Search

The influences of the binding of a urea and a phenolic herbicide on the T4 mutant of Blastochloris viridis, focusing on the recombination reactions and the redox potential of Q(sub A) are studied. It is observed that the influence of the herbicides can be considered to be on the equilibrium of the two reaction center forms.

TrmFO is a tRNA methyltransferase that uses methylenetetrahydrofolate (CH2THF) and flavin adenine dinucleotide hydroquinone as cofactors. We have recently shown that TrmFO from Bacillus subtilis stabilizes a TrmFO-CH2-FADH adduct and an ill-defined neutral flavin radical. The adduct contains a unique N-CH2-S moiety, with a methylene group bridging N5 of the isoalloxazine ring and the sulfur of an active-site cysteine (Cys53). In the absence of tRNA substrate, this species is remarkably stable but becomes catalytically competent for tRNA methylation following tRNA addition using the methylene group as the source of methyl. Here, we demonstrate that this dormant methylating agent can be activated at low pH, and we propose that this process is triggered upon tRNA addition. The reaction proceeds via protonation of Cys53, cleavage of the C-S bond, and generation of a highly reactive [FADH(N5)═CH2]+ iminium intermediate, which is proposed to be the actual tRNA-methylating agent. This mechanism is fully supported by DFT calculations. The radical present in TrmFO is characterized here by optical and EPR/ENDOR spectroscopy approaches together with DFT calculations and is shown to be the one-electron oxidized product of the TrmFO-CH2-FADH adduct. It is also relatively stable, and its decomposition is facilitated by high pH. These results provide new insights into the structure and reactivity of the unique flavin-dependent methylating agent used by this class of enzymes.

The flavoprotein TrmFO catalyzes the C5 methylation of uridine 54 in the TΨC loop of tRNAs using 5,10-methylenetetrahydrofolate (CH(2)THF) as a methylene donor and FAD as a reducing agent. Here, we report biochemical and spectroscopic studies that unravel the remarkable capability of Bacillus subtilis TrmFO to stabilize, in the presence of oxygen, several flavin-reduced forms, including an FADH(•) radical, and a catalytic intermediate endowed with methylating activity. The FADH(•) radical was characterized by high-field electron paramagnetic resonance and electron nuclear double-resonance spectroscopies. Interestingly, the enzyme exhibited tRNA methylation activity in the absence of both an added carbon donor and an external reducing agent, indicating that a reaction intermediate, containing presumably CH(2)THF and FAD hydroquinone, is present in the freshly purified enzyme. Isolation by acid treatment, under anaerobic conditions, of noncovalently bound molecules, followed by mass spectrometry analysis, confirmed the presence in TrmFO of nonmodified FAD. Addition of formaldehyde to the purified enzyme protects the reduced flavins from decay by probably preventing degradation of CH(2)THF. The absence of air-stable reduced FAD species during anaerobic titration of oxidized TrmFO, performed in the absence or presence of added CH(2)THF, argues against their thermodynamic stabilization but rather implicates their kinetic trapping by the enzyme. Altogether, the unexpected isolation of a stable catalytic intermediate suggests that the flavin-binding pocket of TrmFO is a highly insulated environment, diverting the reduced FAD present in this intermediate from uncoupled reactions.

Superoxide dismutases (SODs) are proteins specialized in the depletion of superoxide from the cell through disproportionation of this anion into oxygen and hydrogen peroxide. We have used high-field electron paramagnetic resonance (HFEPR) to test a two-site binding model for the interaction of manganese-SODs with small anions. Because tyrosine-34 was thought to act as a gate between these two sites in this model, a tyrosine to phenylalanine mutant of the superoxide dismutase from R. capsulatus was constructed. Although the replacement slightly reduced activity, HFEPR measurements demonstrated that the electronic structure of the Mn(II) center was unaffected by the mutation. In contrast, the mutation had a profound effect on the interactions of fluoride and azide with the Mn(II) center. It was concluded that the absence of tyrosine-34 prevented the close approach of these anions to the metal ion. This mutation also enhanced the formation of a hexacoordinated water-Mn(II)SOD complex at low temperatures. Together, these results showed that the role of Y34 is unlikely to involve redox tuning but rather is important in regulating the equilibria between the anionic substrate in solution and the two binding sites near the metal. These observations further supported the originally proposed mutually exclusive two-binding-site model.

The Mn4Ca cluster of photosystem II (PSII) goes through five sequential oxidation states (S0-S4) in the water oxidation process that also involves a tyrosine radical intermediate (TyrZ*). An S2TyrZ* state in which the Mn4Ca cluster and TyrZ* are magnetically coupled to each other and which is characterized by a distinct "split-signal" EPR spectrum can be generated in acetate-treated PSII. This state was examined by high-field EPR (HFEPR) in PSII from Thermosynechococcus elongatus isolated from a D2-Tyr160Phe mutant to avoid spectral contributions from TyrD*. In contrast to the same state in plants, both antiferromagnetic and ferromagnetic spin-spin couplings were observed. The intrinsic g values of TyrZ* in the coupled state were directly measured from the microwave frequency dependence of the HFEPR spectrum. The TyrZ* gx value in the antiferromagnetic centers was 2.0083, indicating that the coupled radical was in a less electropositive environment than in Mn-depleted PSII. Two gx values were found in the ferromagnetically coupled centers, 2.0069 and 2.0079. To put these values in perspective, the second redox-active tyrosine, TyrD*, was examined in various electrostatic environments. The TyrD* gx value changed from 2.0076 in the wild type to 2.0095 when the hydrogen bond from histidine 189 to TyrD* was removed using the D2-His189Leu mutant, indicating a change to a significantly less electropositive environment. BLY3P/6-31+G** density functional calculations on the hydrogen-bonded p-ethylphenoxy radical-imidazole supermolecular model complex showed that the entire range of Tyr* gx values, from 2.0065 to 2.0095, could be explained by the combined effects of hydrogen bonding and the dielectric constant of the local protein environment.

The effect of the substrate analogues azide and fluoride on the manganese(II) zero-field interactions of different manganese-containing superoxide dismutases (SOD) was measured using high-field electron paramagnetic resonance spectroscopy. Two cambialistic types, proteins that are active with manganese or iron, were studied along with two that were only active with iron and another that was only active with manganese. It was found that azide was able to coordinate directly to the pentacoordinated Mn(II) site of only the MnSOD from Escherichia coli and the cambialistic SOD from Rhodobacter capsulatus. The formation of a hexacoordinate azide-bound center was characterized by a large reduction in the Mn(II) zero-field interaction. In contrast, all five SODs were affected by fluoride, but no evidence for hexacoordinate Mn(II) formation was detected. For both azide and fluoride, the extent of binding was no more than 50%, implying either that a second binding site was present or that binding was self-limiting. Only the Mn(II) zero-field interactions of the two SODs that had little or no activity with manganese were found to be significantly affected by pH, the manganese-substituted iron superoxide dismutase from E. coli and the Gly155Thr mutant of the cambialistic SOD from Porphyromonas gingivalis. A model for anion binding and the observed pK involving tyrosine-34 is presented.

Various tyrosyl radicals generated by reaction of both native and indomethacin-inhibited ovine prostaglandin H synthase-1 with ethyl hydrogen peroxide were examined by using high-field/high-frequency EPR spectroscopy. The spectra for the initially formed tyrosyl radical commonly referred to as the "wide-doublet" species and the subsequent "wide-singlet" species as well as the indomethacin-inhibited "narrow-singlet" species were recorded at several frequencies and analyzed. For all three species, the g-values were distributed. In the case of the wide doublet, the high-field EPR spectra indicated that dominant hyperfine coupling was likely to be also distributed. The g(x)-values for all three radicals were found to be consistent with a hydrogen-bonded tyrosyl radical. In the case of the wide-doublet species, this finding is consistent with the known position of the radical and the crystallographic structure and is in contradiction with recent ENDOR measurements. The high-field EPR observations are consistent with the model in which the tyrosyl phenyl ring rotates with respect to both the protein backbone and the putative hydrogen bond donor during evolution from the wide-doublet to the wide-singlet species. The high-field spectra also indicated that the g-values of two types of narrow-singlet species, self-inactivated and indomethacin-inhibited, were likely to be different, raising the possibility that the site of the radical is different or that the binding of the inhibitor perturbs the electrostatic environment of the radical. The 130 GHz pulsed EPR experiments performed on the wide-doublet species indicated that the possible interaction between the radical and the oxoferryl heme species was very weak.

The radical forms of two cofactors and an amino acid in the photosystem II (PS II) reaction center were studied by using high-field EPR both in frozen solution and in oriented multilayers. Their orientation with respect to the membrane was determined by using one-dimensionally oriented samples. The ring plane of the stable tyrosyl radical, Y(D)(*), makes an angle of 64 degrees +/- 5 degrees with the membrane plane, and the C-O direction is tilted by 72 degrees +/- 5 degrees in the plane of the radical compared to the membrane plane. The semiquinone, Q(A)(*)(-), generated by chemical reduction in samples lacking the non-heme iron, has its ring plane at an angle of 72 degrees +/- 5 degrees to the membrane plane, and the O-O axis is tilted by 21 degrees +/- 5 degrees in the plane of the quinone compared to the membrane plane. This orientation is similar to that of Q(A)(*)(-) in purple bacteria reaction centers except for the tilt angle which is slightly bigger. The pheophytin anion was generated by photoaccumulation under reducing conditions. Its ring plane is almost perpendicular to the membrane with an angle of 70 degrees +/- 5 degrees with respect to the membrane plane. This is very similar to the orientation of the pheophytin in purple bacteria reaction centers. The position of the g tensor with respect to the molecule is tentatively assigned for the anion radical on the basis of this comparison. In this work, the treatment of orientation data from EPR spectroscopy applied to one-dimensionally oriented multilayers is examined in detail, and improvements over previous approaches are given.

The Mn cluster of Photosystem II (PSII) from Synechococcus elongatus was studied using EPR. A signal with features between g = 5 and g = 9 is reported from the S2-state. The signal is attributed to the manganese cluster in a state with a spin 5/2 state. Spectral simulations of the signal indicate zero field splitting parameters where the |E/D| was 0.13. The new signal is formed by irradiating PSII samples which contain the spin = 1/2 S2-state using 813 nm light below 200 K. This effect is attributed to a spin-state change in the manganese cluster due to absorption of the IR light by the Mn-cluster itself. The signal is similar to that reported recently in PSII of plants [Boussac, A., Un, S., Horner, O., and Rutherford, A. W. (1998) Biochemistry 37, 4001-4007]. In plant PSII the comparable signal is formed at a lower temperature (optimally below 77 K), and gradual warming of the sample in the dark leads to the formation of the state responsible for the well-known g = 4.1 signal prior to formation of the spin 1/2 multiline signal. In the present work using cyanobacterial PSII, warming of the sample in the dark leads to the formation of the spin 1/2 multiline signal without formation of the g = 4 type signal as an intermediate. These observations provide a partial explanation for the long-standing "mystery of the missing g = 4 state" in cyanobacterial PSII. The observations are rationalized in terms of three possible states which can exist for S2: (i) the spin 1/2 multiline signal, (ii) the state responsible for the g = 4.1 signal, and (iii) the new spin 5/2 state. The relative stability of these states differs between plants and cyanobacteria.

The Mn4 complex which is involved in water oxidation in photosystem II (PSII) is known to exhibit two types of EPR signals in the S2 state, one of the five redox states of the enzyme cycle: either a multiline signal (S = 1/2) or a signal at g = 4.1 (S = 3/2 or S= 5/2). The S = 1/2 state can be converted to that responsible for the g = 4.1 signal upon the absorption of near-infrared (IR) light [Boussac, A., Girerd, J.-J., and Rutherford, A.W. (1996) Biochemistry 35, 6984-6989]. It is shown here that a third state gives rise to signals at g = 10 and 6. This state is formed by IR illumination of the S = 1/2 state at 65 K, a temperature where IR illumination leads to the loss of the S = 1/2 signal but to no formation of the g = 4.1 state. On the basis of the corresponding decrease of the S = 1/2 state, the new state can be trapped in approximately 40% of the PSII centers. Warming of the sample above 65 K, in the dark, leads to the loss of the g = 10 and 6 resonances with the corresponding appearance of the g = 4.1 signal. It is suggested that the IR-induced conversion of the S = 1/2 state into the g = 4.1 state at 150 K involves the transient formation of the new state. The new state is attributed to a S = 5/2 state of the Mn4 complex (although a S value > 5/2 is also a possibility). Spectral simulations indicate an E/D ratio of -0.05 with D

The photosynthetic reaction center of photosystem I (PS I) contains a phylloquinone molecule (A1) which acts as a transient electron acceptor. In PS I form the cyanobacterium Synechocystis PCC 6803 under reducing conditions, we have photoaccumulated an EPR signal assigned to the phylloquinone radical anion. The phylloquinone EPR spectrum has been studied in oriented multilayers of PS I using EPR at 9 GHz. In addition, the phyllosemiquinone spectrum has been obtained at 283 GHz using high-field, high-frequency EPR spectroscopy. From the orientation dependence of the spectrum at 9 GHz and the resolved g values obtained at 283 GHz, the phyllosemiquinone ring plane was determined to be almost perpendicular to the membrane (76 degrees) while the oxygen-oxygen (O-O) axis of the quinone was found to make an approximate 63 degrees angle to the membrane plane. The orientation of the ring plane is similar to that determined for the quinone electron acceptor (QA) in the purple bacterial reaction center, while the orientation of the O-O axis is significantly different. The new orientation information, when taken with data in the literature, allows the position of the phylloquinone in the reaction center to be better defined.

A 245 GHz 8.7 T high-field EPR study of tyrosine-D (TyrD zero) and tyrosine-Z (TyrZ zero) radicals of photosystem II (PSII) from Synechocystis PCC 6803 was carried out. Identical principal g values for the wild-type Synechocystis and spinach TyrD zero showed that the two radicals were in similar electrostatic environments. By contrast, the principal g values of the TyrD zero in the D2-His189Gln mutant of Synechocystis were different from those of the wild-type and spinach radicals and were similar to those of the tyrosyl radical in ribonucleotide reductase. These comparisons indicate that the D2-His189Gln mutant TyrD zero is not hydrogen-bonded or is only weakly so. The HF-EPR spectrum of TyrZ zero was obtained from the D2-Tyr160Phe mutant that lacks TyrD zero. The principal g values were nearly identical to those of the wild-type TyrD zero. The low-field edge of the TyrZ zero spectrum was much broader than at the other two principal g values and was also much broader than the TyrD zero spectrum. From the identical g values and previous work on tyrosyl radical g values [Un S., Atta M., Fontecave, M., & Rutherford, A. W. (1995) J. Am. Chem. Soc. 117, 10713-10719], it was concluded that TyrZ zero, like TyrD zero, is hydrogen-bonded The broadness of the gx component was interpreted as a distribution in strength of the hydrogen-bonding due to disorder in the protein environment about TyrZ zero.