Evidence for aluminum-binding erythropoietin by size-exclusion chromatography coupled to electrothermal absorption atomic spectrometry

Erythropoietin (EPO) is a glycoprotein that stimulates erythropoiesis and is clinically used for treating anemia during chronic renal failure and for anemia in preterm infants. EPO formulations usually have elevated rates of contamination due to aluminum (Al), which is toxic to both types of patients. Size-exclusion chromatography (SEC) coupled with graphite furnace atomic absorption spectrometry (GF AAS) was employed to separate proteins and to quantify the amount of aluminum present in the elution volume corresponding to EPO and, therefore, to evaluate possible binding. Because EPO formulations contain human serum albumin (HSA), a chromatographic method was optimized for the separation of these proteins. Subsequent to the chromatographic separation, 1-mL fractions of the column effluent were collected, and the Al content in these aliquots was measured by GF AAS. EPO and HSA samples were incubated with Al for 4h at 4°C and 37°C as well as for 16 h at 4°C and 37°C. Afterwards, they were injected into the chromatographic system. These samples were also submitted to ultrafiltration (10 and 50 kDa membranes), and Al was measured in the ultrafiltrates. The results showed that Al was present in the eluent volume corresponding to the EPO peak but not in the HSA peak in the chromatograms. Temperature strengthened the interaction because the Al present in the EPO fraction was 3 times higher at 37°C compared to 4°C. Thirty-eight percent of the Al present in a 2.4 μg/mL EPO standard solution, and approximately 50% of the Al in formulation samples containing approximately 11 μg/mL EPO and either citrate or phosphate, were non-ultrafiltrable, which suggests that EPO is an effective Al acceptor in vitro.

Al and Si: their speciation, distribution, and toxicity

OBJECTIVES

In dialysis patients both aluminum (AI) and silicon (Si) may accumulate. Whereas the toxic effects of AI within this population are clearly established, little is known on the role of Si in the development/protection of particular dialysis-related diseases. A clear insight in the protein binding and speciation of trace elements is important to better understand the mechanisms underlying their toxicity/essentiality. Research in this field however is complex and often prone to analytical difficulties and inaccuracies.

DESIGN AND METHODS

In the first part of this review techniques used for speciation studies of AI and Si in biological fluids are discussed. Notwithstanding recent technical advances (a) extraneous metal contamination, (b) unrecognized aspecific binding of metals to proteins, and (c) unwanted interactions with separation equipment such as chromatography columns and ultrafiltration membranes remain important pitfalls and often lead to erroneous conclusions. The factors that determine the speciation of AI and Si and their ultimate tissue distribution and toxicity are dealt with in the second part. Here, experimental data obtained with various speciation techniques are linked to in vivo data on the tissue distribution, localization/toxicity of both elements.

CONCLUSIONS

A model in which the AI tissue distribution/toxicity is mediated by either its citrate or transferrin bound form is proposed.

Low percentage of aluminoxamine and ferrioxamine in uremic serum after desferrioxamine administration

HPLC was used to study the effectiveness of two different desferrioxamine (DFO) administration strategies (15 mg/kg DFO, 1 h or 44 h before dialysis) on generation of aluminoxamine and ferrioxamine in five hemodialysis patients. The percentage of ultrafilterable aluminum and iron in these patients was also investigated by electrothermal atomic absorption spectrometry. The administration of DFO in both schemes increased the ultrafilterable serum aluminum concentrations from a mean of 17.1 ± 1.6% to a mean of 75.7 ± 14.1%. However, 1 h after DFO infusion, only 38.8 ± 7.7% of the total serum aluminum was bound to DFO; 44 h after DFO infusion, only 15.8 ± 8.0% was bound. Similar results were obtained for ferrioxamine. These results suggest that the ultrafilterable serum fraction contains aluminum and iron chelated by DFO and by DFO metabolites, which retain similar metal-chelating abilities.

Aluminum exposure and metabolism

Aluminum (Al) is a nonessential, toxic metal to which humans are frequently exposed. Oral exposure to aluminum occurs through ingestion of aluminum-containing pharmaceuticals and to a lesser extent foods and water. Parenteral exposure to aluminum can occur via contaminated total parenteral nutrition (TPN), intravenous (i.v.) solutions, or contaminated dialysates. Inhalation exposure may be important in some occupational settings. The gut is the most effective organ in preventing tissue aluminum accumulation after oral exposure. Typically gastrointestinal absorption of aluminum from diets is < 1%. Although the mechanisms of aluminum absorption have not been elucidated, both passive and active transcellular processes and paracellular transport are believed to occur. Aluminum and calcium may share some absorptive pathways. Aluminum absorption is also affected by the speciation of aluminum and a variety of other substances, including citrate, in the gut milieu. Not all absorbed or parenterally delivered aluminum is excreted in urine. Low glomerular filtration of aluminum reflects that most aluminum in plasma is nonfiltrable because of complexation to proteins, predominantly transferrin. The importance of biliary secretion of aluminum is debatable and the mechanism(s) is poorly understood and appears to be saturable by fairly low oral doses of aluminum.

Spectroscopic study of the interaction of aluminum ions with human transferrin

Transferrin is the plasma protein responsible for transporting Fe3+ from the absorption to the utilization site. Interactions of apo- and holo-transferrin with Al3+ were studied by circular dichroism (CD), UV-visible, and fluorescence spectrometry. Binding of Al3+ to both metal-ion binding sites of apo-transferrin was confirmed by fluorescence studies. No interaction of Al3+ with holo-transferrin was observed, indicating that Al3+ cannot displace Fe3+ under the experimental conditions employed. An increase in tryptophan fluorescence (lambda max at 330 nm) by excitation at either 280 or 295 nm was observed after Al3+ interaction with apo-transferrin. There was no shift in wavelength of the fluorescence band of apo-transferrin after interaction with Al3+, but the intensity did increase. Since excitation at 295 nm is specific for tryptophan residues, tryptophan but not tyrosine must be responsible for the change in fluorescence intensity. Decreased fluorescence is the result of Fe3+ binding to apo-transferrin. The CD spectrum of apo-transferrin was slightly affected in the far UV by Al3+ binding, but a salient change was noted in the near UV at approximately 288 nm where tyrosine and tryptophan absorb. It is concluded that a small conformational change in the protein was induced by Al3+ binding to apo-transferrin.

Aluminium and silicon speciation in human serum by lon-exchange high-performance liquid chromatography-electrothermal atomic absorption spectrometry and gel electrophoresis

Speciation of aluminium and silicon in serum was studied by a reliable and sensitive high-performance liquid chromatographic-electrothermal atomic absorption spectrometric (HPLC-ETAAS) hybrid method, based on the use of a polymeric anion-exchange column (Protein-Pak DEAE-5PW). This polymer-based column minimizes the risk of aluminium losses and of silicon contamination from the column during separation. The results obtained were compared with the results of previous studies carried out using different, complementary techniques including ultramicrofiltration, gel filtration and silica-based column for HPLC. In order to ascertain which protein(s) of serum actually bind(s) aluminium, gel electrophoresis was employed for the further separation of the column fractions obtained by HPLC and aluminium was determined in separate aliquots of the same fractions. From all the experiments, it appears that transferrin (Tf) is the only serum protein that binds aluminium and it contains about 90% of total serum aluminium. It was also confirmed that in the presence of desferrioxamine (DFO). aluminium is partly displaced from its complex with transferrin to a low molecular mass AL-DFO complex. Aluminum citrate seems to be the main low molecular mass aluminium species in serum, amounting to about (12 +/- 5% of the total aluminium in an aluminium-loaded serum sample. The proposed speciation procedure permits the simultaneous identification and determination of three aluminium species in metal-spiked serum (Al-Tf, Al-DFO and AI-citrate). The result for silicon suggest that it seems to be unspecifically adsorbed to several serum proteins and its speciation is not affected by the presence of DFO.(ABSTRACT TRUNCATED AT 250 WORDS)

Aluminum exposure and excretion

Occupational exposure to aluminum can be associated with increases in both urinary aluminum excretion and serum aluminum. In most studies, the increases in urinary aluminum are proportionately greater than the changes in serum aluminum. A similar pattern of response follows increases in dietary aluminum intake. Thus, there is ample evidence for systemic aluminum absorption from occupational exposure to airborne aluminum as well as dietary intake. Although both circumstances are accompanied by a similar renal response, there is little information explaining how normal kidneys augment renal excretion with only trivial changes in serum aluminum concentrations. In addition, it is not understood how airborne exposure to microgram amounts of aluminum produces significant increases in urinary aluminum. The latter observation suggests the presence of a sensitive uptake process for aluminum from airway exposure.

Binding of aluminium to plasma proteins: comparative effect of desferrioxamine and deferiprone (L1)

Ultramicrofiltration techniques were used to study both the binding of aluminium to high molecular weight proteins in the presence of different concentrations of desferrioxamine and deferiprone (L1) and the kinetics of aluminium release from human serum proteins. Human serum from healthy volunteers was used in all studies. The serum was spiked with aluminium (100 micrograms/l) and different concentrations of chelators. Ultramicrofiltration was performed with Amicon YMT membranes which had a nominal cut-off of 30,000 Da. Aluminium was measured by graphite furnace atomic absorption spectrometry in total serum and ultrafiltered fluid. Deferiprone shows a higher capability to displace aluminium from serum proteins (80%) than desferrioxamine (60%) at equivalent concentrations of the chelators. The kinetics of the release were also faster for deferiprone, taking 20 min to achieve its maximum effect, whereas, desferrioxamine achieved only 80% of its maximum effect after 2 h. Thus, deferiprone could be an attractive alternative to desferrioxamine, as an aluminium chelator agent.

Hematological effects of aluminum on living organisms

1. Aluminum has been of great interest for many researchers over a number of years; its biochemical and physiological role is not yet fully clear. There are few papers describing the hematological consequences of its excess in living organisms and most of their data are cited in this paper. 2. Aluminum reduced the deformability of erythrocytes, and such cells are rather frequently retained in the reticuloendothelial system of the spleen and eliminated faster from the blood stream. 3. Aluminum produces peroxidative changes in the erythrocytes membrane, leading to hemolysis. Therefore, the depressed erythrocyte count in animals intoxicated with aluminum may be the consequence of both the hemolytic action of aluminum and the shortened time of survival of erythrocytes. 4. It was demonstrated that aluminum inhibits heme biosynthesis in vitro. This problem requires, however, further studies and observation. 5. Changes occurring under the influence of Al3+ on the leukocyte system of animals suggest the influence of this element on the resistance of the organism, but the mechanism of the action of Al3+ still requires elucidation. 6. Cell metabolism including blood cells may be affected by aluminum in many ways, the more so as the element may combine in vitro with amino acids, peptides, proteins, enzymes, substrates, cofactors, nucleotides and carbohydrates. Aluminum stimulates NADPH oxidation and takes part in the process of free radical formation.

Transferrin receptors of rat and human brain and cerebral microvessels and their status in Alzheimer's disease

We studied the regional distribution of specific [125I]transferrin binding to transferrin receptors in the brains and cerebral microvessels of humans and rats. We also assessed transferrin receptors in subjects with Alzheimer's disease. Human diferric [125I]transferrin bound to regional brain and cerebral microvessels with high affinity (dissociation constants of 1-10 nM), and the maximal binding densities ranged from 30 to 90 pmol/mg protein in the brain and were several-fold higher in cerebral microvessels. In Alzheimer's disease, transferrin receptor densities were significantly reduced in the hippocampus and the temporal and occipital cortex but were unchanged in the frontal and parietal cortex and the cerebellum. Although [125I]transferrin binding was higher in cerebral microvessels from subjects with Alzheimer's disease than in those of age-matched controls, this difference did not attain statistical significance. These results suggest that transferrin receptor density was decreased in some cortical areas including the hippocampus in Alzheimer's disease but relatively unchanged in cerebral microvessels.

Aluminum-binding serum proteins: desferrioxamine alters serum aluminum speciation

The aluminum content of four size classes of protein (high and low molecular weight, transferrin/albumin and a fraction provisionally termed albindin) in sera from healthy volunteers (group I) and from aluminum workers with normal (group II) and high (group III) total serum aluminum was compared using size exclusion chromatography and electrothermal atomic absorption spectroscopy. In the absence of any drug treatment the transferrin/albumin fraction was the major carrier, containing 29% to 33% of the aluminum recovered, in all three subject groups. Desferrioxamine treatment of groups II and III significantly decreased the proportion of aluminum bound by albumin/transferrin (P less than 0.05 in group III) and increased that bound by albindin (P less than 0.05 in groups II and III). The albindin fraction contained over 40% of the aluminum recovered from sera of group III subjects during desferrioxamine treatment. We conclude that the albindin fraction contains a protein or proteins that can form stable complexes with aluminum which may be important in preventing aluminum toxicity.

High performance liquid chromatography methods for studying protein binding of aluminium in human serum in the absence and in the presence of desferrioxamine

Ion-exchange high performance liquid chromatography of serum proteins was combined with aluminium determination by electrothermal-atomisation-atomic-absorption spectroscopy and fluorimetry for studying the distribution of aluminium in human serum in the absence and in the presence of desferrioxamine. Aluminium was eluted as a single peak in the same fraction as transferrin. However, following the addition of desferrioxamine most of the aluminium was liberated from transferrin and become attached to the chelator.

Aluminum speciation studies in biological fluids. Part 3. Quantitative investigation of aluminum-phosphate complexes and assessment of their potential significance in vivo

Following the discovery that specific health disorders affecting patients with renal disease were due to their excessive body accumulation of aluminum, it was established that aluminum toxicity was mainly due to the ingestion of aluminum-containing phosphate binders. Suspicion of toxicity was thus cast on aluminum-containing antacids, and subsequent tests held on healthy subjects did reveal that aluminum hydroxide gels were also potential oral sources of aluminum, especially in the presence of citric acid. Nevertheless, authors of these tests concluded that there was only marginal absorption of aluminum phosphate. In contrast with these clinical conclusions, it has recently been contended on theoretical grounds that aluminum phosphate represents a serious health hazard. To help elucidate this issue, this paper first deals with a quantitative investigation of aluminum-phosphate equilibria under physiological conditions. Then appropriate computer simulations based on corresponding results are used to assess the actual extent to which phosphate can influence aluminum bioavailability. These simulations confirm that aluminum phosphate is not expected to induce absorption of high amounts of aluminum when administered by itself. Nevertheless, this result may no longer apply in the presence of food, whose various acidic components are likely to modify the involved chemical equilibria. Moreover, it is shown that rising blood plasma phosphate levels should tend to increase aluminum tissue penetration and hence favor its potential toxicity.

Aluminum speciation studies in biological fluids. Part 2. Quantitative investigation of aluminum-citrate complexes and appraisal of their potential significance in vivo

Because of the recent implications of aluminum in the pathogenesis of various disease states, its in vivo chemistry has been receiving growing attention from bioinorganic chemists over the last few years. In this context, the elucidation of the main factors that govern aluminum bioavailability constitutes an urgent objective. Clearly, prevention measures require that mechanisms of aluminum absorption be definitely characterized, whereas specific sequestering agents are needed to detoxify patients with high-aluminum-body burdens. In particular, speciation studies are necessary to discriminate among the chemical forms under which aluminum predominates in vivo. Low molecular weight (LMW) species, which are the most active in terms of bioavailability, cannot be assessed by analytical techniques, and so computer simulations must be used. In recent clinical studies as well as in preliminary simulations dealing with aluminum distribution in blood plasma, citrate has been recognized as the most important LMW ligand of aluminum. The present paper thus reports a quantitative investigation of aluminum-citrate equilibria, carried out at 37 degrees C in NaCl 0.15 mol dm-3 in accordance with the experimental protocol defined in our previous study on aluminum hydrolysis. The ML, MLH, ML2, M3L3H-4, M2L2H-2, ML2H-1, and ML2H-2 species have been characterized over the whole physiological pH range using as large reactant concentration ratios as possible. Corresponding formation constants have then been used to investigate the role of citrate towards aluminum bioavailability. Blood plasma simulations reveal that citrate can promote aluminum urinary excretion, which substantiates recent clinical observations made on mice. However, the higher plasma aluminum concentrations are, the less effective citrate is to be expected. Gastrointestinal simulations confirm that the electrically neutral ML complex does represent an important risk of aluminum absorption in the upper region of the gastrointestinal tract at usual therapeutic doses. At moderate- and low-aluminum concentrations, citrate is also capable of dissolving the aluminum trihydroxide precipitate, which may combine with the capacity of other ligands to complex Al3+ into absorbable complexes at less acidic pH.

Aluminum and bone disorders: with specific reference to aluminum contamination of infant nutrients

Aluminum (Al) impairment of bone matrix formation and mineralization may be mediated by its direct effect on bone cells or indirectly by its effect on parathyroid hormone and calcium metabolism. Its toxic effects are proportional to tissue Al load. Al contamination of nutrients depends on the amount of Al present naturally in chemicals or from the manufacturing process. Intravenous calcium, phosphorus, and albumin solutions have high Al (greater than 500 micrograms/L), whereas crystalline amino acid, sterile water, and dextrose water have low Al (less than 50 micrograms/L) content. Enteral nutrients including human and whole cow milk have low Al, whereas highly processed infant formulas with multiple additives, such as soy formula, preterm infant formula, and formulas for specific disorders are heavily contaminated with Al. Healthy adults are in zero balance for Al. The gastrointestinal tract excludes greater than 95% of dietary Al, and kidney is the dominant organ for Al excretion. However, even with normal renal function, only 30-60% of an Al load from parenteral nutrition is excreted in the urine, resulting in tissue accumulation of Al. The risk for Al toxicity is greatest in infants with chronic renal insufficiency, recipients of long term parenteral nutrition, i.e., no gut barrier to Al loading, and preterm infants with low Al binding capacity. The rapid growth of the infant would theoretically potentiate Al toxicity in all infants, although the critical level of Al loading causing bone disorders is not known. To minimize tissue burden, Al content of infant nutrients should be similar to "background" levels, i.e., similar to whole milk (less than 50 micrograms/L).

Direct spectrophotometric determination of the two site binding of aluminum to transferrin

Difference UV spectrophotometry is used to determine the conditional binding constants for aluminum to human transferrin in the presence of HCO3- of initial concentration 18 mM according to a two site model. The values obtained at 37 degrees C and pH 7.4 are: K1 = 3.0 (+/- 0.8) X 10(15) M-2, K2 = 2.9 (+/- 0.7) X 10(15) M-2 for the reactions: Tr + Al3+ + HCO3- = Tr-Al-HCO3 and Tr-Al-HCO3 + Al3+ + HCO3- = Tr-Al2-(HCO3)2 respectively. Possible consequences arising for the transport of aluminum in human serum are discussed.