A brief history of brain iron research

Abstract

Brain iron research began in the second half of the 19th century when Perls [1] described the histochemical visualization of ‘‘iron oxide’’ in tissues by exposure of sections to a mixture of potassium ferrocyanide (1%) and hydrochloric acid (1%). The reaction causes an alluring blue (‘‘Prussian Blue’’) and continues to be the main method for the staining of iron in tissues everywhere. It is simple and quite specific but lacks sensitivity, especially with paraffin-embedded tissues. Enhancement by subsequent incubation in a mixture of diaminobenzidine and hydrogen peroxide greatly improves sensitivity [2] and reveals reaction product in much greater detail, especially in optimally fixed brain [3]. The distinction of ‘‘heme-’’ and ‘‘non-heme-’’ iron in the brain must be attributed to Zaleski [4] who observed that the iron in hemoglobin did not react with potassium ferrocyanide, hence the blue color had to arise from non-heme iron. He also concluded that brain iron was present in linkage to a protein and that it was largely ferric rather than ferrous. This discovery antedated the recognition of ferritin as the main iron-carrying protein in brain by 69 years [5]. Guizzetti [6] and Spatz [7] used immersion of fresh or fixed animal and human brain slices in Perls’ solution to visualize iron in various gray and white matter structures of the brain. The long paper by Spatz [7] offered the first color illustrations of the unusual differential distribution of iron. The most intense macrostaining occurred in the globus pallidus, the red nucleus, and the substantia nigra (including the melanin-free pars reticularis), followed by the caudate nucleus, the putamen, the subthalamic nucleus, and the dentate nucleus of the cerebellum. The reaction was less intense in the thalamus but was still stronger than in the cerebral cortex and the subcortical white matter. In vibratome or frozen sections of normal rat and rabbit brain, the enhanced iron stain [2] shows iron in microglia, oligodendroglia, astrocytes (especially the tanycytes), nerve cells, and possibly even axons and myelin sheaths. The staining of oligodendroglia in the cerebral white matter is not uniform. Reaction product occurs in clusters of ironreactive cells, but the significance of this distribution is not known. Iron is readily lost from sections unless the tissue is fixed promptly by aldehydes or ethanol. One of the reasons is the high solubility of ferritin in aqueous solvents. As in other organs, brain ferritin contains light (L-) and heavy (H-) subunits, and its biosynthesis is subject to translational control by the iron-regulatory proteins 1 and 2 (IRP) [8– 10]. Ferritin is readily detected by immunocytochemistry or immunofluorescence. The distribution is quite similar to that of iron when detected by histochemistry with Perls’ [1] ingredients. Reaction product for IRP-1 (the ‘‘ferritin repressor protein’’ in Ref. [8]) occurs mostly in astrocytes [11]. Bloch et al. [12] showed that brain contains transferrin and transferrin messenger ribonucleic acid (mRNA). Antisera to transferrin yield reaction product that does not coincide with that of iron or ferritin [13], and there is considerable variability among species [14]. However, transferrin is abundant in the oligodendroglial cytoplasm of many mammals and has become a reliable biochemical and immunocytochemical marker for development and pathology of these cells. In addition to transferrin, brain also contains transferrin receptor protein (TfR) [15]. Antisera to TfR reveal reaction product in the endothelium where the protein may serve in iron uptake. However, the mechanism of brain iron homeostasis and exchange is likely much more complex than the interaction of transferrin and TfR on the surface of endothelial cells. In recent years, systematic investigations of intestinal iron absorption and hemochromatosis have led to the characterization of several proteins that accomplish iron transport across membranes of cells and organelles. While there is no universal agreement on their names, ‘‘divalent metal transporter 1’’ (DMT1) (syn. divalent cation transporter 1 [DCT1]; natural resistance-associated macrophage protein 2 [NRAMP2]) and ferroportin (syn. metal transporter protein 1 [MTP1], ironregulated transporter 1 [IREG1]) are now widely recognized (brief review in Ref. [16]). Immunocytochemistry reveals a wide distribution of DMT1 in neurons, glial cells, and the