The antitumor agent amsacrine, 4'-(9-acridinylamino)methanesulfon-m-anisidide (m-AMSA), when bound to double-stranded DNA, particularly poly(deoxyadenylicthymidylic acid), reduced the fluorescence of bound ethidium without physically displacing it from DNA. Fluorescence lifetime measurements showed that the reduction of fluorescence was not due to reduction of the lifetime of the excited state of ethidium. Rather, a proportion of the DNA-bound ethidium changed to a state where the fluorescence was highly quenched. Several other 9-anilinoacridine derivatives, and also 9-hydroxyellipticine, caused quenching of ethidium-DNA fluorescence, whereas 9-aminoacridine, proflavin, and ellipticine had no effect. Resonance energy transfer (Förster transfer) is not responsible for the effect since there is no spectral overlap between the absorption spectrum of any of the agents and the fluorescence emission spectrum of ethidium. It is suggested that quenching may occur as a result of reversible formation of electron-transfer complexes between the intercalating drug and the excited state of ethidium.

SynopsisThe distribution of counterions around a charged polyion cylinder is calculated by several methods. First, the Debye-Huckel approximation is used, and it is shown that Manning's condensation hypothesis is necessry to avoid overneutralization of the polyion charges by the counterions when the linear-charge-density parameter, t, of the polyion exceeds the critical value of unity. However, it appears that this method of getting this result involves inconsistent application of Debye-Huckel theory. Therefore, we turn to the analytical solution of the Poisson-Boltzmann equation that was obtained by Alfrey, Berg, and Morawetz for a polyion cylinder plus a neutralizing number of counterions but without added salt. One of the integration constants of this solution is a radius, which we call Rw, within which lies precisely the fraction of counterions that Manning assumes to condense in his theory. This radius can be rather large, however, so that the "Manning fraction" of condensed ions actually forms a diffuse cloud whose size varies with the polyelectrolyte concentration; Rw varies as K -I '~, where K is the Debye-Huckel screening parameter. The Manning fraction, 1 -l/[, and its associated radius are unique in their behavior with dilution; smaller fractions stay within finite radii, while with larger fractions the corresponding radii increase as K -~. Thus, the condensation hypothesis does have a simple mathematical foundation in the Poisson-Boltzmann equation. Finally, by comparison with numerical solutions, we find that these conclusions are not significantly changed even when salt is added to the polyelectrolyte. A short table of numerical solutions of the Poisson-Boltzmann equation in cylindrical geometry is given, together with tables of coefficients tht enable one to discover the particular solution that applies for a given polyion radius and charge density.

As a first step in the synthesis and the study of DNA polyintercalating drugs, dimers of acridines have been prepared. Their DNA binding properties have been studied. It has been determined that wen the chain separating the two aromatic rings is longer than a critical distance, bisintercalation is actually observed and that the DNA binding affinity becomes quite large (>108-109 M-1). It is shown also that the optical characteristics of these molecules are dependent on the sequences of DNA. The fluorescence intensity of one of these dimers when bound to DNA varies as the fourth power of its A+T content. This derivative could be used as a fluorescent probe of DNA sequence.

SynopsisWe report a calculation of the distribution of small ions around a charged cylinder representing a polyelectrolyte molecule in solution. The Monte Carlo method of Metropolis, Rosenbluth, and Teller was used to avoid the inaccuracies known to be associated with the Poisson-Boltzmann equation. The systems examined contained a long polyelectrolyte cylinder with charge parameter, x , equal to 4.2, corresponding approximately to a DNA molecule.In one model, the cylinder had charges on its axis and an exclusion radius to the center of the small ions equal to 10 A, while the small ions had various radii in the range from 1 to 10 8, and one or two protonic charges. Various systems were studied; some had one species of small ion alone, others had mixtures of different types. The results showed good agreement with the solution of the Poisson-Boltzmann equation when only the species with 1-8, radius was present, but considerable discrepancies appeared with larger ions as a result of excludedvolume interactions between the latter. Deviations from the Poisson-Boltzmann equation also appeared when both positive and negative small ions were present; the deviations were in the direction of a higher concentration of both counter-and co-ions, but particularly co-ions, close to the polyelectrolyte. In another model, the charges were arranged along two helices on the surface of the cylinder; the resulting radial distribution of small ions was not much different from that found when the charges were situated on the axis. In all cases there was a striking accumulation of counterions in a layer of concentration exceeding 1 molL a t the surface of the polyion.

Our prior studies showed that polyhydroxylated styrylquinolines are potent HIV-1 integrase (IN) inhibitors that block the replication of HIV-1 in cell culture at nontoxic concentrations. To explore the mechanism of action of these inhibitors, various novel styrylquinoline derivatives were synthesized and tested against HIV-1 IN and in cell-based assays. Regarding the in vitro experiments, the structural requirements for biological activity are a carboxyl group at C-7, a hydroxyl group at C-8 in the quinoline subunit, and an ancillary phenyl ring. However the in vitro inhibitory profile tolerates deep alterations of this ring, e.g. by the introduction of various substituents or its replacement by heteroatomic nuclei. Regarding the ex vivo assays, the structural requirements for activity are more stringent than for in vitro inhibition. Thus, in addition to an o-hydroxy acid group in the quinoline, the presence of one ortho pair of substituents at C-3' and C-4', particularly two hydroxyl groups, in the ancillary phenyl ring is imperatively required for inhibitory potency. Starting from literature data and the SARs developed in this work, a putative binding mode of styrylquinoline inhibitors to HIV-1 IN was derived.

Condensation of the counter-ions around a highly charged infinitely long cylindrical molecule, such as DNA, can be described in detail in terms of the solutions of the Poisson-Boltzmann (Gouy-Chapman) equation. By using the Alfrey-Berg-Morawetz (1951) solution of this equation one can show that a certain fraction of the counter-ions remain within finite distances of the poly-ion even when the volume of the system is expanded indefinitely; these ions can be appropriately called "condensed". The fraction of the macromolecule's charge represented by these ions is just 1-1/xi, where xi is the linear charge-density parameter of the macromolecule; this is also the value given by Manning's theory. The question arises: Is this property unique to the infinite cylinder? Using the same PB equation, we can consider the infinite charged plane and a large finite charged sphere for comparison. In the case of the plane all of the counter-ions are condensed in the above sense, not just a fraction, for any surface charge density of the plane. These ions form the classical Gouy double layer. On the other hand, none of the counter-ions of the charged sphere are condensed in the above sense, no matter how high the surface charge density. Thus the cylinder is a unique intermediate case in which a fraction of the counter-ions are condensed if the linear charge density is higher than the critical value of unity.

An ethidium homodimer and acridine ethidium heterodimer have been synthesized (Gaugain, B., Barbet, J., Oberlin, R., Roques, B. P., & Le Pecq, J. B. (1978) Biochemistry 17 (preceding paper in this issue)). The binding of these molecules to DNA has been studied. We show that these dimers intercalate only one of their chromophores in DNA. At high salt concentration (Na+ greater than 1 M) only a single type of DNA-binding site exists. Binding affinity constants can then be measured directly using the Mc Ghee& Von Hippel treatment (Mc Ghee, J. D., & Von Hippel, P. H. (1974) J. Mol. Biol. 86, 469). In these conditions the dimers cover four base pairs when bound to DNA. Binding affinities have been deduced from competition experiments in 0.2 M Na+ and are in agreement with the extrapolated values determined from direct DNA-binding measurements at high ionic strength. As expected, the intrinsic binding constant of these dimers is considerably larger than the affinity of the monomer (ethidium dimer K = 2 X 10(8) M-1; ethidium bromide K = 1.5 X 10(5) M-1 in 0.2 M Na+). The fluorescence properties of these molecules have also been studied. The efficiency of the energy transfer from the acridine to the phenanthridinium chromophore, in the acridine ethidium heterodimer when bound to DNA, depends on the square of the AT base pair content. The large increase of fluorescence on binding to DNA combined with a high affinity constant for nucleic acid fluorescent probes. In particular, such molecules can be used in competition experiments to determine the DNA binding constant of ligands of high binding affinity such as bifunctional intercalators.

The electrostatic contribution to the persistence length is computed through the integration of the average electrostatic potential at the surface of a toroidal polyion over its charge parameter, when specific interaction between the mobile ions of the salt and the fixed charges of the polyion are ignored. When the radius of curvature of the polyion is very large, we assumed that the electrostatic potential may be developed in series as a function of the curvature. With this assumption, it may be computed from one-dimensional integrations only. The assumption is then numerically verified by performing a painstaking two-dimensional integration for finite values of the radius of curvature. Values for the electrostatic contribution to the persistence length are given in two limit cases: in the presence of an excess of added salt and in the total absence of added salt. The role of the dielectric constant and of the mobility of the ’’fixed’’ charges of the polyion is also determined.