THE DISSIMILATORY SULFITE REDUCTASE FROM DESULFOSARCINA-VARIABILIS ISA DESULFORUBIDIN CONTAINING UNCOUPLED METALATED SIROHEMES AND S = 9 2IRON-SULFUR CLUSTERS/

The active site of Escherichia coli NADPH-sulfite reductase has previo usly been modeled as a siroheme with its iron bridged to a nearby iron -sulfur cubane, resulting in antiferromagnetic exchange coupling betwe en all iron atoms. The model has been suggested to hold also for other sulfite reductases and nitrite reductases. We have recently challenge d the generality of the model with the finding that the EPR of Fe/S in dissimilatory sulfite reductase (desulfoviridin) from Desulfovibrio v ulgaris indicates that an S = 9/2 system is not subject to coupling. S iroheme in desulfoviridin is to a large extent demetalated, and theref ore coupling is physically impossible. We have now studied examples fr om a second class of dissimilatory sulfite reductases, desulforubidins , which have their siroporphyrins fully metalated. Desulforubidin from Desulfosarcina variabilis is a 208-kDa alpha2beta2gamma2 hexamer. The alpha- and beta-subunits are immunologically active with antibodies r aised against the corresponding subunits from D. vulgaris desulfovirid in, whereas the gamma-subunit is not. The desulforubidin contains two fully metalated sirohemes and a total of almost-equal-to 15 Fe and alm ost-equal-to 19 S2-. Quantification of high-spin plus low-spin heme EP R signals accounts for all sirohydrochlorin. The frequency-independent (9-35 GHz) effective perpendicular g-values of the high-spin S = 5/2 siroheme (6.33, 5.19) point to quantum mixing with an excited (almost- equal-to 770 cm-1) S = 3/2 multiplet. Similar anomalous g-values are o bserved with sulfite reductases from Desulfovibrio baarsii and Desulfo tomaculum acetoxidans. The D. variabilis enzyme exhibits very approxim ately stoichiometric S = 9/2 EPR (g = 16). None of the EPR signals giv e indication for dipolar and/or exchange coupling between siroheme and iron-sulfur clusters. S = 9/2 EPR is not detected in concentrated sam ples of assimilatory sulfite reductases from E. coli and from D. vulga ris. Thus, the functional difference between dissimilatory and assimil atory sulfite reductases appears to have a structural parallel in the presence or absence, respectively, of an S = 9/2 EPR iron-sulfur clust er.