In this report we demonstrate that prion infection induces a state of iron imbalance in diseased human, hamster, and mouse brains as revealed by a phenotype of iron deficiency in the presence of increased total iron. The phenotype of iron deficiency appears early in the disease process, probably with the initiation of PrPSc formation, and is not an outcome of end-stage disease. An important underlying cause of iron imbalance is sequestration of cellular iron in detergent insoluble PrPSc-ferritin complexes, thus rendering it bio-unavailable and creating a phenotype of apparent iron deficiency. Consequent up-regulation of iron uptake proteins Tf and TfR result in increased cellular iron uptake, creating a state of cellular iron imbalance. Since iron is potentially toxic due to its redox-active nature, these observations have significant implications for prion disease associated neurotoxicity.
In evaluating iron homeostasis in healthy and diseased brains, we have focused primarily on total brain iron and levels of Tf and TfR. Although these parameters do not represent changes in iron levels in individual cell types that comprise the complex milieu of the brain, these are reliable markers of neuronal iron levels, the principal cell type affected in prion disorders 
. The mechanisms underlying brain iron homeostasis are complex, and details of iron import and export from various cells within the brain are still emerging. Normally, iron enters the brain through the Tf/TfR pathway across capillary endothelial cells, and is delivered to the brain interstitium using mechanisms that are still debated. Once in the brain, iron is bound by citrate, ascorbate, and Tf present in the interstitial fluid. Citrate and ascorbate are mainly released by astrocytes, and brain Tf is derived from cells of the choroid plexus and oligodendrocytes and is saturated with iron under normal conditions. Neurons take up most of their iron from Tf through TfR mediated uptake. Under conditions of iron deficiency, neurons up-regulate TfR levels to internalize increased amounts of iron saturated Tf to make up for the deficiency. The opposite scenario holds true when iron is in excess. In contrast to other cell types such as astrocytes, oligodendrocytes, and microglia, neurons in most brain regions lack the iron storage protein ferritin, necessitating efflux of excess iron through the iron export protein ferroportin. Astrocytes, microglia, and oligodendrocytes lack TfR, and take up most of their iron through other mechanisms 
. Thus, TfR levels in the brain reflect the iron status of neurons, the most vulnerable cell population in prion disorders. However, experimental manipulation of brain tissue often results in the degradation of TfR, providing inconsistent results from post-mortem brains and animal brains unless the tissue is handled with extreme care. Tf, on the other hand, is as resilient to degradation as PrPSc
, and reflects TfR levels as demonstrated by our immunohistochemistry and immunoblot data from animal and human brains. It is for these reasons that we have focused on Tf and iron levels, and where possible, TfR levels to evaluate brain iron status in prion disease affected brains.
Using the above criteria, our data demonstrate the presence of iron imbalance in prion disease affected brains from three different species, human, hamster, and mice. Although alteration of brain iron has been described in other neurodegenerative diseases such as AD and PD, it is notable that the changes in iron homeostasis in prion disease affected brains are distinct 
. A prominent difference is the apparent phenotype of iron deficiency in prion disease affected brains despite increased brain iron content. A direct correlation between PrP and Tf levels in both sCJD and infected hamster samples suggests that iron deficiency arises as a direct consequence of PrPSc
accumulation, perhaps by forming PrPSc
-ferritin complexes as noted in scrapie infected cells and reported earlier 
. An increase in Tf and TfR levels is also noted in human biopsy samples, indicating that brain iron levels decrease much before the onset of end-stage pathology. The presentation of AD and PD brains, however, is strikingly different. In the pool of CJD− cases examined in this report, several samples contained high levels of iron, including cases of AD, PD, dementia of unknown origin, and vascular disorders. As opposed to prion disease affected brains, these cases demonstrated a consistent down-regulation of Tf and TfR as expected, suggesting that the increased iron content was probably derived from insoluble deposits in the extra-cellular space. Brains from sCJD cases, on the other hand, show a modest and a surprisingly similar increase in iron content across twenty samples and a significant increase in Tf levels, suggesting a common pathway of iron accumulation and sequestration by scrapie infection.
The data on scrapie infected hamsters support the above conclusions. In these animals, iron deficiency appears almost at the same time as PK-resistant PrPSc
, and worsens with disease progression as indicated by increasing levels of Tf and TfR till end stage disease. Since this analysis was performed on experimental brains that were snap frozen after harvesting, the increase in TfR levels with disease progression provides a direct measure of neuronal iron deficiency, further supported by an increase in Tf that is likely to reflect TfR levels as observed in the Purkinje cell neurons of CJD+ samples. Since most neurons lack ferritin iron stores 
, such a situation is likely to result in compromised neuronal health and toxicity. It is unlikely that advanced age contributes to the iron imbalance observed in prion disease affected brains since the hamster brains used for this evaluation range in age from ten to sixteen weeks, a relatively young age considering the life-span of hamsters. It is interesting to note that the extent of iron accumulation varies with the strain of prions used for infecting the experimental animal. Hamsters infected with TME accumulate much higher levels of total and redox-active iron than those infected with the 263K strain of scrapie (unpublished observations). However, the increase in Tf levels is similar in both cases, suggesting that iron in some cases precipitates in the extra-cellular space and does not alter the intracellular iron status.
The most likely cause of iron deficiency in prion disease affected brains is the sequestration of iron in PrPSc
-ferritin aggregates as noted in scrapie infected cells and sCJD brain tissue. In a previous report we demonstrated that PrPSc
from sCJD and scrapie infected mouse tissue forms a PK-resistant, detergent insoluble complex with ferritin 
. Here we show that ferritin from sCJD affected brains sequesters iron, and is not denatured by boiling in 2% SDS for ten minutes. Ferritin from normal human brains is denatured by this treatment and releases bound iron. A similar observation is noted in scrapie infected cell lines, though cellular ferritin is less resistant to SDS. These observations suggest that ferritin co-aggregates with PrPSc
to form a detergent insoluble complex and sequesters the associated iron, creating a state of iron deficiency. Such an occurrence would explain the development of iron deficiency with the appearance of PrPSc
, increased levels of total iron in diseased brains, a direct correlation between PrP and Tf levels, and an overall state of iron imbalance in diseased brains. Although ferritin is mainly cytosolic, it is known to undergo degradation within the lysosomes, the cellular compartment where PrPSc
and ferritin are observed in scrapie infected cells, and are hence likely to form the complex 
To summarize, this report demonstrates that brain iron imbalance is a common feature of prion disease affected brains, and is likely to contribute to prion disease associated neurotoxicity due to the redox-active nature of iron. The underlying cause of iron imbalance is the formation of PrPSc-ferritin complexes, a pathogenic process that may lend itself to therapeutic manipulation, thus providing a means to reduce disease associated neurotoxicity.