A kinetic analysis of cyanine selectivity: CCyan2 and Cyan40 intercalation into poly(dA-dT) x poly(dA-dT) and poly(dG-dC) x poly(dG-dC)

A T-jump investigation of the binding of Cyan40 [3-methyl-2-(1,2,6-trimethyl-4(1H)pyridinylidenmethyl)-benzothiazolium ion] and CCyan2 [3-methyl-2-[2-methyl-3-(3-methyl-2(3H)-benzothiazolylidene)-1-propenyl]-benzothiazolium ion] with poly(dA-dT) x poly(dA-dT) and poly(dG-dC) x poly(dG-dC) is performed at I = 0.1M (NaCl), 25 degrees C and pH 7. Two kinetic effects are observed for both systems. The binding process is discussed in terms of the sequence D + P <==> P,D <==> PD(I) <==> PD(II), which leads first to fast formation of a precursor complex P,D and then to a partially intercalated complex PD(I) which converts to the fully intercalate complex PD(II). Concerning CCyan2 the rate parameters depend on the polymer nature and their analysis shows that in the case of poly(dG-dC) x poly(dG-dC) the most stable bound form is the fully intercalated complex PD(II), whereas in the case of poly(dA-dT) x poly(dA-dT) the partially intercalated complex PD(I) is the most stable species. Concerning Cyan40, the rate parameters remain unchanged on going from A-T to G-C indicating that this dye is unselective.     

Cyanine dyes as intercalating agents: kinetic and thermodynamic studies on the DNA/Cyan40 and DNA/CCyan2 systems

The interaction of cyanines with nucleic acids is accompanied by intense changes of their optical properties. Consequently these molecules find numerous applications in biology and medicine. Since no detailed information on the binding mechanism of DNA/cyanine systems is available, a T-jump investigation of the kinetics and equilibria of binding of the cyanines Cyan40 [3-methyl-2-(1,2,6-trimethyl-4(1H)pyridinylidenmethyl)-benzothiazolium ion] and CCyan2 [3-methyl-2-[2-methyl-3-(3-methyl-2(3H)-benzothiazolylidene)-1-propenyl]-benzothiazolium ion] with CT-DNA is performed at 25 degrees C, pH 7 and various ionic strengths. Bathochromic shifts of the dye absorption band upon DNA addition, polymer melting point displacement (DeltaT = 8-10 degrees C), site size determination (n = 2), and stepwise kinetics concur in suggesting that the investigated cyanines bind to CT-DNA primary by intercalation. Measurements with poly(dA-dT).poly(dA-dT) and poly(dG-dC).poly(dG-dC) reveal fair selectivity of CCyan2 toward G-C basepairs. T-jump experiments show two kinetic effects for both systems. The binding process is discussed in terms of the sequence D + S left arrow over right arrow D,S left arrow over right arrow DS(I) left arrow over right arrow DS(II), which leads first to fast formation of an external complex D,S and then to a partially intercalated complex DS(I) which, in turn, converts to DS(II), a more stable intercalate. Absorption spectra reveal that both dyes tend to self-aggregate; the kinetics of CCyan2 self-aggregation is studied by T-jump relaxation and the results are interpreted in terms of dimer formation.     

Interaction of synthetic analogues of distamycin with poly(dA-dT): role of the conjugated N-methylpyrrole system

Two synthetic analogues of distamycin (Dst), PPA and PAP, containing a saturated beta-alanine moiety substituting for an N-methylpyrrole chromophore were studied for their interactions with the double-stranded alternating copolymer poly(dA-dT).poly(dA-dt) [abbreviated as poly(dA-dT)], with UV absorption and circular dichroism (CD) spectroscopy. The distinctive feature of these analogues is the difference in the extents of extended conjugation due to contiguous pyrrole rings: it decreases in the order Dst greater than PPA greater than PAP. Both these analogues bind to poly(dA-dT) in a way similar to Dst, as suggested from the observed red shift in the UV spectra of the ligands upon complexation and the appearance of induced Cotton effects (in the 290-350-nm region) in the CD spectra of the complexes. A comparative study of (i) the spectral features of the complexes between these ligands, Dst and netrospin (Nt) and poly(dA-dT), and (ii) the binding parameters for the association with the polynucleotide suggests that the number and relative positions of the pyrrole moieties influence the spectral features and thermodynamic stabilities of the complexes, and the latter show a progressive decrease in the order Dst greater than Nt greater than PPA greater than PAP. Implications of these results vis-à-vis the molecular basis of Dst-DNA interaction are discussed.     

Influence of cyanine dye structure on self-aggregation and interaction with nucleic acids: a kinetic approach to TO and BO binding

Kinetics and equilibria of cyanine dyes thiazole orange (TO) and benzothiazole orange (BO) self-aggregation and binding to CT-DNA are investigated in aqueous solution at 25 degrees C and pH 7. Absorbance spectra and T-jump experiments reveal that BO forms J-aggregates while TO forms more stable H-aggregates. Fluorescence and absorbance titrations show that TO binds to DNA more tightly than BO. TO stacks externally to DNA for low polymer-to-dye concentration ratios (C(P)/C(D)) while dye intercalation occurs for high values of C(P)/C(D). T-jump and stopped-flow experiments performed at high C(P)/C(D) agree with reaction scheme D+S <=> D,S <=> DS(I) <=> DS(II) where the precursor complex D,S evolves to a partially intercalated complex DS(I) which converts to the more stable intercalate DS(II). Non-electrostatic forces play a major role in D,S stabilization. Last step is similar for both dyes suggesting accommodation of the common benzothiazole residue between base pairs. Experiments using poly(dA-dT).poly(dA-dT) and poly(dG-dC).poly(dG-dC) confirm base pair preference for TO.     

Binding of pyrene to DNA, base sequence specificity and its implication.

Solubilization as well as spectral studies of pyrene in natural DNA and synthetic deoxypolynucleotide solutions at neutral pH reveal at least two binding modes. Sites I are predominant in native DNA and in poly(dA-dT): poly(dA-dT) whereas sites II are found with denatured DNA and other polynucleotides such as poly(dA):poly(dT) and three different types of guanine containing copolymers which solubilize pyrene to a lesser extent. Spectral comparison with the covalent adducts of trans-7,8-dihydroxy-anti-9,10-epoxy-7,8,9,10- tetrahydro-benzo(a)pyrene (anti-BPDE) and the physical complexes of its tetraols lead to the suggestion of a base sequence specific binding model for this carcinogenic metabolite to account for the puzzling fact that although its physical binding is predominantly intercalative, the covalent adducts appear not to be intercalated. It is speculated that in neutral solutions, intercalation may have little, if any, to do with the chemical lesion of this metabolite to the guanine base of the DNA and may, on the contrary, provide an efficient pathway for detoxification.     

Ultraviolet resonance Raman spectroscopy of distamycin complexes with poly(dA)-poly(dT) and poly(dA-dT): role of H-bonding

Raman spectra are reported for distamycin, excited at 320 nm, in resonance with the first strong absorption band of the chromophore. Qualitative band assignments to pyrrole ring and amide modes are made on the basis of frequency shifts observed in D2O. When distamycin is dissolved in dimethyl sulfoxide or dimethylformamide, large (30 cm-1) upshifts are seen for the band assigned to amide I, while amides II and III shift down appreciably. Similar but smaller shifts are seen when distamycin is bound to poly(dA-dT) and poly(dA)-poly(dT). Examination of literature data for N-methylacetamide in various solvents shows that the amide I frequencies correlate well with solvent acceptor number but poorly with solvent donor number. This behavior implies that acceptor interactions with the C = O group are more important than donor interactions with the N-H group in polarizing the amide bond and stabilizing the zwitterionic resonance form. The resonance Raman spectra therefore imply that the distamycin C = O groups, despite being exposed to solvent, are less strongly H-bonded in the polynucleotide complexes than in aqueous distamycin, perhaps because of orienting influences of the nearby backbone phosphate groups. In this respect, the poly(dA-dT) and poly(dA)-poly(dT) complexes are the same, showing the same RR frequencies. Resonance Raman spectra were also obtained at 200-nm excitation, where modes of the DNA residues are enhanced. The spectra were essentially the same with and without distamycin, except for a perceptable narrowing of the adenine modes of poly(dA-dT), suggesting a reduction in conformational flexibility of the polymer upon drug binding.     

Two binding modes of netropsin are involved in the complex formation with poly(dA-dT).poly(dA-dT) and other alternating DNA duplex polymers

Using CD measurements we show that the interaction of netropsin to poly(dA-dT).poly(dA-dT) involves two binding modes at low ionic strength. The first and second binding modes are distinguished by a defined shift of the CD maximum and the presence of characteristic isodichroic points in the long wavelength range from 313 nm to 325 nm. The first binding mode is independent of ionic strength and is primarily determined by specific interaction to dA.dT base pairs. Employing a netropsin derivative and different salt conditions it is demonstrated that ionic contacts are essential for the second binding mode. Other alternating duplexes and natural DNA also exhibit more or less a second step in the interaction with netropsin observable at high ratio of ligand per nucleotide. The second binding mode is absent for poly(dA).poly(dT). The presence of a two-step binding mechanism is also demonstrated in the complex formation of poly(dA-dT).poly(dA-dT) with the distamycin analog consisting of pentamethylpyrrolecarboxamide. While the binding mode I of netropsin is identical with its localization in the minor groove, for binding mode II we consider two alternative interpretations.     

Binding of Hg(II) to poly(dA): poly(dT) and its component single strands

Mercuric binding studies at pH 10 revealed that poly(dA): poly(dT) exhibits a more dramatic absorption spectral alteration than the alternating polymer poly(dA-dT):poly(dA-dT) and induces a unique intense positive CD band at 296 nm during the spectral titrations. Comparative studies with its component single strands suggest that the spectral alterations exhibited by poly(dA): poly(dT) are consistent with a binding model in which the mercuric ions initially bind to thymines and cause the eventual strand separation of the duplex, with subsequent high cooperative binding to the poly(dA) strands. This interpretation is supported by the binding isotherms indicating much stronger mercuric binding to poly(dT) than to poly(dA), with saturation binding densities of 1 Hg(II) per 2 bases and 1 Hg(II) per base, respectively, and very high binding cooperativity for poly(dA). Striking spectral alterations are exhibited by the mercuric binding to poly(dA), likely the consequence of binding to the amino group of dA in an alkaline solution. The mononucleoside dA exhibits minor spectral alterations upon similar mercuric chloride additions whereas the dinucleoside monophosphate d(AA) exhibits significant spectral changes, albeit less pronounced than those of poly(dA). Some sequence effects on the mercuric binding are observed in the dinucleotide studies. Our CD results on the mercuric binding to polynucleotides do not support the contention of (psi)-type condensed complex formation.     

The CD of ligand-DNA systems. 2. Poly(dA-dT) B-DNA.

A systematic theoretical study of the CD of [poly(dA-dT)]2 and its complexes with achiral small molecules is presented. The CD spectra of [poly(dA-dT)]2 and of poly(dA):poly(dT) are calculated for various DNA structures using the matrix method. The calculated and experimental spectra agree reasonably well for [poly(dA-dT)]2 but less well for poly(dA):poly(dT). The calculated CD spectrum of [poly(dA-dT)]2 fails to reproduce the wavelength region of 205-245 nm of the experimental spectrum. This discrepancy can be explained by a magnetic dipole allowed transition contributing significantly to the CD spectrum in this region. The induced CD of a transition moment of a molecule bound to [poly(dA-dT)]2 is also calculated. As was the case for [poly(dG-dC)]2, the induced CD of a groove bound molecule is one order of magnitude stronger than that of an intercalated molecule. The calculations also show considerable differences between pyrimidine-purine sites and purine-pyrimidine sites. Both signs and magnitudes of the CD induced into ligands bound in the minor groove agree with experimental observations.     

Conformational transitions of a synthetic DNA poly(dA-dU). poly(dA-dU) in concentrated solutions of caesium fluoride

Chiroptical properties of poly(dA-dU).poly(dA-dU) were studied in concentrated NaCl and CsF solutions to reveal the role of the alternating B conformation in the CsF-induced alternating B-X conformational transition of poly(dA-dT).poly(dA-dT). Poly(dA-dU).poly(dA-dU) has been chosen for this purpose because it has, instead of the alternating B conformation, a regular conformation like poly(dG-dC).poly(dG-dC) in low-salt solution. It was found that poly(dA-dU).poly(dA-dU) did not assume that Z form at high NaCl concentrations but exhibited extensive CsF-induced changes in the circular dichroism spectra like poly(dA-dT).poly(dA-dT). The changes of reflect two consecutive two-state conformational transitions of the polynucleotide, both taking place with fast kinetics and low cooperativity. The transition were interpreted as involving the regular and alternating B conformation at lower CsF concentrations and the alternating B and X conformation at higher CsF concentrations. A comparison of the behaviour of poly(dA-dU).poly(dA-dU) and poly(dA-dT).poly(dA-dT) in CsF solutions demonstrates that the thymine methyl groups promote the X form but are not crucial for its existence. On the other hand, the alternating B conformation appears to be the inevitable starting structure for DNA isomerization into the X form.     

Conformation of poly(dG-dC)·Poly(dG-dC) and Poly(dA-dT)·Poly(dA-dT) interacting with the antitumor drug cis-[Pt(NH32]Cl2

The interaction of cis-dichlorodiammine platinum(II) with poly(dG-dC)·poly(dG-dC) and poly(dA-dT) ·poly(dA-dT) was studied by circular dichroism. Significant conformational changes were induced in both alternating polymers: in the case of poly(dG-dC) ·poly(dG-dC) the spectra were not conclusive in terms of a well defined conformation, even if the presence of left-handed helices could be suggested. For poly(dA-dT)·poly(dA-dT) the data were interpreted in terms of a dimer-helix → single hairpin helix transition induced by the metal. The results obtained are discussed with reference to the antitumor activity of the drug.     

Interaction between double stranded poly(dA-dT)·poly (dA-dT) and lucanthone

270 MHz ¹H NMR and theoretical studies indicate that the drug lucanthone forms intercalated complexes with the synthetic DNA poly(dA-dT)·poly(dA-dT). In the intercalated complex the long axis of the drug is perpendicular to the helix axis and parallel to the base pair axis, i.e., the long axis is perpendicular to the dyad axis.     

Two Binding Modes of Netropsin are Involved in the Complex Formation with Poly(dA-dT)·Poly(dA-dT) and other Alternating DNA Duplex Polymers

Abstract

Using CD measurements we show that the interaction of netropsin to poly(dA-dT)·poly(dA-dT) involves two binding modes at low ionic strength. The first and second binding modes are distinguished by a defined shift of the CD maximum and the presence of characteristic isodichroic points in the long wavelength range from 313 nm to 325 nm. The first binding mode is independent of ionic strength and is primarily determined by specific interaction to dA·dT base pairs. Employing a netropsin derivative and different salt conditions it is demonstrated that ionic contacts are essential for the second binding mode. Other alternating duplexes and natural DNA also exhibit more or less a second step in the interaction with netropsin observable at high ratio of ligand per nucleotide. The second binding mode is absent for poly(dA)·poly(dT). The presence of a two-step binding mechanism is also demonstrated in the complex formation of poly(dA-dT)·poly(dA-dT) with the distamycin analog consisting of pentamethylpyrrolecarboxamide. While the binding mode I of netropsin is identical with its localization in the minor groove, for binding mode II we consider two alternative interpretations.     

Binding of mithramycin to DNA in the presence of second drugs

Comparative DNA equilibrium binding studies with mithramycin (MTR) and ethidium bromide in the presence and in the absence of second drugs were investigated by spectral titrations. Unusual curvatures (in contrast to those due to neighbor exclusion or anticooperativity) are found in the Scatchard plots of MTR-DNA titrations in the presence of netropsin, a minor-groove binder. Parallel studies with ethidium bromide indicate that although the presence of netropsin significantly reduces the binding ability of ethidium, no unusually curved Scatchard plots are obtained. The unusual curvature exhibited by the Scatchard plots of MTR titrations in the presence of netropsin indicates that the binding of netropsin greatly affects the MTR binding to DNA and can be simulated by an explicit incorporation of the second drug-DNA interaction in the binding formalism. Since netropsin is a minor-groove binder, its interference with the binding of MTR is in accord with the notion that MTR also binds at this groove. The observation of negligible effects on the DNA binding ability of MTR in the presence of either a major-groove or a phosphate group binder lends further support to this conclusion. Consistent with its guanine specificity, studies with synthetic polynucleotides suggest that MTR exhibits negligible affinity for poly(dA-dT).poly(dA-dT) or poly(dA).poly(dT).(ABSTRACT TRUNCATED AT 250 WORDS)     

Binding of nonintercalative antitumor drugs to DNA-polymers: structural effects of bisquaternary ammonium heterocycles

The binding of the antitumor agents SN-16814 nd SN-13232 to various DNA's in solution was monitored by CD and UV absorption measurements. In addition comparative studies with dA.dT containing duplex DNA of the related ligands SN-6136 and SN-6324 were included with respect to effects of structural variations. In general all four ligands show a dA.dT preference in their binding affinity to DNA. Differences were observed for the reaction of SN-16814 which contains bicyclic ring system: it has a lower base pair selectivity, shows some affinity to poly(dG-dC).poly(dG-dC), poly(rA).poly(rU) and poly(rU). The binding mechanism of SN-16814 is associated with a significant time dependent binding effect in CD spectra and UV absorption in case of reaction with poly(dA).poly(dT) and poly(dI).poly(dC) indicating a slow kinetics. The preferred binding to dA.dT base pairs in DNA decreases in the order from SN-61367 greater than SN-13232 greater than SN-6324,SN-16814 as judged from CD titration studies, salt dissociation and melting temperature data. Competitive binding experiments with netropsin (Nt) or distamycin-5 revealed that SN-16814 and SN-13232 are displaced from poly(dA.dT).poly(dA-dT) suggesting that both ligands are less strongly bound than Nt and Dst-5 within the minor groove of B-DNA. These studies are consistent with results of the DNAse I cleavage of poly(dA-dT).poly(dA-dT) which show the same relative order of inhibition of the cleavage reaction due to ligand binding. The results suggest that the variability of the DNA binding and dA.dT sequence specificity may reside in the adaptability of benzamide-type ligands in the helical groove which is influenced by distinct structural modifications of the ligand conformation.     

Interactions of water-soluble metalloporphyrins with nucleic acids studied by resonance Raman spectroscopy.

The resonance Raman spectra of water-soluble porphyrins, M(TMpy-P4) (M = Cu(II), Ni(II) and Co(III] and their mixtures with poly(dG-dC)2, poly(dA-dT)2 and calf thymus and salmon DNAs were measured using a divided rotating cell to determine the magnitudes of frequency shift and intensity variation resulting from M(TMpy-P4)-nucleic acid interactions. Bands II(C beta-H bending, approximately 1100 cm-1) and VIII(C beta-C beta stretch, approximately 1570 cm-1) show a large and small upward shift, respectively, when Cu(TMpy-P4) and Ni(TMpy-P4) are intercalated at the G-C sites. In contrast, these bands show a small upward and downward shift, respectively, when Co(TMpy-P4) is groove-bound at the A-T sites of nucleic acids. Both Bands V (approximately 1260 cm-1) and IX (approximately 1646 cm-1) which originate in the N-methylpyridyl group always show small downward shifts due to coulombic interaction between the N-CH3+ group of TMpy-P4 and the PO2 group of the nucleic acid.     

Circular dichroism of polynucleotides: Interactions of NiCl2 with poly(dA-dT).poly(dA-dT) and poly(dG-dC).poly(dG-dC) in a water-in-oil microemulsion

The thermal behavior of the synthetic, high molecular weight, double stranded polynucleotides poly(dA-dT).poly(dA-dT) [polyAT] and poly(dG-dC).poly(dG-dC) [polyGC] solubilized in the aqueous core of the quaternary water-in-oil cationic microemulsion CTAB|n-pentanol|n-hexane|water in the presence of increasing amounts of NiCl(2) at several constant ionic strength values (NaCl) has been studied by means of circular dichroism and electronic absorption spectroscopies. In the microemulsive medium, both polynucleotides show temperature-induced modifications that markedly vary with both Ni(II) concentration and ionic strength. An increase of temperature causes denaturation of the polyAT duplex at low nickel concentrations, while more complex CD spectral modifications are observed at higher nickel concentrations and ionic strengths. By contrast, thermal denaturation is never observed for polyGC. At low Ni(II) concentrations, the increase of temperature induces conformational transitions from B-DNA to Z-DNA form, or, more precisely, to left-handed helical structures. In some cases, at higher nickel concentrations, the CD spectra suggest the presence of Z'-type forms of the polynucleotide.     

Solution Structure of Poly(dA-dT)· Poly(dA-dT) in Low and High Salt: A 500 MHz 1H NMR Study Using One-Dimensional NOE

Abstract

CD spectra of poly(dA-dT)· poly(dA-dT) in low salt (10–100 mM NaCl) and high salt (4–6 M CsF) are different i.e. 275 nm band gets inverted in going from low to high salt (Vorhickova et. al.MarJ. Mol. Biol. 166, 85, 1983). However, from CD spectra alone it is not possible to decipher any structural differences that might exist between the low and high salt forms of poly(dA-dT)? poly(dA-dT). Hence, we took recourse to high resolution NMR spectroscopy to understand the structural properties of poly(dA-dT)? poly(dA-dT) in low and high salt. A detailed analysis of shielding constants and extensive use of NOE studies under minimum spin diffusion conditions using C(8)-deuterated poly(dA-dT)? poly(dA-dT) enabled us to come up with the following conclusions (i) base-pairing is Watson-Crick under low and high salt conditions, (ii) under both the conditions of salt the experimental data can be explained in terms of an equilibrium blend of right and left-handed B-DNA duplexes with the left-handed form 70% and the right-handed 30%. In a 400 base pairs long poly(dA-dT)? polyidA-dT) (as used in this study), equilibrium between right and left-handed helices can also mean the existence of both helical domains in the same molecule with fast interchange between these domains or/and unhindered motion/propagation of these domains along the helix axis, (iii) However, there are other structural differences between the low and high salt forms of poly(dA-dT) ? poly(dA-dT); under the low salt condition, right-and left-handed B-DNA duplexes have mononucleotide as a structural repeat while under the high salt conditions, right-and left-handed B-DNA duplexes have dinucleotide as a structural repeat. In the text we provide the listing of torsion angles for the low and high salt structural forms, (iv) Salt (CsF) induced structural transition in poly(dA-dT)? poly(dA-dT) occurs without any breakage of Watson- Crick pairing, (v) The high salt form of poly(dA-dT)? poly(dA-dT) is not the left-handed Z-helix.

Although the results above from NMR data are quite unambiguous, a question still remains i.e. what does the salt (CsF) induced change in the CD spectra of poly(dA-dT)? poly(dA-dT) really indicate? Interestingly, we could show that the salt (CsF) induced change in poly(dA-dT)? poly(dA-dT) is quite similar to that caused by a basic polypeptide viz. poly-L(Lys₂-Ala)_n i.e. both the agents induced a ψ-structure in DNA. And it was also demonstrated that the changes in poly(dA-dT)? poly(dA-dT) as caused by CsF and poly-L-(Lys₂-Ala)_n could be reverted back by ethidium bromide-a relaxing agent.

To minimize complications from spin diffusion in this study we have used very small presaturation pulse lengths and C(8)-deuterated poly(dA-dT)? poly(dA-dT) of 400 ± 150 bp long. Even though deuteration of a primary site of diffusion such as C(8)H substantially decreases diffusion, in order to make sure that our conclusions are not compromised by possible diffusion in such a long fragment under small presaturation times, we have repeated our experiments using the six base pair long duplex of d(A-T-A-T-A-T) and found the results to be strikingly similar to that from the polymer.     

Interaction of water-soluble Cu(II), Ni(II), and Co(III) porphyrins with polynucleotides

The interaction of the Cu(II), Ni(II) and Co(III) complexes of the following six water-soluble cationic porphyrins with calf thymus DNA, poly(dG-dC)2 and poly(dA-dT)2 was studied by UV-visible and resonance Raman spectroscopy: tetrakis(2-N-) and (3-N-methylpyridyl) porphyrin (1, 2); monophenyl-tris(4-N-methylpyridyl)porphyrin (4); cis- and trans-diphenyl-bis (4-N-methylpyridyl)porphyrin (5, 6). The binding to nucleic acids was compared with that of tetrakis(4-N-methylpyridyl)porphyrin (3). If the N(+)-CH3 group is moved from the para (3) to the meta position (2), binding of the free porphyrin as well as that of the metal complexes is only gradually modified; thus, the square-planar Cu- and Ni-2 are intercalated at the G-C site whereas Co-2 is groove-bound at A-T. Additionally, Ni-2 is probably also intercalated at the A-T site. When the N(+)-CH3 group is located at ortho position (1), the high rotation barrier of the 2-N-methylpyridyl group prevents intercalation of Cu- and Ni-1, resulting in weak outside binding. At ionic strength mu = 0.2, there is no evidence of significant interaction of Co-1 with any of the polynucleotides. When the charged N-methylpyridyl groups in 3 are subsequently replaced by phenyl groups (4, 5/6), the tendency of the Cu(II) and Ni(II) complexes to bind to the outside of the helix or to intercalate only partially increases at the expense of full intercalation. The coulombic attraction remains strong, no significant differences can be detected between 3, 4, 5, and 6. Ni-4 binds to poly(dA-dT)2 in the same complicated manner as Ni-3. The outside-binding in Co-4, -5 and -6 differs slightly from that in Co-2 and Co-3.