Experimental precedent for the need to involve the primary hydration layer of DNA in lead drug design

The increase in fluorescence on binding of m-phenyl substituted hydroxy derivatives of Hoechst 33258 with poly[d(A-T)], d(CGCGAATTCGCG)(2), and with the corresponding T-4-looped 28-mer AATT hairpin was used to monitor binding by equilibrium titrations and stopped-flow kinetics. Replacing the p-OH substituent of Hoechst 33258 (association constant K-a = 5.2 x 10(8) M-1 for 28-mer hairpin) by m-OH increases the AATT site binding energy by 1.1 kcal mol(-1), K-a = 3.8 x 10(9) M-1. Addition of a second m-hydroxy group (bis-m-OH Hoechst) further strengthens binding, giving K-a = 1.9 x 10(10) M-1, and the binding energy increases by about 2.1 kcal mol(-1) compared to p-OH Hoechst. The value of K-a determined at equilibrium equaled that determined from the ratio of association and dissociation rate constants from stopped-flow studies. The increase in affinity of the monohydroxy Hoechst analogue (m-OH) may originate from water-mediated hydrogen bonding with the minor groove. The further increase in affinity of the bis-m-OH derivative (whose second m-OH group must be directed away from the DNA minor groove floor) may arise from a hydrogen-bonded network of water molecules. The potential to increase binding strength through relayed water molecules is proposed as an additional possible input for lead drug design at DNA targets.