A-T base sequences: a theoretical and an experimental study
Abstract
As a result of the study undertaken, besides predicting the relative stabilities of polynucleotide duplexes containing A–T base pairs, the following conclusions can be drawn (Rajagopalan et al., 1984):
a) Base–base interactions depend on both helical parameters and base parameters.
b) Hydrogen bonding plays an important role not only in determining stacking patterns but also in defining the pairing scheme.
c) Particularly for the D form, the A–T–A–T sequence prefers a right handed structure, whereas the T–A–T–A sequence prefers a left handed structure.
These energy calculations highlight the role of base sequence, orientation, and disposition of bases, enabling theoretical predictions of energetically stable structures.
Chapter 4 will deal in detail with the structure of the synthetic polynucleotide poly d(A–T) in the D form. The relevance of the energy calculation results to its structure-compared with available X ray data-will be discussed.
The results of physicochemical studies on both the impure and pure trimer and the impure dimer have suggested certain criteria important for the interaction of such compounds with DNA. It must be emphasized that purity is essential for quantitatively determining the extent of interaction with various DNAs and for effective comparison with distamycin (Dst). Since the dimer could not be purified, only qualitative studies of its interactions were carried out, and no quantitative data are reported. Work on the dimer-aimed at identifying the reasons for its resistance to purification and possible solutions-is in progress.
For the trimer–DNA interaction, only a qualitative analysis was conducted, supporting the idea that compounds resembling A–T specific probes can be synthesized.
However, quantitative studies on the interaction of certain trimer analogues of Dst are currently underway in this laboratory (Dasgupta et al., 1987). These compounds resemble Dst, except that either the second or third consecutive methylpyrrole ring is replaced by alanine. The effect of this modification-and the resulting curvature-on binding to natural and synthetic DNAs, in comparison with Dst, has been determined.
A few additional analogues of Dst have been synthesized (Hao & Ekambareshwar, 1987) to better understand the role of curvature in Dst analogue–DNA interactions. In these analogues, the N methylpyrrole rings of Dst or netropsin (Nt) have been replaced by meta or para aminobenzoic acid units. These analogues:
show specificity for B DNA only,
exhibit no affinity for A or Z DNA,
display reduced A–T selectivity compared with Dst, and
bind via the minor groove.
Preliminary antiviral and antitumor studies indicate that these analogues are active against bovine herpes simplex virus and BUK 2 cells.
The antibacterial activity of both the dimer and trimer showed interesting results (as mentioned in Chapter 8), even though these preliminary results do not yet establish the comparative strength or specificity of binding of these analogues relative to Dst. Antiviral and antitumor studies on these Dst analogues are ongoing.
Thus, it can be concluded that multiple factors contribute to ligand–DNA interactions, influencing both base specificity and binding strength. These factors include:
electrostatic attraction,
hydrogen bonding,
van der Waals radii,
curvature, and
the chemical structure of the ligand.
The design of A–T specific probes using conformational principles is therefore of significant importance.
It has been shown that substituting consecutive pyrrole rings of Dst with alanine reduces A–T specificity (Dasgupta et al., 1987), relative to native Dst. A similar loss of A–T specificity was observed in trimer–DNA interactions in the present study. Further investigations are needed to determine whether this reduced specificity arises from chemical substitution, changes in curvature, or a combination of both. These studies are being actively pursued in this laboratory.

