Search

In this report, we illustrate through the study of two allosteric heme proteins the contribution of time-resolved absorption spectroscopy to the understanding of fundamental biological mechanisms. The first studied protein is the endogenous nitric oxide receptor (guanylate cyclase, sGC) whose activation and deactivation mechanisms are not yet fully resolved. We show that the rebinding of the proximal histidine occurs in ~100 picoseconds in sGC, which is the very first step of its deactivation following NO release. We also show that synergistic action of CO together with an allosteric activator induces the cleavage of the bond between heme iron and proximal histidine. The second one is the prototype of allosteric protein, the dioxygen transporter hemoglobin (Hb). In Hb, we show that the motion of the iron atom, central to the heme, moves in ~18 picoseconds after NO binding; this motion represents the very first step of the allosteric T → R transition.

Transient absorption spectra on cytochrome c’ and their kinetics were recorded to identify the formation of 5-coordinate (5c)-NO and 5c-His hemes from 4c-heme (99% and 1% amplitudes; 7-ps and 100-ps time constants, respectively). We demonstrate that proximal histidine precludes NO rebinding at the proximal site.

We present here how ultrafast time-resolved spectroscopy improves our understanding of a new class of proteins: Nitric Oxide sensors. Nitric oxide (NO) is a small, short-lived, and highly reactive gaseous molecule and it acts as a second messenger in several physiological systems. NO sensors are proteins which bind NO and are able to translate this binding into a signal for mammal cells as well as in bacteria. We have studied NO-sensors with the goal of understanding the activation and deactivation mechanism of the human NO-receptor, the enzyme guanylate cyclase (sGC), which is involved in communication between cells. Some bacterial sensors of NO (SONO) have structural homologies and common properties with sGC, but also have differences with sGC which make them valuable system to get structural and physiological information on sGC. To understand how NO-sensors interact with NO and control its reactivity, it is essential to probe dynamics and interactions when NO is present within protein core and what are the associated structural changes. For this purpose, we have used time-resolved absorption spectroscopy in the picoseconds (10− 12s) time domain. NO can be photodissociated from heme by the pulse of femtosecond laser. Time-resolved transient absorption spectra on NO-sensors were recorded and NO-protein interacttion were recorded. In case of cytochrome c’, we identified the formation of 5-coordinate (5c)-NO and 5c-His hemes from 4c-heme and demonstrate that proximal histidine precludes NO rebinding at the proximal site. In bacteria, the adaptation of SONO to temperature changes was not achieved by a simple temperature-dependent NO binding equilibrium, but by a change of the proportion between 5c-NO and 6c-NO species. This amplifies the response to temperature changes since a fast NO rebinding is the only property of a 5c-NO leading to 4c-heme after dissociation. Our results of NO dynamics provide a model for the regulation at molecular level in NO-sensing function.

The primary events in bacteriorhodopsin are investigated by coherent infrared emission spectroscopy of oriented purple membranes. Long-lived vibrational motions involving charge displacements are observed following sudden (

International audience; We used a femtosecond visible-infrared pump-probe experiment to follow ligand dissociation in carbonmonoxy-myoglobin. The CO mid-infrared absorption was found to be significantly delayed with respect to the photodissiciation. Assuming a gradual vibrational frequency change, the mid-infrared differential transmission can be interpreted in terms of a 1.2 ps ligand transfer time from the heme to the docking site.

International audience; We have measured coherent infrared emission from myoglobin crystals Jiganded with CO.

The mechanism of primary charge separation in bacterial reaction centers is still subject to debate [1]. Amongst other questions, the functional role of coherent motions from non-thermalized vibrational states has to be evaluated. The potential energy surface of the excited donor state P* is strongly displaced with respect to the ground state for a number of low-frequency vibrational modes. This has been extensively documented by photochemical holeburning [2, 3, 4] and resonance Raman [5,6] studies. Using femtosecond spectroscopy to monitor oscillations in the stimulated emission band shape, we have previously shown that phase correlations for those modes are maintained on the timescale of electron transfer [7,8]. In the usual assumption that vibrational dephasing precedes or at most is concomitant with the processes of energy exchange, these observations imply that the entire vibrational manifold of P* is not thermalized on the timescale of ET.

Visible or near-infrared (hereafter called simply visible) femtosecond pulses make possible the generation of far-infrared (THz) [1] and even mid-infrared [2] femtosecond pulses. A powerful technique used to characterize these pulses is Fourier-transform spectroscopy [3], yielding the power spectrum through a Fourier-transform of the recorded linear autocorrelation. Although such a measurement can be performed using an infrared Michelson interferometer, a sequence of two collinear infrared pulses separated by a variable time delay can more simply be generated using other techniques [4]. Here, we propose and demonstrate a new Fourier-transform spectrometer in which two time-delayed visible pulses are focused on a single infrared generator, as shown in fig. 1, thus generating a sequence of two infrared pulses. The signal recorded as a function of time delay yields the linear autocorrelation of the infrared pulse. However, when the two incident visible beams are collinear, their interaction in the generator results in an additional emission. As a consequence, the signal then recorded as a function of time delay, shown in fig. 2(a), resembles the interferometric second-order autocorrelation of the visible pulse, with a contrast larger than 2 and high-frequency fringes. Fortunately, a non-collinear geometry eliminates this crossed emission, while preserving a nearly collinear geometry in the infrared. Indeed, since the natural angle of diffraction is larger in the infrared than in the visible, we can choose an angle θ between the two incident beams such that two non-collinear incident visible beams yield two strongly overlapping infrared beams, as shown in the inset of fig. 1. Therefore, the two infrared beams can interfere on the detector, nearly as much as if they were truly collinear. In this case, the detected signal is the linear autocorrelation of the infrared pulse, with a constrast close to 2, as expected (fig. 2(b)). A Fourier transform provides the power spectrum of the infrared radiation (inset). The broad spectral width results from the use of 15-fs optical pulses on GaAs as the infrared generator [2], and the sharp cut-off above 14 µm is due to our detector spectral response.

A sub-picosecond time resolved resonance Ram: n experiment is used to measure the time scale ’ f conformational changes in hemoglobin following liga d photolysis. Comparison with an unligated species allows t e heme photophysical contribution to the signal to be isolated

Now that the structure of the RC from two different purple photosynthetic bacteria has been solved at atomic resolution [1,2], the understanding of the remarkable speed and efficiency of the light-induced primary charge separation step constitutes a challenging task in the field of electron transfer reactions. The core of the RC is made of two homologous polypeptides (denoted L and M), each containing five transmembrane α-helices. An approximate C-2 symmetry around a transmembrane axis relates the L and M α-helices as well as the various cofactors. The pigments are organized in two branches starting at a pair (P) of closely interacting bacteriochlorophylls and extending, via the monomeric bacteriochlorophylls (BA, BB, towards the bacteriopheophytins (HA, HB) and the quinones (QA, QB). The interpretation of the spectroscopic data pertaining to the initial step of electron transfer, notably the femtosecond and picosecond time evolution of the RC absorption following excitation, is a subject of intense interest [3–5]. Irrespective of earlier experiments performed under sub-optimal conditions in terms of pulse duration and excitation wavelength, most of the more recent experiments agree that the initial step of electron transfer (i.e. the transition P*→P+) requires about 3 ps at room temperature [6–9] and about 1 ps at cryogenic temperature [6, 10].

One of the most prominent features of photosynthetic reaction centers is the high quantum efficiency achieved in a sequence of transmembrane electron transfer steps. This implies an evolutionary optimization aiming at forward electron transfer rates which are sufficiently fast to overcome adverse processes such as the decay of the excited donor species into products other than charge-separated states and the recombination thereof.

In photosynthetic bacteria the light is first absorbed by antenna proteins, which contain chlorophyll-like chromophores. The initially trapped excitation very rapidly migrates to a membrane protein called the reaction center where the primary charge separation subsequently occurs. This light-induced electron transfer reaction is extremely fast even at cryogenic temperatures (700 fs at 10K in Rps. viridis [1]) and occurs with a near-unity quantum efficiency. The reaction center of different bacteria has been crystallized and high-resolution X-ray structures have been obtained for two different photosynthetic bacteria, Rps. viridis [2,3] and Rb. sphaeroides [4,5]. The primary electron donor in both species is the excited singlet state (P*) of a dimer of bacteriochlorophyll (Bchl), the special pair P. These two Bchl molecules are strongly coupled electronically, and this coupling is responsible, at least in part, for striking optical properties, like the large red shifts of the infrared Qy states as compared to monomer spectra. The relative contributions of the protein moiety and of the pigment-pigment interaction to these shifts have still to be determined. Besides the red shift, two other unusual features are observed in the special pair absorption bands: a strong exciton splitting (850 cm−1 in Rps. viridis) and a broadening of the P band.

The occurrence of differences among spatially segregated populations of a species is, since Darwin’s time, essential in any reflection on species evolution and adaptation. With the development of multivariate analysis methods the study of variation in external characters has been greatly enhanced. Indeed, “geographic variation is not likely to be due to a single environmental variable, but is doubtless a multidimensional process involving the adaptation of many characters to a variety of independant environmental factors whose gradients and ranges probably overlap in rather complex fashion” (Sokal and Rinkel 1963). Progress has also been favoured by advances in ecomorphology, which allows a better understanding of the relations existing between geographic variation in morphology and variation in ecology over the distributional range of a species or a group of species (see Hespenheide 1973 or the review by Leisler and Winkler 1985).

Dynamical analysis of photolysis products in the picosecond time scale have allowed new insights in the early processes of heme-ligand interactions in myoglobin and hemoglobin [1–4]. In a preliminary experiment SHANK et al [3] demonstrated that photolysis of HbCO occurs in less than 0.5 ps.

With the advent of powerful and ultrafast optical pulses, many physical phenomena, non linear and time dependent in the picosecond and subpicosecond time scale, become potentially directly accessible to study.

There has been recently important efforts devoted to the amplification of subpicosecond pulses, with laser pumped dye amplifier, in view of numerous applications in the domain of time-resolved spectroscopy. But still, little has been done for the theoretical understanding of such experiments. We develop here a theoretical model of laser pumped dye amplifiers to be compared to experimental arrangements which amplify subpicosecond pulses to gigawatt peak power [1,2].

The mechanisms governing the localization and solvation of excess electrons in liquids and glasses are one of the fundamental and still unsolved problems in the physics and chemistry of condensed phase. Laser photolysis of solute molecules and pulse radiolysis of polar fluids have been extensively used to generate excess electron and to investigate the initial steps leading to the long lifetime state of solvated electron. However, very little is known about the dynamics of electron solvation in liquid water. The best available measurement gives an upper limite of 0.3 ps for the appearance of solvated electron after a UV photolysis of an aqueous solution of ferrocyanure[1]. It has not been possible up to now to clearly demonstrate whether the infrared absorbing state, identified in alcools as a localized and a precursor of the fully relaxed solvated state[2], has its analog in pure liquid water.

The ability to investigate ultrafast photophysics and photochemistry of reaction centers and antenna pigments of photosynthetic bacteria and of bacteriorhodopsin is a direct result of the advancing technology for the generation and the amplification of ultrashort light pulses [1]. The dual-beam femtosecond absorption spectrometer with which the data discussed here were obtained is described in detail elsewhere [2]. For our investigations of reaction centers and antenna pigments of photosynthetic bacteria, a tunable excitation beam was obtained by amplification [2–4] of a selected portion of a white-light continuum. For systems with several interacting chromophores, such as reaction centers, tunability of the excitation and the probe beams is equally as important as temporal resolution [3,5].

Since 1975 different kinds of dye lasers have been developed to generate subpicosecond pulses: CW pumped passively mode locked oscillators [1], synchronously pumped lasers or more recently CW mode locked ring lasers [2] which generate pulses of duration less than 100 femtoseconds. Many more applications become accessible when these pulses are amplified to powers in the gigawatt range [3,4] and particularly if one can utilize nonlinear optical phenomena to generate pulses at different wavelengths. In this paper, we describe new techniques for producing high power wavelength tunable subpicosecond pulses, starting from the amplified outputs of CW mode locked linear or ring dye lasers, generating a white light continuum pulse and amplifying any selected spectral part of it to GW peak power.

While there has been a considerable effort devoted to two-photon absorption spectroscopy in solids, (in which process the intermediate state of the transition is virtual), relatively little has been done in the two-photon resonant case, where one of the incident frequencies is tuned to an absorption level of the system. This situation may provide informations upon the real intermediate state, particularly its lifetime and this aspect has already been exploited in resonant Raman, spectroscopy [1]. With available light sources of subpicosecond duration, it is possible to perform a time-resolved study of the transition process; in this way, the pure two-photon signal can be discriminated from the two-step process, and a measure of the intermediate state lifetime is obtained directly. (By contrast, resonant Raman spectroscopy measures the variation of intermediate state lifetime with input frequency, but not its magnitude).

The development of short laser pulses in the last decade provided new tools for spectroscopic studies of phenomena taking place on a picosecond time scale.