Nitrogenase — After The Structure

Abstract

The publication, in 1992, of the 3-dimensional X-ray structures of the Fe and MoFe proteins of Azotobacter vinelandii nitrogenase represented a major step forward in our understanding of this enzyme. These structures and their subsequent refinement are described elsewhere in this volume (Smith et al., 1995) and herald a new era in nitrogenase biochemistry and chemistry. We are now in a position to pose mechanistic questions on a much more informed basis, e.g.:

• (a) Where do substrates bind - at Fe or Mo or S or some combination of these? Precedents for N₂ being reduced at Fe or Mo sites exist and C₂H₂ can be reduced on S sites. Furthermore, there are many models in which N₂ is bound in a bridging mode between two metal sites. This type of complex can sometimes be reduced to form ammonia, although more frequently hydrazine is formed. There is little information on the reduction of nitrogenase substrates at sites in metal-sulphur cluster complexes. Since there is very strong evidence that the FeMoco cluster acts as the substrate binding and reducing site in nitrogenase, then it will be necessary to develop an understanding of nitrogenase substrate interactions with metal-sulphur clusters in order to attempt to understand the enzyme mechanism.

• (b) How is MgATP hydrolysis coupled to electron transfer? Nucleotide binding to each protein has been observed (Eady et al., 1995). Furthermore, MgATP hydrolysis only occurs when both proteins are present. The crystal structure of Av2 (Georgiadis et al., 1992) shows half-site occupancy by an ADP molecule. This provides very good evidence for the binding site of nucleotides to the Fe protein and yet this site is a considerable distance from the surface of the protein and cannot therefore be regarded as a possible bridging site where a nucleotide binds both to the Fe and the MoFe protein. Is it possible that more than one type of site is involved during turnover? For example, the crystallographically observed nucleotide binding site in the Fe protein could be important for modifying the interaction of the Fe protein with the MoFe protein, but another site on the MoFe protein could be involved in other stages of the substrate reduction process and might be associated with a nucleotide bridge between the two proteins. If there were two distinct sites on nitrogenase for nucleotide binding, and hydrolysis occurred at both sites, then one puzzling feature of nitrogenase function, viz. that two ATP molecules are hydrolysed at every electron transferred, may be explained.

• (c) What is the role of the P clusters? Are they directly on the electron transfer path from the Fe protein to FeMoco or do they act as suppliers of electrons only at critical points during turnover? There is now developing evidence (Kim, Rees, 1992; Thorneley et al., 1993; Peters et al., 1994) for the interaction surfaces between the Fe protein and the MoFe protein which indicates that the P clusters could be on the electron transfer path from the Fe protein to the FeMoco centres. However, there is no evidence available in the literature to indicate that the P clusters become further reduced from their dithionite-reduced form at any point during turnover. Thus there is no evidence that electrons always pass from the Fe protein 4Fe4S cluster to the P clusters. However there is evidence that the P clusters can become oxidized releasing electrons at a critical point during turnover when reducing the natural substrate N₂ or protons. This oxidation appears not to take place in the presence of carbon monoxide or of acetylene (Lowe et al., 1993). These observations of course do not exclude the possibility that the P clusters are involved in passing all electrons to FeMoco. However this hypothesis appears unnecessary on current data.