Ferredoxins structures, crystallographic

Three-iron clusters. Despite the availability of an X-ray crystallographic structure determination of Azotobacter vinelandii ferredoxin I (27), which contains both a normal 4Fe-4S cluster and a 3Fe cluster, the fundamental properties and even the structures... 

Furthennore, resonance Raman studies of aconitase by Johnson et al. (53) demonstrated homologous spectra for both inactive and active aconitase. This suggests similar vibrational modes and thus similar core structures for the two forms. Finally, a cubane structure for the [3Fe-4S] cluster is supported by recent protein crystallographic studies of inactive aconitase by Robbins and Stout (54). (Recent results from Jensen s group (55) on the redetermination of the crystal structure of the Azotobacter ferredoxin I clearly show that the 3Fe cluster does not have a [3Fe-3S] ring structure, as originally determined (37), but has a [3Fe-4S] cubane structure.)... 

A brief historical note on the structure of the iron-sulfur clusters in ferredoxins is relevant. After the first analytical results revealed the presence of (nearly) equimolar iron and acid-labile sulfur, it was clear that the metal center in ferredoxins did not resemble any previously characterized cofactor type. The early proposals for the Fe S center structure were based on a linear chain of iron atoms coordinated by bridging cysteines and inorganic sulfur (Blomstrom et al., 1964 Rabino-witz, 1971). While the later crystallographic analyses of HiPIP, PaFd, and model compounds (Herskovitz et al., 1972) demonstrated the cubane-type structure of the 4Fe 4S cluster, the original proposals have turned out to be somewhat prophetic. Linear chains of sulfide-linked irons are observed in 2Fe 2S ferredoxins and in the high-pH form of aconitase. Cysteines linked to several metal atoms are present in metallothionein. The chemistry of iron-sulfur clusters is rich and varied, and undoubtedly many other surprises await in the future. 

Figure 10.3 Crystallographic structures of the most important iron-sulphur centres in proteins (a) dimeric centre of ferrodoxin from Spinacia oleracea [54], (b) trimeric centre of ferredoxin from Bacillus schlegelii [55], (c) cubane cluster of nitrogenase reductase from Azotobacter vinelandii [56], (d) nitrogenase octameric cluster from Azotobacter vinelandii [57], (e) nitrogenase octameric cluster from Clostridium pasteurianum [58], (f) MoFe cluster of nitrogenase from Azotobacter vinelandii [59], and (g) active centre of sulphite reductase from Escherichia coli [60]...

Azotobactor vinelandii ferredoxin I has two Fe/S clusters, [4Fe-4S] and a [3Fe-xS], with redox potentials at —0.42 and +0.32 V versus NHE, respectively (36). The existence of a normal Fe4S42+ core and a [3Fe-3S](S-cys)s(oxo) core has been considered. In the ferredoxin I the unique sequence Cys-Val-Glu-Val-Cys has been suggested as a tridentate ligand for the [3Fe-3S] cluster (37). Recently, the structure has been crystallographically (38) corrected, and the iron-sulfur centers are now believed to consist of Fe4S42+ and [3Fe-4S] clusters, as proposed by EXAFS (39). One possibility is that the above-mentioned [3Fe jcS] structure is formed during isolation by oxidative removal of an Fe ion from the [4Fe-4S] cluster (40). 

Fig. 11. The crystallographic structure within 5 A of the Fe4S4 cluster I ligated by Cys-Ile-Ala-Cys-Gly-Ala-Cys in P. aerogenes ferredoxin (6, 43). Open arrows show the direction of NH groups.

Structural data on the organization of the active clusters and the various subunits can be obtained from biochemical, biophysical and genetic studies. However, the final word is in the mouth of the crystallographer, after the biochemist hands him the crystals. Only one piece of solid information is available to date, and this is the amino acid sequences of subunits la and Ib [78,79]. These two subunits should accommodate the P-700, the primary electron acceptor (Ai), and probably the secondary electron acceptor A2 which is the nonheme iron cluster X. The amino acid sequences of subunits I and I, make it unlikely that these subunits also contain one of the bound ferredoxins A or B because subunit I, contains 4 cysteine residues and subunit I, contains only 2 cysteines [79]. These numbers are hardly sufficient for the formation of the nonheme iron cluster X which is supposed to be a nonheme iron center [74,88]. Therefore, it may be likely that sub-... 

A situation similar to that outlined above for rubredoxin occurred for the 7Fe-7S ferredoxin of A. vinelandii. Thus spectroscopic studies, including iron K-edge EXAFS data for this and related systems (63, 64), caused crystallographers to reconsider and correct the structural details for the [3Fe-4Sl center (65). 

Fig. 21. Stereogram ofthe nicotinamide binding site in the FNR(Y308->S)NADP complex. For clarity, only the isoalloxanzine and the ribityl portion of FAD and the nicotinamide mononucleotide portion of NADP are shown. Figure source Deng. Aliverti. Zanetti, Arakaki, Ottado, Orellano, Calcaterra, Ceccarelli, Carrillo and Karplus (1999) A productive NADP binding mode of ferredoxin-NADP reductase revealed by protein engineering and crystallographic studies. Nature Structural Biology 6 848...

X-ray structure of the ferredoxin NADP-l-reduetase from the cyanobacterium Anabaena PCC 7119 at 1.8 A resolution, and crystallographic studies of NADP binding at 2.25 A resolution. J. Mol. Biol 263, 20-39. 

The crystals of several clostridial ferredoxins are shown in Figure 2. It indicates some of the problems involved in the determination of the structure of the protein. The crystalline forms from different clostridial species differ, but all are very small crystals. The largest ones I have seen are these from Clostridium acidi-uiici, but even these are very small as far as a crystallographer is concerned, and unfortunately, they are multiple crystals. There has been very little work done on the structure by x-ray crystallography. One problem is the small size of the crystals the second diflSculty is the fact that it has not been possible to obtain any heavy metal replacement for these proteins. However, Jensen has gotten some information on clostridial type ferredoxin that I will discuss a httle later. The plant type ferredoxin has not been obtained in crystalline form that is suitable for any x-ray work. 

One of the two important parts of enzyme is a flavo-iron-sulfur protein with NADH-dependent oxidoreductase activity. The reductase is a monomeric 34 kDa (in case of phthalate dioxygenase reductase) flavo-iron-sulfur protein containing flavin mononucleotide (FMN) and a plant-ferredoxin-type [2Fe-2S] center in a 1 1 ratio [372]. Structure of this part of the enzyme has been studied recently by X-ray crystallographic analysis [375], low-temperature EPR [383], and kinetically [378]. 

---