Previous studies have shown that the room temperature photocycle of the photoactive yellow protein (PYP) from Ectothiorhodospira halophila involves at least two intermediate species: I1, which forms in <10 ns and decays with a 200-micros lifetime to I2, which itself subsequently returns to the ground state with a 140-ms time constant at pH 7 (Genick et al. 1997. Biochemistry. 36:8-14). Picosecond transient absorption spectroscopy has been used here to reveal a photophysical relaxation process (stimulated emission) and photochemical intermediates in the PYP photocycle that have not been reported previously. The first new intermediate (I0) exhibits maximum absorption at approximately 510 nm and appears in

The primary light-induced events in the photosynthetic retinal protein bacteriorhodopsin (bR) are investigated by ultrafast optical spectroscopy over the 440-1000 nm spectral range. The study compares the early dynamics of the native all-trans pigment bR 570 with those of two synthetic analogues, bR5.12 and bR5.13, in which isomerization around the critical C 13 dC 14 bond is blocked by a five-membered ring into all-trans and 13-cis configurations, respectively. Nearly identical spectral evolution is observed in both native and artificial systems over the first 100-200 fs of probe delay. During this period stimulated near-IR (∼900 nm) emission, and intense ∼460 nm absorption bands, due to analogous fluorescent I states (denoted as I 460 , I5.12 and I5.13, respectively), appear concurrently within 30 fs. In all systems continuous spectral shifting over tens of femtoseconds is observed in the 500-700 nm range. Native bR goes on to produce the J 625 absorption band within ∼1 ps, which is accompanied by disappearance of the I 460 emission and absorption features. In bR5.12 and bR5.13, aside from minor spectral modifications, the analogous dramatic changes associated with I5.12 and I5.13 are arrested beyond the first ∼100 fs, reverting uniformly to the initial ground state with exponential time constants of 19 ps and 11 ps, respectively. Analysis of the data calls for a major revision of models previously put forward for the primary events in bacteriorhodopsin. The close likeness of initial transient spectral evolution in both native and artificial pigments, despite the locking of the active isomerization coordinate in the synthetic chromophores, demonstrates that in bR 570 the ultrafast changes in transmission leading to I 460 , previously believed to involve C 13 dC 14 torsion, must be associated with other modes. The detailed comparison conducted here also identifies which of the later spectral changes in the native system requires torsional flexibility in C 13 dC 14. Similarity of 660 nm probing data in both synthetic and native chromophores demonstrates that the sub-picosecond dynamic features uncovered at this probing wavelength commonly attributed to the evolution of J 625 , are not, as previously thought, reliable measures of all-trans S 13-cis isomerization dynamics.

The vibrational spectrum (800-1700 cm -1 region) of the J-625 intermediate, formed within 200-500 fs (3.5 ps decay time to K-590) in the room-temperature bacteriorhodopsin (BR) photocycle, is measured via picosecond time-resolved coherent anti-Stokes Raman spectroscopy (PTR/CARS). An examination of the excitation conditions and BR photocycle kinetics, as well as the vibrational CARS data, convincingly demonstrates that these PTR/CARS spectra can be quantitatively analyzed in terms of primarily BR-570 and J-625 by using third-order nonlinear susceptibility (χ (3) ) relationships. The resultant background-free (Lorentzian line shapes) CARS spectrum contains 24 distinct vibrational features which provide the most complete structural characterization of J-625 yet reported. Comparisons of the J-625 vibrational spectrum with those of groundstate BR-570 and the K-590 intermediate show that J-625 maintains some structural similarities with BR-570 while it has a significantly different structure than that of K-590. Specifically, J-625 has (i) an all-trans retinal configuration, (ii) increased electron density in the CdC stretching modes as manifested by increased CdC stretching frequencies relative to those in both BR-570 and K-590, (iii) significant delocalized hydrogen out-of-plane motion not observed in any other BR species, (iv) decreased C-CH 3 in-plane wagging motion, and (v) a Schiff-base bonding environment similar to that of BR-570 and distinctively different from that in K-590. Comparisons between the PTR/CARS spectra of J-625 and T5.12, an intermediate found in the photoreaction of the artificial BR pigment, BR5.12, containing a five-membered ring spanning the C 12 -C 13 dC 14 bonds (thereby blocking C 13 dC 14 isomerization), support the conclusion that the J-625 structure reflects the reaction coordinates in the BR photocycle that precede C 13 dC 14 isomerization. Since these PTR/ CARS data show J-625 to have an all-trans retinal, C 13 dC 14 isomerization cannot be the primary reaction coordinate described in numerous models for the BR photocycle. The all-trans to 13-cis isomerization must occur as J-625 transforms into K-590, and other changes in the retinal structural and/or retinal-protein interactions must comprise the primary reaction coordinates that precede C 13 dC 14 isomerization. These results require that significant changes in the mechanistic model describing the room-temperature BR photocycle be considered.