than in the control cell population. The drug potently inhibits the incorporation of BrdU into DNA and therefore reduces the entry of cells into the vulnerable DNA synthesis phase of the cell cycle. Cells with DNA content less than G1 can be also detected with this compound, and to a lower extent with the chlorinated analog AT2433-A1 (Fig. 11). Sub-G1 cells are usually considered as apoptotic cells (Kluza et al., 2000). No such effects were observed with JDC-277 and with the diastereoisomer iso-AT2433-B1. The cell cycle profiles also remained unchanged, even when using a high drug concentration (data not shown).

View larger version (36K): [in this window] [in a new window]   Fig. 11.   Cell cycle analysis of CEM human leukemia cells treated for 24 h with the different compounds at 0.5 µM (A) and increasing concentrations of AT2433-B1 (B). Cells were analyzed with the FACScan flow cytometer.

View larger version (43K): [in this window] [in a new window]   Fig. 12.   Modification of bromodeoxyuridine incorporation in CEM cells treated with AT2433-B1. Cells were treated for 24 h with the test drug at the indicated concentrations before labeling with the BrdU-fluorescein isothiocyanate conjugate for 1 h and counterstained with PI.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The study reported herein was initiated on the basis of the structural analogy between the indolocarbazole diglycoside AT2433-A1 and the indolocarbazole monoglycoside rebeccamycin. We hypothesized that the replacement of the methoxyglucose moiety of rebeccamycin with a 2,4-dideoxy-4-methylamino-L-xylose disaccharide could preserve the capacity of the drug to interact with DNA and topoisomerase I. Not surprisingly, the cationic aminosugar subunit of the AT2433 antibiotics is a positive element for DNA recognition but, unexpectedly, it is unfavorable for topoisomerase I poisoning. In the following section, these two aspects are discussed in turn because we know from our previous studies with different series of indolocarbazole glycosides that DNA interaction and topoisomerase I inhibition are two different and essentially unrelated aspects of the molecular mechanism of action of these compounds (Anizon et al., 1997; Bailly et al., 1997, 1999a,b, 2000). In the NB-506 series, it is clear that the drug needs not to intercalate into DNA to inhibit topoisomerase I and exert its cytotoxic activity (Bailly et al., 1999a).

DNA Binding and Sequence Recognition. The mechanism of action of the antibiotic AT2433-A1 had never been studied before the present work. For the last 20 years, this indolocarbazole diglycoside was in some way an orphan drug with no identified molecular target(s). We show herein that DNA is a receptor for AT2433-A1 and its dechlorinated analog AT2433-B1. In a very recent study, we disclosed that AT2433-B1 preferentially recognizes GC-rich sequences in DNA to form very stable, slow-dissociating complexes, whereas its diastereoisomer iso-AT2433-B1 exhibits a very weak sequence preference (Carrasco et al., 2002). The present study extends these observations to confirm that the configuration of the amino-xylose subunit of AT2433-B1 is essential to the DNA interaction. The presence of a chlorine atom on the indolocarbazole ring system slightly reduces the capacity of the drug to interact with DNA. The Tm and SPR data concur to show that AT2433-B1 binds more tightly to DNA than its chlorinated analog AT2433-A1, which, however, maintains the capacity to discriminate between AT- and GC-rich sequences. These observations are reminiscent to those previously made with rebeccamycin: the removal of the two chlorine atoms significantly promoted DNA intercalation (Bailly et al., 1997). There is no doubt that the cationic diglycoside moiety of the AT2433 antibiotics plays a major role in their tight interaction with DNA. Both the charge and the configuration are important. Altering the orientation of the sugar residue, as in iso-AT2433-B1, and the substitution of a melibiose disaccharide (as in JDC-277) for the aminosugar subunit of AT2433-B1, decreases DNA affinity by a factor of 10 or more, depending on the target sequence.

Stimulation of Topoisomerase I-Mediated DNA Strand Breaks. The replacement of the monoglycoside of rebeccamycin-type compounds with a cationic diglycoside residue is attractive for at least two essential reasons: 1) it potentially offers expanded DNA sequence recognition capacities, and 2) it increases considerably the aqueous solubility of the compounds and therefore should facilitate drug formulation and administration. But these two key points are valid only if the main target of the drug, topoisomerase I, is preserved (unless another drugable target could be identified). This is not the case here. Neither AT2433-A1 nor AT2433-B1 was capable of stimulating topoisomerase I-mediated DNA cleavage. The monoglucoside derivative JDC-108 is a weak inhibitor of topoisomerase I. The addition of a second uncharged sugar residue does not reinforce the interaction with DNA but has a positive impact on the capacity of the drug to inhibit topoisomerase I. Indeed, JDC-277 and JDC-108 exhibit comparable affinities for DNA (as judged from the Tm data) but the galactose-glucosyl (i.e., melibiose) diglycoside unit of JDC-277 potentiates the DNA nicking activity of topoisomerase I slightly more than the glucosyl monoglycoside JDC-108 (compare the extent of DNA cleavage in Fig. 10A). In sharp contrast, the incorporation of the methylamino-xylose carbohydrate unit abolishes topoisomerase I inhibition in spite of the reinforced DNA interaction. The mode of DNA binding rather than the binding affinity may be a critical determinant of topoisomerase I inhibition by indolocarbazole glycosides.

Relation to Cytotoxicity. There is no direct relationship between cytotoxicity and ability to stimulate DNA cleavage by topoisomerase I. The only diglycoside that stimulates DNA cleavage by topoisomerase I, JDC-277, is 10 times less cytotoxic than the parent antibiotic AT2433-B1 devoid of activity against the nuclear DNA relaxing enzyme. Parenthetically, it is worth mentioning that none of the drugs used in this study stabilize topoisomerase II-DNA complexes (data not shown). In contrast, there may be a relationship between cytotoxicity and affinity constants for DNA. The two AT2433 antibiotics that display the highest affinity for GC-rich sequences in DNA are the most cytotoxic compounds. The modification of the stereochemistry of the aminosugar residue provides iso-AT2433-B1, which exhibits a decreased DNA affinity and is also substantially less potent at inhibiting cell growth than the parent antibiotic. To paraphrase the common dogma in the anthracycline series (Arcamone et al., 1997), we can conclude that DNA complexation is certainly an essential step in the sequence of events leading to the inhibition of tumor cells by the indolocarbazole diglycoside antibiotic AT2433. The consideration that DNA interaction contributes to the cytotoxic action of the indolocarbazole diglycosides sets a direction for further work to enhance our understanding of the mechanism of action of these compounds and the development of a new generation of indolocarbazoles that are more effective against tumor cells.