Pyrrole (Py) and imidazole (Im) polyamides can be designed to target specific DNA sequences.The effect that the pyrrole and imidazole arrangement, plus DNA sequence, have on sequence specificityand binding affinity has been investigated using DNA melting (
TM), circular dichroism (CD), and surfaceplasmon resonance (SPR) studies. SPR results obtained from a complete set of triheterocyclic polyamidesshow a dramatic difference in the affinity of f-ImPyIm for its cognate DNA (
Keq = 1.9 × 10
8 M
-1) and f-PyPyImfor its cognate DNA (
Keq = 5.9 × 10
5 M
-1), which could not have been anticipated prior to characterizationof these compounds. Moreover, f-ImPyIm has a 10-fold greater affinity for CGCG than distamycin A hasfor its cognate, AATT. To understand this difference, the triamide dimers are divided into two structuralgroupings: central and terminal pairings. The four possible central pairings show decreasing selectivityand affinity for their respective cognate sequences: -ImPy > -PyPy-
-PyIm-
-ImIm-. Theseresults extend the language of current design motifs for polyamide sequence recognition to include theuse of "words" for recognizing two adjacent base pairs, rather than "letters" for binding to single basepairs. Thus, polyamides designed to target Watson-Crick base pairs should utilize the strength of -ImPy-and -PyPy- central pairings. The f/Im and f/Py terminal groups yielded no advantage for their respectiveC/G or T/A base pairs. The exception is with the -ImPy- central pairing, for which f/Im has a 10-foldgreater affinity for C/G than f/Py has for T/A.