Despite its importance in iron-overload diseases, little is known about the composition of plasma non-transferrin-bound iron (NTBI). Using 30-kDa ultrafiltration, plasma from thalassemic patients consisted of both filterable and non-filterable NTBI, the filterable fraction representing less than 10% NTBI. Low filterability could result from protein binding or NTBI species exceeding 30 kDa. The properties of iron citrate and its interaction with albumin were therefore investigated, as these represent likely NTBI species. Iron permeated 5- or 12-kDa ultrafiltration units completely when complexes were freshly prepared and citrate exceeded iron by tenfold, whereas with 30-kDa ultrafiltration units, permeation approached 100% at all molar ratios. A g = 4.3 electron paramagnetic resonance signal, characteristic of mononuclear iron, was detectable only with iron-to-citrate ratios above 1:100. The ability of both desferrioxamine and 1,2-dimethyl-3-hydroxypyridin-4-one to chelate iron in iron citrate complexes also increased with increasing ratios of citrate to iron. Incremental molar excesses of citrate thus favour the progressive appearance of chelatable lower molecular weight iron oligomers, dimers and ultimately monomers. Filtration of iron citrate in the presence of albumin showed substantial binding to albumin across a wide range of iron-to-citrate ratios and also increased accessibility of iron to chelators, reflecting a shift towards smaller oligomeric species. However, in vitro experiments using immunodepletion or absorption of albumin to Cibacron blue-Sepharose indicate that iron is only loosely bound in iron citrate-albumin complexes and that NTBI is unlikely to be albumin-bound to any significant extent in thalassemic sera.

Despite the availability of deferoxamine (DFO) for more than three decades, its rates of interaction with cellular iron pools in different tissues, and the effects of its pharmacokinetics on the interaction with plasma iron pools, remain incompletely understood. The positive charge of DFO, together with the negative resting potential in vertebrate cells, favors cellular uptake, whereas the low lipophilicity and high molecular weight counter this effect. The findings presented suggest a facilitated uptake of DFO into hepatocytes, being several hundred-fold faster than into red cells. Antibodies that selectively recognize ferrioxamine (FO) show that initial hepatocellular iron chelation is cytosolic, but later transposes to lysosomal and ultimately canalicular compartments. Strong FO staining is visible in myocytes within 4-8 h after commencing a subcutaneous DFO infusion, indicating effective chelation of myocyte iron. A methodology was developed to study the interaction of DFO and its metabolites with plasma iron pools by stabilizing DFO with aluminum ions, thereby preventing iron shuttling from non-transferrin-bound iron (NTBI) onto DFO after plasma collection. DFO removes only about a third of NTBI rapidly, and NTBI is rarely cleared completely. Increasing DFO dosing does not increase NTBI removal, but instead leads to a greater rebound in NTBI on cessation of intravenous infusion. Thus, intermittent infusions of high-dose DFO are less desirable than continuous infusions at low doses, particularly in high-risk patients. Here the benefits of continuous DFO on heart function occur before changes in T2*-visible storage iron, consistent with early removal of a toxic labile iron pool within myocytes.

The causes and consequences of ageing are likely to be complex and involve the interaction of many processes. It has been proposed that the decline in mitochondrial function caused by the accumulation of oxidatively damaged molecules plays a significant role in the ageing process. In agreement with previous reports we have shown that the activities of NADH CoQ1 reductase and cytochrome oxidase declined with increasing age in both rat liver and gastrocnemius muscle mitochondria. However, only in the liver were the changes in lipid peroxidation and membrane fluidity suggestive of an age-related increase in oxidative stress. After 12 weeks on a vitamin E deficient diet, vitamin E levels were undetectable in both gastrocnemius muscle and liver. In skeletal muscle, this was associated with a statistically significant increase in lipid peroxidation, a decrease in cytochrome oxidase activity after 48 weeks, and an exacerbation in the age-related rate of decline of NADH CoQ1 reductase activity. This was consistent with the suggestion that an imbalance between free radical generation and antioxidant defence may contribute to the mitochondrial dysfunction with age. In contrast to this, vitamin E deficiency in the liver caused a significant increase in mitochondrial respiratory chain activities with increasing age despite evidence of increased lipid peroxidation. Comparison of other features in these samples suggested vitamin E deficiency; did not have a significant impact upon mtDNA translation; induced a compensatory increase in glutathione levels in muscle, which was less marked in the liver, but probably most interestingly caused a significant decrease in the mitochondrial membrane fluidity in muscle but not in liver mitochondria. These data suggest that while increased lipid peroxidation exacerbated the age-related decline in muscle respiratory chain function this relationship was not observed in liver. Consequently other factors are likely to be contributing to the age-related decline in mitochondrial function and specific stimuli may influence or even reverse these age-related effects as observed with vitamin E deficiency in the liver.

Rats fed a vitamin E-depleted diet for 48 weeks had undetectable levels of vitamin E in the gastrocnemius muscle and liver, leading to elevated malondialdehyde levels in both tissues and an elevated GSH level in muscle. Skeletal-muscle mitochondria showed decreased mitochondrial respiratory chain (MRC) activities, whereas liver MRC activities were increased. Exposure of normal rat liver submitochondrial particles (SMPs) to an in vitro NADPH-dependent lipid peroxidation system resulted in a dose-dependent increase in lipid peroxidation and inhibition of complex I and complex IV activities. Complex I exhibited greater sensitivity to lipid peroxidation than complex IV. At low and high NADPH concentrations, the rate of lipid peroxidation and the level of enzyme inhibition were essentially the same in liver SMPs from both vitamin E-deficient and control rats, suggesting that under these conditions, the loss of vitamin E did not exacerbate the effects of either lipid peroxidation or enzyme inhibition. These results indicate that normal vitamin E levels in liver mitochondria are not required for protection against lipid peroxidation and are consistent with the normal liver mitochondrial function in vitamin E-deficient animals. This suggests other antioxidants, such as ubiquinol and GSH, may be more important in protecting liver mitochondria and MRC from lipid peroxidation.