Book of Abstracts: Albany 2003
June 17-21 2003
On the Cooperativity of Metal Ion Binding to DNA
DNA involving biological processes are governed by cooperative binding of various molecules (ligands). The cooperativity is a result of different types of long-range and short-range interactions between ligands bound to DNA. There are the three main types of interactions between bound ligands: i) Contact interactions between ligands that occupy adjacent base pairs, ii) long-range interactions between neighboring ligands located at a distance of up to several tens of base pairs, iii) infinite-range interactions between all the ligands bound to a DNA molecule. The striking difference between these types of interaction is that only the last type of interactions is able to give rise to adsorption of ligands on DNA with the character of phase transition (1). In the present work we discuss the possibility of distinguishing between short-range and long-range interactions on the basis of experimental titration curves. Such an analysis is carried out then for divalent metal ion binding to double stranded and single stranded DNA.
There are two main approaches for graphical analysis of binding of ligands to DNA. The first one is the Scatchard plot (ν/[L] against [L], where ν is degree of binding of ligands to DNA, [L] is concentration of free ligands in solution); the second one is the Hill plot (Log[Y/(1-Y)] against Log[L], where Y is the fractional saturation). These presentations have been successfully used to distinguish between homogenous vs. inhomogeneous, and cooperative vs. non-cooperative or anti-cooperative binding. However, these two plots, as well as standard ν against [L] plot can not help to distinguish between long-range vs. contact cooperativity.
Such an analysis may be carried out using plot of logarithm of apparent binding constant, Kapp, against ν. In our recent paper (2) we have described a method for calculation of Kapp(ν) dependence for short- and long-range models of interaction between bound ligands. Calculations show, that the behavior of Log(Kapp(ν)) changes dramatically for different models. On the other hand, Kapp may be extracted directly from experimental binding data without assuming any definite type of cooperativity. Comparison of theoretical curves calculated according to a given model with experimental one allows us to choose between different cooperativity models. In particular, analysis of data in literature on DNA-divalent metal binding reveals the following:
1) Binding of divalent metal ions to double-stranded DNA is anticooperative. This binding is described by infinite-range model in which all ligands bound to DNA influence each other by decrease in DNA charge density.
2) Binding of Cu2+ to single-stranded polynucleotides (poly(C)), as well as binding of Mg2+ and Ca2+ to bases of single-stranded genomic DNA may be described by the McGhee and von Hippel model of contact cooperativity.
Contact cooperativity may be explained by metal ion cross-linking distant DNA segments. It is more favorable to form a new cross-link close to already existing one in order to prevent formation of a new DNA loop, which it entropically unfavorable. Therefore, cross-links tend to form blocks as in the case of ordinary contact cooperativity of ligand binding. The idea that metal ion cross-linking may cause contact cooperativity has been in the literature for some time. However, to the best of our knowledge this is the first direct proof that it may be described in the frame of the McGhee and von Hippel contact interaction model.
This work was supported by INTAS (YSF2002-141) and BFFI (B02M-091).
Institute of Bioorganic Chemistry