Nailing Down Iron Assimilation

Iron is both a nutritious and potentially noxious element. Its precise regulation in cells is necessary to assure synthesis of several proteins of which the metal is an integral part. At the same time, the cell must avoid generation of tissue-damaging oxygen radicals, which are known to arise from certain transition metals, such as iron. Environmental iron is mostly in a form (the 3+ oxidation state) that is characterized by extreme insolubility at biological pH. In order to dissolve iron, many bacterial and fungal species synthesize a line of relatively low molecular weight, virtually ferric-specific chelating, or binding, agents generically called siderophores (Gr. “iron bearer,” ISI Atlas of Science: Biochemistry, 1, 53-6, 1988). The affinity of siderophores for iron (III) is such that they can compete effectively with hydroxyl ion for the metal. Microorganisms have also evolved highly efficient transport systems for ferric siderophores that enable them to recognize trace amounts of the complexed metal ion and make it available to metabolic demands within the cell. A remarkable feature of the siderophore biosynthesis an transport pathway is its ability to adapt to the external level of iron At less than molecular concentrations in the growth medium, the systems are usually fully induced whereas at iron levels of 10 um or greater, all components of the siderophore are barely detectable. Biochemists now purport to understand, at the molecular level, how this iron pump is regulated. All strains of a genetically accessible bacterium, Escherichia coli, form a siderophore called enterobactin, a catechol derivative. Clinical isolates of E. coli synthesize enterobactin and a second siderophore, a hydroxamic acid, called aerobactin. The latter is an important component of the virulence mechanism of hospital isolates of E. coli. Cloning of the aerobactin gene cluster was possible because receptors for ferric siderophores are frequently also found to serve as receptors for lethal agents such as bacteriocins (“killer proteins”). One such bacteriocin—cloacin—uses the same receptor as ferric aerobactin. The acquisition of sensitivity to cloacin could be used to select for cells of E. coli which had been successfully transformed with DNA coding for the ferric aerobactin receptor protein. Whereas the genetic determinants for enterobactin were known to be spread over more than 20 kb of DNA and organized into several transcriptional units, the cloned aerobactin genes turned out to be restricted to a single cluster comprised of four genes for biosynthesis of the siderophore and one for transport of its ferric complex. Characterization of the promoter driving the aerobactin cluster revealed it to be of the typical E. coli “strong” variety. The iron-sensing region was shown to be confined to a particular site near where RNA polymerase binds and initiates transcription. Although iron, whether ferric or ferrous, is generally six-coordinate, it is still scarcely conceivable that the naked metal ion could be sufficiently “semantic,” or structurally complex, to achieve sequence-specific binding and hence regulation of transcription of the aerobactin cluster. Ten years ago a mutation was described in Salmonella typhimurium, a close relative of E. coli, which resulted in a failure of iron to regulate siderophore synthesis; The mutation was designation fur, for ferric uptake regulation. The duplication of this constitutive type mutation in E. coli led to identification of the product of the fur gene as a 17 kDa protein rich in histidine. By fusion of a gene for a reporter enzyme into the aerobactin cluster, it was demonstrated that the Furprotein required Fe(II) for recognition of its operator site. In fact, a number of other divalent heavy metal ions such as those of Mn, Co, Cu, and Cd could, with varying degrees of efficiency, substitute for Fe(II) in in vitro tests. By use of a standard “footprinting” analysis in which the isolated Fur protein was used to protect the DNA against enzymatic digestion, the actual bases contacted by the repressor could be identified. A comparison of this array with the upstream regions of several akeady sequenced genes in E. coli known to be iron regulated afforded a consensus “iron box” operator recognition site. Application of the same type of analysis to the upstream section of the gene for the Fur protein disclosed a relatively weak “iron box” overlapping the polymerase binding site and, farther upstream, a functional CAP or cyclic AMP--protein binding, site. From these data it appears that the siderophore-dependent iron assimilation system of E. coli operates the following way. When internal iron, which must be a least in part simple hexa-aquo iron (II), is at a low level, then RNA polymerase can freely transcibe the siderophore biosynthetic and transport genes which, in turn, results in rapid accumulation of internal iron. The metal is released from the siderophore by reduction and, in some cases, by hydrolysis of the ferri-siderophore complex. The reduced metal ion then binds to the Fur protein which, like most repressors, interacts as a dimer with the sequence specific “iron box” operator to negatively block transcription of the complete iron uptake apparatus. The virtual absence of free iron in host tissue apparently acts as an environmental signal enabling invasive strains of E. coli to express freely the aerobactin cluster, the products of which allow the cells to scavenge iron from transferrin. Like siderophores, bacterial toxins are often only excreted under conditions of low iron growth and, in one case, that of the Shiga-like toxin of E. coli, an “iron box” has been shown to occur in the operator. The efficiency with which Mn(ll) can substitute for Fe(II) as an activator of the Fur protein may account for the ability of an excess of the former metal to cause anemia in experimental animals and ehlorosis (iron deficiency) in plants. However, it has not yet been established that the repressor-mediated negative control system just described for E. coli applies to higher forms of life. Nonetheless, it has been known from the early work of McCance and Widdowson that all animal species that are replete with respect to iron do not uptake more of the mineral. As if to underline the confounding complexity of regulatory circuits, even in a genetically “simple” organism such as E. coli, all fur mutants isolated to date fail to grow on certain carbon sources such as acetate, succinate, and fumarate. A function loosely mapped near the fur locus on the E. coli chromosome is involved in dicarboxylic acid transport, but its relation,if any,to the fur gene has not been established. At a recent UCLA Symposium on Molecular and Cellular Biology held April 4-10 in Frisco, Cob., E. Neiderhoffer, C. Naranjo, and J. A. Fee reported (Journal of Cellular Biochemistry, Supplement, 12D, 340, 1988) that fur mutants of E. coli produce none of the iron form of superoxide dismutase. This is surprising since the gene for the manganese form of the enzyme, which is also present in E. coli, has a strategically located and quite reasonable “iron box” consensus sequence. Apparently it is necessary to perform footprinting and gel shift assays in order to confirm repressor binding and to obtain an estimate of the affinity of the regulatory protein for the DNA. In the case of the aerobactin cluster the very small dissociation constant for the Fur-metal-operator complex, of the order of nanomolar, strongly suggests that the proposed control mechanism is operative in viva. It now remains to ascertain the extent to which the regulatory system for iron assimilation in E. coli can be applied to other microorganisms and, ultimately, to plants and animals. Beyond this task and still open to exploration are the uptake systems for the several other bio-essential mineral elements. J.B. Neilands is a professor of biochemistry at the University of California, Berkeley

Nailing Down Iron Assimilation

Interested in reading more?

Become a Member of