Detection of iron-containing proteins contributing to the cellular labile iron pool by a native electrophoresis metal blotting technique

Abstract

The labile iron pool (LIP) plays a role in generation of free radicals and is thus the target of chelators used for the treatment of iron overload. We have previously shown that the LIP is bound mostly to high molecular weight carriers (MW>5000). However, the iron does not remain associated with these proteins during native gel electrophoresis. In this study we describe a new method to reconstruct the interaction of iron with iron-binding proteins. Proteins were separated by native gradient polyacrylamide gel electrophoresis and transfered to polyvinilidene difluoride membrane under native conditions. The immobilized iron-binding proteins are then labeled by ⁵⁹Fe using a ‘titrational blotting’ technique and visualized by storage phosphorimaging. At least six proteins, in additon to ferritin and transferrin, are specifically labeled in cellular lysates of human erythroleukemic cells. This technique enables separation and detection of iron-binding proteins or other metal–protein complexes under near-physiological conditions and facilitates identification of weak iron–protein complexes. Using a new native metal blotting method, we have confirmed that specific high molecular weight proteins bind the labile iron pool.

Introduction

Iron plays vital catalytic and structural roles in numerous metalloproteins [1]. Under physiological conditions, ferric iron is highly insoluble and rapidly autooxidizes to ferrous iron, catalyzing the formation of highly damaging oxygen radicals [2], [3]. Thus, efficient homeostatic mechanisms have developed to tightly regulate iron uptake, transport and storage. Under physiological conditions iron can not exist as a ‘free’ ion and is coordinated to carrier ligands. Most organic moieties, including hydroxy, carboxy, keto, alkyl and nitrogen groups, bind iron to some extent. In mammalian cells, the proportion of total cellular iron was shown to be extractable by chelatation and constitutes the ‘labile iron pool’ (LIP). Low molecular weight cytosolic compounds have been postulated to bind the LIP and mediate specific transfer of iron among various proteins [4], [5], but have so far not been specifically identified [6].

Identification of the source of the intracellular labile iron pool (LIP) has important consequences. Chelators are used to treat iron overload target the LIP [7], [8] and are being tested as cytotoxic, antimalaric and antiinflammatory drugs [7], [8], [9], [10]. Iron containing enzymes such as 5-, 12- and 15-lipoxygenases, prolyl-4-hydroxylase, lysine-4-hydroxylase, deoxyhyposyl hydroxylase, phenylalanine hydroxylase, tyrosine hydroxylase, tryptophan hydroxylase and ribonucleotide reductase are directly inhibited by iron chelators [11]. Similarly, the mechanism of action of some clinically used cytotoxic agents (anthracyclines and hydroxyurea) may be chelation of iron, resulting in the inhibition of iron containing enzymes [12]. Despite the physiological and medical significance of the LIP, the biochemical properties and cellular function(s) of LIP remain obscure.

We have found, using native gradient gel electrophoresis of erythroid cells metabolically labeled with ⁵⁹Fe, that LIP is bound predominantly to high molecular weight carriers (MW>5000) [13], [14]. In contrast with the commonly accepted view that LIP is an ‘intermediate’ or ‘transient’ pool of iron, we demonstrated that the LIP may be an end product, since the amount of iron in this pool increases linearly over time in cells exposed to a short pulse of transferrin iron. The only iron-binding cellular constituent with kinetic properties of an early intermediate we identified was a multiprotein membrane complex containing transferrin receptor, transferrin and additional proteins [13].

During native electrophoretic separation of cellular lysates, weakly bound iron is stripped from its original (non-ferritin, non-transferrin) carriers by Tris and glycine, two chelating agents used in electrophoresis. This causes the LIP iron pool to migrate as a diffuse band at the front of electrophoretograms. When desferrioxamine or 1,2-dimethyl-3-hydroxypyrid-4-one (L1) chelators are added to the lysates prior to the electrophoresis, the LIP iron disappeared from electrophoretograms due to inability of the desferrioxamine and L1 iron complexes to enter the gel [15], [16].

Previously, other researchers have utilized a ‘metal blotting’ technique to identify metal binding proteins separated by denaturing SDS electrophoresis. This consisted of transferring proteins to a support membrane and probing with a radioactive metal isotope. Such metal blotting techniques have been used to detect manganese, calcium, cadmium, zinc [17], [18], [19], [20] and iron-binding proteins [21], [22], [23]. The physiological relevance of these studies is questionable because sodium dodecyl sulfate (SDS) electrophoresis denatures the separated proteins. It has been proposed that proteins partially or completely renature upon transfer to immobilized membranes. However, this has not been demonstrated. Furthermore, the presence of metal binding moieties on the vast majority of proteins confound whether this metal binding represents the actual in vivo situation.

We have developed a ‘native metal blotting’ procedure to circumvent the denaturing electrophoretic approach used in previous studies. We used native, nondenaturing electrophoresis of cellular lysates, to identify iron carriers contributing to LIP in K562 human erythroleukemic cells. The polyacrylamide gels were blotted to polyvinilidene difluoride membranes under non-denaturing conditions and subsequently exposed to a ⁵⁹Fe metal labeling solution. In this way, macromolecules binding iron even with low affinity can be visualized under near-native conditions.