Biosynthesis of Iron-Sulfur Clusters

It is not known whether biosynthesis of [Fe-S] clusters occurs through a spontaneous self-assembly process or an enzymatic process. However, in the Azotobacter vinelandii nitrogenase system, it has been proposed that NifS and NifU are involved in the mobilization of sulfur and iron necessary for nitrogenase-specific [Fe-S] cluster assembly. The NifS protein has been shown to have cysteine desulfurase activity and can be used to supply sulfur for the in vitro catalytic formation of [Fe-S] clusters. The activity of the NifU protein has not yet been established, but NifU could have functions complementary to NifS by mobilizing iron or serving as an intermediate site necessary for nitrogenase-specific [Fe-S] cluster assembly. A second iron-binding site within NifU was predicted to serve these functions because two identical [2Fe-2S] clusters that had previously been identified within the homodimeric NifU are tightly bound, and the NifU primary sequence is rich in cysteine residues. In this dissertation, I examined the possibility that NifU might mobilize iron or serve as an intermediate site for [Fe-S] cluster assembly, as well as the possibility that NifU could work in concert with NifS.

Primary sequence comparisons, amino acid substitution experiments, and biophysical characterization of recombinantly-produced NifU fragments were used to show that NifU has a modular structure. One module is contained in approximately the C-terminal half of NifU and provides the binding site for the [2Fe-2S] cluster previously identified (the permanent [2Fe-2S] cluster). Cysteine residues Cys¹³⁷, Cys¹³⁹, Cys¹⁷⁵, and Cys¹⁷⁵ serve as ligands to the [2Fe-2S] cluster. Another module (referred to as NifU-1) is contained in approximately the N-terminal third of NifU and provides a second iron-binding site (rubredoxin-like Fe(III)-binding site). Cysteine residues Cys³⁵, Cys⁶², Cys¹⁰⁶>, and a putative non-cysteine ligand of unknown origin provide coordination to the iron at this site. The significance of these iron-binding sites was also accessed by showing that cysteine residues involved in providing the rubredoxin-like Fe(III)-binding site and those that provide the [2Fe-2S] cluster binding site are all required for the full physiological function of NifU. The two other cysteine residues contained within NifU, Cys²⁷² and Cys²⁷⁵, are neither necessary for binding iron at either site nor are they required for the full physiological function of NifU.

These results provide the basis for a model where iron bound at the rubredoxin-like sites within NifU-1 (one iron per monomer) is proposed to be destined for [Fe-S] cluster formation. It was possible to find in vitro evidence supporting this idea. First, it was demonstrated that NifU and NifS are able to form a transient complex. Second, in the presence of NifS as well as L-cysteine and a reducing agent, the Fe(III) contained at the rubredoxin-like sites within the NifU-1 or NifU homodimer can rearrange to form a transient [2Fe-2S] cluster between the two subunits. Finally, a mutant form of NifU-1 was isolated that appears to be trapped in the [2Fe-2S] cluster-containing form, and this [2Fe-2S] cluster (the transient [2Fe-2S] cluster) can be released from the polypeptide matrix upon reduction with dithionite.

Previous work has shown that the permanent [2Fe-2S] clusters of as-isolated NifU are in the oxidized form but can be reduced chemically. The transient [2Fe-2S] cluster formed between rubredoxin-like sites, in contrast, is reductively labile. If the transient cluster serves as an intermediate [Fe-S] cluster to be destined for [Fe-S] cluster assembly, I propose that the permanent [2Fe-2S] clusters could have redox roles participating in either one or all of the following events. The permanent [2Fe-2S] clusters could have a redox function in the acquisition of iron for initial binding at the mononuclear sites. They could also provide reducing equivalents for releasing the transient [2Fe-2S] cluster. In addition, upon releasing the transient [2Fe-2S] cluster, the permanent [2Fe-2S] clusters could provide the appropriate oxidation state of the irons to be destined to nitrogenase metallocluster core formation.

Finally, because proteins homologous to NifU and NifS are widely distributed in nature, it is suggested that the mechanism for NifU and NifS in the formation of nitrogenase-specific [Fe-S] clusters could represent a general mechanism for [Fe-S] cluster synthesis in other systems.