Self-Assembly of an Oligodeoxyribonucleotide Harboring the Estrogen

Departments of Environmental and Community Medicine and Medicine, The ... Occupational Health Sciences Institute, and The Cancer Institute of New Jers...
0 downloads 0 Views 166KB Size
Download PDF
Biomacromolecules 2000, 1, 339-349

339

Self-Assembly of an Oligodeoxyribonucleotide Harboring the Estrogen Response Element in the Presence of Polyamines: Ionic, Structural, and DNA Sequence Specificity Effects Joan S. Lewis,†,‡,§ T. J. Thomas,|,⊥,§ Akira Shirahata,# and Thresia Thomas*,†,‡,⊥ Departments of Environmental and Community Medicine and Medicine, The Environmental and Occupational Health Sciences Institute, and The Cancer Institute of New Jersey, University of Medicine and Dentistry of New JerseysRobert Wood Johnson Medical School, New Brunswick, New Jersey 08903; Nutritional Sciences Graduate Program, RutgerssThe State University of New Jersey, New Brunswick, New Jersey 08903; and Faculty of Pharmaceutical Sciences, Josai University, Sakado, Saitama 350-02, Japan Received February 28, 2000; Revised Manuscript Received May 16, 2000

Estrogenic regulation of gene expression is mediated by the binding of the hormone to its specific receptor, estrogen receptor (ER), which undergoes structural and conformational alterations to recognize specific DNA sequences, estrogen response elements (ERE), in responsive genes to trigger a series of events culminating in the transcription of these genes. Polyamines are ubiquitous cellular cations that are important for cell growth and differentiation, and have been shown to participate in estrogenic regulation of gene expression. Polyamine-mediated DNA condensation/aggregation has been studied to understand the ionic and structural requirements for the compaction of DNA. DNA condensation/decondensation may also play a role in transcription and replication. We studied the aggregation of a 38-mer oligonucleotide duplex (ODN) in the presence of natural and synthetic polyamines under different ionic conditions (NaCl, KCl, and K glutamate). Our results showed that an ODN harboring the consensus ERE (ODN1) was 2-fold more susceptible to precipitation by spermine compared to ODN2 containing scrambled sequences, or a mutant ODN (ODN3). The nature of the monovalent cations (Na+ vs K+), and anions (Cl- vs glutamate) also played an important role in the efficacy of a polyamine to precipitate ODNs: potassium glutamate being the least effective in suppressing the ability of spermine to precipitate ODNs. The concentration of polyamines required for precipitating the ODNs increased with monovalent ion concentration in the buffer. With ODN1, a plot of log[spermine4+] at the 50% precipitation concentrations against log[Na+/K+] yielded a straight line, with a slope of 1.8 ( 0.18, a value comparable to that predicted by the counterion condensation theory (1.85). We also observed significant structural specificity effects of spermine and its analogues [NH2(CH2)3NH(CH2)nNH(CH2)3NH2, where n ) 2-9; n ) 4 for spermine] on aggregating the ODN1. These results demonstrate DNA sequence and polyamine structural specificity effects on the aggregation of ODNs, and suggest that the gene regulatory function of ERE may be linked to its ability to undergo facile condensation/ decondensation in the presence of biological cations, such as polyamines. Introduction In eukaryotic cells, the packaging of DNA within the cell nuclei is accomplished by its interaction with histones, nonhistone proteins and polyamines.1-5 During replication * Corresponding author. Clinical Academic Building, Room 7090, UMDNJ-Robert Wood Johnson Medical School, 125 Paterson Street, New Brunswick, NJ 08903. Telephone: (732) 235-8458. Fax: (732) 235-8473. E-mail: [email protected]. † Department of Environmental and Community Medicine, University of Medicine and Dentistry of New JerseysRobert Wood Johnson Medical School. ‡ The Environmental and Occupational Health Sciences Institute, University of Medicine and Dentistry of New JerseysRobert Wood Johnson Medical School. § Nutritional Sciences Graduate Program, RutgerssThe State University of New Jersey. | Department of Medicine, University of Medicine and Dentistry of New JerseysRobert Wood Johnson Medical School. ⊥ The Cancer Institute of New Jersey, University of Medicine and Dentistry of New JerseysRobert Wood Johnson Medical School. # Josai University.

and transcription, however, the highly compacted DNA of the nucleosome becomes decondensed so that polymerases and transcription factors can interact with it.1,2 This transient accessibility might be regulated, in part, by changes in polyamine levels.2-5 Furthermore, studies by Krasnow and Cozzarelli5 showed that the catenation of DNA by topoisomerases was dependent on the ability of DNA to be aggregated by polyamines. The natural polyamines, putrescine [H2N(CH2)4NH2], spermidine [H2N(CH2)4NH(CH2)3NH2], and spermine [H2N(CH2)3NH(CH2)4NH(CH2)3NH2], are ubiquitous cellular cations which are present in millimolar concentrations in all cells and are implicated in many biological processes.6-9 The intracellular levels of polyamines are highly regulated by stimulatory and inhibitory agents and their presence has been shown to be obligatory for cell growth and differentiation. Under physiological pH and ionic conditions, polyamines are positively charged, and hence, a target of their interaction

10.1021/bm000010s CCC: $19.00 © 2000 American Chemical Society Published on Web 06/29/2000

340

Biomacromolecules, Vol. 1, No. 3, 2000

is the negatively charged phosphate groups of DNA.10-12 Polyamine-DNA interaction is known to stabilize duplex DNA and induce structural alterations, such as condensation, aggregation, and ψ-DNA formation in random sequence DNA.13-17 There is also some evidence to the concept that polyamines interact with DNA in a sequence-specific manner,18,19 resulting in the induction and stabilization of unusual DNA conformations, including left-handed Z-DNA,20-22 triplex DNA,23 A-DNA,24 and bent DNA.25,26 In addition, several studies have reported that DNA sequences which are capable of undergoing unusual DNA conformations in the presence of polyamines are also more prone to polyamineinduced aggregation as compared to their random sequence counterparts.27,28 Interestingly, several genes contain enhancer elements which are potentially capable of adopting unusual DNA conformations in the presence of suitable environmental conditions, including the presence of the endogenous cations, polyamines.29,30 We and others have previously reported the involvement of polyamines in the mechanism of action of estrogens in breast cancer cells.31-33 Estrogens are the most often implicated agents in the origin and progression of breast cancer.34 Estrogens exert their biological effects through the estrogen receptor (ER), a ligand-activated transcription factor present in the majority of breast tumors. ER stimulates the transcription of an array of estrogen-responsive genes by recognizing and binding with high affinity to EREs.35-37 The EREs are highly conserved palindromic sequences, usually located upstream of the promoter of estrogen-regulated genes and contain one or more copies of a core inverted repeat sequence, 5′-GGTCANNNTGACC-3′.37,38 Polyamines have already been shown to enhance ER-ERE interactions.34,38 This enhanced interaction appears to be a result of polyamineinduced conformational changes within the ERE.39 In this study, we further explored the specificity of polyamine interactions with ERE. We studied the effect of natural and synthetic polyamines (differing in structure by either chain length modification or ethyl group substitution at the pendant amino groups) on the aggregation of three oligonucleotide duplexes; one containing the ERE (ODN1), one control consisting of scrambled sequences (ODN2), and another control containing a two base pair mutation in the ERE sequence (ODN3). Our results showed that ODN1, containing ERE, was 2-fold more susceptible to undergo spermine-induced aggregation than ODN2 or ODN3. The ability of spermine to aggregate DNA was strongly influenced by environmental and structural factors, such as the cationicity, ionic strength, and the number and disposition of the methylene bridging region between the primary and secondary amino groups of polyamines. Materials and Methods Polyamines. Putrescine‚2HCl, spermidine‚3HCl, and spermine‚ 4HCl were purchased from Sigma Chemical Co. (St. Louis, MO). All spermine analogues, 1,10-diamino-4,7-diazadecane (3-2-3), 1,11-diamino-4,8-diazaundecane (3-3-3), N,N′-bis(3-aminopropyl)-1,5-diaminopentante (3-5-3), N,N′-bis(3-aminopropyl)-1,6diaminohexane (3-6-3), N,N′-bis(3-aminopropyl)-1,7-diaminoheptane (3-7-3), N,N′-bis(3-aminopropyl)-1,8-diaminooctane (3-

Lewis et al.

Figure 1. Chemical structures of natural and synthetic polyamines.

8-3), N,N′-bis(3-aminopropyl)1,9-diaminononane (3-9-3), and N1,N12-bis(ethyl)spermine (BE-3-4-3) were synthesized according to procedures previously described.40,41 The structures and purity of all polyamines were confirmed by elemental analysis, NMR, HPLC, and mass spectrometry. The abbreviated names and chemical structures of these compounds are shown in Figure 1. Concentrated stock solutions of the polyamines were prepared in 10 mM cacodylate buffer (10 mM sodium cacodylate, pH 7.4, and 0.5 mM EDTA) and were stored at 4 °C. For precipitation experiments, appropriate dilutions of stock polyamine concentrations were made in the same buffer. Oligonucleotides. HPLC-purified oligonucleotides (ODNs) were purchased from Oligos, Etc., Inc. (Wilsonville, OR). The ERE consensus sequence from the vitellogenin gene and its flanking sequences, which are important in ER-ERE binding, were used for this study.42 The control oligonucleotide consisted of a similar distribution of purines and pyrimidines and GC content, arranged in a scrambled sequence (ODN2). Another oligonucleotide (ODN3) containing a two base pair rearrangement in the consensus ERE (referred to as mutant ERE) was also used. The base sequences of the oligonucleotides used in this study are as follows

ODN1: 5′-GATCCAGGTCAGAGTGACCTGAGCTAAAATAACACATTCAG-3′ 3′-GGTCCAGTCTCACTGGACTCGATTTTATTGTGTAAGC-5′ ODN2: 5′-AAAGCTCGCTTCCTGAAGACGTTCTCGAAGAGAAATCTCTT-3′ 3′-CGAGCGAAGGACTTCTGCAAGAGCTTCTCTTTAGAGAA-5′ ODN3: 5′-GATCCAGGTCAGAGTGCACTGAGCTAAAATAACACATTCAG-3′ 3′-GGTCCAGTCTCACGTGACTCGATTTTATTGTGTAAGTC-5′ The ODNs were dissolved in 10 mM cacodylate buffer and dialyzed three times against the same buffer before use. The concentration of ODNs was determined by measuring the absorbance of solutions in 10 mM cacodylate buffer at 260 nm, and

Self-Assembly of Oligonucleotides by Polyamines molar concentrations were calculated by using the extinction coefficients of 10 243 and 9210 M-1 cm-1, for upper and lower strands of ODN1, and 9512 and 9736 M-1 cm-1 for upper and lower strands of the ODN2, and 10 146 and 9713 M-1 cm-1 for upper and lower strands of ODN3, respectively. Centrifugation Assay. Samples for precipitation experiments were prepared by mixing 1 µM of each oligomer in 10 mM cacodylate buffer with appropriate concentrations of polyamines, NaCl, KCl, and K glutamate (KGlu) in 400 µL of buffer. After vortexing for 15 s, samples were boiled for 10 min, cooled for 30 min at room temperature, and then allowed to equilibrate overnight at 4 °C. Samples were then centrifuged for 10 min at 11300g, and the amount of DNA in the supernatant was determined by measurement of the absorbance at 260 nm. All absorbance measurements were performed with a Beckman DU 640 spectrophotometer, using a 350 µL quartz microcuvette (path length ) 1 cm). The amount of ODN in solution was calculated using the equation: percent DNA in solution ) (As/Ao) x 100; where As and Ao are the absorbance of the supernatant and control (absorbance of ODN in 10 mM sodium cacodylate buffer without any polyamines) at 260 nm, respectively. EC50 values were calculated from polyamine concentrations vs percent DNA in solution plots. EC50 values were reproducible within 5% in repeated experiments. Polyacrylamide Gel Electrophoresis. Single-stranded ERE oligonucleotides were labeled with 32P-γ-ATP using a 5′ end labeling kit from Boehringer Mannheim (Indianapolis, IN). The labeled oligonucleotide was purified using Chroma Spin-10 column from Clontech (Palo Alto, CA). Equimolar concentrations of labeled single-stranded ERE and its unlabeled complementary strand were mixed together in 10 mM sodium cacodylate buffer (pH 7.4) to prepare the duplex ODN. The labeled ODN was mixed with appropriate concentrations of spermine, samples were boiled for 10 min, cooled, and allowed to equilibrate overnight at 4 °C, as in the case of the precipitation experiments. The samples were loaded on a 12% polyacrylamide gel, and electrophoresis was performed at 4 °C in 0.5× Tris-borate buffer (89 mM Tris base, 89 mM boric acid, 2 mM EDTA, pH 8.3) for 3 h at 200 V. The gels were then placed in fixer buffer (20% methanol and 10% acetic acid in ddH2O) for 15 min and dried in gel drying buffer (10% glycerol and 20% ethanol) for 30 min. The gels were dried overnight and exposed to Kodak Biomax MR-1 film for autoradiography for 1224 h. Melting Temperature Measurements. The Tm studies were carried out using a Beckmann DU640 spectrophotometer interfaced with an IBM computer.22 The Tm block consists of six cells, each with a volume of ∼0.35 mL, of which the first was filled with buffer and used as the blank. All the measurements were done at a heating rate of 0.5 °C/min using a Beckman high performance temperature controller, with the absorbance and temperature recorded every 30 s. For Tm measurements, duplex ERE (prepared as described in the centrifugation experiments) was incubated with increasing concentrations of spermine and various concentrations of NaCl in a buffer containing 10 mM Na cacodylate (pH 7.4) and 0.5 mM EDTA. In a parallel set of experiments, the absorbancetemperature profile of the individual strands of ODN1 was also recorded to examine the presence of possible secondary structures, including hairpins and cruciforms, in these single strands in the presence of spermine. Tm was taken as the temperature corresponding to half-dissociation of the duplex and the reproducibility was within 1 °C between individual experiments.

Results Effect of Spermine on the Aggregation of ODN1 under Different Ionic Conditions. Figure 2A shows the effect of

Biomacromolecules, Vol. 1, No. 3, 2000 341

Figure 2. Aggregation of ERE-containing oligonucleotide ODN1 by spermine (3-4-3) in the presence of 10 (b), 25 (O), 35 (1), 50 (3), 75 (9), 100 (0), and 150 ([) mM of NaCl (A), KCl (B), and KGlu (C). A logarithmic scale is used for spermine concentrations. All experiments were conducted in 10 mM sodium cacodylate buffer (pH 7.4). The solubility of DNA was calculated using the following equation: % DNA in solution ) (As/Ao) × 100, where As and Ao are the absorbance of the supernatant and control (absorbance of DNA in 10 mM sodium cacodylate buffer without any polyamines) at 260 nm, respectively.

spermine on the precipitation/aggregation of ODN1 in the presence of 10, 25, 35, 50, 75, 100, and 150 mM NaCl. An increase in Na+ concentration in the buffer caused a significant increase in the concentration range at which spermine precipitated ODN1. In the presence of 10 mM Na+ (cacodylate buffer alone), precipitation of ODN1 was initiated at 25 µM spermine, and the EC50 value (concentration of polyamine necessary to precipitate 50% ODN from solution) was 45 µM spermine. Eighty-five percent of the ODN1 was precipitated by 100 µM spermine. An increase in Na+ concentration to 25 mM caused a 2-fold reduction in the ability of spermine to aggregate ODN1 (Table 1). Complete precipitation of ODN1 did not occur until 500 µM spermine was used. As the concentration of Na+ was increased further, the ability of spermine to aggregate ODN1 became progressively less effective, and EC50 increased by 80-fold, from 0.045 to 3.6 mM, as Na+ concentration

342

Biomacromolecules, Vol. 1, No. 3, 2000

Lewis et al.

Table 1. Spermine Concentrations Required to Precipitate 50% of the Oligonucleotides (EC50) under Different Na+ Concentrations

Table 2. EC50 Values of Spermine in the Presence of Potassium Salts

EC50, µM Na+ concentration, mM 10 25 35 50 75 100

ODN1 45 90 170 650 1600 3600

ODN2 55 190 250 950 3500 6500

EC50, µM ODN3 65 175 250 760 2200 4500

Figure 3. Concentrations of spermine which caused 50% precipitation (EC50) of ODN1 plotted as a function of NaCl (b), KCl (O), and KGlu (1) concentrations on a log-log scale. The straight lines were obtained by least-squares linear regression, and slopes were calculated. The slope values were 1.98, 1.6, and 1.5 for NaCl, KCl, and KGlu, respectively, with a >97% goodness of fit.

increased from 10 to 100 mM (Table 1). At 150 mM Na+, there was no detectable precipitation of ODN1 up to 5 mM spermine. A plot of log[spermine4+] at the EC50 concentrations against log[Na+] gave a straight line (Figure 3) with a slope of 1.98. This value is a quantitative measure of the concentration dependence between the multivalent and monovalent ions in precipitating the ODNs and is related to the coefficient of ln c2 in the counterion condensation theory (please see Discussion). Figure 2B shows the effect of spermine on the precipitation of ODN1 at different concentrations of KCl. As with Na+, increase in K+ decreased the ability of spermine to precipitate ODN1; however, KCl was less effective than NaCl in blocking the effect of spermine. The maximal difference between KCl and NaCl was observed at 100 mM concentrations of these monovalent ions. The EC50 for spermineinduced precipitation of ODN1 was 1.6 mM spermine in the presence of 100 mM KCl, compared to that of 3.6 mM with NaCl. Overall, the EC50 values of spermine increased by 40fold as K+ increased from 10 to 100 mM (Table 2). The slope of the log[spermine4+] vs log[K+] plot (Figure 3) was 1.59 in this case. Figure 2C shows the effect of spermine on the precipitation of ODN1 at various concentrations of KGlu. We used KGlu in these experiments because of its physiological importance and the observation that it facilitates ER-ERE interactions.43,44 KGlu had the least inhibitory effect on spermine-

concentration of K+, mM

anion

ODN1

ODN2

10 25 35 50 75 100

Cl-

40 80 160 250 750 1600

70 160 250 750 2150 3500

10 25 35 50 75 100 150

Glu

25 80 95 195 600 950 1500

50 95 210 285 780 2500 4200

induced precipitation of ODN1 compared to KCl and NaCl. The concentration range of spermine at which 50% precipitation of ODN1 occurred was significantly reduced when KGlu was substituted for NaCl. EC50 value of spermine increased from 0.025 to 0.95 mM as the concentration of KGlu increased from 10 to 100 mM. Spermine precipitated ODN1 at 150 mM KGlu with an EC50 value of 1.5 mM. The slope of the log[spermine4+] vs log [KGlu] plot (Figure 3) was 1.5. These results indicate significant differences between the monovalent ions in suppressing the ability of spermine to precipitate the ODN1. We also examined the ability of spermine to aggregate the ODNs in cacodylate buffer containing the approximate physiologic composition of cations (120 mM KCl, 10 mM NaCl, 2 mM MgCl2, and 0.1 mM CaCl2). The EC50 value of spermine determined under this condition was 4.5 mM. Studies on trivalent spermidine and divalent putrescine on the precipitation of ODN1 showed that these polyamines had no significant effect on the precipitation even at concentrations as high as 10 and 15 mM, respectively (data not shown). DNA Sequence Specificity of Spermine-induced Aggregation of ODNs. We next examined the role of sequencespecific interaction of spermine in inducing the aggregation of the oligonucleotides using the control oligonucleotide, ODN2. Figure 4A shows the effect of spermine on the precipitation of ODN2 at various concentrations of NaCl. As it was observed for ODN1, there was a general increase in EC50 values with increasing NaCl concentrations (Table 1). More interestingly, however, spermine was less effective in precipitating ODN2 than ODN1 at all the ionic concentrations examined, except at 10 mM Na+. The EC50 value increased 118-fold from 0.055 to 6.5 mM spermine as the concentration of Na+ was increased from 10 to 100 mM. A plot of log[spermine4+] against log[Na+] gave a straight line (Figure 5) with a slope of 2.1. Figure 4B shows the effect of spermine on the precipitation of ODN2 at various KCl concentrations. Similar to the observation with ODN1, the effect of KCl in blocking spermine-induced precipitation was lower than that of NaCl. The EC50 values for precipitating ODN2 were consistently higher than those of ODN1 at all KCl concentrations. The EC50 increased 70-fold from 0.050 to 3.5 mM as the KCl

Self-Assembly of Oligonucleotides by Polyamines

Biomacromolecules, Vol. 1, No. 3, 2000 343

Figure 5. Concentrations of spermine which caused 50% precipitation (EC50) of ODN2 plotted as a function of NaCl (b), KCl (O), and KGlu (1) concentrations on a log-log scale. The straight lines were obtained by least-squares linear regression. The slope values of these lines were 2.13, 1.92, and 1.90 for NaCl, KCl, and KGlu, respectively, with a >95% goodness of fit.

Figure 4. Precipitation/aggregation of control oligonucleotide (ODN2) by spermine in the presence of 10 (b), 25 (O), 35 (1), 50 (3), 75 (9), 100 (0), and 150 ([) mM of NaCl (A), KCl (B), and KGlu (C). A logarithmic scale is used for spermine concentrations. All experiments were conducted in 10 mM sodium cacodylate buffer (pH 7.4).

concentration increased from 10 to 100 mM. The slope of the log[spermine4+] against log[KCl] plot was 1.9. Figure 4C shows the effect of spermine on the aggregation of ODN2 at various KGlu concentrations. The general trend of requiring higher concentration of spermine to precipitate ODN2 compared to ODN1 was persistent in the presence of KGlu. In addition, KGlu was less effective than KCl and NaCl in blocking spermine-induced precipitation of ODN2. The EC50 values increased 50-fold from 0.05 to 2.5 mM as the concentration of KGlu increased from 10 to 100 mM. The slope of the log[spermine4+] against log[KGlu] plot (Figure 5) was 1.85. The EC50 value of spermine under the approximate physiologic cation composition was 10.5 mM. We next examined the effect of spermine on the precipitation of ODN3 and the results are presented in Figure 6 and Table 1. The EC50 values of ODN3 were intermediate to those of ODN1 and ODN2. There was a 113-fold increase in the EC50 of spermine in precipitating ODN3 as the concentration of Na+ was increased from 10 to 100 mM. A straight line was obtained by plotting log[spermine4+] against log[Na+], with a slope of 2.11 (Figure 6B).

Figure 6. Precipitation/aggregation of mutant ERE containing oligonucleotide ODN3 by spermine in the presence of 10 (b), 25 (O), 35 (1), 50 (3), 75 (9), 100 (0), and 150 ([) mM of NaCl (A). A loglog plot of [spermine] vs [Na+] concentrations at the EC50 concentrations is shown in panel B. The slope of the straight line was 2.11, with a 96% goodness of fit.

Effect of Chain Length Modification and Bis(ethyl) Substitution of Spermine on the Precipitation of ODN1. To differentiate between the ionic and structural specificity effects of spermine on the precipitation/aggregation of

344

Biomacromolecules, Vol. 1, No. 3, 2000

Figure 7. Precipitation/aggregation of ODN1 by 3-3-3 (A), 3-5-3 (B), 3-6-3 (C), and 3-7-3 (D) in the presence of 10 (b), 25 (O), 35 (1), 50 (3), and 75 (9) mM NaCl. A logarithmic scale is used for spermine analogue concentrations. All experiments were conducted in 10 mM sodium cacodylate buffer (pH 7.4).

ODN1, we studied a series of spermine analogues [NH2(CH2)3NH(CH2)nNH(CH2)3NH2 (where n ) 2-9; n ) 4 for spermine)] in our precipitation experiments. These studies were conducted in cacodylate buffer containing 10-75 mM NaCl and/or increasing concentrations of spermine analogues. Figure 7A shows the effect of 3-3-3, which has one -CH2- group less than that of spermine, on the precipitation of ODN1. Removal of one methylene group did not affect the ability of spermine to precipitate ODN1, and 3-3-3 was very efficient at precipitating ODN1 at 10-75 mM Na+ concentrations. Interestingly, 3-3-3 was about 2-fold more effective than spermine in precipitating ODN1 in the presence of 50 mM NaCl. Efficacy of 3-3-3 was higher than spermine at 75 mM NaCl also. The EC50 value of 3-3-3 in precipitating ODN1 increased 32-fold as the Na+ concentration was increased from 10 to 75 mM. In this case also, a straight line was obtained by plotting log[polyamine4+] against log[Na+], with a slope of 1.66 (Table 4). Panels B, C, and D of Figure 7 show the effect of 3-53, 3-6-3, and 3-7-3, respectively, on the precipitation of ODN1. These analogues, which have one, two or three -CH2- groups more than spermine, were 2-, 3-, and 4-fold less effective than spermine at inducing the aggregation of ODN1 at 25 mM Na+. The EC50 values for these analogues at 10 to 50 mM NaCl concentrations ranged between 0.085 and 0.95 mM, 0.085 and 2 mM, and 0.090 and 2.5 mM for 3-5-3, 3-6-3, and 3-7-3, respectively (Table 3). We also determined the efficacy of 3-2-3, 3-8-3, and 3-9-3 on the precipitation of ODN1 under different concentrations of Na+. These analogues were much less efficient than spermine in precipitating the ODN1 at 10 and 25 mM Na+ (Table 3), and when Na+ concentrations were >25 mM, ODN1 did not undergo precipitation up to 2 mM concentrations of the polyamine analogues. The low efficacy of 3-2-3 may be due to its partial protonation at pH 7.2. On the basis of the EC50 values, the order of efficacy of

Lewis et al.

spermine homologues in causing ODN precipitation is as follows: 3-4-3-3-3-3 > 3-5-3 > 3-6-3 . 3-7-3 .> 3-8-3 > 3-9-3 > 3-2-3. These results represent the structural specificity effects on the interaction of polyamines analogue with ODN1, and the differences in the ability of Na+ in displacing the analogues from their interaction sites on ODN1. We also plotted log[polyamine4+] against log[Na+] and determined the slopes of the straight lines thus generated (plots not shown). Surprisingly, the slope values of these plots were very close (1. 7 ( 0.1) (Table 4). Effects of Bis(ethyl)substitution of Spermine on ODN Precipitation. We next examined the effect of bis(ethyl)spermine (BE-SPM or BE-3-4-3) on the precipitation of both ODN1 and ODN2. BE-3-4-3 and related compounds are being currently studied as anticancer agents in several laboratories.45,46 These compounds are reported to inhibit tumor growth by depleting the intracellular levels of natural polyamines. Figure 8 (parts A and B) shows the effect of BE-3-4-3 on the precipitation of ODN1 and ODN2 at various Na+ concentrations. BE-3-4-3 was significantly less effective than spermine in precipitating ODN1 and ODN2. Furthermore, the EC50 values of BE-3-4-3 in precipitating ODN1 were 2- to 4-fold lower than that of ODN2, ranging from 0.070 to 2.3 mM for ODN1 and from 0.180 to 5.75 mM for ODN2 as the concentration of Na+ was increased from 10 to 50 mM (Table 5). The slopes of the log[BE-3-4-34+] vs log[Na+] plots were 1.46 and 2.1, respectively, for ODN1 and ODN2 (Table 4). These results show that BE-3-4-3 retains its ability to discriminate DNA in a sequence-specific manner, although it is less efficient at aggregating the ODNs. Structural Integrity of ODN1 in the Presence of Spermine. ODN1 is likely to form hairpin and cruciform structures because of the reverse palindromic nature of the ERE sequence. To examine whether the precipitation of ODN1 proceeded as a duplex DNA structure or through other forms, including the cruciform structures, we conducted polyacrylamide gel electrophoresis experiments of ODN1 incubated with different concentrations of spermine. 32Plabeled single-stranded ODN (20 000 cpm) was incubated with an equimolar amount (10 nmol of nucleotide/L) of unlabeled complementary ODN strand and different concentrations of spermine. Our results (Figure 9) showed that >90% of the labeled strand existed in the single-stranded form (lane 1), whereas >90% of the annealed, equimolar mixture of the strands existed in duplex form (lane 2). In the presence of increasing concentrations of spermine, the single-stranded form disappeared and the duplex form was stabilized. However, the intensity of the duplex band decreased at higher concentrations of spermine because of precipitation of the DNA and its adherence to the plastic tubes used for performing these experiments. Figure 10 shows the autoradiogram of our electrophoresis experiment performed under ODN1 concentrations comparable to that used in our precipitation experiments. The results are comparable to that found in Figure 9; ODN1 existed in both the duplex and single-stranded forms in the absence of spermine (lane 1), and the duplex form was stabilized by

Biomacromolecules, Vol. 1, No. 3, 2000 345

Self-Assembly of Oligonucleotides by Polyamines Table 3. EC50 Values of Spermine Analogues for Precipitating ODN1 EC50, µM [Na+],

a

mM

3-2-3

3-3-3

3-4-3

3-5-3

3-6-3

3-7-3

3-8-9

3-9-3

10 25 35 50 75

670 1950 ----

40 90 185 360 1280

45 90 170 650 1600

85 175 375 950 1850

85 370 550 2000 3550

90 450 725 2500 --

250 2150 ----

400 - -a ----

Dashes indicate that ODN precipitation was