The simultaneous excitation of paramagnetic molecules with optical (laser) and microwave radiation in the presence of a magnetic field can cause an amplitude, or phase, modulation of the transmitted light at the microwave frequency. The detection of this modulation indicates the presence of coupled optical and ESR transitions. The phenomenon can be viewed as a coherent Raman effect or, in most cases, as a microwave frequency modulation of the magnetic circular dichroism by the precessing magnetization. By allowing the optical and magnetic properties of a transition metal ion centre to be correlated, it becomes possible to deconvolute the overlapping optical or ESR spectra of multiple centres in a protein or of multiple chemical forms of a particular centre. The same correlation capability also allows the relative orientation of the magnetic and optical anisotropies of each species to be measured, even when the species cannot be obtained in a crystalline form. Such measurements provide constraints on electronic structure calculations. The capabilities of the method are illustrated by data from the dimeric mixed-valence CuA centre of nitrous oxide reductase (N2OR) from Paracoccus pantotrophus.
- coherent Raman
- electron spin resonance (ESR)
- magnetic circular dichroism (MCD)
Historical development and principles
Our work to develop an optical-microwave double-resonance experiment for the study of metalloproteins was stimulated by the wish to obtain enhanced chemical and orientational selectivity of individual paramagnetic metal centres [1,2]. The microwave and optical resonance conditions depend on the chemical nature of the species. In an optical-microwave double-resonance experiment, this can be used to identify the optical spectrum of the species responsible for a particular ESR signal, or conversely measure the ESR spectrum of the species generating a specific optical absorption band. The means to deconvolute overlapping optical or ESR spectra is especially attractive in the case of proteins with multiple centres, or samples containing multiple forms of a given centre.
The microwave resonance condition and the intensity of optical bands are also dependent on molecular orientation. If the microwave resonance condition selects molecules with specific orientations, then the double-resonance measurement will allow one to obtain a single-crystal-like optical spectrum of these centres. Obtaining the relative orientation of the optical (e.g. transition polarization) and magnetic (e.g. g-value) anisotropies in this way allows more rigorous tests of chemical and electronic structural models. This approach may be especially attractive for the study of unstable transient species, such as enzyme intermediates which are often impossible to obtain in an orientated crystalline form.
Double-resonance spectroscopy connects transitions resonant with the two radiation frequencies that share a common quantum state. The simplest approach relies upon perturbing the population of the common state by ‘pumping’ one transition with one radiation field and detecting this change via a modified response at the second ‘probe’ frequency. ENDOR (electron nuclear double resonance) spectroscopy and NOESY  are examples of this class. This was also the approach taken in early optical-microwave double-resonance metalloprotein work at the University of East Anglia [4,5]. Some considerable success was obtained, e.g. in identifying the principal optical bands of the CuA centre in cytochrome oxidase . However, magnetic relaxation phenomena [6,7], such as cross-relaxation between species and spectral diffusion, would often remove the desired chemical or orientational selectivity. In addition, exploitation of the orientational selectivity required a detailed quantitative knowledge of the magnetic relaxation anisotropy. Rarely is such knowledge available for metalloprotein systems where this form of spectroscopy has its greatest potential.
These difficulties led us to investigate the feasibility of a second mode of double-resonance experiment, namely coherence transfer spectroscopy [3,8]. Coherence transfer involves a triangle of transitions between three states (Figure 1A). If radiation fields excite two of the three transitions, coherence transfer will also excite the third transition, even though there is no radiation field resonant with it. Provided this transition has a non-zero transition dipole moment, then radiation will be emitted. The frequency of the emitted light will be the sum or difference of the two incident fields depending on the ordering of the three states in energy. Such processes are usually referred to as coherent Raman effects  and are employed in some well-known spectroscopic techniques, such as ESEEM (electron spin-echo envelope modulation) spectroscopy , COSY , and CARS (coherent anti-Stokes Raman spectroscopy) . However, unlike those techniques, it is not necessary to employ pulsed measurements in the optical-microwave double-resonance experiment.
Derivation of selection rules and optimal experimental geometry from a coherent Raman viewpoint can be complex unless the optical line widths are much larger than the microwave frequency. Then we can ‘decouple’ the relatively long microwave-timescale and short optical-timescale components in the theory, and the experiment becomes a microwave frequency modulation of the optical properties, or, to be more specific, a modulation of the MCD (magnetic circular dichroism)  by the microwave-induced precession of the (effective) spin . The oscillation is largest in the plane normal to the applied magnetic field, and the largest optical modulation is observed for a circularly polarized beam propagating in this plane. The frequency-shifted coherent Raman light is equivalent to modulation sidebands in the microwave-frequency modulation picture (Figure 1).
The key advantage is that only molecules resonant with the optical and microwave radiation contribute to the signal. Although magnetic relaxation processes will, as in conventional ESR, contribute to line broadening, they do not affect the total integrated signal intensity. Unlike the earlier population-transfer-based experiments, it is not necessary to know the details of the magnetic relaxation processes to extract useful information from the data. Even more importantly, chemical and orientational selectivity will be present as long as the chemically or orientationally induced differences can be spectrally resolved. The experiment is therefore much better suited to the study of poorly defined biochemical samples. Although coherence transfer is conceptually more difficult than population transfer, the theory, simulation and interpretation of the experiment are much more tractable.
However, in a sample containing randomly oriented magnetically anisotropic centres, only a small proportion of the centres will contribute to the signal. For typical values of transition metal ion g-value anisotropy, the modulations of the transmitted light beam are a few parts in a million at most. Signals of this magnitude would be too small to detect in a conventional MCD instrument . However, because lasers are exceptionally stable on a microwave-frequency timescale, we can achieve near ideal performance using microwave-frequency optical heterodyne detection methods . Sensitivities of approx. 1 part in 108 in absorbance modulation can be achieved in typical measurements. This is sufficient sensitivity to study almost all ESR active transition metal ion centres found in proteins. Coherence transfer is thus a more widely applicable detection method than population transfer.
We have now demonstrated experimentally the wide applicability of coherent Raman detected ESR spectroscopy. The first experiments were on ruby (Cr3+:Al2O3) , an exceptionally well understood material that can be used as a calibration standard. Metalloproteins with a wide range of physical properties have now been studied by us: low-spin ferric haem cytochrome c551 [11,14]; a type-1 ‘blue’ copper protein azurin [15,16]; the CuA centre of N2OR (nitrous oxide reductase) ; rubredoxin, a mononuclear iron–sulfur high-spin ferric ion ; and a [3Fe–4S] cluster with spin 1/2 in ferredoxin I, Azotobacter chroococcum (L. Udovicic, D. Suter, A.J. Thomson and S.J. Bingham, unpublished work). Excellent agreement with theory was obtained.
Here we illustrate the CuA centre of Paracoccus pantotrophus N2OR . The sample is a cryogenic glass to provide optimal spectral resolution and signal magnitude. The applied magnetic field is varied at a series of fixed laser wavelengths. The ESR spectra obtained thus differ in several ways from the conventional microwave-detected spectrum. First, it is more sensitive to measure the dispersion signal, the component of the signal that oscillates in phase with the microwave field, rather than the phase-quadrature absorption signal more commonly seen in ESR spectroscopy. The conventional magnetic modulation scheme that produces derivative spectra is not employed for sensitivity reasons. Centres with different orientations will be resonant with the microwaves at different fields and hence the spectral lineshape depends on the MCD variation with orientation. CuA has essentially axially symmetrical g-values, giving two limiting cases: when the observable MCD arises along the unique g-value symmetry axis (conventionally labelled as z), and along the plane normal to this axis. Simulations of these two cases are shown in Figure 3 (labelled as Czz and Cperp respectively), together with the corresponding conventional microwave-detected lineshape. The experimental data can only be a linear combination of the two limiting cases. Each optical band has a unique ESR lineshape owing to the different nature of the excited optical level. The MCD anisotropy data indicate the nature of these levels and therefore can be used elucidate the electronic and chemical structure of the centre. The interpretation of the observed MCD anisotropy in terms of the structure of the CuA centre is given in .
Now that the theory and instrumentation of coherent Raman detected ESR has been successfully developed and tested, it can be used in a wide range of biochemical applications. The method can measure far more subtle differences in the MCD anisotropy than variable-field variable-temperature MCD spectroscopy. The ability to study selectively different centres within multi-centre enzymes is also very clear.
Recent support of the U.K. research councils, EPSRC (Engineering and Physical Sciences Research Council) and BBSRC (Biotechnology and Biological Sciences Research Council), is gratefully acknowledged. Sample preparation and characterization was performed by Tim Rasmussen, Jaqui Farrar and Myles Cheesman at UEA.
Transition Metals in Biochemistry: A joint Biochemical Society meeting with the Inorganic Biochemistry Discussion Group to honour Professor Andrew Thomson FRS, held at University of East Anglia, Norwich, U.K., 24–26 June 2008. Organized and Edited by Steve Chapman (Edinburgh, U.K.), David Richardson (University of East Anglia, U.K.) and Nick Watmough (University of East Anglia, U.K.).
Abbreviations: MCD, magnetic circular dichroism; N2OR, nitrous oxide reductase
- © The Authors Journal compilation © 2008 Biochemical Society