H-1-NMR studies, of Rh(NH3)(4)phi(3+) (phi = 9,10-phenanthrenequinone diimine) bound to d(TGGCCA)(2) indicate that the octahedral complex binds to the duplex in a manner consistent with classical intercalation. Upfield shifts of the phi protons of Rh(NH3)(4)phi(3+) upon binding to the oligonucleotide are observed with the greatest shift seen for the (2,7) protons followed by (1,8) > (3,6) similar to (4,5). From the chemical shift differences of the inequivalent phi (2,7) proton resonances (determined at 285 K) the exchange rate is estimated to be approximately 250 s(-1) at the coalescence temperature of 305 K and similar to 10 s(-1) at 285 K. The NOESY of the 1:1 Rh(NH3)(4)phi(3+)-d(TGGCCA)(2) complex reveals a complete absence of an NOE cross peak between the C(4)H6 and the G(3)H2’/H2" protons, and instead NOE cross peaks between phi protons and protons from the G(3) and C-4 residues are observed. The one- and two-dimensional NMR data art : consistent with the rhodium complex intercalated into the G(3)C(4) site from the major groove. Molecular modeling based on these data leads to a structure of Rh(NH3)(4)phi(3+) intercalated from the major groove with the phi ligand fully inserted and stacked in the column of base pairs. In this model the axial ammines of Rh(NH3)(4)phi(3+) are well positioned for hydrogen bonding to the guanine O6 atoms. A comparison of this NMR study to those with Delta-Ru(phen)(3)(2+) and Delta-Rh(phen)(2)phi(3+) is made; the tetraammine complex appears to be intercalated more deeply in the intercalation pocket. In general, based upon these data, the DNA binding mode of Rh(NH3)(4)phi(3+) may be considered as classical intercalation. Octahedral metallointercalators, however, may be considered as nonclassical; their sequence-selectivity is derived from the nonintercalating functionalities of the synthetic, octahedral complex. These structural studies provide the basis for the design of sequence-specific DNA-binding complexes in the major groove.