The photochemical reaction that plants and photosynthetic bacteria use to capture the energy of sunlight occurs in pigment-protein "reaction centers." Bacterial reaction centers typically contain four molecules of bacteriochlorophyll (BChl), two bacteriopheophytins (BPh) and two quinones (Q_A and Q_B) bound to three polypeptides. Two of the four BChls are very close together. When this special pair (P) is excited with light, it passes an electron to the quinones by way of the neighboring BChl and BPh.

Our research group is using fast spectroscopic techniques and computational approaches to study the initial electron-transfer steps in reaction centers. A homebuilt Ti/S laser is used to excite a sample with pulses of light lasting less than 10^-13 s. A second pulse of light probes the absorption spectrum of the sample as a function of time after the excitation. The spectrum changes as an electron moves from one molecule to another. The excited BChl dimer (P*) appears to transfer an electron to one of the neighboring BChls in about 3 ps, creating a radical-pair (P⁺BChl^-). An electron moves from BChl^- to a BPh in about 1 ps and from BPh^- to Q_A in 200 ps. In addition to being very fast and efficient, these reactions have the unusual property of speeding up with decreasing temperature.

The reaction center provides an ideal system for addressing the general question of how enzymes work because the kinetics can be measured with high precision under a wide range of conditions in a protein with a well-defined structure, and because the initial reactions occur on a time scale that is amenable to molecular-dynamics simulations.

To explore the way in which pigment-protein interactions affect the energies of P⁺BChl^- and P⁺BPh^-, and thus control the rate, temperature dependence and specificity of electron transfer, we are measuring the energies and electron-transfer kinetics in bacterial mutants with amino acid substitutions near the pigments. Mutations of a particular tyrosine residue were found to change the rate and temperature dependence of the initial reaction dramatically. In a current project, the electrostatic symmetry of the protein is being modified by mutations of ionizable residues. A variety of computational approaches also are being used, including molecular dynamics simulations and electrostatics and quantum calculations. By combining accurate electrostatics and molecular dynamics calculations, it is possible to calculate the free energy surfaces of the reactant and product states as functions of the reaction coordinate and to identify vibrational motions of the protein that are coupled to the electron-transfer reactions. Comparisons of measured and calculated effects of mutations are used to refine the computational methods

We are also using experimental and computational techniques to study antenna complexes that absorb light and transfer energy to the reaction center. These are integral membrane proteins that typically contain 24 to 36 BChls. Topics currently under investigation include the extent to which excitations are delocalized over the entire complex, the manner in which excitations move from complex to complex, and the structural organization of the antenna complexes and reaction centers in the membrane.

In other recent work, we have used resonance Raman and IR spectroscopy to study the vibrational modes that are coupled to excitation of green fluorescent protein (GFP).

The lab recently has embarked on a new study of catechol-O-methyltransferase (COMT), an enzyme that plays important roles in inactivation of the neurotransmitter dopamine and other catechols, including catechol estrogens. COMT occurs in both soluble and membrane-bound forms. In addition, variants that differ only in the substitution of methionine at position 108 of the soluble protein or residue 158 of the membrane-bound protein occur commonly. The 108/158Met variant is unstable under physiological conditions, while the Val variant is stable. The instability of 108/158Met COMT appears to have serious consequences for individuals with chromosome 22q11.2-deletion syndrome (velocardiofacial syndrome), who are missing the COMT gene on one copy of chromosome 22. We are using a combination of experimental and computational techniques to explore the reasons why 108/158Met COMT is unstable, and to search for ways to stabilize the enzyme.