A direct real-time spectroscopic investigation of the mechanism of open complex formation by T7 RNA polymerase.

Initiation of transcription occurs through a series of steps starting with the binding of RNA polymerase to a promoter DNA and formation of a closed complex. The closed complexes, then isomerize to open complexes. In the open complexes a portion of the promoter DNA is unwound. Using fluorescence spectroscopy, we have investigated in real-time the mechanism of unwinding of promoter DNA during the transition from closed to open complexes of T7 RNA polymerase. We synthesized DNA templates containing the fluorescent base analog 2-aminopurine in place of adenine at specific positions in a T7 RNA polymerase promoter. We located the 2-aminopurine residues in the presumed melting domain of the promoter at -1, -4, and at -6. The fluorescence of 2-aminopurine increases when the DNA goes from a double-stranded form to a single-stranded form. By spectroscopically monitoring the increase in fluorescence of 2-aminopurine in DNA-T7 RNA polymerase complexes, we obtained kinetic and thermodynamic information for DNA unwinding. In the presence of the initiating nucleotide GTP, conformational transitions in the polymerase-promoter complex leading to strand opening were slower than in its absence. The rate of base pair disruption at -1, -6, and at -4 was also slower in the presence of GTP than in its absence. At 37 degrees C, base pair disruption occurred first at -1 followed by -6 and finally at -4. Open complex formation was temperature-sensitive. Temperature effects at -1, -6, and at -4 were consistent with this order of base pair disruption. The apparent activation energies (Ea) for base pair disruption around -1 and -6 were 14 kcal mol-1 and 50 kcal mol-1, respectively, also suggesting this order of base pair disruption. Transcription initiation assays using G-ladder synthesis revealed that initiation rates were almost the same on all three templates containing the modified base. Unlike strand opening, we did not observe lag times for G-ladder synthesis. We suggest that facile base pair disruption at -1 is sufficient for transcription initiation. Based on these data, it is proposed that the polymerase makes contacts at or near -1 and -6 resulting in untwisting of these base pairs thus creating at least two base pair disruption events at -1 and at -6, which are followed by bidirectional propagation to -4.